Fuel Cell System Modeling and Analysis

R. K. Ahluwalia, X. Wang, and J-K Peng

U.S. DOE Hydrogen and Fuel Cells Program 2018 Annual Merit Review and Peer Evaluation Meeting

Washington, D.C.June 13-15, 2018

This presentation does not contain any proprietary, confidential, or otherwise restricted information.

Project ID: FC017

2

Overview

Timeline Start date: Oct 2003 End date: Open Percent complete: NA

BarriersB. CostC. PerformanceE. System Thermal and Water

ManagementF. Air ManagementJ. Startup and Shut-down Time,

Energy/Transient Operation

Budget FY17 DOE Funding: $500 K Planned DOE FY18 Funding: $250 K Total DOE Project Value: $250 K

Partners/Interactions Eaton, Ford, Honeywell,

UDEL/Sonijector SA, Aalto University (Finland) 3M, Ballard, Johnson-Matthey Fuel

Cells (JMFC), UTRC, FC-PAD, GM IEA Annex 34 Transport Modeling Working Group Durability Working Group U.S. DRIVE fuel cell tech team

This project addresses system, stack and air management targets for efficiency, power density, specific power, transient response time, cold start-up time, start up and shut down energy

3

Objectives and RelevanceDevelop a validated system model and use it to assess design-point, part-load and dynamic performance of automotive (primary objective) and stationary (secondary objective) fuel cell systems (FCS) Support DOE in setting technical targets and directing component

development Establish metrics for gauging progress of R&D projects Provide data and specifications to DOE projects on high-volume manufacturing

cost estimation Impact of FY2018 work Projected 46.0* ± 0.7 $/kWe FCS cost at 500,000 units/year and 8.5 ± 0.4

kWe/gPt FCS Pt utilization with SOA d-PtCo/C cathode catalyst, reinforced 14-µm 850 EW membrane, and Q/∆T = 1.45 kW/oC constraint Verified that the SOA catalyst system can achieve 1180 ± 55 mW/cm2 stack

power density exceeding the target at low Pt loading (0.125 mg-Pt/cm2 total) Projected <5% penalty in power density if the cathode humidifier is removed,

and ~15% penalty if stack inlet pressure is lowered to 2 atm from 2.5 atm Showed that parasitic power approaches 25 kWe if the compressor discharge

pressure is raised to 4 atm Modified the reference system configuration to include valves and controls for

protected shutdown, safe startup from sub-freezing temperatures, and limiting cell voltage to 0.85-0.875 V during idle

*51 $/kWe at 100,000 units/year; Q: Stack heat load; ∆T: Stack coolant exit T – Ambient T

4

ApproachDevelop, document & make available versatile system design and analysis tools GCtool: Stand-alone code on PC platform GCtool-Autonomie: Drive-cycle analysis of hybrid fuel cell systemsValidate the models against data obtained in laboratories and test facilities inside and outside Argonne Collaborate with external organizationsApply models to issues of current interest Work with U.S. DRIVE Technical Teams Work with DOE contractors as requested by DOE

1 Evaluate the advantages of operating PEMFC stack at elevated pressures up to 4 atm using 2-stage centrifugal compressors. 12/17

2 Determine the comparative performance of PEMFC stacks with and without cathode humidifier. 03/18

3

Evaluate the performance and durability of MEAs with de-alloyed PtCo cathode catalyst on high surface area carbon with tailored pore size distribution relative to the targets of 0.44 A/mg-PGM mass activity, 1000 mW/cm2 at rated power, 300 mA/cm2 at 800 mV, and 5000 h lifetime.

06/18

4Update the performance and cost of an automotive fuel cell system with an advanced low-PGM catalyst relative to 2020 targets of 65% peak efficiency, Q/∆T of 1.45 kW/K, and $40/kW cost.

09/18

5

Technical Accomplishments: SummaryStack: Collaboration with FC-PAD and GM in obtaining data to develop and validate model for pressures up to 3 atm State-of-the-art (SOA) dispersed de-alloyed PtCo/C catalyst systems De-alloyed PtCo/C catalyst system: durability on drive cycles Air Management: Investigating integrated air management system with high speed centrifugal compressors and expanders including cooling requirements of motor and airfoil bearings (AFB) and two-stage compressors (Honeywell patent)Water Management: Optimizing cost of integrated PEFC stack and cross-flow humidifier Investigating FCS performance without

cathode humidifier (dispersed catalyst electrodes)

Fuel Management: Evaluating the performance of anode system with a pulse injector in lieu of H2recirculation blower (collaboration with Ford & UDEL)Thermal Management: Optimizing system performance and cost subject to Q/∆T constraintSystem startup and shutdown: Modified reference system to incorporate controls for protected shutdown, safe startup from sub-freeze temperatures, and limiting cell voltage at idle

Argonne 2018 FCS Configuration

∆T: Stack coolant exit T – Ambient T

6

Model Framework and Single Cell HardwareUS-EU Differential Cell Hardware* P: 1-3 atm; T: 45-95oC, RH: 30-150%; X(O2): 1-21% Random Tests: Semi-statistical

randomized with multiple variables, forward scans, 3-min hold at each cell voltage

Controlled Tests: Model guided single-variable with some two-variable tests, forward scans, 3-min hold at each cell voltage

Cathode AnodeCatalyst d-Pt3Co/C Pt/CCatalyst Support HSAC VulcanIonomer Equivalent Weight 950 950Pt Loading 0.1 mg/cm2 0.025 mg/cm2

ECSA 45 m2/g 60 m2/gElectrode Thickness 7 µm 5 µmDiffusion Medium Thickness 200 µm 200 µmMembrane 18 µm Reinforced

Representative State-of-the-Art Low-PGM MEA

*All tests conducted at GM (S. Arisetty Lead), FCPAD-FC156 collaboration (S. Kumaraguru PI)

4. Expanded Polarization Data

OperatingConditions

Cell Design

, , , ,2. Overpotential Breakdown

iL(P, T, RH, XO2)ηm(P, T, RH, XO2, i/iL)

3. ηm Correlation

Rm (P, T, RH, XO2, E, i)5. Mass Transfer Resistance

1+1D or 2+1D7. Integral Cell Model

(T, RH)CCL Conductivity

Θ(E)PtOx Formation

Rcf (T, RH, E, i)CCL Resistance

Rg(P,T,RH,XO2)Gas Resistance

, (E, i), δl/δd

GDL Resistance Rd: Pressure DependentRcf: Pressure Independent

6. Resistance Breakdown

FC-PAD

GCtool Model

7

Kinetics of ORR on d-PtCo/C CatalystElectrode Resistance

Electrode (σ𝑐𝑐) / membrane conductivities (σ𝑚𝑚) from Galvanostatic impedance data in H2/N2

Distributed ORR Kinetic Model For Tafel kinetics, the ORR and CCL

Ohmic overpotentials are separableη𝑐𝑐 = η𝑠𝑠𝑐𝑐 + 𝑖𝑖𝑅𝑅Ω𝑐𝑐 ( 𝑖𝑖𝛿𝛿𝑐𝑐

𝑏𝑏𝜎𝜎𝑐𝑐)

𝑖𝑖 + 𝑖𝑖𝑥𝑥 = 𝑖𝑖0𝑆𝑆𝑃𝑃𝑃𝑃 1 − θ 𝑒𝑒−ωθ𝑅𝑅𝑅𝑅𝑒𝑒

α𝑛𝑛𝑛𝑛𝑅𝑅𝑅𝑅 η𝑠𝑠

𝑐𝑐

𝑖𝑖0 = 𝑖𝑖0𝑟𝑟𝑒𝑒−∆𝐻𝐻𝑠𝑠

𝑐𝑐

𝑅𝑅1𝑅𝑅−

1𝑅𝑅𝑟𝑟 𝑃𝑃𝑂𝑂2

γ Φβ

0

20

40

60

80

100

20 40 60 80 100

Con

duct

ivity

, mS.

cm-1

Relative Humidity, %

Electrode

Membrane

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 0.5 1.0 1.5 2.0

Imag

inar

y -Z

", Ω

.cm

2

Real Z', Ω.cm2

122% 95% 75% 50% 35% 20%

T:80oC, P:1.5 atmH2/N2

RH

PtOx Formation on d-PtCo/C Catalyst Transient solid solution model from measured

CV reduction charge after hold for 12 s to 2 h at constant potential1

0.0

0.2

0.4

0.6

0.8

1.0

0.6 0.7 0.8 0.9 1.0

Oxi

de C

over

age

Potential, V

80⁰C

Φ100%

35%

50%

θPtOH

θPtO

0.0

0.2

0.4

0.6

0.8

1.0

0.6 0.8 1.0

50⁰C

θPtOH

θPtO

[1] Arisetty et. al. (2015), ECS Transactions, 69, 273-289

Kinetic data from controlled tests

8

Calibration of ORR Kinetic Model*

𝑖𝑖0 calibration: 1 data point per pol curve

Data: 𝑖𝑖 + 𝑖𝑖𝑥𝑥 = 𝑖𝑖0𝑆𝑆𝑃𝑃𝑃𝑃 1 − θ 𝑒𝑒−ωθ𝑅𝑅𝑅𝑅𝑒𝑒

α𝑛𝑛𝑛𝑛𝑅𝑅𝑅𝑅 η𝑠𝑠

𝑐𝑐

Model: 𝑖𝑖0 = 𝑖𝑖0𝑟𝑟𝑒𝑒−∆𝐻𝐻𝑠𝑠

𝑐𝑐

𝑅𝑅1𝑅𝑅−

1𝑅𝑅𝑟𝑟 𝑃𝑃𝑂𝑂2

γ Φβ

η𝑠𝑠𝑐𝑐 calibration: all data in kinetic region Data: η𝑠𝑠𝑐𝑐 = 𝐸𝐸𝑁𝑁 − 𝐸𝐸 − 𝑖𝑖𝑅𝑅Ω𝑚𝑚 − 𝑖𝑖𝑅𝑅Ω𝑐𝑐

Model: η𝑠𝑠𝑐𝑐 = 𝑏𝑏[ln 𝑖𝑖+𝑖𝑖𝑥𝑥𝑖𝑖0𝑆𝑆𝑃𝑃𝑃𝑃

− ln 1 − θ + 𝜔𝜔𝜔𝜔𝑅𝑅𝑅𝑅

]

0

20

40

60

80

0 20 40 60 80

33.7 kJ/molγ: 0.647β: 0.9ω: 3 kJ/mol

i0 (Data),

i 0(M

odel

),

Open symbols: Random tests (BOT); Closed symbols: Controlled tests (EOT: end of test)

*Kinetic model based on random test data; BOT: Beginning of tests; EOT: End of tests

250

300

350

400

250 300 350 400

38.8 kJ/molγ: 0.629β: 0.524ω: 3 kJ/mol

(Data), mV

(Mod

el),

mV

9

Mass Activity for ORR on d-PtCo/C Catalysts d-PtCo/C has 2X modeled mass activity of a-Pt/C that has nearly the same

particle size d-PtCo/C and d-PtNi/C alloy have comparable mass activities Both low-PGM alloy catalysts (d-PtNi/C and d-PtCo/C) meet the mass

activity targets of 440 A/gPt

745

659

760

432

312

500

180

330

380

651

763

623

661

456

319

650

190

392334

0

200

400

600

800

1000

Mas

s Ac

tivity

, A/g

Pt

Project: FC156 FC-PAD FC106 FC104

d-Pt

Ni/C

a-Pt

/C

Pt/C

Pt60

Co 4

0/C

Pt70

Co 3

0/C

Pt85

Co 1

5/C

Pt68

(CoM

n)32

/NST

F

d-Pt

3Ni 7/

NST

F

d-Pt

3Ni 7/

NST

F +

CI

Modeled Mass Activity

d-Pt

3Co/

C

Solid Fill: Modeled mass activity from ORR kinetic modelPattern Fill: Measured mass activity in H2/O2

10

Oxygen Mass Transfer: Limiting Current DensityDetermined limiting current density (iL) and correlated mass transfer overpotential (ηm) with reduced current density (i/iL) η𝑚𝑚 = 𝐸𝐸𝑁𝑁 − 𝐸𝐸 − 𝑖𝑖𝑅𝑅Ω𝑚𝑚 − η𝑐𝑐 − η𝑎𝑎 𝑖𝑖𝐿𝐿 defined as current density at which η𝑚𝑚

equals 400 mV

Model Variables 𝑃𝑃: Pressure 𝑃𝑃𝑂𝑂2: O2 partial pressure in gas

channel Tc : CCL temperature, function of

T, 𝐸𝐸𝑁𝑁, 𝐸𝐸, and 𝑖𝑖 Φc : CCL RH, function of Φ, 𝑖𝑖, and

water transport across membraneSymbolsT: Bipolar plate temperature; Tc: CCL temperatureΦ: Gas channel RH; Φc: CCL RH

11

Limiting Current Density: Effect of PressureAt constant P(O2), 𝑖𝑖𝐿𝐿 decreases with increase in P because of the inverse dependence of O2 gas phase diffusivity on P. The decrease in 𝑖𝑖𝐿𝐿 is less than

proportional to 1/P, implying that non-Fickian diffusion controls mass transport resistance

At constant X(O2), 𝑖𝑖𝐿𝐿 increases with increase in P because of higher P(O2). The increase in 𝑖𝑖𝐿𝐿 is somewhat less

than proportional to P(O2), implying that mass transport resistance also increases with P(O2).

The plots include the experimental data for iL at different P, X(O2), T=80oC, and Φ=100%. For comparison, the model was used to rescale the iL data to P(O2)=0.15 atm or X(O2)=10%, Tc=80oC, and 𝜙𝜙𝑐𝑐=100%.

12

Limiting Current Density: Effect of Temperature and RHDependence of 𝑖𝑖𝐿𝐿 on Tc >> 𝑇𝑇 ⁄3 2, confirming that processes other than Fickian diffusion are rate controlling. O2 permeability through the ionomer

film on the catalyst particles

The plot includes all the experimental data for 𝒊𝒊𝑳𝑳at different T, P=1.6 atm, X(O2)=10%, and Φ=100%. For comparison, the model was used to rescale the 𝒊𝒊𝑳𝑳 data to P(O2)=0.12 atm, and 𝜙𝜙𝑐𝑐=100%.

For given Tc, 𝑖𝑖𝐿𝐿 is highest at an intermediate RH in CCL (𝜙𝜙𝑐𝑐∗). CCL flooding for 𝜙𝜙𝑐𝑐 > 𝜙𝜙𝑐𝑐∗

𝜙𝜙𝑐𝑐∗ Increases at higher Tc

The plot includes all the experimental data for iL at different Φ, T=45, 70 and 95oC, P=1.6 atm and X(O2)=10%. For comparison, the model was used to rescale the iL data to P(O2)=0.12 atm and different Tc=45, 70 and 95oC.

13

Performance of Automotive FCS with SOA d-PtCo/C Cathode CatalystModeled optimal beginning of life (BOL) performance of automotive FCS subject to Q/∆T=1.45 kW/oC constraint: 0.125 mg/cm2 total Pt loading; 850 EW, 14-µm chemically-stabilized, reinforced membraneFY2018 model based on differential cell data supports FY2017 landmark result 46.0 ± 0.7 $/kWe projected cost* at 2.5 atm stack inlet pressure and 95oC stack

coolant outlet temperature for high volume manufacturing Removing membrane humidifier slightly lowers the system cost at 2.5 atm stack

inlet pressure

*Using 2018 cost correlations from Strategic Analysis (SA), 500,000 units/year, no H2 blower. Includes $2.01 cost increase in 2018 for manufacturing bipolar plates and MEAs, added controls, and CEMM price inflation.

Results are preliminary pending model validation using data from 50-cm2 integral cell Error bars reflect variance of

kinetic data in random and controlled tests and include degradation between the two series of tests

Effect of manufacturing volume on projected cost (SA) 500,000 units/year: 46 $/kWe

100,000 units/year: 51 $/kWe

10,000 units/year: 88 $/kWe

14

FCS with SOA d-PtCo/C Cathode Catalyst: Pt UtilizationModeled optimal beginning of life (BOL) performance of automotive FCS subject to Q/∆T=1.45 kW/oC constraint: 0.125 mg/cm2 total Pt loading; 850 EW, 14-µm chemically-stabilized, reinforced membraneFY2018 model based on differential cell data supports FY2017 landmark result Modeled stack Pt utilization (9.5 ± 0.5 kWe/gPt) exceeds the target (8.0 kWe/gPt) Modeled FCS Pt utilization (8.5 ± 0.4 kWe/gPt) also exceeds the target Stack inlet pressure ≥ 2.0 atm needed to meet the Pt utilization target

15

FCS with SOA d-PtCo/C Cathode Catalyst: Power DensityModeled optimal beginning of life (BOL) performance of automotive FCS subject to Q/∆T=1.45 kW/oC constraint: 0.125 mg/cm2 total Pt loading; 850 EW, 14-µm chemically-stabilized, reinforced membraneFY2018 model based on differential cell data supports FY2017 landmark result Modeled gross stack power density (1180 ± 55 mW/cm2) exceeds the target

(1000 mW/cm2) at low Pt loading (0.125 mg-Pt/cm2 total) Stack inlet pressure ≥ 2.0 atm needed to meet the power density target

16

Air Management SystemModeled CEMM with Bleed Air Recovery

Compressor ExpanderMotor

Filter SpentAirAir

LT Coolant

Controller

AFB

MH

From FC

To FC

Simulation results: 73 g/s air flow rate, and 2.5 atm compressor discharge P; 2.3 atm expander inlet P, 85oC inlet T, fully saturated; 40oC ambient T

CEMM w/o Bleed Air Recovery

Compressor Expander

MotorFilter Spent

AirAir

LT Coolant

Controller

AFB

MHPrecooler

From FC

To FC

CoolingAir

AFB Cooling Valve

0

2

4

6

8

10

0 20 40 60 80 100

CEM

Coo

ling

Air,

g/s

Air to FC, g/s

0

30

60

90

120

1.0

1.5

2.0

2.5

3.0

0 50 100

N, k

rpm

P, a

tm

Air to FC, g/s

Discharge Pressure

ShaftSpeed

Projected cooling air loss using Honeywell data for CEMM

*AFB: Air foil bearing

17

Performance of Two-Stage Centrifugal CompressorMixed axial and radial flow compressors on a common shaft with AFBs and 3-phase brushless DC motor; Honeywell US Patent 2015/0308456 AFB*/motor cooling air extracted

downstream of the pre-cooler and is not recovered

Compressor power ~25 kWe at 4-atm discharge pressure

Results for existing compressor but resized motor and motor controller

0

5

10

15

20

25

30

0

1

2

3

4

5

0 25 50 75 100 125

Com

pres

sor P

ower

, kW

Disc

harg

e Pr

essu

re, a

tm

Dry Air Flow Rate, g/s

Ambient T: 40°CAmbient RH: 50%

Solid: PressureDashed: Power

100k

90k

80k

70k

60k50k

40k

30k

100k

90k

80k

70k

60k50k

40k30k 50

60

70

80

90

100

0

10

20

30

1.5 2.0 2.5 3.0 3.5 4.0

Effic

ienc

y, %

Com

pres

sor P

ower

, kW

e

Discharge pressure, atm

Ambient: 40°C (T), 50% (RH)Air for AFB cooling: 8.1 g/sAir to FC: 73 g/s

CompressorStage 1

Stage 2

Motor Controller

Motor

Efficiency

Power

CompressorStage 1

Motor

FilterAir

LT Coolant

Controller

AFB

MHFrom FC

CompressorStage 2

To FCAir Cooler

AFB Cooling Valve

*AFB: Air foil bearing

18

Proposed Argonne 2018 FCS Configuration with Controls Stack idling: Limit cell voltage by operating at low SR(c) and low O2 concentration Stack shutdown: Avoid H2-air front during subsequent start-up by depleting oxygen Sub-freeze start: Prevent icing by maximizing in-stack heat production*

*M. Toida, Y. Naganuma, and T. Ogawa, “Fuel Cell System and Method for Discharging the System,” US 2016/0141672 A1, May 19, 2016.

19

Mitigation of High Cell Voltages

0.5

0.6

0.7

0.8

0.9

0.0 0.5 1.0 1.5 2.0

Cell

Volta

ge, V

Current Density, A.cm-2

0 to 50 kW MapFC Mapping HLB5xWOT and 6.3 kW25% GradeCold StartSSS HEV @ 1-6% Grade

SOA stack limits cell voltage to 0.85 V during idle

Reducing cathode stoichiometry may not provide adequate control of cell voltage

Lowering stack temperature may also be inadequate and has thermal inertia

Reducing O2 concentration by recycling spent air provides most control

0.75

0.80

0.85

0.90

0.95

0.00 0.05 0.10 0.15 0.20

Cell

Volta

ge, V

Current Density, A.cm-2

T: 60⁰CX(O2): 21%

SR(c): 5

1.05

1.5

48 mA.cm-2

3.0 kWe

0.75

0.80

0.85

0.90

0.95

0.00 0.05 0.10 0.15 0.20

Cell

Volta

ge, V

Current Density, A.cm-2

SR(c): 1.5X(O2): 21%

T: 60⁰C80⁰C70⁰C

45⁰C

43 mA.cm-2

2.8 kWe

0.75

0.80

0.85

0.90

0.95

0.00 0.05 0.10 0.15 0.20

Cel

l Vol

tage

, V

Current Density, A.cm-2

T: 60⁰CSR(c): 1.5

Xi(O2): 21%

10%5% 15%

21 mA.cm-2

1.4 kWe

M. Sato, “Fuel cell System and method of Controlling the Same,” US 2017/0047602 A1, Feb. 16, 2017.S. Kawahara, K. Suematsu, M. Toida, and R. Akaboshi, “Fuel Cell System and Control Method Thereof,” US 9,531,022 B2, Dec. 27, 2016.

20

Controlling O2 Concentration at Stack InletControlling O2 concentration at stack inlet by recycling cathode spent air and bypassing stack*

Recycle and bypass depend on desired reduction in O2 concentration at stack inlet and SR(c)

Increase in compressor flow rate to reduce inlet O2 concentration below 21%

0

20

40

60

80

100

5 10 15 20

Rec

ycle

/ B

ypas

s, %

O2 Mole Concentration at Stack Inlet, %

SR(c): 1.5

Recycle

Bypass

SR(c): 2.5

0

5

10

15

5 10 15 20

Com

pres

sor F

low

Rat

e M

ultip

lier

O2 Mole Concentration at Stack Inlet, %

SR(c): 1.5

2.5

Model for mixed cathode potential at low cathode stoichiometry

Anode Spent

Air

ExpanderMotorCompressor

c

Air Recycle Valve

StackBypassValve

Air Inlet

Valve

TP

H2 H+ e-

e-

𝑅𝑅 = 𝑚𝑅𝑅/𝑚𝑐𝑐

𝐵𝐵 = 𝑚𝐵𝐵/𝑚𝑐𝑐

*M. P. Wilson, V. Yadha and C. A. Reiser, “Low Power Control of Fuel Cell Open Circuit Voltage,” US 8,808,934 B2, Aug. 19, 2014.

Air inlet and recycle valves not included in Argonne 2018 FCS configuration

0.5

0.6

0.7

0.8

0.9

0.001 0.01 0.1 1 10 100

Pres

sure

, atm

Time, ks

Initial P: 0.85(a)/0.68(c) atmInitial Gas: H2(a)/N2(c)T: 25⁰C, RH: 100%Air Intake Valve Leakage: 2ml/min/atm

Cathode

Anode

21

Protected ShutdownProtected shutdown algorithm

Step 1: Isolate the cathode circuit by closing the air intake valve and opening the air recycle valve

Step 2: Deplete oxygen by applying load while supplying hydrogen

Step 3: Shut the compressor (if on) and isolate H2 circuit by closing the H2 main valve

Step 4: Allow H2 and N2 in anode and cathode circuits to equilibrate

Step 2: Depletion of oxygen with cathode air recycled by compressor

Step 4: Equilibration of anode and cathode pressures

Step 4: Equilibration of H2 partial pressures in anode and cathode circuits

0.4

0.5

0.6

0.7

0.8

0.9

0.000.050.100.150.20

Cel

l Vol

tage

, V

Oxygen Partial Pressure, atm

Current Density: 50 mA/cm2

T: 60⁰C

0.0

0.2

0.4

0.6

0.8

1.0

0.001 0.01 0.1 1 10 100

H2

Part

ial P

ress

ure,

atm

Time, ks

Initial P: 0.85(a)/0.68(c) atmInitial Gas: H2(a)/N2(c)T: 25⁰C, RH: 100%Air Intake Valve Leakage:

2ml/min/atm

Cathode

Anode

S. U Kwon, S. H. Choi, N. W. Lee, S. P. Ryu, and S. S. Park, “Idle Stop-Start Control Method of Fuel Cell Hybrid Vehicle,” US 2010/0009219 A1, Jan. 14, 2010.

22

Collaborations

Argonne develops the fuel cell system configuration, determines performance, identifies and sizes components, and provides this information to SA for high-volume manufacturing cost estimation

Honeywell: Cost and Performance Enhancements for a PEM Fuel Cell Turbocompressor (FC27) Eaton: Roots Air Management System with Integrated Expander (FC103)

Stack 3M: High Performance, Durable, Low Cost Membrane Electrode Assemblies for Transportation (FC104)Ballard/Eaton: Roots Air Management System with Integrated Expander (FC103)JMFC and UTRC: Rationally Designed Catalyst Layers for PEMFC Performance Optimization (FC106)FC-PAD: Fuel Cell Performance and Durability Consortium (FC135, FC136, FC137, FC138, FC139)GM: Highly-Accessible Catalysts for Durable High-Power Performance (FC144)GM: Durable High-Power Membrane Electrode Assemblies with Low Pt Loadings (FC156)

Water Management Gore, Ford, dPoint: Materials and Modules for Low-Cost, High-Performance Fuel Cell Humidifiers (FC067)

Thermal Management 3M, Honeywell Thermal SystemsFuel Management 3M, University of Delaware (Sonijector)Fuel Economy ANL-Autonomie (SA044), Aalto University (Fuel Cell Buses)H2 Impurities 3M

System Cost SA: Manufacturing Cost Analysis of Fuel Cell Systems and Transportation Fuel Cell System Cost Assessment (FC163)

Dissemination IEA Annex 34, Transport Modeling Working Group, Durability Working Group, Catalysis Working Group

Air Management

23

Proposed Future Work1.Continue to support DOE development effort at system, component, and

phenomenological levels2.Continue to support SA in high-volume manufacturing cost projections, collaborate in life-

cycle cost studies Optimize system parameters considering costs at low-volume manufacturing Life cycle cost study for medium and heavy duty vehicles (Ballard, Eaton, SA)3.Alternate MEAs with advanced alloy catalysts State-of-the-art low PGM Pt and Pt alloys (FC-PAD collaboration) Alternate electrode structures (FC-PAD FOA projects collaboration) Durability models (FC-PAD and GM collaboration)4.System architecture and balance-of-plant components Air management system with centrifugal and Roots compressors and expanders

(Honeywell/Eaton collaboration) Fuel and water management systems: anode gas recirculation, internal/external

humidification Bipolar plates and flow fields for low pressure drops and uniform air/fuel distribution, cell

to stack performance differentials Strategies and controls for stack idling, startup and shutdown, subfreezing temperatures5.Incorporate durability considerations in system analysis System optimization for cost, performance, and durability on drive cycles (Advanced alloy

catalyst systems)Any proposed future work is subject to change based on funding levels.

2424

Project SummaryRelevance: Independent analysis to assess design-point, part-load and

dynamic performance of automotive and stationary FCSApproach: Develop and validate versatile system design and analysis tools

Apply models to issues of current interestCollaborate with other organizations to obtain data and apply models

Progress: Projected 46.0 ± 0.7 $/kWe FCS cost at high volume manufacturing and 8.5 ± 0.4 kWe/gPt FCS Pt utilization with SOA d-PtCo/C cathode catalyst, reinforced 14-µm 850 EW membrane, and Q/∆T = 1.45 kW/oC constraintVerified that the stack can achieve 1180 ± 55 mW/cm2 power density exceeding the target at low Pt loading (0.125 mg-Pt/cm2)Projected <5% penalty in power density if the cathode humidifier is removed, and ~15% penalty if stack inlet pressure is lowered to 2 atm from 2.5 atm Showed that parasitic power approaches 25 kWe if the compressor discharge pressure is raised to 4 atmModified the reference system configuration to include valves and controls for protected shutdown, safe startup from sub-freezing temperatures, and limiting cell voltage to 0.85 V during idle

Collaborations: 3M, Eaton, GM, Gore, JMFC, SA, UTRC, UDEL/Sonijector

Future Work: Fuel cell systems with emerging high activity catalystsAlternate balance-of-plant components System analysis with durability considerations on drive cycles

25

Sample comments and feedback ANL employs sound approach in forecasting performance but need more validation Impressive range of accomplishments, excellent study of catalyst systems, but needs

more focus on Ni loss and other degradation mechanisms Excellent collaborations, but needs better definition of engagement with FC-PAD. Closer collaboration with OEMs to obtain stack and system data The project addresses multiple barriers including cost, performance, thermal/water/air

subsystems, but more validation needed. Future work appears to be appropriate. More emphasis on durability in future.Work scope consistent with above recommendations√ On-going work on differential cell data for PtCo/C dispersed catalysts in collaboration

with FC-PAD and an industrial partner. Initial results on performance and durability are included in FC-PAD presentations. The PI is FC-PAD coordinator for modeling and validation thrust area.

√ Examining durability during idling, startup/shutdown and sub-freeze start. Added valves and controls to mitigate catalyst degradation.

√ Analyzing system simplification (compressor air bleed), component elimination (humidifier, H2 recirculation blower) and alternative operating conditions (higher pressures)

√ Expanded collaboration with an OEM on model validation using 10-50 cm2 cells. Analyzing data from a SOA automotive stack and fuel cell system.

√ ANL is a subcontractor to SA on FC-163 project, responsible for supplying performance and design data. Plans and recent results are discussed in bi-weekly calls.

Reviewers’ Comments

26

Technical Back-Up Slides

27

Electrode ResistanceElectrode (σ𝑐𝑐) and membrane conductivities (σ𝑚𝑚) from Galvanostatic impedance data in H2/N2 at 0.4 V with 5 mV perturbation ZVIEW transmission line model (100

repeat units)

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 0.5 1.0 1.5 2.0

Imag

inar

y -Z

", Ω

.cm

2

Real Z', Ω.cm2

122% 95% 75% 50% 35% 20%

T:80oC, P:1.5 atmH2/N2

RH

Nyquist plot consistent with RC circuit

Electrode and membrane conductivities

Effective electrode resistance

0

20

40

60

80

100

20 40 60 80 100

Con

duct

ivity

, mS.

cm-1

Relative Humidity, %

Electrode

Membrane

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0Current Density, A.cm-2

100%

75%

50%

30%

Relative Humidity

28

Oxide Formation on d-PtCo/C Catalyst Solid solution model for PtOx formation

developed from measured CV reduction charge after 45-min hold at constant potential1

𝑃𝑃𝑃𝑃 + 𝐻𝐻2𝑂𝑂 = 𝑃𝑃𝑃𝑃𝑂𝑂𝐻𝐻 + 𝐻𝐻+ + 𝑒𝑒−

𝑃𝑃𝑃𝑃𝑂𝑂𝐻𝐻 = 𝑃𝑃𝑃𝑃𝑂𝑂 + 𝐻𝐻+ + 𝑒𝑒−

θ = θ𝑃𝑃𝑃𝑃𝑂𝑂𝑃𝑃 + θ𝑃𝑃𝑃𝑃𝑂𝑂

𝐾𝐾𝑖𝑖 = 𝑐𝑐𝑃𝑃+𝜃𝜃𝑖𝑖𝜃𝜃𝑖𝑖+1

𝑒𝑒−𝐹𝐹𝑅𝑅𝑅𝑅 𝐸𝐸−𝐸𝐸0𝑖𝑖 = 𝐾𝐾𝑖𝑖0𝑒𝑒

−𝜔𝜔𝑖𝑖𝑅𝑅𝑅𝑅𝜔𝜔𝑖𝑖

𝑥𝑥𝑖𝑖

[1] Arisetty et. al. (2015), ECS Transactions, 69, 273-289

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.6 0.7 0.8 0.9 1.0

Red

uctio

n C

harg

e, Q

/ΓS P

t

Potential, V

100% 50% 35%

80⁰CΦ

0.00.20.40.60.81.01.2

0.6 0.8 1.0

50⁰C

0.0

0.2

0.4

0.6

0.8

1.0

0.6 0.7 0.8 0.9 1.0

Oxi

de C

over

age

(θPt

OH

+ θ

PtO)

Potential, V

80⁰C

0.00.20.40.60.81.0

0.6 0.8 1.0

50⁰CΦ

100%

35%

50%

0.0

0.2

0.4

0.6

0.8

1.0

0.6 0.7 0.8 0.9 1.0

Oxi

de C

over

age

Potential, V

80⁰C

Φ100%

35%

50%

θPtOH

θPtO

0.0

0.2

0.4

0.6

0.8

1.0

0.6 0.8 1.0

50⁰C

θPtOH

θPtO

29

Model ValidationWork in Progress Validation tests on 50-cm2 integral cell with controlled SR(c) and SR(a) Differential cell tests with 0.05 and 0.2 mg/cm2 Pt loadings in cathode Extracting resistances for O2 transport in GDL, CCL pores and ionomerFuture Work Catalyst accelerated cell tests to model durability

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Cel

l Vol

tage

, V

Current Density, A.cm-2

3.2 atm35% RH

10 cm2 Active AreaAnode: 3.5 slpm H2SR(c): 1.5T: 95°C

2.7 atm 51% RH

2.2 atm41% RH

1.7 atm35% RH

0.0

0.2

0.4

0.6

0.8

1.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Cel

l Vol

tage

, V

Current Density, A.cm-2

SR=4.550% RH

SR=3.555% RH

10 cm2 active AreaAnode Flow: 3.5 slpmP: 2.7 atmT: 95°C

SR=1.551% RH

Differential cell operated in integral mode: Constant SR(c), variable P

Differential cell operated in integral mode: Variable SR(c), constant P

30

Projected Performance of CEMM with Bleed Air RecoveryAssumption: CEM cooling air can be recovered and combined with humidified air upstream of the PEFC stack Data are for as-built components.

Efficiencies and performance can be improved by resizing the components to match the actual operating conditions

Compressor ExpanderMotor

Filter SpentAirAir

LT Coolant

Controller

AFB

MH

From FC

To FC

64

73

88

87

0 20 40 60 80 100

Compressor

Expander

Motor

MotorController

Efficiency, %0 2 4 6 8 10 12

Compressor

Expander

Shaft

Motor

MotorController

Power

11.0 kW

4.2 kW

6.8 kW

7.6 kWe

8.8 kWe

Simulation results for 73 g/s air flow rate and 2.5 atm compressor discharge P; 2.3 atm expander inlet P, 85oC inlet T, fully saturated; 40oC ambient T

31

Comparative Performance of CEMM w/o Bleed Air RecoveryA control valve used to split pre-determined amount of air downstream of the pre-cooler to cool AFB and motor; this cooling air is lost irrecoverably. Data are for as-built components.

With resizing, the parasitic power can be reduced to 7.0 kWe with bleed air recovery and 8.1 kWe w/o bleed air recovery.

Simulation results for 73 g/s air flow rate and 2.5 atm compressor discharge P; 2.3 atm expander inlet P, 85oC inlet T, fully saturated; 40oC ambient T

11.8

4.5

7.4

8.3

9.5

0 2 4 6 8 10 12 14

Compressor

Expander

Shaft

Motor

MotorController

Power

11.0 kW

4.5 kW

6.5 kW

7.4 kWe

8.5 kWe

64

71

88

87

68

72

89

88

0 20 40 60 80 100

Compressor

Expander

Motor

MotorController

Efficiency, %

Open Bars: As BuiltShaded Bars: Resized

Compressor Expander

MotorFilter Spent

AirAir

LT Coolant

Controller

AFB

MHPrecooler

From FC

To FC

CoolingAir

AFB Cooling Valve