do more batteries make a plugin hybrid better
TRANSCRIPT
Do More Batteries Make A Plugin Hybrid Better?Implications from Optimal Vehicle Design and Allocation
Lawrence Berkeley Laboratory SeminarJune 18, 2010
ChingShin Norman ShiauPostdoctoral Research FellowMechanical EngineeringCarnegie Mellon University
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
15
20
25
30
35
40
1978 1988 1998 2008 2018
EISA35 mpg
Obama 35.5 mpg
32.5
24.5
CA
FE s
tand
ards
(mpg
)
Year
CAFE Policy
Research Overview
2
Product Design
Public Policy
Market Systems
Vehicle Design Decisions for Market and Policy
Design for Market Systems
Implications of VehicleElectrification
Plug-in Hybrid
Market Competitionand Structure
w1a
M1
Consumer
CRa CRb
M2w1bw2a w2b
p1a p1bp2a p2b
M1
Consumer
CS1w1 = p1
M2CS2
w2 = p2
Company store
Franchised RetailerM1
FR1 FR2
w1
p1
w2
p2
M2
Consumer
Multiple Common Retailers
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Plugin Hybrid Technology
3
What are the economic and environmental implications of plug-in hybrid vehicles?
• Public concerns: global warming on GHG emissions and foreign oil dependency in the US transportation sector
• Plug-in hybrid electric vehicle (PHEV) is considered as a potential technology to address these issues
• Net effects of PHEVs depend critically on vehicle design and battery technology
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Vehicle Technologies
4
CV HEV PHEV BEV
X X
X X
XGasoline
Electricity
Power Convertor Engine MotorEngine & Motor Engine & Motor
Battery Pack None Small Medium Large
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
PHEV Powertrain System
5
Control motor
Traction motor
Planetary gear set
Power to
wheels
Electrical Energy Flow
Gasoline Energy Flow
Engine
Fuel tank
Power inverter
100%Max SOC
Target SOCMin SOC
DOD(s)
Phys
ical
cap
acity
lim
it
Rate
d ca
paci
ty
Swin
g
Battery pack
0%Driving distance s
Stat
e of
Cha
rge
(SO
C)
Charge-Deplete
(CD)
Charge-Sustain (CS)
AER
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
PHEV Literature Review Prior PHEV Studies◦ EPRI Report (2001)◦ Simpson [NREL] (2006)◦ EPRI-NRDC Report (2007)◦ Lemoine, Kammen, and Farrell [UC Berkeley] (2008)◦ Samaras and Meisterling [CMU-EPP] (2008)◦ Kromer and Heywood [MIT] (2009)◦ Sioshansi and Denholm [NREL] (2009)◦ NRC PHEV analysis report (2009)◦ Plotkin and Singh [ANL] (2009)
This study focuses on the implications of optimal PHEV designs and driver allocations on life cycle performance
6
What are the best vehicle choices to reducing life cycle GHGs, cost and fuel consumption?
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
PHEV Optimization Framework
7
Benevolent dictator to determine optimal PHEV designs and driver allocations for social objectives
Optimization ModelMinimize total fuel/cost/GHGs
with respect to design and allocationsubject to vehicle design constraints
SolutionsOptimal vehicle types, drivervehicle allocations and design decisions
PHEV PerformanceMetamodel for simulation data
National Driving DataDistribution for driving distance
Battery DegradationModel for battery life
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Vehicle Simulation
8
Use the Powertrain Systems Analysis Toolkit (PSAT) vehicle simulator developed by Argonne National Laboratory (ANL)
Start with a model of a Toyota Prius Use EPA UDDS cycles to test PHEV performance
Optimization ModelMinimize total fuel/cost/GHGs
with respect to design and allocationsubject to vehicle design constraints
SolutionsOptimal vehicle types, drivervehicle allocations and design decisions
PHEV PerformanceMetamodel for simulation data
National Driving DataDistribution for driving distance
Battery DegradationModel for battery life
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Optimization ModelMinimize total fuel/cost/GHGs
with respect to design and allocationsubject to vehicle design constraints
SolutionsOptimal vehicle types, drivervehicle allocations and design decisions
PHEV PerformanceMetamodel for simulation data
National Driving DataDistribution for driving distance
Battery DegradationModel for battery lifeDaily travel distribution
9Federal Highway Administration, 2010, "National Household Travel Survey 2009," Department of Transportation, Washington, DC.
•Use the 2009 NHTS data to estimate the national average distance driven per day over the population of drivers•Size of the 2009 survey: 136,410 households
0 50 100 150 2000
0.01
0.02
0.03
Daily driving miles
Prob
abilit
y de
nsity
func
tion
S ; 0sf s e s λ = 0.0296 (maximum likelihood)
•We fit the weighted driving data using the exponential distribution
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
0 0.2 0.4 0.6 0.8 1Ba
ttery
lfie
cha
rge
cycl
eDepth of discharge per cycle
(%)
103
104
105
106
Battery degradation
10
0.88
0.92
0.96
1
0 1 2 3 4
Rel
ativ
e en
ergy
capa
city
Normalized energy processed(Wh processed / Wh capacity)
1×103 2×103 3×103 4×103
Peterson energy-based model
Peterson, S.B., Whitacre, J.F., and Apt, J. (2010) Lithium-Ion Battery Cell Degradation Resulting from Realistic Vehicle and Vehicle-to-Grid Utilization. Journal of Power Sources. 195(8) 2385–2392.
Rosenkranz, K. (2003) Deep-Cycle Batteries for Plug-in Hybrid Application. Presentation in EVS-20 Plug-In Hybrid Vehicle Workshop, November 16-19, Long Beach, CA
RosenkranzDOD-based model
Markel and Simpson [NREL] (2006)Simpson [NREL] (2006)Kromer and Heywood [MIT] (2009)
LiFePO4 cell
Optimization ModelMinimize total fuel/cost/GHGs
with respect to design and allocationsubject to vehicle design constraints
SolutionsOptimal vehicle types, drivervehicle allocations and design decisions
PHEV PerformanceMetamodel for simulation data
National Driving DataDistribution for driving distance
Battery DegradationModel for battery life
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Estimated battery life using two models• Rosenkranz model • Not account for the battery
degradation in CS mode because of no DOD variation
• Can be too optimistic for short distance driving
11Plots using a PHEV40 with battery
9.5kWh and swing at 80%
Battery energy status
swin
g
CD mode CS mode
Driving distance
Sta
te o
f ene
rgy
Charging
AER
Fully charged
Target SOC
• Peterson model• Tested under variable C-rate• Account for energy processed
in CS mode and charging
0
100
200
300
400
0 50 100 150 200
Batte
ry li
fe in
VM
T (m
iles)
Daily driving distance (miles)
Battery life in VMT
Optimization ModelMinimize total fuel/cost/GHGs
with respect to design and allocationsubject to vehicle design constraints
SolutionsOptimal vehicle types, drivervehicle allocations and design decisions
PHEV PerformanceMetamodel for simulation data
National Driving DataDistribution for driving distance
Battery DegradationModel for battery life
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
BAT NN VEH NC OP
BAT NR
Buy: ,, , 1,Lease: ,,
c CRF r TCRF r T c CRF r Tf s c
c CRF r BCRF r T D D
x
Objective Functions
12
Fuel consumption per day:
GG
G
, sf s
x = efficiency (miles/kWh or miles/gal)
Equivalent annualized cost (EAC) per day:
operating vehicle battery
Average lifecycle GHG emissions per day:
BATVEHV OP
BAT
Buy:,
Lease:v T
f sv BTD
x
operatingvehicle battery
Design variables:x = vehicle design variable (engine, motor, battery and swing)si = vehicle allocation range
Constraints:060mph acceleration time, minimum SOC after multiple US06 cycles
T: vehicle life yearB: battery life yearD: driving days per year
rN: nominal (market) discount raterR: real (inflation-free) discount rate
Optimization ModelMinimize total fuel/cost/GHGs
with respect to design and allocationsubject to vehicle design constraints
SolutionsOptimal vehicle types, drivervehicle allocations and design decisions
PHEV PerformanceMetamodel for simulation data
National Driving DataDistribution for driving distance
Battery DegradationModel for battery life
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Core concept of the optimization model
13
1O S, 1, , 1
minimize ,i
ii i
n s
iss i n i
f s f s ds
x
x
Objective function per day {$, GHG, gal}
pdf of distance driven per day
sis
< si per daydrive #1
> si per daydrive #2
Vehicle #1Vehicle #2 ,i s x
,i s x Population-weighted objective
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Assumptions
Fuel and electric efficiency estimated by UDDS cycles $400 per kWh Li-ion battery pack cost (PHEV) $600 per kWh NiMH battery pack cost (HEV) $3.30 per gal gasoline (2008 average) $0.11 per kWh electricity (2008 average) 0.69 kg-CO2-eq/kWh grid emission (US average mix) $0 per ton-CO2-eq allowance price 5% discount rate EPRI powertrain cost model Buy-lease battery replacement Peterson battery degradation model
14
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Optimal PHEV Designs and Allocations
15
Base case settings: Buy-lease battery replacement, Peterson battery degradation model, $400/kWh Li-ion battery cost, $600/kWh NiMH cost, $3.30/gal gasoline, $0.11/kWh electricity, grid emission 0.69 kg-CO2-eq/kWh, $0/ton CO2-eq allowance price, and 5% discount rate
Pop
ulat
ion-
wei
ghte
d ob
ject
ive
per v
ehic
le p
er d
ay
Minimum Fuel Consumption
Minimum Lifecycle GHGs
Minimum Lifecycle Costs
PHEV87(21.6kWh*80%)
0.000
0.004
0.008
0.012
0 100 200
f G(x
* ,s)f S
(s)
CV
HEV
PHEV870.00
0.04
0.08
0.12
0.16
0 100 200
HEV
PHEV40
PHEV25
CV
f V(x
* ,s)f S
(s)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0 100 200
f C(x
* ,s)f S
(s)
CV
PHEV20
HEV
Daily travel milesPHEV25(7.0kWh*68%)PHEV40(9.6kWh*80%)
PHEV20(5.4kWH*67%)
31 50
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Breakdown analysis
16
Lifecycle GHGs Lifecycle EAC
0
2
4
6
8
10
12
14
CV HEV PHEV25 PHEV40
Lifecycle
GHG
s pe
r day (k
g‐CO
2‐eq
)
GHGs on operationGHGs on batteryGHGs on vehicle
2% 4%
87%
76%73% 71%
13% 23%
1%
25% 25% 0
2
4
6
8
10
CV HEV PHEV20
EAC pe
r day ($
)
EAC on operationEAC on batteryEAC on vehicle
50% 69% 70%
3% 10%
50%28% 20%
Assume travel distance 30 miles a day
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Sensitivity analysis for min. GHGs
17
• US average grid mix: 0.69 kg-CO2e/kWh (Weber et al., 2010)• Nuclear: 0.07 kg-CO2e/kWh (Sovacool, 2008)• IGCC w/ CCS: 0.25 kg-CO2e/kWh (Jaramillo et al., 2009)• Natural gas: 0.47 kg-CO2e/kWh (Weisser, 2007)• Coal: 0.90 kg-CO2e/kWh (Weisser, 2007)
0% 20% 40% 60% 80% 100%
Coal 8.19 kgCO2e/vehicleday
Natural gas 6.20 kgCO2e/vehicleday
IGCCCCS4.60 kgCO2e/vehicleday
Nuclear3.22 kgCO2e/vehicleday
Min. GHGs base case7.73 kgCO2e/vehicleday
Population (%)
PHEV3637 miles
34 miles
62 miles
Daily distance (miles) 0 7.5 17.3 31.0 54.5 200
HEV
PHEV60PHEV87
PHEV27 PHEV87PHEV4783 miles
VMT (%) 0% 2% 10% 24% 49% 100%
PHEV3637 miles 62 miles
PHEV59 PHEV87
PHEV3437 miles 65 miles
PHEV57 PHEV87
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Sensitivity analysis for min. cost
18
• Consider 3-vehicle segment as base case
• High battery cost, low gas price and high elec. price make PHEV not cost competitive
• To make PHEVs part of the least-cost solution:Li-ion < $590/kWh at 5%Li-ion < $410/kWh at 10%
0% 20% 40% 60% 80% 100%
Carbon price $100/ton$8.05/verhicleday
Nm. discount rate 10%$9.31/vehicleday
Elec. price $0.30/kWh$7.39/vehicleday
Elec. price $0.06/kWh$7.02/vehicleday
Gas price $6.00/gal$7.89/vehicleday
Gas price $1.50/gal$6.18/vehicleday
Liion battery $1000/kWh$7.39/vehicleday
Liion battery $250/kWh$6.99/vehicleday
Min. cost base case$7.23/vehicleday
Population (%)
PHEV1620 miles
31 miles 65 miles
Daily distance (miles) 0 7.5 17.3 31.0 54.5 200
51 milesPHEV25 HEV
PHEV19 PHEV25PHEV35
HEV
HEV
HEV
PHEV17 PHEV49PHEV3220 miles 38 miles
PHEV17 PHEV32 HEV23 miles 69 miles
VMT (%) 0% 2% 10% 24% 49% 100%
PHEV16 PHEV20 PHEV2620 miles 50 miles
CV HEVPHEV188 miles 32 miles
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
ARRA tax credit vs. carbon tax
19
Tax credit in the ARRA bill (Page 212-219)
Light-duty vehicles: Vehicle base amount is $2,500 for a PHEV can draw propulsion energy from a battery with not less than 5 kWh of capacity (base $417), plus $417 for each kWh of capacity in excess of 5 kilowatt hours. Total amount for battery shall not exceed $5,000. Thus the tax credit can be calculated by min($7500, $2500 + $417 + $417*(kWh − 5))
ARRA: http://fdsys.gpo.gov/fdsys/pkg/BILLS-111hr1ENR/pdf/BILLS-111hr1ENR.pdf
• Tax credit in the American Recovery and Reinvestment Act (ARRA) for battery size in PHEV/EV
•ARRA incentivizes PHEVs with larger batteries (AER 33-49), which results in 6% higher social costs than under a $100/ton CO2 tax scenario
0
1000
2000
3000
4000
5000
6000
7000
8000
0 5 10 15 20Tax c
redit ($)
Battery size (kWh)
GM Volt
Priusplug-in
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
More sensitivity tests
20
•Alternative vehicle base costs and powertrain cost models The base case optimal solution is robust
•Battery life must outlast vehicle life Optimal battery swing reduced 2-5%
•Battery leasing (prorated: paid per portion)Larger battery packs with reduced swings
•Battery end of life (EOL) at 20% capacity fade Makes PHEVs less cost competitive
•Rosenkranz DOD-base degradation modelOptimal PHEV11 with a 5.2 kWh battery at 39% swing
(a battery size equivalent to a PHEV23 at 80% swing)
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Take away 1
• Large PHEVs with AER > 50 miles can reduce petroleum consumption
• PHEVs with AER 25-50 miles can reduce GHGs
• PHEVs with AER 15-25 miles for short distance travel and HEVs for longer distance travel can save cost
• Li-ion battery pack cost below $590/kWh at a 5% discount rate (or below $410/kWh at 10%) to make PHEV part of least-cost choice
21
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Take away 2
Battery swing in excess of 60% with LiFePO4technology should be utilized to achieve minimum life cycle cost, GHGs, and petroleum consumption
Carbon allowance prices have marginal impact on optimal design or allocation of PHEVs
Battery tax credit in the ARRA bill may result in higher net costs than under a $100/ton CO2 tax scenario
22
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Future work
Consumer and market behaviors PHEV performance on real-world driving cycles Uncertainty in GHG emissions◦ Regional effects, charge timing, and marginal dispatch
Battery technology◦ Thin-electrode for high-power and thick-electrode for high-energy
batteries◦ Calendar (storage) degradation◦ Blended-mode PHEVs
Gasoline prices and grid characteristics can become endogenous after significant PHEV market penetration
23
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Publications
24
Shiau, C.-S.N. and J.J. Michalek (2009) Optimal product design under price competition. ASME Journal of Mechanical Design, 131(7), 071003.
Shiau, C.-S.N. and J. J. Michalek (2009) Should designers worry about market systems? ASME Journal of Mechanical Design, 131(1), 011011.
Shiau, C.-S.N., C. Samaras, R. Hauffe and J. J. Michalek (2009) Impact of battery weight and charging patterns on the economic and environmental benefits of plug-in hybrid vehicles. Energy Policy, 37(7), 2653-2663.
Shiau, C.-S.N., J.J. Michalek and C.T. Hendrickson (2009) A structural analysis of vehicle design responses to Corporate Average Fuel Economy policy. Transportation Research Part A: Policy and Practice, 43(9-10), 814-828.
Shiau, C.-S.N., J.J. Michalek, N. Kaushal, C.T. Hendrickson, S.B. Peterson and J.F. Whitacre(2010) Optimal plug-in hybrid electric vehicle design and allocation for minimum life cycle cost, petroleum consumption and greenhouse gas emissions. ASME Journal of Mechanical Design: Special Issue on Sustainable Design, In review.
And 9 peer-reviewed conference papers.
CS Norman Shiau – Implications of Optimal PHEV Design and Allocation
Acknowledgements
25
I would like to thank:Professor Jeremy MichalekProfessor Chris HendricksonProfessor Jay WhitacreDr. Paulina JaramilloNikhil KaushalScott PetersonCamilo ResendeDr. Constantine SamarasElizabeth Traut
This research was supported in part by a grant from the NSF MUSES program: Award #0628084, the NSF CAREER Award #0747911, CMU CIT Liang Ji-Dian Graduate Fellowship, and grants from Ford Motor Company and Toyota Motor Corporation.