ANL/ESD/02-2

Hybrid Electric Vehicle Technology Assessment:

Methodology, Analytical Issues, and Interim Results
Center for Transportation Research
Argonne National Laboratory
Operated by The University of Chicago,
under Contract W-31-109-Eng-38, for the
United States Department of Energy
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 The University of Chicago, 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 The University of Chicago.
Argonne National Laboratory, with facilities in the states of Illinois and Idaho, is
owned by the United States Government and operated by The University
of Chicago under the provisions of a contract with the U.S. Department of Energy.
Available electronically at http://www.doe.gov/bridge
Available for a processing fee to U.S. Department of
Energy and its contractors, in paper, from:
U.S. Department of Energy
Office of Scientific and Technical Information
P.O. Box 62
Oak Ridge, TN 37831-0062
phone: (865) 576-8401
fax: (865) 576-5728
email: [email protected]
ANL/ESD/02-2
Hybrid Electric Vehicle Technology Assessment:
Methodology, Analytical Issues, and Interim Results
by S. Plotkin, D. Santini, A. Vyas, J. Anderson, M. Wang, J. He,
*
and D. Bharathan
*
Center for Transportation Research, Energy Systems Division,
Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439
October 2001
Work sponsored by the United States Department of Energy,
Assistant Secretary for Energy Efficiency and Renewable Energy,
Office of Transportation Technologies
*
J. He is currently affiliated with Lucent Technologies, Naperville, Illinois. D. Bharathan is currently affiliated with the
National Renewable Energy Laboratory, Golden, Colorado.
NOTICE
This technical report is a product of Argonnes Energy Systems Division.
For information on the divisions scientific and engineering activities,
contact:
Director, Energy Systems Division
Argonne National Laboratory
Argonne, Illinois 60439-4815
Telephone (630) 252-3724
Publishing support services were provided by Argonnes Information
and Publishing Division (for more information, see IPDs home page:
http://www.ipd.anl.gov).
This report is printed on recycled paper.
iii
Contents
Acknowledgments.................................................................................................................... vii
Notation.................................................................................................................................... ix
Summary .................................................................................................................................. 1
1 Introduction ....................................................................................................................... 11
Section 1 Appendix A: At What Types of Vehicles Are We Looking?........................... 15
Section 1 Appendix B: What Is a Vehicle?...................................................................... 17
2 Hybrid-Electric Vehicles: Theory and Design .................................................................. 21
2.1 Background .............................................................................................................. 21
2.2 Energy Use in Conventional Vehicles ..................................................................... 23
2.3 Energy Savings Potential of Hybrid Drivetrains...................................................... 25
2.4 Emissions Trade-Offs............................................................................................... 27
2.5 Elements of Hybrid Design ...................................................................................... 28
2.5.1 Parallel Hybrids ............................................................................................. 28
2.5.2 Series Hybrids................................................................................................ 30
2.5.3 Component Sizing.......................................................................................... 31
2.5.3.1 A Dated Thermostat Concept ....................................................... 31
2.5.3.2 Grid-Independent Hybrids ............................................................... 32
2.5.3.3 Grid-Connected Hybrids.................................................................. 33
2.5.3.4 Limited Performance Hybrids.......................................................... 34
2.5.4 Role of Component Efficiency ...................................................................... 35
2.5.5 Trends in Hybrid Efficiency over Time......................................................... 36
2.6 References ................................................................................................................ 37
3 Methodology and Modeling Issues ................................................................................... 39
3.1 Modeling the Vehicle............................................................................................... 39
3.1.1 ADVISOR Model and Validation.................................................................. 39
3.1.2 Rules for Comparing Vehicles and Modeling Difficulties ............................ 40
3.1.2.1 Selecting a Conventional Engine/Hybrid Engine Pair
for Comparison................................................................................ 42
3.1.2.2 Engine Maps.................................................................................... 44
3.1.2.3 Adequacy of Transmission Modeling............................................. 46
3.1.2.4 Zero to 60 mph Acceleration Times................................................ 48
3.1.2.5 Hybrid Design and the Impact of Cost Considerations................... 51
3.1.2.6 Fair Treatment of Time and Context............................................... 52
3.1.2.7 Calculating the Energy Use of Grid-Connected Hybrids................ 52
iv
3.2 Cost Modeling.......................................................................................................... 56
3.2.1 Common Components Retail Price Equivalent ............................................ 57
3.2.2 Allocation of Indirect Costs........................................................................... 59
3.2.3 Component Cost Information......................................................................... 60
3.2.3.1 OEM Manufactured Components.................................................... 61
3.2.3.2 Outsourced Electric Powertrain Components.................................. 61
3.2.3.3 Outsourced Battery Pack.................................................................. 62
3.2.4 Modeling Concerns........................................................................................ 62
3.3 References ................................................................................................................ 63
4 Results of Vehicle Modeling............................................................................................. 65
4.1 Description of the Vehicles Evaluated..................................................................... 65
4.2 Vehicle Control Strategy.......................................................................................... 68
4.3 Driving Cycles.......................................................................................................... 68
4.4 Other Critical Assumptions...................................................................................... 73
4.5 Estimating Energy Use from Grid-Connected HEVs............................................... 73
4.6 Model Results........................................................................................................... 74
4.7 References ................................................................................................................ 89
Section 4 Appendix A: Table of Results: HEV Cost and ADVISOR Modeling,
11/1999-3/2000................................................................................................................ 91
Section 4 Appendix B: Reality Check on ADVISOR Results: Highway Cycle
Fuel Savings for Parallel Hybrids ................................................................................... 105
Section 4 Appendix C: Power and Mass Computations for Initial Vehicle Sizing........... 109
Figures
2-1 Energy Flow for a Midsize Automobile ..................................................................... 25
2-2 One Example of a Parallel Hybrid Drivetrain............................................................. 29
2-3 One Configuration of a Series Hybrid Drivetrain....................................................... 30
3-1 Saturn 1.9-L Fuel Use Map with FTPRP Coverage.................................................... 47
3-2 Plots of Measured and Estimated Z60 Time vs. the Vehicle Power-to-Weight
Ratio for Conventional Vehicles and Hybrids............................................................ 50
3-3 A Days Driving for a Grid-Connected Hybrid Vehicle: SOC vs.
Time or Distance......................................................................................................... 53
3-4 Cumulative Share of Vehicles Traveling vs. Miles Traveled,
from the Nationwide Personal Transportation Survey. .............................................. 56
4-1a Speed Profile of the New York City Cycle................................................................. 70
4-1b Speed Profile of the Japan 10/15 Cycle...................................................................... 71
v
4-1c Speed Profile of the Federal Test Procedure .............................................................. 71
4-1d Speed Profile of the U.S. No. 6 Cycle ........................................................................ 72
4-1e Speed Profile of the Highway Cycle........................................................................... 72
4-2a HEV Fuel Economy Advantage over CVs, for Different Driving Cycles.................. 75
4-2b HEV Fuel Economy Advantage over CVs, Measured in Gallons Saved in
10 Hours of Driving.................................................................................................... 77
4-3 Fuel Economy Variation of CV and HEVs with Z60 Requirements.......................... 78
4-4a Hybrid Vehicles Fuel Economy Improvement over CVs as a Function of
Acceleration Times ..................................................................................................... 79
4-4b CV and HEV Gradeability at 55 mph, Grid-Independent Parallel Hybrids ............... 80
4-5 Projected Year 2010 Cost of CVs and Hybrids .......................................................... 81
4-6 Vehicle Cost as a Function of Acceleration Requirements ........................................ 82
4-7 Breakeven Fuel Prices for CVs and Parallel Grid-Independent Full Hybrids,
for Three Performance Levels .................................................................................... 83
4-8 Breakeven Fuel Prices for CVs and Parallel Grid-Independent Mild Hybrids,
for Three Performance Levels. ................................................................................... 84
4-9 Fuel + Electricity Cost: CVs vs. Parallel Grid-Connected and Grid-
Independent HEVs ...................................................................................................... 85
4-10 Vehicle Cost Comparison: Grid-Connected and Grid-Independent Parallel
HEVs vs. CV............................................................................................................... 85
4-11 Life Cycle Costs: CVs vs. Grid-Connected and Grid-Independent Parallel HEVs.... 86
4-12 Comparison of Gasoline Use: CV vs. Grid-Independent and Grid-Connected
HEVs in Urban Driving. ............................................................................................. 87
4-13 CO
2
Emissions from Similar Conventional and Hybrid Vehicles in Urban
Driving. ....................................................................................................................... 87
vi
Tables
3-1 Common and Dissimilar Components of CV and HEV............................................. 58
3-2 Allocation of MSRP to Midsize Car Groups and Common Components .................. 59
3-3 Vehicle Manufacturing and Retail Price Structure..................................................... 60
3-4 Methodologies Used for Estimating OEM-Manufactured Component Costs ............ 61
3-5 Fixed and Variable Terms Used for Estimating Electric Drive Component Costs .... 61
3-6 Nickel Metal Hydride Battery Characteristics and Costs ........................................... 62
4-1 Midsize HEV Characteristics...................................................................................... 66
4-2 Nickel Metal Hydride Battery Characteristics............................................................ 67
4-3 Specific Power and Efficiency Values for HEV Drivetrain Components.................. 68
4-4 Characteristics of Seven Driving Cycles .................................................................... 70
4-5 Component Power Requirements of HEV Drivetrains............................................... 81
vii
Acknowledgments
Funding for this study was provided predominantly by Dr. Philip Patterson, manager of
Planning and Analysis for Tom Gross, Deputy Assistant Secretary of the Department of Energys
Office of Transportation Technologies. Some funding was also provided by Paul McArdle, then
of the Office of Technology Utilization within the Office of Transportation Technologies, whose
director is David Rodgers.
The authors would like to acknowledge the strong support and valuable advice of Dr. Philip
Patterson of the Office of Transportation Technologies, U.S. Department of Energy. Keith
Wipke and Mathew Cuddy of the National Renewable Energy Laboratory, the designers of the
ADVISOR model, were extremely generous with their time and advice and were invaluable to
this study. Authors appreciate the time and effort put in by the reviewers: Dr. Andrew Burke of
University of California at Davis; David Bauner of MTC AB, Sweden; Dr. Ata Khan of Carlton
University, Ottawa, Canada; Rogelio Sullivan of the Office of Advanced Automotive
Technologies, U.S. Department of Energy; and Dr. Feng An of Argonnes Center for
Transportation Research. Their constructive comments were very helpful. Thanks are also due
to our technical editor, Madonna Pence, and Argonnes Information and Publishing Division
personnel for preparing the manuscript and graphics. Although several persons helped make this
report possible, the authors are solely responsible for its contents.
viii
ix
Notation
4WD four-wheel drive
ac alternating current
ADVISOR Advanced Vehicle Simulator
ANL Argonne National Laboratory
APU auxiliary power unit
AWD all-wheel drive
AWMA Air and Waste Management Association
BMEP brake mean effective pressure
CAFE corporate average fuel economy
CARB California Air Resources Board
C
d
coefficient of drag
CH
4
methane
CO carbon monoxide
CO
2
carbon dioxide
C
r
coefficient of rolling resistance
CV conventional vehicle
CVT continuously variable transmission
dc direct current
DI direct injection
DISC direct injection stratified charge
DOE U.S. Department of Energy
EIA Energy Information Administration
EPA U.S. Environmental Protection Agency
EV electric vehicle
FHEV full hybrid vehicle
FTP federal test procedure
FTPRP Federal Test Procedure Revision Project
FUDS federal urban driving schedule
HC hydrocarbon
HEV hybrid electric vehicle
HEVCOST Hybrid Electric Vehicle Component Sizing and Vehicle Cost (model)
HEVGrid grid-connected hybrid
HEVTA Hybrid Electric Vehicle Technology Assessment
HVAC heating, ventilation, and air-conditioning
ICE internal combustion engine
JPL Jet Propulsion Laboratory
kg kilogram
L liter
LDV light-duty vehicle
MHEV mild hybrid vehicle
x
mph miles per hour
N
2
O nitrous oxide
NCHRP National Cooperative Highway Research Project
NiMH nickel metal hydride
NO
X
nitrogen oxides
NREL National Renewable Energy Laboratory
NVH noise, vibration, and harshness
NYCC New York City driving cycle
OBPU onboard power unit
OEM original equipment manufacturer
OTA Office of Technology Assessment
OTT Office of Transportation Technologies
PNGV Partnership for a New Generation of Vehicles
PU power unit
R range
REP05 Representative No. 5 driving cycle
RFG reformulated gasoline
SI spark ignition
SO
2
sulfur dioxide
SOC state of charge
SOHC single overhead cam
TA technology assessment
TDI turbocharged direct injection
ULEV ultra-low emission vehicle
ULS ultralightweight steel
US06 U.S. No. 6 driving cycle
VOC volatile organic compound
VVC variable valve control
W watt
W/h watt per hour
Z60 0-60 mph acceleration
ZEV zero emission vehicle
1
Summary
Introduction
This report presents the interim results of the Argonne National Laboratory (ANL) Hybrid
Electric Vehicle Technology Assessment (HEVTA). Hybrid electric vehicles, or HEVs, are
vehicles with drivetrains that combine an electric drive (including electric motor and some form
of electricity storage) with a refuelable power plant (e.g., an internal combustion engine). This
combination is capable of significantly improving vehicle efficiency (see Box 1). Toyotas Prius
and Hondas Insight cars are both HEVs, the first commercial examples.
A full technology assessment (TA) examines the effects on society of introducing a new
technology or expanding the use of an existing technology, including a full range of costs and
benefits. This assessment focuses particularly on the energy impacts and costs of HEVs. The
focus is on the individual vehicle. We do not attempt to project the impacts of alternative
scenarios when large numbers of HEVs gradually penetrate the vehicle fleet.
Box 1. How a Hybrid Drivetrain Saves Energy
The primary goal of using a hybrid drivetrain is saving energy. By using an electric motor
and battery to augment a vehicles engine, energy can be saved by several means:
Regenerative braking. The motor is used in generator mode to brake the vehicle,
generating electricity to be stored in the battery and thus recapturing some of the energy
normally lost as heat in conventional brakes.
Engine downsizing. The power added by the electric drive may allow the engine to be
downsized, maintaining most engine operation at a higher percentage of rated torque, which
is generally more efficient.
Idle-off. Use of the motor as a starting motor, with the battery available to run the
accessories, allows the engine to be turned off during stops (and perhaps also during braking
and coasting), saving energy that would otherwise be lost during these events.
Electric launch. In some configurations, the motor alone can be used to accelerate from a
stop, avoiding a driving mode where the engine in a conventional drivetrain would be
particularly inefficient.
The HEVTA uses two primary models:
1. ADVISOR (Advanced Vehicle Simulator), built by the National Renewable Energy
Laboratory, estimates the second-by-second energy use and emissions of a vehicle over
a defined driving cycle. Later versions of ADVISOR also allow the power capacity of
the drivetrain components to be defined from a set of performance requirements, e.g.,
0-60 mph acceleration (Z60) times. ANL has also built a performance model that sizes
the drivetrain based on performance requirements; this model is embedded in
HEVCOST (Hybrid Electric Vehicle Component Sizing and Vehicle Cost model).
2
2. HEVCOST first sizes the components of the vehicle, then estimates its retail price by
individually pricing the drivetrain components in an HEV that are not shared with a
similarly performing conventional vehicle (CV), e.g., battery, engine, and other
drivetrain components, using equations based on their power and energy requirements,
and adding these prices to a pre-established price for the common components. In other
words, the model assumes that hybrids and CVs are essentially identical except for
some minor structural additions to the hybrid to accommodate the electric drive, plus
the drivetrain itself.
Description of the Vehicles Evaluated
The vehicles are midsized passenger cars, with the interior space of such cars as the
Chrysler Intrepid and Ford Taurus and the external dimensions of the slightly smaller Chrysler
Cirrus and Chevrolet Malibu. The current versions of the Intrepid and Taurus weigh about
1,418 kg (3,125 lb); the 2010 CV versions will weigh about 1,225 kg (2,767 lb). The gliders
(bodies without drivetrains) are projected to weigh about 922 kg (2,029 lb) in 2010. The HEV
versions will have gliders that are 5% heavier to account for structural reinforcement needed to
accommodate heavier drivetrain components.
The CVs and HEVs share all nondrivetrain components,
1
including basic body structure,
tires, and aerodynamic shape. Four types of hybrid design are examined:
1. Parallel grid-independent full hybrid (FHEV), with engine sized to maintain a
specified constant gradeability and battery/motor sized to satisfy (in concert with the
engine) the Z60 requirement. Grid independence means that all energy requirements
are satisfied by fuel carried onboard; batteries are recharged by electricity from engine-
driven generation and regenerative braking.
2. Parallel grid-independent mild hybrid (MHEV), with engine power midway
between full hybrid and CV and battery and motor sized to satisfy the Z60 requirement.
3. Series grid-independent FHEV, with engine and electric motor (continuous rating)
sized for gradeability, battery sized to provide boost for acceleration requirements, and
electric motor (maximum power) sized for full acceleration power requirements.
2
4. Parallel and series grid-dependent hybrids, designed to operate part of the time as
full electric vehicles, so battery and electric motor are sized to satisfy a separate
electric vehicle (EV) acceleration requirement, and the engine is sized to provide the
minimum specified gradeability.
1
Except for the small weight gain of the HEV body.
2
In other words, there are two separate sizing criteria for the electric motor. Although the gradeability
criterion requires less power than the acceleration criterion, the motors continuous power rating will be
lower than its peak rating. Therefore, actual motor size may be determined by either the acceleration or
the gradeability criterion, depending on motor design and the gradeability and acceleration criteria used.
3
The HEVs examined use simple control strategies. Parallel vehicles use the motor to launch
the vehicle and provide all torque below a specified launch speed, with the engine turned off.
Above that speed, the engine is turned on and used as a load follower, generally satisfying all
torque requirements up to the engines maximum capability. The exception occurs if torque
demanded falls below a set limit. For example, during a downhill coast or during braking, the
engine shuts off. When the engine is turned on, the motor assists the engine only if the desired
torque is higher than the engines maximum capability (e.g., during high acceleration). During
braking, the motor provides all braking force up to its maximum capacity, feeding the electricity
generated to the batteries; the mechanical brakes provide any additional braking force needed.
Finally, if the battery state of charge (SOC) goes below a specified lower bound, the engine is
designed to increase its torque output above the level demanded by the tractional load and
accessory power load, with the excess torque passing through the motor (acting as a generator) to
charge the battery.
Results
If compared to recent widely publicized studies that evaluate hybrid vehicle technology, this
one is unique with respect to the variability in types of hybrids, potential levels of performance,
and variety of driving cycles evaluated. It is narrow in the sense that it emphasizes cost only,
ignoring issues of emissions of criteria pollutants and the rich variability of consumer
preferences. Thus, it evaluates only the critical aspect of the potential marketability of hybrids.
The results are encouraging enough to recommend that examinations of their value for purposes
in addition to saving fuel for the owner are important, and could make a significant difference to
market success. For example, there might be a change of consumer valuation of a capability for
all-electric operation in an emergency, in light of the terrorism of September 11. Thus, this study
has some strengths and weaknesses that readers should be aware of. Wider reading on the
attributes of hybrids is recommended, but this study does complement others available.
With regard to the focus of this study costs the findings imply that research to reduce
dollar cost per unit of performance of electric drive components remains highly desirable. Thus,
the aggressive research goals of DOE which if successful would drop costs to a level where
hybrids would be attractive on a net cost basis only remain important. In our analysis, we do
not see such costs naturally emerging from market forces. Without any cost reduction
breakthroughs, we estimate that parallel mild hybrids should be within striking distance of
widespread marketability after several years of experience in producing the vehicles, and after
production and sales reach hundreds of thousands of vehicles. It is easy to imagine niche
market success of such hybrids based on factors other than cost that are important to consumers
and society. However, it is difficult to imagine widespread displacement of the conventional
vehicle at the costs that we estimate, unless (as we discuss in the report) fuel prices rise
considerably from levels of the last decade, and remain high. The latter alternative is less
desirable than achievement of cost-for-performance breakthroughs through continuing research
and development of drivetrain components. Specifics on our results follow.
The ADVISOR model was used to evaluate the vehicles performance over seven driving
cycles, ranging from the stop-and-go NYCC (New York City Cycle) to the fast and high-power
4
REP05 (Representative No. 5 Cycle), with year 2010 vehicle characteristics. The HEVCOST
model evaluated vehicle costs for years 2005-2020. Key results are as follows:
1. HEVs achieve the largest percentage gain in fuel economy over CVs on the slowest
driving cycles. This is not a surprising result, since the inefficiencies of slow, stop-
and-go driving high idling losses, high braking losses, highly inefficient engine
operation are exactly the losses hybrid drivetrains are designed to combat. The
implication of this conclusion appears to be that HEVs will be most advantageous in
dense urban environments. This conclusion must be tempered, however, by the
possibility that miles driven per vehicle will be lower in areas where traffic is slower; if
so (and available data imply that it is so), gallons of gasoline saved per vehicle per
year, the most relevant measure of benefit, will not vary as much across different
driving environments as implied by the percentage gain results.
2. Parallel HEVs are likely to be more efficient than series HEVs, with both having
fuel economy that is less sensitive to performance than a CV. The clearest
difference between the two types of hybrid is that the parallel HEVs engine drives the
wheels directly through the differential and transmission for much of its operation,
whereas the series HEVs engine must first generate electricity through the generator
and then drive the wheels through the motor. The latter is a less efficient route, given
existing electric drive component efficiencies. This could change, of course. The lack
of sensitivity of HEV fuel economy to required performance may be largely an artifact
of the design rule used in this study for FHEVs; a faster Z60 time is obtained by
increasing the power capacity of the electric motor and battery rather than that of the
engine, whereas vehicle efficiency is quite dependent on engine rated power relative to
average load. Generally, the higher the ratio of rated power to average power, the less
efficiently the engine operates during normal driving.
3. Hybrids mpg improvement over CVs increases as the acceleration times go down.
The obvious result of HEV fuel economy insensitivity to Z60 time is that moving
toward faster vehicles will tend to increase the efficiency advantage of the hybrid
drivetrain, since CV fuel economy is very sensitive to engine power requirements.
However, it is important to stress that the faster HEV will retain its limited
gradeability and towing capability, while the faster CV will increase its previous
advantage over the HEV in these performance factors.
4. Parallel hybrids are cheaper than series. The crucial cost difference between series
and parallel hybrids is caused by the electric motor the parallel HEVs motor
provides boost power to supplement the engine, while the series HEVs motor must
provide all needed tractive power and thus must be much more powerful (and more
expensive). The parallel HEV requires a more expensive automotive (multigear)
transmission because the engine drives the wheels, whereas a series HEV requires at
most a two-gear transmission (and possibly just one), because the motor, which solely
drives the wheels in this configuration, can provide high torque through a wide speed
range. Nevertheless, the motor is a very expensive component, and its cost far
outweighs the effect of the transmission difference.
5
5. Demanding better acceleration performance is expensive. The U.S. automobile
market has been steadily moving toward lower and lower Z60 times. This is expensive
in a CV, but is considerably more expensive on the order of twice as expensive to
move from a 12-second Z60 to an 8-second Z60 for a parallel HEV, using this studys
design rule (engine sized for gradeability, battery/motor for Z60). This observation led
us to add mild hybrids (hybrids with more engine and less motor and battery) to the
analysis, because we concluded that the trend to faster vehicles would drive designers
of HEVs in this direction.
6. HEV life cycle costs are extremely sensitive to assumptions, but they appear likely
to exceed CV life cycle costs unless fuel prices rise to well over todays levels.
Electric drivetrain components are extremely expensive, but prices should be reduced
over time with learning and higher production rates. Thus, breakeven fuel prices
prices for which CV and HEV life cycle costs would be even will come down over
time. However, our best guess at component prices yields very high breakeven gasoline
prices for our hybrids about $2.50/gallon for a year 2020 parallel hybrid with
8-second Z60, and considerably higher for lower performing vehicles and for earlier
years. Prospects look better for mild hybrids (breakeven prices below $2.00/gallon by
2020 for an 8-second Z60 mild hybrid), but lower performing vehicles have breakeven
prices well above $2.00/gallon. We note that these estimates are for the average
U.S. driver. For those hybrid vehicle owners who spend more hours driving than the
average, the breakeven point will be a bit lower.
7. In comparison to CVs, grid-connected HEVs (HEVGrids) can achieve large
reductions in oil use, and lesser but significant reductions in greenhouse emissions
and total energy use, but at very high purchase and life cycle costs. On the other
hand, grid-connected HEVs do not cost a great deal more than grid-independent HEVs
(if the latter must match their maximum acceleration performance with their engines
on), yet they use substantially less oil and result in a moderate reduction in greenhouse
emissions. Grid-connected HEVs are attractive conceptually because they are capable
of operating part-time as EVs and can refuel a portion of their travel from the electric
grid, which offers significant air quality benefits, as well as the potential to capture
large savings in oil use, and can significantly reduce overall energy use and greenhouse
emissions. Our estimates imply, for example, that a grid-connected HEV with 33-mile
range might use one-third as much gasoline as a CV that matched its hybrid-mode
acceleration performance (that is, with its engine on), and about half as much as a CV
matching its all-electric acceleration performance. However, these benefits come at a
high price, both in initial investment and in life cycle costs. Comparing the same
vehicles, we estimate the CV to be from 20 to 24% cheaper than the comparable
HEVGrid in 2020, with life cycle costs lower by 5 cents or more per mile in 2020.
8. Long-run (i.e., 2020) incremental prices of the parallel HEVs examined in this
study vary from a 12% increment to a 32% increment, increasing as one proceeds
from a mild hybrid to a full hybrid to a grid-connected hybrid. The estimates for
parallel hybrids range from 12 to 14% for the three MHEV cases, from 16 to 19% for
6
the FHEV cases, and from 25 to 32% for the HEVGrids with all-electric range from 23
to 33 miles.
The percentage increase estimates in this study may be compared to those developed in
the recent study by Weiss et al. (2001), done at MIT, and by Graham et al. (2001),
under EPRI management. The one comparable case from the MIT study is an MHEV.
The estimated increase in price is 9%, less than estimated in this study. The base
method price increases from the EPRI study for the two FHEV cases examined are 20
and 21%. Though published earlier, the price estimates in the EPRI study were
completed later than for this study. ANLs HEV cost estimation team worked with the
EPRI study group to generate price increase estimates that reflect revised electric drive
costs from those in this study. More optimism about component characteristics was
included. The ANL method price increase estimates for the two EPRI FHEV cases
were 12 and 13%, well below the 16 to 19% estimated here. The largest difference
resulted from more optimism concerning battery costs. The EPRI base method price
increase estimates for the hybrid with about 20 miles of all-electric range was 32% at
the top end of this studys estimate of 25 to 32%. The revised ANL method gave
21% for this EPRI case.
Compared to the one MHEV case in the MIT study, this studys MHEV price increase
estimates are a bit higher. Compared to the two comparable FHEV and one
comparable HEVGrid cases from the EPRI study, the estimates in this study are a bit
lower. The revisions of the ANL method initiated in this study, and included in the
EPRI study, are more optimistic than this studys estimates. A significant part of the
difference arose from observation of rapid improvement of components used in
marketed hybrids. This rapid improvement, occurring over less than two years,
suggests that this study may be unduly pessimistic, at least about the rate of cost
reduction that is possible for hybrid vehicle components.
The key hybrid components battery (or other storage device), electric motor, and
controller are all under continuing development. DOE has established aggressive
goals to reduce energy storage costs to levels well below those used in this study.
Potential exists to substantially reduce motor and power electronics costs as well.
Although the probability of achieving very high levels of cost reduction is not clear,
success would significantly improve hybrids cost-effectiveness and their prospects for
mass market penetration.
Methodology Issues
The tools and methods available to the study team have limitations that the reader should
understand in order to judge the robustness of the results and to be guided in future analysis. Key
issues are:
1. Comparing CVs to HEVs. A key goal of the study is to compute unambiguously the
changes in energy use and emissions associated with introducing hybrid vehicles. The
two HEVs introduced to the U.S. marketplace both have changes from current vehicles
that go beyond substituting a hybrid drivetrain for a conventional one, e.g., the Insight
7
has a low-mass aluminum body, and the Prius has acceleration performance somewhat
below typical vehicles of its class. The HEVTA attempts to avoid comparing apples to
oranges by assuming that all physical aspects of a CV and HEV being compared, aside
from those dictated by hybridization, are identical. In other words, we assume that both
vehicles have identical bodies and identical tires except for a small weight gain by the
hybrid to beef up its structure for electric drive components.
3
Furthermore, we assure
the comparability of a measure of performance by designing the hybrid drivetrain to
match the Z60 capability of the CV to which it is being compared. However, there are a
number of problems with creating a perfect CV/HEV comparison, some of which are
associated with limitations to the tools we have, and some of which are associated with
inevitable differences between CVs and HEVs.
Gradeability. Most hybrids will have limitations to their gradeability because
of the combination of engine downsizing and sharp limits to battery capacity. If
long-term gradeability is limited to what can be achieved by the engine
operating alone (to avoid draining the battery), an HEV with Z60 capability
equal to that of a CV will have substantially less grade-climbing capability (as
well as towing capacity). Thus, an HEV/CV pair will rarely be performance-
equivalent.
Establishing equivalent acceleration performance. In our analysis, the Z60
of the vehicles is the crucial variable for achieving performance equivalence.
However, there is no industry standard for measuring Z60, and our research
shows that different test organizations have arrived at significantly different
Z60 estimates for identical vehicles, apparently because of different test
protocols. If anything, the sharp differences in CV and HEV operations make it
more difficult to match Z60 performance. In other words, there is no
unambiguous way to define a matched CV/HEV pair of equal Z60 performance.
Selecting a CV/HEV engine pair. In comparing a paired CV and HEV, it is
important to avoid attributing to hybridization any effects that are due to
technology differences that are irrelevant to hybridization. In some cases, this is
easy. For example, in comparisons between the Toyota Prius and Toyota
Corolla, the Prius has variable valve control (VVC) and the Corolla does not.
Because hybridization is not a prerequisite for VVC and does not even make its
use more logical, its positive effect on fuel economy should not be counted as a
hybridization benefit. In many cases, however, defining a paired set of CV/HEV
power plants is not easy. It is difficult even when the conceptual decision is
made to use two engines of identical technology but different power,
4
because it
is unlikely that two such engines with fuel consumption and emissions maps of
3
Admittedly, this is a simplistic assumption because it disregards variations in the amount of additional
structural materials caused by differences in hybrid performance and potential electric range.
4
Moreover, it is not clear that identical technology is in fact an appropriate goal for a CV/HEV engine
pair, since the demands on the engine are quite different in CV and HEV drivetrains.
8
the appropriate size will be available. The more practical method of analyzing
such an engine pair is to use mathematical scaling techniques to construct pairs
of engine maps from the maps of a single engine. Unfortunately, such scaling
may introduce significant errors into the maps.
2. Engine maps. Aside from the scaling problem discussed above, there are significant
problems with obtaining adequate engine maps for HEV engines. First, there are few
emissions maps available; in attempting to model vehicle emissions, we were forced to
rely on maps constructed from second-by-second emissions data that stretched the
measurement state-of-the-art. Second, available fuel consumption maps contain few
data at idle and none at negative load (during braking and coasting) despite the
substantial fuel consumed during these conditions on the standard driving cycles.
We filled in the fuel consumption maps with data we obtained from government
testing programs.
3. Emissions modeling. Emissions rates are extremely sensitive to engine and catalyst
changes that are difficult to model accurately. For example, small variations in engine
conditions can cause large changes in emissions, but available transmission models for
multiple vehicle applications are unlikely to accurately represent where on the engine
map the engine will operate as commanded by the transmission. In fact, the original
transmission model in ADVISOR moved the engine into regions of the engine maps in
which real-world engines would not operate. Also, tailpipe emissions are sensitive to
temperature and other variations in the catalyst that are hard to predict and likely to
vary with changes in drivetrain design strategy. In light of continuing skepticism about
the emissions results, as well as real-world results for the Prius and Insight hybrids that
contradict early (high) hybrid estimates from our early modeling work,
5
ANL has
chosen to report only energy use results in this report.
4. Accounting for costs and other design factors. ANLs analysis of equivalent CVs
and HEVs is based on a simple set of performance targets and design rules rather than
on the sophisticated design process that would actually occur in developing a
commercial vehicle. This basis will tend to distort some of the conclusions about the
way energy use varies with changes in performance targets and other factors, since the
simple design rules may cause changes in the vehicle configuration that are quite
different from the way real-world designers might react to the same target changes.
5. Accounting for technological change. Because this analysis attempts to compare CVs
and HEVs in the future (all the ADVISOR runs and the summary cost results presented
in Section 4 are keyed to the year 2010), ideally the analysis would account for likely
technological changes that would be expected to have occurred by that date. Although
the analysis incorporates projections about change for some technologies, such as
nickel metal hydride (NiMH) batteries, other technologies are left unchanged from
5
Santini, D.J., J. Anderson, J. He, S. Plotkin, A. Vyas, and D. Bharathan, 1999, Gasoline-Fueled Hybrid
vs. Conventional Vehicle Emissions and Fuel Economy, Ninety-Second Annual Meeting and Exhibition
of the Air and Waste Management Association, Paper 99-851, St. Louis, Mo., June.
9
their 1990s versions. For example, the difficulty of changing engine fuel consumption
maps to account for technological change has led ANL to leave the available maps
unchanged; the maps used are from a 1995 Saturn. One effect of this particular decision
may be to increase the apparent fuel economy boost of the HEV, because a more
modern engine might reduce some of the inefficiencies that the HEV drivetrain is
designed to target. On the other hand, we did not assume that hybridization might allow
use of a more efficient engine. In fact, the Prius uses a more efficient Atkinson-cycle
engine one that could not be used in a conventional vehicle. In addition, the great
uncertainty of the future rate of technological change could affect the accuracy of the
boost estimate in either direction. For example, the assumed year 2010 energy density
of the HEVs NiMH batteries is actually substantially below the actual energy density
of the recently introduced MY2000 Prius; technological progress has moved much
faster than we assumed when we began this analysis.
10
11
Section 1
Introduction
This report presents the results of the first phase of Argonne National Laboratorys (ANLs)
examination of the costs and energy impacts of light-duty hybrid electric vehicles (HEVs).
6
We
call this research an HEV Technology Assessment, or HEVTA. HEVs are vehicles with
drivetrains that combine electric drive components (electric motor, electricity storage) with a
refuelable power plant (e.g., an internal combustion engine). The use of hybrid drivetrains is
widely considered a key technology strategy in improving automotive fuel efficiency. Two
hybrid vehicles Toyotas Prius and Hondas Insight have been introduced into the U.S.
market, and all three auto industry participants in the Partnership for a New Generation of
Vehicles (PNGV) have selected hybrid drivetrains for their prototype vehicles.
The HEVTA is being conducted for the U.S. Department of Energy (DOE), Office of
Transportation Technologies (OTT), in support of its basic mission statement:
The Office of Transportation Technologies will work in partnership with the domestic
transportation industry, energy supply industry, and research and development
organizations to develop and promote user acceptance of advanced transportation
vehicles and alternative fuel technologies which will reduce oil import requirements,
and reduce criteria pollutant emissions and greenhouse gases. The Office will also
develop a strong transportation technology base to enable this industry to assure strong
competition in the domestic and world markets.
7
The study methodology incorporated the following goals:
A. An examination of HEV design options, on the basis of an extensive literature review,
discussions with analysts and component and vehicle developers, and technical
analysis.
B. Selection of a set of design options for analysis. In particular, our two DOE sponsors,
the Office of Transportation Technologies Planning and Assessment Group and its
Office of Technology Utilization, have different interests:
6
This analysis involves, as much as possible, constant technology and comparable performance
comparisons of HEVs and conventional vehicles (CVs). Argonne has made use of the Advanced Vehicle
Simulator (ADVISOR) model developed by the National Renewable Energy Laboratory (NREL) and,
throughout this project, with the participation of D. Bharathan of NREL, has worked with NREL to
improve ADVISOR, especially its simulation of CVs.
7
OTT website, www.ott.doe.gov/ottover.html
12
Planning and Assessment is interested in HEVs that closely resemble conventional
vehicles and will not rely on grid electricity; all of their energy is obtained from
their onboard fuel. These are the type of vehicles being designed by the PNGV.
Technology Utilization is interested in HEVs that obtain a portion of their energy
from the grid and probably will operate part of the time in a pure electric mode.
C. Simulation of the energy use and air emissions of vehicle operation. This task includes
development of necessary simulation models; selection of appropriate vehicle driving
cycles to simulate real-world driving; analyzing vehicle physical characteristics, such
as total weight, engine and battery size, and power (including development of models
to calculate these characteristics); and executing the simulation models. The key
simulation tool we use is the Advanced Vehicle Simulator, or ADVISOR model,
developed by the National Renewable Energy Laboratory. ADVISOR provides
estimates of a vehicles energy use and emissions for a range of driving cycles by
calculating the second-by-second forces on the vehicle and the power demanded from
each component of the drivetrain, from the tires to the engine, using maps of efficiency,
fuel use, and/or emissions for each component. ADVISOR is also capable of doing
forward calculations starting with throttle behavior and estimating the effect on each
drivetrain component in turn, to finally calculate the effect on vehicle movement.
D. Calculation of the purchase and life cycle costs of the vehicles analyzed, as well as
exploration of some of the design trade-offs implied by cost considerations.
As of the end of Phase I, we have succeeded in our original goals for A and B and made
substantial progress on goals C and D. In regard to goal C, we have found the task of vehicle
modeling to be more challenging than we had anticipated. We have worked to overcome a
number of problems; including (1) a scarcity of publicly available engine fuel economy and
emissions maps, with those available having limited or no data in important operating areas, and
(2) limits to the capabilities of the early versions of the ADVISOR vehicle simulation model we
began working with. Among other things, the ADVISOR version with which we began did not
accept engine maps with negative torque data, it did not inform the user when available engine
plus battery power could not match the requirements of the more vigorous driving cycles, and its
transmission model caused the engine to operate in an unrealistic fashion. We now are satisfied,
for the most part, with our ability to use the upgraded versions of ADVISOR to model vehicle
energy use, but we remain skeptical of our emissions results, and so we do not report them here.
For goal D, we have completed but not yet documented a cost and component sizing model,
HEVCOST (Hybrid Electric Vehicle Component Sizing and Vehicle Cost Model). We have used
the present version of HEVCOST to conduct several cost analyses of conventional and hybrid
electric vehicles for this report and will continue to update and modify it.
We intend this report to accomplish the following:
Present the preliminary results of our vehicle modeling and the necessary
background information related to the modeling exercise. Appendix A to
13
Section 1 describes the basic vehicle designs, component choices, and types of
driving we have examined in this phase of the study.
Describe the modeling improvements we have made, specifically, development of
the HEVCOST model that computes, in addition to cost, hybrid vehicle weight
and component power characteristics.
Describe what the study team has learned about hybrid vehicle modeling
methodology.
This report is organized as follows:
Section 2, Hybrid-Electric Vehicles: Theory and Design We discuss the energy
implications of hybrid systems, the basic types of vehicle designs, and the relationship between
design choices and vehicle performance.
Section 3, Methodology and Modeling Issues
Vehicle modeling We discuss in detail the models we use (the HEVCOST design model to
build the vehicle, ADVISOR to cross-check HEVCOST and to simulate vehicle operations)
and various methodology issues: choosing control strategies for the vehicles, problems
encountered in modeling vehicle operations and the work done toward achieving solutions, and
constructing a methodology to project the magnitude of vehicles use of electricity from the grid
when the option of ordinary refueling also exists and competes with electric refueling.
Estimating the electricity and gasoline use of grid-connected hybrids We describe the
assumptions and method used for estimating the split between fuel-driven and grid electricity-
driven miles for hybrids with grid-connection capability.
Section 4, Fuel Consumption and Cost Results (for both CVs and HEVs) We present a
definition of the scenarios analyzed, key caveats, and the vehicle modeling results.
In presenting our results, we have tried to be careful to present fair comparisons of
conventional and hybrid electric vehicles. In our experience, many reports describing the benefits
and costs of new types of vehicles, e.g., electric vehicles, have failed to account for important
differences between the new vehicles and conventional ones or have failed to configure the new
vehicles being analyzed in such a way that they fulfill the same functions as the conventional
vehicle. Appendix B to Section 1 describes some of the important concerns that arise in
comparing different vehicles.
14
15
Section 1 Appendix A
At What Types of Vehicles
Are We Looking?
In this phase of the HEVTA, we have examined a series of vehicles designed to be as
identical as possible from the drivers point of view. That is, to the extent possible, the driver
wouldnt know which vehicle he was driving except when he had to refill (or recharge) the
vehicle. In every example, the basic vehicle is a midsized passenger car that retains the same
basic platform (the car minus its drivetrain). That is, regardless of whether the vehicle is a
conventional vehicle or one of several different types of hybrids, and regardless of its
performance, the body structure is identical (exception: the hybrids 5% greater structural weight
accounts for added structure needed for batteries, etc.), the aerodynamic drag coefficient is
identical, the assumed accessory load is identical, and the rolling resistance coefficient is
identical. Only the drivetrain changes.
Furthermore, although CVs and hybrids will not perform identically, performance is held to
be identical or near-identical in some important ways. In particular, vehicles being compared will
have either identical Z60 times if they are independent of the grid, or identical Z60 times in at
least one mode of operation if the hybrids in the group are grid-dependent and will operate part-
time as electric vehicles. Although hybrids with downsized engines will have long-term
gradeability inferior to that of their matching CVs, the hybrids are held to minimum standards
developed for the Partnership for a New Generation of Vehicles 55 mph up a 6.5% grade for
20 minutes. Some of the hybrids evaluated here exceed this minimum considerably, but still are
not as capable as CVs.
The hybrid vehicles examined in this analysis use nickel metal hydride batteries whose
performance is based on existing batteries used in the Toyota Prius (Japanese market version)
and RAV-4, and a permanent magnet motor based on a Unique Mobility design.
The components of a hybrid drivetrain can be sized in a number of ways to satisfy
performance requirements. The baseline hybrid drivetrain in this analysis is sized by matching
engine power to that needed to achieve long-term gradeability, with the motor and battery then
sized to allow achievement of the Z60 requirement. This type of design is designated a full
hybrid in that it produces a vehicle with substantial electric boost. The engine is considerably
smaller than that of the equivalent CV, because the power required for the PNGV gradeability
limit is considerably less than acceleration power requirements; generally, the downsizing is of
the order of 40% or so for a 12-second Z60 requirement, and somewhat more for more
demanding Z60 requirements. We also examine a design with a larger engine and smaller battery
and motor, which is designated a mild hybrid.
16
The following types of vehicles are examined:
Conventional vehicles
Parallel and series grid-independent full hybrids
Parallel grid-independent mild hybrids
Parallel and series grid-connected hybrids
Because high power has been a valuable commodity in the U.S. automobile market, we examine
a range of Z60 times for the vehicles. For CVs and grid-independent hybrids we examine 12-, 10-
, and 8-second Z60s; for grid-connected hybrids we examine 12-, 14-, and 16-second Z60s in all-
electric mode, tested with HEV battery at 0.2 state of charge (SOC).
Hybrid vehicles may employ a variety of strategies to control the use of engine, battery, and
electric motor in powering the vehicle. Parallel hybrids, in particular, split the responsibility for
driving the wheels between the engine/transmission pathway and the battery/electric motor
pathway; furthermore, the engine can be used to recharge the battery. In the parallel hybrid
vehicles examined in this phase of the study, we assume that the vehicles are launched
electrically (with electricity from the battery) to a preset speed, with the engine off; the engine is
then turned on, and power responsibility is then shifted to it, with the motor available to provide
a boost if the power demanded exceeds the engines capability. The engine is automatically
turned off when the vehicle is braked or is at idle or coasting. Braking is accomplished with the
motor to the maximum extent possible, with electricity generated by regenerative braking fed to
the battery. Other strategies are possible, of course, and will be investigated as the study
proceeds.
17
Section 1 Appendix B
What Is a Vehicle?
It is not difficult to design a vehicle that attains a fuel economy of several hundred miles per
gallon; actually, a few thousand miles per gallon is possible under ideal conditions. However,
such vehicles have little or no utilitarian value more than likely, they will accommodate one
person (the driver) under extremely confined conditions (lying down inside a lightweight shell to
minimize aerodynamic drag and weight); they will have very low power in order to minimize
engine size and weight and to ensure that the engine operates generally at an efficient load; they
will therefore have minimal performance and, in fact, might have little or no hill-climbing
capability and low top speed; they will have no power accessories, and possibly minimal braking
capability; and they will offer little or no protection in a crash, again to minimize weight.
Although vehicles designed to garner fuel efficiency records may be an extreme example, their
example should serve as a warning that vehicle analysis should examine carefully whether
vehicle design options satisfy the range of criteria that define a vehicle acceptable to the current
or some future marketplace in terms of performance, safety, comfort, and other characteristics.
Vehicle criteria of importance to the marketplace include:
Performance range, acceleration capability, gradeability, load-carrying ability,
including towing ability, and top speed/cruising speed;
Emissions compliance with regulatory requirements;
Safety crashworthiness, handling/braking capability, protection from hazardous
materials, and compliance with regulatory requirements;
Moderate purchase cost and operating costs;
Convenience of refueling;
Reliability and cost of repair;
Comfort and convenience, including adequate passenger and storage space, heating and
cooling capacity, etc.
Moving to hybrid vehicles raises a number of issues with respect to these criteria:
Repeatability of performance. A hybrids reliance on a battery or other energy storage
system of limited capacity to supplement its onboard power unit (OBPU) when maximum power
is demanded implies that the vehicles performance capability may be reduced if the energy
storage is depleted or otherwise limited. For example, a hybrid may use its engine to satisfy high
power requirements of long duration (high speed cruise, ascent up long grades) and a battery to
add supplemental power for rapid acceleration. If a number of rapid accelerations is demanded in
quick succession, the battery may offer inadequate power because of charge depletion or
18
overheating. Some hybrid storage devices are especially vulnerable to charge depletion because
of their low storage capacity, e.g., ultracapacitors. Thus, a hybrid may offer differing levels of
protection against reduced performance caused by depletion of its stored energy, with different
implications for the required power capability of the OBPU (Tamor, M.A., 1996, Ford Motor
Company, unpublished material).
Full protection all operation can be supported by the OBPU acting alone (here, the
OBPUs power would have to match the vehicles maximum power requirement);
Steady-state protection all constant-speed operations (high speed cruise, very long
grades) can be supported by the OBPU, energy storage supports some transients that do
not last long, e.g., accelerating to highway speeds (OBPU power between 60 and 100%
of maximum required); both the Prius and Insight hybrids appear to meet this
requirement, although the Prius has a dashboard light that alerts the driver to a low SOC
in the battery;
Statistical protection performance failure can occur extremely rarely, or only in non-
normal operations (OBPU power between 30 and 60% of maximum);
8
and
No protection performance failure can occur, but vehicle will warn driver before it
happens, driver must adjust accordingly (OBPU power less than 30% of maximum)
(Tamor 1996).
According to Tamor, vehicle designers will likely avoid both full protection because it
negates much of the potential efficiency benefit of the hybrid (with no weight savings from the
OBPU and the added weight of energy storage and motor), and no protection because it is
unlikely to be acceptable to drivers. The market viability of statistical protection is not clear,
particularly for mass market vehicles.
Emissions performance. Aside from their potential to improve vehicle energy efficiency,
hybrid configurations have been advocated as a means of substantially reducing vehicle
emissions by eliminating or reducing engine excursions into load/speed regimes where emissions
are high, reducing engine size, providing time to electrically heat the catalyst before cold starts,
accommodating the use of OBPU technologies that are inherently low emitters, and electrifying
miles, as discussed elsewhere. Another potential emissions advantage, not foreseen prior to the
commercialization of hybrid vehicles, is rapid catalyst warm-up regardless of driving conditions
following start-up. This is characteristic of the new Prius hybrid. On the other hand, some
concerns were previously raised about possible emissions surges caused by multiple hot restarts
(for hybrids that turn the engine off at idle and low loads, or when the battery reaches a SOC
maximum). However, emissions tests have thus far failed to detect significant emissions surges
on restart.
8
Presumably, the vehicle would warn the driver in the rare event of a potential performance loss.
19
Although both Prius and Insight have excellent emissions performance, it is not clear that
the nature of the drivetrain will be the deciding factor in this performance (vehicles with
conventional drivetrains have been shown to be capable of extremely low emissions). Hybrids
may, of course, be no less invulnerable to control system malfunctions than conventional
vehicles, although some hybrid advocates have proposed that the reduced speed/load range
(assuming the energy storage will absorb most transient power requirements) of the OBPU will
reduce malfunctions.
9
Performance in electric vehicle (EV) mode. Some hybrid configurations can operate for
several miles using only their stored energy, with the OBPU off. As discussed elsewhere, this
mode might be considered standard operating procedure for the beginning of all trips, or as a
special mode for use in EV-only zones or during a gasoline/diesel fuel emergency. The storage
device will have to be extremely large or have very high specific power, however, to be able to
satisfy vehicle performance criteria expected under full power (OBPU plus storage), so the
development of high-specific-power batteries may be crucial to this type of design. The market
acceptance of a reduced power mode (if an ultra-high-specific-power battery is not available) is
not certain, though clearly it will depend on the circumstances that cause the EV mode to be
selected and the precise nature of the performance compromises.
Safety concerns. The primary safety concerns with hybrid drivetrains are associated with
the specific characteristics of any new drivetrain components that may be added (e.g., high
temperature materials in certain battery types, volatility of new fuels, high pressure storage
tanks,
10
etc.). In general, however, there appear to be few safety concerns that are unique to
hybrid drivetrains.
The PNGV has established a variety of vehicle targets associated with fuel economy and
emissions, performance, vehicle dynamics and structure, and other characteristics. A portion of
these is presented in Table 1. It is possible that diverse design efforts will tend to coalesce around
these targets, because they represent characteristics that are proven in the marketplace. That is,
they are designed to match the overall performance of successful conventional vehicles, though
with some compromises associated with the characteristics of hybrid drivetrains. It is important
to note, however, that the characteristics of the current light-duty fleet have evolved as an
interaction between consumer needs and the characteristics of conventional internal combustion
engine (ICE) drivetrains. Designers of hybrid drivetrains can squeeze the integrated
physical/operating design of their drivetrains to match those of current ICE drivetrains, or they
might instead choose, at some market risk, to accept the differences between hybrid and
conventional vehicles and design to the hybrids strengths, accentuating short-term acceleration,
9
Although the hybrids engine will be on fewer hours than a conventional vehicles engine, and will
experience fewer transients, it will be turned on and off considerably more often. The effect on the
potential for emission control malfunctions needs to be determined.
10
Although new fuels may be used with conventional drivetrains (i.e., the safety concerns are not specific to
hybrids), their use in hybrids may be more likely because hybrid configurations can better accept power
plant/fuel combinations that may be ill-suited to conventional use because of their poor static and/or
dynamic matching with vehicle power requirements.
20
for example, at some cost in long-duration gradeability or other characteristic. Only time and
market experience will tell which design direction is most likely to succeed.
Table 1 Selected PNGV Performance
Guidelines
Vehicle Performance at 77% State of Charge
Peak acceleration 17 ft/sec
2
0-60 mph accel 12 seconds
0-85 mph accel 25 seconds
40-60 mph pass time 5.3 seconds
Maximum speed (minimum)
continuous
85 mph
Gradeability, forward &
reverse launch
30% at gross vehicle weight
(GVW)
Gradeability, forward 6.5% @55 mph for 20 minutes
(GVW w/full accessory load)
21
Section 2
Hybrid-Electric Vehicles:
Theory and Design
2.1 Background
Hybrid electric vehicles (HEVs) represent a cross between a conventional automobile and
an electric vehicle. They combine an electric drivetrain, including battery or other energy storage
device, with a quickly refuelable power source such as a gasoline or diesel engine, fuel cell, or
gas turbine. This refuelable power source, called an onboard power unit or OBPU, generally can
recharge the storage device and may drive the wheels either directly (as can the electric motor)
through a mechanical drivetrain, or indirectly by providing electric power to the motor. If the
refuelable source can drive the wheels directly (in parallel with the electric motor), this is a
parallel hybrid; if the refuelable sources function is to supply electricity to the motor (and to
recharge the storage device), with only the motor driving the wheels, this is a series hybrid.
The hybrid concept is by no means a new one. A Woods gasoline-electric coupe selling for
$2700 was introduced in 1916. This early hybrid electric vehicle combined a four-cylinder
gasoline engine with an electric motor and a battery half the size of those used in contemporary
electric cars. Like todays Toyota Prius, it used its electric drive for low speeds and the gasoline
engine for higher speeds and to recharge the batteries (Schiffer 1994). Sixty years later, the
Electric and Hybrid Vehicle Research, Development, and Demonstration Act of 1976 authorized
the U.S. Department of Energy (DOE) to encourage and support research and development of
hybrid vehicles (U.S. DOE 1994). DOE subsequently sponsored major studies beginning in the
late 1970s with teams led by the Jet Propulsion Laboratory (JPL) and the Aerospace Corporation.
These studies included both in-depth computer simulations of hypothetical vehicles and the
design, construction, and testing of prototype vehicles (Burke 1992). The JPL/GE Near-Term
Hybrid Vehicle program, which ran from 1978-82, developed a working vehicle, called the
HTV-1, with a parallel hybrid drivetrain (Burke 1992).
11
The target performance of the vehicles contemplated and built in these early programs was
between that of conventional gasoline vehicles and electric vehicles of the time, and comparable
to the relatively low-powered diesel vehicles then available. The Department of Energy has since
embarked on programs that aim to produce vehicles with performance more likely to compete
with conventional gasoline-powered vehicles. The Department has pursued both its own research
program, the Electric and Hybrid Vehicles Program,
12
and a joint government/industry program,
the Partnership for a New Generation of Vehicles (PNGV), initiated in 1993 by the
11
At the time this vehicle was built, the available motors, controllers, and inverters were much heavier than
those available today. The 80-90 kW that would have been demanded with a series configuration
which demands considerably more power in the electric drivetrain than does a parallel configuration
would have been difficult to package.
12
Established in response to the Electric and Hybrid Vehicle Research, Development, and Demonstration
Act of 1976.
22
U.S. government and the United States Council for Automotive Research (representing Chrysler,
Ford, and General Motors). The PNGVs primary goal is to develop a fully competitive family
car with fuel economy up to 3 times current levels; the PNGV research team has settled on a
hybrid drivetrain as an integral part of this vehicle.
The DOE and PNGV have been driven largely by concerns about oil savings. Interest in
hybrid vehicles stems also from the State of Californias Zero Emission Vehicle (ZEV)
requirements, which demand that automakers doing business in California begin producing
substantial quantities of ZEVs within the next decade.
13
The ZEV requirements are driven by air
quality concerns, and the California Air Resources Board (CARB) proposes allowing hybrids to
attain partial ZEV credits (that is, each hybrid vehicle will be counted as a fraction of one
electric vehicle). The previous version of the ZEV proposals were as follows (Air Resources
Board 1999):
The vehicle must first meet Super ultra low emissions vehicle (ULEV) emissions
requirements
14
at 150,000 miles, to establish a base 0.2 allowance
A qualifying vehicle accumulates additional allowance above 0.2 if it:
Has a zero emission range of 20 miles (0.3 allowance) to 100 miles or more
(0.6 allowance)
Within its zero emission range, employs a strategy to maximize grid recharging
(0.1 allowance, but total ZEV range allowance is limited to 0.6)
Uses fuel with inherently low fuel cycle emissions (up to 0.2 allowance)
Employs advanced ZEV componentry (e.g., advanced batteries) but doesnt
otherwise qualify for zero emission range credits (0.1 allowance).
Under this system, a Super ULEV hybrid with a 20-mile ZEV range and some mechanism to
maximize grid recharging would earn an allowance of 0.2+0.3+0.1 = 0.6. The ZEV system is
currently being modified, with potential provisions including a reduction in the minimum electric
vehicle (EV) range for a grid-connected hybrid, and initial award (e.g., for 2002) of 1.8 credits
for a 20-mile range grid-connected hybrid.
15
13
The original requirements demanded that 2% of vehicles offered for sale would be ZEVs by 1998, with
the percentage rising to 10 by 2002. The 1998 requirement was eliminated, and the 2002 requirement is
being reconsidered.
14
Including zero evaporative emissions, meeting onboard diagnostic requirements for 150,000 miles, and
extending emission controls warranty (including the battery warranty) to 150,000 miles.
15
Personal communication, Chuck Shulock, Air Resources Board, 8/22/2001.
23
More recently, interest in hybrids has also been driven by concerns about global warming.
Hybrids increased fuel efficiency and their enhanced potential to use alternative fuels (see
discussion in Section 2.3) may yield significant reductions in emissions of carbon dioxide, the
key greenhouse gas.
2.2 Energy Use in Conventional Vehicles
To understand how a hybrid may save energy, it is necessary first to examine how
conventional vehicles use energy:
In order to maintain movement, vehicles must produce power at the wheels to overcome
aerodynamic drag (air friction on the body surfaces of the vehicle, coupled with pressure forces
caused by the air flow), rolling resistance (the resistive forces between tires and the road
surface), and any resistive gravity forces associated with climbing a grade. Further, to accelerate,
the vehicle must overcome the natural resistance of its mass to acceleration, called inertia most
of the energy expended in acceleration is then lost as heat in the brakes when the vehicle is
brought to a stop.
16
And in addition, the vehicle must provide power for accessories such as
heating fan, lights, power steering, and air conditioning.
Finally, a vehicle will need to be capable of delivering power for acceleration with very
little delay when the driver depresses the accelerator, which may necessitate keeping the power
source in a standby (energy-using) mode.
A conventional engine-driven vehicle uses its engine to translate fuel energy into shaft
power, directing most of this power through the drivetrain to turn the wheels. Substantial
amounts of energy are lost along the way. Within the engine, for example, moving parts
especially pistons, crankshaft, and valves create friction; there are a number of aerodynamic
and fluid drag losses (pumping losses) because air must be pumped through air cleaner, intake
manifold, valves, and exhaust system, and, most importantly, because spark-ignition engines
reduce their power output by throttling the air flow which causes additional aerodynamic losses
that are very high even at light loads. Much of the heat generated by combustion cannot be used
for work and is wasted, both because heat engines have theoretical efficiency limits, and because
attaining even these limits is impossible because some heat is lost through cylinder walls before
it can do work, and some fuel is burned at less than the highest possible pressure (OTA 1995).
Fuel is also burned while the engine is experiencing negative load (during braking) or when
the vehicle is coasting or at a stop, with the engine at idle.
Although part of engine losses would occur under any circumstances, part occur because in
conventional drivetrains, engines are sized to provide very high levels of peak power for the
acceleration capability expected by consumers
17
perhaps 10 times the power required to cruise
16
Some of this energy is also lost as aerodynamic drag and rolling resistance losses.
17
For a mass market family car in todays market, the ability to accelerate from 0 to 60 mph in 12 seconds
may be a minimum capability.
24
at 60 mph but are operated at most times at a small fraction of peak power where they are quite
inefficient. Having such a large engine also increases the amount of fuel needed to keep the
engine operating when the vehicle is stopped or during braking or coasting, and increases losses
due to the added weight of the engine, which increases rolling resistance and inertial losses. Even
gradeability requirements (example: 55 mph up a 6.5% grade) require only about 60 or 70% of
the power needed to accelerate from 0 to 60 mph in under 12 seconds.
18
Multispeed
transmissions allow the engine to operate within a fairly narrow speed regime across the range of
vehicle speeds, allowing the engine to stay in the most efficient parts of the engine map more of
the time than would be the case with fewer gears but at the cost of losses in the transmission
itself.
Figure 2-1 shows how fuel energy is translated into work at the wheels for a typical midsize
vehicle in urban and highway driving as represented by U.S. Environmental Protection Agency
(EPA) driving cycles.
19
The part of the figure at the right represents the tractive losses incurred
by the vehicle as a fraction of the total fuel energy. Some highlights of the figure are:
At best, only one-fifth of the fuel energy reaches the wheels and is available to overcome
the tractive forces, and this is on the highway cycle when idling losses are at a minimum,
braking loss is infrequent, and shifting is far less frequent.
Braking and idling losses are extremely high in (EPA cycle) urban driving and even
higher in more congested driving, e.g., within urban cores during rush hour. Braking loss,
that is, the shedding of the kinetic energy of motion through heat generated by the
brakes, represents 46% of all tractive losses in urban driving. Idling losses represent
about one sixth of the fuel energy on this cycle.
Losses to aerodynamic drag, a fifth or less of tractive losses in urban driving, are more
than half of the tractive losses during highway driving.
18
Depending on relative inertia and other losses and on the vehicle load specified for both acceleration and
gradeability requirements.
19
Note that the box labeled kinetic represents the potential energy built up in accelerating the vehicle.
This energy is lost as heat in braking and in the rolling and aerodynamic losses that occur as the vehicle
decelerates. In other words, regenerative brakings target for recapture is not the full potential energy
of the vehicle, because rolling resistance and aerodynamic drag will always be responsible for part of the
braking forces when the vehicle is slowed. Another factor that can limit the amount of kinetic energy
recoverable is the number of axles driven. For safety, both axles must brake, so only four-wheel drive
cars can target the full braking energy. Front-wheel drive cars fare a bit better than rear-wheel drive cars
because, in most cars, the front brakes do the larger share of the braking.
25
Figure 2-1 Energy Flow (from fuel energy to power at the wheels) for a Midsize
Automobile
2.3 Energy Savings Potential of Hybrid Drivetrains
In terms of overall energy efficiency, the conceptual advantages of a hybrid over a
conventional vehicle are:
Regenerative braking. A hybrid can capture some of the energy normally lost as heat to
the mechanical brakes by using its electric drive motor(s) in generator mode to brake the
vehicle;
More efficient operation of the onboard power unit (OBPU), including elimination
(or sharp reduction) of idle. A hybrid can avoid some of the energy losses associated
with engine operation at speed and load combinations where the engine is inefficient by
using the energy storage device to either absorb part of the OBPUs output or augment it
(or even substitute for it), allowing it to operate only at speeds and loads where it is most
efficient. When an HEV is stopped, rather than running the engine at idle, where it is
extremely inefficient, the control system may either shut off the engine, with the storage
device providing auxiliary power (for heating or cooling the vehicle interior, powering
headlights, etc.), or run the engine at a higher-than-idle (more efficient) power setting
and use the excess power (over auxiliary loads) to recharge the storage device. When the
vehicle control system can shut the engine off at idle, the drivetrain can be designed so
that the drive motor also serves as the starter motor, allowing extremely rapid restart due
to the motors high starting torque.
Smaller, lighter OBPU. Because the storage device can take up part of the load, a
hybrids OBPU can be downsized. In some cases, the OBPU can be sized for the highest
26
sustained loads, not for (higher) short-term acceleration loads. Consequently, the OBPU
can have a significantly lower power rating than the engine in a conventional vehicle.
This allows the engine to be run at a higher fraction of its rated power, generally at
higher efficiency, during most driving. Also, the reduced engine weight is mildly
beneficial to fuel economy.
20
Potential for alternative OBPU technologies. Conventional drivetrains use piston
engines because such engines do a fair job, when coupled with multispeed transmissions,
of efficiently matching vehicle load requirements (static matching), and an excellent job
of rapidly boosting or reducing power to match the vehicles changing loads (dynamic
matching). Most alternative power sources do not share these matching characteristics.
For example, turbine engines are extremely inefficient at the low loads typical of normal
driving, and are slow to respond to changing load, but they can burn a wide variety of
fuels and are small and lightweight in relationship to their power output. In a hybrid
drivetrain, the storage device could assume the load-following role, compensating for the
turbines slow response, and allow the turbine to operate in a high output, efficient mode
by absorbing its excess energy output. Remaining roadblocks for turbines include high
NO
X
emissions at high loads and the need for further development of the ceramic
materials used to increase their efficiency.
There are counterbalancing factors reducing hybrids energy advantage, including:
Potential for higher weight. Although the fuel-driven energy source on a hybrid
generally will be of lower power and weight than the engine in a conventional vehicle of
similar performance, total hybrid weight is likely to be higher than the conventional
vehicle it replaces because of the added weight of the storage device, electric motor(s),
and other components. This depends, of course, on the storage mechanism chosen, the
vehicle performance requirements, and so forth. The hybrid configurations examined in
this report [conventional internal combustion engines (ICEs) and nickel metal hydride
batteries] were consistently heavier than their conventional vehicle (CV) counterparts.
Electrical losses. Although individual electric drivetrain components tend to be quite
efficient for one-way energy flows, in many hybrid configurations, electricity flows back
and forth through components in a way that leads to cascading losses. Further, some of
the components may be forced to operate under conditions where they have reduced
efficiency. For example, like ICEs, most electric motors have lower efficiency at the
low-speed, low-load conditions often encountered in city driving.
21
Without careful
component selection and a control strategy that minimizes electric losses, much of the
theoretical efficiency advantage often associated with an electric drivetrain can be lost.
20
Although, as discussed below, this weight benefit will likely be more than offset by the weight of the
electric drivetrain components.
21
With parallel hybrids, the motor may be quite small, so that low-speed low vehicle load conditions may
actually be fairly high load when measured against the motor power rating.
27
2.4 Emissions Trade-Offs
Because hybridization changes the way a vehicles engine is operated, emissions
performance should change as well. For the most part, hybrid features seem to offer several
advantages in controlling emissions: the power available from the electric motor can help keep
engine operation away from high-emission regions; engine downsizing will mean that normal
vehicle loads will be at a higher percentage of engine rated power (which is generally more
efficient and less polluting) than in a conventional drivetrain; the power absorbing capability of
the battery allows the engine to be run in a startup mode that shortens the cold start period,
reducing starting emissions; and electric heating of the catalyst is more feasible than in a
conventional drivetrain. One early emissions concern with hybrid operation was the possibility of
emissions spikes from multiple hot and warm restarts. Many of the hybrid powertrain control
strategies discussed by researchers envision the engine being turned off and restarted several
times during a trip. For example, some hybrids will have electric launch, with the engine turned
on only when the vehicle reaches a predetermined speed or when total vehicle load reaches a
predetermined level; this strategy is used in the Prius. Also, both Prius and Insight, and probably
most future hybrids, will use idle off, where the engine is shut down during a stop (it can also
be shut down during braking and coasting). The engine is then restarted when positive torque is
demanded (or when a minimum speed is attained after a stop). Thus far, the excellent emissions
performance of both Prius and Insight imply that the concern about emissions spikes may not be
a problem, although it is worth further study.
22
It is worth noting that the high level of control of
engine operation available in a hybrid can also help prevent emissions spikes, for example, the
Insights control system uses the electric motor to bring engine rpm up to operating levels before
injecting fuel during an engine restart, reducing the possibility of any emissions spike.
23
In hybrids with substantial all-electric range, the capability to operate in a zero emission
mode yields emissions advantages whether or not the battery is recharged by the grid or by the
vehicles own engine. The ZEV capability can allow dramatic changes in the amount, timing, and
location of emissions that might be used to advantage to improve air quality. The key advantages
are locational and temporal:
Locational. This type of hybrid can operate, at least for a time, in an electric vehicle
(EV) mode within a dense urban core. With grid recharge, the recharge electricity may
be generated at a power plant quite distant from the urban area. Engine recharge for at
least a portion of these vehicles will occur outside of the core (for example, a suburban-
based hybrid might switch its engine off on entering the core, and restart it when it
returns to the suburbs in the evening).
22
Highly sophisticated emissions measurement technology, allowing detailed study of emissions variation
on a virtually instantaneous basis, has become available in the last few years. This is the type of
technology needed to examine issues such as emissions spikes after warm restarts, and Argonne National
Laboratory currently is using this technology to examine the emissions trade-offs made possible or
caused by hybridization. Industry is undoubtedly conducting similar research.
23
Personal communication, John German, Honda North America.
28
Temporal. For many areas, there is a theoretical advantage to having vehicle emissions
shifted from early morning to late afternoon because of the role of sunlight in producing
smog. Hybrids with substantial EV range may be able to operate in EV mode during the
morning commute or in other morning driving, switching their engines on only in the
afternoon when emissions may have less impact on ozone concentrations.
We stress that the temporal advantage is theoretical because, to the best of our knowledge,
no air quality modeling research has been conducted to verify that a shift of morning emissions
to the afternoon will provide an ozone air quality benefit. Results should vary depending on
whether the battery recharge for the ZEV operation comes from the grid or from afternoon (or
overnight) engine recharging. Results will also depend strongly on the location and emissions
properties of recharging power plants and on the specific ozone formation chemistry of the air
basin where the vehicles are operating.
2.5 Elements of Hybrid Design
In hybrids, the major design choices involve selecting the type and size of OBPU,
motor/generator, storage device, and other components, and designing an operating strategy that
directs the operation of the components and channels energy flows through the drivetrain.
Although there are numerous ways to configure a hybrid drivetrain and a variety of ways to
characterize various groups of hybrids, dividing hybrids into series and parallel types is an
obvious first-order characterization, and the one adopted here to describe the design choices
associated with hybrids.
2.5.1 Parallel Hybrids
In parallel hybrids, both the OBPU and the electric motor(s) can drive the wheels, either
separately (each takes one drive shaft), or through a transmission or gear set with two or more
input shafts, or through use of an integrated motor/engine set where the motor can also serve as
the starter. Figure 2-2 shows a simplified layout of a parallel hybrid with engine and
battery/electric motor(s) powering separate driveshafts. A commonly discussed mode of
operation for parallel hybrids has the vehicle operating like an EV at low speeds, with only the
motor(s) engaged (this is called electric launch); the OBPU engages only at higher speeds. This
operating strategy keeps engine operations away from low load operations that generally produce
high hydrocarbon (HC) and carbon monoxide (CO) emissions and inefficient fuel use. Within
this basic design, there are a number of choices of alternative operating strategies. For example,
one operating strategy would limit motor usage to electric launch, regenerative braking, and
power boosting when the OBPU could not handle the power demand by itself. In other words,
once the OBPU is engaged after an electric launch, it would provide all of the power needed to
run the vehicle, with the motor disengaged except during braking (to be used as a generator for
regenerative braking) or under high load (e.g., during high acceleration) to provide boost power
to the engine. When the vehicle is stopped, the engine can be turned off and the accessories run
by the battery. Toyotas Prius HEV operates this way. Although Toyota calls the Prius a
parallel/series hybrid, from the basis of energy flows it may be most appropriately called a
parallel hybrid.
29
Figure 2-2 One Example of a Parallel Hybrid Drivetrain
An alternative operating strategy is to use the motor as a more frequent supplement to the
engine, to smooth transients or keep the engine out of operating regions where it is less efficient
or likely to generate higher emissions. To maximize efficiency, the vehicle control system could
be programmed to allow the motor to supplement the engine whenever its net round trip
efficiency (including battery charge and discharge efficiency, motor efficiency, etc.) exceeded
that of the engine pathway. Refinements of this strategy would smooth the load by allowing the
engine to lag the instantaneous load by a short time interval and respond to the time-averaged
load, with the storage device providing either the power boost or power absorption function as
needed to satisfy the instantaneous load. This strategy reduces the transient losses but increases
battery losses compared to the instantaneous load-follower (Cuddy and Wipke 1996).
An alternative, sometimes called a mild hybrid system, is to use the motor only as a power
booster and for regenerative braking, with no all-electric operation; in this case, the motor and
storage device may be relatively small and thus less expensive, but there is less opportunity to
keep the engine out of less-efficient operating regions and lessened capability for recapturing
braking energy. The Honda Insight HEV fits into this category.
Until fairly recently, parallel hybrid designs had generated less interest than series hybrid
designs among automakers and researchers, perhaps because they are harder to design and
analyze. A parallel hybrid operating in the city will have to turn its OBPU on and off frequently,
and smoothly and efficiently combine the changing torques of OBPU and motor, decoupling and
recoupling one or the other from the driveshaft when needed (Burke 1992) not an easy task for
either the control system or the transmission.
Although series hybrids are mechanically simpler than parallel hybrids and would seem to
be easier to design, an important advantage of the parallel hybrid is that it can obtain the
efficiency advantages of the series hybrid (use of regenerative braking, engine downsizing, and
maintenance of OBPU operations in the better parts of its operating map) with a more efficient
connection of OBPU shaft power to the wheels. In other words, the combination of transmission,
torque converter, and differential is more efficient than the series hybrids shaft-to-wheel path of
generator/alternator, (possibly) inverter, motor/controller, transmission or reduction gear and
(unless direct drive wheel motors are used) differential. Another major advantage is that a
parallel hybrids electric motor will be significantly smaller than that required on a series hybrid,
since in the series case the motor provides the sole motive power to the wheels. This yields a
30
significant cost savings. Because most conceptions of the parallel hybrid have the engine doing
more load-following than in a series layout, and the engine being frequently turned on and off,
there had been concerns that attaining low emissions might be substantially more difficult than
with a series layout (Burke 1992). As noted above, this concern now appears to have been
misplaced.
Toyotas and Hondas recent successes with the Prius parallel/series hybrid (Toyota) and
Insight parallel hybrid (Honda) indicates that anticipated barriers are not insurmountable. Both
vehicles attain low emissions, high fuel economy, and smooth operation. Good design and the
rapid advances in onboard computing capability appear to have overcome the parallel systems
operational complexity.
2.5.2 Series Hybrids
In a series hybrid, the OBPU drives a generator, whose electrical output powers an electric
motor driving the wheels and any accessories when needed, and charges a storage device when
the devices state of charge (SOC) drops below a desired range. Figure 2-3 shows a simplified
24
diagram of a series hybrid layout.
Figure 2-3 One Configuration of a Series Hybrid Drivetrain
In a series configuration, all of the tractive power needs are met by the motor, which obtains
electricity from either the engine/generator directly or from the battery. As with the parallel
hybrid, the control system must decide how to trade off the operating advantages and
disadvantages of the various energy paths available to the system in this case, the engine to
generator to motor path and the battery to motor path, with the battery being recharged from
the engine, regenerative braking, and for grid-connected hybrids, the grid. For a grid-independent
hybrid, frequent use of the battery to drive the motor may force the engine to recharge the
battery, which adds the batterys in/out charging losses to the engine to motor path; the control
system will generally avoid engine recharge unless battery charge is getting dangerously low or
the engine would otherwise be forced to operate quite inefficiently. But in a grid-connected
24
Missing are inverters and controllers; if alternating current (ac) motors are used, an alternator could be
used in place of a generator (with an inverter between the alternator and battery).
31
hybrid, the system will often favor the battery to motor path, to maximize the amount of grid-
recharging that can be obtained.
2.5.3 Component Sizing
The size of the powertrain components, that is, their rated peak and continuous power and,
where applicable, energy storage capacity, depends on the basic operating strategy of the
powertrain, vehicle characteristics that determine loads (vehicle mass, aerodynamic efficiency,
tire efficiency), and performance requirements established for the vehicle, e.g., requirements for
minimum Z60 time.
2.5.3.1 A Dated Thermostat Concept
We should first clear up a common misperception of what a hybrid will look like. A
common early conception of a hybrid was of a vehicle with a small engine operating at a constant
output, at its maximum efficiency design point, providing the average energy needed by the
vehicle, with a battery (or flywheel or ultracapacitor or other storage device) acting as a load
follower, providing any power needed above the average load (e.g., for acceleration or hill-
climbing) and absorbing excess engine power when the power needed was below the average
load (e.g., during idle or moderate speed cruise).
One problem with this conception is that some trips are made under higher-than-average
load conditions for long time periods, e.g., trips with the vehicle heavily loaded, at very high
speeds, and/or with significant uphill grades. Consequently, to be able to handle such trips, the
engine output power for the hybrid described above would have to be much higher than
average, or else the storage device would have to be large enough to provide the extra power
needed without losing its charge an expensive proposition.
In fact, the performance requirements for normal vehicles make this type of
thermostat/constant engine power operation impractical. In the U.S. market, conventional light-
duty vehicles are high performance vehicles, satisfying, virtually without exception, a number of
demanding standards for acceleration performance, gradeability (hill-climbing ability), sustained
high speeds, and long range. For example, the requirement for gradeability (e.g., 55 mph up a
6.5% grade) demands the sustained use of power (under PNGV requirements, for 20 minutes)
well above the average encountered in normal driving. Operating constantly at its best
efficiency point, an engine capable of satisfying gradeability requirements by itself would likely
have to be larger than the engine in the conventional drivetrain it replaced (OTA 1995; Murrell
1995).
25
Thus, obtaining the hoped-for small engine requires either that the thermostat strategy
25
For a midsized auto with a Z60 time of about 12 seconds and gradeability of 60 mph at 6-percent grade
with 2 passengers, gradeability requires about 60% of acceleration power (OTA 1995). For a group of
modern 4-valve engines, power at the maximum efficiency point averaged 40% of peak power, with a
range from 20 to 68%; for 2-valve engines, maximum efficiency power averaged 50% of peak power,
with a 35-71% range (Murrell 1995). Assuming a 40-percent maximum efficiency point and a 60-percent
gradeability point implies that a constant speed engine in a hybrid drivetrain, operating at its best
efficiency point, would have to be rated at 50-percent higher power than the engine in the conventional
drivetrain it replaces.
32
be abandoned and the engine be allowed to follow the load up to its full rated capacity (with
some cost in efficiency and emissions performance), or that the storage device be capable of
supplying the extra power needed.
Incorporating a battery large enough to provide substantial power assist to the engine over
long periods would not be easy. Currently available batteries are capable of providing their peak
power only for short periods, because they generate a great deal of heat on rapid discharge.
Although batteries could simply be made large enough to provide adequate power for
gradeability on a sustained basis (at well below their peak capacity), the trade-off in size, weight,
and cost for most batteries would be unacceptable. Some recent battery designs appear to
combine reasonable energy storage capability with very high specific power, which might help to
solve this problem; however, these batteries are just beginning to become commercially
available.
26
Storage devices such as ultracapacitors and flywheels can provide high power more
readily than batteries of equal size and weight, but they generally have low specific energy, so
that anything but a very large device would have limited capability of sustaining a long hill climb
or other extended-duration high power episode. Unless drivers are willing to accept substantial
deterioration of performance on such occasions, a design that requires battery power assist for
extended events like hill climbing will not be commercially feasible.
2.5.3.2 Grid-Independent Hybrids
Given the impracticality of the small average power engine, how should the powertrain
components of a hybrid be sized? Although various combinations of engine/battery/motor size
will work, one obvious starting point is an engine sized to at least satisfy gradeability
requirements, since the long duration of these requirements implies that we may not want to rely
on battery power to satisfy them. For the case examined by OTA (midsized car, lightly loaded
gradeability of 60 mph up a 6% grade, zero to 60 mph acceleration (Z60) time of 12 seconds), a
hybrid vehicle engine sized to just meet the requirements would have about 60% of the peak
output of the conventional engine drivetrain it replaces (OTA 1995; Murrell 1995). The battery
and motor would then be sized to allow the vehicle to meet acceleration requirements, which
typically are considerably more demanding in terms of maximum power than the gradeability
requirements.
27
One way to operate this grid-independent drivetrain is to allow the control system to select
whatever combination of engine, motor, and battery output satisfies a preselected objective
highest net efficiency, lowest emissions, etc. Selecting as an objective the minimization of
energy use is likely to result in the engine following the load above a minimum load, thereby
keeping engine operations away from low load operations that generally produce inefficient fuel
use and high emissions. At loads below the minimum (for example, during braking or an actual
stop, or low speed cruise), either the engine would be shut off and the storage device would
26
As discussed in Section 3, new NiMH battery designs have shown substantial increases in specific
power, with batteries in Toyotas Prius hybrid attaining nearly 500 kW/kg.
27
Motors have both continuous and peak power ratings because of cooling limitations. For acceleration
requirements, the motors peak rating can be used.
33
power the vehicle, or the engine would be run within its most efficient region (which would
produce more power than demanded by the vehicle) and the storage device would absorb the
engine's excess power. At loads above the engine's rated power, or above the engine's efficient
operating region (for example, during rapid acceleration), the storage device would supplement
the engine. This operating strategy seeks to use the storage device only when its net round trip
efficiency is greater than that of the engine-to-wheels pathway.
This basic type of design (engine sized to provide hill-climbing capability, storage device
sized to provide boost power for short-term acceleration) is, of course, not the only design
option. For example, if time limits are placed on the requirement for continuous power for grade
climbing, such as the PNGV-specified 20 minutes, the storage device might provide a portion of
this power, leading to a design with a higher power storage device (with possibly higher storage
capacity as well) and smaller engine.
28
Recent development of high power storage devices with
good energy density
29
make this option appear more feasible.
In a parallel system, another alternative design, a mild hybrid, would be one in which the
engine was more powerful and the motor and storage device less powerful, so that the
motor/storage device provided only some boost power and a smaller regenerative braking
capability as well as the ability to turn the engine off during idle or braking. This type of design
represents a trade-off of fuel economy boost versus cost. Using a larger engine will reduce the
fuel economy boost of the hybrid, but at a savings in vehicle cost the battery/electric motor
portion of the drivetrain is more expensive, on a marginal $/kW basis, than the engine/
transmission portion of the drivetrain.
2.5.3.3 Grid-Connected Hybrids
An alternative advocated by some especially those seeking to maximize zero emissions
capability is to design the vehicle as a full performance electric vehicle, with an engine to
provide extended range. This vehicle could run in a pure electric mode on days when its total
miles traveled were below its all-electric range, or for the first portion of daily travel on days
with longer travel (or during some intermediate portion of the travel, if this is advantageous). The
number of miles thus electrified then would be primarily a function of the size and specific
energy and power of the storage device. For storage devices with substantial energy storage (and
thus high all-electric range), most drivers might use the engine only rarely, recharging the storage
device each night. To assure full performance on extended-length trips, however, the engine
would have to have as much power as the power booster engine above.
30
For designs with
28
Note, however, that there would be a theoretical possibility that such a design would occasionally deplete
its reserve storage capacity, forcing the vehicle to limp home in an engine-only mode.
29
The original Japanese version of the Toyota Prius hybrid had NiMH batteries attaining approximately
500 kW/kg specific power and 45 Wh/kg specific energy (Hermance 1998). The 2000MY U.S. Prius has
a redesigned battery pack achieving a 880 W/kg specific power (Toyota 2000).
30
That is, sized for gradeability on engine power alone. On a long trip, the battery would eventually be
substantially depleted, and would then be incapable of providing a lengthy power boost for a long grade.
34
enough EV range to allow the majority of miles traveled to be fueled by the grid, the engines
size and weight would be more important considerations than its fuel efficiency; a small,
lightweight gas turbine or two-stroke spark ignition (SI) engine might be a good choice for such
designs.
The advantages of a grid-connected hybrid include:
The vehicle can operate in zero emissions mode for a considerable distance, allowing
entry to areas that might otherwise be restricted to conventional vehicles and providing a
substantial emissions benefit over a grid-independent hybrid.
It can use grid electricity for a significant fraction of its total travel, reducing gasoline
use.
In the event of an oil crisis, it could be used as a short-range EV, providing
transportation capability and thus supplementing domestic oil supplies.
The key disadvantage is that the design will be substantially more expensive than a grid-
independent hybrid the increased battery size and, for parallel designs, increased motor size
31
dictate considerable additional costs with no offsetting reduction in engine requirements.
An important design point for grid-connected hybrids is the relationship between battery
longevity and the operating SOC range of the battery during hybrid operation. After the vehicle
has traveled its full EV range and the engine is turned on, the control system will seek to
maintain battery charge within a predetermined range as the vehicle operates in hybrid mode.
The lower this range is, the more grid electricity can be used, since the amount of grid recharge is
proportional to beginning SOC minus ending SOC. However, there is some possibility that
battery longevity may suffer if the operating range is too low. Vehicle designers must select an
operating range based on trading off the potential for more grid recharging versus the (potential)
loss of battery longevity at lower SOC operating ranges.
As discussed in Section 2.1, the State of California Air Resources Board has specified that
hybrid vehicles with substantial EV range can obtain ZEV credits toward vehicle manufacturers
ZEV goals an added advantage of a grid-connected design.
2.5.3.4 Limited Performance Hybrids
Relaxation of requirements for full performance and range equivalence allows more
flexibility in vehicle design, though perhaps at the cost of reduced consumer satisfaction. For
example, a hybrid designed to operate as a full function EV (often referred to as a range
extender) could use an engine sized only to allow high speed cruise capability, forgoing full
gradeability for longer trips to minimize engine size and weight (and cost) and maximize fuel
31
To allow reasonable performance in EV mode.
35
efficiency.
32
Similarly, relaxing gradeability requirements for a grid-independent hybrid could
allow the engine to be considerably smaller, with the improved fuel economy this implies. If the
Z60 requirement were retained, however, the vehicle designer would have to either retain the
larger engine or install a larger, more powerful battery and motor, at considerable added cost.
2.5.4 Role of Component Efficiency
In hybrid-electric powertrains, the engine is used as a generator of electricity (its only role
for series hybrids) and producer of power that flows to the wheels through a transmission
(parallel hybrids only). When engines are used for electricity generation, their shaft power turns
a generator or alternator, producing electricity that flows either directly to the electric motor(s)
(which turns the wheels through direct drive, reduction gears, or a transmission) or to the storage
device, where it is stored for later use. When the motors are used for braking, they produce
electricity that must be stored by the storage device, also for later use to power the vehicle. When
battery storage and alternating current (ac) motors are used, the electricity flowing from the
engine to either the motors or battery must be inverted, since the battery requires direct current
(dc) and the motors, ac; and flow from the battery to the motors must be inverted to produce the
required ac. Similarly, electricity produced by ac motor braking would have to be inverted before
being stored by the battery, then inverted again later when it is required for use by the motor.
Consequently, the hybrid drivetrains electric path from the engine shaft to the wheels loses
energy in the alternator or generator, inverters (if needed), motor/controller acting both as motor
and as generator/alternator, reduction gear or transmission, and storage device (in both charging
and discharging as well as storage). Generally, the electric pathway from engine to motor
(partly direct, partly through the storage device) is less efficient than the mechanical pathway of
transmission, differential, etc. it complements (in a parallel hybrid) or replaces (in a series
hybrid). The hybrid gains its advantage from regenerative braking, the smaller engine operating
in a more efficient mode than the engine it replaces, and the elimination of idle fuel flow.
However, the need to obtain relatively high energy efficiencies from each of the components in
the hybrid drivetrain is obvious if hybrids are to attain fuel economy superior to conventional
vehicles, even with these latter advantages.
Each component in the drivetrain can use alternative technologies with significant variation
in efficiency, cost, size, and weight. Because most batteries have relatively poor in/out efficiency
80% is typical for a lead acid battery hybrids with battery storage should be designed to
minimize production of excess electricity from the engine that must be stored. This can be
translated into a maxim that average engine output should be the minimum possible consistent
with the need to maintain reasonable levels of engine efficiency (which may fall off dramatically
at light load). Were batteries to be developed with much higher in/out efficiency DOE has
measured 1998 SAFT America Li-ion batteries designed for power-assist hybrids at 93% in/out
efficiency (Sutula et al. 2000), for example the need to minimize use of the storage pathway
would diminish.
32
If grade climbing requirements were anticipated in advance, the engine could be turned on early to
maintain charge and allow the vehicles gradeclimbing ability to be sustained longer.
36
The efficiency of capture of regenerative braking is also controlled by component efficiency
as well as design choices. Assuming four-wheel braking for safety, a portion of braking energy
will be lost if the vehicle has two-wheel drive, since only the powered wheels can capture
braking energy. Since the motor(s) can exert braking force equal only to their horsepower, a
portion of the energy exerted during hard braking will be lost to the mechanical brakes,
33
and a
fraction will be lost to the motor gearing or transmission. And a fraction of that captured by the
motor/generators (alternators) will be lost at the battery and possibly in inversion to dc power
and back again to ac.
2.5.5 Trends in Hybrid Efficiency over Time
Over time, improvements in component technology and substantial redesign of some key
hybrid system elements, particularly engines, should yield improved efficiency. Ongoing
research efforts in electronic components and electric motors should yield important
improvements in component efficiency as well as reductions in component size and weight, both
yielding improved vehicle efficiency. Further, if hybrid sales grow to sufficient levels, vehicle
manufacturers will substitute engines specifically designed for hybrid service for the off-the-
shelf designs that dominated early prototypes. Note again that the Prius engine was purposely
built for hybrid operation, with a unique combination of Atkinson Cycle design and lightened
components to take advantage of its restriction to operating below 4,500 rpm (Toyota 2000).
On the other hand, it is far from certain that hybrid drivetrains incremental improvement
over baseline conventional drivetrains will grow with time, and it may shrink. This is because
many ongoing improvements to conventional drivetrains focus on reducing the same
inefficiencies that hybridization addresses. For example, variable valve timing and lift systems
and cylinder deactivation found in the Mitsubishi MIVEC system and being introduced in other
engines dramatically cut pumping loss, leaving less energy waste for a hybrid drivetrain to
reduce. If direct injection stratified charge (DISC) engines are introduced to the U.S. market
(they currently are attaining a substantial share in the Japanese market) in the future, the waste
target for hybrid drivetrains to help reduce will shrink further. In the OTA study, for example,
a midsized vehicle was examined for three timeframes (1995, 2005, and 2015) in both series
hybrid (with advanced lead acid battery) and advanced conventional form where vehicle
designs were identical in all aspects except the drivetrain so that all efficiency differences can
be attributed to the drivetrain. Despite assumed improvements in hybrid component efficiency,
the efficiency advantage of the hybrid over the advanced conventional vehicle was found to
shrink over time from 30% in 1995 to 25% in 2005 to below 23% in 2015 (OTA 1995). On the
other hand, assumptions about component efficiency are extremely important to these results. In
the OTA analysis, for example, substituting a flywheel for the battery in the 2015 vehicle, which
greatly boosts storage in/out efficiency and reduces overall vehicle weight, increased the hybrid-
to-advanced conventional efficiency advantage to 38% (OTA 1995). In a real-world example of
improvement in HEV fuel efficiency, Toyota managed to boost the performance on the EPA
33
Maximum braking power will likely be higher than maximum motor power, since 60-0 mph braking time
and distance should be expected to be substantially shorter than 0-60 mph acceleration time and distance,
and acceleration uses both motor and engine (for parallel hybrids).
37
Highway cycle of its Prius HEV from about 54 mpg in 1998 to 58 mpg in 2000 with an
improvement in Z60 time from about 14.5 seconds to 12.7 seconds (Hellman et al. 1998; Toyota
2000). In other words, trends in the efficiency boost afforded by hybridization will depend on the
balance between improvements in hybrid designs and components and improvements in
conventional drivetrains that attack the same sources of inefficiency addressed by hybridization.
The net effect of these two opposing forces over time is uncertain.
2.6 References
Air Resources Board, 1999, California Exhaust Emission Standards and Test Procedures for
2003 and Subsequent Model Zero-Emission Vehicles, and 2001 and Subsequent Model Hybrid
Electric Vehicles, in the Passenger Car, Light-Duty Truck and Medium-Duty Vehicle Classes,
California Environmental Protection Agency, Aug. 5.
Burke, A.F., 1992, Hybrid/Electric Vehicle Design Options and Evaluations, SAE Technical
Paper Series 920447, in Electric and Hybrid Vehicle Technology, SAE Report SP-915.
Cuddy, M.R., and K.B. Wipke, 1996, Analysis of the Fuel Economy Benefit of Drivetrain
Hybridization, SAE Paper.
German, J., Honda North America, personal communication.
Hellman, K., M.R. Peralta, and G.K. Piotrowski, 1998, Evaluation of a Toyota Prius Hybrid
System (THS), U.S. Environmental Protection Agency Report EPA-420-R-98-006, Office of
Mobile Sources, Ann Arbor, Mich., Aug.
Hermance, D., 1998, unpublished material (Luncheon Keynote Presentation on Toyota Hybrid
System: The Worlds First Mass-Produced HEV, SAE/NESEA Making Hybrid Electric Vehicles
Commercially Viable TOPTEC).
Murrell, J.D., 1995, unpublished material.
Office of Technology Assessment, 1995, Advanced Automotive Technology: Visions of a Super-
Efficient Family Car, OTA-ETI-638 (U.S. Congress, Washington, D.C., Sept.). Full report
unpublished due to OTA closing.
Reuyl, J.S., and P.J. Schuurmans, 1996, NEVCOR, Inc., Policy Implications of Hybrid-Electric
Vehicles, Final Report to NREL, Figure 11, April 22.
Schiffer, M.B., 1994, Taking Charge; The Electric Automobile in America, Smithsonian
Institution Press, Washington and London.
Shulock, C., 2001, Air Resources Board, personal communication, Aug. 22.
Sutula, R.A., et al., 2000, Electric and Hybrid Vehicle Energy Storage R&D Programs of the
U.S. Department of Energy, 17
th
Electric Vehicle Symposium, Montreal, Canada, Oct.
38
Tamor, M.A., 1996, Ford Motor Company, unpublished material.
Toyota, 2000, Prius Product Information 2000, brochure PR-E-00UJ03, July.
U.S. Department of Energy, 1994, Electric and Hybrid Vehicles Program, DOE/EE-059, April.
39
Section 3
Methodology and Modeling Issues
3.1 Modeling the Vehicle
This section describes the modeling of the energy use and emissions of vehicle operation.
The question we are trying to address in our vehicle modeling sounds deceptively simple: What
is the impact on fuel consumption, emissions, and performance of hybridizing a drivetrain, that
is, shifting from a conventional drivetrain to a hybrid one? From a hardware standpoint, we are
trying to measure the energy, emissions, and performance impact of removing the conventional
drivetrain and substituting a hybrid one, that is, removing the engine/transmission/differential
and substituting one of a potential variety of hybrid drivetrains (for example, an engine/
generator/battery/electric motor/power split device combination), with whatever structural
changes might be necessary.
3.1.1 ADVISOR Model and Validation
Vehicle performance (acceleration capability, gradeability, fuel use, and emissions) are
estimated using the ADVISOR (Advanced Vehicle Simulator) model developed by the National
Renewable Energy Laboratory (NREL) in support of the Department of Energys hybrid
propulsion system program. ADVISOR is programmed in the MATLAB

/Simulink

environment using a modeling approach that combines backward- and forward-facing

calculations.
The backward approach begins with a required speed trace that the vehicle must meet. The
model first calculates the force that must be exerted at the wheels to move the vehicle as the trace
requires, and checks to see whether the tires can deliver the required force without slipping. It
then works backwards through the powertrain, calculating what each component must do to
allow the vehicle to meet the trace, for example, the power that the engine must exert. For each
calculation, the model enforces limits on what can be achieved, for example the motor/controller
has speed, torque, and current limits. When the powertrain cannot achieve what is demanded, the
vehicle will fall behind the trace.
For this backward approach, the model estimates the capabilities and efficiency of key
components by using tables or maps of efficiency or fuel use (or emissions) versus output torque
and speed (for engines and motors). These tables and maps normally are produced by steady-
state testing, so they reflect conditions after warmup. ADVISOR simulates the effect of engine
and catalyst warmup, as well.
In the forward-facing approach to vehicle modeling, the computational path starts with
driver behavior, that is, throttle and braking behavior, calculates the effect on engine power, and
continues through the transmission and the remainder of the drivetrain to compute the tractive
force at the tire/road interface, which then results in an effect on vehicle movement. As in the
backward-facing approach, limits on wheel slip are applied to ensure that the vehicle does not
40
accelerate faster than the wheels adhesion capabilities will allow. Inertia effects of the drivetrain
are also accounted for.
ADVISOR acts as a purely backward-facing model as long as none of the torque or other
limits of the powertrain components are exceeded by the requirements of the speed/time trace.
The model is very fast in the backward-facing mode, so that multiple iterations of vehicle cases
can be run quickly. In our evaluations, examination of the difference in the desired vs. the actual
speed-time trace was used to determine whether a vehicle could match the trace. All cases
reported here do so; in earlier reported cases (Wang et al. 1997), one vehicle did not meet the
most aggressive trace.
When the work reported here was done, few actual hybrid vehicles had been tested, and
necessary details for broad sensitivity analysis of varying hybrid electric vehicle (HEV)
performance capabilities, e.g., a library of engine and motor maps, were rather limited. Further,
since only the Prius HEV had been thoroughly tested when this work was conducted, validation
of ADVISOR against real-world hybrid vehicle performance was very limited. However, ANL
and NREL tested the model in a number of ways:
ANL has validated ADVISOR against numerous conventional vehicles for Z60 and
against two Saturn SL1 models [1.9-L single overhead cam (SOHC) with manual and
automatic transmission] and a VW Passat [1.9-L turbocharged direct injection (TDI)
manual] for fuel economy.
Since the model is capable of operating in a backward-facing mode and partially in a
forward-facing mode, NREL has created tests in which simulated vehicles are measured
using both modes. The first test asks the vehicle to meet a time/speed trace that it cannot
satisfy, forcing the model into a forward-facing mode. The second test simulates the
vehicles performance on speed traces that are iteratively changed until the trace is found
that is exactly the fastest that the vehicle can accelerate. For this trace, ADVISOR works
strictly as a backward-facing model. ADVISOR achieved very similar results for both
0-60 mph acceleration and energy consumption in both modes (Wipke et al. 1999).
NREL tested earlier versions of ADVISOR against industry models using identical
inputs, with similar results (Wipke et al. 1999).
Researchers at Virginia Polytechnic Institute ran ADVISOR using data from their
FutureCar competition entry, a series hybrid, simulating the speed trace used by the
actual vehicle and comparing actual to modeled results for fuel use and battery energy
use. The ADVISOR results matched the test results within the uncertainty of the test
measurements (Wipke et al. 1999).
3.1.2 Rules for Comparing Vehicles and Modeling Difficulties
Ideally, the hybrid vehicle/conventional vehicle pair being compared would follow a set of
rules designed to make the comparison as fair as possible. The following rules are designed to
ensure a fair assessment of the impact of shifting from a conventional drivetrain to a hybrid one:
41
1. Everything outside the drivetrain, with the exception of any needed structural
reinforcements, will be the same. In other words, both the HEV and the conventional
vehicle (CV) to which it is being compared should have the same aerodynamics, same
basic structure and materials (except for reinforcements forced by the heavier hybrid
drivetrain components), and the same tires (with allowances for differences in vehicle
weight). This rule would be relaxed if the layout and space demanded by the hybrid
drivetrain either forces negative changes or allows positive ones in vehicle size and/or
aerodynamics.
2. The engines in the conventional drivetrain and hybrid drivetrain generally will be
different in rated power but of the same design except for changes made possible or
logical by the hybridization. Changes could range from minor design changes that take
advantage of the narrower engine operating regime the hybrid design allows, to totally
changing the auxiliary power unit (APU) technology, e.g., using a gas turbine in the
hybrid if the hybrid operating strategy did not require the APU to follow the load.
3. The two vehicles should perform as similarly as possible, except where performance
differences are inevitable because of problems too expensive to fix or advantages too
attractive not to take advantage of. The primary performance indicators we used are
zero to 60 mph acceleration (Z60) time and gradeability, matching within reasonable
limits of accuracy the Z60 times of the two vehicles, but allowing hybrids to meet a
minimum gradeability standard considerably below that achievable by the conventional
vehicle. We have not at this time estimated and compared passing acceleration for the
two vehicles, though this is another performance indicator that is of importance to
vehicle designers.
Where these rules are not followed, it is our contention that differences in fuel consumption,
emissions, and performance between a hybrid vehicle and conventional vehicle should not be
attributed solely to hybridization. For example, the 1999 Japanese version of the Toyota Prius
has often been compared to the 1999 U.S. Toyota Corolla, a vehicle of roughly similar size made
by the same company. However, the Prius has several features different from the Corolla (low
rolling resistance tires, a reduced coefficient of aerodynamic drag but larger frontal area, and
variable valve control) that impact fuel economy but have nothing to do with hybridization. That
is, these features could just as easily have been applied to vehicles with conventional drivetrains.
Also, the Prius is significantly slower than the Corolla, with Z60 time of about 14.5 seconds;
were the Corollas gear ratios adjusted and engine size reduced to match the Priuss performance,
its fuel economy would be significantly increased. The effects of these differences should be
accounted for separately from the effects of hybridization. On the other hand, the effect of three
additional, unique Prius features probably should be included as an effect of hybridization. These
features are Priuss use of an Atkinson cycle in the engine,
34
its lighter engine components (the
availability of boost power from the motor allows engine speed to be limited to 4,500 rpm,
34
The Atkinson cycle, if used in a gasoline engine of a given displacement, will sharply reduce the power
of the engine while increasing its efficiency. This cycle is not appropriate for a conventional drivetrain,
but a hybrid can overcome its power disadvantage and thereby obtain the benefit of its higher efficiency.
42
yielding less stress on the engine), and its continuously variable transmission, whose unique
design is made possible only by the combination of electric motor, generator, and engine.
35
3.1.2.1 Selecting a Conventional Engine/Hybrid Engine Pair for Comparison
A key problem in modeling the effects of hybridization is the difficulty of modeling a hybrid
engine that represents only those features (particularly, lower power and size/weight) allowed by
the hybridization. There are both philosophical and practical problems.
A philosophical problem is that reducing engine size while maintaining all else equal, that
is, maintaining the same level of technology and the same design, may not represent the best
comparison between conventional and hybrid drivetrains. Instead, the degree of operating control
afforded by a hybrid drivetrain may make it attractive to the vehicle designer to eliminate some
engine enhancements normally used in a conventional drivetrain to maintain high engine
efficiency over a wide range of operating conditions. For example, although the Toyota Prius has
variable valve control and the baseline Toyota Corolla does not, it may be more likely that a
hybrid designer would remove a costly pre-existing variable valve system in hybridizing a
conventional drivetrain. This is because some hybrid engine management strategies keep the
engine within fairly narrow operating limits negating much or all of the benefits of a variable
valve system, which is designed to overcome the efficiency trade-offs inherent in an engine with
fixed valve timing and widely varying operating conditions. Note that a hybrid drivetrain is
substantially more complex than a conventional drivetrain and the added components such as
motor/controller and battery are very expensive. Thus, a primary task facing the designer of a
hybrid vehicle is to remove costs wherever possible, because the vehicle will be competing with
simpler, less expensive conventional vehicles.
The issue of making an apples to apples comparison becomes even more difficult if the
hybrid engine is a gas turbine or other device that becomes practical because of the hybrid
drivetrains ability to use the battery/electric motor to follow the load, with the engine providing
baseload power. For such a hybrid, it is difficult to identify an appropriate baseline engine for
comparisons sake. In this report, however, we deal only with hybrids using conventional
gasoline engines.
There is a practical problem with the goal of keeping all else equal. In moving from a
baseline conventional drivetrain to a hybrid drivetrain with a matched (but smaller) engine, the
modeler must downsize the engine without losing the basis for a valid comparison. The two most
obvious downsizing methods are to scale the engine down (mathematically) or to find fuel
consumption and emissions maps of a smaller engine whose design is similar enough to the
original that any fuel economy and emissions differences between the two engines are primarily
a function of engine size rather than design differences. An examination of the fuel economy and
emissions performance of engines with essentially equivalent power but different designs reveals
35
The inclusion of the latter as a benefit of hybridization may be controversial because continuously
variable transmissions (CVTs) can be added to conventional drivetrains; however, the transmission of the
Prius is both less expensive and more efficient than a conventional CVT would be, thus, hybridization
makes this type of CVT use more likely.
43
that design differences can yield large differences in emissions and fuel economy. Unfortunately,
there are few engine families that retain constant design over a range of displacements and power
output. Consequently, modeling the hybrids engine by selecting a conveniently-sized engine
map from the library makes it difficult to distinguish which part of any energy and emissions
differences between the conventional and hybrid vehicles are due to hybridizing the drivetrain
and which are due to differences in engine design.
Another approach to modeling a reduction in power is to reduce peak rpm (modeled by
deleting the engine map above the selected peak rpm), as done in the Prius. However, it appears
more likely that, in the long term, automakers will place physically smaller engines in their
hybrids, and obtain them by redesigning off-the-shelf engines, designing new engines, or
purchasing smaller engines from suppliers. We could model reduced displacement by assuming a
reduction in the number of cylinders, but there would be changes in noise, vibration, and
harshness (NVH) in an engine downsized in this manner, and this would still provide limited size
options. Finally, we could model a downsized engine by reducing the peak torque line and
associated isobars of a baseline engine by a constant. This option is attractive because it offers an
infinite number of power levels. However, the approach ignores the real-world discontinuities
that arise when cylinders are reduced in size, which is the physical analog of this approach.
We have chosen the scaling approach of reducing the peak torque line and associated
isobars of the baseline CV engine by a constant to achieve the map of a downsized engine of the
same technology. We note, however, that this approach automatically introduces errors into the
modeling process.
In this study, we seek to project the effect of hybridization in the future, given the likelihood
that large numbers of hybrid vehicles will not enter the U.S. market until 2005 or later. Some
components of hybrid and conventional drivetrains are changing at a faster pace than others,
partly because they are at different stages of development and partly because there are different
levels of resources being expended toward their improvement. Consequently, current state-of-
the-art data may not reflect likely future increases in component efficiencies, complicating the
modeling process. Although higher future efficiencies can be assumed in the modeling process, it
is not possible to recreate the effects of the unknown design and materials changes that will
likely accompany the increase.
Finally, although this analysis attempts to compare conventional and hybrid vehicles that are
close to equivalent in performance, in practice the two will not have identical performance
characteristics, and in some cases differences may be large. In particular, using our design rules,
an HEVs gradeability will be inferior to the performance equivalent CVs grade-climbing
ability. Also, grid-connected hybrids expected to operate part-time as EVs may have substantially
different performance from CVs. For example, in the cases we examined, the hybrids
acceleration capability in EV mode was inferior to the CV, although the development of higher-
specific-power batteries may eventually allow equivalent performance; in HEV mode, for the
parallel designs, the HEVs acceleration was somewhat superior to the CV. Some HEV designs
with limited battery energy storage may occasionally encounter driving situations in which the
batteries have been discharged and only engine power is available, reducing performance
capability temporarily. The Toyota Prius has a warning light to alert drivers to low battery charge
44
and reduced performance. In other words, readers of this and other analyses of HEV capability
should be aware of the potential for some apples to oranges comparisons where side-by-side
comparisons are between vehicles whose performance capability is not completely equivalent.
Further, the drivetrain changes may make the performance of hybrids very sensitive to design
details, including control strategies, with subsequent uncertainty in predicting what performance
actually will be. For example, the Z60 time of a hybrid will depend strongly on whether the
control strategy allows the engine to be left on when the vehicle is stopped (which would,
however, increase fuel use) or, when the vehicle is electrically launched, whether the engine will
be turned on at low vehicle speeds when high power is demanded. If the engine is turned off
when the vehicle is stopped, acceleration time will also depend on the delay between the power
demand signal (depressed accelerator) and engine start and attainment of full power. The
potential for vehicle-to-vehicle variation in these factors further complicates the attainment of
fair comparisons between vehicles.
The following discussions lay out more of the modeling difficulties we encountered.
3.1.2.2 Engine Maps
Engine maps are representations of an engines fuel consumption or emissions
characteristics in which brake-specific fuel consumption or emissions are plotted as a function of
rpm on the x axis and torque [or brake mean effective pressure (BMEP)] on the y axis. What
appears on the fuel consumption map are iso-bsfc lines, that is, lines of constant brake-specific
fuel consumption; the maps look similar to topographic maps showing lines of constant
elevation. Similarly, the emission maps have iso-emissions lines with the same x and y axes. For
conventional vehicles, ADVISOR and other second-by-second vehicle simulation models
combine these maps with a transmission model (which determines where on the map the engine
will be operating for a given demand for power) and, for emissions calculations, a catalyst model
to estimate the instantaneous fuel consumption and emissions of the vehicle. For hybrids, maps
of electric motors, controllers, batteries, and/or generators are also necessary.
Having accurate engine maps is crucial to obtaining credible fuel consumption and
emissions estimates for hybrids and conventional vehicles, but obtaining adequate maps is far
from simple. Aside from the problem of scaling discussed above (Section 3.2.2.1), we
encountered the following problems:
1. Availability of emissions maps. A reasonable number of fuel consumption maps for
engines of various sizes and technology levels are available in the scientific literature
or informational literature prepared by engine and vehicle manufacturers. There are
very few maps of emissions available, however. We did have access to maps and data
from the Federal Test Procedure Revision Project (FTPRP) and the National
Cooperative Highway Research Program (NCHRP). These projects measured second-
by-second emissions and fuel consumption as the vehicles followed prescribed driving
cycles. This project stretched the measurement state-of-the-art, particularly with regard
to equipment response times and dealing with lags between engine changes and tailpipe
emission responses. Perhaps as a result of these issues, we found that our estimates of
fuel consumption using FTPRP and NCHRP data disagreed with published engine map
45
information for the same engines by a few percent, generally giving optimistic results.
Presumably, there are similar problems with the emissions maps, although we did not
have the same opportunities to compare these with manufacturers data. Despite these
problems, however, we had no other source of emissions data, and our emissions maps
were derived directly from the FTPRP and NCHRP emissions data.
2. Gaps in the available fuel consumption maps. Given the measurement difficulties
noted above with FTPRP and NCHRP fuel consumption data, we chose to use
manufacturers fuel consumption maps as the baseline for use in ADVISOR. However,
these maps provided accurate information only for positive load conditions, and then
only for moderate and higher load levels. Much of the time, however, engines operate
at low or even negative load, i.e., at idle and during braking or coasting (negative load
implies engine braking is occurring). Although the rate of fuel consumption at idle or
during braking or coasting is low, the time a conventional vehicle spends at these
conditions is so high during most cycles that the fuel consumed during these conditions
is important (e.g., between 15 and 20% of total fuel consumed during the EPA FTP
cycle). Because a large portion of a hybrids fuel economy advantage over a
conventional vehicle is due to the way it handles engine operations during low and
negative load (it turns the engine off or maintains engine output at an efficient level,
using any excess output to recharge the battery), accurately measuring energy use for
both CVs and hybrids during these periods is important to accurately gauging hybrids
fuel economy advantage. For this reason, we used FTPRP and NCHRP fuel
consumption data to fill in the low and negative load areas on the published maps. Note
that CVs use fuel during these conditions, whereas hybrids often do not (their engines
usually are turned off).
3. No accounting for engine transient effects. The published fuel consumption maps we
used are constructed by running the engine at constant speed and load, measuring fuel
consumption rates, and then repeating the measurements over the desired range of
speed and load. Thus, the maps measure steady-state fuel consumption and provide no
information on effects of engine transients in operations such as acceleration and
deceleration. This may be important to the modeled comparisons of hybrids and
conventional vehicles because some hybrid designs moderate or eliminate most engine
transients. On the other hand, accurate fuel economy results are obtainable for CVs
without accounting for engine transients, since validated simulations of conventional
vehicles have used the same type of steady-state engine maps that are used here for
simulating hybrid vehicles. At this time, we are not able to estimate the magnitude of
the effect of not accounting for engine transients.
4. Difficulty of modeling emissions. Emissions rates for hydrocarbons and carbon
monoxide can vary by thousands of percentage points as engine and catalyst conditions
change, demanding a level of precision in tracking these conditions that is difficult to
obtain in modeling. For example:
Using a one size fits all transmission model means that, however accurate the
emissions map, potentially large errors can be introduced by differences between
46
the rpm/torque points selected by the model and the actual points that would have
been selected by the physical transmission.
Emissions control state of the art changes quite rapidly, so available data are often
behind the curve.
Tailpipe emissions are a function of both engine out emissions and catalyst
efficiency; the latter is highly sensitive to temperature variations, variations from
stoichiometric exhaust conditions that may occur during deceleration, and other
conditions that are hard to predict and are likely to vary with changes in
manufacturer design strategy.
Runs of ADVISOR predicted that a small emissions spike occurs at hot or warm
restart of the engine. Some limited experimentation on a single engine/control
system combination found that such spikes were small for the time between engine
starts typical of hybrid operation. Control strategies for hybrids often demand that
the engine be turned off and on several times during a driving cycle, yielding
multiple hot or warm restarts. Thus, further experiments to ensure accurate
prediction of restart emissions may be important in modeling hybrid emissions.
However, the nature of the emissions reaction to restarts will depend on the
behavior of the catalyst and the precise details of the engine and emissions control
strategy and hardware (aside from the catalyst, monitors and computer controls may
also be important). These details are not likely to be known for new or hypothetical
hybrid vehicles; control system data are far more proprietary than engine maps.
We do note that the 1999 Prius hybrid is as clean as many clean gasoline vehicles,
so the aforementioned spikes do not appear to be a problem for this vehicle
(Santini et al. 1999).
We concluded that the emissions estimates produced by our ADVISOR runs probably were an
order of magnitude less reliable than the fuel consumption estimates. We have decided in this
report to present only the fuel economy results until the reliability of the emissions estimates
made with ADVISOR has at least been verified by ongoing second-by-second emissions
testing at ANL.
3.1.2.3 Adequacy of Transmission Modeling
At every moment of the driving cycle, the transmission determines the operating point of the
engine, that is, the combination of BMEP and rpm that will produce the required load. This
means that an accurate transmission model is needed to provide accurate estimates of both fuel
consumption and emissions, since these are functions of where on the fuel consumption and
emissions maps the engine is operating.
To test the validity of ADVISORs transmission model, we compared data clouds (plotted
second-by-second data points that show where the engine is operating across the entire driving
cycle) for vehicle dynamometer tests from the Federal Test Procedure Revision Program
(FTPRP) and for modeled results from ADVISOR. We discovered that the original version of
47
ADVISORs 5-speed transmission model tended to put the modeled engine into a portion of the
maps where real engines do not operate. Figure 3-1 shows a fuel consumption map for the Saturn
1.9-L SOHC engine, with the shaded area illustrating those portions of the map where the
FTPRP-tested engine operated. The ADVISOR-modeled engine often operated in region C on the
map, a lower efficiency region outside the shaded area, on the same driving cycle tested in the
FTPRP. The differences between the actual tested engine and the ADVISOR-modeled engine
apparently occurred because the ADVISOR transmission model emphasized increases in engine
rpm to obtain higher power, whereas a real conventional vehicle transmission emphasizes
increases in torque to obtain more power. In other words, where a real transmission would
increase power by pushing the engine operating point up in the figure, the early ADVISOR
transmission simulation pushed the operating point to the right, into region C.
36
Figure 3-1 Saturn 1.9-L Fuel Use Map with FTPRP Coverage
36
Region A is an area of the map that empirical information implied cannot be reached by a conventional
transmission. The Toyota Prius, with a unique continuously variable transmission, can reach region A.
Region B represents deceleration. Point I represents idle.
48
After the Hybrid Electric Vehicle Technology Assessment (HEVTA) project developed this
evidence and developed a basis for appropriate changes to the model, NREL, ADVISORs
developer, incorporated recommended revisions to the shift logic of the transmission model in
ADVISOR. The revised transmission model now reasonably reproduces a real transmissions
behavior in a conventional vehicle. This project also provided funding for development of the
automatic transmission model added to ADVISOR. The transmission model was developed by
Dill Murrell.
3.1.2.4 Zero to 60 mph Acceleration Times
Zero to 60 mph acceleration time represents a key factor both in defining vehicle
performance, to ensure that vehicle-to-vehicle comparisons are truly between equivalent vehicles,
and in sizing the powertrain correctly.
37
Consequently, it is extremely important to be able to
consistently and accurately translate powertrain (and other vehicle) characteristics into Z60 times
and vice versa.
We encountered significant problems with measuring and estimating Z60 time. First, there
appears to be no industry standard for physically measuring Z60 time, so that different testing
organizations consistently obtain different Z60 values for identical vehicles. The differences are
particularly stark for manual transmission cars, which is not surprising since the degrees of
freedom afforded the drivers are that much greater than with an automatic transmission. Second,
estimating Z60 time requires a complex calculation with several judgment inputs, so that
estimated Z60 time values appearing in the literature should not be used without understanding
the underlying test protocols. Two vehicles of apparently identical performance, as indicated by
estimated Z60 times obtained under different test protocols from different sources, would
probably perform quite differently if tested in the field.
Z60 measurements. We examined Z60 test times from Consumer Reports, Car and Driver,
and Daniel Herauds Carpoint website, and found significant differences among the three
sources. The largest differences were between manual transmission cars tested by Consumer
Reports and Car and Driver, with Consumer Reports test times about 17% slower than Car and
Drivers. For higher performance cars with Z60 times of about 6 seconds, Car and Drivers tests
brought the cars to 60 mph about 1 second faster than Consumer Reports. The gap widened to
about 2 seconds for Z60 times approaching 12 seconds.
38
37
With regard to powertrain sizing, Argonnes baseline method has been first to size the engine to achieve
vehicle minimum continuous gradeability requirements of 6.5% grade @ 55 mph for 20 minutes, and
then to size the battery and electric motor to achieve the Z60 time goal (8, 10, or 12 seconds).
38
These values reflect plotted statistical fits of recent (1994-1998) values of Z60 time versus inertia
weight/peak power rather than compared tests of identical vehicles, but the results resemble an earlier
examination of test results from Consumer Reports and enthusiast magazines for identical vehicles
(Santini and Anderson 1993).
49
A comparison of test results for automatic transmission-equipped cars between Consumer
Reports and the Heraud website show considerably smaller differences, which is not surprising
given the fewer degrees of freedom available with an automatic transmission.
There are limited performance tests for hybrids available. EPA conducted Z60 tests on the
Japanese version of the Prius during extensive testing for fuel economy and emissions
performance (Hellman et al. 1998). EPA found that the Japanese Priuss Z60 times varied
significantly with the state-of-charge of the battery. With the battery fully charged, this Priuss
Z60 time was about 14 seconds. After deliberately discharging the battery by coasting down
rather than braking, thus avoiding regenerative braking, the Z60 times increased gradually to
about 20 seconds or so, with the 20-second Z60 probably about the acceleration capability of the
engine alone, without motor assist.
Car and Driver conducted similar tests of the Japanese Prius, at full and zero battery state of
charge (SOC). More recently, Consumer Reports and Toyota have published Z60 test results for
the U.S. version, presumably at full SOC. Finally, both Car and Driver and Consumer Reports
have tested the Insight, with Car and Driver testing it at varying states of charge.
Figure 3-2 plots the Z60 times for both conventional and hybrid vehicles from Consumer
Reports, Heraud (conventional only), and Car and Driver as a function of the vehicle power-to-
weight ratio, and adds a standard EPA equation of the relationship (for conventional drivetrains)
for comparisons sake. The figure also shows some results from ADVISOR runs, for
conventional vehicles. These tests confirmed that ADVISOR predicted slower acceleration with
automatic transmissions than with manual transmissions, though the difference may not be quite
as great as in Consumer Reports tests.
Modeling Z60 time. We also discovered that alternative models (and methods) we had been
using to estimate Z60 time for identical vehicles gave different estimates. We had been using two
models, ADVISOR and Hybrid Electric Vehicle Component Sizing and Vehicle Cost model
(HEVCOST), to estimate either Z60 time given vehicle characteristics, or drivetrain component
power given required Z60 time (and gradeability). Our initial versions of HEVCOST and
ADVISOR were close in predictions of Z60 times for series HEVs (within ~ 0.5 second
difference) but were far apart in Z60 predictions for conventional vehicles and parallel hybrids
with HEVCOST predicting significantly faster Z60 times (or lower power requirements to obtain
the same Z60 time). The differences in results led to careful reexamination of both approaches.
After these evaluations, we made changes to both models, bringing them closer together in their
50
5
7
9
11
13
15
17
19
21
40 50 60 70 80 90 100 110 120
W/kg (curb wt. + load)
Z
6
0
t
i
m
e
(
s
e
c
)
CU / AT fit
CU / 5spd fit
C&D / 5spd fit
EPA equation
Heraud AT fit
ANL Cost Model as of 12/21/98
Advisor / 5spd
Advisor / AT
Prius (Japan): C&D @ Full & No SOC
Prius (US): CU & Toyota
Insight C&D @ Full, Half, and No SOC
This point for CU @ full
SOC
Other points from C&D
U.S. Prius (Full
SOC)
CU = 12.6,
Toyota=12.7
Japan Prius
No SOC
Full SOC
CV
HEV
Figure 3-2 Plots of Measured and Estimated Z60 Time vs. the Vehicle Power-to-Weight
Ratio (watts per kilogram, curb weight plus load) for Conventional Vehicles and Hybrids
estimates, and better matching them to measured Z60 times. Some of the obvious areas of
concern in estimating Z60 time are:
For hybrids, is the engine on or off at start? The need to turn the engine on will cause a
delay in reaching full power, although the electric motor can launch the vehicle (at lower
acceleration) even with the engine off. ADVISOR assumed the engine was on in series
HEV calculations, but off in parallel HEV calculations; in its original version,
HEVCOST assumed the engine is on for both types of hybrids. In all likelihood, hybrids
will be designed to turn the engine off at stop, at least for a default mode, to conserve
fuel.
39
However, hybrid designs may allow drivers to override the engine off feature.
This raises the question of whether announced Z60 times will reflect normal (engine off
at stop) driving conditions or competitive (engine on) conditions. Further, if the engine is
on, different testing organizations testing a manual transmission HEV may use a
different starting rpm. We suspect that Consumer Reports starts its tests at idle and Car
and Driver starts at higher rpm. The relative positions of the best Car and Driver test of
39
However, for effective emissions control, after first start-up the engine probably will not turn off until the
catalyst reaches lightoff temperature; the Prius engine behaves this way.
51
the Insight, and the single Consumer Reports test of the Insight (Figure 3-2) are
consistent with this expectation.
For manual transmission cars, how long are the transmission shift delays? This is a
function of driver and transmission capability. Vehicle simulation models are likely to
have a single transmission model for all vehicles. The ADVISOR model does allow the
analyst to specify shifting delay, gear ratios, final drive ratio, and torque converter
characteristics; any good model should do so.
Does the model consider wheel slip? For full power acceleration, wheel slip can add
significantly to Z60 times, since tire/pavement adhesion limits affect the amount of
power that can be used for acceleration. Note that the type and size tires used are a
critical determining factor. As part of our model verification and calibration process, we
modified the wheel slip submodel in ADVISOR, increasing the adhesion limits and
reducing achievable Z60 times.
Are the transmission shift points always at the redline? Note that a typical Z60 test will
have one or two upshifts, with the second upshift often occurring quite close to 60 mph
(so that small differences in the shift point can place the second shift inside or outside of
the Z60 test); the difference in delay time between one or two shifts is not insignificant.
In an actual test run, an aggressive tester might avoid the second shift by leaving the
transmission in second gear and briefly exceeding the redline.
Is engine inertia accounted for? At startup (unless the test driver races the engine in
neutral) and after each shift, there is a small delay for the engine to reach the high rpm
required for maximum power. We added an inertial component to our early version of
ADVISOR to improve the accuracy of its Z60 calculation; the effect was to increase
slightly estimated Z60 times.
3.1.2.5 Hybrid Design and the Impact of Cost Considerations
Our modeled comparisons of equivalent conventional and hybrid vehicles have the
disadvantage that the hybrid vehicles design is based on a simplistic set of performance targets
and design rules rather than the sophisticated design process that would actually occur in
developing a commercial vehicle (for example, accounting for cost and a range of performance
requirements including start-up acceleration, passing acceleration, top speed, shift feel, etc). In
this study, the base vehicles are designed to meet simple Z60 and 55 mph gradeability targets,
with the sizing of drivetrain components generally obeying simple rules: engine sized to meet the
gradeability target; battery and motor sized to provide the additional power needed to meet the
acceleration target. These rules yield a vehicle that superficially performs in a manner similar to
a conventional vehicle and whose overall performance is likely to be acceptable to most users.
However, the modeled vehicle design lacks two important considerations:
1. The fuel economy and emissions outputs have not been used to redesign the vehicle in
an iterative fashion. It is quite possible that other design rules could produce a more
52
efficient and/or lower emitting vehicle that still performed adequately or, alternatively,
that strategies to hold emissions or fuel consumption down could diminish performance.
2. Cost estimates were similarly not used in the initial set of model runs to redesign the
vehicle. Thus, following the design rules, acceleration performance improvements were
obtained primarily by increasing battery and electric motor power capacity, even though
increasing engine power probably would be cheaper at U.S. fuel prices. The selected
approach does result, however, in better fuel economy.
The HEVTA has conducted some sensitivity analysis to allow for the second consideration, e.g.,
allowing engine power to exceed gradeability requirements (with reduced motor and battery
power requirements) to produce a less expensive vehicle. However, in general it must be
remembered that the hybrid designs studied to date are likely to be significantly different from
eventual commercial designs.
3.1.2.6 Fair Treatment of Time and Context
A fair treatment of the hybrid/conventional comparison will ensure that assumptions about
technological progress and context will apply to both drivetrains. Thus, if the hybrid is assumed
to be equipped with an advanced, efficient battery and electric motor that might not be available
until 2005, comparing it to a 1999 conventional vehicle introduces a bias against the CV. It
would be fairer to the comparison to assume improvements to the conventional drivetrain that
could be expected to be made by 2005, although this clearly introduces additional uncertainties.
The likely result would be that the relative fuel-economy advantage of the hybrid will shrink,
because advanced engines and transmissions will reduce some of the same losses that are
targeted by hybridization.
A more subtle issue is context: if the hybrid is commercially feasible despite its higher cost,
this probably means that, in the scenario being modeled, fuel economy is valued highly. This
might mean, for example, that turning the (conventional) engine off during zero road load
situations (braking, coasting, standing) would be acceptable to consumers. Turning the engine off
during stops, braking, and coasting is a typical strategy for a hybrid and a significant source of its
fuel savings; if the conventional drivetrain could use the same strategy, this would narrow the
fuel consumption gap between the two. But if the design for idle-off involved replacing the
starter and alternator with an integrated motor/generator and using a larger battery, this might
stretch the traditional meaning of conventional drivetrain. Such a design could also be
considered an extreme example the mildest of mild hybrids.
3.1.2.7 Calculating the Energy Use of Grid-Connected Hybrids
Computing the energy use of grid-connected HEVs is complicated by their dual mode of
operation and potential for recharge from the electrical grid. For grid-independent HEVs,
computing energy use is simply a matter of modeling fuel use over selected driving cycles and
reporting vehicle fuel use either separately for each cycle or for a time period consisting of
several cycles of different types, strung together. Either way, this is a straightforward exercise,
with all energy supplied by a single fuel stored onboard. That computing the energy use of a
53
grid-connected hybrid is not straightforward can be demonstrated by following the course of a
days travel (Figure 3-3):
We can assume that the vehicle begins the day as an electric vehicle (EV), with the
battery fully charged (e.g., at 0.95 SOC) and the engine not operating. This may not
always be the case, since vehicle operators might want to save the vehicles EV range
for later in the trip (or day) when the vehicle might have to operate in an urban area open
only to zero emission vehicles. Some European cities have established zero emission
zones, and conceivably some American cities might do the same. However, where
ozone reduction is the primary focus, reducing early morning emissions requiring
initial use of the vehicle in zero emissions mode seems most likely.
Assuming an EV range R, if the vehicles total travel is less than R miles, the vehicle
will operate all day on battery power and end the day at a battery SOC greater or equal to
the minimum allowed (e.g., 0.2 SOC, which is the value we have assumed in our
analysis). If the vehicle is recharged at night, all of that days miles ( R) may be said to
have been powered from the grid.
Figure 3-3 A Days Driving for a Grid-Connected Hybrid Vehicle:
SOC vs. Time or Distance
54
If the vehicles travel on that day is greater than R, after R miles the engine will turn on
and the vehicle will operate in hybrid mode. For a time after the engine starts, the engine
will work to recharge the battery until it reaches a predetermined state-of-charge range
selected to allow both long battery life and high potential for grid recharging.
40
After this
range is reached, the vehicles control system will seek to keep the battery SOC within
this (narrow) operating range, to extend battery life. When the battery is recharged
overnight, the grid can provide only that electricity needed to raise the batterys charge
from the SOC when the vehicle was turned off (either somewhere between 0.2 and the
bottom of the operating range if the vehicle is turned off before the operating range is
reached, or somewhere within the operating range if the vehicle is turned off later) to the
fully charged state. On these days, the grid will have provided the energy to power some
number of miles less than R, with the rest powered by the vehicles onboard fuel.
Uncertainty about the SOC range likely to be chosen for normal HEV operations introduces
difficulty in estimating the amount of grid electricity that a grid-connected hybrid can use. As
noted above, on trips that significantly exceed the hybrids EV range, the amount of recharging
will depend on the difference between the SOC at the end of the trip (somewhere in the SOC
operating range) and the maximum SOC, say 0.95. The lower the SOC operating charge range,
the more potential there is in the battery for recharging. Thus, the choice of operating range is
likely to be a trade-off between maximizing battery longevity, which might require a fairly high
SOC range (e.g., 0.5-0.6
41
), and maximizing the amount of grid electricity used, which would
require a low range (e.g., 0.2-0.3). At this time, we do not know enough about the quantitative
effect on battery longevity of different operating SOC ranges to provide a reliable assumption
about the range likely to be chosen by manufacturers. It is our understanding, however, that a
wide swing in SOC from almost fully charged (0.95) to almost fully depleted (0.2) will have a
highly detrimental effect on battery life compared to a modest 0.5-0.6 swing.
Individual driving characteristics of grid-connected hybrids present another uncertainty
about their energy use. A simple example will illustrate the dilemma. If a grid-connected HEV
were to travel 10,000 miles per year, it would average about 27 miles per day. If its EV range
were 30 miles and its travel extremely regular, for example, if it were a mail delivery vehicle
with a 27-mile daily route, or primarily a commuting vehicle with a 27-mile round trip, it might
rarely be forced out of its EV mode and could fuel most of its miles from the grid. On the other
hand, if its travel were highly irregular, with a mix of short travel days (during which it was
idle or traveled only a few miles) and days when it was used intensively, with daily distances
well beyond its EV range, half or more of its travel could be fueled by its onboard fuel stores.
40
Maximizing grid recharge means selecting as low an SOC operating range as possible (but still avoiding
SOCs below the point where specific power experiences sharp declines, about 0.2 SOC in the NiMH
batteries that we simulated), e.g., 0.2-0.3 SOC. Maximizing battery life may imply a higher SOC
operating range, but we are not aware of data on the interplay of battery life and operating range.
41
The Prius HEV uses an SOC range in this vicinity. ANL tests show that battery SOC stays within a range
of 0.55 and 0.60 on the Japanese 10.15 cycle (personal communication, Aymeric Rousseau, Argonne
National Laboratory).
55
Consequently, understanding the usage patterns of grid-connected HEVs will be crucial to
estimating the nature of their energy use.
In other words, a grid-connected HEV will use both grid electricity and fuel to power its
miles, with the split between the two determined in large part by the type of driving being done,
the vehicles electric range, and the operating SOC range. A vehicle that travels short distances
daily might obtain all its energy from the grid; one that travels long distances might obtain
considerably less than half its energy from the grid.
HEV energy use can be modeled from the daily miles traveled data available from the
Nationwide Personal Transportation Survey (FHWA 1997). Figure 3-4 (cumulative share of
vehicles traveling vs. total miles traveled daily) shows that, on an average day, about 40% of
vehicles driven travel less than 20 miles.
42
We might assume that HEVs conform to the daily
travel distance distribution reported by the survey. However, using this data implicitly assumes
that the travel patterns of grid-connected HEVs will be identical to the patterns of all vehicles,
and this seems unlikely. Grid-connected HEVs will be more expensive than both CVs and grid-
independent HEVs in initial costs and, at U.S. gasoline and electricity prices, probably in life
cycle costs as well, and their performance is somewhat limited in EV mode. They are likely to be
purchased by individuals who either require, or value highly, their EV capability, and these will
not be average drivers. Although we can, and will, speculate about their tripmaking
characteristics,
43
computing electricity and fuel use involves adding another uncertainty to the
already large number of uncertainties present in our other calculations of HEV energy use.
Assuming that the full range R can be recharged electrically when the days travel distance
exceeds R (that is, ignoring the previously discussed loss of rechargeable distance that occurs
after the HEV mode commences), HEVs with a 20-mile EV range could recharge 40% of all
miles traveled electrically (this potential was first recognized by Reuyl and Schuurmans 1996).
This percentage of recharge may be reduced substantially when engine-induced recharge is
considered, with the amount depending on the minimum SOC and the HEV operating SOC
range. In all likelihood, our accounting for engine recharge (assuming a 0.5-0.6 SOC operating
range) would reduce the 20-mile HEV to about 25-30% electrification. If battery longevity can
be maintained with a very low operating SOC range, then this percentage would move upwards
toward 40%.
42
Note that the graph represents only a snapshot of what happens on an average day, and on a different
day, it is quite possible that a different group of vehicles will be traveling less than 20 miles. In other
words, the graph does not mean that 40% of vehicles travel less than 20 miles per day as an annual
average (that is, 6,300 miles/year); it is more likely that far fewer vehicles travel this little over the
course of a year. Strictly speaking, the graph means only that the average driver drives less than 20 miles
about 4 days out of every 10 that he or she drives.
43
We would guess that purchasers of grid-connected hybrid vehicles would spend more-than-average time
in congested urban driving conditions (where HEVs achieve maximum benefit/mile) or at least in non-
highway conditions, drive more miles than most such drivers, and regularly return to a location where the
vehicle can be recharged.
56
0%
20%
40%
60%
80%
100%
0 50 100 150 200 250 300 350 400 450 500 550 600
Miles per vehicle per day
C
u
m
m
u
l
a
t
i
v
e
s
h
a
r
e
o
f
v
e
h
i
c
l
e
s
Figure 3-4 Cumulative Share of Vehicles Traveling vs. Miles Traveled, from
the Nationwide Personal Transportation Survey
3.2 Cost Modeling
The Hybrid Electric Vehicle Component Sizing and Vehicle Cost model, HEVCOST, is a
direct extension of the conventional and electric vehicles manufacturing and retailing cost
analysis described by Cuenca, Gaines, and Vyas (2000) and Vyas et al. (1998). A few aspects
were adapted for hybrid electric vehicles (HEVs). The analysis relies on first identifying those
vehicle components that would be common between a CV and a hybrid electric vehicle, and then
estimating the CV retail price share of these common components. The analysis then develops a
separate estimate of the cost of those components unique to the HEV primarily the powertrain.
Using a relationship between vehicle component manufacturing cost and retail price developed
by the study team, the retail price of the HEV components (mostly the powertrain) are
determined and added to the price of the shared components to yield total vehicle price.
Argonne developed a separate procedure (described in Appendix C) for estimating
powertrain component power requirements and mass, and cost equations were derived for the
powertrain components as a function of the component power ratings. The resulting estimates
serve as a useful initial value for the ADVISOR model, which then iterates to a more exact set.
The cost estimation procedure is designed to be applied to high-volume production of hybrid
electric vehicles. It assumes that the body and chassis components of a hybrid electric vehicle
57
will be very similar to a conventional vehicle and will be mass-produced by an original
equipment manufacturer (OEM).
3.2.1 Common Components Retail Price Equivalent
The cost analysis separates the conventional vehicle into four groups: (1) body, (2) engine,
(3) transmission, and (4) chassis. The retail price contributed to electric vehicles by each group
was determined by Cuenca, Gaines, and Vyas (2000), who also classified each component or
group of components as fully common, partially common, or not common between conventional
and electric vehicles. Subsequently, that analysis was extended to apply to hybrid electric
vehicles and is reported in Table 3-1.
With the exception of the instrument panel and the heating, ventilating, and air-conditioning
(HVAC) system, the body group is expected to remain practically the same between a CV and an
HEV. The HEV instrument panel will have specific displays for new components and HEV
operation. Its HVAC system ducts and blowers will be very similar to those of a CV, but the
system would be powered and configured differently. Although the initial HEVs are expected to
have an internal combustion engine (ICE) power plant, the entire engine group is identified as
not common because proposed HEV engine systems are different from those used in the CVs.
Aside from having lower power ratings than their CV counterparts, HEV engines are not likely to
have a separate alternator and a starter. They may have one alternator/starter or they may use,
when one is available, a generator for starting the engine. HEV engines may operate within a
narrower range of speeds than a CV engine and may use combustion cycles, such as the Atkinson
cycle used by the Toyota Prius, that a CV engine would not.
The transmission group, too, is marked not common. The parallel HEVs transmission
must coordinate two power sources, an ICE and an electric motor. Several novel approaches are
being developed for this configuration, much different from the present CV transmissions. In the
series configuration, the motor could be connected to a simple gear drive to transmit its power,
although some applications might require a two-speed transmission. In either configuration, a
separate cost estimate is required. Within the chassis group, several subgroups are marked
partially common. The exhaust system, fuel storage, and fluids subgroups are likely to be
smaller and perhaps different from the CVs systems. The steering and brake subgroups may
differ in their power source and design. The chassis electric system may use a different power
source and voltage, depending upon the OEM design choices.
The analysis discussed in this report assigns shares of retail price to each group of the CV
components for a midsize car a four-door sedan with a four-cylinder engine, automatic
transmission, and air-conditioning (HEVCOST also can evaluate minivan and subcompact
vehicles). Table 3-2 provides a summary of the allocation by major groups.
58
Table 3-1 Common and Dissimilar Components of CV and HEV
CV Groups and Subgroups
Fully
Common
Partially
Common
Not
Common
Body Group
Body-in-white X
Paint and Coatings X
Glass X
Interior Body Trim X
Exterior Body Trim X
Seats X
Instrument Panel X
Restraint System X
Body Electrical Components X
Heating, Ventilating, Air-Conditioning (HVAC) X
Engine Group
a
Base Engine X
Emissions Control X
Engine Accessories X
Engine Electrical Components X
Cooling System X
Transmission Group
b
Transaxle X
Clutch and Actuator X
Transmission Controls X
Chassis Group
Frame X
Suspension X
Steering X
Brakes X
Exhaust System X
Fuel Storage X
Final Drive X
Wheels and Tires X
Bumpers, Fenders, and Shields X
Chassis Electrical Components X
Accessories and Tools X
Fluids X
a
Although the HEV has an engine, the entire engine group will be designed differently.
Hence it is treated as not common.
b
An HEV transmission could range from a continuously variable transmission to a set of
reduction gears.
59
Table 3-2 Allocation of MSRP to Midsize Car Groups and
Common Components
Vehicle Group CV Price Share (%)
Share Common
to HEV (%)
Body group 28.46 28.46
Engine group 21.53 0.00
Transmission group 5.03 0.00
Chassis group 23.94 22.78
Vehicle assembly 21.04 21.04
Total 100.00 72.28
With the assumption that the common body and chassis components will be mass produced
and assembled the same way for both CV and HEV, the shares also provide an estimate of the
retail price of these components in a midsize HEV. The analysis adds the cost of powertrain
components to this common component price to yield an estimate of the potential HEV retail
price.
3.2.2 Allocation of Indirect Costs
A vehicles retail price includes costs that are not associated directly with its manufacturing,
but were incurred in other areas of vehicle manufacturing and retailing. These costs include
research and development, engineering, depreciation and amortization, corporate overhead,
retirement and health benefits, vehicle distribution, advertising, dealer support, and profit. A
vehicles retail price structure can be broken down as shown in Table 3-3.
The cost of manufacturing, including assembly, typically accounts for about 50% of the
retail price. This relationship can be extended to individual components. The analysis doubles
the factory gate cost of a component manufactured by an OEM to arrive at its contribution to
the vehicle retail price. Several components of the HEV, such as motor, motor controller and
power electronics, and generator, are not likely to be produced by an OEM, however. These
components would likely be procured from independent suppliers who would include warranty,
R&D and engineering, and depreciation and amortizing costs in their component prices. These
non-OEM HEV powertrain component prices are multiplied by a factor of 1.5 to arrive at their
contributions to the vehicle retail price. The two multiplication factors are similar to those
resulting from two other methodologies (Borroni-Bird 1996; OTA 1995).
The battery pack is a unique component. Though most likely to be procured from an
independent supplier, it will require only assembly and testing. We assume that a multiplication
factor of 1.15 applied to the battery pack cost is the minimum feasible cost multiplier for this
expensive HEV component. It is lower than the 1.50 factor applied to the other outsourced
components, and this admittedly low factor is used in the analysis.
60
Table 3-3 Vehicle Manufacturing and Retail Price Structure
Retail Price Share (%)
Major Cost Category Cost Subcategory
By Major
Category
By Sub-
category
Production 67.0
Manufacturing 50.0
Warranty 5.0
R&D and Engineering 6.5
Depreciation and Amortization 5.5
Selling 23.5
Distribution 20.0
Advertising and Dealer Support 3.5
Administration and Profit 9.5
Corporate Overhead 5.0
Retirement and Health Benefits 2.0
Gross Profit 2.5
Total 100.0 100.0
Bear in mind that there are substantial differences among vehicle manufacturers in
allocating indirect costs and determining retail prices. The prices derived by this analysis should
be viewed as representative of average prices in a market operating in a manner similar to todays
retail market, with the implicit assumption that manufacturers do not need to artificially
subsidize HEVs in order to make them more acceptable to a risk-averse buying population.
3.2.3 Component Cost Information
The model uses a default midsize car cost of $22,500 for a vehicle capable of accelerating
from 0-60 mph in 10 seconds. The resulting retail price equivalent share of the common
components (between the conventional car and the HEV) is $16,270. The cost model then adds
to this the estimated retail price contribution of the power unit and associated system, motor,
inverter and power electronics, optional generator, transmission and/or gear drive, battery pack,
system control, and other components for HVAC and electric auxiliaries.
The database containing the required component cost and performance information was
developed using data from published sources coupled with technical judgment on learning and
production volume-related improvements. The mass and power computation procedure uses the
performance information and determines the ratings of each powertrain component. The cost
modeling procedure uses the resulting power ratings and a set of equations to compute cost of the
powertrain components. The information presented here is from the default database. The cost
model uses four points in time. The first year assumes an annual production of 25,000 vehicles.
The next three points in time represent rising annual production levels of 50,000-100,000,
100,000-200,000, and 250,000 or more. Thus, the cost estimation procedure does not estimate a
potential HEV price for an initial production phase when several components would be
manufactured in small quantities using less-than-optimal methods. In the description below, we
use cost information for the 50,000-100,000 production level.
61
The component cost data reported below are manufacturing costs at the OEM factory
gate. The final price to the consumer is computed through application of the above-mentioned
multiplication factors.
3.2.3.1 OEM Manufactured Components
The vehicle body, power unit and associated system, transmission and/or gear drive, system
control, and HVAC and other components are assumed to be manufactured by the OEM.
Table 3-4 shows fixed and variable terms used by the cost model.
Table 3-4 Methodologies Used for Estimating OEM-Manufactured
Component Costs
OEM Manufactured
Component
Cost Function
Type
Minimum
Power (kW) Fixed Cost ($)
Variable Cost
($/kW)
Power unit and system Linear
a
394 24.9
Transmission: parallel Step 50
b
270 5.0
Motor gear drive Step 50
b
85 1.7
System control Fixed 202
HVAC and other Fixed 250
a
A minimum 20 kW power unit size assumed.
b
The variable cost is zero up to this rating.
The second column in the table, cost function type, indicates the method of cost
computation. The methods are: (1) linear, in which fixed plus variable cost per kW values are
used; (2) step, in which the cost remains unchanged until a minimum power rating listed in the
third column is reached and, thereafter a variable cost per each additional kW is added; and
(3) fixed, in which the cost does not change.
3.2.3.2 Outsourced Electric Powertrain Components
The cost model assumes that all electric powertrain components, motor/generator and
inverter/controller with power electronics, are procured from outside suppliers. We used data
supplied by a DOE contractor to develop a set of linear functions (Duleep 1998). Table 3-5
shows fixed and variable terms used by the cost model.
Table 3-5 Fixed and Variable Terms Used for
Estimating Electric Drive Component Costs
Outsourced
Component
Fixed
Cost ($)
Variable Cost
($/kW)
Motor/generator 200 13.7
Inverter & power electronics 425 19.0
62
3.2.3.3 Outsourced Battery Pack
The battery cost numbers presented in this report are for the nickel metal hydride (NiMH)
battery only. At the time of developing the HEV cost methodology, NiMH batteries had just been
introduced in both Toyotas RAV-4 electric vehicle and its Prius HEV. The characteristics of
these EV and HEV battery packs are very different. The EV battery pack has high specific
energy while the HEV battery pack has high specific power. Since the motor in a power-assist
(grid-independent) HEV is used intermittently and must be capable of producing high power for
short periods of time (e.g., during maximum acceleration), its battery pack should be optimized
for high power. Consequently, we assumed that the battery packs for grid-independent hybrids
would be similar to the Priuss battery back. We also analyzed dual-mode (grid-connected)
HEVs, which required battery packs with high energy storage as well as power output. For this
vehicle type, we assumed an intermediate level of the nickel metal hydride battery with energy
and power density characteristics in between the RAV-4 type and Prius type batteries. For
estimating cost per kilowatt-hour, we used a report by the California Air Resources Board
(CARB) Battery Advisory Panel (Kalhammer et al. 1995) and an ANL Delphi study (Vyas et al.
1997) for guidance. The characteristics and costs for the three batteries are shown in Table 3-6.
Table 3-6 Nickel Metal Hydride Battery
Characteristics and Costs
Battery
Type
Specific
Power (W/kg)
Specific Energy
(Wh/kg)
Cost
($/kWh)
High power 520 46 639
Intermediate 350 52 567
High energy 184 73 426
We note that the newly introduced U.S. version of the Prius has a significantly improved
battery pack with specific power of 880 W/kg (Toyota 2000), 75% higher than the previous
version. The cost of the new pack is reported as half of the previous version.
44
3.2.4 Modeling Concerns
Projecting the costs and retail prices of future vehicles and vehicle components has tended
to be problematic even for conventional vehicles, and there is substantial controversy about how
much hybrid vehicles will cost. Obvious problem areas for this analysis include:
1. Direct information on component prices to original equipment manufacturers is rarely
available; the OEMs generally hold this information as confidential. In addition, the
manufacturers treat the prices they pay to component manufacturers as confidential
information. Another potential source of information, retail prices for particular
components, may be available only in limited situations, because many components of
interest are integral to the vehicle and are not separately priced. Even when prices are
44
Nakamura, N., Toyota Motor North America, Inc., personal communication.
63
available, they often do not reflect actual cost plus profit, but instead reflect the
manufacturers estimate of how the market values the components. Thus, the existing
price data are of limited help in supporting cost projections.
2. Hybrid engine costs are problematic for two reasons: first, they are extrapolations from
larger, more powerful engines designed for CVs; and second, hybrid duty cycles are
different from CV cycles, so the design of a purpose-built hybrid engine will be
different from that of an engine built for a CV. In the Prius, for example, Toyota has
restricted the rpm level of the engine and has been able to cut costs by using some
lighter, less expensive components as a result.
45
3. The electric drive components presently are built in small quantities at high prices that
do not reflect savings associated with mass production. Further, there is a vigorous
R&D program aiming at improving (altering) the designs for key components, further
complicating the task of projecting costs and retail prices. The HEVCOST
methodology ignores current costs and estimates high volume costs by starting from the
materials in the components, building in normal markups. However, uncertainties
associated with changes in materials, production methods, and power densities
(yielding reductions in materials) remain.
46
4. Estimating fuel and electricity costs for both grid-independent and grid-dependent
hybrids run into a number of problems. We have modeled HEV onroad fuel economy
by multiplying the ADVISOR-estimated corporate average fuel economy (CAFE)
values by the current EPA onroad degradation factor of 0.85 (we used a lower value,
0.80, as the degradation factor for conventional vehicles; this reflects the shift in
driving to more congested conditions, a shift that will have less impact on HEVs).
However, the 0.85 degradation factor was developed for vehicles with conventional
drivetrains and is unlikely to be accurate for HEVs. We have adopted the 0.85
degradation factor as a default value for HEVs until better data are available. Another
concern for grid-dependent HEVs is uncertainty about the percentage of energy use
supplied by the grid.
3.3 References
Borroni-Bird, C., 1996, Automotive Fuel Cell Requirements, Proceedings of the 1996
Automotive Technology Development Customers Coordination Meeting, U.S. Department of
Energy, Office of Transportation Technologies, Washington, D.C.
Cuenca, R.M., L.L. Gaines, and A.D. Vyas, 2000, Evaluation of Electric Vehicle Production and
Operating Costs, Argonne National Laboratory Report ANL/ESD-41, Argonne, Ill.
45
Duleep, K.G., unpublished report to U.S. Department of Energy on Prius costs, 1998.
46
As noted above, the new Prius battery back is much smaller and lighter than the previous version,
implying that cost estimates based on the previous packs power density and weight would be much too
high.
64
Duleep, K.G., 1998, Briefing on Technology and Cost of Toyota Prius, Energy and
Environmental Analysis, Inc., Arlington, Va.
FHWA, 1997, 1995 Nationwide Personal Transportation Survey (NPTS): Compact Disk and
Code Book, U.S. Department of Transportation, Federal Highway Administration,
Washington, D.C.
Hellman, K.H., M.R. Peralta, and G.K. Piotrowski, 1998, Evaluation of a Toyota Prius Hybrid
System (THS), U.S. Environmental Protection Agency Technical Report EPA 420-R-98-006,
Ann Arbor, Mich., Aug.
Kalhammer, F.R., A. Kozawa, C.B. Moyer, and B.B. Owens, 1995, Performance and Availability
of Batteries for Electric Vehicles: Report of the Battery Technology Advisory Panel, California
Air Resources Board, El Monte, Calif.
Nakamura, N., Toyota Motor North America, Inc., personal communication.
OTA, 1995, Advanced Automotive Technology: Visions of a Super-Efficient Family Car, United
States Congress, Office of Technology Assessment Report OTA-ETT-638, Washington, D.C.
Reuyl, J.S., and P.J. Schuurmans, 1996, NEVCOR, Inc., Policy Implications of Hybrid-Electric
Vehicles, Final Report to NREL, Figure 11, April 22.
Santini, D., and J. Anderson, 1993, Determinants of Multiple Measures of Acceleration,
SAE 931805, SAE Transactions.
Santini, D., J. Anderson, J. He, S. Plotkin, A. Vyas, and D. Bharathan, 1999, Gasoline-Fueled
Hybrid vs. Conventional Vehicle Emissions and Fuel Economy, Annual Meeting of the Air and
Waste Management Association, St. Louis, Mo., June 20.
Vyas, A.D., H.K. Ng, D.J. Santini, and J.L. Anderson, 1997, Electric and Hybrid Electric
Vehicles: A Technology Assessment Based on a Two-Stage Delphi Study, Argonne National
Laboratory Report ANL/ESD-36, Argonne, Ill.
Vyas, A., R. Cuenca, and L. Gaines, 1998, An Assessment of Electric Vehicle Life Cycle Costs to
Consumers, Proceedings of the 1998 Total Life Cycle Conference, pp. 161-172, SAE
International Report P-339, Warrendale, Pa.
Wang, M.Q., S. Plotkin, D.J. Santini, J. He, L. Gaines, and P. Patterson, 1997, Total Energy-
Cycle Energy and Emissions Impact of Hybrid Electric Vehicles, Presented at the 14th
International Electric Vehicle Symposium, Orlando, Fla., Dec. 15-17.
Wipke, K.B., M.R. Cuddy, and S.D. Burch, 1999, ADVISOR 2.1: A User-Friendly Advanced
Powertrain Simulation Using a Combined Backward/Forward Approach, National Renewable
Energy Laboratory Technical Report NREL/JA-540-26839, Sept.
65
Section 4
Results of Vehicle Modeling
This section describes the results of the Advanced Vehicle Simulator (ADVISOR) and
Hybrid Electric Vehicle Component Sizing Cost and Vehicle model (HEVCOST) modeling of
vehicle energy use and cost (see discussion in Section 3 for a discussion of why we have chosen
not to present criteria emissions results).
4.1 Description of the Vehicles Evaluated
The HEVTA has focused on a midsized passenger car, primarily because this is the focus of
the current Partnership for a New Generation of Vehicles (PNGV) research and because the
results can be applied, with some caution and modification, to most of the light-duty vehicles in
the U.S. fleet. The current vehicles on which the HEVTA vehicles are based the Chrysler
Intrepid, Chevrolet Lumina, and Ford Taurus are assumed to be slightly downsized (in external
dimensions) for the timeframe of the analysis, 2005-2020. They are also assumed to have
advanced low rolling resistance tires and greatly improved aerodynamics.
For sizing, we have chosen the current Chrysler Cirrus and Chevrolet Malibu as having the
approximate external dimensions of a future midsize car that has undergone a process of interior
space and structural redesign accompanied by substantial weight reduction. The current cars
weigh about 1,418 kg (3,125 lb). By 2010, with an ultralight steel (ULS) body, these cars would
be expected to weigh about 1,255 kg (2,767 lb), with the body and chassis (basically, the car
without its drivetrain) assumed to contribute 73.5% of the vehicle mass, or 922 kg (2,029 lb).
When these vehicles are converted to hybrid electric vehicles (HEVs), their body and chassis
weight is assumed to increase by 5% to account for any additional structural strength needed to
support the battery pack and power-electronic components.
47
As discussed previously, to obtain a fair comparison between hybrid and conventional
vehicles, we have assumed that all the vehicles share the same body characteristics and similar
performance, measured as zero to 60 mph acceleration (Z60) time. All of the vehicles are
essentially identical except for differences in their drivetrains and the 5% structural weight
difference. The hybrids also must satisfy a gradeability requirement, although the conventional
vehicles have better gradeability due to their larger, more powerful engines. Table 4-1 shows the
basic body characteristics assumed for the 2010 and 2020 vehicle.
47
Note that this 5% factor is somewhat arbitrary, for two reasons: first, it is possible to design the battery
pack in such a way that it becomes an integral and supporting part of the body structure, thus reducing or
eliminating the need for added structural weight; and second, the size and weight of the battery pack and
electronic components, thus the added structural weight required, will vary significantly depending on
the vehicles performance requirements and on whether it is expected to be capable of operating all-
electrically, with some recharging from the grid.
66
Table 4-1 Midsize HEV Characteristics
Vehicle Characteristics 2010 2020
Body and chassis mass in kg (lb in parenthesis)
Ultralight steel body 969 (2,136) 959 (2,114)
Aerodynamic and rolling
Frontal area (square meters) 2.06 2.01
Coefficient of drag (Cd) 0.26 0.24
Coefficient of rolling resistance (Cr) 0.0075 0.0065
We examine a number of different types of hybrid design, including:
Parallel grid-independent full hybrid, with engine sized to provide the power to maintain
gradeability requirements, and battery/motor sized to provide adequate boost to the engine to
allow attainment of the Z60 requirement.
Parallel grid-independent mild hybrid, with less battery and motor power and more engine
power than the parallel grid-independent hybrid. Engine power is precisely midway between the
conventional vehicles (CVs) and full hybrids engine power, with battery and motor power
consequently sized to provide enough boost to satisfy the Z60 requirement.
Series grid-independent full hybrid, with engine and electric motor (continuous rating) sized for
gradeability, battery sized to provide boost for acceleration requirements, and electric motor
(maximum power) sized for full acceleration power requirements.
48
Parallel and series grid-dependent hybrids, designed to operate part of the time as full electric
vehicles, so battery and electric motor are sized to satisfy a separate electric vehicle (EV)
acceleration requirement, and the engine is sized to provide full HEV gradeability. For the
parallel hybrids we examined, this combination tends to allow especially rapid acceleration when
both engine and battery/motor are engaged simultaneously, because the battery and motor are
considerably larger than those found in a parallel grid-independent hybrid.
For the grid-independent vehicles, we examine power levels that will allow Z60 times of 8,
10, and 12 seconds. The original PNGV guidelines call for Z60 times of 12 seconds, but this
value has been outrun by changes in the fleet. The current average Z60 time for midsize cars is
10.5 seconds, vs. 11.4 seconds in 1994 when the guidelines were established (Heavenrich and
Hellman 1999). Trends of increasing power and decreasing Z60 times appear to be robust. For
grid-connected vehicles, we examine power levels that will allow all-electric Z60 times of 12,
14, and 16 seconds. Keeping all-electric Z60 times for these vehicles more modest makes sense
because the powerful and very expensive electric motors needed to allow faster acceleration
48
In other words, there are two separate sizing criteria for the electric motor. Although the gradeability
criterion requires less power than the acceleration criterion, the motors continuous power rating will be
lower than its peak rating. Therefore, actual motor size may be determined by either the acceleration or
the gradeability criterion, depending on motor design and the gradeability and acceleration criteria used.
67
would make these vehicles extremely expensive (they are already considerably more expensive
than the grid-independent hybrids).
The drivetrain components of both conventional and hybrid vehicles will likely be
substantial improvements over those available today. For example, it seems quite possible that,
by 2010, the engines driving both conventional and hybrid vehicles will be mature versions of
recently-introduced direct injection stratified charge (DISC) gasoline engines, direct injection
(DI) diesels (if emissions problems are solved), or other advanced technologies. Recent progress
in electrical drive technology implies that batteries, motors, and power electronics will have both
higher efficiency and higher specific power than those available today. All such developments
will increase the efficiency of hybrid vehicles, although use of advanced engines and
transmissions in conventional vehicles may shrink the relative efficiency advantage of the
hybrids (see Section 2.5.5).
Tables 4-2 and 4-3 show the assumed characteristics of the batteries and other hybrid
drivetrain components for 2010 and 2020. The study team decided not to try to simulate an
advanced engine, in part because appropriate engine maps were not available. The base engine,
from which the HEV and CV engines were scaled, is a 1994 version of a 63 kW (85 horsepower)
single overhead cam (SOHC) 1.9-L Saturn engine. The battery characteristics are based on the
results of a Delphi study conducted by ANL as well as published characteristics of the batteries
used in Toyotas Prius hybrid (Japanese market version
49
) and RAV-4 electric vehicles. The
motor, generator, and inverter characteristics are based on the performance of a Unique Mobility
SR218H permanent magnet motor and CA40-300L inverter, with assumed improvements over
time. The efficiencies shown in Table 4-3 are those attained during high-power acceleration (that
is, during peak loading of the components) and are used to determine the battery power capacity
needed to allow the vehicle to achieve acceleration goals. Future analysis will explore the effects
of changing the various component assumptions.
Table 4-2 Nickel Metal Hydride Battery Characteristics
High-Energy/
Low-Power Intermediate
Low-Energy/
High-Power
Battery
Characteristic 2010 2020 2010 2020 2010 2020
Specific power (W/kg) 184 203 350 386 520 573
Specific energy (Wh/kg) 73 79 54 58 46 50
Cost ($/kWh) 426 382 533 478 567 508
Cost ($/kW) 169 149 82 72 50 44
49
The version of the Prius introduced into the U.S. market has an upgraded battery pack attaining a specific
power of 880 W/kg, versus the earlier Japanese versions 500 W/kg. We based our battery on the earlier
Japanese version.
68
Table 4-3 Specific Power and Efficiency Values for HEV
Drivetrain Components
Component Type 2010 2020
Specific Power (W/kg)
Motor and generator Permanent magnet 1,300 1,400
Auxiliary power unit Gasoline 330 340
Motor with inverter 1,025 1,110
Transmission (parallel HEV) 1,320 1,360
Transmission (series HEV) 1,650 1,700
Efficiency (%)
Motor and inverter Permanent magnet 90 93
Battery (one-way) Nickel metal hydride 93 95
Transmission 93 95
4.2 Vehicle Control Strategy
Ideally, the computer onboard a hybrid vehicle will respond to the drivers power demands
by combining engine and battery power capabilities in an optimum manner, with optimum
defined as a predetermined function of energy use and emissions. For example, optimum might
be defined as minimizing fuel consumption with emissions held below Tier 2 limits. To
accomplish this, the computer would have to continuously examine the energy losses associated
with each path for tractive power (e.g., engine to transmission to differential to wheels, battery to
motor to reduction gear to differential to wheels) and select the most efficient combination that
delivered the required power while checking to ensure that emissions constraints were not
exceeded. The National Renewable Energy Laboratory (NREL), the developer of ADVISOR, has
examined such control strategies (Johnson 2000).
Up to this point, the Hybrid Electric Vehicle Technology Assessment (HEVTA) analyses
have used a simpler control strategy. For parallel vehicles, the motor is used to launch the vehicle
and provides all torque below a specified launch speed, with the engine turned off. Above that
speed, the engine is turned on and used as a load follower, generally satisfying all torque
requirements up to the engines maximum capability. The exception occurs if torque demanded
falls below a set limit, for example during a downhill coast or during braking, and the engine
shuts off. When the engine is turned on, the motor assists the engine only if the desired torque is
higher than the engines maximum capability, e.g., during high acceleration. During braking, the
motor provides all braking force up to its maximum capacity, feeding the electricity generated to
the batteries; the mechanical brakes provide any additional braking force needed. Finally, if the
battery state of charge (SOC) goes below a specified lower bound, the engine deliberately
increases its torque output above the level demanded by the tractional load and accessory power
load, with the excess torque passing through the motor (acting as a generator) to charge the
battery.
4.3 Driving Cycles
All of the vehicles are tested under a variety of driving conditions to evaluate HEV costs
and benefits under a range of conditions, identify potential market niches for HEVs, and evaluate
69
HEV and CV performance on the standardized government tests used to measure official
vehicle fuel economy. These driving conditions are represented quantitatively by driving
cycles.
Driving cycles are defined by second-by-second profiles of vehicle speed; Figures 4-1a-e
show five of the seven cycles examined in our analysis. Each profile determines the second-by-
second power requirements of the vehicle being tested, which in turn (in conjunction with
vehicle efficiency attributes) determine the vehicle operation regimes (rpm, torque, etc.) and
vehicle fuel economy and emissions performance.
Standard driving cycles are used for certifying motor vehicles for compliance with fuel
economy and emissions regulations. In the United States, emissions compliance for a light-duty
vehicle is measured by driving the vehicle (on a dynamometer) over a driving cycle called the
federal testing procedure (FTP, see Table 4-4). Under this procedure, the vehicle is first driven
through the FUDS (federal urban driving schedule) cycle, hot-soaked for about 10 minutes, and
then driven through the first 505 seconds of the FUDS. Emissions are collected in three separate
bags. The first bag covers the first 505 seconds of the FUDS, which represents cold-started
operations; the second bag covers the rest of the FUDS, which represents hot-stabilized
operations (the bag is switched with the engine running); and the third bag covers the repeated
(after the 10-minute hot soak) first 505 seconds of the FUDS, which represents hot-started
operations. Emissions of the three bags are weighted together and combined with the results of a
supplementary driving schedule recently added to the testing (see discussion below) to determine
if a given vehicle meets applicable emission standards.
For compliance with corporate average fuel economy (CAFE) standards, the fuel economy
of motor vehicles is measured under both the FTP and the Highway cycle. A composite
city/highway fuel economy is calculated by applying the formula
1/fuel economy = (0.55/FTP fuel economy) + (0.45/Highway fuel economy)
The FUDS and Highway cycles were developed in the early 1970s to represent urban and
highway driving, respectively. However, the Highway cycle was developed with a focus on
accommodating the capability of vehicle dynamometer testing facilities available at the time,
with less focus on representing actual freeway driving patterns. Also, since the late 1980s,
studies have found that both the FUDS and the Highway cycle under-represent the
aggressiveness of actual on-road driving patterns (Austin et al. 1992; U.S. EPA 1993).
Consequently, the U.S. EPA and California Air Resources Board (CARB) have made efforts to
identify on-road driving behaviors and develop driving cycles that are more representative of
real-world driving (Effa and Larsen 1993; Enns et al. 1995). The Representative No. 5 cycle
(REP05), developed by the U.S. EPA using data from Atlanta, Baltimore, and Spokane, is
intended to represent aggressive urban freeway driving patterns. The U.S. No. 6 cycle (US06) is
a short version of the REP05 which EPA has included as an additional bag (Bag 4) in a revised
FTP to address emission impacts of aggressive driving (U.S. EPA 1995). Compared with the
FUDS and the Highway cycle, these two newly developed cycles have higher average speeds,
maximum speeds, maximum acceleration rates, and maximum deceleration rates.
70
Table 4-4 Characteristics of Seven Driving Cycles
NYCC JAPAN10/15 FUDS FTP US06 HWY REP05
Time duration, sec 600 660 1372 1877 600 765 1400
Distance, miles 1.2 2.59 7.5 11 8 10.2 20
Avg. Speed, mph 7.1 14.1 19.5 21.2 48 48.2 51.5
Max. Speed, mph 27.7 43.5 56.7 56.7 80.3 59.9 80.3
Max. decel, mph/sec -5.9 -3.3 -3.3 -6.9 -3.3 -7.1
Max. accel, mph/sec 6 1.6 3.3 3.3 8.5 3.2 8.5
0 100 200 300 400 500
0
10
20
30
40
50
60
70
80
mph
nycc
NYCC
Figure 4-1a Speed Profile of the New York City Cycle
71
0 100 200 300 400 500 600
0
10
20
30
40
50
60
70
80
mph
japan
Japan 10.15 cycle
Figure 4-1b Speed Profile of the Japan 10/15 Cycle
0 200 400 600 800 1000 1200
0
10
20
30
40
50
60
70
80
mph
ftp
Federal Test Procedure
Figure 4-1c Speed Profile of the Federal Test Procedure (FTP)
72
0 100 200 300 400 500 600
0
10
20
30
40
50
60
70
80
mph
us06
US06
Figure 4-1d Speed Profile of the U.S. No. 6 Cycle
0 100 200 300 400 500 600 700
0
10
20
30
40
50
60
70
80
mph
hwy
HWY
Figure 4-1e Speed Profile of the Highway Cycle
73
Table 4-4 presents key characteristics of the above-described seven cycles, which are used
in the HEVTA to evaluate each CV and HEV.
Other cycles have been developed to represent driving patterns in downtown areas of large
cities. The New York City cycle (NYCC), developed in the 1970s, was intended to represent
particularly congested inner cities. A sixth cycle, the Japan 10/15 cycle, also represents
congested driving. It is particularly relevant for an evaluation of hybrid vehicles because the two
hybrids that have been commercially introduced thus far Prius and Insight are both Japanese
and had, until 1999, been tested only on this cycle.
4.4 Other Critical Assumptions
Most of the ADVISOR modeling examined 2010 vehicle configurations. Because of the
(assumed) sharp changes in HEV purchase costs from 2005-2020, we wished to examine how
life cycle costs changed over this same time period. To roughly approximate the changes in fuel
costs over time, we ran ADVISOR for parallel grid-independent hybrids and their competing
CVs using projected 2020 vehicle parameters (which included lighter weight and more efficient
electrical components than were presumed to be available in 2010). The ratio of ADVISOR-
derived 2020 mpg and fuel costs to 2010 mpg and fuel costs for the parallel grid-independent
hybrids and CVs was presumed to apply to all other models (series grid-independent, parallel
mild hybrids, and both parallel and series grid-connected HEVs). Further, 2005 and 2015 values
for all vehicles were estimated by interpolating from the 2010 and 2020 values.
4.5 Estimating Energy Use From Grid-Connected HEVs
As discussed in Section 3.1.2.7, uncertainties in the trip characteristics and battery
management strategies of grid-connected HEVs complicate the accurate estimation of their
energy use, particularly the split between grid-powered and fuel-powered operations. At this
point, we have modeled grid-connected HEVs in hybrid operation (that is, after the period of EV
operation and engine recharge) for the six driving cycles discussed above, and in EV operation
over the FUDS cycle. Aside from reporting on these results, we will estimate average energy use
results based on the assumption that the grid supplies energy for an average daily distance equal
to 60% of the vehicles EV range, with the remainder of daily travel supplied by on-board fuel.
In measuring life cycle energy costs, we assume that both CVs and HEVs in hybrid operation
travel the CAFE cycle, and grid-connected hybrids in EV mode travel the FUDS cycle.
50
The 60-percent recharge value was obtained by examining the three possible options for a
grid-connected HEVs daily travel, assuming that the battery SOC range for hybrid operation is
either 0.4-0.5 or 0.5-0.6
51
:
50
This is primarily because our initial modeling runs modeled the EV-mode energy use only for the FUDS
cycle. In the future, we will model the EV-mode energy use over the range of driving cycles, and make
our vehicle-to-vehicle comparisons using multiple cycles.
51
This is a conservative value for the state-of-charge range, selected out of concern for battery longevity.
See discussion in Section 3.1.2.7.
74
1. Days when the vehicle travels less than its EV range R: All miles are recharged by the
grid; the vehicle is awarded something less than R grid-miles on those days (zero if the
vehicle does not travel at all).
2. Days when the vehicle travels more than its EV range but less than the distance to full
hybrid operation: For 0.4-0.5 SOC operation, the vehicle is recharged by the grid from
0.325 SOC
52
to 0.95 SOC, or 0.625 SOC, or 0.625/.75 = 0.83R; for 0.5-0.6 SOC
operation, the vehicle is recharged from 0.375 SOC to 0.95 SOC, or 0.575 SOC, or
0.77R.
3. Days when the vehicle travels farther than the distance to full hybrid operation: The
vehicle is recharged from the grid, for 0.4-.5 SOC operation, from 0.45 SOC to 0.95
SOC, or 0.5 SOC, or 0.5/0.75 = 0.67R; for 0.5-0.6 SOC operation, the vehicle is
recharged from 0.55 SOC to 0.95 SOC, or 0.53R.
Given these three options, daily grid-energized operations for 0.6R miles seems a reasonable
value, although definitive assumptions about the type of driving patterns most likely for such
vehicles would, of course, push this estimate in one direction or the other. Also, if the battery
SOC range for hybrid operation can be kept low (e.g., at 0.2-0.3 SOC) without unduly
compromising battery longevity, the fraction of grid-recharged miles will be somewhat higher
than 0.6R.
4.6 Model Results
The full results of the ADVISOR and HEVCOST modeling are presented in the tables in
Appendix A. The following discussion presents some key conclusions drawn from these results.
1. HEVs gain the largest fuel economy advantage over CVs on the slowest driving
cycles.
Figure 4-2a shows how hybrids fuel economy advantage over conventional vehicles
depends sharply on driving cycle. As the figure shows, HEVs appear to do best, in relative terms,
on slow cycles with a great deal of stop-and-go traffic.
52
Assuming that, on average, the vehicle is shut down when the battery SOC is midway between 0.2 SOC
and the middle of hybrid operation, 0.45 SOC.
75
1
1.2
1.4
1.6
1.8
2
2.2
NYCC
(7.1)
Japan
10.15
(14.1)
FTP
cycle
(city)
(21.2)
CAF
(29)
US06
(48)
Hwy
(48.2)
REP5
(51.5)
Driving cycle (avg. MPH)
M
P
G
i
m
p
r
o
v
e
m
e
n
t
o
v
e
r
C
V
Full hybrid
Mild hybrid
Figure 4-2a HEV Fuel Economy Advantage over CVs, for Different Driving Cycles
(2010 parallel grid-independent ultralightweight steel [ULS] HEV, 10-second Z60)
These results are logical when one examines the sources of fuel savings by HEVs. We
expect hybrid vehicles to excel relative to CVs under driving conditions where CVs lose
substantial amounts of energy in ways that hybrids do not. In slow, congested traffic, CVs lose
large amounts of energy to braking losses, idling, and engine operation at low efficiency points.
Hybrids recapture part of braking losses, minimize idling by turning off the engine when the
vehicle isnt moving, and, in most hybrid configurations, avoid using the engine at its lowest
efficiency points either by using only the battery and motor to drive the vehicle at very low
speeds, or always keeping engine power high enough to remain efficient, bleeding off excess
power to recharge the battery.
The modeled parallel grid-independent ultralightweight steel (ULS) hybrid approximately
doubles the efficiency of a comparable CV on the very slow NYCC and Japan 10/15 cycles. This
result is compatible with the reported fuel economy gain of the Toyota Prius hybrid on the
Japanese cycle. Although these cycles are far from typical for U.S. driving, the result implies that
hybrids will be particularly effective in congested city driving, for example, for urban taxicabs,
76
transit buses, and even garbage trucks,
53
all of which stop frequently and operate during much of
the day in crowded, slow traffic.
In contrast, the hybrids fuel economy gains on the Highway and other fast cycles are far
more modest about 30% for the full hybrid and 20% for the mild hybrid, for the 10-second Z60
case. If anything, these gains seem high, because there is little braking and idling on the cycle,
and engines in conventional drivetrains tend to run fairly efficiently at highway speeds, yielding
less opportunity for savings compared to city driving, especially when the hybrids weight gain is
factored in. In fact, we would expect that most of any (Highway cycle) savings to be had from
hybridization would come from the effects of engine downsizing, that is, on the efficiency effects
of running at a higher percentage of the engines rated power. Because of our initial skepticism
about the highway cycle fuel economy results, we examined the effects of engine downsizing on
conventional drivetrains (see Section 4 Appendix B) to gain some perspective on the relative
benefits of such downsizing. As shown in Appendix B, the ADVISOR Highway results seem
reasonably consistent with the mpg benefits associated with engine downsizing in CVs.
At first glance, the implications of this conclusion are that maximum fuel savings from
hybrids would be obtained from drivers in congested urban areas. This conclusion must be
tempered by the possibility that vehicle miles traveled per vehicle may vary among different
kinds of areas. Santini and Vyas (1999) point out that different data sources (e.g., the Nationwide
Personal Transportation Survey, cross-national data on miles driven/vehicle and average driving
speeds) imply that the number of hours/day most drivers are willing to spend in their cars falls
within a fairly narrow range. On a fixed time budget, vehicle miles traveled per vehicle varies
inversely with average driving speed. In other words, personal vehicles based in congested urban
areas may accumulate fewer miles of driving per year than suburban-based vehicles. Newman
and Kenworthy (1989) previously reached precisely this conclusion. Thus, drivers of hybrid
vehicles living in congested areas may drive less than hybrid owners living in suburban areas,
partially nullifying the large fuel economy advantage they hold over comparable CVs.
Figure 4-2b shows the implications in fuel savings of assuming that vehicles are driven the
same number of hours per day, by showing the gallons of gasoline saved by full and mild hybrids
(compared to a CV) in 10 hours of driving. With this assumption, full hybrids superiority in
slow, stop-and-go urban driving essentially disappears; driving in higher speed conditions saves
more gasoline even though gallons saved per mile goes down. For mild hybrids, the relationship
between type-of-driving and fuel savings is a bit different, with savings on the slowest cycles
showing a small advantage.
53
In a separate analysis for the DOE Office of Heavy Vehicle Technologies, ANL has evaluated the effect
of hybridizing Class 3-7 heavy-duty vehicles. See An et al. (2000).
77
Fuel Saved in 10 Hours of Driving
0
0.5
1
1.5
2
2.5
3
3.5
4
NYCC Jap10.15 FTP CAF US06 HWY REP5
Driving cycle
Full
Mild
G
a
l
l
o
n
s
o
f
g
a
s
o
l
i
n
e
Figure 4-2b HEV Fuel Economy Advantage over CVs, Measured in Gallons Saved in
10 Hours of Driving
2. Parallel HEVs are likely to be more efficient than series HEVs, with both having
fuel economy that is less sensitive to performance than a CV.
Figure 4-3 shows the fuel economy of comparably-performing CVs and grid-independent
HEVs as a function of acceleration performance.
Our modeled results show a clear efficiency advantage for parallel HEVs over series HEVs.
Examining the way the alternative configurations use energy, the clearest difference is that the
parallel HEVs engine drives the wheels directly through the differential and transmission for
much of its operation, whereas the series HEVs engine must first generate electricity through the
generator and then drive the wheels through the motor. The parallel HEVs direct engine/
transmission/differential route is more efficient than the series HEVs engine/generator/motor
route, accounting for much of the efficiency difference between the two vehicles. Further, the
series vehicle will likely be heavier than the parallel vehicle because the series motor is by far
the more powerful of the two, having to provide all motive power to the vehicle, and the series
engine is also somewhat larger (compared to the full hybrid) to compensate for the less efficient
driveline. The series hybrid may, however, be capable of capturing more braking energy, because
the larger motor is capable of exerting far more braking force, leaving less work for the
mechanical brakes; this capability gives the series hybrid a small offset to the efficiency
advantages of parallel hybrids.
78
YEAR 2010 ULS GRID-INDEPENDENT VEHICLES
20
25
30
35
40
45
50
55
12 10 8
Z60 time
C
o
r
p
o
r
a
t
e
a
v
e
r
a
g
e
f
u
e
l
e
c
o
n
o
m
y
,
m
p
g
PARALLEL HEV
PARALLEL MHEV
SERIES HEV
CV
Figure 4-3 Fuel Economy Variation of CV and HEVs with Z60 Requirements
(year 2010 ultralight steel grid-independent vehicles)
We emphasize that our results apply to passenger cars in normal driving. Urban buses may
yield different results because their stop-and-go, low-speed duty cycles may make the series
hybrid configuration more competitive.
54
The results also show that hybrids fuel economy levels will not vary with vehicle
performance to the same extent as do CVs. In going from a 12-second Z60 time to an 8-second
time, the CVs fuel economy loss is 28.3%, in contrast to the mild (parallel) hybrids 10.7%, the
full (parallel) hybrids 8.4%, and the series hybrids 5.5%. The results generally conform well
with our understanding of how CVs and hybrids use energy. In a CV, a more powerful engine
used to obtain better acceleration performance will, aside from adding weight, have increased
idle losses and tend to operate further away from its maximum efficiency point during most
driving (assuming identical engine design). This yields a significant penalty in fuel economy.
Hybrids minimize idle losses, so the increased idle losses that occur when a CV engine gets
larger should not occur with a hybrid. This, plus the fact that the mild hybrids engine increases
in power less than does the CVs engine (the mild hybrid gains performance by increasing both
engine power and motor/battery power) may account for why the mild hybrid pays such a small
penalty in fuel economy. Finally, both the full (parallel) hybrid and series hybrid designed
according to our convention engine sized for gradeability, motor/battery for acceleration gain
acceleration performance primarily by increasing motor and battery power, so their engines will
54
We note that the two hybrid buses tested in Northeast Advanced Vehicle Consortium, 2000, were both
series hybrids.
79
get only slightly bigger. Thus, there will be little in the way of added losses from a decline in
engine efficiency over the driving cycle. The major fuel economy penalty will come from added
weight, which is a small increment.
3. Hybrids mpg improvement over CVs increases as the acceleration times go down.
Figure 4-4a shows the degree to which our simulated HEVs attain a fuel economy
improvement over CVs, as a function of acceleration time.
An obvious result of the CV fuel economy sensitivity to acceleration requirements and the
hybrids lack of such sensitivity is that the hybrids relative fuel economy advantage over CVs
increases as acceleration times decrease. In other words, the more powerful the vehicle, the more
fuel hybridization will save. This is important because PNGVs original 12-second 0-60 mph
acceleration time target has become outmoded. The U.S. fleet has grown increasingly powerful
and faster during the past decade, and 10 seconds now seems a more likely target for a midsized
car right now, and 8 seconds might well be the target by the time PNGV cars are actually
marketed. Thus, in terms of acceleration performance, the market is driving light-duty vehicles
(LDVs) in a direction that is favorable for hybridization.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
12 seconds 10 seconds 8 seconds
Z60 time
C
A
F
E
i
m
p
r
o
v
e
m
e
n
t
o
v
e
r
C
V
full hybrid
mild hybrid
Figure 4-4a Hybrid Vehicles' Fuel Economy Improvement over CVs
as a Function of Acceleration Times (year 2010 ULS parallel HEV)
80
It is important to avoid overgeneralizing this conclusion. Hybrids do not perform identically
to CVs. In particular, as shown in Figure 4-4b, grid-independent hybrids will suffer in long-
duration gradeability, because over time their batteries cannot sustain speeds on steep grades that
CV drivetrains can easily sustain. The same holds true of towing capability hybrids cannot tow
as much weight as CVs with the same acceleration capability, because of their downsized
engines. Thus, drivers concerned primarily with acceleration may view HEVs as plainly superior
to CVs because of their large fuel economy boost; drivers with wider needs who are buying more
power for towing or other reasons aside from acceleration capability might not be interested in
hybrids despite their fuel economy advantage.
0
5
10
15
20
25
12 seconds 10 seconds 8 seconds
Z60 time
G
r
a
d
e
a
b
i
l
i
t
y
,
p
e
r
c
e
n
t
s
l
o
p
e
CV
MHEV
FHEV
Figure 4-4b CV and HEV Gradeability at 55 mph, Grid-Independent
Parallel Hybrids
4. Parallel hybrids are cheaper than series.
Figure 4-5 compares the projected sales prices of a year 2010 ULS CV with comparably-
performing
55
(10-second Z60 time) HEVs.
55
However, as discussed above, the hybrids will have less gradeability and towing capacity than CVs with
the same acceleration capability.
81
Year 2010 ULS Grid-Independent Vehicles
10 Second Z60
15
20
25
30
35
CV Mild Parallel
HEV
Full Parallel
HEV
Series HEV
V
e
h
i
c
l
e
c
o
s
t
,
1
0
0
0
$
Figure 4-5 Projected Year 2010 Cost of CVs and Hybrids
Without sharper-than-expected decreases in the cost of electrical system components, hybrid
vehicles will be more expensive than CVs, and series hybrids will be more expensive than
comparable parallel hybrids. This is because the hybrids cost savings from their smaller engines
and transmissions (and, for the series hybrid, elimination of a multi-gear transmission) do not
fully offset the high costs of the hybrids motor/controller and battery. As shown in Table 4-5,
the series hybrid actually has a larger engine than a comparable parallel (full) hybrid (though no
transmission) and requires a much larger motor/controller, virtually assuring (with our price
expectations) that it will be more expensive than the parallel hybrid.
Table 4-5 Component Power Requirements
of HEV Drivetrains (10-second Z60, year 2010
ULS vehicles)
mild parallel full parallel series
Motor Size, kW 22 43 87
Engine, kW 65 46 54
Battery, kW 25 49 46
82
5. Demanding better acceleration performance is expensive.
Figure 4-6 shows how vehicle sales price varies with acceleration.
The cost model results show that obtaining better acceleration performance is expensive for
all vehicles, but especially so for hybrids. For the case shown, moving from 12-second Z60 to
8 seconds costs $3,900 for a CV but $7,200 for a parallel full hybrid and $5,700 for a series
hybrid. The high cost of performance for a full parallel hybrid led us to examine the mild hybrid,
because it became apparent that vehicle designers aiming at higher performance would have to
trade off fuel economy performance and cost. Because the mild hybrid obtains increased
performance by increasing both engine and motor/battery power, it can obtain this performance
less expensively than a full hybrid, which increases only motor/battery power (because an engine
kW is less expensive than a motor/battery kW).
1 5
2 0
2 5
3 0
3 5
4 0
1 2 1 0 8
0 - 6 0 M P H A C C E L E R A TI O N TI M E , S E C O N D S
V
E
H
I
C
L
E
P
R
I
C
E
,
1
0
0
0
$
S E R I E S H E V
P A R A L L E L M H E V
P A R A L L E L H E V
C V
Figure 4-6 Vehicle Cost as a Function of Acceleration Requirements
(2010 ULS grid-independent vehicles)
6. HEV life cycle costs are extremely sensitive to assumptions but appear likely to
exceed CV life cycle costs unless fuel prices rise to well over todays levels.
As shown in Appendix A, under our baseline assumptions, life cycle costs for full HEVs
exceed those of competing CVs in all cases. Because fuel costs are a modest proportion of total
life cycle costs, it would take a large increase in gasoline price to create an incentive (equal life
cycle costs for HEVs and CVs) to purchase HEVs. Figure 4-7 shows the breakeven gasoline
prices (prices at which life cycle costs for the two vehicles would be equal) for CVs and parallel
grid-independent ultralight steel hybrids with batteries that last the vehicle lifetime, at three
performance levels (Z60 times of 8, 10, and 12 seconds). Parallel grid-independent hybrids have
the lowest life cycle costs of the full hybrids evaluated thus far. The higher performance HEVs
83
Figure 4-7 Breakeven Fuel Prices for CVs and Parallel Grid-
Independent Full Hybrids, for Three Performance Levels
are more cost-effective, relative to their competing CVs, than the lesser performers, so their
breakeven gasoline prices are lower but still high in comparison to todays gasoline prices.
Although the difference in purchase price between the HEV and CV grows considerably in
moving toward higher performance, this is more than offset by growth in the HEVs fuel savings
over the CV.
The breakeven gasoline prices get considerably lower over time, primarily because the cost
of the hybrids electrical components are presumed to drop considerably during this time period
due to the effects of higher production (economies of scale) and learning.
As noted, the values in Figure 4-7 are based in part on the assumption that continuing
battery development and the use of control strategies that maximize battery life will eliminate the
need to replace the battery during the vehicle lifetime. Failure to reach this goal will increase life
cycle costs significantly. For example, one battery replacement during the lifetime of the
12-second full hybrid above increases the breakeven gasoline price by about $0.50/gal in 2005
and $0.20/gal in 2020.
Figure 4-8 examines breakeven costs for mild variants of the parallel grid-independent
hybrid. The mild hybrids have somewhat lower purchase costs than the full hybrids and pay a
moderate penalty in fuel economy. The 12-second vehicles are slightly less cost-effective than
the full hybrids; the faster, more expensive 10- and 8-second vehicles have superior cost-
effectiveness. By 2015, the 8-second Z60 mild hybrids breakeven fuel costs dip well below
$2/gallon. Although $2/gallon is considerably higher than the 2015 gasoline price projected by
the Energy Information Administration (EIA) Annual Energy Outlook, it appears a plausible
price in the context of recent oil price fluctuations. In other words, it appears possible that these
hybrids may be cost-effective in the 2015-2020 time frame, if their modeled fuel economy
advantage and life cycle costs prove correct.
0
1
2
3
4
5
6
7
2 0 0 5 2 0 1 0 2 0 1 5 2 0 2 0
Y e a r
B
r
e
a
k
e
v
e
n
f
u
e
l
p
r
i
c
e
,
$
/
g
a
l
1 2 s e c o n d s
1 0 s e c o n d s
8 s e c o n d s
84
0
2
4
6
8
2005 2010 2015 2020
Year
B
r
e
a
k
e
v
e
n
f
u
e
l
p
r
i
c
e
,
$
/
g
a
l
12 seconds
10 seconds
8 seconds
Figure 4-8 Breakeven Fuel Prices for CVs and Parallel Grid-
Independent Mild Hybrids, for Three Performance Levels
7. Grid-connected HEVs can achieve large reductions in oil, and significant
reductions in total energy use and greenhouse emissions, in comparison to CVs,
but at very high purchase and life cycle cost.
Grid-connected HEVs are attractive conceptually because they are capable of operating
part-time as EVs and can refuel a portion of their travel from the electrical grid, which offers
significant air quality benefits as well as the potential to capture large savings in oil use and
overall energy use. Figure 4-9 compares the total vehicle energy cost/mile of comparable parallel
grid-connected and grid-independent HEVs and CVs (with the grid-connected HEVs, HEVGrid
in the figure, presumed to travel a daily distance equal to 60% of their EV range on the FTP with
grid-recharge and the remainder on the CAFE cycle, and the CVs and grid-independent HEVs
traveling 100% of their miles on the CAFE cycle). Two CVs and two grid-independent HEVs are
portrayed because it is not clear which offers the best comparison to the grid-connected vehicle.
The CV8 and FHEV8 vehicles (attaining 8-second Z60 times) match the HEVGrids hybrid-
mode (combined engine/electric motor operation) acceleration capability, and the CV12 and full
HEV (FHEV) 12 match the HEVGrids EV-mode acceleration capability. The HEVGrid energy
cost is only about half of the CV12s energy cost and a still smaller fraction of the CV8s energy
cost, partly because it is more efficient in hybrid operation and partly because it is much more
efficient in EV operation. However, neither of the grid-independent FHEVs have substantially
higher energy costs than the HEVGrid, primarily because the heavier HEVGrid is less efficient
in ordinary hybrid operations than either of the grid-independent HEVs, and the HEVGrids less
expensive all-electric operation is not so much less expensive than the FHEVs gasoline
operation that it lowers the total energy costs all that much.
As shown in Figures 4-10 and 4-11, both the initial purchase price and the life cycle costs of
the grid-connected HEVs are much higher than that of the CVs, which is not surprising because
the HEVGrids require substantially larger batteries and motor/controllers than grid-independent
HEVs the battery to provide the range and power needed to operate in EV mode, and the
motor/controller to provide enough power to attain the required 12-second Z60 time. However,
85
0
1
2
3
4
5
6
7
2005 2010 2015 2020
E
n
e
r
g
y
c
o
s
t
s
,
c
e
n
t
s
/
m
i
l
e
HEVGRID
FHEV8
FHEV12
CV8
CV12
Figure 4-9 Fuel + Electricity Cost: CVs vs. Parallel Grid-Connected (EV
Z60 = 12 seconds) and Grid-Independent HEVs
$0
$10,000
$20,000
$30,000
$40,000
$50,000
2005 2010 2015 2020
HEVgrid
FHEV8
FHEV12
CV8
CV12
Figure 4-10 Vehicle Cost Comparison: Grid-Connected and Grid-
Independent Parallel HEVs vs. CV
86
0
10
20
30
40
50
60
2005 2010 2015 2020
L
i
f
e
c
y
c
l
e
c
o
s
t
s
,
c
e
n
t
s
/
m
i
l
e
HEVGRID
FHEV8
FHEV12
CV8
CV12
Figure 4-11 Life Cycle Costs: CVs vs. Grid-Connected and Grid-
Independent Parallel HEVs
the 8-second Z60 grid-independent hybrid (FHEV8) costs almost as much as the HEVGrid and
has life cycle costs nearly as high.
Figure 4-12 compares the gasoline use of performance-equivalent CVs, grid-independent
parallel HEVs,
56
and grid-connected parallel HEVs in average long-term operation (that is,
assuming that the grid-connected HEV operates part-time in all-electric mode and recharges its
batteries overnight) in urban driving. Because the grid-connected HEV recharges a substantial
portion of its miles from the grid, with virtually all grid electricity from nonpetroleum sources, it
uses only about 1/3 as much gasoline as the CV and a bit more than half as much as the grid-
independent HEV.
Finally, Figure 4-13 compares the fuel-cycle greenhouse emissions of performance-
equivalent CVs and (parallel) grid-independent and grid-connected HEVs.
57
56
All five vehicles have the same 0-60 mph time (8 seconds) when their engines are operating (e.g., the
HEVGrid is in hybrid mode); the grid-connected HEV attains a 12-second 0-60 mph time in EV mode.
57
Assumptions: national average grid electricity, low sulfur reformulated gasoline (RFG) with no
oxygenates, half of the grid-connected hybrids miles are grid-recharged (based on about
13,000 miles/yr, 32.6 mile all-electric range calculated by ADVISOR run), 194,803 gCO
2
/10
6
Btu
electricity (from Table B48, Vol 3, GM 2001); gasoline well-to-tank emissions of 21,619 gCO
2
/10
6
Btu
(Table B5, Vol 3, GM 2001); gasoline combustion emissions of 98,096 gCO
2
/10
6
Btu and energy
content of 115,500 Btu/gallon (from GREET 1.6 spreadsheet).
87
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
c
v
1
2
m
h
e
v
1
2
f
h
e
v
1
2
c
v
8
m
h
e
v
8
f
h
e
v
8
G
R
I
D
h
e
v
1
2
/
8
G
a
l
l
o
n
s
/
1
0
0
M
i
l
e
s
Figure 4-12 Comparison of Gasoline Use: CV vs. Grid-Independent and
Grid-Connected (32.6-mile range) HEVs in Urban Driving
CO2 Emissions in Urban Driving, Year 2010 Vehicles
0
0.1
0.2
0.3
0.4
0.5
0.6
C
V
1
2
M
H
E
V
1
2
F
H
E
V
1
2
C
V
8
M
H
E
V
8
F
H
E
V
8
G
R
I
D
H
E
V
1
2
/
8
(
E
V
)
G
R
I
D
H
E
V
1
2
/
8
C
O
2
e
m
i
s
s
i
o
n
s
/
m
i
l
e
,
k
g
HEV Grid EV range is 32.6 miles, EV Z60 is
12 seconds, hybrid Z60 is 8 seconds. EV bar
refers to the vehicle operating as battery
only. Assumptions: in day-to-day urban
driving, half of miles are grid-recharged,
national average electricity mix, low sulfur
RFG without oxygenates.
Figure 4-13 CO
2
Emissions from Similar Conventional and Hybrid Vehicles in
Urban Driving (FTP driving cycle, kilograms per mile)
HEV Grid EV range is 32.6
miles, EV Z60 is 12 seconds,
hybrid Z60 is 8 seconds.
Assumptions: in day-to-day
urban driving, half of miles are
grid-recharged.
88
The grid-connected HEV offers a moderate reduction in greenhouse emissions over the grid-
independent hybrids for this case, with the one exception being the 12-second Z60 full HEV; in
the latter comparison, the HEVs Grid advantage in all-electric operation (next to the last column
in the figure) is neutralized by a disadvantage in hybrid operation (because the heavier battery
and motor required by the grid-connection reduce overall efficiency). The key question here is
whether the 12-second FHEV represents a fair comparison to the Grid HEV, which matches the
FHEVs performance when operating all-electrically but far exceeds it in hybrid operation.
These results are particularly sensitive to the type of recharge electricity assumed and the
hybrid-mode SOC operating range of the battery. Graham et al. (2001) assert that battery
longevity will not suffer unduly by operating the battery within a low SOC range, offering more
rechargeable miles than assumed here, and a greater advantage in greenhouse emissions for the
HEV Grid (note the next-to-the-last column in the figure, which shows the rate of greenhouse
emissions when the vehicle is operating electrically). Also, the national average electricity
assumed here is fairly heavily weighted toward coal; if grid recharge were weighted toward
natural gas combined cycle plants or toward renewables, the greenhouse gases associated with
the grid recharge would be considerably less than assumed here. For a thorough analysis of
possible attributes of grid-connected HEVs, including cost, oil use, electricity use, emissions, and
marketability, see Graham et al. (2001).
8. Long-run (i.e., 2020) incremental prices of the parallel HEVs examined in this
study vary from a 12% increment to a 32% increment, increasing as one proceeds
from a mild hybrid to a full hybrid to a grid-connected hybrid. The estimates for
parallel hybrids range from 12 to 14% for the three MHEV cases, from 16 to 19%
for the FHEV cases, and from 25 to 32% for the HEVGrids with all-electric range
from 23 to 33 miles.
The percentage increase estimates in this study may be compared to those developed in the
recent study by Weiss et al. (2001), done at MIT, and by Graham et al. (2001), under EPRI
management. The one comparable case from the MIT study is an MHEV. The estimated
increase in price is 9%, less than estimated in this study. The base method price increases from
the EPRI study for the two FHEV cases examined are 20 and 21%. Though published earlier, the
price estimates in the EPRI study were completed later than for this study. ANLs HEV cost
estimation team worked with the EPRI study group to generate price increase estimates that
reflect revised electric drive costs from those in this study. More optimism about component
characteristics was included. The ANL method price increase estimates for the two EPRI
FHEV cases were 12 and 13%, well below the 16 to 19% estimated here. The largest difference
resulted from more optimism concerning battery costs. The EPRI base method price increase
estimates for the hybrid with about 20 miles of all-electric range was 32% at the top end of this
studys estimate of 25 to 32%. The revised ANL method gave 21% for this EPRI case.
Compared to the one MHEV case in the MIT study, this studys MHEV price increase
estimates are a bit higher. Compared to the two comparable FHEV and one comparable
HEVGrid cases from the EPRI study, the estimates in this study are a bit lower. The revisions of
the ANL method initiated in this study, and included in the EPRI study, are more optimistic than
this studys estimates. A significant part of the difference arose from observation of rapid
89
improvement of components used in marketed hybrids. This rapid improvement, occurring over
less than two years, suggests that this study may be unduly pessimistic, at least about the rate of
cost reduction that is possible for hybrid vehicle components.
The key hybrid components battery (or other storage device), electric motor, and
controller are all under continuing development. DOE has established aggressive goals to
reduce energy storage costs to levels well below those used in this study. Potential exists to
substantially reduce motor and power electronics costs as well. Although the probability of
achieving very high levels of cost reduction is not clear, success would significantly improve
hybrids cost-effectiveness and their prospects for mass market penetration.
4.7 References
An, F., F. Stodolsky, and D. Santini, 1999, Hybrid Options for Light-Duty Vehicles, SAE
Technical Paper 1999-01-2929.
An, F., et al., 2000, Scenario Analysis of Hybrid Class 3-7 Heavy Vehicles, SAE Technical Paper
2000-01-0989, March.
Austin, T.C., F.J. DiGenova, J.M. Lee, and M. Carlock, 1992, An Analysis of Driving Patterns in
Los Angeles During 1992, paper presented at CRC-APRAC On-Road Vehicle Emissions
Workshop, Dec. 3.
Effa, R., and L.C. Larsen, 1993, Development of Real-World Driving Cycles for Estimating
Facility-Specific Emissions from Light-Duty Vehicles, paper presented at the Air and Waste
Management Specialty Conference on the Emission Inventory - Perception and Reality,
Pasadena, Calif., Oct. 18-20.
Enns, P., J. German, and J. Markey, 1995, EPAs Survey of In-Use Driving Patterns:
Implications for Mobile Source Emission Inventories, U.S. EPA, Office of Mobile Sources,
Certification Division, Ann Arbor, Mich.
EPA, 1993, Federal Test Procedure Review Project: Preliminary Technical Report,
U.S. Environmental Protection Agency, Office of Mobile Sources, Certification Division, Ann
Arbor, Mich., May.
EPA, 1995, Proposed Regulations for Revisions to the Federal Test Procedure for Emissions
from Motor Vehicles; Proposed Rule, Federal Register, Vol. 60, No. 25, pp. 7404-7424, Feb. 7.
Graham et al., 2001, Comparing the Benefits and Impacts of Hybrid Electric Vehicle Options,
1000349, Electric Power Research Institute, Palo Alto, Calif., July.
GM et al., 2001, General Motors Corporation, Argonne National Laboratory, Well-to-Wheel
Energy Use and Greenhouse Gas Emissions of Advanced Fuel/Vehicle Systems North
American Analysis, June.
90
Heavenrich, R.M., and K.H. Hellman, 1999, Light-Duty Automotive Technology and Fuel
Economy Trends through 1999, U.S. Environmental Protection Agency EPA420-R-99-018, Sept.
Johnson, V.H., K.B. Wipke, and D.J. Rausen, 2000, HEV Control Strategy for Real-Time
Optimization of Fuel Economy and Emissions, Society of Automotive Engineers OOFCC-67.
Newman, P.G., and J.R. Kenworthy, 1989, Cities and Automobile Dependence: A Sourcebook,
Hants, England: Gower Publishing Co.
OTA, 1995, Advanced Automotive Technology: Visions of a Super-Efficient Family Car, United
States Congress, Office of Technology Assessment Report OTA-ETT-638, Washington, D.C.
Santini, D.J., and A.D. Vyas, 1999, Commercial Viability of Hybrid Vehicles: Best Household
Use and Cross National Considerations, USAEE/IAEE 20
th
Annual North American Conference,
Aug. 20-Sept. 1.
Weiss et al., 2000, On the Road in 2020: A Life-Cycle Analysis of New Automobile Technologies,
Energy Laboratory Report #MIT EL 00-003, Massachusetts Institute of Technology, Energy
Laboratory, Cambridge, Mass., Oct.
91
Section 4 Appendix A
Table of Results: HEV Cost and ADVISOR
Modeling, 11/1999-3/2000
92
Designtargets
0to60time, 12sec 12sec 12sec 10sec 10sec 10sec 8sec 8sec 8sec
0.2startingSOC
55mphgradeability "float" "float" 6.50% "float" "float" 6.50% "float" "float" 6.50%
Vehicletype convent- "mild" HEV "full" HEV convent- "mild" HEV "full" HEV convent- "mild" HEV "full" HEV
ional (CV) mid-sz.apu parallel ional (CV) mid-sz.apu parallel ional (CV) mid-sz.apu parallel
parallel gridindep. parallel gridindep. parallel gridindep.
gridindep. gridindep. gridindep.
Vehicleattributes
Cd, Crr, frontal area notes notes notes notes notes notes notes notes notes
empty mass (kg) 1175 1246 1247 1248 1321 1328 1366 1453 1466
body material (notes) uls uls uls uls uls uls uls uls uls
APU(kW) (notes) 66 55 44 85 65 46 116 82 49
battery(@20%soc) kW n/a 16 28 n/a 25 49 n/a 43 85
motor - peak kW n/a 14 25 n/a 22 43 n/a 39 75
motor - cont. kW n/a 10 17 n/a 15 30 n/a 27 51
Hevcost 0-60time(soc) 12 12 12 10 10 10 8 8 8
Hevcost 55mph N/E N/E 6.5% N/E N/E 6.5% N/E N/E 6.5%
gradeability
Advisor (A2.1.1) results (notes)
55mphgradeability 12.4% 9.0% 6.6% 16.0% 10.7% 6.7% 21.4% 13.2% 6.8%
0-60time(initial soc) 12sec 12.2sec (0.2) 12sec (0.2) 9.9sec 10sec (0.2) 9.8sec (0.2) 8.1sec 8.1sec (0.2) 8sec (0.2)
Advisor (A2.1.1) fuel economy results ( mpg)
(SOCmax/min=.7/.6- notes)
DrivingCycle
NYCC 15.7 24.4 26.7 13.3 23.6 26.2 10.6 21.9 24.6
Japan1015 28.1 47.1 50.7 24.0 45.7 48.6 19.3 41.3 46.1
FTPcycle(city) 32.2 42.7 46.7 27.7 41.3 45.2 22.6 37.9 42.5
CAF 37.5 46.6 50.9 32.6 45.0 49.4 26.9 41.6 46.6
US06 33.1 35.2 38.3 30.2 35.1 38.5 26.4 34.7 37.4
Hwy 47.1 52.4 57 41.6 50.4 55.7 35.0 47.2 52.8
REP5 38.0 40.3 44 34.3 39.9 43.8 29.6 38.6 41.7
A2rundate: 11/24/99 11/24/99 11/24/99 11/24/99 11/24/99 11/24/99 11/28/99 11/28/99 11/28/99
Performanceandfuel economydatafor CV, "mild" HEV, and"full" HEV
Notes: For all vehicles : Cd = 0.26, Crr = 0.0075, frontal area = 2.06 sq.m
For all vehicles : 5-speed manual transmission (gears of 3.45, 1.94, 1.29, 0.97, 0.75 ; axle of 3.89)
ULSs = ultralight steel; al (not applicable here) = aluminum
APU for "mild hybrids" is halfway between CV and full HEV APU sizes
HEV ADVISOR accel. & grade runs made at initial SOC = 0.2
HEV ADVISOR mpg runs made for SOC range 0.7 to 0.6 .B93
mass, APU, battery, and motor specified by HEVCOST, but after iterations to match with ADVISOR. Both
ADVISOR and HEVCOST were modified repeatedly until they matched empirical data, as discussed in text.
ADVISOR version used was A2.1.1 (of 13 Apr99). Current versions may not produce identical results.
93
Design targets
0 to 60 mph time(sec) float float float 12 10 8 12 12 12
grade (%) float float float 6.5 6.5 6.5 6.5 6.5 6.5
EV range (mi) 23.2 26.3 30.5 n/a n/a n/a 21.1 24.0 28.2
All electric 0-60 time (sec) 16.0 14.0 12.0 n/a n/a n/a 16.0 14.0 12.0
Vehicle type parallel parallel parallel series series series series series series
grid grid grid grid indep. grid indep. grid indep. grid grid grid
connected connected connected connected connected connected
Vehicle attributes
Cd, Crr, frontal area notes notes notes notes notes notes notes notes notes
empty mass (kg) 1439 1491 1566 1341 1404 1513 1468 1504 1559
APU (kW) 48 49 50 52 54 56 56 57 59
battery(@20%soc) kW 62 73 89 30 46 75 60 70 85
interm.power interm.power interm.power high power high power high power interm.power interm.power interm.power
motor - peak kW 54 64 78 71 87 115 70 72 75
motor - cont. kW 37 44 53 49 60 79 48 49 51
Advisor (A2.1.1) results
grade (%) 6.7 6.7 6.6 6.9 7 6.9 7 7.1 7.2
Z60 time (init. soc) 9.1 (0.2) 8.6 (0.2) 8.1 (0.2) 11.8 (0.2) 9.7 (0.2) 7.9 (0.2) 12.3 (0.2) 12.2 (0.2) 12.1 (0.2)
all EL Z60 time (init. soc) 15.9 (0.30) 13.7 (0.30) 12.0 (0.30) n/a n/a n/a 15.6 (0.30) 13.6 (0.30) 12.2 (0.30)
all EL Z60 time (init. soc) 15.9 (0.25) 13.7 (0.25) 12.0 (0.25) 16.2 (0.25) 14.0 (0.25) 12.3 (0.25)
all EL Z60 time (init. soc) 16.1 (0.20) 13.9 (0.20) 12.0 (0.20) 17.4 (0.20) 15.0 (0.20) 12.7 (0.20)
all EL range (miles) 23.2 27 32.6 n/a n/a n/a 21.1 24.2 29.9
(SOC max/min=.95/.2)
A2 run date: 12/15/99 12/15/99 12/15/99 12/15/99 12/15/99 12/15/99
Advisor (A2.1.1) fuel economy results
(mpg) (SOC max/min=.7/.6)
Driving Cycle
NYCC 25.3 24.2 23.7 23.0 23.8 23.2 24.6 24.9 22.8
Japan 1015 46.4 45.2 44.1 40.7 39.7 37.4 38.7 38.8 34.0
FTP cycle (city) 43.3 42.2 41 38.8 38.1 37.0 38.7 39.1 38.2
CAF 47.5 46.4 45.1 43.9 42.9 41.5 43.2 43.3 42.2
US06 37.7 36.9 36.2 31.1 31.2 31.9 32.0 31.8 31.7
Hwy 53.8 52.8 51.5 52.2 50.9 48.5 50.5 50.0 48.2
REP5 42.2 41.5 40.8 38.2 37.5 36.6 36.9 37.1 36.6
pDD pEE pFF sGG sHH sII sJJ sKK sLL
A2 run date: 12/10/99 12/10/99 12/10/99 11/30/99 12/1/99 12/1/99 12/10/99 12/10/99 12/10/99
Notes: For all vehicles : Cd = 0.26, Crr = 0.0075, frontal area = 2.06 sq.m
For all vehicles : 5-spd manual trans. (gears of 3.45, 1.94, 1.29, 0.97, 0.75 ; axle of 3.89)
HEV ADVISOR accel. & grade runs made at initial SOC = 0.2
HEV ADVISOR mpg runs made for SOC range 0.7 to 0.6
all electric (EL) Z60 and range for vehicle with APU = 0 kW.
all electric range made for SOC range 0.95 to 0.20
mass, APU, battery, and motor specified by HEVCOST
ADVISOR version is A2.1.1 (of 13 Apr99)
grid connected vehicles use intermediate power battery (350W/kg, 52Wh/kg at 20% soc)
grid independent vehicles use high power battery (520W/kg, 46Wh/kg at 20% soc)
Performance and fuel economy data for various parallel and series
HEVs
94
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs estimated for the HEVTECA project.)
HEVTECA (Parallel, Grid-independent, UL Steel, HP Batt, 8-12s Z60 Time, APU for 6.5% Grade): ADVISOR-2 Components Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
Z60 Time: 12 seconds Z60 Time: 10 seconds Z60 Time: 8 seconds
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 2,940 $ 2,980 $ 2,970 $ 3,070 $ 3,030 $ 3,080 $ 3,030 $ 3,120 $ 3,220 $ 3,230 $ 3,180 $ 3,290 $
Generator 690 $ 590 $ 550 $ 520 $ 710 $ 610 $ 550 $ 540 $ 740 $ 630 $ 570 $ 550 $
Inverter & Power Electronics 1,870 $ 1,440 $ 1,170 $ 1,020 $ 2,660 $ 2,040 $ 1,570 $ 1,370 $ 4,140 $ 3,060 $ 2,290 $ 1,950 $
Motor 970 $ 820 $ 710 $ 680 $ 1,480 $ 1,190 $ 1,010 $ 960 $ 2,370 $ 1,870 $ 1,530 $ 1,430 $
Transmission 900 $ 860 $ 800 $ 790 $ 1,130 $ 1,060 $ 980 $ 970 $ 1,520 $ 1,420 $ 1,310 $ 1,280 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 2,320 $ 1,840 $ 1,590 $ 1,430 $ 3,960 $ 3,210 $ 2,700 $ 2,460 $ 7,020 $ 5,570 $ 4,650 $ 4,170 $
Total HEV Price 26,900 $ 25,710 $ 24,940 $ 24,610 $ 30,180 $ 28,370 $ 26,990 $ 26,520 $ 36,220 $ 32,960 $ 30,680 $ 29,770 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.04 2.97 2.88 2.79 3.12 3.06 2.98 2.90 3.34 3.24 3.13 3.02
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 2.81 2.11 1.76 1.60 4.80 3.69 2.99 2.76 8.51 6.40 5.16 4.67
Total cost (c/mi) 11.40 10.71 10.36 10.28 13.47 12.38 11.69 11.55 17.39 15.28 14.01 13.58
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 24.12 23.08 22.25 22.09 25.73 24.32 23.14 22.92 28.65 26.48 24.80 24.39
Operating Cost (c/mi) 11.40 10.71 10.36 10.28 13.47 12.38 11.69 11.55 17.39 15.28 14.01 13.58
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.16) (0.15) (0.14) (0.14) (0.18) (0.16) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 35.37 33.65 32.47 32.23 39.03 36.55 34.69 34.33 45.87 41.59 38.66 37.82
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 26 25 23 23 46 43 40 40 81 76 70 68
Motor power (kW): Constant 18 17 16 16 32 30 28 28 56 52 48 47
APU power (kW) 46 44 41 40 48 46 42 41 52 49 45 44
Generator power (kW) 15 14 14 13 16 15 14 14 17 16 15 15
Battery power (kW) 31 28 26 25 53 49 44 43 94 85 76 73
Battery energy (kWh) 2.7 2.5 2.3 2.2 4.6 4.4 3.9 3.7 8.2 7.6 6.7 6.3
Battery mass (kg) 62 54 48 44 106 94 81 75 188 164 139 127
Total vehicle mass (kg) 1,276 1,247 1,217 1,201 1,364 1,328 1,284 1,265 1,525 1,468 1,407 1,377
Zero-to-60 mph time(s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 7 7 6 7 12 11 11 11 19 19 18 18
Parallel Grid-independent FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 11/28/99, PNGV p1& CV5 (with default gear ratio) runs .
The default battery life used, requires one battery replacement.
Run date: 3/30/2000
95
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs estimated for the HEVTECA project.)
HEVTECA (Parallel, Grid-independent, UL Steel, HP Batt thru HEV Life, 8-12s Z60 Time, APU for 6.5% Grade): ADVISOR-2 Components
Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
Z60 Time: 12 seconds Z60 Time: 10 seconds Z60 Time: 8 seconds
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 2,940 $ 2,980 $ 2,970 $ 3,070 $ 3,030 $ 3,080 $ 3,030 $ 3,120 $ 3,220 $ 3,230 $ 3,180 $ 3,290 $
Generator 690 $ 590 $ 550 $ 520 $ 710 $ 610 $ 550 $ 540 $ 740 $ 630 $ 570 $ 550 $
Inverter & Power Electronics 1,870 $ 1,440 $ 1,170 $ 1,020 $ 2,660 $ 2,040 $ 1,570 $ 1,370 $ 4,140 $ 3,060 $ 2,290 $ 1,950 $
Motor 970 $ 820 $ 710 $ 680 $ 1,480 $ 1,190 $ 1,010 $ 960 $ 2,370 $ 1,870 $ 1,530 $ 1,430 $
Transmission 900 $ 860 $ 800 $ 790 $ 1,130 $ 1,060 $ 980 $ 970 $ 1,520 $ 1,420 $ 1,310 $ 1,280 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 2,320 $ 1,840 $ 1,590 $ 1,430 $ 3,960 $ 3,210 $ 2,700 $ 2,460 $ 7,020 $ 5,570 $ 4,650 $ 4,170 $
Total HEV Price 26,900 $ 25,710 $ 24,940 $ 24,610 $ 30,180 $ 28,370 $ 26,990 $ 26,520 $ 36,220 $ 32,960 $ 30,680 $ 29,770 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.04 2.97 2.88 2.79 3.12 3.06 2.98 2.90 3.34 3.24 3.13 3.02
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 2.28 1.78 1.51 1.36 3.89 3.10 2.57 2.34 6.89 5.38 4.43 3.97
Total cost (c/mi) 10.87 10.38 10.12 10.04 12.55 11.80 11.27 11.14 15.78 14.26 13.28 12.88
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 24.12 23.08 22.25 22.09 25.73 24.32 23.14 22.92 28.65 26.48 24.80 24.39
Operating Cost (c/mi) 10.87 10.38 10.12 10.04 12.55 11.80 11.27 11.14 15.78 14.26 13.28 12.88
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.16) (0.15) (0.14) (0.14) (0.18) (0.16) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 34.84 33.31 32.23 31.99 38.12 35.97 34.27 33.92 44.25 40.58 37.93 37.12
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 26 25 23 23 46 43 40 40 81 76 70 68
Motor power (kW): Constant 18 17 16 16 32 30 28 28 56 52 48 47
APU power (kW) 46 44 41 40 48 46 42 41 52 49 45 44
Generator power (kW) 15 14 14 13 16 15 14 14 17 16 15 15
Battery power (kW) 31 28 26 25 53 49 44 43 94 85 76 73
Battery energy (kWh) 2.7 2.5 2.3 2.2 4.6 4.4 3.9 3.7 8.2 7.6 6.7 6.3
Battery mass (kg) 62 54 48 44 106 94 81 75 188 164 139 127
Total vehicle mass (kg) 1,276 1,247 1,217 1,201 1,364 1,328 1,284 1,265 1,525 1,468 1,407 1,377
Zero-to-60 mph time(s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 7 7 6 7 12 11 11 11 19 19 18 18
Parallel Grid-independent FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 11/28/99, PNGV p1& CV5 (with default gear ratio) runs .
The battery life increased to last the life of HEV.
Run date: 3/30/2000
96
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs estimated for the HEVTECA project.)
HEVTECA (Parallel, Mild Grid-independent, UL Steel, HP Batt, 8-12s Z60 Time, APU=6.5% Grade+0.5*Gap): ADVISOR-2 Components
Matching
Vehicle type: Midsize Car (MHV) Vehicle type: Midsize Car (MHV) Vehicle type: Midsize Car (MHV)
Z60 Time: 12 seconds (B) Z60 Time: 10 seconds (B) Z60 Time: 8 seconds (B)
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,460 $ 3,530 $ 3,500 $ 3,620 $ 3,980 $ 4,030 $ 3,970 $ 4,170 $ 4,880 $ 4,870 $ 4,800 $ 4,930 $
Generator 690 $ 590 $ 530 $ 520 $ 870 $ 740 $ 660 $ 630 $ 1,020 $ 840 $ 740 $ 700 $
Inverter & Power Electronics 1,370 $ 1,100 $ 900 $ 800 $ 1,730 $ 1,350 $ 1,080 $ 960 $ 2,480 $ 1,870 $ 1,460 $ 1,250 $
Motor 690 $ 590 $ 530 $ 520 $ 890 $ 760 $ 660 $ 630 $ 1,350 $ 1,110 $ 920 $ 870 $
Transmission 900 $ 860 $ 800 $ 790 $ 1,100 $ 1,040 $ 960 $ 960 $ 1,470 $ 1,380 $ 1,270 $ 1,240 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 1,270 $ 1,050 $ 860 $ 800 $ 2,020 $ 1,640 $ 1,350 $ 1,260 $ 3,590 $ 2,820 $ 2,390 $ 2,110 $
Total HEV Price 25,590 $ 24,900 $ 24,270 $ 24,150 $ 27,800 $ 26,740 $ 25,830 $ 25,710 $ 32,000 $ 30,070 $ 28,730 $ 28,200 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.32 3.24 3.15 3.05 3.42 3.36 3.28 3.19 3.74 3.63 3.51 3.38
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 1.54 1.21 0.95 0.90 2.45 1.88 1.50 1.41 4.35 3.24 2.65 2.37
Total cost (c/mi) 10.41 10.08 9.82 9.84 11.42 10.88 10.49 10.49 13.64 12.51 11.88 11.64
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 23.86 23.06 22.30 22.25 25.30 24.27 23.32 23.30 27.88 26.34 25.10 24.86
Operating Cost (c/mi) 10.41 10.08 9.82 9.84 11.42 10.88 10.49 10.49 13.64 12.51 11.88 11.64
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.16) (0.15) (0.14) (0.14) (0.18) (0.16) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 34.12 33.00 31.99 31.95 36.56 34.99 33.67 33.64 41.34 38.69 36.82 36.34
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 15 14 13 13 23 22 20 20 41 39 35 34
Motor power (kW): Constant 11 10 9 9 16 15 14 14 28 27 24 24
APU power (kW) 57 55 51 50 68 65 60 60 87 82 76 74
Generator power (kW) 15 14 13 13 22 21 20 20 28 26 25 24
Battery power (kW) 17 16 14 14 27 25 22 22 48 43 39 37
Battery energy (kWh) 1.5 1.4 1.2 1.2 2.3 2.2 1.9 1.9 4.2 3.8 3.4 3.2
Battery mass (kg) 34 31 26 24 54 48 40 38 96 83 71 65
Total vehicle mass (kg) 1,270 1,246 1,214 1,202 1,351 1,321 1,280 1,269 1,504 1,453 1,400 1,373
Zero-to-60 mph time(s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 4 4 4 4 6 6 5 6 11 10 10 10
Parallel Grid-independent MHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 11/28/99. The default battery life used, requires one battery replacemen
Run dated: 3/30/2000 (The MHVs' APU are sized in-between the corresponding CV and FHV capable of 6.5% grade.)
97
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs estimated for the HEVTECA project.)
HEVTECA (Parallel, Mild Grid-independent, UL Steel, HP Batt thru HEV Life, 8-12s Z60 Time, APU=6.5% Grade+0.5*Gap): ADVISOR-2
Components Matching
Vehicle type: Midsize Car (MHV) Vehicle type: Midsize Car (MHV) Vehicle type: Midsize Car (MHV)
Z60 Time: 12 seconds (B) Z60 Time: 10 seconds (B) Z60 Time: 8 seconds (B)
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,460 $ 3,530 $ 3,500 $ 3,620 $ 3,980 $ 4,030 $ 3,970 $ 4,170 $ 4,880 $ 4,870 $ 4,800 $ 4,930 $
Generator 690 $ 590 $ 530 $ 520 $ 870 $ 740 $ 660 $ 630 $ 1,020 $ 840 $ 740 $ 700 $
Inverter & Power Electronics 1,370 $ 1,100 $ 900 $ 800 $ 1,730 $ 1,350 $ 1,080 $ 960 $ 2,480 $ 1,870 $ 1,460 $ 1,250 $
Motor 690 $ 590 $ 530 $ 520 $ 890 $ 760 $ 660 $ 630 $ 1,350 $ 1,110 $ 920 $ 870 $
Transmission 900 $ 860 $ 800 $ 790 $ 1,100 $ 1,040 $ 960 $ 960 $ 1,470 $ 1,380 $ 1,270 $ 1,240 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 1,270 $ 1,050 $ 860 $ 800 $ 2,020 $ 1,640 $ 1,350 $ 1,260 $ 3,590 $ 2,820 $ 2,390 $ 2,110 $
Total HEV Price 25,590 $ 24,900 $ 24,270 $ 24,150 $ 27,800 $ 26,740 $ 25,830 $ 25,710 $ 32,000 $ 30,070 $ 28,730 $ 28,200 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.32 3.24 3.15 3.05 3.42 3.36 3.28 3.19 3.74 3.63 3.51 3.38
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 1.25 1.02 0.82 0.76 1.98 1.59 1.29 1.20 3.52 2.73 2.28 2.01
Total cost (c/mi) 10.12 9.89 9.69 9.70 10.96 10.58 10.28 10.28 12.81 11.99 11.50 11.28
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 23.86 23.06 22.30 22.25 25.30 24.27 23.32 23.30 27.88 26.34 25.10 24.86
Operating Cost (c/mi) 10.12 9.89 9.69 9.70 10.96 10.58 10.28 10.28 12.81 11.99 11.50 11.28
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.16) (0.15) (0.14) (0.14) (0.18) (0.16) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 33.83 32.81 31.85 31.81 36.09 34.69 33.46 33.43 40.52 38.17 36.45 35.99
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 15 14 13 13 23 22 20 20 41 39 35 34
Motor power (kW): Constant 11 10 9 9 16 15 14 14 28 27 24 24
APU power (kW) 57 55 51 50 68 65 60 60 87 82 76 74
Generator power (kW) 15 14 13 13 22 21 20 20 28 26 25 24
Battery power (kW) 17 16 14 14 27 25 22 22 48 43 39 37
Battery energy (kWh) 1.5 1.4 1.2 1.2 2.3 2.2 1.9 1.9 4.2 3.8 3.4 3.2
Battery mass (kg) 34 31 26 24 54 48 40 38 96 83 71 65
Total vehicle mass (kg) 1,270 1,246 1,214 1,202 1,351 1,321 1,280 1,269 1,504 1,453 1,400 1,373
Zero-to-60 mph time(s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 4 4 4 4 6 6 5 6 11 10 10 10
Parallel Grid-independent MHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 11/28/99. The battery life increased to last the life of HEV.
Run dated: 3/30/2000 (The MHVs' APU are sized in-between the corresponding CV and FHV capable of 6.5% grade.)
98
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs estimated for the HEVTECA project.)
HEVTECA (Parallel, Mild Grid-independent, UL Steel, HP Batt thru HEV Life, 8-12s Z60 Time, APU=6.5% Grade+0.5*Gap): ADVISOR-2
Components Matching
Vehicle type: Midsize Car (MHV) Vehicle type: Midsize Car (MHV) Vehicle type: Midsize Car (MHV)
Z60 Time: 12 seconds (B) Z60 Time: 10 seconds (B) Z60 Time: 8 seconds (B)
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,460 $ 3,530 $ 3,500 $ 3,620 $ 3,980 $ 4,030 $ 3,970 $ 4,170 $ 4,880 $ 4,870 $ 4,800 $ 4,930 $
Generator 690 $ 590 $ 530 $ 520 $ 870 $ 740 $ 660 $ 630 $ 1,020 $ 840 $ 740 $ 700 $
Inverter & Power Electronics 1,370 $ 1,100 $ 900 $ 800 $ 1,730 $ 1,350 $ 1,080 $ 960 $ 2,480 $ 1,870 $ 1,460 $ 1,250 $
Motor 690 $ 590 $ 530 $ 520 $ 890 $ 760 $ 660 $ 630 $ 1,350 $ 1,110 $ 920 $ 870 $
Transmission 900 $ 860 $ 800 $ 790 $ 1,100 $ 1,040 $ 960 $ 960 $ 1,470 $ 1,380 $ 1,270 $ 1,240 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 1,270 $ 1,050 $ 860 $ 800 $ 2,020 $ 1,640 $ 1,350 $ 1,260 $ 3,590 $ 2,820 $ 2,390 $ 2,110 $
Total HEV Price 25,590 $ 24,900 $ 24,270 $ 24,150 $ 27,800 $ 26,740 $ 25,830 $ 25,710 $ 32,000 $ 30,070 $ 28,730 $ 28,200 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.32 3.24 3.15 3.05 3.42 3.36 3.28 3.19 3.74 3.63 3.51 3.38
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 1.25 1.02 0.82 0.76 1.98 1.59 1.29 1.20 3.52 2.73 2.28 2.01
Total cost (c/mi) 10.12 9.89 9.69 9.70 10.96 10.58 10.28 10.28 12.81 11.99 11.50 11.28
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 23.86 23.06 22.30 22.25 25.30 24.27 23.32 23.30 27.88 26.34 25.10 24.86
Operating Cost (c/mi) 10.12 9.89 9.69 9.70 10.96 10.58 10.28 10.28 12.81 11.99 11.50 11.28
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.16) (0.15) (0.14) (0.14) (0.18) (0.16) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 33.83 32.81 31.85 31.81 36.09 34.69 33.46 33.43 40.52 38.17 36.45 35.99
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 15 14 13 13 23 22 20 20 41 39 35 34
Motor power (kW): Constant 11 10 9 9 16 15 14 14 28 27 24 24
APU power (kW) 57 55 51 50 68 65 60 60 87 82 76 74
Generator power (kW) 15 14 13 13 22 21 20 20 28 26 25 24
Battery power (kW) 17 16 14 14 27 25 22 22 48 43 39 37
Battery energy (kWh) 1.5 1.4 1.2 1.2 2.3 2.2 1.9 1.9 4.2 3.8 3.4 3.2
Battery mass (kg) 34 31 26 24 54 48 40 38 96 83 71 65
Total vehicle mass (kg) 1,270 1,246 1,214 1,202 1,351 1,321 1,280 1,269 1,504 1,453 1,400 1,373
Zero-to-60 mph time(s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 4 4 4 4 6 6 5 6 11 10 10 10
Parallel Grid-independent MHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 11/28/99. The battery life increased to last the life of HEV.
Run dated: 3/30/2000 (The MHVs' APU are sized in-between the corresponding CV and FHV capable of 6.5% grade.)
99
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Series Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs Estimated for the HEVTECA Project.)
HEV (Series, Grid-Independent, UL Steel, HP Batt, 8-12s Z60 Time) ADVISOR2 Components Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
Z60 Time: 12 seconds Z60 Time: 10 seconds Z60 Time: 8 seconds
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,410 $ 3,380 $ 3,390 $ 3,400 $ 3,500 $ 3,430 $ 3,440 $ 3,510 $ 3,690 $ 3,580 $ 3,600 $ 3,620 $
Generator 1,680 $ 1,330 $ 1,150 $ 1,050 $ 1,730 $ 1,350 $ 1,150 $ 1,080 $ 1,810 $ 1,410 $ 1,200 $ 1,110 $
Inverter & Power Electronics 3,890 $ 2,920 $ 2,290 $ 1,930 $ 4,610 $ 3,410 $ 2,630 $ 2,270 $ 5,900 $ 4,320 $ 3,300 $ 2,790 $
Motor 2,240 $ 1,760 $ 1,520 $ 1,410 $ 2,680 $ 2,090 $ 1,780 $ 1,640 $ 3,490 $ 2,670 $ 2,240 $ 2,070 $
Gear Drive 280 $ 250 $ 240 $ 230 $ 340 $ 300 $ 290 $ 280 $ 450 $ 400 $ 380 $ 370 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 2,470 $ 1,970 $ 1,720 $ 1,540 $ 3,810 $ 3,010 $ 2,630 $ 2,400 $ 6,280 $ 4,910 $ 4,280 $ 3,830 $
Total HEV Price 31,200 $ 28,800 $ 27,500 $ 26,700 $ 33,900 $ 30,800 $ 29,100 $ 28,300 $ 38,900 $ 34,500 $ 32,200 $ 30,900 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.53 3.44 3.34 3.24 3.63 3.52 3.40 3.31 3.75 3.64 3.52 3.39
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 2.99 2.26 1.90 1.73 4.62 3.46 2.92 2.69 7.61 5.65 4.75 4.29
Total cost (c/mi) 12.07 11.34 10.97 10.85 13.79 12.62 12.04 11.89 16.90 14.92 13.98 13.57
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 28.19 25.94 24.56 23.97 29.53 26.87 25.22 24.68 32.01 28.61 26.60 25.79
Operating Cost (c/mi) 12.07 11.34 10.97 10.85 13.79 12.62 12.04 11.89 16.90 14.92 13.98 13.57
Less Scrappage Value (c/mi) (0.18) (0.16) (0.15) (0.15) (0.19) (0.17) (0.16) (0.15) (0.20) (0.18) (0.16) (0.16)
Total Life-cycle Cost (c/mi) 40.08 37.11 35.38 34.68 43.13 39.32 37.10 36.42 48.71 43.35 40.42 39.20
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 76 71 69 67 93 87 84 81 125 115 110 107
Motor power (kW): Constant 52 49 47 46 64 60 58 56 85 79 75 73
APU power (kW) 56 52 49 46 58 53 50 48 62 56 53 50
Generator power (kW) 54 50 48 45 56 51 48 47 59 54 51 49
Battery power (kW) 33 30 28 27 51 46 43 42 84 75 70 67
Battery energy (kWh) 2.9 2.7 2.5 2.3 4.4 4.1 3.8 3.6 7.3 6.7 6.1 5.8
Battery mass (kg) 66 58 51 47 102 89 79 73 168 144 128 117
Total vehicle mass (kg) 1,386 1,341 1,310 1,281 1,457 1,401 1,364 1,336 1,591 1,513 1,465 1,426
Acceleration time 0-60 mph (s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 5 5 5 5 8 8 8 8 14 13 12 13
Series Grid-independent FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 12/1/99. The default battery life used, requires one battery replacement.
Run date: 3/30/2000
100
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Series Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs Estimated for the HEVTECA Project.)
HEV (Series, Grid-Independent, UL Steel, HP Batt thru HEV Life, 8-12s Z60 Time) ADVISOR2 Components Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
Z60 Time: 12 seconds Z60 Time: 10 seconds Z60 Time: 8 seconds
Battery type: NiMH (High Power) Battery type: NiMH (High Power) Battery type: NiMH (High Power)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,200 $ 21,200 $ 21,200 $ 22,500 $ 22,500 $ 22,500 $ 22,500 $ 25,100 $ 25,100 $ 25,100 $ 25,100 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,410 $ 3,380 $ 3,390 $ 3,400 $ 3,500 $ 3,430 $ 3,440 $ 3,510 $ 3,690 $ 3,580 $ 3,600 $ 3,620 $
Generator 1,680 $ 1,330 $ 1,150 $ 1,050 $ 1,730 $ 1,350 $ 1,150 $ 1,080 $ 1,810 $ 1,410 $ 1,200 $ 1,110 $
Inverter & Power Electronics 3,890 $ 2,920 $ 2,290 $ 1,930 $ 4,610 $ 3,410 $ 2,630 $ 2,270 $ 5,900 $ 4,320 $ 3,300 $ 2,790 $
Motor 2,240 $ 1,760 $ 1,520 $ 1,410 $ 2,680 $ 2,090 $ 1,780 $ 1,640 $ 3,490 $ 2,670 $ 2,240 $ 2,070 $
Gear Drive 280 $ 250 $ 240 $ 230 $ 340 $ 300 $ 290 $ 280 $ 450 $ 400 $ 380 $ 370 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 2,470 $ 1,970 $ 1,720 $ 1,540 $ 3,810 $ 3,010 $ 2,630 $ 2,400 $ 6,280 $ 4,910 $ 4,280 $ 3,830 $
Total HEV Price 31,200 $ 28,800 $ 27,500 $ 26,700 $ 33,900 $ 30,800 $ 29,100 $ 28,300 $ 38,900 $ 34,500 $ 32,200 $ 30,900 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.26 4.28 4.28 4.21 4.87 4.93 4.95 4.91 6.02 5.97 5.89 5.72
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Hybrid Electric Vehicle
Electricity cost (c/mi) - - - - - - - - - - - -
Fuel cost (c/mi) 3.53 3.44 3.34 3.24 3.63 3.52 3.40 3.31 3.75 3.64 3.52 3.39
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 2.42 1.90 1.64 1.47 3.74 2.91 2.51 2.29 6.16 4.75 4.08 3.65
Total cost (c/mi) 11.50 10.98 10.70 10.59 12.91 12.07 11.63 11.49 15.46 14.02 13.31 12.93
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.50 20.20 20.20 22.08 21.75 21.44 21.44 24.63 24.27 23.91 23.91
Operating Cost (c/mi) 9.16 9.26 9.33 9.41 9.77 9.90 10.00 10.11 10.92 10.94 10.94 10.92
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.14) (0.14) (0.13) (0.13) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 29.84 29.62 29.40 29.48 31.71 31.52 31.31 31.41 35.39 35.06 34.71 34.69
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 28.19 25.94 24.56 23.97 29.53 26.87 25.22 24.68 32.01 28.61 26.60 25.79
Operating Cost (c/mi) 11.50 10.98 10.70 10.59 12.91 12.07 11.63 11.49 15.46 14.02 13.31 12.93
Less Scrappage Value (c/mi) (0.18) (0.16) (0.15) (0.15) (0.19) (0.17) (0.16) (0.15) (0.20) (0.18) (0.16) (0.16)
Total Life-cycle Cost (c/mi) 39.51 36.76 35.11 34.42 42.26 38.76 36.69 36.01 47.27 42.45 39.75 38.56
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 76 71 69 67 93 87 84 81 125 115 110 107
Motor power (kW): Constant 52 49 47 46 64 60 58 56 85 79 75 73
APU power (kW) 56 52 49 46 58 53 50 48 62 56 53 50
Generator power (kW) 54 50 48 45 56 51 48 47 59 54 51 49
Battery power (kW) 33 30 28 27 51 46 43 42 84 75 70 67
Battery energy (kWh) 2.9 2.7 2.5 2.3 4.4 4.1 3.8 3.6 7.3 6.7 6.1 5.8
Battery mass (kg) 66 58 51 47 102 89 79 73 168 144 128 117
Total vehicle mass (kg) 1,386 1,341 1,310 1,281 1,457 1,401 1,364 1,336 1,591 1,513 1,465 1,426
Acceleration time 0-60 mph (s) 12.0 12.0 12.0 12.0 10.0 10.0 10.0 10.0 8.0 8.0 8.0 8.0
Estimated all-elec range (mi) 5 5 5 5 8 8 8 8 14 13 12 13
Series Grid-independent FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 12/1/99. The battery life increased to last the life of HEV.
Run date: 3/30/2000
101
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs are based on the ANL Delphi study.)
HEVTECA (Parallel, Grid-Connected, UL Steel, Intermediate Batt, 12-16s All-electric) ADVISOR-2 Components Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
All-electric Z60 Time: 16 seconds All-electric Z60 Time: 14 seconds All-electric Z60 Time: 12 seconds
Battery type: NiMH (Intermediate) Battery type: NiMH (Intermediate) Battery type: NiMH (Intermediate)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 23,700 $ 23,700 $ 23,700 $ 23,600 $ 24,300 $ 24,300 $ 24,300 $ 24,300 $ 25,000 $ 25,000 $ 25,000 $ 25,000 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,170 $ 3,180 $ 3,240 $ 3,340 $ 3,220 $ 3,230 $ 3,290 $ 3,400 $ 3,310 $ 3,280 $ 3,390 $ 3,510 $
Generator 740 $ 630 $ 570 $ 550 $ 740 $ 630 $ 590 $ 550 $ 760 $ 630 $ 590 $ 570 $
Inverter & Power Electronics 3,200 $ 2,410 $ 1,910 $ 1,640 $ 3,670 $ 2,720 $ 2,130 $ 1,820 $ 4,320 $ 3,180 $ 2,490 $ 2,110 $
Motor 1,810 $ 1,440 $ 1,250 $ 1,160 $ 2,090 $ 1,640 $ 1,410 $ 1,310 $ 2,470 $ 1,950 $ 1,660 $ 1,540 $
Transmission 1,290 $ 1,190 $ 1,160 $ 1,130 $ 1,410 $ 1,300 $ 1,260 $ 1,230 $ 1,580 $ 1,460 $ 1,420 $ 1,390 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 7,520 $ 6,010 $ 5,350 $ 4,820 $ 8,960 $ 7,080 $ 6,250 $ 5,580 $ 10,950 $ 8,630 $ 7,700 $ 6,850 $
Total HEV Price 34,940 $ 32,040 $ 30,630 $ 29,740 $ 37,300 $ 33,780 $ 32,080 $ 30,990 $ 40,600 $ 36,310 $ 34,400 $ 33,070 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 5.33 5.34 5.32 5.22 5.60 5.59 5.55 5.42 5.95 5.91 5.83 5.66
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 10.23 10.32 10.37 10.42 10.50 10.57 10.59 10.62 10.85 10.89 10.88 10.86
Hybrid Electric Vehicle
Electricity cost (c/mi) 0.93 0.88 0.84 0.80 1.11 1.03 0.98 0.95 1.36 1.26 1.20 1.17
Fuel cost (c/mi) 1.84 1.76 1.70 1.65 1.63 1.56 1.51 1.46 1.38 1.30 1.26 1.21
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 12.15 9.31 7.93 7.16 14.47 10.97 9.27 8.29 17.69 13.37 11.42 10.18
Total cost (c/mi) 20.47 17.58 16.19 15.51 22.77 19.20 17.48 16.59 25.97 21.56 19.60 18.45
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 23.26 22.91 22.58 22.49 23.85 23.49 23.15 23.15 24.53 24.17 23.82 23.82
Operating Cost (c/mi) 10.23 10.32 10.37 10.42 10.50 10.57 10.59 10.62 10.85 10.89 10.88 10.86
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.15) (0.15) (0.14) (0.14) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 33.34 33.09 32.81 32.76 34.20 33.91 33.60 33.63 35.23 34.91 34.55 34.54
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 26.91 25.16 24.09 23.74 27.81 25.81 24.61 24.21 29.10 26.76 25.44 24.98
Operating Cost (c/mi) 20.47 17.58 16.19 15.51 22.77 19.20 17.48 16.59 25.97 21.56 19.60 18.45
Less Scrappage Value (c/mi) (0.17) (0.16) (0.15) (0.15) (0.17) (0.16) (0.15) (0.15) (0.18) (0.17) (0.16) (0.15)
Total Life-cycle Cost (c/mi) 47.21 42.59 40.12 39.10 50.40 44.85 41.94 40.66 54.89 48.16 44.88 43.28
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 59 55 54 52 70 65 63 61 85 80 77 75
Motor power (kW): Constant 41 38 37 36 48 45 43 42 58 55 53 51
APU power (kW) 51 48 46 45 52 49 47 46 54 50 49 48
Generator power (kW) 17 16 15 15 17 16 16 15 18 16 16 16
Battery power (kW) 68 62 59 57 81 73 69 66 99 89 85 81
Battery energy (kWh) 9.9 9.2 8.6 8.2 11.8 10.9 10.1 9.5 14.4 13.2 12.5 11.7
Battery mass (kg) 202 177 160 148 241 209 188 171 294 255 231 210
Total vehicle mass (kg) 1,495 1,441 1,404 1,374 1,556 1,494 1,451 1,417 1,647 1,570 1,527 1,488
Accl time on APU+Battery (s) 9.1 9.1 9.1 9.2 8.6 8.6 8.7 8.7 8.1 8.1 8.1 8.1
Accl time on Battery-only (s) 16.0 16.0 16.0 16.0 14.0 14.0 14.0 14.0 12.0 12.0 12.0 12.0
Estimated all-elec range (mi) 22.7 23.2 22.4 21.8 25.9 26.3 25.4 24.3 30.0 30.5 29.7 28.1
Parallel Grid-Connected FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 12/10/99. The default battery life used, requires one battery replacement.
Run date: 3/30/2000
102
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Parallel Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs are based on the ANL Delphi study.)
HEVTECA (Parallel, Grid-Connected, UL Steel, Int Batt Lasts Veh Life, 12-16s All-electric) ADVISOR-2 Components Matching
Vehicle type: Midsize Car Vehicle type:Midsize Car Vehicle type:Midsize Car
All-electric Z60 Time:16 seconds All-electric Z60 Time:14 seconds All-electric Z60 Time:12 seconds
Battery type:NiMH (Intermediate) Battery type:NiMH (Intermediate) Battery type:NiMH (Intermediate)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 23,700 $ 23,700 $ 23,700 $ 23,600 $ 24,300 $ 24,300 $ 24,300 $ 24,300 $ 25,000 $ 25,000 $ 25,000 $ 25,000 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,170 $ 3,180 $ 3,240 $ 3,340 $ 3,220 $ 3,230 $ 3,290 $ 3,400 $ 3,310 $ 3,280 $ 3,390 $ 3,510 $
Generator 740 $ 630 $ 570 $ 550 $ 740 $ 630 $ 590 $ 550 $ 760 $ 630 $ 590 $ 570 $
Inverter & Power Electronics 3,200 $ 2,410 $ 1,910 $ 1,640 $ 3,670 $ 2,720 $ 2,130 $ 1,820 $ 4,320 $ 3,180 $ 2,490 $ 2,110 $
Motor 1,810 $ 1,440 $ 1,250 $ 1,160 $ 2,090 $ 1,640 $ 1,410 $ 1,310 $ 2,470 $ 1,950 $ 1,660 $ 1,540 $
Transmission 1,290 $ 1,190 $ 1,160 $ 1,130 $ 1,410 $ 1,300 $ 1,260 $ 1,230 $ 1,580 $ 1,460 $ 1,420 $ 1,390 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 7,520 $ 6,010 $ 5,350 $ 4,820 $ 8,960 $ 7,080 $ 6,250 $ 5,580 $ 10,950 $ 8,630 $ 7,700 $ 6,850 $
Total HEV Price 34,940 $ 32,040 $ 30,630 $ 29,740 $ 37,300 $ 33,780 $ 32,080 $ 30,990 $ 40,600 $ 36,310 $ 34,400 $ 33,070 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 5.33 5.34 5.32 5.22 5.60 5.59 5.55 5.42 5.95 5.91 5.83 5.66
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 10.23 10.32 10.37 10.42 10.50 10.57 10.59 10.62 10.85 10.89 10.88 10.86
Hybrid Electric Vehicle
Electricity cost (c/mi) 0.93 0.88 0.84 0.80 1.11 1.03 0.98 0.95 1.36 1.26 1.20 1.17
Fuel cost (c/mi) 1.84 1.76 1.70 1.65 1.63 1.56 1.51 1.46 1.38 1.30 1.26 1.21
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 7.38 5.81 5.10 4.59 8.79 6.84 5.95 5.32 10.75 8.34 7.34 6.53
Total cost (c/mi) 15.70 14.08 13.35 12.94 17.08 15.07 14.17 13.62 19.03 16.54 15.51 14.80
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 23.26 22.91 22.58 22.49 23.85 23.49 23.15 23.15 24.53 24.17 23.82 23.82
Operating Cost (c/mi) 10.23 10.32 10.37 10.42 10.50 10.57 10.59 10.62 10.85 10.89 10.88 10.86
Less Scrappage Value (c/mi) (0.15) (0.14) (0.14) (0.14) (0.15) (0.15) (0.14) (0.14) (0.15) (0.15) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 33.34 33.09 32.81 32.76 34.20 33.91 33.60 33.63 35.23 34.91 34.55 34.54
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 26.91 25.16 24.09 23.74 27.81 25.81 24.61 24.21 29.10 26.76 25.44 24.98
Operating Cost (c/mi) 15.70 14.08 13.35 12.94 17.08 15.07 14.17 13.62 19.03 16.54 15.51 14.80
Less Scrappage Value (c/mi) (0.17) (0.16) (0.15) (0.15) (0.17) (0.16) (0.15) (0.15) (0.18) (0.17) (0.16) (0.15)
Total Life-cycle Cost (c/mi) 42.44 39.09 37.29 36.53 44.72 40.73 38.62 37.68 47.94 43.13 40.79 39.63
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 59 55 54 52 70 65 63 61 85 80 77 75
Motor power (kW): Constant 41 38 37 36 48 45 43 42 58 55 53 51
APU power (kW) 51 48 46 45 52 49 47 46 54 50 49 48
Generator power (kW) 17 16 15 15 17 16 16 15 18 16 16 16
Battery power (kW) 68 62 59 57 81 73 69 66 99 89 85 81
Battery energy (kWh) 9.9 9.2 8.6 8.2 11.8 10.9 10.1 9.5 14.4 13.2 12.5 11.7
Battery mass (kg) 202 177 160 148 241 209 188 171 294 255 231 210
Total vehicle mass (kg) 1,495 1,441 1,404 1,374 1,556 1,494 1,451 1,417 1,647 1,570 1,527 1,488
Accl time on APU+Battery (s) 9.1 9.1 9.1 9.2 8.6 8.6 8.7 8.7 8.1 8.1 8.1 8.1
Accl time on Battery-only (s) 16.0 16.0 16.0 16.0 14.0 14.0 14.0 14.0 12.0 12.0 12.0 12.0
Estimated all-elec range (mi) 22.7 23.2 22.4 21.8 25.9 26.3 25.4 24.3 30.0 30.5 29.7 28.1
Parallel Grid-Connected FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 12/10/99. The battery life increased to last the life of HEV.
Run date: 3/30/2000
103
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Series Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs are based on the ANL Delphi study.)
HEV (Series, Grid-Connected, UL Steel, Int Batt, 12-16s All-electric) ADVISOR2 Components Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
All-electric Z60 Time: 16 seconds All-electric Z60 Time: 14 seconds All-electric Z60 Time: 12 seconds
Battery type: NiMH (Intermediate) Battery type: NiMH (Intermediate) Battery type: NiMH (Intermediate)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,100 $ 21,100 $ 21,100 $ 21,300 $ 21,100 $ 21,100 $ 21,100 $ 21,800 $ 21,900 $ 21,600 $ 21,600 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,600 $ 3,530 $ 3,550 $ 3,620 $ 3,650 $ 3,580 $ 3,600 $ 3,620 $ 3,690 $ 3,630 $ 3,650 $ 3,670 $
Generator 1,760 $ 1,390 $ 1,180 $ 1,110 $ 1,780 $ 1,410 $ 1,200 $ 1,110 $ 1,810 $ 1,440 $ 1,220 $ 1,130 $
Inverter & Power Electronics 3,600 $ 2,640 $ 2,200 $ 1,890 $ 3,670 $ 2,690 $ 2,220 $ 1,910 $ 3,780 $ 2,810 $ 2,270 $ 1,950 $
Motor 2,320 $ 1,740 $ 1,570 $ 1,460 $ 2,370 $ 1,780 $ 1,590 $ 1,480 $ 2,450 $ 1,870 $ 1,620 $ 1,510 $
Transmission 510 $ 430 $ 430 $ 410 $ 520 $ 440 $ 430 $ 420 $ 530 $ 460 $ 440 $ 430 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 7,410 $ 5,820 $ 5,170 $ 4,650 $ 8,630 $ 6,790 $ 6,070 $ 5,410 $ 10,400 $ 8,240 $ 7,250 $ 6,510 $
Total HEV Price 36,500 $ 32,800 $ 31,300 $ 30,300 $ 37,900 $ 33,900 $ 32,300 $ 31,100 $ 39,900 $ 35,700 $ 33,600 $ 32,300 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.25 4.20 4.20 4.14 4.27 4.22 4.19 4.13 4.51 4.56 4.45 4.40
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.15 9.17 9.25 9.34 9.17 9.19 9.24 9.33 9.41 9.53 9.50 9.60
Hybrid Electric Vehicle
Electricity cost (c/mi) 0.92 0.85 0.81 0.77 1.07 0.99 0.94 0.89 1.29 1.20 1.13 1.07
Fuel cost (c/mi) 2.09 2.03 1.97 1.91 1.88 1.82 1.77 1.71 1.60 1.52 1.48 1.43
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 11.97 9.01 7.66 6.91 13.94 10.52 9.00 8.04 16.80 12.77 10.75 9.67
Total cost (c/mi) 20.53 17.53 16.16 15.47 22.44 18.96 17.43 16.53 25.23 21.12 19.08 18.06
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.40 20.10 20.10 20.90 20.40 20.10 20.10 21.39 21.17 20.58 20.58
Operating Cost (c/mi) 9.15 9.17 9.25 9.34 9.17 9.19 9.24 9.33 9.41 9.53 9.50 9.60
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.13) (0.13) (0.12) (0.12) (0.13) (0.13) (0.13) (0.13)
Total Life-cycle Cost (c/mi) 29.83 29.44 29.23 29.32 29.94 29.47 29.22 29.31 30.67 30.57 29.95 30.05
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 28.55 26.08 24.90 24.44 28.72 26.21 24.99 24.48 28.95 26.55 25.11 24.57
Operating Cost (c/mi) 20.53 17.53 16.16 15.47 22.44 18.96 17.43 16.53 25.23 21.12 19.08 18.06
Less Scrappage Value (c/mi) (0.18) (0.16) (0.15) (0.15) (0.18) (0.16) (0.15) (0.15) (0.18) (0.17) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 48.89 43.45 40.90 39.76 50.98 45.01 42.27 40.86 54.00 47.50 44.03 42.48
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 79 70 72 70 81 72 73 71 84 76 75 73
Motor power (kW): Constant 54 48 49 48 56 49 50 49 58 52 51 50
APU power (kW) 60 55 52 50 61 56 53 50 62 57 54 51
Generator power (kW) 57 53 50 49 58 54 51 49 59 55 52 50
Battery power (kW) 67 60 57 55 78 70 67 64 94 85 80 77
Battery energy (kWh) 9.7 8.9 8.4 7.9 11.3 10.4 9.8 9.3 13.7 12.6 11.7 11.1
Battery mass (kg) 199 172 155 143 232 200 182 166 280 243 218 200
Total vehicle mass (kg) 1,538 1,464 1,429 1,395 1,577 1,500 1,461 1,419 1,634 1,554 1,503 1,459
Accl Time on APU+Battery (s) 12.0 12.3 12.3 12.2 12.0 12.2 12.3 12.3 11.1 11.1 11.4 11.4
Accl Time on Battery-only (s) 16.0 16.0 16.0 16.0 14.0 14.0 14.0 14.0 12.0 12.0 12.0 12.0
Estimated all-elec range (mi) 20.9 21.1 20.4 20.1 23.7 24.0 23.5 23.0 27.6 28.2 27.3 26.9
Series Grid-Connected FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 12/10/99. The default battery life used, requires one battery replacement.
Run date: 3/31/2000
104
Argonne National Laboratory: Center for Transportation Research
A Methodology for Projecting Series Hybrid Electric Vehicle Cost
(All costs are in 1995 dollars. Battery costs are based on the ANL Delphi study.)
HEV (Series, Grid-Connected, UL Steel, Int Batt Lasts Veh Life, 12-16s All-electric) ADVISOR2 Components Matching
Vehicle type: Midsize Car Vehicle type: Midsize Car Vehicle type: Midsize Car
All-electric Z60 Time: 16 seconds All-electric Z60 Time: 14 seconds All-electric Z60 Time: 12 seconds
Battery type: NiMH (Intermediate) Battery type: NiMH (Intermediate) Battery type: NiMH (Intermediate)
Item 2005 2010 2015 2020 2005 2010 2015 2020 2005 2010 2015 2020
Purchase Price Purchase Price Purchase Price
Conventional Vehicle 21,200 $ 21,100 $ 21,100 $ 21,100 $ 21,300 $ 21,100 $ 21,100 $ 21,100 $ 21,800 $ 21,900 $ 21,600 $ 21,600 $
Hybrid Electric Vehicle
Common components 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $ 16,270 $
Additional aluminum cost - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $ - $
Auxiliary Power Unit 3,600 $ 3,530 $ 3,550 $ 3,620 $ 3,650 $ 3,580 $ 3,600 $ 3,620 $ 3,690 $ 3,630 $ 3,650 $ 3,670 $
Generator 1,760 $ 1,390 $ 1,180 $ 1,110 $ 1,780 $ 1,410 $ 1,200 $ 1,110 $ 1,810 $ 1,440 $ 1,220 $ 1,130 $
Inverter & Power Electronics 3,600 $ 2,640 $ 2,200 $ 1,890 $ 3,670 $ 2,690 $ 2,220 $ 1,910 $ 3,780 $ 2,810 $ 2,270 $ 1,950 $
Motor 2,320 $ 1,740 $ 1,570 $ 1,460 $ 2,370 $ 1,780 $ 1,590 $ 1,480 $ 2,450 $ 1,870 $ 1,620 $ 1,510 $
Transmission 510 $ 430 $ 430 $ 410 $ 520 $ 440 $ 430 $ 420 $ 530 $ 460 $ 440 $ 430 $
System control 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $ 420 $ 410 $ 390 $ 350 $
Other body parts 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $ 400 $ 390 $ 380 $ 370 $
Additional HVAC Cost 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $ 120 $ 110 $ 110 $ 110 $
First Battery Price 7,410 $ 5,820 $ 5,170 $ 4,650 $ 8,630 $ 6,790 $ 6,070 $ 5,410 $ 10,400 $ 8,240 $ 7,250 $ 6,510 $
Total HEV Price 36,500 $ 32,800 $ 31,300 $ 30,300 $ 37,900 $ 33,900 $ 32,300 $ 31,100 $ 39,900 $ 35,700 $ 33,600 $ 32,300 $
Operating Cost Operating Cost Operating Cost
Conventional Vehicle
Fuel cost (c/mi) 4.25 4.20 4.20 4.14 4.27 4.22 4.19 4.13 4.51 4.56 4.45 4.40
Non-fuel cost (c/mi) 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20 4.90 4.97 5.05 5.20
Total cost (c/mi) 9.15 9.17 9.25 9.34 9.17 9.19 9.24 9.33 9.41 9.53 9.50 9.60
Hybrid Electric Vehicle
Electricity cost (c/mi) 0.92 0.85 0.81 0.77 1.07 0.99 0.94 0.89 1.29 1.20 1.13 1.07
Fuel cost (c/mi) 2.09 2.03 1.97 1.91 1.88 1.82 1.77 1.71 1.60 1.52 1.48 1.43
Non-fuel cost (c/mi) 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89 5.55 5.63 5.72 5.89
Battery cost (c/mi) 7.27 5.63 4.93 4.43 8.47 6.56 5.78 5.15 10.21 7.97 6.91 6.20
Total cost (c/mi) 15.82 14.14 13.42 12.99 16.97 15.01 14.21 13.65 18.64 16.32 15.24 14.59
Life-cycle Cost Life-cycle Cost Life-cycle Cost
Conventional Vehicle
Purchase Price (c/mi) 20.80 20.40 20.10 20.10 20.90 20.40 20.10 20.10 21.39 21.17 20.58 20.58
Operating Cost (c/mi) 9.15 9.17 9.25 9.34 9.17 9.19 9.24 9.33 9.41 9.53 9.50 9.60
Less Scrappage Value (c/mi) (0.13) (0.13) (0.12) (0.12) (0.13) (0.13) (0.12) (0.12) (0.13) (0.13) (0.13) (0.13)
Total Life-cycle Cost (c/mi) 29.83 29.44 29.23 29.32 29.94 29.47 29.22 29.31 30.67 30.57 29.95 30.05
Hybrid Electric Vehicle
Purchase Price - Battery (c/mi) 28.55 26.08 24.90 24.44 28.72 26.21 24.99 24.48 28.95 26.55 25.11 24.57
Operating Cost (c/mi) 15.82 14.14 13.42 12.99 16.97 15.01 14.21 13.65 18.64 16.32 15.24 14.59
Less Scrappage Value (c/mi) (0.18) (0.16) (0.15) (0.15) (0.18) (0.16) (0.15) (0.15) (0.18) (0.17) (0.15) (0.15)
Total Life-cycle Cost (c/mi) 44.19 40.06 38.16 37.28 45.51 41.06 39.05 37.97 47.41 42.70 40.19 39.01
Important HEV Parameters Important HEV Parameters Important HEV Parameters
Motor power (kW): Peak 79 70 72 70 81 72 73 71 84 76 75 73
Motor power (kW): Constant 54 48 49 48 56 49 50 49 58 52 51 50
APU power (kW) 60 55 52 50 61 56 53 50 62 57 54 51
Generator power (kW) 57 53 50 49 58 54 51 49 59 55 52 50
Battery power (kW) 67 60 57 55 78 70 67 64 94 85 80 77
Battery energy (kWh) 9.7 8.9 8.4 7.9 11.3 10.4 9.8 9.3 13.7 12.6 11.7 11.1
Battery mass (kg) 199 172 155 143 232 200 182 166 280 243 218 200
Total vehicle mass (kg) 1,538 1,464 1,429 1,395 1,577 1,500 1,461 1,419 1,634 1,554 1,503 1,459
Accl Time on APU+Battery (s) 12.0 12.3 12.3 12.2 12.0 12.2 12.3 12.3 11.1 11.1 11.4 11.4
Accl Time on Battery-only (s) 16.0 16.0 16.0 16.0 14.0 14.0 14.0 14.0 12.0 12.0 12.0 12.0
Estimated all-elec range (mi) 20.9 21.1 20.4 20.1 23.7 24.0 23.5 23.0 27.6 28.2 27.3 26.9
Series Grid-Connected FHV: HEVCOST Model Run with ADVISOR mpg (for CV & HEV) dated 12/10/99. The battery life increased to last the life of HEV.
Run date: 3/31/2000
105
Section 4 Appendix B
Reality Check on ADVISOR Results: Highway
Cycle Fuel Savings for Parallel Hybrids
The ADVISOR results for the parallel grid-independent hybrids show strong improvements
(relative to the conventional drivetrain) in fuel economy on the Highway cycle. For example,
among vehicles achieving a 10-second Z60 time, the full hybrid achieves a 21% improvement in
fuel economy, and the mild hybrid (MHEV) achieves an 11% improvement. These values appear
high in relation to other results we are aware of, though comparisons should be made with
caution because of differences in assumptions and vehicle details. For example, OTAs analysis
of a series hybrid yielded an 8-percent improvement on the highway cycle for a Taurus-sized
(full) hybrid with 12-second Z60 time (OTA 1995). An (1999) estimated that shifting from a
Corolla conventional vehicle to a Prius-like hybrid (which is part series, part parallel) and
adjusting for performance yielded about a 4% gain on the Highway cycle. Note, however, that
series hybrids should be inferior to parallel hybrids in steady highway driving because of higher
electrical losses.
To obtain a better sense of whether the ADVISOR results are plausible, we examined the
energy flows and engine efficiency results of the ADVISOR runs in more detail and collected
and examined fuel economy test results from a group of vehicles with alternative engine options
with different power levels, to examine the effect of engine power on Highway cycle fuel
economy.
Table 1 presents key ADVISOR results for the Highway cycle for the vehicles. The results
indicate that the Highway mpg improvements increase with increased CV engine displacement
and power and with vehicle performance. The mild hybrids appear to obtain more improvement
in fuel economy than the full hybrids for each percent reduction in engine power. If correct, this
implies diminishing returns as the degree of hybridization increases. Both full and mild hybrids
are significantly heavier than the equivalent CV, but there is little weight difference between the
two hybrids. This has implications for the fuel economy savings achievable by the hybrids, since
added weight yields higher inertia and rolling resistance forces that the vehicles must overcome.
However, as we discuss elsewhere, the effects on fuel economy of mass changes are not as great
for a hybrid as for a conventional vehicle.
106
Table 1 ADVISOR Results for Parallel Grid-Independent ULS Hybrids
12 seconds 10 seconds 8 seconds
Factor CV MHEV FHEV CV MHEV FHEV CV MHEV FHEV
Mass, kg 1175 1246 1247 1248 1321 1328 1366 1453 1466
Engine power, kW 66 55 44 85 65 46 116 49 82
% Power drop 17 33 24 46 29 58
Hwy fuel eco, mpg 47 52 57 42 50 56 35 47 53
% Fuel eco increase 11 21 24 34 35 51
f.e, %/kW, % .65 .64 1.00 .74 1.21 .88
ADVISOR output includes a list of energy flows through various parts of the drivetrain,
e.g., the energy storage system, engine, clutch, motor/controller, gearbox and wheel/axle, etc.
Examining these flows during the Highway cycle for the 12-second CV and the pair of parallel
grid-independent hybrids yields the following conclusions:
1. The largest part of the HEVs fuel economy gain comes from increases in engine
efficiency, including (according to our interpretation of the ADVISOR accounting
system) the effects of idle-off.
2. The engine efficiency gain that comes with reductions in displacement and power is
highly nonlinear, so that the gain achieved by the MHEVs reduction from the CV is
nearly as great as that achieved by the FHEV, even though the FHEVs power reduction
is much greater.
3. Regenerative braking gains for the hybrids are relatively small in the Highway cycle,
though there is little difference between the FHEV and MHEV despite the MHEVs
smaller motor and battery. This implies that the braking events on the Highway cycle
are mild enough that the MHEVs small motor and battery can handle most of them
with little need for the mechanical brakes.
4. Losses in the energy storage system are significantly higher in the MHEV than the
FHEV. This is not surprising because, with its smaller battery, swings in SOC are wider
and more frequent in the MHEV.
The overall implication of these conclusions is that the largest share of energy savings on
the Highway cycle will come from increased efficiency of operation of the engine, due to more
frequent operation at points of lower fuel consumption. Given the speed profile of the cycle,
most of this should come from the effects of engine downsizing and the resulting frequency of
operation at a higher fraction of wide open throttle, where this engine technology is most
efficient. A small amount also comes from turning the engine off during braking and coasting.
Thus, an examination of the fuel economy effects of engine downsizing should then give us a
better idea of the realism of the ADVISOR results for the hybrids.
107
One way of examining the effects of engine downsizing on fuel economy is to examine the
fuel economy ratings of car models that offer multiple engine choices. There are a considerable
number of vehicles listed in the Fuel Economy Guide with two engine options for the same
vehicle. Figure 1 shows a plot of the percentage change in highway mpg versus the percentage
change in engine displacement for a number of such vehicles. All the vehicles have manual
transmissions (to match the simulated ADVISOR vehicle) and none have four-wheel drive
(4WD) or all-wheel drive (AWD). In addition, engine pairs were excluded when it was clear that
the pair had substantially different technology, e.g., where only one of the pair was turbocharged.
Displacement was used as a proxy for horsepower (which is not given in the guide), which
conforms with the scaling routine in ADVISOR.
58
This means that the last row in Table 1,
f.e,%/kW,%, will be identical to f.e,%/displacement,%.
There are a number of ways to draw the curve of f.e,% versus displacement,%, as shown
in the figure. The ratio of averages, which is appealing because it passes through the origin,
yields a value of 0.623 for f.e,%/displacement,% and f.e,%/kW,%. This matches well with
the ADVISOR results for the 12-second vehicles (0.63 for the full hybrid) but not with the results
for the faster vehicles. Note, however, that this ratio accounts only for engine downsizing, and
captures no other effects.
It is not obvious whether or not the net effects of hybridizing, aside from engine
downsizing, will be positive. These effects are:
Engine off during idling (only 6 seconds, or less than 1% of the cycle time, on the Highway
cycle) and deceleration
Regenerative braking gains
Losses in the energy storage system
Losses associated with added weight
As discussed above, downsizing the engine seems likely to be the primary factor affecting
the change in Highway fuel economy when one shifts from a CV to an HEV of equivalent
performance. Consequently, the relative similarity between the f.e,%/kW,% obtained from
ADVISOR and that obtained from the comparison of paired vehicles from the Fuel Economy
Guide is reassuring.
58
This is an approximation, at best.
108
0%
5%
10%
15%
20%
25%
30%
0% 5% 10% 15% 20% 25% 30% 35%
40% 45% 50%
Percent Reduction in Displacement
Developed by Ratio of Averages (not
regression fit forced through origin)
Developed by automatic
linear trendline fit in Excel
Developed by automatic
logarithmic trendline fit in Excel
Figure 1 HWY Cycle Percentage Gain in Fuel Economy as Function of Engine
Displacement Reduction
109
Section 4 Appendix C
Power and Mass Computations
for Initial Vehicle Sizing
Hybrid electric vehicles (HEVs) are expected to meet two performance criteria in order to
compete successfully with conventional vehicles. The first criterion is the time required to
accelerate from zero to 60 mph. The vehicles must also be able to negotiate a minimum grade at a
constant speed. Argonne developed a model to compute power requirements associated with
these criteria. Each drivetrain component is sized to meet the power requirements and its mass is
then computed. The model is described in this section.
Power Requirements
The procedure presented here estimates power requirements for accelerating on a flat road
(no grade) and negotiating a grade represented by an angle at a constant speed. We assume
that the air is still and vehicles are not required to accelerate from a stop to the maximum speed
up a hill or a ramp.
Acceleration Power
A hybrid vehicle that has an inertia mass of M
v
and is accelerating on a flat road (i.e., 0
grade) would require a power P
a
specified by the following equation.
r v d v a gvC M v AC
dt
dv
v M P + + =
3
2
1
(1)
where
v = Vehicle speed,
= Air density,
A = Vehicle frontal area,
C
d
= Coefficient of aerodynamic drag,
g = Gravity, and
C
r
= Coefficient of rolling resistance.
The power in equation 1 is at the wheels. After the acceleration power is determined, the
drivetrain components would be sized to allow for losses at various levels. Since the vehicle is
accelerating from a stop to a maximum speed v
m
, the acceleration power requirements in
equation 1 can be restated by integrating the first term from zero to v
m
.
110
dt v
t
AC
vdt g C v
t
M
P
m m t
m
d
t
r
m
m
v
a

+
)
`

+ =
0
3
0
2
2 2
1
(2)
Where t
m
is the time taken to reach the maximum speed.
The speed v in equation 2 is a function of time t. Under conditions involving a smooth
acceleration, the speed and time relationship can be plotted as shown in Figure 1. Assuming that
the vehicle speed and time relationship is approximated by a hyperbolic function, speed v in
equation 2 can be expressed as:
x
m
m
t
t
v v
|
|
|
.
|

\
|
= (3)
The exponent x in the above equation has a value in the range of 0.5-0.66 for zero to 60 mph
acceleration (Z60) times of 8-13 seconds.
Time
S
p
e
e
d
Figure 1 Vehicle Speed and Time Relationship during Acceleration
Two approaches were explored for computing the acceleration power requirements: (i) by
integrating equation 2, and (ii) by solving the power equation between time t
m
and t
m
+0.1.
The first approach provides an average value for the acceleration power. The integrated
equation 2 has two terms. The first term provides the power requirement for moving the vehicle
mass and the second term provides the average power required for overcoming air resistance.
The simplified equation for acceleration power P
a
can be written as shown below.
1 1
b M a P v a + = (4)
111
where
m r
m
m
t
m
r
m
m
gv C
x t
v
vdt
t
g C
t
v
a
m
+
+ = + =

1
1
2 2
2
0
2
1
(5)
and
3
0
3
1
) 2 1 ( 2
1
2
m
d
t
m
d
v AC
x
dt v
t
AC
b
m

+
= =

(6)
Equation 4 shows that the acceleration power has two parts: one is linear to vehicle mass
and the other is a constant that depends on vehicle design. The term a
1
is a function of
acceleration specifications (i.e., maximum speed and time to reach it) and rolling resistance. The
term b
1
represents power required to overcome the aerodynamic drag and is a function of drag
coefficient and frontal area. The aerodynamic power requirement does not depend on vehicle
mass.
The second approach provides an estimate of passing power requirement at the maximum
speed v
m
. Since the vehicle accelerates for only 0.1 seconds, this estimate would be very close to
the maximum power required to reach the target speed. This approach would provide a power
value higher than the first approach.
Equation 2 is used for this approach. The value of v is v
m
, here for aerodynamic drag and
rolling resistance, the time range is t
m
to t
m
+0.1, and dt has a value of 0.1. The equation can be
used in the following form.
r m v m d
x
m
m
m
x
m
m
m m
v a C gv M v AC
t
t
v
t
t
v v M P + +
(
(
(

|
.
|

\
| +
=
|
|
|
|
.
|

\
|
3
2
1
1 . 0
1 . 0
(7)
r m v m d
x
m
m m
v
C gv M v AC
t
t v
M + +
(
(

|
|
.
|

\
| +
=
3
2
2
1
1
1 . 0
1 . 0
(8)
This equation can be simplified in a form similar to that of equation 4:
b aM P
v a
+ = (9)
where
m
r
x
m
m m
gv C
t
t v
a +
(
(

|
|
.
|

\
| +
= 1
1 . 0
1 . 0
2
(10)
112
3
2
1
m d
v AC b = (11)
Grade-Climbing Power
Power (P
g
) required for negotiating a grade that is represented by angle at a constant speed
v
g
could be written as follows:
cos sin
2
1
3
r
g
v
g
v
g
d g C gv M gv M v AC P + + = (12)
This equation can be simplified as:
P cM d g v = + (13)
where
( ) c gv C
g
r = + sin cos and (14)
3
2
1
g
dv AC d = (15)
The grade-climbing power requirement also has two parts, one linear to vehicle mass and
the other a constant dependent on vehicle design. The term c represents the effects of grade
specifications (i.e., speed and grade angle) and rolling resistance, while the term d represents the
power required to overcome aerodynamic drag.
Both the acceleration and grade climbing power requirements are dependent on vehicle
mass. The acceleration power requirement is usually higher than the grade-climbing power
requirement because the inertial forces that must be overcome during rapid acceleration generally
will outweigh the weight force of a grade climb (unless the grade is extremely steep). The
Partnership for a New Generation of Vehicles (PNGV) has compiled a set of vehicle
specifications. Under these specifications a high fuel economy vehicle should accelerate from
zero to 60 mph in 12 seconds and should be able to sustain a constant speed of 55 mph for
20 minutes while climbing a 6.5% grade. Figure 2 shows the general relationship between the
two power requirements and vehicle mass under these specifications. The gap between the two
power requirements increases as vehicle mass increases.
113
Vehicle Mass
P
o
w
e
r
R
e
q
u
i
r
e
m
e
n
t
For acceleration
For grade climbing
Figure 2 Power Requirements as a Function of Vehicle Mass
Vehicle Mass
A vehicle has three distinct mass groups: (1) body, (2) chassis, and (3) drivetrain. The body
and chassis for the hybrid and conventional vehicles would be nearly identical. The conventional
vehicle mass and contributions by individual group have been analyzed earlier (Stodolsky et al.
1995a; OTA 1995). Conventional steel vehicles have 73-74% of their total vehicle mass in body
and chassis groups. The optimal use of ultralight steel might reduce the total vehicle mass by
10-12% while optimal use of aluminum would reduce the total mass by 31% (Stodolsky et al.
1995b; OTA 1995; AISI 1998). In the case of a hybrid vehicle, a smaller power unit (PU), a
motor and an inverter, a generator, a battery pack, and a gear-drive or transmission (depending
upon the hybrid design) will replace the conventional engine and transmission.
The total inertia mass M
v
is expressed as follows:
M M M M M v b ch dt l = + + + (16)
where:
M
b
= Body mass,
M
ch
= Chassis mass,
M
dt
= Drivetrain mass, and
M
l
= Load.
The body and chassis mass, M
b
and M
ch
, depend on vehicle design and their sum M
f
can be
treated as fixed.
114
The drivetrain has several components, each with its own efficiency and specific power.
The efficiency helps determine the power rating for the component, and the specific power
determines its mass. The drivetrain mass is specified as:
tran gen mot bat pu dt M M M M M M + + + + = (17)
where
M
pu
= PU mass,
M
bat
= Battery mass,
M
mot
= Motor and inverter mass,
M
gen
= Generator mass, and
M
tra
= Transmission mass.
Let S
pu
, S
bat
, S
mot
, S
gen
, and S
tran
be the specific power for each of the five components and P
pu
, P
bat
,
P
mot
, P
gen
, and P
tran
be the power ratings. Then mass of each component can be computed as
power divided by specific power:
pu pu pu S P M = (18)
M P S bat bat bat = (19)
M P S mot mot mot = (20)
M P S gen gen gen = (21)
M P S tran tran tran = (22)
These mass and specific power values are for the complete component assembly including
auxiliary and/or supporting units. For example, the motor assembly includes the inverter.
Component Sizing
The acceleration and grade-climbing power estimates from the above described procedure
represent power delivered at the wheels. Each component has its own power conversion
efficiency and some losses are involved in mechanical components such as bearings. A design
factor, k, is used to account for other losses and contingencies.
The power rating of a drivetrain component depends upon the HEV system configuration,
series or parallel. The motor delivers all the required power through the transmission in a series
HEV while both the power unit and motor deliver power through the transmission in a parallel
HEV. A series HEVs transmission is simple, consisting of a few reduction gears, while a
parallel HEVs transmission is relatively complex, requiring linking of the two power sources.
Also, the power units link to the drive axle requires greater control compared to the link of a
115
motor. The component sizing procedures are different for the two configurations with some
assumptions common to both. We assume that the power unit supplies the total power necessary
for grade climbing in both the configurations. The battery usually supplies the difference
between the power required to accelerate from zero to 60 mph and that for grade climbing. The
battery power would be higher if the HEV is required to have some all-electric travel capability
unless much lower acceleration capability was acceptable for the all-electric vehicle operations.
When HEVs internal combustion engine (ICE) power unit is sized to meet the minimum
grade-climbing power and the battery pack is sized to meet the difference between acceleration
and grade-climbing power values, the battery-supplied power is usually greater than 25% of the
total HEV power. Such HEVs are termed full HEVs. The ICE represents a mature and
affordable technology while the electric drive, consisting of motor, inverter/controller, and
battery pack, represents an evolving technology. Consequently, it would be economical to
reduce the size of the electric drive in some cases. By specifying a higher grade-climbing
requirement and keeping the Z60 time unchanged, a user may increase the power rating of the
ICE power unit. Since traction motor and battery provide the difference between acceleration
and grade-climbing power requirements, their sizes are reduced with increase in the gradeability
specification. Consequently, the battery power share of total power drops. Such HEVs are
termed mild HEVs.
Series HEV
Figure 3 shows a schematic diagram of the series HEV drivetrain. The power unit is
connected to the generator. The generator supplies power to either the motor or the battery pack.
The battery pack supplies power to the motor and also receives some recharge electricity fed
back from the motor (acting as a generator) during regenerative braking. The motor is connected
to the transmission (or reduction gears) to drive the wheels. The gears also transmit power back
to the motor during regenerative braking.
Power Unit Generator
Motor Battery
G
e
a
r
D
r
i
v
e
Figure 3 Series HEV Drivetrain Components and Their Connections
116
Let
gen
,
mot
, and
tran
be the average power efficiency of the generator, motor, and transmission,
respectively. These efficiencies are for the component assemblies including supporting units.
The formulas for computing acceleration power requirement P
a
and grade-climbing power
requirement P
g
are specified in equations 9 and 13, respectively. The component power ratings
for a series HEV that is not required to have any all-electric acceleration capability can be
computed as follows:
gen mot tran g pu P P = (23)
P P gen g tran mot = (24)
P P mot a tran = (25)
( ) P P P bat a g tran mot = (26)
The acceleration power requirement is assumed to be larger than the grade-climbing power
requirement in the above equations. Alternatively, the PU power can be computed on the basis of
the smaller of P
a
and P
g
, with motor power based on the larger of the two values, and battery
power on the absolute value of the difference. The grade-climbing power requirement will
exceed the acceleration power requirement when the grade to be negotiated is high and the time
allowed to accelerate from zero to 60 mph is also high. In such a case, the battery would be used
only at low speeds and during grade climbing. The battery would also serve its usual function of
a sink to absorb excess energy during the periods of low power demand and braking. The motor
power should always be based on the larger of the two power requirements to satisfy the highest
power demand.
A series HEV that is required to have an all-electric acceleration capability will have a
larger battery pack. This battery pack size is determined by the time required to accelerate from
a stop to the maximum speed v
m
on battery power alone. Let P
e
be the all-electric acceleration
power requirement:
P a M b e v = + ' ' (27)
Where a' and b' are parameters that match the required all-electric acceleration time t
m
in
equation 8.
The battery power is computed as follows:
P P bat e tran mot = (28)
The equations 9 and 13 for power computation require that the vehicle inertia mass M
v
be
known. However, the vehicle inertia mass has four contributors (see equation 16): body, chassis,
drivetrain, and load. The body and chassis mass depend on the type of vehicle (i.e., small or
midsize car) and type of material (i.e., conventional steel, ultralight steel, or aluminum) while the
drivetrain mass depends on the power requirements. The load is usually fixed, but may have
different values for acceleration and grade climbing (the PNGV vehicle criteria measure
117
acceleration with a load of 300 pounds, whereas gradeability is measured with a higher load of
1100 pounds). Thus, the power and mass computation, in part, depend on the drivetrain
component mass.
Mass for a Series HEV with No All-Electric Acceleration Capability
The total vehicle mass M
veh
is defined as:
M M M veh f dt = +
where
M
f
= Fixed mass (i.e., sum of body and chassis mass),
M
dt
= Drivetrain mass,
Let
P
a
= Power requirements for acceleration,
P
g
= Power requirements for grade climbing,
M
c
= Mass of drivetrain component c,
P
c
= Power rating of drivetrain component c,
S
c
= Specific power of drivetrain component c, and

c
= efficiency of drivetrain component c.
Then the drivetrain mass is estimated as shown below:
tran bat gen pu mot dt M M M M M M + + + + =
tran
tran
bat
bat
gen
gen
pu
pu
mo
mot
S
P
S
P
S
P
S
P
t S
P
+ + + + =
tran tran
a
bat mot tran
g a
gen mot tran
g
pu mot gen tran
g
mot tran
a
S
P
S
P P
S
P
S
P
S
P

+

+ + + =
|
|
|
|
.
|

\
|
+ +
|
|
.
|

\
|
+ + =
bat mot tran gen mot tran pu gen mot tran
g
tran tran bat mot tran mot tran
a
S S S
P
S S S
P

1 1 1 1 1 1
= + e P f P s a s g (29)
118
Here
|
|
.
|

\
|
+ + =
|
|
.
|

\
|
+ + =
tran bat mot mot trans tran tran bat mot tran mot tran
s
S S S S S S
e
1 1 1 1 1 1 1

|
|
.
|

\
|
+ = + =
|
|
|
|
.
|

\
|
bat mot gen mot pu gen mo trans bat mot tran gen mot tran pu gen mot tran
s
S S S t S S S
f

1 1 1 1 1 1 1
Total vehicle mass g s a s f dt f veh P f P e M M M M + + = + = .
Note that acceleration power P
a
and grade-climbing power P
g
are computed with inertia mass
values that include some load. The design factor k is applied to account for other losses.
M M e k a M M b f k c M M d veh f s veh aload s veh gload = + + + + + + [ ( ) ] [ ( ) ]
where
M
aload
= the load during acceleration and
M
gload
= the load during grade climbing.
M e kaM f kcM M e k aM b f k cM d veh s veh s veh f s aload s gload = + + + + ( ) ( )
( ) ( )
( )
M
M e k aM b f k cM d
k e a f c
veh
f s aload s gload
s s
=
+ + + +
+ 1
(30)
Mass for a Series HEV with All-Electric Acceleration Capability
The above procedure applies to a series HEV that is not required to accelerate on battery
power. A series HEV may be required to have some all-electric acceleration capability to reduce
emissions. Such an HEV would function as an electric vehicle as long as its battery maintains a
state of charge above a predetermined minimum level. The power unit would turn on when the
battery charge reaches this predetermined level. The State of California appears to prefer such an
HEV because it would not have any tailpipe emissions while running on battery power.
The drivetrain mass of such a series HEV is determined as follows.
Let P
e
= Power requirements for all-electric acceleration (P
e
<= P
a
)
119
Then
tran bat gen pu mot dt M M M M M M + + + + =
tran
tran
bat
bat
gen
gen
pu
pu
mo
mot
S
P
S
P
S
P
S
P
t S
P
+ + + + =
tran tran
a
bat mot tran
e
gen mot tran
g
pu mot gen tran
g
mot tran
a
S
P
S
P
S
P
S
P
S
P

+ + + + =
=
|
|
.
|

\
|
+ + +
|
|
.
|

\
|
+
|
|
|
|
.
|

\
|
bat mot tran
e
gen mot tran pu gen mot tran
g
tran tran mot tran
a
S
P
S S
P
S S
P

1 1 1 1 1
=
s
a
s
g
s
e e P f P h P
'
'
'
+ + (31)
where
|
.
|

\
|
+ =
tran mot tran
s
S S
e
1 1 1
'

,
|
|
.
|

\
|
+ =
gen mot pu gen mot tran
s
S S
f

1 1 1
'
, and
bat mot tran
s
S
h

1
'
= .
Total vehicle mass
e s g s a s
f dt f veh P h P f P e M M M M
' ' '
+ + + = + =
.
The power requirements P
a
and P
e
are computed with a load of M
aload
and P
g
is computed with a
load of M
gload
. Also, the design factor k is applied:
] ' ) ( ' [
'
] ) ( [
'
] ) ( [
'
b M M a k
h
d M M c k f b M M a k
e
M M aload veh
s
gload veh
s
aload veh
s
f veh + + + + + + + + + =
where a' and b' are the parameters that match the all-electric acceleration time requirement.
+ + + + + = ) (
'
) (
'
'
'
'
'
d cM k f b aM k
e
M M ka
h
kcM f kaM
e
M gload
s
aload
s
f veh
s
veh
s
veh
s
veh
) ' ' (
'
b M a k
h
aload
s
+
|
.
|

\
|
+ +
+ + + + + +
=
'
'
'
'
1
) ' ' (
'
) (
'
) (
'
a
h
c f a
e
k
b M a k
h
d cM k f b aM k
e
M
M
s
s
s
aload
s
gload
s
aload
s
f
veh (32)
120
Parallel HEV
A schematic diagram of the parallel HEV drivetrain is shown in Figure 4. In a parallel
HEV, both power unit and motor supply power to the wheels. A generator is optional because
the motor can be reversed during episodes of low power demand to recharge the battery.
Normally, the motor is the primary source of power during low speeds and congested conditions.
Since extended periods of congested driving under such a control strategy could drain the battery
and the motor cannot generate any power to recharge the battery when the vehicle is stopped,
parallel HEVs may be equipped with a generator. In our analysis, we assumed that all parallel
HEVs would be equipped with a generator.
Power Unit
Motor
Generator
Battery
T
r
a
n
s
m
i
s
s
i
o
n
G
e
a
r
s
Figure 4 Parallel HEV Drivetrain Components and Their Connections
The power rating of the generator would depend on the vehicle control strategy. It could
supply power directly to the motor or only to the battery pack. Thus, a generators power rating
could be equal to or lower than the power rating of the motor. We assumed the generator to have
one-third the power rating of the power unit in our computations. Using the earlier described
convention, the power ratings of parallel HEV drivetrain components are computed as follows,
assuming no all-electric acceleration capability:
tran g pu P P = (33)
mot tran g gen P P 3 = (34)
( ) P P P mot a g trant = (35)
121
( ) P P P bat a g tran mot = (36)
In the above computations, acceleration power demand is assumed to be larger than the
grade-climbing power demand. This assumption is true for most driving conditions in the United
States. The battery supplies the difference between acceleration and grade-climbing power
demands. The motor is connected to the transmission and uses power from either the battery or
the generator. The motor may draw power from both the battery and the generator under some
control strategies.
A parallel HEV may be required to have some all-electric acceleration and travel capability.
Such a parallel HEV would have a larger battery pack, a larger motor, and could benefit from a
larger generator. However, the generator size is limited by the PU power rating, its prime mover.
The power requirements for these three components are determined as follows:
All-electric acceleration power requirement P a M b e v = + ' ' .
where a' and b' satisfy the all-electric acceleration time requirement.
mot tran e bat P P = (37)
P P mot e tran = (38)
mot tran g gen P P 3 / = (39)
Mass for a Parallel HEV with No All-Electric Acceleration Capability
The total vehicle mass M M M veh f dt = +
Where M
f
is the fixed mass and M
dt
is the drivetrain mass.
tran bat gen pu mot dt M M M M M M + + + + =
tran
tran
bat
bat
gen
gen
pu
pu
mot
mot
S
P
S
P
S
P
S
P
S
P
+ + + + =
tran tran
a
bat mot tran
g a
gen mot tran
g
pu tran
g
mot tran
g a
S
P
S
P P
S
P
S
P
S
P P

+

+ + +

=
3
+
|
|
.
|

\
|
+ + =
tran tran bat mot tran mot tran
a
S S S
P

1 1 1
|
|
.
|

\
|
+
bat mot tran gen mot tran mot tran pu tran
g
S S S S
P

1
3
1 1 1
= + e P f P p a p g (40)
122
where
|
|
.
|

\
|
+ + =
tran ba mot mot tran
p
S t S S
e
1 1 1 1

and
|
|
.
|

\
|
+ =
bat mot gen mot mot pu tran
p
S S S S
f

1
3
1 1 1 1
.
The vehicle mass M
veh
can be computed the same way as was done for the series HEV (see
equation 30):
( ) ( )
( )
M
M e k aM b f k cM d
k e a f c
veh
f p aload p gload
p p
=
+ + + +
+ 1
(41)
Mass for a Parallel HEV with All-Electric Acceleration Capability
A parallel HEV that is required to have some all-electric acceleration capability would
likely have a larger battery pack, motor, and generator. The drive train mass is computed as
follows:
bat mot tran
e
gen mot tran
g
mot tran
e
pu tran
g
tran tran
e g
dt
S
P
S
P
S
P
S
P
S
P P
M

+ + + +
+
=
3
|
|
.
|

\
|
+ + +
|
|
.
|

\
|
+ + =
bat mot mot tran tran
e
gen mot pu tran tran
g
S S S
P
S S S
P 1 1 1
3
1 1 1
e
p
g
p
P
h
P f
'
'
+ =
(42)
where
|
|
.
|

\
|
+ + =
gen mot pu tran tran
p
S S S
f
3
1 1 1 1
'
and
|
|
.
|

\
|
+ + =
bat mot mot tran tran
p
S S S
h

1 1 1 1
'
.
Notice that
'
p
h is computed the same way as e
p
in equation 40.
The vehicle mass M
veh
is computed as follows:
e
p
g
p
f veh kP
h
kP f M M
'
'
+ + =
] ' ) ( ' [
'
)] ) ( [
'
b M M a k
h
d M M c k f M aload veh
p
gload veh
p
f + + + + + + =
123
|
.
|

\
|
+
+ + + +
=
'
'
'
1
) ' ' (
'
) (
'
a
h
c f k
b M a k
h
d cM k f M
p
p
aload
p
gload
p
f
(43)
Application of the Procedure
The following example demonstrates the application of the above-described power and mass
computing procedure to a midsize car. Data for the example were gathered from several sources
and were complemented with technical judgment in some cases.
Midsize HEV
Two midsize cars (1998 model year Chrysler Cirrus and Chevrolet Malibu) provide a
baseline conventional vehicle (CV) for this example. The cars have an average mass of 1,418 kg
(3,125 lb). Assuming that the future cars will use ultralight steel, we can assign a mass of
1,322 kg (2,915 lb) in 2005, an 11% reduction. We assume that more improvements will result
in a 0.5% reduction in mass every 5 years after 2005. We also assume that the body and chassis
mass contribute 73.5% of the total vehicle mass (Stodolsky et al. 1995a). An HEVs body and
chassis are assumed to have 5% higher mass than a CVs due to additional components for power
electronics and stiffeners to support the battery pack. The resulting body and chassis mass
values are shown in Table 1. Three body types are listed in Table 1: (1) ultralight steel body,
(2) partial aluminum body, and (3) maximum aluminum body. A partial aluminum vehicle could
weigh 10% less than its ultralight steel body counterpart and a maximum aluminum vehicle could
weigh 22% less than its ultralight steel body counterpart (Stodolsky et al. 1995b).
The selected baseline vehicles are 70.5 inches wide and 55.3 inches tall. These dimensions
were kept unchanged in this analysis. A factor is applied to the vehicle cross section to account
for ground clearance and side curvatures. This factor was 0.83 in 2005, 0.82 in 2010, 0.81 in
2015, and 0.80 in 2020. The resulting frontal area ranged from 2.09 square meters in 2005 to
2.01 square meters in 2020. We also assume that the future cars will have much lower
aerodynamic drag coefficients and their tires will have very low rolling resistance. Table 1 also
lists aerodynamic and rolling resistance parameters.
Some design criteria are necessary for computing power requirements. The vehicle is
required to accelerate from zero to 60 mph (26.82 m/s) in some fixed time. It should also have
sufficient power to climb a predetermined grade at a constant speed. Table 2 lists maximum
grade and acceleration requirements. The vehicles characterized here are capable of negotiating
124
Table 1 Midsize Vehicle Characteristics
Characteristics 2005 2010 2015 2020
Body & Chassis Mass in kg (lb in parenthesis)
Ultralight steel body 974 (2,147) 969 (2,136) 964 (2,125) 959 (2,114)
Partial aluminum body 876 (1,931) 872 (1,922) 868 (1,914) 863 (1,903)
Maximum aluminum body 759 (1,673) 756 (1,667) 752 (1,658) 748 (1,649)
Aerodynamic and Rolling
Frontal area (m
2
) 2.09 2.06 2.04 2.01
Coefficient of drag (Cd) 0.27 0.26 0.25 0.24
Coefficient of rolling resistance (Cr) 0.0080 0.0075 0.0070 0.0065
Table 2 Other Design Parameters
Item 2005 2010 2015 2020
Maximum grade at 55 mph (%) 6.5 6.5 6.5 6.5
Time to accelerate from zero to 60 mph (s) 12 12 12 12
Time to accelerate from zero to 60 mph all-electrically
(s) (where applicable)
16 16 16 16
Loading during acceleration (kg) 136 136 136 136
Loading during grade climbing (kg) 499 499 499 499
the maximum grade on PU power only. They do not have enough electrical power to negotiate
these grades. We characterize two vehicles each for parallel and series hybrids. One of these two
vehicles does not have any all-electric acceleration capability while the other has such a
capability. The vehicles with all-electric acceleration capability would have larger battery packs
that would be charged from the electricity grid. They are often referred to as grid-connected
HEVs. The all-electric acceleration time is specified the same for both series and parallel grid-
connected HEVs in our example. The table also shows vehicle loading during acceleration and
grade climbing. The acceleration load of 136 kg is adapted from USEPAs standard procedure.
The grade-climbing load of 499 kg represents 6 passengers and 91 kg of luggage and is adapted
from the vehicle specifications compiled by the Partnership for a New Generation of
Vehicles (PNGV).
Battery
This example uses a modified version of the ANLs Delphi Study data for the nickel metal
hydride battery (Vyas et al. 1997). Delphi respondents appear to have specified battery
technologies that were prevalent in the early 1990s. These data should now reflect the
availability of high specific-power batteries for HEV use (Table 3). This update was done by
125
Table 3 Nickel Metal Hydride Battery Characteristics
for CV-like, Grid-Independent HEV
Characteristic 2005 2010 2015 2020
Specific power at 20% SOC (W/kg) 500 520 546 573
Specific energy (Wh/kg) 43 46 48 50
applying factors to the Delphi Study data. The factors for specific power and specific energy
were computed by using the battery data for Toyota Prius and Toyota RAV-4.
59
An alternative set of battery characteristics that has lower specific power and higher specific
energy was developed for the grid connected HEV. Such batteries would have characteristics
that fall somewhere in the middle of the characteristics of Toyota Prius and RAV-4 batteries.
Factors were developed to arrive at a set of battery characteristics that would provide a range of
20 miles in 2005. The characteristics in Table 4 incorporate these factors.
Component-Specific Power and Efficiencies
The power computation procedure requires specific power and efficiency information for
each drivetrain component. Table 5 lists the values used in this analysis. We analyzed data on
Unique Mobility motor SR218H and inverter CA40-300L. The motor has a peak specific power
of 1,110 W/kg and the two units have a combined specific power of 875 W/kg. We assumed a
10% increase in the specific power by 2005.
Power and Mass Computations
In this analysis, we assumed that the hybrid cars would have ultralight steel bodies. Values
for such fixed parameters as gravity and air density are taken from standard physical tables. The
value for gravity is 9.8 m/s
2
and air density is 1.23 kg/m
3
. Also, we apply a design factor, k, in
computing power to account for other mechanical losses and contingencies. The value of k is 1.1
in these examples.
Table 4 Nickel Metal Hydride Battery Characteristics for
EV-like, Grid-Connected HEV
Characteristic 2005 2010 2015 2020
Specific power at 20% SOC (W/kg) 335 350 370 385
Specific energy (Wh/kg) 49 52 54 56
59
This calculation was completed prior to the U.S. introduction of the Prius; the U.S. version is equipped
with an improved battery that attains a higher specific power than used here.
126
Table 5 Specific Power and Efficiency Values for Drivetrain Components
Component Type 2005 2010 2015 2020
Specific Power (W/kg)
Motor & generator Permanent magnet 1,225 1,300 1,350 1,400
Power unit Gasoline 325 330 335 340
Motor with inverter 960 985 1,010 1,035
Transmission For parallel HEV 1,300 1,320 1,340 1,360
For series HEV 1,625 1,650 1,675 1,700
Efficiency (%)
Motor & inverter Permanent magnet 90 92 92 93
Generator Permanent magnet 95 95 95 96
Transmission-during acceleration 90 92 92 92
Parallel HEV with No (or Minimal) All-Electric Acceleration
First we apply the procedure to the year 2005 parallel HEV that is not required to have any
all-electric acceleration capability. The HEV will have the high specific-power battery with the
characteristics shown in Table 3. The vehicle will accelerate from zero to 60 mph in 12 seconds.
The value of exponent x (in equation 3) was set at 0.56.
The power requirement P
a
for acceleration from zero to 60 mph (26.82 m/s) = ( ) k aM b v + .
82 . 26 8 . 9 008 . 0 1
12
1 . 0 12
1 . 0
82 . 26
1
1 . 0
1 . 0
56 . 0
2 2
+
(
(

|
.
|

\
| +
= +
(
(

|
|
.
|

\
| +
= m r
x
m
m m
gv C
t
t v
a
= 33.5 + 2.1 = 35.6 W/kg
2 . 6695 82 . 26 27 . 0 09 . 2 23 . 1
2
1
2
1
3 3
= = =
m
dv AC b W
The power requirement P
g
for grade climbing at a constant speed of 55 mph (24.59 m/s) =
( ) k cM d v + .
( ) 6 . 17
065 . 0 1
1
008 . 0
065 . 0 1
065 . 0
59 . 24 8 . 9 sin
2 2
=
|
|
.
|

\
|
+
+
+
= + = Cos C gv c r g W/kg
1 . 5160 59 . 24 27 . 0 09 . 2 23 . 1 5 . 0
2
1
3 3
= = = g dv AC d W
( )
M
M e k aM b f k cM d
k e a f c
veh
f p aload p gload
p p
=
+ + + +
+
( ) ( )
1
127
3
10 481 . 4
1300
1
500 9 . 0
1
960
1
9 . 0
1 1 1 1 1

=
|
|
.
|

\
|
+

+ =
|
|
.
|

\
|
+ + =
tran bat mot mot tran
p
S S S
e

3
10 128 . 0
500 9 . 0
1
1225 9 . 0 3
1
960
1
325
1
9 . 0
1 1
3
1 1 1 1

=
|
|
.
|

\
|

+ =
|
|
.
|

\
|
+ =
bat mot gen mot mot pu tran
p
S S S S
f

( )
kg 257 , 1
6 . 17 10 128 . 0 6 . 35 10 481 . 4 1 . 1 1
) 1 . 5160 499 6 . 17 ( 1 . 1 10 128 . 0 ) 2 . 6695 136 6 . 35 ( 1 . 1 10 481 . 4 974
3 3
3 3
=
+
+ + + +
=


veh M
kW 9 . 61 909 , 61 ] 2 . 6695 ) 136 1257 ( 6 . 35 [ 1 . 1 ) ( = = + + = + = W b aM k P v a
kW 6 . 39 575 , 39 ] 1 . 5160 ) 499 1257 ( 6 . 17 [ 1 . 1 ) ( = = + + = + = W d cM k P v g
kW 44 9 . 0 6 . 39 = = = tran g pu P P
kW 3 . 16 ) 9 . 0 9 . 0 3 ( 6 . 39 3 / = = = mot tran g gen P P
kW 8 . 24 9 . 0 ) 6 . 39 9 . 61 ( ) ( = = = tran g a mot P P P
kW 6 . 27 ) 9 . 0 9 . 0 ( ) 6 . 39 9 . 61 ( ) ( = = = mot tran g a bat P P P
kg 3 . 135 325 44000 = = = pu pu pu S P M
kg 8 . 25 960 24800 = = = mot mot mot S P M
kg 3 . 13 1225 16300 = = = gen gen gen S P M
kg 2 . 55 500 27600 = = = bat bat bat S P M
kg 9 . 52 ) 1300 9 . 0 ( 61900 = = = tran tran a tran S P M
Parallel HEV with All-Electric Acceleration Capability
In this example, the parallel HEV is expected to accelerate, all-electrically, from zero to
60 mph in 16 seconds. The battery power is assumed to be available immediately and no
gearshifts are necessary. The power unit is sized to provide the grade climbing power. The
value of exponent x in equation 3 is 0.49 for all-electric acceleration. The combined power unit
and battery power would be much higher than what is needed for accelerating the vehicle from
zero to 60 mph in 12 seconds. The vehicle is assumed to be equipped with a battery pack that
has the characteristics listed in Table 4. The battery pack has lower specific power and higher
specific energy compared to the battery pack used for the parallel HEV that has no all-electric
acceleration capability.
Power for accelerating the vehicle, all-electrically, from zero to 60 mph (26.82 m/s)
is Pe k a Mv b = + ( ' ' ) .
128
W/kg 1 . 24 1 . 2 22 82 . 26 8 . 9 008 . 0 1
16
1 . 16
1 . 0
82 . 26
1
16
1 . 0 16
1 . 0
'
49 . 0
2 2
= + = +
(
(

|
.
|

\
|
= +
(
(

|
.
|

\
| +
= m r
x
m
gv C
v
a
W b b 2 . 6695 ' = =
The values of c and d remain unchanged.
) '
' '
( 1
) ' ' (
'
) (
'
a h c f k
b M a k h d cM k f M
M
p p
aload
p
gload
p
f
veh
+
+ + + +
=
3
10 609 . 4
1225 9 . 0 3
1
325
1
1300
1
9 . 0
1
3
1 1 1 1
'
= |
.
|

\
|

+ + =
|
|
.
|

\
|
+ + =
gen mot pu tran tran
p
S S S
f

|
|
.
|

\
|
+ + =
bat mot mot tran tran
p
S S S
h

1 1 1 1
'
3
10 697 . 5
335 9 . 0
1
960
1
1300
1
9 . 0
1

=
|
|
.
|

\
|

+ + =
kg 457 , 1
) 1 . 24 10 697 . 5 6 . 17 10 609 . 4 ( 1 . 1 1
) 2 . 6695 136 1 . 24 ( 1 . 1 10 697 . 5 ) 1 . 5160 499 6 . 17 ( 1 . 1 10 609 . 4 974
3 3
3 3
=
+
+ + + +
=


Mveh
kW 6 . 49 W 582 , 49 ] 2 . 6695 ) 136 1457 ( 1 . 24 [ 1 . 1 ] ' ) ( ' [ = = + + = + + = b M M a k P aload veh e
kW 4 . 43 W 441 , 43 ] 1 . 5160 ) 499 1457 ( 6 . 17 [ 1 . 1 ] ) ( [ = = + + = + + = d M M c k P gload veh g
kW 3 . 48 9 . 0 4 . 43 = = = tran g pu P P
kW 1 . 55 9 . 0 6 . 49 = = = tran e mot P P
kW 9 . 17 ) 9 . 0 9 . 0 3 ( 4 . 43 3 / = = = mot tran g gen P P
kW 2 . 61 9 . 0 9 . 0 6 . 49 = = = mot tran e bat P P
kg 5 . 148 325 48300 = = = pu pu pu S P M
kg 4 . 57 960 55100 = = = mot mot mot S P M
kg 6 . 14 1225 17900 = = = gen gen gen S P M
kg 7 . 182 335 61200 = = = bat bat bat S P M
129
kg 5 . 79
1300
103400
= =
+
=
tran
mot pu
tran
S
P P
M
Because this HEV has a large battery pack and a power unit that is capable of negotiating a 6.5%
grade at a constant speed of 55 mph, its performance on combined PU and battery power would
be very good. The acceleration time can be computed by solving the equation for time (t
m
). The
power P
a
is replaced by the sum of grade climbing power P
g
and all-electric acceleration power
P
e
.
|
|
.
|

\
|
+ +
(
(

|
|
.
|

\
| +
= +
3
2
2
1
1
1 . 0
1 . 0
m
d v m r
x
m
m
v
m
g e v AC M gv C
t
t M v
k P P
( )
|
|
.
|

\
|

+
(

+
= 1 1 1 . 0
5 . 0
1 . 0
/ 1
2
3
x
v m
v m r m d e g
m
M v
M gv C v AC k P P
t

Because the Z60 time on combined PU and battery power is expected to be close to 8 seconds,
we assign a value of 0.6 to the exponent x.
|
|
.
|

\
|

)
`

+
(

+
+
= 1 1 1 . 0
) 136 1457 ( 82 . 26
) 136 1457 ( 82 . 26 8 . 9 008 . 0 82 . 26 27 . 0 09 . 2 23 . 1 5 . 0 1 . 1 / 93000
1 . 0
6 . 0 / 1
2
3
m t
= 9.2 seconds
Series HEV with No All-Electric Acceleration Capability
The year 2005 series HEV would be equipped with the high specific power battery listed in
Table 3. The acceleration power requirement P k aM b a v = + ( ) :
W/kg 6 . 35 1 . 2 5 . 33 82 . 26 8 . 9 008 . 0 ) 004658 . 0 (
1 . 0
82 . 26
1
12
1 . 12
1 . 0
2
56 . 0
2
= + = + = +
(
(

|
.
|

\
|
=
m r
m
gv C
v
a
The value of b, c, and d remain unchanged at 6695.2, 17.6, and 5160.1, respectively.
( ) ( )
( ) c f a e k
d cM k f b aM k e M
M
s s
gload s aload s f
veh
+
+ + + +
=
1
3
10 13 . 4
1625
1
500 9 . 0
1
960
1
94 . 0
1 1 1 1 1

=
|
|
.
|

\
|
+

+ =
|
|
.
|

\
|
+ + =
tran bat mot mot trans
s
S S S
e

130
3
10 43 . 2
500 9 . 0
1
1225 9 . 0
1
325 95 . 0 9 . 0
1
94 . 0
1 1 1 1 1

= |
.
|

\
|

+

=
|
|
.
|

\
|
+ =
bat mot gen mot pu gen mot trans
s
S S S
f

kg 344 , 1
) 6 . 17 10 43 . 2 6 . 35 10 13 . 4 ( 1 . 1 1
) 1 . 5160 499 6 . 17 ( 1 . 1 10 43 . 2 ) 2 . 6695 136 6 . 35 ( 1 . 1 10 13 . 4 974
3 3
3 3
=
+
+ + + +
=


Mveh
kW 3 . 65 W 331 , 65 ] 2 . 6695 ) 136 1344 ( 6 . 35 [ 1 . 1 = = + + = a P
kW 3 . 41 262 , 41 ] 1 . 5160 ) 499 1344 ( 6 . 17 [ 1 . 1 = = + + = W Pg
kW 3 . 51 95 . 0 9 . 0 94 . 0 3 . 41 = = pu P
kW 8 . 48 9 . 0 94 . 0 3 . 41 = = gen P
kW 5 . 69 )] 675 . 0 94 . 0 ( 3 . 41 , 4 9 . 0 3 . 65 [ = = Max Pmot
The above motor power shows its peak power rating based on the higher of the two power needs:
(1) for acceleration and (2) for grade climbing. We used the motors peak power rating for the
acceleration power requirement, which is to be met for 12 seconds. Because the motor is
required to provide constant grade climbing power for 20 minutes or more, we used the motors
constant power rating for grade climbing. A ratio of 0.675 between constant and peak specific
power, derived from the runs of Advanced Vehicle Simulator (ADVISOR) for the Unique
Mobility motor SR218H, is used here:
kW 5 . 28 9 . 0 94 . 0 ) 3 . 41 3 . 65 ( = = bat P
kg 158 325 51300 = = pu M
kg 8 . 39 1225 48800 = = gen M
kg 4 . 72 960 69500 = = mot M
kg 57 500 28500 = = bat M
kg 8 . 42 ) 94 . 0 1625 ( 65300 = = tran M
Series HEV with All-Electric Acceleration Capability
This series HEV will accelerate, all-electrically, from zero to 60 mph in 16 seconds. The
battery, which has the characteristics in Table 4, is sized to provide the power for this
acceleration. The motor is sized to provide constant power for grade climbing because its
constant power rating is assumed to be only 67.5 of the peak power rating.
131
W/kg 1 . 24 82 . 26 8 . 9 008 . 0
16
1 . 16
1 . 0
82 . 26
1
1 . 0
1 . 0
'
49 . 0
2 2
= + |
.
|

\
|
= +
(
(

|
.
|

\
| +
=
m r
x
m
m
m
gv C
t
t v
a
2 . 6695 ' = = b b
The values of a, b, c, and d remain unchanged at 35.6, 6695.2, 17.6, and 5158.4, respectively.
|
.
|

\
|
+ +
+ + + + + +
=
'
'
'
'
1
) ' ' (
'
) (
'
) (
'
a
h
c f a
e
k
b M a k
h
d cM k f b aM k
e
M
M
s
s
s
aload
s
gload
s
aload
s
f
veh
3
10 76 . 1
1625
1
960
1
94 . 0
1 1 1 1
'

= |
.
|

\
|
+ = |
.
|

\
|
+ =
tran mot tran
s
S S
e

3
10 79 . 4
1225 9 . 0
1
325 95 . 0 9 . 0
1
94 . 0
1 1 1 1
'

=
|
|
.
|

\
|

+

=
|
|
.
|

\
|
+ =
gen mot pu gen mot tran
s
S S
f

3
10 53 . 3
335 9 . 0 94 . 0
1 1
'

=

= =
bat mot tran
s
S
h

) 1 . 24 10 53 . 3 6 . 17 10 79 . 4 6 . 35 10 76 . 1 ( 1 . 1 1
) 2 . 6695 136 1 . 24 ( 1 . 1 10 53 . 3 ) 4 . 5158 499 6 . 17 ( 1 . 1 10 79 . 4 ) 2 . 6695 136 6 . 35 ( 1 . 1 10 76 . 1 974
3 3 3
3 3 3
+ +
+ + + + + +
=


x
M
veh
=1,488 kg
kW 71 W 70982 ] 2 . 6695 ) 136 1488 ( 6 . 35 [ 1 . 1 = = + + =
a
P
kW 44 W 041 , 44 ] 4 . 5158 ) 499 1488 ( 6 . 17 [ 1 . 1 = = + + =
g
P
kW 4 . 50 W 414 , 50 ] 2 . 6695 ) 136 1488 ( 1 . 24 [ 1 . 1 = = + + =
e
P
kW 8 . 54 ) 9 . 0 95 . 0 94 . 0 ( 44 = = =
mot gen tran g pu
P P
kW 5 . 75 94 . 0 71 ) , ( = = =
tran g tran a mot
C P P Max P
(C = constant to peak power ratio = 0.675)
kW 1 . 52 ) 9 . 0 94 . 0 ( 44 = = =
mot tran g gen
P P
kW 6 . 59 ) 9 . 0 94 . 0 ( 4 . 50 = = =
mot tran e bat
P P
kg 6 . 168 325 54800 = =
pu
M
132
kg 7 . 78 960 75500 = =
mot
M
kg 5 . 42 1225 52100 = =
gen
M
kg 9 . 177 335 59600 = =
bat
M
kg 5 . 46 ) 1625 94 . 0 ( 71000 = =
tran
M
This grid-connected HEV has a combined power of 111 kW from its generator and battery that
can be delivered to its wheels. However, its motor has a peak power of only 75.5 kW. Its Z60
time on combined power unit and battery power is limited by the size of its motor. Because the
motor is sized to meet a 12-second Z60 time, the HEV cannot accelerate any faster.
References
AISI, 1998, UltraLight Steel Auto Body Final Report, American Iron and Steel Institute,
Washington, D.C.
Duleep, K.G., 1998, Briefing on Technology and Cost of Toyota Prius, prepared for the Office of
Transportation Technology, U.S. Department of Energy, Washington, D.C., April.
OTA, 1995, Advanced Automotive Technology: Visions of a Super-Efficient Family Car, OTA
Report OTA-ETT-638, Office of Technology Assessment, U.S. Congress, Washington, D.C.
Stodolsky, F., A. Vyas, and R. Cuenca, 1995a, Lightweight Materials in the Light-Duty Passenger
Vehicle Market: Their Market Penetration Potential and Impacts, presented at the Second World
Car Conference, University of California at Riverside, Riverside, Calif., March.
Stodolsky, F., A. Vyas, R. Cuenca, and L. Gaines, 1995b, Life-Cycle Energy Savings Potential from
Aluminum-Intensive Vehicles, in Proceedings of the 1995 Total Life Cycle Conference (P-293),
SAE International, Warrendale, Pa., Oct.
Vyas, A.D., H.K. Ng, D.J. Santini, and J.L. Anderson, 1997, Electric and Hybrid Electric Vehicles:
A Technology Assessment Based on a Two-Stage Delphi Study, Argonne National Laboratory
Report ANL/ESD-36, Argonne, Ill., Dec.
Toyota, 1997, Toyota Hybrid System, Toyota Press Information 97, Tokyo, Japan.