Eliminating EV Range Anxiety, Predicting Personalized Fuel ... · PDF file19/11/2014 · Eliminating EV Range Anxiety, Predicting Personalized Fuel Economy, ... EV range anxiety, Predicting

Eliminating EV Range Anxiety,Predicting Personalized Fuel Economy, and Enabling Vehicle-Grid Integration

Samveg Saxena, Ph.D

November 19, 2014

v2gsim.lbl.gov | powertrains.lbl.gov

MyGreenCarV2G‐Sim

Samveg Saxenav2gsim.lbl.gov

November 19, 2014 2Eliminating EV range anxiety,

Predicting personalized fuel economy,Enabling vehicle-grid integration

V2G-Sim | MyGreenCar

Charging Ahead onClean Transportation

Vehicles account for ~¼ of global GHG emissions Substantial urban pollutants from tailpipe emissions

Transportation electrification is needed soon, at large scale to meet climate goals, but many obstacles: Range anxiety, available charging infrastructure, cost, etc.

If not electrification, at least needmarked increases in fuel economy

MyGreenCar





V2G-Sim: Understanding how cleantransport enables a clean grid

If ¼ US vehicles were PHEVs= Grid storage for >1 hour of full US grid operations (~1,000 GWh) Grid storage becomes a side-effect of deploying clean transportation Enables substantial renewables integration… Clean transportation enables clean grid Many coupled variables, each with large uncertainties…

VehiclePowertrain

Models

ChargingModels

Automateddrive-cycleGenerationAutomated

TravelPattern

Generationfor each PEV

Responseto managed

charging/V2G

User inputs(fleet usagestatistics,i.e. NHTS)

Aggregatebi-directionalcharging of

each vehicle

Grid-scaleimpacts &

opportunitiesfrom PEVs

DetailedBattery Models

BatteryDegradation

Models

1

2

3a

3b

4a

4b

5a

5b

6

V2G-Sim models the driving and charging behavior of individual PEVs to generate

temporal and spatial grid-scale impact/opportunity predictions

Individual PEV driving/charging/V2G profile

PEV 1 PEV 2 PEV N

Grid‐scale impacts

V2G‐Sim





Vehicle Powertrains &Grid-Integration Research at LBNL

Grid Integration

Sepa

rato

rCathodeAnode

ee

e

Electrolyte

Electrochemical Technologies

Electricity Markets and Policy

Vehicle Powertrain Systems

Vehicle electrification &Vehicle-grid integration Advanced

EngineResearch

PowertrainInnovations

for DevelopingCountries

LDRD: seed funding approved by Congress

for high risk, high reward ideas that can

transform society

V2G-Sim development funded by laboratory directed research & development (LDRD)





Methodology: Travel Itinerary Inputs

1) Statistical approach: Categorize vehicleusage into finite number of “bins” Describe how/when vehicles are used

2) Deterministic approach: Direct simulation with actual vehicle usage data E.g. National Household Travel Survey / Real-time data

2.5) Automatically deriving inputs statisticsfrom actual vehicle usage data





Methodology: Travel Itinerary Inputs

0 5 10 15 201

2

3

Time (hours)

Veh

icle

sta

te

Activity profile of vehicle 48621, 3=plugged in, 2=parked, 1=driving

0 5 10 15 201

2

3

Time (hours)

Veh

icle

sta

te

Activity profile of vehicle 97199, 3=plugged in, 2=parked, 1=driving

0 5 10 15 201

2

3

Time (hours)

Veh

icle

sta

te

Activity profile of vehicle 92615, 3=plugged in, 2=parked, 1=driving Generatetrip-specificdrive cycles





Methodology: Vehicle Powertrain Models

Methods: several possible approaches Simple: Wh/km for each vehicle Detailed: Physics-based vehicle

powertrain models

Output: Vehicle energy use, SOC while driving

PowertrainController

Propulsion ControllerBrake Controller

PowertrainController

Propulsion ControllerBrake Controller

Example:EV model

Example:Powersplit PHEV

Source:ANL, Autonomie





Methodology: Charging Models &Managed Charging/Discharging Controllers

Methods: charging rate with exponential decay: Simple model: Pcharge(t) = f ( charger type , SOC , vehicle )

Managed charging model: considers simple model +signal from managed bi-directional charging algorithms

Output: charging rate & SOC while charging

Example: ChargePoint CT503, Level 2 charger

Source: INL





Methodology: Battery Models

Electrochemical Models

Equivalent Circuit Models

, int,out chg OC coul out chgV V I R

, int,out dis OC out disV V I R

int, ,dis cellR f SOC T

int, ,chg cellR f SOC T

,OC cellV f SOC T

max

max

3600 initIAh Ah

SOCAh

2, int, 1gen chg out chg out out coulQ I R V I

2, int,gen dis out disQ I R

module air module

Thermal resistancecoolingT TQ

module

module p,module

gen coolingcell

Q Q dtT T

m C

Prof. Scott Moura, UC Berkeley





Applications: Cross-disciplinary research

CarbonAnode

Li+

Li+

Li+

SEI – Solid Electrolyte Interphase

SEI Growth

Electrolyte

Li+

Intercalatedlithium ions

Lithium bondedwith solventmolecules

SEI Cracking,leads to new SEI formation

V2GResourceBidding=f(x1,..,xn)

x4=f(LSEI,..)

RegulationMarketImpact





V2G-Sim Demonstration Cases

1. Predicting PEV charging demand & uncertainty

2. Spatial resolution of grid impacts

3. Battery degradation from driving & grid services

4. PEVs for renewables integration

5. EVs for demand response

6. Adequacy of inexpensive charging infrastructure

7. Redefining the Useful Lifetime of EV Batteries





Case 1: Forecasting PEV ChargingDemand from Stochastic Inputs

0 5 10 15 200

200

400

600

800

1000

1200

Time (hours after 12 am)

Tota

l grid

dem

and

from

PEV

cha

rgin

g (k

W)

Grid power demand from charging of 1000 vehicles

0 5 10 15 200

200

400

600

800

1000

1200

Time (hours after 12 am)

Tota

l grid

dem

and

from

PEV

cha

rgin

g (k

W)

V2G-Sim allows any number of differentvehicle or charger types to be considered

Case 1a 1,000 EVs Mix of L1 & L2 charging Favoring evening charging

Case 1b 1,000 EVs Mix of L1 & L2 charging Day and evening charging

V2G-Sim forecasts PEV loadssecond-by-second

with uncertainty estimates(enabler for PEV aggregation)





Case 2: Spatially resolvingPEV loads and resources

0 5 10 15 200

50

100

150

200

250

300

350

Time (hours)

Tota

l PEV

cha

rgin

g lo

ad a

t eac

h lo

catio

n (k

W) Charging load at each location

Workplace chargingHome charging

Case 2a 659 vehicles, SF Bay

data from NHTS L1 at home, L2 at work

Case 2b 80% PEV adoption

uniformly across U.S. L1 at home, L1 at work

Alternatively, can spatially-resolve down todistribution level, neighborhood, GPS coordinate, etc.





0 1 2 3 4 5 6 7 8 9 1075

80

85

90

95

100

Elapsed time (years)

Perc

ent b

atte

ry c

apac

ity re

mai

ning

(%)

PEV2 no grid dischargePEV5 no grid dischargePEV7 no grid dischargePEV10 no grid dischargePEV12 no grid discharge

Case 3: Predicting battery degradationfrom driving vs. vehicle-grid services

Case 3a 12 EVs Capacity fade of a

LiFePO4 battery pack (driving only)

0 5 10 15 200.4

0.5

0.6

0.7

0.8

0.9

1

Time (hours)

Vehi

cle

Batte

ry S

OC

Battery SOC profile for 12 electric vehicles

PEV1 PEV2 PEV3 PEV4 PEV5 PEV6 PEV7 PEV8 PEV9 PEV10 PEV11 PEV12

Individual EV SOCprofiles from driving only

(Repeated for 10 years)

Preliminary results

Estimated capacity fade due tobattery use from driving only





0 1 2 3 4 5 6 7 8 9 1075

80

85

90

95

100

Elapsed time (years)

Perc

ent b

atte

ry c

apac

ity re

mai

ning

(%)

PEV2 no grid dischargePEV5 no grid dischargePEV7 no grid dischargePEV10 no grid dischargePEV12 no grid dischargePEV2 with grid dischargePEV5 with grid dischargePEV7 with grid dischargePEV10 with grid dischargePEV12 with grid discharge

0 5 10 15 200

0.2

0.4

0.6

0.8

1

Time (hours)

Vehi

cle

Batte

ry S

OC

Battery SOC profile for 12 electric vehicles

PEV1 PEV2 PEV3 PEV4 PEV5 PEV6 PEV7 PEV8 PEV9 PEV10 PEV11 PEV12

Case 3: Predicting battery degradationfrom driving vs. vehicle-grid services

Case 3b 12 EVs Capacity fade of a

LiFePO4 battery pack (driving + grid service)

Individual EV SOC profiles for driving and grid services

Example: Discharging 6-9 pm(Repeated for 10 years)

Preliminary results

Quantifies degradation impact from any grid service scenario

(This case: discharging at maximum power from 6-9 pm)





Case 4: PEVs for Renewables Integration

4 6 8 10 12 14 16 18 20 22 24Time of Day (Hours after Midnight)

Duck + L2Home,L2Work (Controlled V1G)Duck + L2Home,L2Work (Controlled V2G)Duck + L2Home,L2Work (Uncontrolled)Original Duck Curve

010

15

20

25

30

Net

Loa

d (G

W)

Net load = Forecasted load – Forecasted wind & solar generation

CAISO “Duck Curve”

V2G‐Sim resultsfor ~3M PEVs(mostly PHEVs)

Plug‐in vehicles canalleviate challengesfrom large‐scale

renewables integration





70%

75%

80%

85%

90%

95%

100%

6:00 AM 9:00 AM 12:00 PM 3:00 PM 6:00 PM 9:00 PMPercen

t Red

uctio

n in EV Ch

arging

Loa

d

Start Time of Demand Response Event

1 hour DR ‐ L2 Home, L2 Work2 hour DR ‐ L2 Home, L2 Work4 hour DR ‐ L2 Home, L2 Work

Case 5: Flexibility of EVs to varycharging to offer demand response

Parametrically simulate DR events for different durations, at different times of day

75-95% EV charging can be shed during DR event without interfering with driver needs

Charging load from 3,166 EVs

12:00 am 4:00 am 8:00 am 12:00 pm 4:00 pm 8:00 pm0

500

1,000

1,500

2,000

2,500

3,000

Time of day (hours)

Uncontrolled chargingControlled charging - DR 5-9pm

Uncontrolled chargingControlled charging - DR 5-9pm

4:00 pm 6:00 pm 8:00 pm 10:00 pm70%

80%

90%

100%

Perc

enta

ge re

duct

ion

in c

harg

ing

load

(%)

Tota

l cha

rgin

g lo

ad (k

W)

Max Avg

Min

Managed charging controller in V2G-Sim set up to reduce PEV charging

during demand response events(without interfering with any drivers’ mobility needs)

4:00 pm 6:00 pm 8:00 pm 10:00 pm0

1,000

2,000

3,000

4,000

5,000

Time of day

Tota

l cha

rgin

g lo

ad (k

W)

Can mitigate post-DR peak by gradually resuming EV charging





Case 6: Quantifying the adequacyof EVs to meet driver needs

Travel patterns from NHTS for the entire United States(2009 data – indicative of how people want to drive if they had no range limitations)o # Samples: 120,495 weekday drivers, 39,349 weekend drivers

Model vehicles with Nissan Leaf specs in V2G-Sim

V2G-Sim predicts SOC profile of each vehicle in each charging scenario





84%

86%

88%

90%

92%

94%

96%

98%

U.S. Weekday travel U.S. Weekend travel

Fractio

n of U.S. travel nee

ds satisfied by

EVs (%

)

L1 Home, No Work

L1 Home, L1 Work

L1 Home, L2 Work

L2 Home, No Work

L2 Home, L2 Work

L1 All Locations

L2 All Locations

50 kW All Locations

L1 = standard 120 V outlet. up to 1.44 kW charge rate.

L2 = up to 7.2 kW chargerate.


Daily travel needs of 85-89% of U.S. drivers are satisfied with EVs charging on 120V wall outlets at homes only






0%

10%

20%

30%

40%

50%

60%

Percen

t of Total EVs (%)

0%

10%

20%

30%

40%

50%

60%Pe

rcen

t of Total EVs (%) L1 Home, No Work

L1 Home, L1 WorkL1 Home, L2 WorkL2 Home, No WorkL2 Home, L2 WorkL1 All LocationsL2 All Locations50 kW All Locations

(A) U.S. Weekday Travel

A1

A2

A3

A4

Fraction of vehicles with over60 km of buffer range for

unexpected trips:A1 + A2 + A3 + A4 = 77%

(B) U.S. Weekend Travel

0 20 40 60 80 100 120 140Minimum Remaining EV Range During Travel Day (km)

Fraction of vehicles with over60 km of buffer range for

unexpected trips:B1 + B2 + B3 + B4 = 79%

B1

B2

B3

B4





40%

50%

60%

70%

80%

90%

100%

Base case scenario

4.8 kW ancillary load

20% loss of batt capacity

3% Uphill Grade for all Trips

Worst case

scenario

Fractio

n of U.S. travel nee

ds satisfied by

EVs (%

)

Base case scenario

4.8 kW ancillary load

20% loss of batt capacity

3% Uphill Grade for all Trips

Worst case

scenario

L1 Home, No WorkL1 Home, L1 WorkL1 Everywhere

(B) U.S. Weekend Travel(A) U.S. Weekday Travel






Case 7: Redefining the useful lifetimeof EV batteries and the start of second life

10% 20% 30% 40% 50% 60% 70% 80% 90%-400

-300

-200

-100

0

100

200

300

400

Battery state of charge (%)

Batte

ry p

ower

out

put (

kW)

UDDS HWFET US06 Max charge Max discharge

100%90%

80%70%

60%50%

40%30%

100%90%

80%70%

60%50%

40%30%

55%

60%

65%

70%

75%

80%

85%

90%

95%

30%40%50%60%70%80%90%100%

Fractio

n of U.S. D

rivers W

hose Daily Travel N

eeds

are Satisife

d Und

er Differen

t Levels o

f Cap

acity

Fad

e

Remaining Usable Battery Capacity (% of 23.8 kWh battery)

L1 Home, No WorkL1 Home, L1 WorkL2 Home, No WorkL2 Home, L1 WorkL2 Home, L2 WorkL1 Everywhere

Impact of energy capacityfade on ability to satisfy

drivers’ daily travel needs

Impact of power fade on ability to meet drive cycle requirements

EV batteries continue to meet driver needs far

longer than expected…

…battery useful life is much longer than

expected





MyGreenCar: Leveraging V2G-Sim to accelerate deployment of clean vehicles

Drivers and fleet managers are often unaware of the benefits of clean vehicles

Many are unaware whether these vehicles even meet their individual needs

o E.g. range anxiety is a key obstacle for EV adoption

MyGreenCar is an end user application of V2G-Sim that provides drivers with personalized information on how clean vehicles meet their needs





How MyGreenCar Works

1. Users record their travel patterns and personalized drive cycles for any duration (e.g. 3 weeks, 3 months, etc.)

2. User chooses vehicle types and data is transmitted to a V2G-Sim server that uses physics-based powertrain models to predict SOC, charging needs, fuel consumption, etc.o Models consider important factors such as traffic, terrain, ancillary power

consumption, etc.3. Analysis results are reported to the user

8:00 am 10:00 am 12:00 pm 2:00 pm 4:00 pm 6:00 pm

65%

70%

75%

80%

85%

90%

95%

Time of day

Vehi

cle

batte

ry S

OC

(%)

Vehicle # 15,301Vehicle # 98,170

Line Styles

10 15 20Time (hours) User Report

1. Fuel econ2. SOC3. Charging

needs4. Etc…





MyGreenCar Application 1: Eliminatingrange anxiety for prospective car buyers

Accelerating EV adoption by eliminating range anxiety for prospective EV ownerso Does an EV provide the range requirements for my normal travel

patterns? Am I ever in danger of running out of charge?o How much range will I have if I need to make an unexpected trip?o Do I need a level 2 charger at home, or is level 1 sufficient?o Do I need a charger at work or other locations?

o How much will I pay for electricity? How does this compare to what I will pay for gasoline in a comparable conventional car?

o What incentives are available?





MyGreenCar Application 2: EliminatingEV range anxiety for current EV owners

Eliminating range anxiety for current EV ownerso At my car’s current SOC, do I have enough charge to make my next

trip? (when considering traffic, terrain, ancillary power consumption, etc.)

o If not, how much do I need to charge my car? How long will this charging take on different types of chargers?

Prediction of trip’s drive cycleIdentify likely route

Mapping / V2G-Sim server Powertrain model calculations

Forecast SOC

Does the vehiclehave sufficient SOC

for the trip?





Application 3: Personalized fuel economy estimates for any driver in any car

Enabling drivers to compare fuel economy across vehicles for their own driving style

(i.e. personalized EPA fuel economy label)o How do these cars compare in terms of fuel economy and operating

costs for my personal driving patterns and style?o Are there comparable cars available that will offer greater fuel

economy for my driving style?

vs.Compare several carsConventional modelsvs. hybrid modelsvs. plugin models, etc.

My PersonalFuel Economy

+ +





MyGreenCar Application 4:Guiding fleet vehicle purchasing

Enabling fleet managers to quantify the benefits of clean vehicles and deploy them to capture the greatest benefitso How much fuel savings will an advanced vehicle get in comparison to

a conventional vehicle on our routes?o How long will it take to recoup the added capital cost of an advanced

vehicle?o Which routes should I deploy the advanced vehicle on to capture the

greatest fuel savings benefits?

vs.

Diesel vs. natural gasvs. diesel hybrid vs. etc. ?

Which routes?





More applications of MyGreenCar…

Schedule a test drive within the MyGreenCar App

Enable drivers to understand economic benefits of clean vehicles Fuel savings, incentives, time-of-use rates for charging, etc.

Research dataset – more detail & samples than any other travel survey

Spatially forecasting PEV deployment for grid infrastructure planning

Enabling real-time data collection to enable vehicle-grid integration MyGreenCar as interface to an aggregator – specify planned travel activity PEV owners + MyGreenCar + V2G-Sim forecasting + aggregation server = grid services

Utilities: a tangible tool for outreach and accelerating PEV adoption

And more…





Transformative applicationsspanning disciplines

MyGreenCar V2G‐Sim Samveg [email protected]

Understand,quantify, andavoid batterydegradation

from VGI

Optimizepowertrain

controlsystemsfor VGI

Develop andunderstandimpacts ofmanagedcharging

algorithms

Understanddistribution &transmission

system impacts

Plan futureinfrastructure

Develop VGIbusinessmodels &

understandtheir impacts

Forecast PEV loads

Quantifyimpact of VGIfor many grid applicationsand markets

Acceleratingclean vehicledeployment

Eliminating EVrange anxiety

Predictingpersonalizedfuel economy

V2G‐Sim Analysis planned release: Nov‐Dec 2014

Documents

Eliminating EV Range Anxiety, Predicting Personalized Fuel ... · PDF file19/11/2014 · Eliminating EV Range Anxiety, Predicting Personalized Fuel Economy, ... EV range anxiety, Predicting