Embed Size (px)
Citation preview
Report ANLESD-1812
Technology Assessment of a Fuel Cell Vehicle 2017 Toyota Mirai
Energy Systems Division
About Argonne National Laboratory Argonne is a US Department of Energy laboratory managed by UChicago Argonne LLC under contract DE-AC02-06CH11357 The Laboratoryrsquos main facility is outside Chicago at 9700 South Cass Avenue Argonne Illinois 60439 For information about Argonne and its pioneering science and technology programs see wwwanlgov
DOCUMENT AVAILABILITY
Online Access US Department of Energy (DOE) reports produced after 1991 and a growing number of pre-1991 documents are available free at OSTIGOV (httpwwwostigov) a service of the US Dept of Energyrsquos Office of Scientific and Technical Information
Reports not in digital format may be purchased by the public from the National Technical Information Service (NTIS)
US Department of Commerce National Technical Information Service 5301 Shawnee Rd Alexandria VA 22312 wwwntisgov Phone (800) 553-NTIS (6847) or (703) 605-6000 Fax (703) 605-6900 Email ordersntisgov
Reports not in digital format are available to DOE and DOE contractors from the Office of Scientific and Technical Information (OSTI)
US Department of Energy Office of Scientific and Technical Information PO Box 62 Oak Ridge TN 37831-0062 wwwostigov Phone (865) 576-8401 Fax (865) 576-5728 Email reportsostigov
Disclaimer This report was prepared as an account of work sponsored by an agency of the United States Government Neither the United States Government nor any agency thereof nor UChicago Argonne LLC nor any of their employees or officers makes any warranty express or implied or assumes any legal liability or responsibility for the accuracy completeness or usefulness of any information apparatus product or process disclosed or represents that its use would not infringe privately owned rights Reference herein to any specific commercial product process or service by trade name trademark manufacturer or otherwise does not necessarily constitute or imply its endorsement recommendation or favoring by the United States Government or any agency thereof The views and opinions of document authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof Argonne National Laboratory or UChicago Argonne LLC
Report ANLESD-1812
Technology Assessment of a Fuel Cell Vehicle 2017 Toyota Mirai
prepared by Henning Lohse-Busch Michael Duoba Kevin Stutenberg Simeon Iliev Mike Kern Energy Systems Division Argonne National Laboratory Brad Richards Martha Christenson Transport Canada Arron Loiselle-Lapointe Environment and Climate Change Canada Prepared for US Department of Energy Fuel Cell Technologies Office June 2018
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
Report ANLESD-1812June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
GOALS OF THE TESTING
The technical objectives of the technology assessment of this advanced production vehicle are to Establish vehicle level energy
consumption efficiency and performance data on varying drive cycles at ambient temperatures ranging from 20F to 95F with a stretch goal of testing a cold start at 0F Measure the performance envelops and
synergies between the fuel cell system and the hybrid system (incl fuel cell system idle) Generate an efficiency map of the fuel cell
systems
Testing of a fuel cell production vehicle in a controlled laboratory environment
SOW_v5 submitted to FCTO in September 2017 2
OVERVIEWTesting of a fuel cell production vehicle in a controlled laboratory environment in order to generate public powertrain data
Collaboration with
Transport Canada provided the vehicle and engaged in technical exchange of ideas Argonne is very grateful for the successful collaboration
3
Abbreviations used in the presentation FC Fuel CellHEV Hybrid Electric VehicleBEV Battery Electric VehicleMpge Miles per gallon gasoline equivalentLe100km Liter gasoline equivalent per 100 kilometers
Vehicle overview and instrumentation Link
Fuel cell hybrid powertrain operation overviewbull On cyclesbull By components
Link
Fuel cell stack and system efficiency (steady load mapping) Link
Certification drive cycle testing and results at 72F bull Results with average efficienciesbull Energy analysis
Link
Fuel cell system operationbull Air amp hydrogen as function of power outputbull FC idle testingbull FC maximum power output (continuous and peak pulses)
Link
Impact of temperatures 0F 20F 72F and 95F+850Wm2
bull Fuel economy results and energy analysis ults for UDDS Highway and US06 across temperatures
bull Fuel cell system shut down from 72F to 0Fbull Fuel cell system cold start from 72F to 0Fbull Power delivery of FC conditioned at 0F bull Hill climb testrsquo at 95F with solar load
Link
Summary and take away points Link
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
About Argonne National Laboratory Argonne is a US Department of Energy laboratory managed by UChicago Argonne LLC under contract DE-AC02-06CH11357 The Laboratoryrsquos main facility is outside Chicago at 9700 South Cass Avenue Argonne Illinois 60439 For information about Argonne and its pioneering science and technology programs see wwwanlgov
DOCUMENT AVAILABILITY
Online Access US Department of Energy (DOE) reports produced after 1991 and a growing number of pre-1991 documents are available free at OSTIGOV (httpwwwostigov) a service of the US Dept of Energyrsquos Office of Scientific and Technical Information
Reports not in digital format may be purchased by the public from the National Technical Information Service (NTIS)
US Department of Commerce National Technical Information Service 5301 Shawnee Rd Alexandria VA 22312 wwwntisgov Phone (800) 553-NTIS (6847) or (703) 605-6000 Fax (703) 605-6900 Email ordersntisgov
Reports not in digital format are available to DOE and DOE contractors from the Office of Scientific and Technical Information (OSTI)
US Department of Energy Office of Scientific and Technical Information PO Box 62 Oak Ridge TN 37831-0062 wwwostigov Phone (865) 576-8401 Fax (865) 576-5728 Email reportsostigov
Disclaimer This report was prepared as an account of work sponsored by an agency of the United States Government Neither the United States Government nor any agency thereof nor UChicago Argonne LLC nor any of their employees or officers makes any warranty express or implied or assumes any legal liability or responsibility for the accuracy completeness or usefulness of any information apparatus product or process disclosed or represents that its use would not infringe privately owned rights Reference herein to any specific commercial product process or service by trade name trademark manufacturer or otherwise does not necessarily constitute or imply its endorsement recommendation or favoring by the United States Government or any agency thereof The views and opinions of document authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof Argonne National Laboratory or UChicago Argonne LLC
Report ANLESD-1812
Technology Assessment of a Fuel Cell Vehicle 2017 Toyota Mirai
prepared by Henning Lohse-Busch Michael Duoba Kevin Stutenberg Simeon Iliev Mike Kern Energy Systems Division Argonne National Laboratory Brad Richards Martha Christenson Transport Canada Arron Loiselle-Lapointe Environment and Climate Change Canada Prepared for US Department of Energy Fuel Cell Technologies Office June 2018
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
Report ANLESD-1812June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
GOALS OF THE TESTING
The technical objectives of the technology assessment of this advanced production vehicle are to Establish vehicle level energy
consumption efficiency and performance data on varying drive cycles at ambient temperatures ranging from 20F to 95F with a stretch goal of testing a cold start at 0F Measure the performance envelops and
synergies between the fuel cell system and the hybrid system (incl fuel cell system idle) Generate an efficiency map of the fuel cell
systems
Testing of a fuel cell production vehicle in a controlled laboratory environment
SOW_v5 submitted to FCTO in September 2017 2
OVERVIEWTesting of a fuel cell production vehicle in a controlled laboratory environment in order to generate public powertrain data
Collaboration with
Transport Canada provided the vehicle and engaged in technical exchange of ideas Argonne is very grateful for the successful collaboration
3
Abbreviations used in the presentation FC Fuel CellHEV Hybrid Electric VehicleBEV Battery Electric VehicleMpge Miles per gallon gasoline equivalentLe100km Liter gasoline equivalent per 100 kilometers
Vehicle overview and instrumentation Link
Fuel cell hybrid powertrain operation overviewbull On cyclesbull By components
Link
Fuel cell stack and system efficiency (steady load mapping) Link
Certification drive cycle testing and results at 72F bull Results with average efficienciesbull Energy analysis
Link
Fuel cell system operationbull Air amp hydrogen as function of power outputbull FC idle testingbull FC maximum power output (continuous and peak pulses)
Link
Impact of temperatures 0F 20F 72F and 95F+850Wm2
bull Fuel economy results and energy analysis ults for UDDS Highway and US06 across temperatures
bull Fuel cell system shut down from 72F to 0Fbull Fuel cell system cold start from 72F to 0Fbull Power delivery of FC conditioned at 0F bull Hill climb testrsquo at 95F with solar load
Link
Summary and take away points Link
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
Report ANLESD-1812
Technology Assessment of a Fuel Cell Vehicle 2017 Toyota Mirai
prepared by Henning Lohse-Busch Michael Duoba Kevin Stutenberg Simeon Iliev Mike Kern Energy Systems Division Argonne National Laboratory Brad Richards Martha Christenson Transport Canada Arron Loiselle-Lapointe Environment and Climate Change Canada Prepared for US Department of Energy Fuel Cell Technologies Office June 2018
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
Report ANLESD-1812June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
GOALS OF THE TESTING
The technical objectives of the technology assessment of this advanced production vehicle are to Establish vehicle level energy
consumption efficiency and performance data on varying drive cycles at ambient temperatures ranging from 20F to 95F with a stretch goal of testing a cold start at 0F Measure the performance envelops and
synergies between the fuel cell system and the hybrid system (incl fuel cell system idle) Generate an efficiency map of the fuel cell
systems
Testing of a fuel cell production vehicle in a controlled laboratory environment
SOW_v5 submitted to FCTO in September 2017 2
OVERVIEWTesting of a fuel cell production vehicle in a controlled laboratory environment in order to generate public powertrain data
Collaboration with
Transport Canada provided the vehicle and engaged in technical exchange of ideas Argonne is very grateful for the successful collaboration
3
Abbreviations used in the presentation FC Fuel CellHEV Hybrid Electric VehicleBEV Battery Electric VehicleMpge Miles per gallon gasoline equivalentLe100km Liter gasoline equivalent per 100 kilometers
Vehicle overview and instrumentation Link
Fuel cell hybrid powertrain operation overviewbull On cyclesbull By components
Link
Fuel cell stack and system efficiency (steady load mapping) Link
Certification drive cycle testing and results at 72F bull Results with average efficienciesbull Energy analysis
Link
Fuel cell system operationbull Air amp hydrogen as function of power outputbull FC idle testingbull FC maximum power output (continuous and peak pulses)
Link
Impact of temperatures 0F 20F 72F and 95F+850Wm2
bull Fuel economy results and energy analysis ults for UDDS Highway and US06 across temperatures
bull Fuel cell system shut down from 72F to 0Fbull Fuel cell system cold start from 72F to 0Fbull Power delivery of FC conditioned at 0F bull Hill climb testrsquo at 95F with solar load
Link
Summary and take away points Link
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
Report ANLESD-1812June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
GOALS OF THE TESTING
The technical objectives of the technology assessment of this advanced production vehicle are to Establish vehicle level energy
consumption efficiency and performance data on varying drive cycles at ambient temperatures ranging from 20F to 95F with a stretch goal of testing a cold start at 0F Measure the performance envelops and
synergies between the fuel cell system and the hybrid system (incl fuel cell system idle) Generate an efficiency map of the fuel cell
systems
Testing of a fuel cell production vehicle in a controlled laboratory environment
SOW_v5 submitted to FCTO in September 2017 2
OVERVIEWTesting of a fuel cell production vehicle in a controlled laboratory environment in order to generate public powertrain data
Collaboration with
Transport Canada provided the vehicle and engaged in technical exchange of ideas Argonne is very grateful for the successful collaboration
3
Abbreviations used in the presentation FC Fuel CellHEV Hybrid Electric VehicleBEV Battery Electric VehicleMpge Miles per gallon gasoline equivalentLe100km Liter gasoline equivalent per 100 kilometers
Vehicle overview and instrumentation Link
Fuel cell hybrid powertrain operation overviewbull On cyclesbull By components
Link
Fuel cell stack and system efficiency (steady load mapping) Link
Certification drive cycle testing and results at 72F bull Results with average efficienciesbull Energy analysis
Link
Fuel cell system operationbull Air amp hydrogen as function of power outputbull FC idle testingbull FC maximum power output (continuous and peak pulses)
Link
Impact of temperatures 0F 20F 72F and 95F+850Wm2
bull Fuel economy results and energy analysis ults for UDDS Highway and US06 across temperatures
bull Fuel cell system shut down from 72F to 0Fbull Fuel cell system cold start from 72F to 0Fbull Power delivery of FC conditioned at 0F bull Hill climb testrsquo at 95F with solar load
Link
Summary and take away points Link
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
GOALS OF THE TESTING
The technical objectives of the technology assessment of this advanced production vehicle are to Establish vehicle level energy
consumption efficiency and performance data on varying drive cycles at ambient temperatures ranging from 20F to 95F with a stretch goal of testing a cold start at 0F Measure the performance envelops and
synergies between the fuel cell system and the hybrid system (incl fuel cell system idle) Generate an efficiency map of the fuel cell
systems
Testing of a fuel cell production vehicle in a controlled laboratory environment
SOW_v5 submitted to FCTO in September 2017 2
OVERVIEWTesting of a fuel cell production vehicle in a controlled laboratory environment in order to generate public powertrain data
Collaboration with
Transport Canada provided the vehicle and engaged in technical exchange of ideas Argonne is very grateful for the successful collaboration
3
Abbreviations used in the presentation FC Fuel CellHEV Hybrid Electric VehicleBEV Battery Electric VehicleMpge Miles per gallon gasoline equivalentLe100km Liter gasoline equivalent per 100 kilometers
Vehicle overview and instrumentation Link
Fuel cell hybrid powertrain operation overviewbull On cyclesbull By components
Link
Fuel cell stack and system efficiency (steady load mapping) Link
Certification drive cycle testing and results at 72F bull Results with average efficienciesbull Energy analysis
Link
Fuel cell system operationbull Air amp hydrogen as function of power outputbull FC idle testingbull FC maximum power output (continuous and peak pulses)
Link
Impact of temperatures 0F 20F 72F and 95F+850Wm2
bull Fuel economy results and energy analysis ults for UDDS Highway and US06 across temperatures
bull Fuel cell system shut down from 72F to 0Fbull Fuel cell system cold start from 72F to 0Fbull Power delivery of FC conditioned at 0F bull Hill climb testrsquo at 95F with solar load
Link
Summary and take away points Link
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
OVERVIEWTesting of a fuel cell production vehicle in a controlled laboratory environment in order to generate public powertrain data
Collaboration with
Transport Canada provided the vehicle and engaged in technical exchange of ideas Argonne is very grateful for the successful collaboration
3
Abbreviations used in the presentation FC Fuel CellHEV Hybrid Electric VehicleBEV Battery Electric VehicleMpge Miles per gallon gasoline equivalentLe100km Liter gasoline equivalent per 100 kilometers
Vehicle overview and instrumentation Link
Fuel cell hybrid powertrain operation overviewbull On cyclesbull By components
Link
Fuel cell stack and system efficiency (steady load mapping) Link
Certification drive cycle testing and results at 72F bull Results with average efficienciesbull Energy analysis
Link
Fuel cell system operationbull Air amp hydrogen as function of power outputbull FC idle testingbull FC maximum power output (continuous and peak pulses)
Link
Impact of temperatures 0F 20F 72F and 95F+850Wm2
bull Fuel economy results and energy analysis ults for UDDS Highway and US06 across temperatures
bull Fuel cell system shut down from 72F to 0Fbull Fuel cell system cold start from 72F to 0Fbull Power delivery of FC conditioned at 0F bull Hill climb testrsquo at 95F with solar load
Link
Summary and take away points Link
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
ARGONNE VEHICLE TECHNOLOGY ASSESSMENTAND ANALYSIS REPORTING
4
Vehicle levelbull Energy consumption
(fuel + electricity)bull Emissionsbull Performancebull Vehicle operation and strategy
lsquoIn-sitursquo component amp system testingbull Component performance
efficiency and operationover drive cycles
bull Component mapping
bull Test summary results
bull 10Hz data of major signals
bull Analysis Presentations2WD chassis dynamometer
bull Up to medium duty
opened 2002 upgraded 2011
opened 2009
ldquoWe assess state-of-the-art transportation technology
for the Department of Energy and Argonne research interestsrdquo
Downloadable Dynamometer Database wwwanlgovd3
Research Oriented Test Facilities Vehicle Technology Assessment
4WD chassis dynamometerbull Thermal
Chamber 0F to 95F
bull Solar emulation
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
5
VEHICLE OVERVIEW AND INSTRUMENTATION
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
2017 TOYOTA MIRAI FUEL CELL VEHICLE
6
Manufacturer data +wwwfueleconomygov^wwwzeroto60timescom
2017 Toyota Mirai Fuel Cell VehicleVehicle architecture Fuel Cell Series Hybrid Vehicle
Test weight 4250 lbs
Power plant Fuel CellSolid Polymer Electrolyte Fuel Cell370 cells in stack114kW 31 kWL 20 kWkgFlow channel 3D fine-mesh flow field
(cathode)Humidification Internal circulation system
(humidifier-less)
Hydrogen storage 10000 psi 5 kg of H2
Battery Nickel-metal Hydride 16 kWh 245V
Climate control Electrical AC compressorElectric heater
EPA Label Fuel Economy (mpge)
66 City 66 Hwy 66 Combined+
Performance Reported 0-60 Time 90s^
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
TOYOTA MIRAI POWERTRAIN COMPONENT LAYOUT
7
The high voltage power distribution system is under the hood along with the inverters for the motors
Power Electronics
The air cooled NiMH battery pack is packaged in the trunk similar to most Toyota Hybrid vehicles
Battery Pack
The electric drive motor and the air compressor are packaged in-line between the front wheels
Electric motors
The fuel cell stack along with the boost converter is under the center of the vehicle
Fuel Cell System
The two hydrogen tanks are under the trunk and the rear passenger seats These tanks are disabled for the testing
On-board hydrogen Storage
httpsyoutubeqofykBQre5o
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
HYDROGEN MEASUREMENT FOR THE MIRAI
8
Different metering technicsIntegrated mass flow meter Yes Ideal gas law PossibleGravimetric No
Outside H2 Storage
Laboratory grade hydrogen (99999 pure) in 12 packs with is equivalent to 6 gallons of gasoline Hydrogen is piped into the building at 245 psi
Safety and Metering Panel
The hydrogen is metered by two Micro Motionreg ELITEreg Coriolis mass flow meters The panel has over pressure safeties automatic shut-off valves hydrogen sensors and a venting system Any hydrogen sensor can trigger the active test cell safeties and alert the fire department
The vehicle hydrogen tanks are disabled and completely by-passed The hydrogen is stepped down at the exit of the panel to 220 psi and fed into the pressure regulator in the middle of the vehicle
Test cell Connection
Vehicle Connection
Note The 30 ft delivery hose and the vehicle piping between the mass flow meters and the
fuel cell stack creates a filtering effect and time delay on the
fuel flow measurement
but more complicated and not the intension of this work
too expensive for the budget of this work
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
POWER MEASUREMENTS EQUIPMENT
Hydrogen FlowMicro Motionreg ELITEreg Coriolis meters Argonnersquos uses the CMF010M and
CMF025M plusmn025 gas flow accuracy 00002 gcc
density accuracy The output of the meters is 4-20mA The
output on both meters is scaled from 4-20mA to 0-3grsec in the data acquisition system Further specifications
httpwwwemersoncomen-uscatalogmicro-motion-elite-coriolis
Electric Power FlowHiokireg High Precision Power Analyzer Hiokireg Power Analyzers PW3390-10 Hioki current clamps CT6841 CT6843 plusmn01 power accuracy Outputs (V I P IH WH) transferred to
DAQ via ethernet V I transfer via analog signal in additional to ethernet Further Specifications
httpswwwhiokicomenproductsdetailproduct_key=6413
9
High voltage tap
Power analyzers
Current clamps
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
SAMPLE OF HIGHLIGHTED CAN MESSAGES (TOTAL 400+)
10
Possible Vehicle Communication Signal Value UnitFC Voltage before Boosting 2391 VFC Voltage after Boosting 3455 VFC Current 29 ATarget FC Output Power 47 kWFC output Power 094 KWFC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -091 kPa(gauge)Intake Air Temperature 26 CMass Air Flow Value 3009 NLminAir Compressor Revolution 6395 rpmAir Compressor Consumption Power 0 WFC Stack Coolant Temperature (Radiator Outlet) 29 CFC Stack Coolant Temperature (FC Stack Outlet) 45 CFC Water Pump Revolution 73525 rpmFC Water Pump Consumption Power 19 VRadiator Fan 1 Driving Request 0 Radiator Fan 2 Driving Request 0 Estimated Radiator Rotary Valve Position 0 FC Stack Air Temperature (FC Stack Inlet) 38 CFC Stack Air Pressure (FC Stack Inlet) -135 kPa(gauge)Mass Air Flow Value 2997 NLminHydrogen Pump Revolution 60735 rpmHydrogen Pump Consumption Power 19 WSmoothed Value of Low-range Hydrogen Pressure 545 kPa(gauge)Hydrogen Injector Injection Number 1FC Voltage before Boosting 2939FC Current 335FC Mode FC WorkingLow Temperature Mode 2FC Stack Power Generation Mode IntermittentFC Stack Cell Average Minimum Voltage 002 VFC Stack Cell Minimum Voltage 0 VFC Stack Cell Minimum Average Voltage Cell Channel No 31 chFC Stack Cell Minimum Voltage Cell Channel No 1 chFC Total Voltage 2939FC Stack Internal Resistance 011 ohmHydrogen Pump Motor Temperature 38 chHydrogen Injector I Injection Request OffHydrogen Injector 2 Injection Request OffHydrogen Injector 3 Injection Request Off
Exhaust Drainage Valve Driving Request OffLow-range Hydrogen Pressure 2907 kPa(gauge)Internal Resistance ROB 0019 ohmInternal Resistance RI17 0019 ohmBattery Pack Current Value (1B Correction) -197 ADelta SOC 0 SOC IG-ON 64 Status of Charge Min 455 Cooling Fan Mode 1 0Drive Motor Revolution 0 rpmDrive Motor Execution Torque -05 NmInverter Water Pump Revolution 3500 rpmDCDC Converter target Pulse Duty 793 Inverter Water Pump Run Control Duty 10 Boost Ratio 315 Boosting Converter Carrier Frequency 955 kHzAC Consumption Power 0 WH20 Switch OFFBattery Charging Request OffFC Startup Request OffFC Stop Request OffFC Converter Available Output Current 500 AHeater Temperature 2187 CWater Heater Drive Status OffFC Converter Input Voltage 2972 VFC Converter Output Voltage 3466 VFC Current 3 AFC Converter Shutdown Request OffFC Converter Request Power 2604 kWFC Air Compressor Motor Temperature 29 CFC Air Compressor Motor Temperature after IG-ON 21 CMax FC Air Compressor Motor Temperature 29 CFC Air Compressor Revolution 638 rpmTarget FC Air Compressor Motor Torque 07 NmFC Air Compressor Motor Torque 06 NmAccel Pedal Position No 1 1568 Auxiliary Battery Voltage 1412 VMotor Cooler Oil Pump Status 0 Target Motor Cooler Oil Pump Duty 10 H20 Indicator OffWater Heater Duty 0
Dat
a fro
m id
ling
the
vehi
cle
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
11
AIR COMPRESSOR POWER CALCULATIONS
The CAN reported air compressor power message (lsquoair_compressor_consumption_power_FC_Wrsquo)
clips at the value of ~397kW
Note The 10 Hz data to generate these graphs is from UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
The air compressor power can be calculated using the compressor motor speed and the compressor
torque reported on CAN A linear efficiency of 85 is applied to achieve a correlation to the reported air
compressor power below 4 kW
Air compressor power in the remainder of the analysis is
the calculated power
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
POWERTRAIN ARCHITECTURE AND MEASUREMENTS
12
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
200A clamp (ch2)500A clamp (ch3)
20A clamp (ch4)
200A clamp (ch6)
200A clamp (ch1)200A clamp (ch5)20A clamp (ch7)
Shielding on Shielding on
CAN voltage current amp power
CAN speed and torque
CAN voltage current amp SOC
CAN power CAN powerCAN speed and torque
Dyno speed and effort
CAN power
H2 mass flow from APRF
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL HYBRID POWERTRAIN OPERATION OVERVIEW
13
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
BASIC FUEL CELL HYBRID SYSTEM OPERATION
14
Time [s]
Spee
d [m
ph]
Pow
er [k
W]
FC provides majority of the power during accelerationNo FC
power while
vehicle stopped
Electric vehicle launch (no FC power)
EV only driving
Regenerative braking
FC recharges
battery
FC open circuit voltage drops
while hydrogen flow is stopped
~1 minute window
FC back up to
OCV at startup
FC provide the power to cruise at steady
state speed and the battery is inactive
During accelerations the fuel power increases and
the battery provides assist power
Note NEDC drive cycle has low power demands
FC air compressor
operates at a few hundred
Watts at most
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL OPERATION ON AGGRESSIVE DRIVING
15
Time [s] Time [s]
Spee
d [m
ph]
Pow
er [k
W]
Regenerative braking
limited by battery
Highly dynamic speed changes while
cruising and the FC power follows the
dynamic load changes
Large peak power demand during acceleration at
high speed
Sequence of heavy accelerations with
power from FC with battery assist
Large peak power demand of
100kW met by FC with battery assist
Fuel deceleration
cut like feature
FC air compressor power ramps to up 8 kW
The Mirai is a fuel cell dominant hybrid electric vehicle with a strong
load following control strategy
Note US06 drive cycle has high power demands
FC air compressor power ramps to up 15 kW
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL AND BATTERY USAGE ENVELOPS
16
10kW
20kW
30kW
40kW
Battery polarization curve
Compared to gasoline HEVs the battery polarization curves is more extensive on the charging side for the fuel cell hybrid
DischargeCharge 10kW20kW30kW40kW50kW
100kW
150kW
Open Circuit Voltage is around 315V The voltage drop while the fuel cell is idled is apparent
Fuel Cell polarization curve
H2
star
vatio
n
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
FC vs Battery power
Three acceleration operations emerge 1) High load fuel cell power with little battery assist 2) Medium load fuel cell power with battery assist 3) Low load battery charging
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL AND BATTERY USAGE ENVELOPS
17
FC vs Battery power
3) Low load battery charging
Note The 10 Hz data to generate these graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
Deceleration
Cha
rge
Dis
char
ge
Acceleration
1
1
2
2
3
1) High load fuel cell power with little battery assist
2) Medium load fuel cell power with battery assist
Regenerative braking
The modeling and simulation team will use the data to
validate their current models and control strategies
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL STACK AND SYSTEM EFFICIENCY (STEADY LOAD MAPPING)
18
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
DEFINING THE SYSTEM BOUNDARIES FOR EFFICIENCY CALCULATIONS
19
Clutch
Mechanical Connection
Electrical Connection
Fuel Connection
Fuel component
High voltage component
Mechanical component
Instrumentation note bullH2 mass flow APRF test cell metering system
bullPower (VampI) Voltage tap and current clamp input to power analyzer
bullPower (TampS) Torque and Speed Sensor
Fuel Cell
System114 kW
stack
NiMH Battery16 kWh
245 V
AC motor113 kW
H2 storage5 kg
160 kWhFC Boost converter
Boosted Power
Electronics Bus
FCCompressor
FC water pump
Battery Power
Electronics Bus
DCDC
AC compressor
Heater
FC water + H2 Pump
245V DC
245V DC
245V DC
245V DC
650V AC
650V AC
FC H2 pump
12V battery
12V loads
Fuel CellStack Efficiency
Fuel Cell System Efficiency
Vehicle Efficiency(SAE J2951trade positive cycle energy based)
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL SYSTEM MAPPING IN VEHICLE ON CHASSIS DYNAMOMETER
20
The dynamometer is locked at 30 mph for the test and serves as an
power absorption device
The driver in the vehicle tips into the accelerator pedal to reach specific power
level based on the instrumentation
Note The low and high power level are alternated to attempt to maintain lsquonormalrsquo operating temperatures and avoid abnormally high powertrain temperatures In this test 10-20-30-20-40-30-50-10 kW FC stack output were targeted
Moving to the next power level
A window for steady operation (cst flow cst
power output cst temperatureshellip) is used to compute the average
values which are used for the efficiency analysis
The driver maintain that power level until power level fuel flow levels temperatures stabilize and lsquoenoughrsquo data is
recorded to calculate the efficiency with certainty
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL STACK AND SYSTEM EFFICIENCY
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
Max continuous power 95F ambient + solar
~50k
WFC
stac
k
15m
ph o
n25
grad
e in
95Fa
mbi
ent+
sola
rload
Max continuous power 72F ambient
~73k
WFC
stac
k
15m
ph o
n25
grad
e in
72Fa
mbi
ent
21
Majority of UDDS and
Highway cycle
0 19 36 68 82 9553
Fuel flow metering
challenges on highest load pointsNet Electric Fuel Cell System Output [kW]
Max electric power outputbull 110kW FC Stackbull 105kW FC Boost converterbull 90kW FC System
Test point only held for limited
time 72F ambient
FC Stack peak efficiency 660
FC System peak efficiency 637
Base
d on
LH
V =
119
96 M
Jkg
FC System peak efficiency at 25 power is 58
Toyota describes the modified their air supply system which increases fuel economy by lower the air compressor power at low loads Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-1185
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL EFFICIENCY IN OPERATING ZONE
22
0 19 36 53
Key finding of efficiency based
on laboratory data
The lsquospreadrsquo in the data is caused by thermal conditions stack conditions from previous test
sequence accessory loads data processing decisions
FC Stack peak efficiency 660
FC System peak efficiency 637
FC does not typically operate under 5 kW
on drive cycle Special idle test
forced this operation
At higher loads the air
compressor other system accessories
affect the system efficiency
Tested the power level
multiple times
Net Electric Fuel Cell System Output [kW]
Base
d on
LH
V =
119
96 M
Jkg
Notebull Analysis based on steady
state speed tests with 0 3 and 6 grade + WOT at 72F ambient temperatures
bull Hood closed test cell fan in vehicle speed match mode Climate control off
2017 Toyota Miraitest data
Net
Ele
ctric
ity O
ut
Hyd
roge
n In
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
DOE TECHNICAL TARGETS FOR FUEL CELL SYSTEMS AND STACKS FOR TRANSPORTATION APPLICATIONS
23
Characteristic Units 2015Status
2020Targets
UltimateTargets
Peak energy efficiencyb 60c 65 70Power density WL 640d 650 850Specific power Wkg 659e 650 650Costf $kWnet 53g 40 30Cold start-up time to 50 of rated power
ndash20degC ambient temperature seconds 20h 30 30+20degC ambient temperature seconds lt10h 5 5
Start-up and shutdown energyi
from -20degC ambient temperature MJ 75 5 5from +20degC ambient temperature MJ ndash 1 1
b Ratio of DC output energy to the lower heating value of the input fuel (hydrogen) Peak efficiency occurs at less than 25 rated powerc W Sung Y Song K Yu and T Lim Recent Advances in the Development of Hyundai-Kiarsquos Fuel Cell Electric Vehicles SAE Int J Engines 31 (2010) 768ndash772 doi 1042712010-01-1089
Table from httpswwwenergygoveerefuelcellsdoe-technical-targets-fuel-cell-systems-and-stacks-transportation-applications on April 2018
Technical Targets 80-kWe (net) Integrated Transportation Fuel Cell Power Systems Operating on Direct Hydrogena
ldquoAs a result of the implemented fuel efficiency strategies the tested Borrego FCEV has attained a fuel cell efficiency of 65 and thereby fuel cell system efficiency of 62 in the mean time the tested Tucson FCEV has achieved a fuel economy of over 72 mpg city The fuel efficiency was estimated while running at about 60 mph on average whereas the hydrogen consumption was in-house measured based on the FTP-72 test cycle and is represented in miles per gasoline equivalent gallons for comparison purposes rdquo W Sung
2008 Hyundai Borrego
2017 Toyota Mirai
Weight 2300 kg reported in paper
1930 kgtest weight
Fuel CellPeak power
PEM100 kW
PEM 114 kW
Fuel Economy on FTP 72 (aka UDDS)
72 mpge
Cold start 915 mpgeHot start953 mpge
FC stack efficiency at 60 mph
65 632
FC system efficiency at 60 mph
62 606
Peak FC stack efficiency
65660 5
to 10 kW stack output
Peak FC system efficiency
62637 5
to 10 kW stack output
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
COMPARISON OF SERIES HYBRID ldquoGENERATORSrdquo
24
Motor
Batte
ry
Generator
Fuel
Effic
ienc
y
SeriesHybrid
Architecture
REx Efficiency (Fuel to Electrical)
Engine Efficiency (Fuel to Mechanical)
Fuel CellHybrid Electric Vehicle
bull Charge sustainbull Fuel Cell dominantbull Generator max output 114 kW
Battery Electric Vehiclewith range extender bull Primarily BEVbull Small 649cc 2cyl enginebull Generator max output 25 kW
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
CERTIFICATION DRIVE CYCLE TESTING AND RESULTS AT 72F AMBIENT TEMPERATURE
25
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
STANDARD TEST SEQUENCE FOR CERTIFICATION TESTS
26
Cold StartUDDS1
Hot StartUDDS2 UDDS3 Highway US06
Report fuel economy
from these cycles
10 minute break
Key
off
Key
off
Key
off
Key
off
Prep UDDS cycle completed followed by
vehicle temperature soak over night in the
test cell
Powertrain thermal conditionsbull lsquoCold Startrsquo test means that
the vehicle and therefore the powertrain was soaked at the target ambient temperature for over 12 hours before the start of the test 16 hours is the typical soak time in the test cell on the dynamometer for this testing
bull lsquoHot Startrsquo test means that the vehicle just completed a test and that the powertrain is already warm Argonne tries to keep 10 minutes between the end of a test and the start of the hot start test
FC lsquoReactorrsquo
temperature
Cooling setup The test cell fan was run dynamically to match the air flow speed to the vehicle speed and the hood was closed for all testing regardless of ambient temperature
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
CERTIFICATION CYCLES TEST RESULT AT 72F AMBIENT TEMPERATURE
27
Fuel Economy Vehicle Efficiency
Hybrid vehicle efficiency range
Conventional vehicle efficiency
range
High fuel economyCaution MPG illusion area
Overall high powertrain efficiencies Efficiency drops for higher driving intensity contrary to
conventional vehicles
Electric vehicle efficiency range
Note bull SAE J2951trade defines
ldquoCycle Energyrdquo as the integration of positivepower at the wheel over the cycle
bull Regenerative braking =ldquofreerdquo energy
bull Thus enabling 100 efficiencies for EVs
Increasing operating
temperature
Increasing driving
intensity
SAE
J295
1
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL ECONOMY COMPARISON TO EPA RESULTS
28
(1) The data comes from the EPA testing database published on their website httpswwwepagovcompliance-and-fuel-economy-datadata-cars-used-testing-fuel-economy(2) The EPA file indicate the ldquoFE_UNITrdquo to be ldquoMPGrdquo It has been suggest that the units are ldquomikgrdquo in which case using a Petroleum Equivalent Factor (PEF) of 102 (H2 kggallon) the EPA fuel economy is 960 mpge for FTP75 and Highway Argonne was not able to confirm if this is a mistake in the EPA file
Test EPA(1) ANLFTP 75 (43 cold start and 57 hot start)
941 mpge (2) 919 mpge
Highway 941 mpge (2) 924 mpge
Testing Manufacturer submitted to EPA
ANL internal
2017 Toyota MiraiLabel Fuel Economy Data
HWY (Highway Driving)
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
UDDS (City Driving)
Argonne test results appear to match the EPA published test results
Note Individual drive cycle fuel economy test results are different from label fuel economy The drive cycles fuel economy results are used to calculate the label fuel economy which intends to represent real world fuel economy
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
WHERE DID THE ENERGY GO
29
(cold start)
Ambient test conditions
72F
Climate control bull At 72F the climate
control was off Fuel cell system losses are very low for low
load drive cycles
On high load drive cycles the
energy for the air compressor
becomes significant
FC auxiliary loads on low load cycles are very low
On high load drive cycles the boost converter
losses are noticeable
Average 12V accessory loads
239W 243W 249W 235W 321W
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL VEHICLE EFFICIENCY CONTEXT
30
Mira
i FC
Priu
s EV
Priu
s H
EVM
azda
3 C
V
2017 Toyota Mirai
Fuel Cell FCV
2017 Toyota Prius Prime
Electric Vehicle EVincl charging efficiency
2017 Toyota Prius Prime
Hybrid Vehicle HEV
2014 Mazda 3
Conventional Vehicle CV
The fuel cell vehicle efficiency is higher compared to vehicles that use an internal
combustion engine
FC M
irai
EV P
rius
Prim
eH
EV P
rius
Prim
eC
V M
azda
3
SAE
J295
1
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL SYSTEM PEAK EFFICIENCY CORRELATES WITH VEHICLE DEMANDS
UD
DS
-90
of a
ll po
wer
dem
and
is b
elow
12k
W
Hig
hway
-90
o
f all
pow
er d
eman
d is
bel
ow 2
0kW
US0
6 -9
0 o
f all
pow
er d
eman
d is
bel
ow 4
0kW
UDDS Highway US06
The higher load demands of the aggressive US06 cycle decrease the fuel cell system efficiency
(lower stack eff and high compressor power)
31
AverageEfficiency
UDDS1
UDDS2
UDDS3
Highway2
US062
FC Stack 648 648 647 634 618
FC System 620 622 618 611 481The fuel cell system is
more efficient at low load demands Most driving scenarios are relatively
low power demand
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
32
FURTHER DRIVE CYCLE RESULTS AT 72F AMBIENT TEMPERATURE
Fuel Cell Stack Output [kW]
Time [s]
Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW] Fuel Cell Stack Output [kW]
Time [s]Time [s] Time [s] Time [s] Time [s]
UDDS Highway US06 JC08 NEDC WLTP
Vehi
cle
spee
d [m
ph]
FC P
ower
freq
uenc
y []
Note Each drive cycle was tested in pairs back to back The results presented in the table are from the second cycle The second cycles were always charge sustain by net energy change (less than 1 of fuel energy) and SOC (delta = 0)
UDDS2 Highway2 US062 JC082 NEDC2 WLTP2Fuel economy [mpge] 935 924 550 948 859 908Hydrogen consumed [kWh] 260 363 476 176 261 634Fuel consumption [kWhmi] 3497 3540 5942 3448 3805 4416Average FC Stack efficiency [] 648 634 618 642 624 618Average FC System efficiency [] 622 611 481 577 593 569Vehicle efficiency [] 657 576 571 650 582 590FC power covering 90 demand [kW] 12 20 40 12 14 22
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CONSUMPTION WITH BATTERY STATE OF CHARGE CORRECTION
33
Battery Charging Battery Discharging
Charge sustaining
1
1
3
3
2
2
Hot start UDDS test data
SOC merge to
same levelsUDDS drive cycles
Same approach as for other Hybrid Electric Vehicles
Note The battery SOC was adjusted to high and low before the beginning of specific UDDS tests to obtain the results below
SOCStart[]
SOCEnd[]
Fuel Consumption [le100km]
Net energy change[Vzc x Ah]
Low initial SOC 51 60 278 -257Charge balanced SOC 59 605 252 -57High initial SOC 73 585 221 242
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
HYDROGEN CONSUMPTION CAN BE ESTIMATED USING THE IDEAL GAS LAW
34
Outside H2 Storage
Air gas HYUHP300 which is a 12-pack of DOT 3AA2400 cylinders The liquid volume for each cylinder is 498L The H2 storage is exposed to outside temperature
Initi
al p
ress
ure
End
pres
sure
Post
reco
very
pre
ssur
e
119875119875119875119875 = 119898119898119898119898119898119898Ideal Gas Law Equation
Ideal gas
constant
Total mass of H2 for the given conditions
It was not possible to measure the gas temperature in the
cylinders The outside temperature on 12517 was ~4C
Volume of 12 3AA2400 cylinders S mall portion
of piping at the high pressure is unknown
High pressure sensor
enclosure
12-pack pressure is measured
Pressure used for
end mass
Pressure used for
initial mass
Temperatures evens out in a recovery
phase between tests
Pressure used to calculate a second post test H2 mass
This analysis is only anecdotal This was not a goal associate
with the testing
Note The compressibility factor should be included in the calculations but was left out due to time constrained
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
HYDROGEN CONSUMPTION BASED ON CYLINDER PRESSURE VOLUME AND ldquoTEMPERATURErdquo
35
Initi
al p
ress
ure
[psi
g]En
d pr
essu
re
[psi
g]Po
st-r
ecov
ery
pres
sure
[psi
g]R
ecov
ery
time
[min
]M
ass
flow
met
er
[g]
PV m
ass
end
of
test
[g]
PV m
ass
post
re
cove
ry [g
]
UDDS1 1011 974 984 10 817 1332 989UDDS2 984 948 955 3 781 1301 1044UDDS3 955 924 931 10 757 1121 854Highway1 931 894 894 025 1104 1359 1340Highway2 894 861 867 17 1090 1177 968US061 867 817 823 15 1474 1801 1594US062 823 773 780 341 1430 1808 1544Total sequence 1011 NA 780 NA 7453 NA 8333
The hydrogen consumed based on the ideal gas law is 12 high than the
integrated hydrogen mass flow meter
Note To truly compare the two measurement methods it would require to have a temperature controlled enclosure for the hydrogen storage and wait until the gas and cylinders stabilized at the enclosure temperature after the test before calculating the mass of hydrogen post test The volume of high pressure piping should be known as well
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL SYSTEM OVERVIEW
36
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL POLARIZATION CURVE
37
095
081
068
054
0 41
027
014
0
Cel
l Vol
tage
[V]
Stac
k Vo
ltage
[V]
2530 042 083 127 169 211
Stack current [A]
Current density [Acm2]
Stack and cell assumptions bull 237 cm2 active area (1)
bull 370 cells in stack
Idle Voltage
Voltage under load
bull These data points come for the steady state mapping analysis (slide 16)
bull Red data cloudbull The 10 Hz data to generate
this graphs is from NEDCx2 UDDSx3 Highwayx2 US06x2 maximum acceleration x3 and steady state speed tests
(1) B James NREL ldquoFuel Cell Systems Analysisrdquo 2017 DOE FCTO AMR Project ID FC163
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
38
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
Toyota describes the modified their air supply system (1) which increases fuel economy and the 3D fine-mesh flow structure (2) which enables the humidifier-less fuel cell system1 Hasegawa T Imanishi H Nada M and Ikogi Y Development of the Fuel Cell System in the Mirai FCV SAE Technical Paper 2016-01-1185 2016 doi1042712016-01-11852 Konno N Mizuno S Nakaji H and Ishikawa Y Development of Compact and High-Performance Fuel Cell Stack SAE Int J Alt Power 4(1)2015 doi1042712015-01-1175
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL STACK OPERATING PARAMETERS AND HUMIDIFICATION CONTROL MECHANISMS
39
H2
Air
Anod
e
Cat
hodeFC air
compressor FC H2 Recirculationpump
FC Water drain valve
Humid air
Wat
er
e-Basic Fuel Cell System
Majority of UDDS and
Highway cycle
The higher the load the higher the water production the higher the water drainage
The valve is open or closed This graph show the average over the test period 0 is closed and 100 is open
The recirculation pump has to work harder at higher loads to overcome the higher hydrogen
flow and pressure
The air pressure is very low below 30 kW and increase with correlated with the FC air
compressor power presented in a previous slide
Plateau of hydrogen flow maybe an artefact of the high load testing
The air and hydrogen mass flow increase
with the electric output or load
Liquid-gas separator
Inje
ct1
Inje
ct2
Inje
ct3
Pres
sure
regu
lato
r
Air p
ress
ure
Reg
ulat
ing
valv
e
Exhaust
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
40
~1 minute window of fuel cell lsquoidle timersquo
FC back up to OCV at startup
FC stack open circuit voltage drops from 315V (085Vcell) to 282V (076Vcell) after 64 secondsFuel Cell
stack making power
Fuel Cell stack making
power
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL IDLE OPERATION ON DRIVE CYCLE
41
Air flow is around 50 NLpm and the compressor slows to an average of
105 rpm
Stack stops producing power
The hydrogen flow is stopped
Note This is an idle period from the NEDC drive cycle which was used earlier to provide an overview of the vehicle operation
Slower hydrogen decay is due to mechanical filter
through flow dynamics in the metering system
Occasional hydrogen
pressure peak from 19 kPa
to 35 kPa
Each hydrogen pressure peak is coupled with
some hydrogen flow The hydrogen
consumption is zero most of the time when the fuel cell lsquoidlesrsquo
H2 ldquopuffrdquo of ~0035 g
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
SPECIAL 1 HOUR IDLE TEST TO QUANTIFY THE IDLE HYDROGEN CONSUMPTION
42
First 505 second of UDDS cycle
(Phase1) to warm up the powertrain
FC system reaches 1-2kW power level
which is enough power to maintain
battery SOC
Vehicle stopped transmission in
park ignition ON Scenario waiting
in parking lot
The battery SOC drops
as accessory loads draw power over
time
Low SOC triggers
restart of FC system
FC system recharge the
battery at 5~6kW
Idle phase
Low power phase
FC voltage drops from
315V to 80V
Idle phase FC system has zero
electric output
Zoom
1
Zoom
2
Idle phase More on next slidebull 1401 seconds duration bull 171 g of H2 consumed bull 74 V Stack voltagebull 000 Wh Stack electric outputIdle fuel flow rate 439 ghr
The fuel cell idle fuel flow rate is 439 ghr That is low compared to 754 g consumption on the UDDS (23
minute test or 198 ghr)
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
MORE DETAILS ON SPECIAL 1 HOUR IDLE TEST
43
Hydrogen pressure bumps from 20 kPa to 35 kPa every ~
40 seconds
Hydrogen pumps step up from
20W to 60W for 10~12 second
every 5 minutes
Air flow steps up from 50 to 200
NLpm for 4 second every 3 minutes
Hydrogen puffs corresponding to
hydrogen pressure bumps
FC starts once the battery
SOC drops to 455
10 to 6 kW stack output until battery charge to
50
1 to 2 kW stack output with fuel flow pulsing and it appear that the stack is air starved (compressor speed 660 rpm) 4 minutes later
the FC settles to 500W to sustain accessory loads
Zoom 2 at start of low power phaseZoom 1 in idle phase
During the idle phases the FC system is
starved of hydrogen to maintain a low open
circuit voltage
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F
44
FC stack peak output of 1116kW
Battery assist (26kW peak)
Battery SOC drop from 625 to 39
FC stack output drops
after ~30 seconds
FC stack output settles at 73kW after 6 minutes Higher
relative air flow to the vehicle could increase the maximum
continuous power output
273 mph is the final speed on a
25 grade
Compressor peak speed
9600 rpm peak power 17 kW
All 3 hydrogen injectors used for
peak power
Drainage valve is
open more often during peak power
Notebull This testing was performed at 72F
test cell temperature bull The test cell fan was matching the air
flow to the vehicle speed
Peak power is110 kW for a limited
time period Continuous power is ~75 kW at 72F
ambient conditions
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash NO FUEL STARVATION
45
Pressure drop due to
higher flows
The actual low range pressure is above
the low range target pressure
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
MAXIMUM POWER TESTING ON A 25 GRADE AT 72F ndash THERMAL CONDITIONS
46
Fuel Cell
System
Mirai Coolant System
Fan
1Fa
n2
Rad
iato
r
Water pump
Rotaryvalve
Heater
Hea
ter c
ore
HVA
C
wat
erpu
mp
3 way valve
Test
Cel
l Fan
Note Simplified cooling systemSome components are missing
The water pump the rotary valve
both fans switch to full cooling mode
FC coolant and component
temperate reach a steady state
The available data appears to show that the fuel cell
system was not limited by thermal conditions at the
vehicle system level
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
REPEATED MAXIMUM ACCELERATIONSTO 80 MPH AT 72F AMBIENT TEMPERATURE
47
The fuel cell system reliably and consistently produces
peak power for repeat accelerations runs
Four repeated maximum
acceleration tests Run 0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Battery assist is reduced as the battery state of charge drops
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
IMPACT OF TEMPERATURES 0F 20F 72F AND 95F+850WM2
48
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
TESTING ACROSS A RANGE OF TEMPERATURES
49
Hot testing 95F + 850 Wm2
(35C)
Cold testing20F
(-7C)
Standard testing 72F
(22C)
Cold testing0F
(-18C)
Climate control72F auto
Climate controlOFF
Climate control72F auto
Climate control72F auto
Environmental Test Cell
Thermal testingbull The hood is closed in all casesbull The facility fan blows air
dynamically across the front of the car at the same speed as the vehicle speed
bull Cold start means that the vehicle was thermally lsquosoakedrsquo at the target temperature for over 12 hours (typically 16 hours)
Test Sequencebull Cold start UDDS 1 + Hot start
UDDS 2 amp UDDS 3bull Pair of Highway cycles (except
for 0F)bull Pair of US06 cycles
(expect for 0F)
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
IMPACT OF AMBIENT TEMPERATURE AND CLIMATE CONTROL ON FUEL ECONOMY
50
Fuel Economy Vehicle Efficiency
Increasing operating temperature
Closer to target cabin temperature
Increasing driving
intensity
Note At 72F the climate control was off At 20F and 95F the climate control was set to 72F auto
+850Wm2+850Wm2
SAE
J295
1
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
WHERE DID THE ENERGY GO
51
No electric heating used after some driving because enough power-train waste heat exists
(cold start)
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
HVAC AC is a permanent energy draw
Climate control penalty is worst on
initial test (cabin temperature pull up or
pull down) +1
57
+42
+60
+28
+23
+37
+17
+6
+9
+12
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
20F 72F 95F
COMPARING A FUEL CELL VEHICLE TO A BATTERY ELECTRIC VEHICLE ACROSS TEMPERATURES
52
(cold start)
+157
+42
Ambient test conditions
20F
72F
95F+850Wm2
Climate control bull At 72F the climate
control was off bull At 20F and 95F the
climate control was set to 72F auto
(cold start)
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
UDDS DRIVE CYCLES TEST SEQUENCE AT 0F 20F AND 72F
53
(cold start)
Ambient test conditions
0F
20F
72F
Climate control bull the climate control
was set to 72F auto in all cases
+235
+157
The heater is still needed after 2 hours
of driving at 0F
+60
+23
+134
+127
Significant energy consumption increase
on cold start at freezing temperatures
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL SYSTEM SHUTDOWN AT 72F
54
72F Ambient temperature
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
FC voltage slowly decays
Driver turns the car OFF after completing a
UDDS drive cycle
No FC stack power
A bit of hydrogen
flow
Purpose of the shut down Leave the FC (esp PEM) in a lsquogoodrsquo state especially from a water management perspective
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL SYSTEM SHUTDOWN AT 20F
55
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Longer (70 seconds) air flow
and hydrogen flow to dry out the stack
Driver turns the car OFF after completing a
UDDS drive cycle
No stack power No
electric heater power
20F Ambient temperature
Purpose of the shut down at freezing temperatures Evacuate the water within the FC stack to prevent internal freezing
Periodically the vehicle may drain excess water This can also be performed manually
Note When transitioning into freezing temperatures the vehicle automatically lsquowakes uprsquo and performs a similar routine
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL SYSTEM SHUTDOWN AT 0F
56
Last lsquohillrsquo of the UDDS cycle
Igni
tion
off
Driver turns the car OFF after completing a
UDDS drive cycle
0F Ambient temperature
Still 70 seconds of air flow and
hydrogen flow to dry out the stack
Higher air flow as
compared to 20F
Still no stack power and no electric heater
power
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FC START UP ON COLD START IN FREEZING TEMPERATURES
57
72F 20F 0F
Immediate and steady FC voltage on start up
FC voltage varies at low levels after start (for ~100 second)
Large amount of H2 are pushed through the stack
FC voltage low with minimal initial power output FC
voltage does not stabilized until 150 seconds
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL START UP ON COLD START AT 72F
58
72F
First 300 seconds of UDDS cold start test
FC stack no producing
power until the wheels spin
Driver turn the car ON after an overnight soak at 72F
FC stack voltage is comes to OCV
as soon it produces power
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL START UP ON COLD START AT 20F
59
20F
Large increase in H2 flow
compared to 72F cold start
FC produces power before wheels spin
Driver turn the car ON after an overnight soak at 20F
First 300 seconds of UDDS cold start test
Electric heater power
consumption
FC stack voltage is low
for the first 100 seconds
FC stack voltage is back to
normal operating ranges
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
FUEL CELL START UP ON COLD START AT 0F
60
0F full weekend soak (2+ days)
Note1 Due to time constraints this test could not be repeated It is unclear if is repeatable
Note2 A 0F cold start UDDS test which only 16 hours of soak was performed as well No drive quality issues were reported on that test
Intermittent power delivery Driver
reported lsquobuckingrsquo of vehicle on dyno
FC stack voltage back to lsquonormalrsquo
Driver turn the car ON after 2+ days of 0F
temperatures The wheel spin after 20 seconds
First 300 seconds of UDDS cold start test
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
MAXIMUM ACCELERATION AT 0F
61
0F
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 167 1099 1006
2 187 1126 1030
3 202 1135 1038
4 203 1140 1046
Run72F
0-80 mph[s]
PeakStack[kW]
AverageStack[kW]
1 29 983 860
2 206 1138 1044
3 185 1138 1055
Note Force cooling 0F for 2 hours after the 5 UDDS cycles
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
HILL CLIMB WITH SOLAR LOAD
62
Hill Climb testbull 62 mph (100 kmh)bull 6 gradebull At 4810 lbs test weight (GVW)bull 95F ambient temperaturebull 850 Wm2 solar loadbull Climate control set to 72F auto Realistic worst case scenario
6 grade
Continuous power of 63kW after 30 minutes
The Mirai maintained the 62 mph speed target in
these conditions
95F + 850Wm2
Note This is higher continuous power compared to the 25 grade test The vehicle speed in grade test was 27 mph That low speed resulted in low air flow from the facility fan in front the vehicle and therefore in a lower heat rejection potential
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
CONCLUSION AND TAKEAWAYS FROM TESTING A 2017 TOYOTA MIRAI FUEL CELL VEHICLE
63
The FC is starved of H2 when not used
The fuel cell idle fuel flow rate is 439 ghr
10Hz data will be available publicly at
wwwanlgovd3
Contrary to BEVs FCVs have enough waste heat
to keep a cabin warm
Peak power is 110 kW for a limited time period Continuous power will
vary from ~50 kW to ~75 kW depending on thermal conditions
and vehicle speed
Fuel Cell peak efficiencybull Stack 660bull System 637
Fuel Cell dominant hybrid electric
vehicle
Fuel Cell Stack and System Efficiency Analysis
Energy Analysis Across Temperatures
0F to 95F Testing Completed
Corelated FC system parameters and
operation to output
Power output of the FC system
conditioning to 0F is limited initially
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
TECHNOLOGY ASSESSMENT
OF A FUEL CELL VEHICLE2017 TOYOTA MIRAI
HENNING LOHSE-BUSCH MICHAEL DUOBA KEVIN STUTENBERG SIMEON ILIEV MIKE KERNArgonne National Laboratory
June 2018 (v10)Argonne National Laboratory IL
WWWANLGOVD3
BRAD RICHARDS MARTHA CHRISTENSONTransport Canada AARON LOISELLE-LAPOINTE Environment and Climate Change Canada
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
BACK UP MATERIAL
65
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
bull Special instrumentation High precision power analyzers
(testing and charging) CAN decoding and recording OCR scan tool recording Direct fuel flow metering Infrared temperature camera In cylinder pressure indicating
systems In-situ torque sensor measurement 5 gas emissions dilute bench with
CVS (modal and bag emissions analysis)
FTIR Mobile Emissions unit Raw and Fast HC and NOx bench Aldehyde bench for alcohol fuels
bull Test cell specifications 4WD chassis dynamometer
- Variable wheel base (180inches max)- 250 hpaxle- 300 to 12000 lb inertia emulation
Radiant sun energy emulation of 850Wm2 (adjustable)
Variable speed cooling fan (0ndash62mph)
Gaseous fuel and hydrogen capable Diesel Dilution tunnel PM HFIDbull Thermal chamber EPA 5 cycle capable
(20oF 72oF and 95oF + 850Wm2 solar load) Demonstrated as low as 0oF Intermediate temperatures possible
Advanced Powertrain Research Facility4WD Chassis Dynamometer Thermal Test Cell
ver April 2013
bull Research features Modular and custom DAQ with real
time data display Process water available for cooling
of experiment components Available power in test cell
- 480VAC 200A- 208VAC 100A
ABC 170 Power supply capable to emulate electric vehicle battery
Custom Robot Driver with adaptive learning
Several vehicle tie downs - chains low profile rigidhellip- 2 3 and 4 wheel vehicle capable
Expertise in testing hybrid and plug-in hybrid electric vehicles battery electric vehicles and alternative fuel vehicles
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
HOW ARE VEHICLES TESTED FOR ENERGY CONSUMPTION
67
Chassis dynamometer test cell A driver accelerates and brakes to follow a drive trace
A drive trace is a defined speed profile as a function of time
Simple physics are used to emulate the proper load at the wheel to emulate the real world
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
EPArsquoS 5 CYCLES TESTS FOR FUEL ECONOMY LABELEPA implemented the new 5 Cycle Fuel Economy Label to close the gap to real world fuel economy the consumer can expect
HWFET 75 F
0 200 400 600 800 1000 1200 1400 16000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
PhaseTrace
UDDS 20 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace
US06 75 F
0 100 200 300 400 500 6000
10
20
30
40
50
60
70
80
90
Time [s]
Spee
d [m
ph]
Phase x10Trace
0 100 200 300 400 500 6000
10
20
30
40
50
60
Time [s]
Spee
d [m
ph]
Phase x10Trace
SC03 95 FFTP UDDS 75 F
0 200 400 600 800 1000 1200 14000
10
20
30
40
50
60
70
80
Time [s]
Spee
d [m
ph]
Phase x10Trace1 Cold start
2 Hot start
Classic cycles Aggressive cycle Extreme Temperatures
Parts of these cycles compute into a
City and a Highway Fuel Economy
CAFE (only city and highway)
+850 Wm2
68
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
PSYCHOLOGY OF FUEL ECONOMY VS FUEL CONSUMPTION
69
FE 50FC 33
FE 25FC 20Truck
Compact HEV
Plug-in Hybrid Electric Vehicle
Fuel Economy = Distance Fuel
125 gal saved over 10000 mi
~84 gal saved over 10000 mi
Fuel consumption= Fuel Distance
Trucks amp SUVs lsquoThe MPG illusionrsquo
If little fuel is used FE lsquoexplodesrsquo to a large number and becomes misleading
FC is a better scale for very efficient vehicles
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
ELECTRIFICATION COMPLICATES EFFICIENCY CALCULATIONS
2012 Focus 2L 2013 Focus BEVBattery
Motor
Fuel
Engine Bi-directional power flow
Irreversible power flow
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]En
ergy
[Wh]
0 200 400 600 800 1000 1200 14000
500
1000
1500
2000
Time [s]
Ener
gy [W
h]
in
out
EnergyEnergyEfficiency =
Regenerative braking reverses the power flow and charges the high voltage battery (In amp Out reverse)
Conventional Vehicle Electric Vehicle
SAE J2951 defines ldquoCycle Energyrdquo as the integration of positive powerat the wheel over the cycle (regenerative braking = ldquofreerdquo energy)
70
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
HIGHER EFFICIENCY POWERTRAINS ARE MORE SENSITIVE
Conventional BatteryElectric(incl charger)
Hybrid Electric
Conventional2012 Ford Focus2013 Jetta TDI2013 Chevy Cruze Diesel
Hybrid Electric
2013 Chevy Malibu Eco2013 VW Jetta HEV2013 Honda Civic HEV2010 Prius 2013 Ford Cmax HEV2014 Honda Accord HEV
Battery Electric
2012 Nissan Leaf BEV2013 Nissan Leaf BEV2015 BMW i3 BEV2015 Chevy Spark BEV2013 Ford Focus BEVData note
UDDS hot start
Peak ICE effPeak ICE eff + regen
Observationsbull Everything matters down to
the accessory loads in electric vehicles
SAE
J295
1 de
finiti
on
71
NET ENERGY CHANGE OF THE BATTERY DETERMINES IF A HEV IS CHARGE SUSTAINING OVER THE TEST CYCLE Net energy change (NEC) = net battery energy delta
over the test cycle A test is considered charge sustaining when the NEC of
battery is less then 1 of fuel energy used over the test cycle If an HEV discharges itrsquos battery over a test cycle the
fuel consumption will be increased
Sample data - 11 Sonata HEV - UDDS
ChargeSustainin
g
Fuel
Con
sum
ptio
n A
B
Net Energy ChangeBattery DischargingBattery Charging
72
SAFETY AND METERING PANEL
73
FL
FH
Display
FH
Display
FL
P
R
OP
PPP
PQC
OP
H2
Controller
Vent lines to the roof
Vehicle delivery pressure sensor
Manual valves
Manual pressure regulator (removed
for Mirai testing before of pressure drop ndash pressure set
the outside panel)
Over pressure relief valve
Quick connect fitting
H2 supply
Hydrogen sensor
Flexible fuel hose to vehicle with manual valve
Over pressure sensor
Pressure gage
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
| UDDS | UDDS | Highway | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Engine Power [kW] | Motor Power [kW] | DOH | ETW | A | B | C | Charge eff | 20F | 72F | 95F | 745 | Positive tractive effort [J] | Net Tractive Effort [J] | Calc Fuel energy [Wh] | Vehicle Eff | ||||||||||||||||||||||||||||||||||||||||||||||
| 56 | Chevy Spark BEV | 0 | 1 | 100 | 3157 | 2336367 | 03946 | 001245 | 086 | 61508026 | 4075271030165 | 61508013 | 1976844638756 | 61508017 | 2374556923239 | 134E+03 | 7402321 | 3036076917473 | 441 | 14727492558734 | 910 | 17690449078129 | 758 | ||||||||||||||||||||||||||||||||||||||
| 51 | Honda Accord HEV | 105 | 125 | 054 | 3900 | 198098 | 066306 | 0008675 | 61506029 | 413500078329 | 61506016 | 611989540753 | 61506022 | 495472462003 | 153E+03 | 7504197 | 60293 | 254 | 40738 | 377 | 50318 | 305 | 61506031 | 4969318168 | 61506018 | 6277281085 | 61506024 | 5769464184 | |||||||||||||||||||||||||||||||||
| 55 | BMW i3 BEV | 0 | 1 | 100 | 3182 | 236 | 06633 | 00117 | 086 | 61505028 | 2909242770484 | 61505019 | 1807534323347 | 61505024 | 2134936607337 | 143E+03 | 8537203 | 21673858640108 | 662 | 13466130708938 | 1065 | 15905277724662 | 901 | ||||||||||||||||||||||||||||||||||||||
| 52 | Chevy Cruze Diesel | 1 | 0 | 000 | 3808 | 22636 | 123944 | 0001929 | 61410029 | 327286752498 | 61410013 | 339760515087 | 61410022 | 255440338748 | 165E+03 | 9438627 | 87171 | 189 | 83970 | 197 | 111689 | 148 | 61410032 | 5105943139 | 61410015 | 5449035803 | 61410024 | 4493367482 | |||||||||||||||||||||||||||||||||
| 53 | Ford Focus BEV | 0 | 1 | 100 | 3948 | 364265 | 051941 | 0015143 | 086 | 61408020 | 4856088605889 | 61408014 | 2378234640621 | 61408024 | 2692166362755 | 176E+03 | 104E+03 | 36177860113871 | 486 | 17717848072629 | 992 | 20056639402523 | 876 | 61408019 | 3600097210449 | 61408013 | 2258314668938 | 61408023 | 2278261648663 | ||||||||||||||||||||||||||||||||
| 54 | 2013 Nissan Leaf BEV | 0 | 1 | 100 | 3302 | 3191 | 011159 | 0017757 | 086 | 61403063 | 3653727601704 | 61403013 | 2008257870394 | 61403023 | 2609745920995 | 144E+03 | 8255759 | 27220270632697 | 529 | 14961521134438 | 962 | 1944260711141 | 740 | ||||||||||||||||||||||||||||||||||||||
| 44 | 2013 Ford Cmax HEV | 105 | 88 | 046 | 3966 | 2175 | 0365 | 001859 | 61309048 | 440355115297 | 61309015 | 60386902158 | 61309033 | 519167672053 | 160E+03 | 8058958 | 56616 | 282 | 41286 | 387 | 48022 | 332 | |||||||||||||||||||||||||||||||||||||||
| 37 | 2013 VW Jetta HEV | 110 | 28 | 020 | 3625 | 2788 | 00543 | 001618 | 61308073 | 389427509299 | 61308040 | 491567289342 | 61308059 | 328539742594 | 144E+03 | 7150819 | 64020 | 225 | 50718 | 284 | 75885 | 190 | 61308076 | 4805457022 | 61308042 | 558665186 | 61308061 | 4378545267 | |||||||||||||||||||||||||||||||||
| 40 | 2013 Honda Civic HEV | 82 | 23 | 022 | 3208 | 309399 | -015033 | 0020877 | 61306025 | 419490752528 | 61306002 | 562149307107 | 61306010 | 417005188764 | 136E+03 | 7427576 | 59432 | 228 | 44350 | 306 | 59787 | 227 | 61306034 | 5129333612 | 61306004 | 603617343 | 61306013 | 5405897749 | |||||||||||||||||||||||||||||||||
| 43 | 2013 Chevy Malibu Eco | 135 | 15 | 010 | 3906 | 304035 | 027111 | 0014436 | 61302096 | 28087890276 | 61302059 | 302732847887 | 61302072 | 23212747593 | 159E+03 | 8235604 | 88762 | 179 | 82354 | 193 | 107404 | 148 | 61302099 | 4447094845 | 61302061 | 4798771055 | 61302075 | 4497794875 | |||||||||||||||||||||||||||||||||
| 38 | 2013 Jetta TDI | 100 | 0 | 000 | 3516 | 301456 | 03765 | 00157 | 61210095 | 3156347848 | 61210113 | 3584 | 61210103 | 258991409 | 154E+03 | 8869525 | 90389 | 170 | 79603 | 193 | 110158 | 140 | |||||||||||||||||||||||||||||||||||||||
| 20 | 2010 Prius | 72 | 60 | 045 | 3375 | 18501 | 00223 | 00181 | 61205034 | 4977323036 | 61205007 | 6842 | 61205018 | 4397723698 | 129E+03 | 591872 | 50090 | 258 | 36438 | 354 | 56691 | 228 | |||||||||||||||||||||||||||||||||||||||
| 31 | 2012 Ford Focus | 119 | 0 | 000 | 3250 | 2718 | 02369 | 00193 | 61210072 | 3103724465 | 61210064 | 3317 | 61210081 | 2619279501 | 145E+03 | 8382989 | 80327 | 180 | 75162 | 193 | 95184 | 152 | |||||||||||||||||||||||||||||||||||||||
| 26 | 2012 Nissan Leaf | 0 | 80 | 100 | 3746 | 4106 | -03082 | 00253 | 61203027 | 4463972548082 | 61203033 | 2342588349312 | 61203054 | 2818446220095 | 160E+03 | 895247 | 33256595483208 | 482 | 17452283202378 | 918 | 20997424339707 | 763 | |||||||||||||||||||||||||||||||||||||||
| Diesel | Gasoline | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 42796 | 4300774 | Jg | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 18399 | 18490 | BTUlbm | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 0851 | 074 | gml | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 20F | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 72F | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 95F+sun | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2013 Volt PHEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2013 Toyota Prius PHEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| BMW i3 Rex | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Smart Electric | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Mazda 3 i-eloop | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Mitsubshi MiEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Honda Accord PHEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2013 Ford Cmax Energi | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Ford Fusion Energi | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Peugeot 3008 Hybrid 4 |
| 1 | 1 | 1 | |||
| 05434782609 | 05434782609 | 05434782609 | |||
| 1 | 1 | 1 | |||
| 0 | 0 | 0 | |||
| 1 | 1 | 1 | |||
| 1 | 1 | 1 | |||
| 04559585492 | 04559585492 | 04559585492 | |||
| 02028985507 | 02028985507 | 02028985507 | |||
| 0219047619 | 0219047619 | 0219047619 | |||
| 01 | 01 | 01 | |||
| 0 | 0 | 0 | |||
| 04545454545 | 04545454545 | 04545454545 | |||
| 0 | 0 | 0 | |||
| 1 | 1 | 1 |
NET ENERGY CHANGE OF THE BATTERY DETERMINES IF A HEV IS CHARGE SUSTAINING OVER THE TEST CYCLE Net energy change (NEC) = net battery energy delta
over the test cycle A test is considered charge sustaining when the NEC of
battery is less then 1 of fuel energy used over the test cycle If an HEV discharges itrsquos battery over a test cycle the
fuel consumption will be increased
Sample data - 11 Sonata HEV - UDDS
ChargeSustainin
g
Fuel
Con
sum
ptio
n A
B
Net Energy ChangeBattery DischargingBattery Charging
72
SAFETY AND METERING PANEL
73
FL
FH
Display
FH
Display
FL
P
R
OP
PPP
PQC
OP
H2
Controller
Vent lines to the roof
Vehicle delivery pressure sensor
Manual valves
Manual pressure regulator (removed
for Mirai testing before of pressure drop ndash pressure set
the outside panel)
Over pressure relief valve
Quick connect fitting
H2 supply
Hydrogen sensor
Flexible fuel hose to vehicle with manual valve
Over pressure sensor
Pressure gage
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
| UDDS | UDDS | Highway | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Engine Power [kW] | Motor Power [kW] | DOH | ETW | A | B | C | Charge eff | 20F | 72F | 95F | 745 | Positive tractive effort [J] | Net Tractive Effort [J] | Calc Fuel energy [Wh] | Vehicle Eff | ||||||||||||||||||||||||||||||||||||||||||||||
| 56 | Chevy Spark BEV | 0 | 1 | 100 | 3157 | 2336367 | 03946 | 001245 | 086 | 61508026 | 4075271030165 | 61508013 | 1976844638756 | 61508017 | 2374556923239 | 134E+03 | 7402321 | 3036076917473 | 441 | 14727492558734 | 910 | 17690449078129 | 758 | ||||||||||||||||||||||||||||||||||||||
| 51 | Honda Accord HEV | 105 | 125 | 054 | 3900 | 198098 | 066306 | 0008675 | 61506029 | 413500078329 | 61506016 | 611989540753 | 61506022 | 495472462003 | 153E+03 | 7504197 | 60293 | 254 | 40738 | 377 | 50318 | 305 | 61506031 | 4969318168 | 61506018 | 6277281085 | 61506024 | 5769464184 | |||||||||||||||||||||||||||||||||
| 55 | BMW i3 BEV | 0 | 1 | 100 | 3182 | 236 | 06633 | 00117 | 086 | 61505028 | 2909242770484 | 61505019 | 1807534323347 | 61505024 | 2134936607337 | 143E+03 | 8537203 | 21673858640108 | 662 | 13466130708938 | 1065 | 15905277724662 | 901 | ||||||||||||||||||||||||||||||||||||||
| 52 | Chevy Cruze Diesel | 1 | 0 | 000 | 3808 | 22636 | 123944 | 0001929 | 61410029 | 327286752498 | 61410013 | 339760515087 | 61410022 | 255440338748 | 165E+03 | 9438627 | 87171 | 189 | 83970 | 197 | 111689 | 148 | 61410032 | 5105943139 | 61410015 | 5449035803 | 61410024 | 4493367482 | |||||||||||||||||||||||||||||||||
| 53 | Ford Focus BEV | 0 | 1 | 100 | 3948 | 364265 | 051941 | 0015143 | 086 | 61408020 | 4856088605889 | 61408014 | 2378234640621 | 61408024 | 2692166362755 | 176E+03 | 104E+03 | 36177860113871 | 486 | 17717848072629 | 992 | 20056639402523 | 876 | 61408019 | 3600097210449 | 61408013 | 2258314668938 | 61408023 | 2278261648663 | ||||||||||||||||||||||||||||||||
| 54 | 2013 Nissan Leaf BEV | 0 | 1 | 100 | 3302 | 3191 | 011159 | 0017757 | 086 | 61403063 | 3653727601704 | 61403013 | 2008257870394 | 61403023 | 2609745920995 | 144E+03 | 8255759 | 27220270632697 | 529 | 14961521134438 | 962 | 1944260711141 | 740 | ||||||||||||||||||||||||||||||||||||||
| 44 | 2013 Ford Cmax HEV | 105 | 88 | 046 | 3966 | 2175 | 0365 | 001859 | 61309048 | 440355115297 | 61309015 | 60386902158 | 61309033 | 519167672053 | 160E+03 | 8058958 | 56616 | 282 | 41286 | 387 | 48022 | 332 | |||||||||||||||||||||||||||||||||||||||
| 37 | 2013 VW Jetta HEV | 110 | 28 | 020 | 3625 | 2788 | 00543 | 001618 | 61308073 | 389427509299 | 61308040 | 491567289342 | 61308059 | 328539742594 | 144E+03 | 7150819 | 64020 | 225 | 50718 | 284 | 75885 | 190 | 61308076 | 4805457022 | 61308042 | 558665186 | 61308061 | 4378545267 | |||||||||||||||||||||||||||||||||
| 40 | 2013 Honda Civic HEV | 82 | 23 | 022 | 3208 | 309399 | -015033 | 0020877 | 61306025 | 419490752528 | 61306002 | 562149307107 | 61306010 | 417005188764 | 136E+03 | 7427576 | 59432 | 228 | 44350 | 306 | 59787 | 227 | 61306034 | 5129333612 | 61306004 | 603617343 | 61306013 | 5405897749 | |||||||||||||||||||||||||||||||||
| 43 | 2013 Chevy Malibu Eco | 135 | 15 | 010 | 3906 | 304035 | 027111 | 0014436 | 61302096 | 28087890276 | 61302059 | 302732847887 | 61302072 | 23212747593 | 159E+03 | 8235604 | 88762 | 179 | 82354 | 193 | 107404 | 148 | 61302099 | 4447094845 | 61302061 | 4798771055 | 61302075 | 4497794875 | |||||||||||||||||||||||||||||||||
| 38 | 2013 Jetta TDI | 100 | 0 | 000 | 3516 | 301456 | 03765 | 00157 | 61210095 | 3156347848 | 61210113 | 3584 | 61210103 | 258991409 | 154E+03 | 8869525 | 90389 | 170 | 79603 | 193 | 110158 | 140 | |||||||||||||||||||||||||||||||||||||||
| 20 | 2010 Prius | 72 | 60 | 045 | 3375 | 18501 | 00223 | 00181 | 61205034 | 4977323036 | 61205007 | 6842 | 61205018 | 4397723698 | 129E+03 | 591872 | 50090 | 258 | 36438 | 354 | 56691 | 228 | |||||||||||||||||||||||||||||||||||||||
| 31 | 2012 Ford Focus | 119 | 0 | 000 | 3250 | 2718 | 02369 | 00193 | 61210072 | 3103724465 | 61210064 | 3317 | 61210081 | 2619279501 | 145E+03 | 8382989 | 80327 | 180 | 75162 | 193 | 95184 | 152 | |||||||||||||||||||||||||||||||||||||||
| 26 | 2012 Nissan Leaf | 0 | 80 | 100 | 3746 | 4106 | -03082 | 00253 | 61203027 | 4463972548082 | 61203033 | 2342588349312 | 61203054 | 2818446220095 | 160E+03 | 895247 | 33256595483208 | 482 | 17452283202378 | 918 | 20997424339707 | 763 | |||||||||||||||||||||||||||||||||||||||
| Diesel | Gasoline | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 42796 | 4300774 | Jg | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 18399 | 18490 | BTUlbm | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 0851 | 074 | gml | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 20F | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 72F | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 95F+sun | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2013 Volt PHEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2013 Toyota Prius PHEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| BMW i3 Rex | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Smart Electric | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Mazda 3 i-eloop | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Mitsubshi MiEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Honda Accord PHEV | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2013 Ford Cmax Energi | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Ford Fusion Energi | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Peugeot 3008 Hybrid 4 |
NET ENERGY CHANGE OF THE BATTERY DETERMINES IF A HEV IS CHARGE SUSTAINING OVER THE TEST CYCLE Net energy change (NEC) = net battery energy delta
over the test cycle A test is considered charge sustaining when the NEC of
battery is less then 1 of fuel energy used over the test cycle If an HEV discharges itrsquos battery over a test cycle the
fuel consumption will be increased
Sample data - 11 Sonata HEV - UDDS
ChargeSustainin
g
Fuel
Con
sum
ptio
n A
B
Net Energy ChangeBattery DischargingBattery Charging
72
SAFETY AND METERING PANEL
73
FL
FH
Display
FH
Display
FL
P
R
OP
PPP
PQC
OP
H2
Controller
Vent lines to the roof
Vehicle delivery pressure sensor
Manual valves
Manual pressure regulator (removed
for Mirai testing before of pressure drop ndash pressure set
the outside panel)
Over pressure relief valve
Quick connect fitting
H2 supply
Hydrogen sensor
Flexible fuel hose to vehicle with manual valve
Over pressure sensor
Pressure gage
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
NET ENERGY CHANGE OF THE BATTERY DETERMINES IF A HEV IS CHARGE SUSTAINING OVER THE TEST CYCLE Net energy change (NEC) = net battery energy delta
over the test cycle A test is considered charge sustaining when the NEC of
battery is less then 1 of fuel energy used over the test cycle If an HEV discharges itrsquos battery over a test cycle the
fuel consumption will be increased
Sample data - 11 Sonata HEV - UDDS
ChargeSustainin
g
Fuel
Con
sum
ptio
n A
B
Net Energy ChangeBattery DischargingBattery Charging
72
SAFETY AND METERING PANEL
73
FL
FH
Display
FH
Display
FL
P
R
OP
PPP
PQC
OP
H2
Controller
Vent lines to the roof
Vehicle delivery pressure sensor
Manual valves
Manual pressure regulator (removed
for Mirai testing before of pressure drop ndash pressure set
the outside panel)
Over pressure relief valve
Quick connect fitting
H2 supply
Hydrogen sensor
Flexible fuel hose to vehicle with manual valve
Over pressure sensor
Pressure gage
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
NET ENERGY CHANGE OF THE BATTERY DETERMINES IF A HEV IS CHARGE SUSTAINING OVER THE TEST CYCLE Net energy change (NEC) = net battery energy delta
over the test cycle A test is considered charge sustaining when the NEC of
battery is less then 1 of fuel energy used over the test cycle If an HEV discharges itrsquos battery over a test cycle the
fuel consumption will be increased
Sample data - 11 Sonata HEV - UDDS
ChargeSustainin
g
Fuel
Con
sum
ptio
n A
B
Net Energy ChangeBattery DischargingBattery Charging
72
SAFETY AND METERING PANEL
73
FL
FH
Display
FH
Display
FL
P
R
OP
PPP
PQC
OP
H2
Controller
Vent lines to the roof
Vehicle delivery pressure sensor
Manual valves
Manual pressure regulator (removed
for Mirai testing before of pressure drop ndash pressure set
the outside panel)
Over pressure relief valve
Quick connect fitting
H2 supply
Hydrogen sensor
Flexible fuel hose to vehicle with manual valve
Over pressure sensor
Pressure gage
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
SAFETY AND METERING PANEL
73
FL
FH
Display
FH
Display
FL
P
R
OP
PPP
PQC
OP
H2
Controller
Vent lines to the roof
Vehicle delivery pressure sensor
Manual valves
Manual pressure regulator (removed
for Mirai testing before of pressure drop ndash pressure set
the outside panel)
Over pressure relief valve
Quick connect fitting
H2 supply
Hydrogen sensor
Flexible fuel hose to vehicle with manual valve
Over pressure sensor
Pressure gage
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
BOOST CONVERTER OPERATION
74
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
RIPPLE AND PULSES IN STEADY OPERATION
75
Note Air compressor power pulses at low FC loads
H2 flow ripple correlated to drainage valve
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
61712006 SSS 72F 15 30 45 60 75 MPH
76
Check periodic air compressor power pulses (parallel h2o drainage)
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
25 GRADE 95F WITH 850 WM2
77
Max FC system power
De-rating power to 50 kW
25 grade 95F with 850Wm2
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
Argonne National Laboratory is a US Department of Energy laboratory managed by UChicago Argonne LLC
Energy Systems Division Argonne National Laboratory 9700 South Cass Avenue Bldg 362 Argonne IL 60439 wwwanlgov
