PROSPECTS FOR ELECTRIC VEHICLES
A STUDY OF LOW-POLLUTION POTENTIAL VEHICLES - ELECTRIC
Department of Health, Education and Welfare / National Center for Air Pollution Control
HJUttkJnc.
-------
PROSPECTS FOR ELECTRIC VEHICLES
A STUDY OF LOW-POLLUTION-POTENTIAL VEHICLES - ELECTRIC
by
J. H. B. George
L. J. Stratton
R. G. Acton
for
DEPARTMENT OF HEALTH, EDUCATION AND WELFARE
NATIONAL CENTER FOR AIR POLLUTION CONTROL
Contract No. PH 86-67-108
May 15, 1968
C-69260
Arthur D. Little, Inc.
15 Acorn Park
Cambridge, Massachusetts
-------
TABLE OF CONTENTS
Page
List of Tables v
List of Figures vii
Acknowledgement ix
I. SUMMARY 1
A. PURPOSE AND SCOPE 1
B. FINDINGS 1
C. RECOMMENDATIONS 4
II. APPROACH 7
III. CRITERIA FOR ELECTRIC VEHICLES 11
A. DEVELOPMENT OF VEHICLE SPECIFICATIONS 11
B. POWER AND ENERGY CALCULATIONS 14
C. ELECTRIC TRANSMISSION PARAMETERS 19
IV. ELECTRIC TRANSMISSION SYSTEM 26
A. MOTORS 26
B. CONTROLS 33
C. SPEED REDUCERS 38
D. COOLING REQUIREMENTS 39
V. ELECTROCHEMICAL POWER SOURCES 40
A. CONVENTIONAL BATTERIES 40
B. BATTERIES UNDER DEVELOPMENT 43
in
Arthur ZD.lUttlc.llnr.
-------
TABLE OF CONTENTS (Continued)
Page
V. ELECTROCHEMICAL POWER SOURCES (Continued)
C. FUEL CELLS
D. ELECTROCHEMICAL HYBRIDS
VI. MECHANICAL-ELECTRICAL HYBRIDS
A. TECHNICAL PROSPECTS
B. COST FACTORS
C. REFUELING
D. SUMMATION AND RECOMMENDATIONS
53
56
58
VII. PROSPECTS AND REQUIREMENTS FOR ELECTRIC VEHICLES 67
67
72
77
78
APPENDIX A NOTES ON ELECTRIC VEHICLE DEVELOPMENT
IN THE UNITED STATES
APPENDIX B NOTES ON FOREIGN ACTIVITY
83
89
IV
3rthur 3l.lUttk.Ilnr.
-------
LIST OF TABLES
Table No. Page
1 Electric Vehicle Specifications 12
2 Public Health Service Cycle 20
3 City Bus Cycle 20
4 Electric Transmission Parameters 25
5 Hybrid Vehicle Specifications 65
6 Generalized Technical Parameters For Electro-
chemical Power Sources 68
B-1 Specifications of Some British Experimental
Electric Vehicles 92
Arthur ZD.3Uttle.3lnr.
-------
LIST OF FIGURES
Figure No. Page
1 Standardized Acceleration Curves 13
2 Acceleration Versus Velocity For Selected Vehicles 16
3 Power Requirements For The Family Car At
Maximum Acceleration 18
4 Power-Speed Characteristic Curves For Electric
Motor 21
5 Motor Selection For Family Car 23
6 Allowable Rotor Speeds 27
7 Weights of DC Motors 30
8 Weight Versus Power of AC Motors 32
9 Weight of Control Unit 36
10 Various Electrical Systems for Electrical-Mechanical
Hybrids 59
11 Power/Weight Ratios For Internal Combustion
Engines 61
12 Approximate Cost Levels For Internal Combustion
Engines 62
13 Weight of Generator, Rectifier, Speed Increaser
and Voltage Regulator 64
VII
-------
LIST OF FIGURES (Continued)
Figure No. Page
B-1 British Electrically Driven Milk Truck 90
B-2 The Scamp: A Developmental Vehicle 90
B—3 Prototype Electric Car Built in Italy By Subsidiaries
of Rowan Controller Corporation 95
VIII
3rthur ZD.HittleJnr.
-------
ACKNOWLEDGEMENT
The assistance of the following organizations in providing material for this study is
gratefully acknowledged:
Allis-Chalmers Manufacturing Company
Allmanna Svenska Elektriska Aktiebolaget (ASEA)
Argonne National Laboratory
Atomics International, a division of North American Rockwell Corporation
Brown Boveri and Company Ltd.
Robert Bosch GmbH
Chloride Electrical Storage Co. Ltd.
Compagnie Fran9aise Thomson-Houston
Eagle-Picher Company
Electric Fuel Propulsion, Incorporated
ESB Incorporated
Esso Research and Engineering Company
Ford Motor Company
General Electric Company
General Motors Corporation
Gould-National Batteries, Incorporated
Gulf General Atomic
Gulton Industries, Incorporated
Lead Development Association
Leesona Corporation
McDonnell-Douglas Corporation
Montecatini-Edison S.p.a.
Rowan Controller Company
Scottish Aviation Ltd.
Shell Research Ltd.
Siemens Schuckertwerke A.G.
Societe Le Carbone Lorraine
Standard Oil Company (Ohio)
Texas Instruments Incorporated
TRW, Incorporated
UCLA (Professor G. A. Hoffman)
Union Carbide Corporation
United Kingdom Electricity Council
U. S. Army Engineer R&D Laboratories
W&E Vehicles
Westinghouse Electric Corporation
Yardney Electric Corporation
IX
Arthur 2D.llittlc.3lnr.
-------
I. SUMMARY
A. PURPOSE AND SCOPE
This report presents the results of a study conducted by Arthur D. Little, Inc. for
the National Center for Air Pollution Control concerning electrically powered vehicles of
low pollution potential. A parallel study concerning unconventional thermal, mechanical,
and nuclear power sources for vehicles has recently been completed by the Battelle
Memorial Institute - Columbus Laboratories.*
The major purpose of the present study is to assess the state of the art in the
technologies on which the future development of electric highway vehicles depends. The
assessment is made on the basis that there will be no major change in the patterns of urban
transportation and focuses upon the technical requirements of the vehicle. As a framework
for the study, technical and, to a lesser extent, cost criteria were established for six classes
of electric vehicles, most of which correspond closely in performance to existing
conventionally powered types. On the basis of these criteria the prospects for vehicle
application of various electric motor and control systems and a wide range of electrical
power sources including batteries, fuel cells, and engine-generator-battery hybrids are
compared. The study identifies the most promising systems for further development and
estimates the magnitude of the shortfall in their present characteristics. Recommendations
are made for the expansion of future effort in these areas, taking into account the expected
contributions from existing programs in the public and private sectors of the economy.
B. FINDINGS
(1) The technical state of the art in motors and controls is generally more advanced
than that of power sources in relation to the needs of electric highway vehicles.
(2) Minimum energy and power densities for electrochemical power sources capable
of giving acceptable performance in the six classes of vehicles are listed below. The figures
are based on reasonable estimates of the weight assignable to the power source and
propulsion system and of the weight of suitable electric motors and controls.
Family Commuter Utility Delivery City City
Car Car Car Van Taxi Bus
Conventional construction
Energy density (w-hr/lb) 135
Power density (w/lb) 94
Lightweight construction
Energy density (w-hr/lb) 87
Power density (w/lb) 60
41
46
28
31
26
40
18
28
50
55
33
36
96
45
64
30
81
36
55
25
* Battelle Memorial Institute, "Study of Unconventional Thermal, Mechanical, and Nuclear Low-Pollution-
Potential Power Sources for Urban Vehicles," by J. A. Hoess, etal.. Summary Report to U.S. Dept. of
Health, Education and Welfare, March 15, 1968.
ZD.ILttttcJnr.
-------
(3) In a hybrid power source the important technical parameter for the battery is
the power density; energy density is of only secondary importance. The power densities
required for batteries in vehicles of conventional construction having 40% of the peak power
supplied by a gas piston engine and 60% by a battery are as follows:
Family Commuter Utility Delivery City City
Car Car Car Van Taxi Bus
Power density (w/lb) 92 38 34 43 35 25
For the family car, the power density requirement is virtually equal to that for a purely
battery powered vehicle; for the other vehicles the requirement is not much lower.
(4) The technical requirements for the family car power source can be met only by
high temperature alkali-metal batteries; these systems, as yet in the early stages of
development, are also the most promising for the commuter car, the delivery van, and the
city taxi. If lightweight construction is used metal-air batteries are potentially capable of
meeting the requirements for all vehicles except the family car. Alkali-metal batteries with
organic electrolytes seem likely to have insufficient power density for use in vehicles, and
combining them with high-power-density batteries is clearly too costly a solution.
(5) Although of limited energy density the lead-acid battery seems likely to be
capable of meeting the requirements of the utility car of lightweight construction.
(6) The fuel cell's prospects as an electric vehicle power source seem unfavorable,
mainly because of its extra complexity and cost and its lower power density. Its much
simpler mode of refueling, however, would be a major advantage.
(7) The most practical present choice of motor for an electrical vehicle is a
high-speed dc machine with mechanical commutator, operating at speeds of up to 19,000
rpm for the smaller vehicles, air cooled, and with integral speed reducers. Such motors are
already being developed for aerospace and military applications.
Control of high-speed dc motors is best achieved by a chopper circuit containing
silicon controlled rectifiers (SCR's). These circuits are at present rather heavy for the
amount of power they control. However, they can be developed for operation at high
voltage (500 volts in the case of the bus and family car) so as to be relatively lightweight.
(8) Because ac motors can be run at higher maximum speeds, they have a higher
output per pound than dc motors and hence permit weight reduction; these will doubtless
be adopted when the technologies for inversion and control are further advanced. The
development of solid-rotor ac motors instead of the squirrel-cage induction type will lead to
greater weight reduction and simplify the problems of inversion and control.
Arthur 21.1Uttlr,llttr.
-------
The controller for the ac motor would be a variable-frequency, variable-voltage SCR
inverter circuit. Experimental circuits developed to date are, however, extremely heavy and
require capacitor banks across the battery when an induction motor is used.
(9) Regenerative braking would reduce the power source requirements somewhat
for vehicles that make frequent stops. However, for the majority of vehicles, the amounts of
energy recovered do not appear likely to justify the added cost and complexity.
(10) The recharging of electric vehicles is best achieved at slow rates. Rapid recharge
would call for large investment in facilities and would introduce a number of severe
technical difficulties. Occasional long trips could be accomplished by replacement of
batteries at service stations.
(11) The first costs of electric vehicles will inevitably be higher than those of
equivalent internal-combustion-powered vehicles. Part of this may be offset by lower
maintenance requirements and longer life, but much of the difference would have to be
regarded as a social cost for the elimination of air pollution. If an additional SI,000 is
acceptable in the retail cost of the family car, the cost target for the battery is in the vicinity
of 95 cents per pound. Allowable costs in the commercial vehicles would be somewhat
higher. High-temperature alkali metal batteries have some prospects of meeting these cost
requirements. Metal-air batteries are likely to be somewhat costlier but might still be
acceptable for commercial vehicles.
(12) The duplication in functional capabilities tends to give hybrid vehicles a
fundamentally higher cost structure than pure electrics. The higher cost may, however, be
acceptable in commercial vehicles, and existing technology would permit low cost lead-acid
batteries to meet the power density requirements of a hybrid city bus, making this vehicle
an attractive possibility.
(13) Present federal spending on research and development on electrochemical
power sources, and exclusive of the major fuel cell hardware programs for space vehicles, is
estimated to be in the vicinity of $10 million annually. However, only a small proportion of
this is directly relevant to the needs of electric vehicle development for civilian use.
Expenditure by private industry on power source development relating to electric vehicles is
estimated at $5-7 million annually.
(14) The time scale for the development of electric vehicles will be determined
mainly by progress in battery technology and by the magnitude of the effort which sustains
it. Assuming that the latter is adequate, high-temperature alkali metal batteries appear
likely to require a period of the order of ten years before they can be considered as practical
and economic power sources for use in vehicles. Certain metal-air systems, having promise
for application in commercial vehicles, might be made available in a somewhat shorter
period. At least comparable periods of time would be needed to develop sufficiently low
cost manufacturing processes for the motors and controls.
Arthur ZD.Itttlc.llnr.
-------
C. RECOMMENDATIONS
(1) The power source requirement for electric vehicles calls for a major and
sustained effort in the development of batteries with high energy density and high power
density. The requirements for the most significant classes of vehicles can be fully met only
by batteries having alkali-metal anodes and molten salt (or possibly ceramic) electrolytes,
which necessitate high temperature operation. If such batteries are to be brought to
maturity within a ten-year period a coordinated program under federal sponsorship should
be set up to do the following:
(a) Explore comprehensively the electrochemistry of the limited number
of elements having the reactivity and potential low cost availability to
function as anode or cathode materials in high-temperature alkali metal
batteries. The physical chemistry of molten salt and ceramic systems
appropriate for use as electrolytes should also be investigated.
(b) Examine the materials selection problems associated with such bat-
teries, giving particular regard to choice of current collectors, separa-
tors, sealants, and structural materials generally.
(c) Investigate suitable low-cost engineering solutions for the problems of
thermal insulation, temperature control, and safety, which are com-
mon to all members of this class of batteries.
(d) Develop, in collaboration with industry, the one or two most practical
battery systems arising from these research programs.
(2) Support should be given to the further development of rechargeable metal-air
batteries. For those of relatively conventional design the principal problems rest with the
basic electrochemistry. Other types, which employ engineering solutions to the fundamental
electrochemical difficulties, should be supported in the development stage.
(3) The development of high rate lead-acid batteries should be supported in view of
their prospects in hybrid-powered commercial vehicles.
(4) The problems of adapting the technologies of high speed air-cooled dc commutator
electric motors and of sophisticated dc chopper controllers for low cost mass production
should be investigated. Studies should be initiated as to the techniques - such as extensive
automation and use of microelectronic, modules — by which cost reductions might best be
achieved.
(5) Possible techniques for the low cost manufacture of solid-rotor synchronous ac
motors and of squirrel-cage ac induction motors with inverters should be investigated. The
technology of these machines (known also as brushless dc motors) would also benefit
directly from the advances in dc chopper controllers and high-speed dc motors.
3rthxtr ZD.ILittlc.Knr.
-------
(6) Some general encouragement should be given to developing the use of lightweight
materials, particularly plastics, reinforced and otherwise, and the lighter metals, for use in
present day conventional automobiles. Advances in this area will help to hasten the advent
of electrically powered vehicles by lessening the severity of the requirements on the power
source. Similar considerations hold for tires designed to minimize rolling resistance.
(7) The potential merit of electric vehicles, relative to other approaches toward
alleviation of the air pollution problem, should be periodically re-evaluated in the light of
the progress and pattern of development taking place in the various contributing
technologies.
Arthur ZD.lUttlc.Knr.
-------
II. APPROACH
The pollution of the atmosphere by the exhaust products of the internal combustion
engine, particularly in the cities, is a subject of increasingly serious concern. Viewed as a
purely technical problem, abatement can be approached in several ways. The most
straightforward is to modify the engine and exhaust system with devices which will either
suppress pollutants or convert them into less noxious products. A great deal of effort is
presently being expended in this approach, and it may well prove to be the most satisfactory
for the immediate future. However, the tendency of an internal combustion engine to
pollute increases with its age, and the effectiveness of conversion devices, usually involving
catalytic combustion, decreases with time; thus the suppression of pollution may be
adequate in a new vehicle but quite unsatisfactory by the time it is two or three years old.
Another approach is to replace the internal combustion engine with a type of prime
mover that has no significant pollution problem. In principle there are a large number of
possibilities, including electric power sources, external combustion systems, inertial energy
systems and a variety of exotic energy conversion devices. In practice, however, the practical
options are quite limited.
A third approach is to alter the transportation system through major changes in
conventional public facilities or through such concepts as the rental of small, self-driven.
taxi-like vehicles for journeys within the urban centers.
The subject matter of this study is concerned with the second general approach and
specifically with the prospects for self-propelled electric vehicles.
It is necessary to make numerous assumptions as the basis for a study of this type,
many of them rather arbitrary, and many of them having a significant bearing on the
findings of the study. We are well aware of this and consider our work as a first
approximation which may suggest directions for further more refined and specific analysis.
The study is based upon a consideration of the requirements for six types of electric
vehicles: a family car, a commuter car. a utility car, a delivery van, a city taxi and a city bus.
With the exception of the utility car these represent the great majority of vehicles that now
provide transportation in the urban environment. The model for the family car is the
familiar full-sized, six-passenger automobile, which represents almost SO'/r of the total
vehicle population of this country. The commuter car is modeled on the much smaller,
four-seater, European-type compact car, whose population has been steadily increasing in
recent years, primarily for commuter travel. The delivery van is somewhat less well defined
but can be regarded as a typical panel truck with a carrying capacity of about one ton. The
city bus is the characteristic full-sized vehicle with a carrying capacity of about 80
passengers including standees. The other two vehicles contain an element of invention. The
city taxi is outwardly similar to the family car but has been arbitrarily assigned lower
performance capabilities in view of its working pattern in the dense traffic of city streets.
The utility car is a small vehicle intended purely for local trips, such as might be taken by a
ai.ltittlc.llnr.
-------
housewife on shopping errands or possibly for short-range commuting. While each of these
vehicles is a candidate for electric propulsion in its own right, it is clear that the greatest
interest attaches to the family car in view of its preponderance in the total vehicle
population.
The first step in the selection of weight and performance parameters for the various
vehicles involved an examination of the published data for virtually all vehicle models
manufactured over the last ten years in the United States and Western Europe. From these
vehicle specifications, reasonably appropriate values were established for weight and
performance in each of the classes. The calculation of power requirements necessitated some
assumptions about frontal cross-sectional area, degree of streamlining, and tire friction. The
values chosen for the drag coefficient were liberally low, reflecting the best present
aerodynamic design practice, while the estimates for tire friction were conservative, since
there is some element of compromise with safety in the design of low-friction-loss tires.
The parameters for acceleration are believed to represent reasonable values for new
vehicles. It would not be fair, however, to insist that the electric vehicle should have an
acceleration equal to that of a typical new, carefully tuned vehicle with a conventional
engine. The performance of an electric vehicle should remain fairly constant with time,
while that of most conventional vehicles is at least fractionally lower than when new after
three or four years on the road. Accordingly, adverse weighting of the vehicles that formed
the basis for the assumed parameters was at least partially offset by considering only
vehicles whose power capabilities were on the lower side of normal. Thus, the acceleration
for the family car is typical of that for a six-cylinder vehicle, even though more than 80% of
the vehicles in this category now have eight-cylinder engines.
Probably of even greater influence on the findings of the study than the assumptions
about performance are those about driving cycles and their relationship to vehicle range.
The energy consumed by a vehicle per mile of driving is very heavily influenced by the
number of stops made, by the rate of acceleration, and by the velocity. The selection of any
particular driving pattern thus has a major effect on the range obtainable from a given
quantity of stored energy. Fortunately, the arbitrariness of the assumption is partially offset
by the fact that vehicles are generally driven at relatively high speed when not in stop-and-go
traffic; the energy consumption per mile is therefore very much more constant than it
would be if people were in the habit of driving at modest speed on the open road.
The technical requirements for the six classes of vehicles have been established by
the following procedure. The maximum power requirements corresponding to the maximum
acceleration performance assumed for the vehicles were first calculated; this provided a basis
for setting the power rating of each vehicle. The requirements for stored energy were
calculated from the energy expenditure for each vehicle when following a standard urban
driving cycle. (Since the driving pattern for the city bus involves an unusual number of stops
per mile, a separate driving cycle was developed and used for this vehicle.) Assumptions
about total range varied with the type of vehicle; in general, the criterion was to take a
relatively low value but one that would be justifiable in the application for which the vehicle
was intended. For example, the family car has a range of 200 miles, the commuter car 100
miles, and the utility car only 50 miles.
ZD.1tittlc.3lnr.
-------
To bring into focus the technical requirements for electric vehicles, our approach has
been to remove conceptually from the conventionally powered vehicles those components
(engine, transmission, radiator, muffler, etc.) which are inseparably associated with the
functioning of the internal combustion engine and to replace them with an electrical power
source, motors, and a control system of equal total weight. This procedure considers electric
vehicles as modifications and neglects the possibilities of weight optimization which would
result from special design. Our basis has been some prior work by Hoffman,* who analyzed
the weights of the components of a great many vehicle models and found a recurring pattern
in the weight distribution. Among passenger cars whose size and power varied over a wide
range, for example, the weight of the replaceable components was never far from 35% of the
curb weight of the vehicle. Hoffman also considered the proportion of vehicle weight which
could be assigned to the power source and propulsion unit if lightweight materials such as
plastic body panels and magnesium frames were used. He concluded that this would raise
the allowable proportion to 50%. Our analysis has been conducted on the bases of both
conventional and lightweight construction, using proportions close to Hoffman's for the cars
and estimated values for the delivery van and city bus.
A major element in our further analysis is the assumption of representative weights
for the motors and controls used in these various vehicles so that a total weight could be
allocated to the power source. The pattern of the study is thus to relate the prospects for
the electric vehicle largely to the requirements of its power source. The justification for this
assumption is that the technology in the electrical and electronic areas is relatively further
advanced toward the vehicle objective than is that of electrochemical power sources.
Projections in these areas can thus be made with relatively greater confidence. The derived
results of the above calculations are then expressed as requirements for energy density
(watt-hours per pound) and power density (watts per pound) for the power source.
Power source volume is also of some significance, but to a first approximation it
would seem secondary to considerations of weight. The great majority of the possible power
sources which are candidates for vehicle application have densities which are broadly
comparable with each other and with the power systems they would replace. Moreover, the
electrochemical power sources would seem to have greater flexibility for utilizing internal
space within the vehicle.
The question of mechanical-electrical hybrid power sources has also been considered
as part of this study, although practical power systems of this type would inevitably
produce some degree of air pollution. No attempt has been made to consider the relative
merits of various hybrid approaches on the basis of their relative contribution to air
pollution; they have been regarded solely as alternative electrical power sources. The
approach was to assign the available power source weight to the engine-generator-battery
combination and from the estimated weights for the first two components determine the
power densities required from the battery.
Much of this report is concerned with a review of the present and projected state of
the art in the technologies of electric motors and control systems and electrochemical power
sources. Using the technical requirements of the electric vehicles as a yardstick for
assessment of present achievement and future progress, we have developed some conclusions
and recommendations concerning the preferred direction for further effort.
* G. A. Hoffman, Transportation Research, Vol. 1, No. 3, 1967.
9
2l.llittle.3lnr.
-------
The question of costs is of comparable importance but is even more difficult to
bring into focus, since it frequently relates to manufacturing processes which in many cases
are barely in the conceptual stage. Such conclusions about costs as can be drawn are,
however, integrated into our assessment of the needs for further research and development
programs.
The conclusions to be drawn from our study, while cautiously optimistic for the
prospects of electric vehicles, are not to be interpreted as indicating the likelihood of their
imminent appearance. A vast amount of further technical effort is called for if the
large-scale use of electric vehicles is to become a reality, even twenty years from now.
Nevertheless it seems that it can be done, and the effort should be made if the social cost of
air pollution is sufficiently great and other approaches to the problem are less attractive.
Such decisions, however, are in the political realm; this report is to be considered primarily
as a contribution to the technical evaluation of the problem.
10
9rthur ZD.lUttle.llnr.
-------
III. CRITERIA FOR ELECTRIC VEHICLES
A. DEVELOPMENT OF VEHICLE SPECIFICATIONS
In this section technical criteria are established for the six classes of electrical
vehicles. As a basis for this, the electric vehicles have been considered as comparable
alternatives to existing vehicles powered by internal combustion engines. Though new
requirements for vehicles do appear from time to time, one year's car generally behaves very
much like another's, and motorists have become accustomed to certain kinds of response
from the vehicles they drive. This is not to say that all of today's vehicles show identical
performance but, rather, that there are minimum standards of acceleration and comfort to
which any new vehicle must conform if it is to achieve commercial success. Furthermore,
any vehicle which cannot perform like most of those with which it mingles presents a
serious hazard on crowded roads.
In addition to performance, the vehicle controls must be similar in "feel" to those of
conventional vehicles. This, however, is largely a matter of proper design of the electrical
controls and does not necessarily involve weight and power consumption.
The choice of vehicles and of vehicle parameters for this study (Table 1) were made
after discussions with representatives of the National Center for Air Pollution Control and
Battelle Memorial Institute. Most of these vehicles represent types of conventional motor
vehicles now used in and around our cities. For completeness, we have included in the study
a "utility" vehicle, whose performance is not comparable with that of today's highway
vehicles but which may find acceptance in specialized uses and locations. Its specifications
correspond closely to those of many of the electric vehicles developed in the United
Kingdom.
The reported performances of a number of selected types of internal combustion
powered vehicles in each of the six classes were examined and norms were established for
their maximum acceleration; these figures are tabulated on lines 1 and 2. Using the actual
acceleration curves as a guide we then drew standardized acceleration curves and normalized
them to the acceleration specifications. These curves, shown in Figure 1, were used as the
basis for all further calculations of maximum power and energy requirements.
The maximum performance of a vehicle can be concisely stated as the relationship
between velocity and time while it is driven from rest at full power. It is not necessary to
consider grade climbing separately, because at any given speed there is a simple relationship
between maximum level-road acceleration) and grade climbing ability.
The weights assigned to the vehicles listed in Table 1 were based on data for current
vehicles of similar capacity and performance. The weights assignable to propulsion, energy
storage, controls, etc. in line 7a were obtained by use of the general procedure of Hoffman*
- i.e., by totaling the weights for all those components of the conventionally powered
op cit.
11
ZD.lUttlc.3Jnr.
-------
TABLE 1
ELECTRIC VEHICLE SPECIFICATIONS
Assumptions
1. Acceleration to
2. in
3. Range
4. Seats or payload
5. Loaded weight*
6. Curb weight
7. Weight assignable to propulsion,
energy storage, controls
a. conventional construction
b. lightweight construction
8. Frontal area
9. Drag coefficient
10. Elec. transmission efficiency
(mph)
(sec)
(mi)
(Ib)
(Ib)
(Ib)
I
(Ib)
(ft2)
(%)
Family
Car
60
15
200
6
4,000
3,500
1,250
1,750
25
0.35
82
Commuter
Car
60
30
100
4
2,500
2,100
750
1,050
18
0.25
77
Utility
Car
30
10
50
2
1,700
1,400
500
700
18
0.25
72
Delivery
Van
40
20
60
2,500
7,000
4,500
1,400
2,000
42
0.85
79
City
Taxi
40
15
150
6
4,000
3,500
1,250
1,750
25
0.35
76
City
Bus
30
15
120
10,000
30,000
20,000
5,000
7,000
80
0.85
85
70
94
85
100
100
122
22
30
29
80
20
26
12
16
17
65
8
11
49
66
62
56
45
57
36
48
47
77
75
99
135
180
159
55
300
353
Derived Parameters
11. Max. power delivered by motors
(kw)
-------
70
60
50
£ 40
e
4-*
o 30
20
10
Circles Represent
Normalization
Points
10
15 20
Time (sec)
25
30
100
90
80
70
60 T
I
50 £
_o
v
30
20
10
35
FIGURE 1 STANDARDIZED ACCELERATION CURVES
13
-------
vehicles which would not be required in the same form in an electric vehicle. The assignable
weight for vehicles with conventional construction is shown on line 7a and for lightweight
construction on line 7b.
The values chosen for the range of the electric vehicle, line 3, are on the low side of
average for internal combustion powered vehicles in the case of the family car, city taxi and
city bus. For the commuter car, utility car, and delivery van, we chose values which might
represent reasonable daily ranges for vehicles which are not likely to be used continuously
for long periods.
B. POWER AND ENERGY CALCULATIONS
Two sets of calculations are needed to define the power and energy requirements of
a vehicle. First, we must know the maximum power and the velocity at which this maximum
occurs. Second, we must calculate the total energy consumed when the vehicle operated
under various driving conditions.
1. Power Requirements
The power expended by the vehicle can be represented by the following expression:
instantaneous power = velocity [mass x acceleration
+ aerodynamic drag + rolling resistance]
The maximum power requirement is determined by expressing all the quantities on the right
hand side in terms of velocity as a variable. The expression is then differentiated and the
resulting function put equal to zero.
a. Acceleration
The first step was to express the maximum linear acceleration in terms of velocity.
This was done by synthesizing simple analytical expressions which fitted the graphical data
reasonably well within the normalization limits of the performance curves. These
expressions,* which are shown graphically in Figure 2, are as follows:
Family Car: a = 20 0.500 v + 5.3 x l(T3v2 - 2.2 x l(Tsv3 - 8(1 + lO'V) "'
* The symbols and units used in the analytical expressions are as follows:
a acceleration (ft sec"2) M weight of vehicle (Ib)
A frontal area (ft2) S distance travelled (ft)
CQ drag coefficient (dimensionless) t time (sec)
g gravitational constant (32.2 ft sec"2) v velocity (ft sec"2)
14
Arthur ZH.1Uttlr.3lnr.
-------
Commuter Car: a = 7.0 - 0.028v - 1.3 x 10'3 v2 + 0.9 x 10's v3
Utility Car and a = 6.6 - 0.046v - 1.3 x 10'V + 0.7 x 10-sv3
Delivery Van:
City Taxi: a = 9.0 - 0.053v - 3.0 x 10"3v2 + 3.0 x 10'5v3
City Bus: a = 5.2 + 0.006v - 5.3 x 10'3v2 + 7.0 x 10'5v3
From these expressions, the power expended on linear acceleration during periods of
maximum acceleration can be expressed as a function of v, namely
Mav
g
If the power required for rotational acceleration is assumed to be 10% of that for linear, the
total power used for acceleration is
1.1 Mav
b. Aerodynamic Drag
Estimates given in the literature for power consumption caused by aerodynamic drag
vary considerably. For example, using a 2000-lb vehicle traveling at 60 mph, Hoffman*
finds 9.7 hp and the Morse panel** finds 17 hp. Basically, the reason for these discrepancies
is varying judgments as to the practicability of employing good aerodynamic body styling.
We take an optimistic view and believe that the arguments of Tenniswood et alJ are sound.
These authors reason that the incentive to reduce air resistance in gasoline-propelled vehicles
is comparatively weak, because no initial cost saving is achieved; in electric vehicles, how-
ever, particularly those designed for high speed, it can affect first cost by reducing the
battery capacity required for a given range. We believe that, compared with the problems of
developing satisfactory power sources, improving the aerodynamic characteristics of vehicle
bodies would be relatively easy. We have therefore employed optimistic values for the drag
coefficients, CQ, in the following expression:
Drag force = 1.19 x 10~3 CDAv2 pounds
The values assumed for frontal area, A (Table 1, line 8) reflect a moderate degree of
streamlining.
*op cit.
**77>e Automobile and Air Pollution: A Program for Progress, Report of the Panel on Electrically Powered
Vehicles (R.S. Morse, Chairman), U.S. Dept. of Commerce, October 1967.
TD.M. Tenniswood and H.A. Graetzel, Ford Motor Co. (SAE paper no. 670177), 1967.
15
3rthur ai.1Littlc.Ilnr.
-------
9.0
8.0
7.0
6.0
8 5.0
.c
Q.
C
o
1 4.0
S
u
3.0
2.0
1.0
Utility Car and
Delivery Van
10
Family Car
Commuter Car
I
13.0
12.0
11.0
10.0
9.0
8.0 r^
I
%
7.0
o
'
6.0 "5
u
u
5.0
4.0
3.0
2.0
1.0
20
30 40
Velocity (mph)
50
60
70
FIGURE 2 ACCELERATION VERSUS VELOCITY FOR SELECTED VEHICLES
16
-------
c. Rolling Resistance
When a vehicle moves at uniform velocity over a smooth road, that portion of
resistance which is not due to aerodynamic drag results principally from the work done in
distorting the tires, together with a small addition from wheel bearing and gearbox friction.
Some authorities, such as the Morse panel, treat rolling resistance as a constant force
dependent upon tire characteristics and vehicle weight, but this does not agree with the data
presented by Tenniswood et al., Hoffman, and StiehJer et al.* Combining the available data,
we have derived an empirical equation relating rolling resistance to vehicle weight and speed.
If the latest developments in low-loss tires are employed, the resistance would be less than
the values calculated from the equation, but we have taken a cautious approach for the
following reasons: 1) much of the reduction in rolling resistance of improved tires is lost if
correct inflation pressures are not maintained; 2) a soft ride may be demanded by buyers of
electric automobiles, and on rough roads this would require that energy be absorbed either
in the tires or by the suspension system; 3) federal safety requirements for tires cannot at
present be met by tires with very low rolling resistance.
Our empirical expression for the rolling resistance, which also includes the losses in
the gearbox and bearings, is:
Rolling resistance =-|1 (1 + 1.4 x l(T3v+ 1.2 x l(Tsv2) pounds.
d. Calculations of Maximum Power
The final expression for instantaneous power under conditions of maximum accel-
eration is thus:
_v_ I J.I Ma + i 19x 10-3 CDAv2 +-M- (1 + 1.4 x 10"3v+ 1.2 x l(T5v2) horsepower
J J \J I jj Jv/ J
where a is related to v by the expressions given above in subsection a and the factor 550 con-
verts ft Ib sec" ' to horsepower. This expression is plotted in Figure 3 for the family car. In
this case the total power requirement does not pass through a maximum value in the velocity
range of 0-60 mph, and the conditions for maximum power are those for maximum accelera-
tion at 60 mph. A number of the other vehicles show maxima at velocities in the 20-40 mph
range. Values for the maximum power, which is that delivered at the output of the gearbox,
are shown in line 11 of Table 1.
R. D. Stiehler, M. N. Steel, G. G. Richey, J. Mandel, and R. H. Hobbs. J. Res. Nat. Bur. Std.,
Vol. 64 0,1960.
17
ZD.IUttlr.Knr.
-------
Linear Plus
Rotational
Acceleration
Rolling Resistance
Plus Aerodynamic
Drag
10
20 30
Velocity (mph)
40
50
60
FIGURE 3 POWER REQUIREMENTS FOR THE FAMILY CAR
AT MAXIMUM ACCELERATION
18
-------
2. Energy Requirements
The energy requirements for each vehicle, corresponding to the values for the range,
listed on line 3, are heavily dependent upon the driving cycle or pattern of use involved.
Thus, for a given quantity of energy stored, the range of an electric vehicle operating
nonstop at constant velocity might be several times that when operating in city traffic with
frequent stops.
The selection of an appropriate driving cycle is somewhat arbitrary. After consulta-
tion with representatives of the National Center for Air Pollution Control, we have adopted,
for all vehicles except the city bus, the "Public Health Service Cycle" derived from a driving
cycle (LA4) based on studies in the city of Los Angeles. This cycle is set out in Table 2.
For the city bus, another cycle was devised which reproduces the statistical observation
that its average velocity is 12 mph and that it makes ten stops per mile; this cycle is shown
in Table 3.
On the basis of these driving cycles and the assumed parameters for the vehicles, we
calculated the delivered energy requirements for the various ranges. These are shown on line
14 of Table 3. The requirements for stored energy, line 15, were derived by dividing the
delivered requirement by the electric transmission efficiency.
C. ELECTRIC TRANSMISSION PARAMETERS
The weights allotted to the motors, control systems, cabling, etc. for the six classes
of electric vehicles have been estimated on the basis that high-speed dc motors with silicon
controlled rectifier chopper control circuits would be the most appropriate system. The
factors affecting this choice are considered later, in Section IV. This subsection contains
details of the specifications for the various components of the electrical transmission for
each class of vehicle.
One of the fundamental characteristics of electric motors and electromagnetic
machines in general is their ability to deliver a constant torque over their entire speed range.
By proper control, a motor can be made to deliver its rated torque or fraction thereof at any
speed from zero up to the physical limit of its rotating member. By overloading the motor,
one can produce 150% to 200% torque for short periods of time; the amount of overload
torque obtainable and the length of time the motor can deliver it without damage depend
on the details of the design of its motor particularly as it affects thermal capacity. Since
horsepower is the product of torque and speed, the maximum horsepower from a given size
motor is obtained by operating it at its highest speed. The rated maximum speed is dictated
by the mechanical stress limit of the rotor. These fundamental motor relationships are
summarized in Figure 4.
19
3rthur ZD.lUttlc.3nr.
-------
TABLE 2
PUBLIC HEALTH SERVICE CYCLE
Velocity Mode
(mph)
0
0-19
19
19-44
44-52
52
52-37
37
37-31
31-0
Mode Time
(sec)
30
8
15
12
7
9
12
16
5
13
Cumulative
(sac)
30
38
53
65
72
81
93
109
114
127
Acceleration Rate
(mph/sec)
+ 2.38
+ 2.08
+ 1.14
- 1.25
- 1.20
-2.38
TABLE 3
CITY BUS CYCLE
Velocity Mode
(mph)
0-15.7
15.7
15.7-19.0
19.0-0
Time
(sec)
6.60
13.30
1.35
7.95
Distance
(ft)
76
306
35
111
Acceleration Rate
(mph/sec)
+ 2.4
+ 2.4
-2.4
29.2
528
Average speed = 12.2 mph
20
2l.1Uttlr.ilnr.
-------
Maximum Power Attained
at Maximum Speed
O
I
Speed
FIGURE 4 POWER-SPEED CHARACTERISTIC CURVES FOR ELECTRIC MOTOR
21
-------
The selection of one or more motors for a given vehicle application involves matching
motor characteristic curves of the type shown in Figure 4 to the power demand profile for
the vehicle, such' as that for the family car in Figure 3. This matching is shown in Figure 5
where the dashed line represents the characteristics of the smallest dc motor which could
meet the power requirements entirely within its rated power capabilities. Since maximum
power demand will in practice be relatively infrequent and of short duration, one could
safely select a motor wriich would have these capabilities on 100% overload. The power-
speed relationship for the chosen motor is shown by the solid sloping line. It can be seen
that the maximum power output of which this motor is capable and which is attained at its
maximum motor speed will be significantly greater than the maximum power demand of the
vehicle. The vehicle cannot, however, use a smaller motor because of its need for substantial
power at lower speeds and cannot utilize the full capability of a larger motor because of the
limited output of the power source. A smaller motor could be used if a gear changing
mechanism or torque converter were inserted into the power train, but such devices are
undesirable because they introduce extra weight, cost, and power inefficiency.
The selection of motors is also influenced by certain general engineering consider-
ations. For example, we have chosen to use a motor for each wheel in the family car,
delivery van, city taxi, and city bus (the latter being a six-wheel vehicle); this has been done
to obtain certain benefits resulting from separate control of the power applied to the
individual wheels.* Normally, the use of several small motors rather than one large motor
would incur a weight penalty: for motors of a given speed, the power output per unit weight
increases with increasing size, since magnetic and thermal designs become more efficient.
This characteristic can be offset, however, by designing the smaller motors to operate at
higher maximum speeds.
Estimates of the weights of these high-speed dc motors have been based on dc
machines now used in aerospace and torpedo propulsion applications. It is assumed that
these motors would be cooled by forced air from a separate blower system. The weights of
the controllers have also been estimated from existing devices, some of which are being used
in lift trucks and other electric motor control applications.
To size the controllers and vehicle cabling, we have had to make assumptions about the
maximum voltage for the electrical system in each of the vehicles. The values selected were
500 volts for the family car and city bus, 250 volts for the commuter car, delivery van, and
city taxi, and 125 volts for the utility car. Allowance has also been made for the weight
associated with the electric cabling, component mounting, cooling system, and a speed
reducer for each motor-wheel system.
Estimates based on the present state of the art have also been made for the efficiency
of the individual motors and controllers. Speed-reducer efficiencies were already considered
in the calculation of energy requirements.
* A compromise of two wheels per motor was made for the commuter car and the utility car.
22
Arthur ai.llittkKnr.
-------
Power-Speed Line for
Motor Capable of
Powering Vehicle
without Overload
94
a
x:
o
a.
Power-Speed Line for
Motor Selected
Power Demand of
Vehicle at Maximum
Acceleration
Power Demand of
Vehicle at Constant
Velocity
60
Vehicle Velocity (mph)
FIGURE 5 MOTOR SELECTION FOR FAMILY CAR
23
-------
The results of these analyses are presented in Table 4 and the total weight of the
electric transmission system is entered in line 16 of Table 1. It can be seen that this weight
amounts to some 5-6% of the curb weight of all vehicles except the family car (for which it
is about 10%) and the city bus (about 3%). The effects of matching the motor to the profile
of vehicle power demand can be seen by comparing the estimates for the maximum possible
output of the motors in the family car (56 hp) with the maximum actually used (23.5 hp).
Each motor thus has a potential output of 1.0 hp per Ib but a useful output of only 0.41 hp
per Ib.
From the values for electric transmission weight presented in Table 4 and included
in Table I, line 16, it is possible to estimate allowable weights for the power source by
subtraction from the weights given on lines 7a and 7b of Table 1. The results are shown on
line 17. Combining these with the calculated power and energy demand of the vehicle
enables us to calculate the required energy densities and power densities for power sources
capable of meeting the vehicle specifications. These values are entered on lines 18 and 19 of
Table I and form the background to the review of the state of the art in electrochemical
power sources in Section V.
The approach taken in this study focuses attention on the energy and power
capabilities of the power source per unit weight rather than per unit volume. The justifi-
cation for this is that the average densities of electric motors, controllers, and electro-
chemical power sources are, with the possible exception of ambient-temperature fuel cells,
fairly similar to those of the components they would replace in an internal-combustion-
powered vehicle. Moreover, since the electrical system is interconnected by cabling rather
than mechanical links, it has greater flexibility for location within the vehicle. Thus it seems
reasonable for a first consideration to assume that w-hr/lb and w/lb can be used as yard-
sticks of the technical merit of a power source for vehicle applications.
The analysis of power and energy requirements carried out here has been concerned
only with the demands of the vehicle for mechanical energy. Present-day automobiles use
significant amounts of energy for a variety of other purposes. Most prominent are the
demands for heating and air conditioning, the former amounting to a maximum of perhaps
five thermal kilowatts and the latter to two electrical kilowatts in the average family car.
The additional demand for lights and various ancillary power equipment can amount to a
sizable fraction of a kilowatt. In our analysis we have neglected these demands, since an
auxiliary gasoline-fed burner is probably the most appropriate type of heating device, and
air conditioning is not yet regarded in most of the United States as a necessity.
24
3rthur
-------
ro
Maximum power delivered to wheel
speed reducers (hp)
(kw)
Number of motors
Maximum power delivered to wheel
speed reducers by each motor (hp)
Maximum motor speed (rpm)
Fraction of maximum motor speed
at which motor delivers its
maximum power* to the wheel
speed reducers
Maximum power output of which each
motor is capable at its maximum
speed
Weight of motors
Motor efficiency
System voltage
Weight of controller
Controller efficiency
Weight of speed reducers
Weight of cables, mounting, etc. (Ib)
Weight of cooling system
Overall efficiency (excluding speed
reducers)
Overall weight of electrical system
(including speed reducers)
'Assuming 100% overload capability.
TABLE 4
ELECTRIC TRANSMISSION PARAMETERS
Family Car Commuter Car Utility Car Delivery Van
City Taxi
City Bus
0.42
56
0.48
31
0.56
14.5
0.48
34.5
(hp)
(Ib)
(%)
(Ib)
(%)
(Ib)
(Ib)
(Ib)
ed
(%)
n
(Ib)
4 x 57 = 228
88
500
50
93
15
45
10
82
348
2 x 33 = 66
85
250
30
91
6
12
4
77
118
2 x 18 = 36-
80
125
28
90
3
12
3
72
82
4 x 36= 144
85
250
66
93
12
30
7
79
259
4x32 = 128
83
250
48
92
8
20
6
76
210
6 x 70 = 420
90
500
90
95
30
60
15
85
615
ta
n
-------
IV. ELECTRIC TRANSMISSION SYSTEM
This section discusses the state of the art of electric transmission systems as it applies
to the criteria for the vehicle classes arrived at in Section III.
The function of the electric transmission system is to drive and control the speed of
the wheels of the vehicle, using the power of the battery, fuel cell, or hybrid power source.
The system consists of one or more motors geared through speed reducers to the wheels of
the vehicle, a control unit for each motor, and electric cables that join the control units to
the power source. Since the control units, motors, and speed reducers have certain
unavoidable losses, the resulting heat must be dissipated by a cooling system, the weight of
which must be assigned to the electric transmission system.
In a battery-powered vehicle, it is extremely important to have high efficiency
throughout the electric transmission system, since battery power is limited. Replacement of
the automatic transmission with a fixed-ratio speed reducer increases the overall efficiency;
however, some weight penalty is imposed upon the motors, since they are called upon to
deliver high torques at extremely low speeds. This weight penalty is not too severe if
high-speed motors and large speed reduction ratios are used, as discussed in subsequent
sections.
A. MOTORS
The maximum rated speed of a motor depends upon the rotor construction, which
varies with the type of motor. The most important limitation is the rotor peripheral
velocity; this is determined by the speed and the rotor diameter, as shown in Figure 6 for
four types of rotor construction. Obviously, the smaller the diameter, the higher the
allowable rotor speed.
For the low-horsepower requirements of small vehicles, motors of relatively small
diameter can be used, and thus very high rotor speeds are possible. This helps to counteract
the intrinsically greater efficiency of large motors: by operating the low-horsepower motors
at higher speeds than the high-horsepower motors, one can largely avoid the penalty in
pounds per horsepower imposed upon smaller machines. Further significance of the
information contained in Figure 6 will be discussed when the relative advantages of ac and
dc motors are discussed later in this section.
Since high speed is a requirement for obtaining a high horsepower per pound, it is
desirable to have a small diameter and hence a long rotor. At the present state of the art,
therefore, only conventionally shaped, cylindrical-rotor dc and ac motors capable of
high-speed operation can be considered for vehicle propulsion.
The two unconventional dc motors described below are discussed for the sake of
completeness but must be excluded from vehicle use consideration at this time.
26
ZD.Hittle.Knr.
-------
fin
40
5 on
ousands of r\
•> r>
3 C
p
0.
to 5
o
oc
4
9
i
x
s x
\ \
"v V
\\
\
X
>
V
\
\v
\N
y
\
\
\
\
s
s
\
\
\
y
V
S
S
^.*i
a?
V
<
!v
>
s
s
\
&
^
*Sj
s
%
1
\?
^
$
^
x
•> X
, x
\ x
\ "^
\\
V
\
x
X
>
y
\
\
\^
\
\
^
\
^
^
s
\
\
\
^
\
\
\
Inductor Rotor
Squirrel Cage
Rotor
dc Motor Rotor
Salient Pole
4 6 8 10
Rotor Diameter (in.
20
40 60
FIGURE 6 ALLOWABLE ROTOR SPEEDS
27
-------
Pancake-shaped motors, of which the most common is the printed-circuit dc motor, are
large in diameter and short in length; consequently they have high rotor stresses at high
speeds and are not acceptable for electric vehicles. Widely used in low-inertia, fast-response
drives, such motors have axial working air gaps instead of the annular gaps of conventional
motors. This produces a different physical configuration, but the torque per ampere is about
the same.
So-called homopolar dc motors are unacceptable because, as explained later, the
characteristics of control systems do not permit use of any motor which requires extremely
high currents and correspondingly low voltages. This is unfortunate, since their voltage
requirement nicely matches the low voltage available from batteries. Homopolar motors also
employ liquid metal brushes, which do not permit the maximum speed that is theoretically
possible for the type of rotor construction used. The only appreciable application of these
machines has been in homopolar generators; these supply extremely heavy currents at a few
tens of volts and are desirable because of the smooth dc power they generate.
1. Direct-Current Traction Motors
Direct-current motors have been used for a number of years in electric locomotives and
have established a good performance record. These motors have carbon brushes and copper
bar commutators and are of series field construction. In this form of construction the
armature current establishes the field flux; proper control of the armature voltage provides
high torque at low speed and decreasing torque at higher speeds. Thus, an approximately
constant horsepower characteristic can be achieved over most of the speed range.
Since dc traction motors are built in relatively large sizes with high horsepower ratings,
they are not usable in their present form for electric vehicle propulsion. However, the
general design technology of traction motors for electric locomotives is adaptable to the
production of reliable dc motors for electric vehicles. Present traction motors have relatively
large rotors and do not operate at the high speeds required for electric vehicles, but the
peripheral velocities of the rotor and commutator establish design guidelines for scaling. It is
customary to have a maximum rated speed which corresponds to commutator peripheral
velocities of 12,000 ft/min, with overspeed during wheel slippage of 20,000 ft/min. This
corresponds to rotor peripheral velocities of approximately 15,000 ft/min and 25,000
ft/min, respectively.
Direct-current traction motors have also been applied to fork lift trucks, delivery vans,
and golf carts. In these applications cost did not justify developing the lightest and highest
speed motor. Instead, the selection was based on available electric equipment. Therefore,
these applications do not represent attempts to extend the state of the art.
The applications where effort has been made to obtain the maximum horsepower per
pound are in electric torpedo propulsion and aircraft dc generators and motors; even these
do not represent the ultimate, however, because their speed is restricted by the nature of the
systems in which they are used. If motors for vehicles were designed like these special
motors and run at speeds approaching their maximum rated values, they could deliver about
one horsepower per pound, at least in the larger sizes.
28
-------
The effect of speed is illustrated in Figure 7. For a given increase in horsepower per
pound, it can be seen that the speed of a small motor must be increased more than that of a
large motor. Fortunately, for the range of vehicle horsepower considered in this study, the
necessary speeds (12-19,000 rpm; see Table 4) are within the capabilities of today's
bearings. The accompanying stresses are within the rated design values, and reasonable brush
wear is maintained.
It must be emphasized that the values plotted in Figure 7 are based on the motors'
maximum horsepower rating - i.e., the horsepower they can deliver at their maximum rated
speed and torque. Normally, aircraft dc generators are not rated in this manner, so published
specifications must be carefully interpreted.
Because of the high performance which can be obtained from dc motors and the ease
with which they can be controlled, they are currently being used in small electric vehicles
and in larger experimental vehicles. In general, they provide a very acceptable interim
solution to the electric transmission system. To obtain higher horsepower per pound and
eliminate the brush wear problem of the dc commutator motor, however, an ac motor must
be used.
2. Alternating-Current Induction Motors
In an induction motor, the rotor is a squirrel-cage arrangement of conductors
imbedded in steel. This robust construction permits a high rotor speed (see Figure 6). In
addition, the active length of the rotor can be longer than that normally used for a dc
commutator motor. These features enable an ac induction motor to have a higher
horsepower per pound.
Since the induction motor requires alternating current and the battery provides direct
current, one of the functions of the control unit is to act as an inverter, controlling the
speed of the motor by varying the frequency of the ac power. The control unit must also
provide the proper voltage-to-frequency ratio to enable the motor to deliver its rated torque
over the entire speed range. The resulting torque vs speed curve differs greatly from that
normally associated with an induction motor; however, it is completely consistent with the
principle that any electromagnetic machine is capable of constant torque over its entire
speed range.
Induction motors for vehicle propulsion are similar in construction to high-perform-
ance aircraft motors of the low-slip variety. Such motors have low-resistance rotors that
minimize losses. Normally, low-slip designs would be unsuitable because of their high
starting currents; with a current-limiting control, however, starting current is not a problem,
and the motor design can be optimized for maximum efficiency.
29
arthur 2J.lLittIc.Knr.
-------
0.1
10
20
Output (hp)
50
100
FIGURE 7 WEIGHTS OF DC MOTORS
30
-------
Recent experimental electric vehicles and rapid transit systems have used induction
motors. By taking maximum advantage of their desirable characteristics, manufacturers have
been able to build induction motors that provide approximately two horsepower per pound.
(This is based on the horsepower at maximum rated speed; induction motors generally lack
the overload capability of dc motors, so the figure is an estimate based on 200% overload
capability for five minutes.) Figure 8 includes data on two recently publicized induction
motors.
Since the above applications have used induction motors of high horsepower, speeds
exceeding the 13-19,000 rpm range have not been necessary. It is doubtful that a high
horsepower-to-weight ratio could be maintained for lower-rated motors by increasing their
speed, because bearing problems would be encountered. However, the use of induction
motors eliminates brush wear at the expense of some increase in control unit weight and
cost; this advantage will be further discussed later.
3. Synchronous Motors
The synchronous motor is another possible contender for powering future electric
vehicles. In this type of motor, ac power is supplied to the stator to produce a rotating
magnetic field, and the rotor revolves synchronously with the field. The most common
design has a "wound field" — i.e., excitation current is supplied to the rotor through slip
rings and creates strong north and south poles in passing through the rotor windings. While
such a motor must have brushes to make connection with the slip rings, the brush wear is far
less than with the commutator of a dc motor.
Nevertheless, a truly brushless motor is desirable because brushes are a possible source
of trouble. The simplest form of brushless synchronous motor uses permanent magnets in
the rotor and, therefore, needs no brushes, but its power output is too low for vehicular
applications. A promising recent development is the solid-rotor brushless machine; much
work has been done on this design for aerospace generators, and various experimental units
have been built. These have characteristically been heavier than comparable motors of
wound-field construction, but they have the offsetting ability to be run at higher speeds.
This advantage cannot be fully utilized in small motors for the lighter vehicles, because their
bearings cannot withstand the speeds permitted by the rotor. (Sophisticated, high-perform-
ance bearings are too expensive for the uses considered here.) The speed capability of the
solid rotor might, however, prove advantageous for applications in vehicles such as buses
that require larger, more powerful motors.
The principal advantage of synchronous motors over induction motors is that the
former require a lighter, less complex control unit. Synchronous motors receive rotor
excitation power directly from the battery; their control units supply power to the stator
only. Induction motors, on the other hand, obtain excitation power from the control unit
via the stator circuit. As a result, their control units must be more powerful and thus
heavier. Furthermore, to prevent excessive heating of the battery, they must provide storage
for the excitation power that cycles back and forth from the motor. This storage usually
takes the form of a capacitor bank across the battery terminals, which involves a further
weight increase.
31
ZD.ltttlr.Knr.
-------
1000
800
600
400
200
•&
1
100
80
60
40
20
10
^
.s
/
'>*'
-------
Because of this advantage of the synchronous motor over the induction motor, it
would be desirable to use synchronous motors. Until the brushless variety is fully developed,
the slip-ring, wound-field synchronous motor is perhaps the most desirable solution for
powering an electric vehicle, if the higher control cost of the ac motor approach can be
justified.
The synchronous motors which might be used could be patterned after aircraft
generators. As with aircraft dc generators, the technology is highly developed and is directly
applicable to air-cooled and special oil-cooled machinery. The weights of a few standard
aircraft generators with slip-ring-fed rotors are shown in Figure 8 when rated as motors. Also
shown is the weight of a solid-rotor machine developed for a high-speed application.
The cost of high-speed air-cooled or oil-cooled motors with integral gear reducers
cannot be closely estimated from available data; the requirements for aerospace applications
differ greatly from those for mass-produced electric vehicle drive motors, even though
speeds, ratings, and technologies are similar. As a rough estimate, the manufacturing cost in
large quantities of motor and gear reducer might ultimately be about $4 per vehicle
horsepower for a dc commutator motor. For an ac induction motor, it might be as low as $2
per vehicle horsepower. The cost of solid-rotor synchronous motors would range between
these limits, approaching the upper figure for those of higher speed.
B. CONTROLS
Historically the speed of dc traction motors has been controlled by inserting a
resistance in series with the motor so as to regulate the current from the dc power source. In
addition to switching resistance into and out of the circuit, switching of the field circuit
between series and parallel connection is also customary. This manner of speed control
causes jerky performance and is inefficient from the standpoint of conserving battery
power.
Because a battery supplies an essentially fixed dc voltage, and because the speed of a dc
machine varies with the input voltage, a control must be used that will effectively vary the
voltage with negligible losses. The advent of solid-state devices has made such control
possible. Instead of reducing voltage by inserting resistance into the circuit, these devices
achieve the same effect by rapidly interrupting the current. The longer the interval of
disconnection, the lower the average voltage that the motor receives. The motor current is
not interrupted during the time that the battery is disconnected; rather, the motor circuit is
closed on itself.
The switching elements which are used to disconnect the motor winding from the
power source are silicon power transistors or silicon controlled rectifiers (SCR's). The
difference between them is that SCR's can be turned off only by reversing their
anode-to-cathode voltage for a short time; silicon power transistors can be shut off simply
by interrupting their control current. Both devices are turned on by applying a low-power
pulse to their control terminals. In the case of the power transistor, the low-power pulse
33
Srtbur ZD.lttttlc.Dnr.
-------
must persist for the duration of power conduction. Power transistors were initially used, but
SCR's are now preferred because of their hjgher power handling capability, even though
they require special turnoff circuits. Both devices have essentially unlimited life.
1. Chopper for Direct-Current Motor
Frequently the control for a dc motor is termed a chopper, since it connects the
battery to the motor intermittently. As explained above, the rapid interruption is timed to
achieve the desired average voltage. A number of different schemes are possible: some use a
fixed on-time with variable off-time, while others use a variable on-time and a fixed
off-time; still others use a combination of these two. In general, it is desirable to operate at
the highest possible chopping frequency so as to minimize the fluctuation of current; some
compromise must be made, however, to reduce switching losses.
The chopper is not only much more efficient than the switch resistor type, but it also
permits very exacting control of current. This, in turn, permits continuous, stepless torque
control. Probably more important, the current can be limited to protect the motor and the
solid-state controls themselves; with proper current limiting, high reliability can be expected
from the chopper and motor. For example, if maximum torque is required from the motor,
the current limit can be set at, say, twice rated motor current. This will enable the motor to
deliver approximately twice its rated torque. Both the motor and control should be designed
to withstand this current for perhaps 5 minutes; in this way, both the motor and the control
are protected even if the motor is stalled.
The 5-minute rating of the motor is an arbitrary figure. It could be designed to be 1 or
2 minutes or even 10 minutes. The particular design will depend upon the overall system
efficiency, weight, driving pattern, etc. However, the solid-state component in the chopper
must be capable of carrying the overload current continuously, since its overload capability
is extremely limited. Most of the other components in the chopper, except those that have a
thermal storage capacity, must also be able to carry the overload current on a continuous
basis.
In principle, the efficiency of the controller can be extremely high, since the process is
basically lossless. However, when the SCR is carrying current, it exhibits a 1-volt drop,
which when multiplied by the current gives the forward conduction loss. In addition, there
are switching losses, blocking losses, and losses associated with the free-wheeling diode that
closes the motor on itself during the times that the SCR blocks the current from the battery
to the motor. Other losses are associated with the turnoff capacitor and its associated
charging circuitry.
The efficiency of the chopper is greatly affected by the voltage of the system, because
conduction loss due to the 1-volt drop is relatively less at higher voltages. Considering this
loss alone in the case of a 12-volt system, the efficiency cannot exceed 92%. When the other
losses are included, a rather low overall efficiency would result. Therefore, it is essential to
use a high system voltage.
34
-------
High voltage is important not only for increasing efficiency, but also for reducing the
weight of the chopper. The physical size of the solid-state components is dependent upon
the current and is unaffected by the voltage. This independence of voltage does not hold for
the capacitive and inductive components in the chopper, however; their size and weight are
in general related to the power handled and also to the precise design of the chopper.
As a rule, it is best to use as high a voltage as possible to reduce the magnitude of the
current. To keep the rated current below 100 amperes and hence the overload current below
200 amperes, it is necessary to use voltages of the order of 150-500V for the range of
vehicles studied. For these voltages the weights which should be achievable are shown in
Figure 9. These weights assume fan cooling of the choppers, which is possible because of
their high efficiencies (90% or more). Heavy contactors and circuit breakers should not be
necessary, because of the current-limiting capability of the chopper.
Considering the small size of the chopper, it might be feasible to attach it directly to
the motor housing. In this way, dissipation of heat from the solid-state power handling
components could be increased, and heavy-current carrying terminals could be avoided. This
concept is analogous to the incorporation of silicon diodes within the housing of present
automobile alternators.
Although the basic principle of the chopper is well understood, the device has only
recently been applied to experimental fork lift trucks, delivery vans, and other vehicles. The
technology is available, but it will require a great deal of development to bring the chopper
from a promising concept to readily available equipment for electric vehicles.
The present manufacturing cost of a chopper control is about S20 per vehicle
horsepower. However, considering the actual material costs and the fact that the
manufacturing processes lend themselves to automation, an order-of-magnitude reduction to
$2 per horsepower would appear to be reasonable.
All the logic functions associated with the timing and control of the chopper could be
done with microelectronic integrated circuit modules. These are incredibly small and can
perform many control functions at an extremely reasonable cost. Therefore, it would seem
that a great deal of ingenuity and development should be directed toward applying chopper
technology to a mass-produced, reliable, low-cost product.
To accomplish regenerative braking, all that is necessary is a slight rearrangement of
the basic power handling component and a small amount of additional logic. Regenerative
braking is only beginning to make an appearance in chopper circuits, and much innovation is
needed; e.g., through appropriate current sensing and making the circuit changeovers at
current zero, it would be possible to do away with all the large contactors or dc current
interrupters which are frequently incorporated in choppers.
Highly efficient regenerative braking can permit the return of power to the battery.
The amount returned depends upon the driving power of the vehicle as well as circuit design
and efficiency. For vehicles such as buses that make frequent starts and stops, it has been
35
3rthur ZD.lUttle.llnr.
-------
1000
800
600
400
200
•E 10°
5
80
60
40
20
10
- X* X|
X X
h- X
X
x
X
-x-
X
J I
10
20
40 60 80 100
Motor Output (hp)
200
FIGURE 9 WEIGHT OF CONTROL UNIT
36
-------
reported that the saving in power can be as much as 30%. For vehicles that use a more
normal driving pattern, however, the energy saving is not large enough to justify the cost of
adding this feature.
Regenerative braking does not make conventional brakes unnecessary, for it is of
little use at low speed; most of the power is returned during the initial period of
deceleration.
2. Inverter for Induction Motor
When an induction motor is used as a propulsion means, an inverter is used as the
control for varying its torque and speed. An inverter takes dc power from the battery and
changes it to ac power of the appropriate voltage and frequency; thus, it combines the
voltage-adjusting function of the chopper with an ability to vary the frequency as well.
Inverters using SCR's for driving induction motors over wide speed ranges are
relatively new. They are just coming onto the market for industrial variable-speed drives and
have been used with a few experimental electric vehicles.
As with the chopper, power SCR's are used in an inverter to connect the three
motor windings to the battery on a time-varying basis. In essence, at least three choppers are
required, one for each of the three phases of the motor, because each phase must be
cyclically energized and each carries the same peak current. In practice, it is often
convenient to use essentially six choppers, each capable of delivering half the required
voltage. As with the chopper, the peak current is the limiting factor in design.
In addition to all the components associated with a chopper, a large bank of
capacitors must be placed across the dc input. This results from the fact that reactive power
must be handled by the inverter. A further component which must be added is a special
speed-sensing tachometer. A signal from this tachometer is used to determine the frequency
of the inverter.
The size and weight of an inverter are typically about three times those of a chopper
of equivalent power and voltage, excluding the capacitor bank across the dc input. The
estimated weights are graphically shown in Figure 9 as a function of voltage and power. A
manufacturing cost of $5 per vehicle horsepower should eventually be possible, again
excluding the capacitor bank and tachometer.
The fact that the inverter has a more complex logic requirement than the chopper
should not increase the ultimate cost; specialized integrated circuits would be developed for
vehicle inverters to accommodate all the complex functions. As in the case of the chopper,
current limiting is essential and must be applied with instantaneous control to protect both
the inverter and the motor, as well as to provide torque control.
If regenerative braking is considered justified by the application, it can be
incorporated within the inverter at a slight increase in cost by using the motor as an
induction generator. This does not require that the solid-state components be shifted
around, as in the case of the chopper circuit, but only that additional logic be incorporated.
37
3rthur ZD.lttttlr.Knr.
-------
To protect both motor and inverter, the controlling of the SCR's must be timed so as to
limit the rate at which power is returned to the battery.
3. Static Commutator for Synchronous Motor
As with the induction motor, the speed and torque of a synchronous motor are
controlled by varying both the voltage and the frequency of the power supplied. However,
the control for a synchronous motor differs somewhat from an inverter; it is called a static
commutator for purposes of distinction as well as to indicate its functional relation to the
copper bar commutator of a conventional dc motor. The commutator and brush assembly
of a dc motor in effect chop the dc input into variable-frequency power for the motor
windings; when this is accomplished by a separate static commutator, the roles of stator and
rotor can be interchanged. This permits a large reduction in the current that the slip rings
must carry, because the current in the rotor of a dc motor is much larger than that in the
stator.
The power circuit of the static commutator is identical to that of the inverter,
except that it needs no capacitor bank across the dc supply. However, the control of the
SCR's is somewhat different: the position-sensing tachometer attached to the motor
determines which windings should be energized at any given moment. The tachometer can
take a number of forms; experimental arrangements have been employed using photoelectric
detectors, Hall detectors, and reluctance detectors. The information from the position
sensor is used to control the main power SCR's.
The size of the static commutator is about the same as that of the inverter. The
weights given in Figure 9 and the cost of $5 per vehicle horsepower also apply for the basic
static commutator; to these figures must be added the small weight of the tachometer and
its associated cost.
C. SPEED REDUCERS
To couple a high-speed motor to the wheel, a speed reducer with a relatively large
reduction ratio is needed. For railroad traction, a gearbox with a low reduction ratio is used;
these are very rugged but heavy units. At the other extreme, very lightweight gearboxes have
been developed for torpedo propulsion systems, but these are designed for short life and
thus are not suitable for vehicular use. It would seem that a compromise between these two
technologies would lead to a gearbox with the desired reliability and light weight.
The large experimental electric vehicles built to date have used planetary gearboxes
that are directly connected to the motor and employ oil cooling. This is a lightweight
arrangement particularly suited to motors of relatively high horsepower.
38
ZD.UtttleJnr.
-------
Another solution is the "motorized wheel," in which the motor and gear reduction
unit are incorporated in the hub of the wheel. This compact arrangement is now used in
some earth-moving equipment and might be well suited to large vehicles like buses.
On small experimental electric cars, double timing belts have been used to achieve
relatively high speed-reduction ratios. Belt drives are an extremely lightweight and
inexpensive means for coupling the motor to the wheels.
In view of the mature technology of speed reduction devices, adapting one or more
of these approaches should not be difficult. The principal task is to make a detailed analysis
that will indicate which offers the lowest cost and best arrangement. For the small utility
car, which would use air-cooled motors of rather low power, the multiple timing belt
approach seems desirable; for the bus, the oil-cooled gearbox or possibly the motorized
wheel might be preferable.
D. COOLING REQUIREMENTS
The heat losses which occur in the motor, the speed reduction unit, and the control
unit must be removed by a cooling system. Air cooling, which is the simplest method,
requires a constant-speed blower and duct arrangement. The more elaborate oil-cooling
system requires an oil pump, heat exchanger, and other components. In both cases, the heat
to be rejected would be used to warm the passenger compartment during the winter and
discharged to the surroundings in warmer weather. The motor may have to operate for
sustained periods at extremely low speed, so it cannot be used to drive the cooling system;
the system must be separately driven.
39
3rthur ZD.lUttlrJnr.
-------
V. ELECTROCHEMICAL POWER SOURCES
This section contains a review of the present state of the art of those electrochemical
power sources, batteries and fuel cells,which have prospects of application as the prime
power source in electric vehicles, together with some prognosis of their future development.
Consideration is also given to the possibility of hybrid electrochemical systems consisting of
battery-battery and fuel cell-battery combinations. (The possible application of batteries in
mechanical-electrical systems is covered in Section VI.) The review of electrochemical power
sources is organized into four subsections covering conventional batteries, batteries under
development, fuel cells, and electrochemical hybrid combinations.
A. CONVENTIONAL BATTERIES
1. Lead-Acid
The lead-acid battery is the oldest of the secondary battery systems and has been in
existence for over 100 years. It is also the most important system economically, accounting
for about 80% of the more than S500 million annual sales of secondary batteries in the
United States. While the largest proportion of these sales is for the specific application of
starting, lighting, and ignition (SLI) in vehicles powered by internal combustion engines, a
substantial volume is used as the power source for non-highway electric vehicles, such as
fork lift trucks, golf carts, and mine locomotives.
SLI batteries are designed for float service with occasional high power discharge and
are not suitable for traction service, which involves repeated deep discharge cycling. Bat-
teries for traction applications have thicker plates and are generally of more rugged con-
struction. Their delivered energy density is now in the range of 10-14 w-hr/lb. While high
power drains cause a decrease in delivered energy, power densities of as much as 35 w/lb can
be sustained for short periods.
The major virtues of the lead-acid system are its high degree of reliability (a typical
lifetime of four to six years in traction service under a continuous daily cycle of full charge
and discharge) at low cost (OEM prices approximate S0.25/lb for SLI and $0.50-0.60/lb for
traction type batteries). While these virtues make lead-acid today's most important practical
battery system for electric traction,* its low energy density eliminates it from consideration
as the primary energy source for almost all of the six classes of vehicles under consideration
in this study. The sole possible exception is the utility car with lightweight construction, the
specifications for which are close to the upper limit of the lead-acid battery's technical
capabilities.
The question arises as to the potential for improvement in the lead-acid system.
Incremental advances in energy density are certainly possible, and in fact there has already
been a substantial improvement (of the order of 30%) in these parameters over the last
15-20 years and this is expected to continue. For example, a major U.S. manufacturer is
The extensive history of lead-acid battery use in highway vehicles, particularly in the United
Kingdom, is reviewed in Appendices A and B.
40
3rtb«r
-------
about to introduce a new industrial battery with an energy density of 16 w-hr/lb at a ten
hour discharge rate. These advances have primarily resulted from the utilization of lighter,
stronger materials for cases, improvement in the design of the plates, new grid alloys, more
porous and durable separators, etc. Somewhat more radical changes are being worked on by
a group in England (U.K. Electricity Council Research Centre), which is investigating im-
provements in capacity that might result from pumping electrolyte through the battery
plates, although the extra complexity may make the scheme unattractive. Some publicity
has been given in the United States to the use of traces of cobalt salts in the electrolyte to
improve energy density (Electric Fuel Propulsion, Inc., see also Appendix A). The presence
of cobalt in closely controlled proportions can lessen grid corrosion, thereby permitting the
use of thinner plates and yielding a higher energy density; it does, however, increase the rate
of shedding of active material from the plates, thus shortening battery life.
With regard to power density, the maxima achieved in today's conventional lead-acid
batteries are only moderately below the demands of the vehicle types as listed in Table 1.
Operation at high power density does lower the energy storage capacity of the battery,
however, and rapid charge and discharge raises the internal temperature of the battery and
tends to shorten its life. An interesting recent development is that of a high rate lead-acid
battery for pulse discharge and having "pile type" construction. This battery, being devel-
oped by Gould-National Batteries, Inc. for the U.S. Navy Electronics Laboratory, has a
power density of around 60-70 w/lb and might possibly find application in hybrid electrical
vehicle systems (see Sections IV-D and VI).
Thus the limitations on its energy density would seem to preclude the large scale use
of lead-acid batteries as a prime energy source for highway vehicles. Much wider use of these
batteries, based on their low cost, rechargeability, and availability is, however, foreseen for
such applications as transportation within industrial plants, hospitals, airports, resorts, and
off-highway shopping centers and communities. Anticipated advances and cost improve-
ments in motors and control systems will also benefit electric vehicles used for these pur-
poses.
2. Nickel-Cadmium
The salient technical features of the nickel-cadmium battery are its excellent service
life, an energy density of 12-14 w-hr/lb, which is closely comparable to that of the lead-acid
battery; and a power density which in certain constructions can be as high as 300 w/lb.
There are two major types of nickel-cadmium batteries: those with pocket plates, which
have cycle lives in excess of 3000 or so, and those with sintered plates, whose cycle lives are
less than half as great but which have very high power densities. Because of its low energy
density it is, like the lead-acid system, outside the scope of our primary power source
specifications for electric vehicles, except perhaps for the small utility vehicle. Its high
power density however makes it an attractive prospect on technical grounds for various
types of hybrid power source, as will be discussed in Sections V-D, VI and VII. This is
especially true of the 'bipolar' nickel-cadmium battery developed by Gulton Industries, Inc.
for high discharge rate applications. This battery has very thin sintered plates attached to
thin metal sheets which act as the connectors between adjacent cells. The thin electrode
design permits the achievement of power densities of around 300 w/lb.
41
Arthur ZD.lUttle.lfnr.
-------
The nickel-cadmium battery has not been used in traction applications, except to a
limited extent in lift trucks in Continental Europe and in a demonstration automobile in
Sweden (Appendix B). The major drawback is its cost, which on a per pound basis is about
five times that of the lead-acid traction battery. Since a significant part of these costs derives
from the costs of nickel and cadmium, they are unlikely to fall sufficiently to change the
nickel-cadmium battery's competitive position. A further argument can be raised against
cadmium on the grounds of availability: supplies of the element are not great enough to
support a major application in electric vehicles, and since it is a by-product of lead and zinc,
its production cannot be economically expanded for its own sake. Thus the nickel-cadmium
battery in any form seems unlikely to play a major role in electric vehicle development.
3. Nickel-Iron
The nickel-iron system has an energy density slightly lower than those of nickel-
cadmium and lead-acid and a lifetime which can be measured in decades. It is also a very
rugged battery, capable of withstanding much physical and electrical abuse. Its power den-
sity, however, is not as great as that of the nickel-cadmium system, and because of the
smaller volume of production and the processing costs involved, its price structure is not
significantly lower. The major reasons why the nickel-iron battery has not found greater
general application are its poor charge retention and low electrical efficiency in charging,
both of which characteristics lead to excessive gassing and a high maintenance requirement.
Although it has found some acceptance in European lift trucks and in U.S. mine loco-
motives, the low energy density of the system gives it little relevance as a potential power
source for electric highway vehicles.
4. Silver Batteries
There are two secondary batteries which use silver oxide cathodes: the silver-zinc
system and the silver-cadmium system. The former battery has a high energy density
amounting in certain configurations to about 60 w-hr/lb, and an excellent power density of
as much as 200 w/Ib. properties which on technical grounds would appear to make it a
significant contender for many electric vehicle applications. It has, in fact, been used
recently in certain experimental electric vehicles, most notably the CM "Electrovair" and
the Yardney Electric converted Renault Dauphine (see Appendix A). However, the cycle life
at full discharge seldom exceeds 40-50 cycles, and the high cost and relative scarcity of silver
rule it out for large scale use. The 500-lb battery in the Electrovair II was reported to cost
some SI 5.000. or S30/lb; and about 30% of the packaged weight is accounted for by silver,
which now has a raw material cost of about S35/lb. Yardney Electric has suggested that the
silver might be leased, but the poor cycle life is a difficult problem and lack of availability
unsurmountable. Its use therefore seems likely to be restricted to demonstration vehicles
and test bed evaluations.
42
Arthur
-------
The properties of silver-cadmium batteries are in many ways intermediate between
those of nickel-cadmium and silver-zinc batteries. Their energy density of about 25 w-hr/lb
and their high power density, comparable to that of silver-zinc batteries, make them
marginal candidates from a technical standpoint for certain of the vehicle applications listed
in Table 1. However, as with silver-zinc, high cost and insufficient availability of materials
eliminate them from serious consideration for large scale use.
B. BATTERIES UNDER DEVELOPMENT
The technical parameters of the conventional batteries described in the previous
subsection clearly fall short, particularly in energy density, of the requirements specified in
Table 1. Since conventional batteries use materials with relatively high equivalent weights,
an obvious route to improvement lies in the use of reactions which have higher energy
release and employ active materials of low equivalent weight. The source of energy in all
batteries is the oxidation-reduction reaction which takes place between the active materials
of the positive (cathode) or oxidizing plate and the negative (anode) or reducing plate. The
energy released, corresponding to the weights of these active materials alone, can be calcu-
lated and sets an upper limit for the energy density of a given battery system. In practice,
when the weights of case, grids, separators, electrolyte, and supporting structures are
considered, the practically attainable energy density is only some 10-20% of this theoretical
maximum. Nonetheless, it is evident that the higher the energy release in the cell reaction,
the higher the energy density of the resulting battery.
Prime candidates for use as anode materials in high energy density batteries are the
alkali metals, in particular lithium and sodium. Their very reactivity, however, imposes some
severe restriction on their mode of utilization. They cannot be used with conventional aque-
ous electrolytes, because of the rapid direct chemical reaction; their use is possible only in
combination with nonreactive electrolytes, in particular molten salts and organic liquids.
The former introduce the complication of a high operating temperature and, in many in-
stances, severe problems in materials selection, while the latter tend to be hampered by low
electrolytic conductivity, poor shelf life, and cathode polarization problems, which limit
their power density.
The other major approach to improvement in energy density is the use of atmos-
pheric oxygen as the cathodic oxidizing agent. The weight of the cathode in the battery is
then limited to that of a catalytically active oxygen electrode, the oxidant (air) being
available in limitless amount with no weight penalty.
Power densities are a reflection of the kinetics of the various electrochemical reac-
tions taking place in the battery. They can be improved by attention to certain physical
factors, particularly the surface area of the electrode, and are increased at elevated tempera-
tures. For this latter reason, the molten salt systems are particularly favored for applications
involving high power density.
This subsection reviews the status of the work presently under way on improved
battery systems for use in electric vehicles.
43
,3rthur ZD.lUttlf.3lnr.
-------
1. High-Temperature Alkali Metal Batteries
a. Lithium-Chlorine
The battery with the highest potential energy density of those now under investiga-
tion is based on lithium and chlorine as reactants and is being developed by the General
Motors Corporation. Its theoretical energy density is approximately 1000 w-hr/lb, and it has
an open circuit potential of 3.5 volts. In its work to date, GM has demonstrated the cell's
capability for generating very high current densities - on the order of several thousand
amperes per square foot.
The principal liabilities of the lithium-chlorine battery are the difficulties associated
with storing chlorine gas under pressure and at temperatures of around 650°C, and the
tendency of the cell to self-discharge. The first problem is being tackled by using highly
adsorbent charcoal to reduce the pressure of the stored chlorine. This approach seems quite
successful, although a great deal more needs to be learned about the cycling capability of
charcoal in this application.
While its electrochemical characteristics are very encouraging, the lithium-chlorine
battery presents many novel and extremely challenging problems in materials engineering,
and extensive development will clearly be needed before the system can be made practical.
The question of hazards arising from the storage of chlorine gas under pressure and at high
temperatures seems, at this stage at least, likely to limit the chances of such a battery's being
adopted as a power source for vehicles.
It is not possible to make any accurate estimates of the effective energy and power
densities, still less the costs, to be expected for the lithium-chlorine battery. Much depends
on the complexity of the safety engineering and packaging that would be necessary. It seems
likely, however, that the system will easily possess the technical capabilities required to
satisfy the most exacting of the electric vehicle's specifications.
With regard to costs, a theoretical minimum quantity of about 25 pounds of lithium
would be required for a battery capable of meeting the requirements of the family car. At
today's price for lithium, $7/lb, this corresponds to SI75. The actual quantity of lithium
needed might be half as much again as the 25 pounds, but with anticipated lower produc-
tion costs of the metal in quantity the resulting materials cost for lithium might not be
much more than $100. While this is a significant item, it would not appear to rule out the
system from further consideration.
Maintenance of the operating temperature is a problem with all high-temperature
battery systems. In a well-insulated system, the self-discharge rate is usually sufficient to
maintain temperature. However, good insulation tends to be an embarrassment when the
battery is working at high power levels, either on charge or on discharge, since the excessive
amount of heat generated causes an undesirable rise in the internal temperature. On the
other hand, if the battery is to be shut down for long periods, it seems essential that some
kind of built-in combustion heater be included in the system.
44
9rthur
-------
b. Sodium-Sulfur
Another approach to the use of alkali metals in high-temperature batteries is being
taken by the Ford Motor Company. Ford's cell uses sodium and sulfur as the reactants, and
the electrolyte is a special ceramic material permeable only to sodium ions. The properties
of the ceramic are critical for the effective functioning of this battery. While the sodium-
sulfur battery does not have the theoretical energy density of the lithium-chlorine system
(300-600 w-hr/lb, depending on the extent of the reaction, versus more than 1000 w-hr/lb),
it does have two major practical advantages. First is the lower temperature of operation,
approximately 300°C. At this temperature, sulfur is a liquid, and containment of the
cathode materials thus presents far fewer engineering problems than in the lithium-chlorine
system. The second advantage is that the battery reactants, sodium and sulfur, are both
extremely abundant and very inexpensive; there are thus no immediate and fundamental
reasons why the battery might not eventually develop into a low-cost power source for
large-scale use.
Since Ford's work is proprietary, relatively little specific information is available.
Present efforts are apparently confined to experiments with single cells, the goals being to
determine the factors affecting electrical efficiency and cycle life and to find solutions for
the various engineering and materials selection problems which are encountered. Since
multi-cell work is bound to raise a whole new series of problems concerned with design and
with maintenance of the electrical balance of cells in the battery, it is not possible to make
an accurate prediction of the time scale for the development of this battery into a practical
power source for vehicles.
c. Capacitive Storage
A third approach to high-temperature battery development has been taken by the
Standard Oil Company (Ohio). SOHIO's work was originally based on the use of carbon
electrodes with very high surface areas to provide capacitive energy storage. All electrolyte-
electrode interfaces have double layer capacitances; however, these are in general quite small
and do not store significant quantities of energy. In the case of porous carbon electrodes,
the internal surface can reach huge values, and the amounts of energy that can be stored
capacitively become comparable with those stored in a battery of equivalent weight. Elec-
trolytic condensers with capacities of the order of farads can be set up by immersing porous
carbon electrodes in an aqueous electrolyte such as sulfuric acid. A very substantial enhance-
ment in capacity and hence in the capability for energy storage, can, however, be obtained
by using a molten salt electrolyte. A further bonus comes from the much higher breakdown
voltages of these electrolytes, which permit the cells to be operated at levels of 3 volts or
more, compared with a maximum of 1.8 volts or so in aqueous systems. Since the energy
storage is proportional to the square of the voltage, this means a substantially greater energy
density.
45
2D.lttttlp.Knr.
-------
In practice, the SOHIO device, which contains a molten potassium chloride-lithium
chloride eutectic as the electrolyte, does not behave as a pure capacitor. It was found
preferable to substitute aluminum for carbon at the negative electrode (to reduce self-
discharge, due to lithium solubility), and the lithium enters the aluminum interstitially. At
the positive electrode it seems likely that the chloride ion is at least partly discharged but is
adsorbed very strongly on the carbon and so never appears as a gas. The system may thus be
partly a capacitor and partly a special type of lithium-chlorine battery in which the active
materials remain within the current collectors. The fall-off in voltage as the cell is discharged
is intermediate between those typical of batteries and capacitors.
At its present stage of development the SOHIO device appears to be capable of
attaining an energy density of 50 w-hr/lb and a power density of 150 w/lb in its packaged
condition. Experimental batteries with 30-kwh storage capacity are now being designed.
Thermal insulation is provided by a vacuum container which will maintain operating temper-
ature for 40 hours without auxiliary heat. The self-discharge rate is low, taking approxi-
mately 20 days from full charge. While the SOHIO battery seems likely to have inherently
lower energy densities than those of other types of high-temperature systems, with improve-
ments in the technology of the carbon cathodes, its energy density might eventually be
raised to about 100 w-hr/lb. Estimates of cost are necessarily very tentative because of the
novelty of some of the fabrication processes, but it appears that in large-scale production
they might approach a level of $1.50/lb. Preliminary indications are that the lifetime will be
about three years. With a pilot manufacturing plant going into operation SOHIO's battery
seems to be most advanced of the high-temperature systems now under development.
d. Lithium Cells With Tellurium, Selenium, and Other Non-Metals
Other types of high-temperature cells involving lithium anodes and molten
nonmetallic elements as the cathodes are under investigation at the Argonne National
Laboratory. The electrolyte is a relatively low melting ternary eutectic mixture of lithium
fluoride, chloride, and iodide made into a solid paste by the carefully controlled addition of
micron sized particles of lithium aluminate. The majority of the work has been focused
upon the lithium-tellurium and lithium-selenium couples, the cell reaction resulting in the
formation of lithium telluride or selenide, which remains dissolved in the molten cathode.
Operating temperatures are necessarily above those of the melting points of the cathode
elements and are in the 470-500° C range for lithium-tellurium and 350-400° C for
lithium-selenium.
The great virtue of these systems is the extremely high degree of reversibility of the
electrode reactions and their capacity for sustaining current densities of approximately
10,000 amperes per square foot. Polarization of the cells is very low and entirely accounted
for by internal resistance. The major constructional material is Armco iron and corrosion is
apparently no great problem. Projections based upon the behavior of single cells having an
electrode area of 10 square centimeters indicate that the packaged energy densities would be
90 w-hr/lb for Li/Te and 120 w-hr/lb for Li/Se. Power densities are very high: 200-500 w/lb
for Li/Te and 300-600 w/lb for Li/Se.
46
3rtbur 3t.ltittlc.Dnr.
-------
Probably the greatest handicap to the development of these cells for vehicle applica-
tions is the cost and relative scarcity of the cathodic elements. Because of this some prelimi-
nary experiments were carried out in recent months with a lithium-phosphorus cell, but the
necessity of working above the melting point of red phosphorus requires operating tempera-
tures above 600°C, and the observed voltage of the cell is only about 1 volt versus 1.75 volts
for Li/Te and 2.4 volts for Li/Se. This initial work with phosphorus has therefore been
unpromising. The most recent work has been with the lithium-sulphur system and has been
very encouraging as far as it has gone. This system appears to have great promise as a
vehicular power source.
Sodium cells are not of interest because of the higher melting points of the sodium
salt electrolytes and the greater tendency of sodium to dissolve in the electrolyte.
2. Alkali Metal Batteries With Organic Electrolytes
The use of nonreactive organic liquids as electrolytic solvents in alkali metal bat-
teries permits ambient-temperature operation and thus avoids many of the difficulties asso-
ciated with molten salt and ceramic electrolytes. It has its full share of disadvantages,
however, since for a variety of reasons, organic liquids are inferior electrolytic solvents.
A great deal of work has been done over the last five years, much of it supported by
NASA, on the selection of cathodes and organic electrolyte combinations for use with
lithium anodes in primary batteries. The major problems have been the identification of
non-aqueous electrolytic solvents compatible with lithium and having sufficiently high con-
ductivity, and of cathode materials capable of sustaining even very small current densities
without extremely high polarization. Some progress has been made, although the best of
these systems can support current densities only 5-10% as great as those possible in conven-
tional aqueous batteries.
The development of rechargeable systems is at an even earlier stage. In work carried
out some three or four years ago, the Lockheed Aircraft Company demonstrated the
rechargeability of a lithium-silver chloride battery having a lithium perchlorate-propylene
carbonate electrolyte. The best of these cells showed some 40 or 50 cycles of operation at
almost 80% discharge before the capacity fell off significantly. This system was not devel-
oped further, because the high equivalent weight of the cathode gave it no substantial
advantage in energy density over many aqueous systems. Other attempts at recharging
organic electrolyte batteries have encountered severe problems with lithium plating and with
the recycling or solubility in the electrolyte of the cathode materials (generally halides or
sulfides of transition metals).
More recently, Gulton Industries, Inc. has engaged in the development of a recharge-
able lithium-nickel fluoride battery having a potassium hexafluorophosphate-propylene car-
bonate electrolyte. Claims have been made that this battery system has the potential to
47
Arthur ZD.1Uttle.llnr.
-------
attain a packaged energy density of 150 w-hr/lb and a power density of 20-30 w/lb; how-
ever, the published technical information* shows that achievements to date fall short of
these goals. On repeated cycling a nominally 0.025 amp-hr cell discharged for 8 hours at 20
ma (the current density at the electrodes being 0.1 ma/cm2) showed a voltage decrease of
from 3.0 to 1.4, the value at the midpoint being 1.75 volts. If these conditions are typical,
then a 13-amp-hr cell which apparently weighs 0.55 pound would deliver 27 w-hr/lb, or
approximately 40 w-hr/lb if the cell were discharged completely. High discharge currents
have been reported, but only for pulse conditions of perhaps a few milliseconds. The falloff
in voltage at a steady discharge current density of only 0.1 ma/cm2 indicates that high
power densities are unlikely to be attained under sustained conditions. A five-hour discharge
rate may well be the practical limit for this type of battery.
Gulton has had more success than other groups in recharging lithium non-aqueous
batteries, but even so they have found it necessary to use low current densities and very thin
coatings of active material on the current collector to achieve significant cycle life. The
latter requirement, of course, tends to decrease the energy density of the cell.
As a class, rechargeable alkali metal batteries with organic electrolytes seem a long
way from practicality for vehicle use. Their difficulties derive from fundamental electro-
chemical problems which even in comparison with the engineering problems faced by the
high-temperature batteries, could take a long time to resolve. Gulton Industries, however, is
optimistic about their prospects and has entered into a vehicle development program with
American Motors Corporation (see Appendix A) to use its battery with a high-rate nickel-
cadmium battery in the hybrid power source described in Section V-D. The long-term cost
estimates for the lithium battery have been estimated to be in the range of $4 - 5/lb.
The Electrochimica Corporation has also advocated the use of organic electrolyte
lithium batteries in electric vehicles. Its work has been concentrated on a lithium-copper
fluoride system, and claims are made that this system has particularly high current densities
for this class of battery. Lacking specific information, we cannot make a detailed assessment
of the system. It is not claimed to be rechargeable, however, and therefore is not yet at a
stage for consideration in vehicles.
3. Metal-Air Batteries
The employment of air cathodes represents another general approach to the develop-
ment of batteries with higher energy density. In most programs in this area, aqueous electro-
lytes have been used in combination with relatively conventional metal anode materials. The
concept is not new; low-rate zinc-air primary batteries have been available for more than
half a century. The novelty in the recent work is in the use of air cathodes capable of
supporting very much higher current densities than those of the conventional cells, an
advance derived from fuel cell technology.
* H. N. Seiger, S. Charlip, A. E. Lyall, and R. C. Shair, Organic Electrolyte Batteries, paper
presented at 21st Annual Power Sources Conference, Atlantic City, N. J., May 16-18, 1967.
48
Arthur ZD.TUttlr.llnr.
-------
The electric vehicle application requires rechargeable batteries, which present addi-
tional problems to air-cell technology.Problems arise with the rechargeability of the anode
and with the oxidation resistance of the air cathode, which has to withstand oxygen evolu-
tion or be supported by an auxiliary third electrode for recharging. The work on zinc-air
cells is hampered by the classic difficulties of recharging the zinc electrode, of which
dendrite growth is the most significant. The following paragraphs review the present activi-
ties in metal-air battery development.
a. Zinc-Air (General Atomic)
One of the most prominent of the programs on zinc-air batteries is that being
conducted at General Atomic, now a division of the Gulf Oil Corporation. The work is
supported in part by the Edison Electric Institute, and the British company, Joseph Lucas
Ltd., has become a partner in the program. The basic approach taken is to avoid the
dendrite growth problem by pumping the electrolyte over the surface of the electrode. The
technique is very effective but results in considerable engineering complexity. On discharge,
the reaction product, zinc oxide, is swept out of the cell and stored separately, being fed
back in again on charge. The cathode is of sintered nickel with a palladium catalyst, and air is
forced through it with a compressor. The system appears to have a near-term energy density
capability of 50-60 w-hr/lb and a power density capability of 30-35 w/lb. These figures are
net after subtracting the parasitic power load for pumps, compressors, and other auxiliary
equipment. This latter drain is estimated at a fairly constant 15% of gross power output in
the size range of greatest interest, 20-80 kw.
The largest modules built to date have been sized at 1 5 kwh, 4-5 kw at 36 volts, and
they will be used to determine the performance of the system on extended cycling. A full
charge can be completed in about two hours. If the system is not abused electrically, it
would appear to have a cycle life comparable to that of the lead-acid battery. Carbon
dioxide contamination of the electrolyte is a problem but may be dealt with by changing
the electrolyte occasionally, perhaps with the same frequency as oil is now changed in a
gasoline engine.
Realistic estimates place the cost of the General Atomic battery in the range from
about $1.50/lb for one designed with good energy storage but only limited power capability
to about $3.30/lb for one with the highest power density of which the system is capable. At
present, the major objective of General Atomic's project is the electric lift truck application.
b. Zinc-Air (Leesona)
The Leesona Corporation is also very active in metal-air battery development. While
much of its effort is in the primary zinc-air field, it is also investigating electrically
rechargeable secondary battery systems. It has developed a mechanically rechargeable
zinc-air battery, which is now being delivered to the Army and Marine Corps. The latter
battery, which is designed as a power source for military communication systems, is far too
costly to consider as a vehicular power source. Even if anode costs were simply those of the
materials, with a credit for the discharged anode, mechanically rechargeable systems would
still appear to be several times too expensive for consideration in vehicles.
49
ZD.lUttleJnr.
-------
The work on secondary zinc-air batteries at Leesona is based on configurations
which are similar to those of the company's primary metal-air batteries. Much of its effort is
directed at improving separators, both organic and inorganic; Leesona is a licensee of the
inorganic separator developed by the Astropower Laboratory of McDonnell-Douglas, Inc.
(see below). Leesona is also investigating the effect of close control of charging current
density and pulse charging in attempts to solve the problem of dendrite growth.
The Leesona primary zinc-air battery delivers as much as 116 w-hr/lb in radio service
and can sustain steady discharge at a power density of 18 w/lb. Such favorable parameters
are unlikely to carry over to electrically rechargeable secondary systems, where good cycle
life is possible only when the battery is not completely discharged.
The future costs of Leesona's electrically rechargeable zinc-air battery, manu-
factured in quantity, can only be guessed, because so much depends on the quantities and
costs of catalysts required for the cathode and the degree of complexity in the battery
structure; a figure of $1.50-S2.00/lb might be reasonable. It is generally agreed that the
technical problems associated with such cells are far from being solved and their intro-
duction lies in the indefinite future.
c. Sodium-Air
An interesting variant on metal-air batteries — one which permits the use of an alkali
metal anode - is the sodium-air battery being developed by the Atomics International
Division of North American Rockwell Corporation. Each cell of this battery is essentially
two cells: one is a sodium-sodium amalgam cell with a molten salt electrolyte, and the other
is formed by the sodium amalgam with an air cathode in an aqueous sodium hydroxide
electrolyte. The system is made feasible by the development of a very low-melting eutectic
sodium salt mixture as the electrolyte for the sodium-sodium amalgam cell, in which the
sodium ion is the sole current-carrying species. Because of the low melting point the
temperature of operation is not much above 100°C, which is quite compatible with the
operation of the sodium amalgam-air cell.
The sodium-air battery is still at an early stage of development. Experience with it is
limited to prototype models which generate about 25 watts and which have completed
about 50 operating cycles. The air cathode, which has not been optimized, is a modified
form of an American Cyanamid electrode containing about 10 milligrams of platinum per
square inch. The relatively low temperature of operation permits the use of plastics, partic-
ularly polypropylene, as materials of construction. The molten salt electrolyte is contained
in a porous alumina disc in some designs and is unconstrained in others.
Projections based on extrapolations from early results indicate that energy densities
for a 20-kw battery would be in the range of 100-150 w-hr/lb and power densities would be
25-35 w/lb, with ultimate capabilities significantly higher than these figures. The power
densities are, however, based on current densities that may necessitate uneconomically large
quantities of platinum in the air cathodes.
50
21.lUttle.3nr.
-------
In addition to the platinum, mercury is also a major material cost. Overall manu-
facturing costs for the battery on a large scale are estimated at being in the $1.50 to
S2.00/lb range, which is similar to those projected for many of the high-temperature bat-
teries. The cost related to the engineering complexity of the sodium-air cell is offset by the
use of economical plastic materials in its construction.
From the operational standpoint the sodium-air concept has a number of attractive
features for possible application as a vehicle power source. It has a much lower temperature
of operation than the other alkali metal-molten salt systems, and its power density is likely
to be higher than those of the organic electrolyte batteries. Moreover.it retains the molten
anode feature and thus avoids the difficulties of dendrite formation on recharge. A further
consequence of the molten and thus structureless anode is that mechanical rechargeability is
more likely to be economically feasible than with other types of metal-air battery.
On the negative side, the cell appears to be rather complex,and operating experience
with it is limited. The long-term stability of the low-melting eutectic mixture may be a
questionable feature, but the self discharge rate at the sodium amalgam-sodium hydroxide
interface is apparently low.
Atomics International has recently formulated a design in which the two cells of the
battery are constructed as separate entities and the sodium amalgam is pumped between
them. While this would require somewhat greater quantities of mercury, cell construction
would be simplified, reliability would be enhanced, and it is predicted that overall costs
would be reduced.
d. Other Metal-Air Batteries
In addition to the work reviewed above,there is a great deal of other activity in the
field of rechargeable metal-air batteries. On the basis of the little information that has been
published, none of it appears to be at a more advanced stage than the programs already
described. Union Carbide and ESB, Inc. are working on electrically rechargeable zinc-air
systems generally similar to that of Leesona. Yardney Electric Corporation, in a program
partly sponsored by the Ford Motor Company, is developing its"Rotoxel"zinc-air cell in
which dendrite formation at the zinc anode is prevented by rotating the zinc anode against a
doctor blade; this cell uses silver catalysts at its air cathode. The Astropower Laboratory of
McDonnell-Douglas Corporation is also investigating rechargeable zinc-air cells, making use
of its novel inorganic separator developed originally for heat sterilizable silver-zinc batteries;
this separator is reported to achieve a significant reduction of dendrite growth on zinc
during recharge.
Globe-Union, Inc., in a program sponsored by the U.S. Army Electronics Command,
is investigating a lithium-moist air battery in nonaqueous solvents. While this might rep-
resent the ultimate in theoretical energy density for metal-air cells, the practical difficulties
appear to be overwhelming.
51
21.ltittlc.Ilnr.
-------
General Telephone and Electronics has a program on the iron-air battery. In view of
the maturity of the technology of rechargeable iron electrodes, this might be a promising
approach.
Zaromb Research Corporation is investigating aluminum-air cells, but in spite of
their attractive theoretical energy densities, they are relatively unpromising. Less than half
the theoretical voltage is obtained, leaving the rest to appear as heat. As the aluminum
electrode cannot be recharged electrically, the system can be considered potentially re-
chargeable only on a mechanical basis. The General Electric Company has developed a
mechanically rechargeable magnesium-air battery conceptually similar to the Leesona zinc-
air system but has not advanced it as a possible vehicle power source. General Electric is also
investigating rechargeable cadmium-air systems, as is Union Carbide. These latter systems,
however, are not being considered as potential power sources for vehicles.
4. Nickel-Zinc
While the high level of present interest in this system is not directly related to
electric vehicle development, the nickel-zinc battery has, potentially at least, performance
parameters which merit consideration for vehicle applications. The substitution of zinc for
cadmium should give it approximately double the energy density of the nickel-cadmium
system (i.e., about 30 w-hr/lb), and the power density should be comparably high. There is
no problem of material availability with zinc; the costs of the battery should be appreciably
less than those of nickel-cadmium batteries, although as a nickel-alkaline system it is still
likely to be expensive compared with lead-acid.
The major difficulty with the nickel-zinc system is the classic problem of obtaining
an adequate cycle life from any rechargeable battery containing a zinc anode. General
Telephone and Electronics, which is one of the companies investigating nickel-zinc cells, has
reported cycle lives of 100-200 but at only 50% discharge, which nullifies its advantage over
nickel-cadmium in energy density. General Electric has reported improved cycle life by
incorporating calcium hydroxide in the electrolyte to limit the concentration of soluble
zincate, but this too lowers the energy density. Also active in the field are Texas Instru-
ments, Eagle Richer, Yardney Electric, and others.
Even if perfected, the nickel-zinc battery would still seem to be too costly as the
prime power source for an electric vehicle, but it might have some promise in hybrid
mechanical-electrical systems (see Section VI). Until a reasonable lifetime has been demon-
strated, however, its prospects as a vehicle power source remain speculative.
52
Arthur ZD.lUttkllttr.
-------
C. FUEL CELLS
Fuel cells are essentially steady-state primary batteries and are of interest for electric
vehicle applications mainly because vehicles so powered would have no requirement for
electrical recharging. They would thus be no more restricted in range than are convention-
ally powered vehicles. Because of this, energy density is not a meaningful parameter for fuel
cells; an evaluation of their applicability to vehicles turns upon their power density, relia-
bility, and potential cost.
Fuel cells, which exist in a wide variety of types, can be differentiated by the nature
of the fuel used - hydrogen; water-soluble substances such as hydrazine, methanol and
ammonia; hydrocarbons — and by their temperature of operation: ambient, moderately
elevated (80-150°C), or high (500-1 OOCfC). Since hydrogen reacts so easily in a fuel cell
but is an inconvenient fuel to handle, several types of fuel cells convert other fuels into
hydrogen as the first step. These are called indirect fuel cells. The types of greatest interest
for vehicle application are ambient-temperature hydrogen systems, both direct and indirect,
and soluble fuel systems. On a conceptual basis the U.S. Army has also considered a
high-temperature fuel cell-battery hybrid system involving a high-temperature hydrocarbon
fuel cell.
1. Ambient-Temperature Hydrogen Systems
Ambient-temperature hydrogen fuel cells have reached a relatively high degree of
reliability, and several thousand hours of trouble-free operation are now commonplace. The
major limitations of these systems for application in electric vehicles are their relatively low
power density and their high estimated production costs. As a general rule, restrictions on
current densities imposed by problems of heat dissipation and electrode polarization limit
power densities to levels of around 30 w/lb or so. More serious is the cost restriction, which
has two major causes: 1) precious metal catalysts are necessary to generate acceptable
current densities and 2) the problems controlling materials transfer across the boundaries of
fuel cells make the engineering problems of the latter more complex than those of batteries.
The present state of the art in ambient-temperature hydrogen-air fuel cells is such
that cells can be operated with no precious metal catalyst at all, but the current densities
would be too low for vehicle applications. A loading of about 5 milligrams of platinum per
square centimeter of electrode presently appears necessary to give a power density of 25
w/lb to a hydrogen-air fuel cell and this may be improved to 50 w/lb in ten years.* Present
costs of such cells are no reliable index of future costs, since they are all essentially hand-
made. The most optimistic estimates for future volume production are in the range of
* The Automobile and Air Pollution: A Program for Progress, Report to the Panel on Electrically
Powered Vehicles (R. S. Morse, Chairman), U. S. Department of Commerce, October 1967.
53
3rrbur
-------
$100-5150 per kw. To this must be added the cost of the fuel supply system. Cylinder
hydrogen is probably impractical due to its bulk and weight; liquid hydrogen is a distinct
possibility but will certainly add significant costs for cryogenic storage. Indirect fuel cells
would require equipment to reform hydrocarbons or methanol or to crack ammonia. None
of these possibilities appears likely to involve the addition of less than S50/kw to the first
cost; and thus, S150-$200/kw seems to be a limiting cost range for such systems.
Operational costs are likely to be quite moderate if hydrocarbons or methanol are
used as fuel, since the conversion efficiency of the fuel cell is much higher than that of the
internal combustion engine. The extent to which there would be a road tax on fuel-cell-
powered vehicles is a major unknown factor here. In addition, fuel cells with alkaline
electrolytes would need occasional changes of electrolyte because of gradual contamination
with atmospheric carbon dioxide.
2. Soluble Fuel Systems
Soluble fuel types will in general involve even higher first costs than will hydrogen
fuel cells. Those using ammonia or methanol call for much higher loadings of platinum
catalysts per unit area, and their power density is not as great. Relatively low costs would be
associated with a hydrazine-fed fuel cell, but this material is presently too costly (hydrazine
hydrate costs approximately Sl.OO/lb) and there seems little prospect for a great enough
cost reduction to make it a practical fuel.
3. Direct Hydrocarbon Fuel Cells
Research activity continues with direct hydrocarbon fuel cells, using aqueous elec-
trolytes, particularly phosphoric acid, at temperatures of about 150°C. While the reactivity
of the hydrocarbons is higher than was earlier thought possible, the quantities of precious
metal catalyst required are still, and are likely to remain, orders of magnitude higher than
those which could be contemplated in an electric vehicle.
4. Platinum Availability
The use of platinum in fuel cells is potentially restricted not only by cost but also by
availability. On the basis of today's technology, a full sized car powered by a fuel cell might
have a platinum requirement in the range of 1 to 10 pounds. This should be compared with
an annual world production of this element of about 60,000 pounds and an estimated
reserve of 10 million pounds. Clearly, the world's present apparent resources of platinum are
not sufficient to support the widespread use of fuel cells in vehicles. Considerable research
effort is being expended in a search for alternative and noncritical catalyst materials, but the
results to date are not particularly promising.
54
3rthur 2l.1Uttle.llnr.
-------
High-temperature fuel cells capable of utilizing hydrocarbons directly do not use
precious metal catalysts and hold promise of a lower ultimate cost structure than those for
other types of fuel cells. They are, unfortunately, excessively heavy, and the problems of
reliability and control are exacerbated by the elevated temperatures.
5. Overload Capability
A consideration of the fuel cell's prospects solely in terms of cost per kilowatt may
lead to an excessively pessimistic outlook, since some allowance should be made for its
overload capabilities. Power rating is a term calling for close definition; if the demand for
maximum power is infrequent, a much lower rating is possible for the fuel cell than would
be possible with a battery power source. For example, Union Carbide's hydrogen-oxygen
fuel cell for the GM Electrovan* had a rating of 32 kw but a maximum output of 160 kw.
The exact degree and duration of overload possible with a given fuel cell system will vary
with its individual peculiarities, but it seems reasonable to take a factor of three as being
representative. In such a case the family car of Table 1 might be adequately powered by a
30-kw cell rather than one of 90 kw. Even so, at an estimated S200/kw, the fuel cell would
be prohibitively costly.
6. Present Applications
While much effort is being expended in the United States and elsewhere on fuel cell
development, relatively little is focused on highway vehicle applications, most probably for
the reasons of cost. Major engineering programs at General Electric. Pratt & Whitney Divi-
sion of United Aircraft, Allis Chalmers, Union Carbide, Texas Instruments, and Monsanto
have been concerned more with the fuel cell's application as a power source for space
satellites, both manned and unmanned, as a means of generating electricity from natural gas
in the home (a program at Pratt & Whitney in which the Institute of Gas Technology also
participates), as a power source for the military in the field, and as a power source for
industrial vehicles, particularly fork lift trucks. Overseas, ASEA in Sweden has built fuel
cells for submarine propulsion. (This and other European activities are reviewed in Appen-
dix B.)
Direct application of the fuel cell to highway vehicle propulsion has been confined
to two experimental vehicles: the GM Electrovan, powered by a Union Carbide hydrogen-
oxygen cell, and an M-37 truck modified by the U.S. Army Engineers R&D Laboratory to
use a Monsanto hydrazine-air cell. Some notes on these two vehicles are contained in
Appendix A. In Europe the French Government is sponsoring an electric car program based
upon a 25-kw hydrogen-oxygen fuel cell system (see Appendix B).
See Appendix A.
55
3rthur 3l.HittIc.3nr.
-------
D. ELECTROCHEMICAL HYBRIDS
Many of the electrochemical power sources reviewed in this section have had certain
favorable properties for the electric vehicle application, but some of their other properties
might be unacceptable. The most common situation, exemplified by fuel cells and alkali
metal-organic electrolyte batteries, is good energy density allied with inadequate power
density. One solution to this difficulty is to use a combination or hybrid power source in
which each component provides properties which complement those of the other. Thus, a
low-energy-density, high-power-density battery such as nickel-cadmium can be combined
with a high-energy, low-power battery or fuel cell.
There are a number of fairly apparent drawbacks to electrochemical hybrid systems.
First, they are more complex than a single power source and thus require significantly more
complicated control equipment. Second, the energy efficiency of the hybrid tends to be low
because of the losses involved in the internal redistribution of charge among the components
of the power source. A further difficulty is that, future high-temperature systems aside, the
most suitable battery for the hybrid systems is nickel-cadmium which has the limitations of
high cost and materials availability problems, previously noted. As mentioned elsewhere in
Sections VI and VII, the development of high-power-density "pile-type" lead-acid batteries
would improve the prospects of hybrid power sources.
In spite of these difficulties, electrochemical hybrid systems offer enough advantages
to have been the subject of a number of detailed paper studies and the power source for at
least one experimental vehicle. There are two subcategories fuel cell-battery hybrids and
battery-battery hybrids.
1. Fuel Cell-Battery Hybrids
Fuel cell-battery hybrids offer possibilities for improving the prospects of the fuel
cell as a vehicle power source on two counts. The battery can be chosen to give the overall
system an adequate power density and, since its specific costs are likely to be lower than
those of the fuel cell, the overall cost per kilowatt will be lessened. The system retains the
fuel cell's advantages of rapid refueling and a range limited only by the size of the fuel tank.
Union Carbide Corporation, which built the hydrogen-oxygen fuel cell for the CM
Electrovan,* has carried out a conceptual study of coupling a fuel cell with nickel-cadmium
batteries for the same application. It concluded that the same performance could be
obtained by reducing the fuel cell stack weight from 1345 to 690 pounds and incorporating
427 pounds of nickel-cadmium batteries. Further weight savings would be effected in other
* See Appendix A.
56
ZD.1Uttle.Knr.
-------
parts of the system. The study was not intended to find the optimum makeup of the system
but was merely a first approach based on the present state of the art. The battery would be
capable of putting out 128 kw of power for short periods; based on a rule of thumb of $3/lb
for nickel-cadmium battery costs, it would give peak power capability for about $10 per kw,
considerably less than could ever be expected from a fuel cell.
Union Carbide has also constructed for demonstration purposes a small electric
motorcycle powered by a hybrid combination of hydrazine-air fuel cell.and nickel-cadmium
batteries. The power rating of the fuel cell in the system was 0.8 kw.
Another approach to fuel cell-battery hybrid power sources is being studied at the
U.S. Army Engineers R&D Laboratories at Fort Belvoir. This concept would combine a
high-temperature hydrocarbon-air fuel cell of the type being developed at Texas Instru-
ments, Inc., with a high-temperature alkali-metal battery (for which the SOHIO system is a
likely candidate). The advantages include the ability to use conventional fuels directly, a
high power density due to the properties of the battery, and the possibility of using waste
heat from the fuel cell to maintain the temperature of the system, thus eliminating the need
for a separate heater.
The high-temperature fuel cell-battery hybrid has certain logistic attractions for the
military because it is designed to operate on standard hydrocarbon fuels, but it seems a
rather remote possibility for general application. If the technical problems of their develop-
ment can be solved adequately, high temperature batteries are likely to be the sole power
source in a vehicle. The fuel cell would add increased range capabilities but at a higher first
cost and much greater complexity.
2. Battery-Battery Hybrids
The most prominent example of the battery-battery hybrid is a system proposed by
Gulton Industries, Inc., presumably to be the power source for the experimental vehicle be-
ing developed in Gulton's joint program with American Motors Corporation. It will consist
of a high-energy-density lithium-nickel fluoride battery (see Section V-B-2) in combination
with a high-power-density "bipolar" nickel-cadmium battery. The latter has a pile-type
construction and has been developed by Gulton specifically for applications involving high
discharge rates. The hybrid system is expected to combine the best features of both bat-
teries.
The General Electric Company has used a hybrid battery system in its small experi-
mental electric vehicle (see Appendix A). The major energy source is a lead-acid battery; to
enhance the vehicle's acceleration, a small nickel-cadmium battery having about 5% of the
capacity of the lead-acid battery is connected in parallel with it. The system apparently
operates successfully, and the vehicle has significantly better performance than it would
have with lead-acid batteries alone.
57
-------
VI. MECHANICAL-ELECTRICAL HYBRIDS
The power source for an electric vehicle can, of course, be an electrical generator
activated by some kind of combustion engine. This is the basic system used in diesel-electric
locomotives and, to a limited extent, in large diesel-powered off-highway vehicles for heavy
construction work. Combustion engines are inevitably a source of some pollution, but it is
reasonable to expect that the pollution could be minimized if they were specially designed
to operate at a constant power output, as would be possible in a hybrid system. The
fluctuating power demand would be met by a battery connected in parallel with the
generator. The system might take any one of the four forms shown in Figure 10.
Hybrid systems can, in principle, have any desired ratio of engine-generator to
battery power. In practice, however, it would seem desirable to have the engine-generator
capable of supplying the average power level used by the vehicle plus sufficient reserve to
ensure that the battery never became fully discharged in normal driving situations. This
would make the hybrid vehicle independent of battery charging facilities and would
eliminate from consideration engines which are uneconomically small,* since the power/cost
ratio of engines decreases significantly as the size falls to a few kilowatts.
An engine much larger than the average power rating is undesirable, however, since if
it were being run at constant power, much energy would be wasted in overcharging the
battery. Shutting off the engine when the battery is fully charged is not a solution to this
difficulty, because the vehicle could not then respond rapidly to a demand for power greater
than could be supplied by the battery alone.
The driving cycle used in the calculations of vehicle performance parameters (Table
2) indicates that the average power demand in the moving vehicle is about 24% of the peak
demand. This may be somewhat on the low side for vehicle performance as a whole, since
the cycle is concerned with city driving conditions. We have therefore arbitrarily chosen to
consider the prospects for mechanical-electrical hybrids on the basis of the engine's having a
continuous power output equal to 40% of the peak power demand. This would provide a
good reserve for unusually prolonged operation at high power demand and also permit the
use of power-consuming auxiliaries such as air conditioning.
Of the four possible systems shown in Figure 10 we consider the ac generator-dc
motor combination (d) the most immediately promising. The ac generator is attractive
because it has no brush maintenance and replacement problems, and because it can run at
high speeds and is thus very compact. Moreover, the technology of mass producing such
* The few hybrid vehicles in commercial production for use on the highway utilize a relatively small engine-
jenerator system manufactured by G&M Power Plant Co. Ltd., of Ipswich, England. Designed for attach-
ment to electric milk trucks and similar vehicles, the engine-generator serves merely to extend the range of
what are essentially battery-powered vehicles.
58
3rthur
-------
(a)
Engine
\\\\\\N
Speed
Increaser
0\\\\V
D-C
Generator
Battery
1
Choppers
D-C
Motors
(b)
Engine
\\\\\\N
Speed
Increaser
s\\\\\\
D-C
Generator
Battery
1
Inverters
A-C
Motors
(c)
Engine
S\\\\
Speed
[\\\\\
A-C
Generator
Rectifier
Battery
1
Inverters
A-C
Motors
(d)
Engine
X\\\>
Speed
Increaser
,\\\V
A-C
Generator
1 Battery
Rectifier
1
Choppers
D-C
Motors
FIGURE 10 VARIOUS ELECTRICAL SYSTEMS FOR
ELECTRICAL-MECHANICAL HYBRIDS
-------
generators is well advanced because of the general introduction of the alternator in
conventionally powered vehicles some five or six years ago. A dc motor is preferred,
however, for traction; as explained in Section IV, the principal reason is the high cost and
weight of the inverter system needed by an ac motor for speed control.
To evaluate the technical requirements of the hybrid system, we follow the same
general procedure as that used for the purely electrochemical power source. From the
weight available for the power sources (line 17 of Table 1), we subtract the weight of an ac
generator, rectifier, speed increaser, and voltage regulator rated at 40% of the final peak
power requirement for each vehicle. The remaining weight can then be distributed between
the prime mover and the battery. Our approach has been to estimate weights for the former
based upon existing technology and then calculate the power density which is required from
the battery. Provided it is sufficient to meet the demand for maximum power, the energy
density of the battery is of secondary importance, since the source of energy is the fuel fed
to the prime mover.
Five different classes of engines could conceivably be employed in a mechanical-
electrical hybrid system. Three of these - the gas piston engine, the diesel piston engine,
and the gas turbine engine - involve internal combustion; the others - the Rankine
cycle (steam) engine and the Stirling engine — involve external combustion. The latter two
can be operated at lower overall pollution levels than the former, since external combustion
can be carried out in a comparative excess of air (although oxide of nitrogen production is
enhanced); nevertheless, the data available on steam engines and Stirling engines in the size
range of interest (reviewed in the recent report* by the Battelle Memorial Institute) are so
meager that a separate analysis of their prospects in hybrids is not justified. Since their
power/weight ratio appears to fall within the same range as the diesel piston engine, the
calculations for this engine can be used as a first approximation for them.
Representative data for the technical performance of gas piston engines and diesel
piston engines in the size range of interest (6-60 kw) are plentiful, and some data exist for
the gas turbine at the upper end of this range. Figure 11 shows some estimates of the
weight-to-power ratio for these types of engines, and Figure 12 shows estimates of the
corresponding factory costs per kilowatt. These data are for engines with anticipated life
times of 2500 operating hours, manufactured in quantities greater than a hundred thousand
units annually and designed for constant power service at their rated load in a hybrid
vehicle. Their parameters therefore correspond more closely to the engine technology of
long-haul truck service and industrial applications, where average operation is close to the
rated power level, rather than to that of conventional automobile engines, whose average
operation calls for only 20-25% of the maximum rated power. For gasoline and diesel
engines, which are very mature products, the power rating is the main influence on the
power/weight ratio, and production volume is the main determinant of costs.
The technology of gas turbines is well advanced, but the translation of this
technology into economic solutions to the problem of automotive propulsion is in its
infancy. Few units exist in the size range of interest, and the estimated values in Figures 1 1
and 12, particularly of cost, must be regarded as very approximate.
* Battelle Memorial Institute, "Study of Unconventional Thermal, Mechanical, and Nuclear Low-Pollution-
Potential Power Sources for Urban Vehicles," by J. A. Hoess, eta/.. Summary Report to U. S. Dept. of
Health, Education and Welfare, March 15, 1968.
60
3l.lLittlf.Hnr.
-------
100
I
•5 10
10
Rating (kw)
100
FIGURE 11 POWER/WEIGHT RATIOS FOR INTERNAL COMBUSTION ENGINES
61
-------
70
60
50
I 40
o
Q
30
CO
20
10
Diesel
Piston
Gasoline
Piston
10
Continuous Rating (kw)
Gas
urbine
\
\
SI
100
FIGURE 12 APPROXIMATE COST LEVELS FOR
INTERNAL COMBUSTION ENGINES
62
-------
The weight estimates for the ac generator are taken from Figure 13. These
correspond to machines running at speeds of 12,000 rpm in the higher end of the power
range and as high as 16,000 rpm in the lower end. They are representative of the technology
used in equipment which is already in existence, particularly for aircraft applications, and
which might be manufactured at low unit cost if sufficient demand developed. (Figure 13
also makes allowance for the relatively small extra weight associated with the rectifier and
the speed which the system requires.)
Using the data from Figures 11 and 13, the weights of generator and engine rated at
40% of the peak power requirement have been calculated for each of the six classes of
vehicle and for each of the three engine types. The permissible weight for the battery and
thus its required power density are derived from these calculations and are presented in
Table 5. The first two lines of the table are, respectively, lines 12 and 17 of Table 1.
The results of the calculations, which at first sight are somewhat surprising, indicate
that the power density requirements for the batteries in hybrid vehicles are only slightly
lower than for those in vehicles powered solely by batteries. These requirements are not
particularly sensitive to the arbitrary choice of a 40/60 prime-mover-to-battery ratio. Thus,
among vehicles with gasoline engines and with conventional construction the family car
shows essentially no change, while in the most favorable case, the city bus, the reduction is
above 30%, from 36 w/lb to 25 w/lb. In the case of the diesel engine, the power density
requirement for many hybrid vehicles is actually greater than when the battery is the sole
power source; this results from the relatively high weight which must be assigned to the
diesel engine.
Although the engines which would be used in the hybrid are smaller than those used
in corresponding conventionally powered vehicles, the weight differences are not great. The
larger maximum power capability of conventional vehicles does not call for substantially
greater weight, because the engine is not built to run steadily at the higher power level. Also,
the power density of internal combustion engines decreases as the size falls, so a small
decrease in average power level does not return a proportional decrease in weight. The
decrease in battery power density requirement for the hybrid involving a gas turbine engine
is somewhat greater than for the two piston engines, but unfortunately the technology of
small gas turbines is not well established, and their cost, which might be in the range of
$40-$ 100 per kilowatt in the size range of interest, is likely to be discouragingly high.
The major differences between the battery requirements in the pure electric and the
hybrid is that, in the latter, 1) energy density is of small importance and 2) the maximum
power density is reduced by about one-fifth. The power density requirements for the family
car, however, still restrict the choice of batteries primarily to the high-temperature alkali
metal type (among which the SOHIO system's high power density capability might be
particularly attractive); the sole exception to this statement is the nickel-cadmium battery,
which has excellent power characteristics but low energy density.
For vehicles other than the family car, the power density requirements of the hybrid
could be satisfied by metal-air battery systems; for various technical reasons, however (such
as corrosion at the air cathode), this class of battery does not appear to be suitable for the
63
3l.ltittlf.3lnr.
-------
£
£ 100
o>
i
10
a
X
/
/
<
/
/
'
A
/
'
S
/
A
^
/
r
X
/
-
'
/
4 6 8 10 20 40 60 80 100 20
Generator Rating (kw)
FIGURE 13 WEIGHT OF GENERATOR, RECTIFIER,
SPEED INCREASER AND VOLTAGE
REGULATOR
64
-------
TABLE 5
HYBRID VEHICLE SPECIFICATIONS3
Maximum output of hybrid
power source
(kw)
Weight assignable to hybrid
power source (Ib)
Conventional construction
Lightweight construction
Weight assigned to generator,
rectifier, speed increaser, and
voltage regulator (Ib)
Weight assigned to engine (Ib)
Gas piston
Diesel piston
Gas turbine
Weight available for battery (Ib)
Conventional construction
Gas piston
Diesel piston
Gas turbine
Weight available for battery (Ib)
Lightweight construction
Gas piston
Diesel piston
Gas turbine
Power density of battery (w/lb)
Conventional construction
Gas piston
Diesel piston
Gas turbine
Power density of battery (w/lb)
Lightweight construction
Gas piston
Diesel piston
Gas turbine
Family
Car
85
902
1,402
70
279
715
89
553
117
743
1,053
617
1,243
92
440
69
48
83
41
Commuter
Car
29
632
932
30
150
395
44b
452
207
558b
752
507
858
38
82
30
23
34
20
Utility
Car
17
418
618
20
105
295
32b
293
103
366b
493
303
566
34
97
27
20
33
18
Delivery
Van
62
1,141
1,741
55
233
600
72
853
486
1,014
1,453
1,086
1,614
43
76
36
25
34
23
City
Taxi
47
1,040
1,540
44
196
512
63
800
484
933
1,300
984
1,433
35
58
30
22
28
20
City
Bus
159
4,385
6,385
115
416
1,010
135
3,854
3,260
4,135
5,854
5,260
6,135
25
29
23
16
18
15
a) 40% of power supplied by prime mover, 60% supplied by battery
b) Power rating for commuter and utility car application considered to approach lower limits of feasible ap-
plication of gas turbine.
65
21.lLtttIr.Ilnr.
-------
"float charge" service characteristic of the hybrid application. More important is the
possibility that the lead-acid battery might qualify for the non-family-car hybrid
applications. The power requirements are somewhat high for it, except for the city bus, but
its power density might be increased with further development of pile-type construction.
From a cost standpoint, the prospects for the hybrid vehicle are generally
unfavorable. In the smaller vehicles there would be virtually no saving in the cost of the
prime mover to offset the extra cost of the generator, electric motors, control system,
battery, cabling, etc. of the hybrid. The nickel-cadmium battery, whose present technical
capabilities make the hybrid vehicle an immediately feasible proposition, is too expensive at
$3/lb and could not be manufactured in the required quantities because of insufficient
supplies of cadmium. Although they obviously face a prolonged period of development,
high-temperature alkali metal batteries could eventually prove feasible for hybrids; however,
their energy density would make them even more attractive as the sole power source of a
purely electric vehicle, which is inherently a less complex and thus less costly solution.
There appears to be one possible exception to these negative conclusions. In the case
of the city bus, the largest of the vehicle types considered, existing lead-acid batteries have a
power density adequate for the hybrid. In the diesel-powered version, the cost of the
battery might be about S3000-S4000, an acceptable figure in comparison with average total
costs of about $30,000 for conventional full-sized city buses. The slight reduction in the size
of the engine would produce a small saving to offset the cost of motors and controls in this
size range, but the economic attractiveness, if any, would probably come from lower fuel
and maintenance costs, which are always important in commercial/industrial applications.
Enthusiasm for the hybrid city bus is, however, tempered by these facts: 1) it is still
basically an internal-combustion-powered vehicle and will contribute to air pollution, and 2)
the number of buses is relatively small, so the elimination of their pollutants would make
only a small contribution to solving the overall problem.
66
Arthur 31.lUttlc-.3lnr.
-------
VII. PROSPECTS AND REQUIREMENTS FOR ELECTRIC VEHICLES
This section reviews the prospects for the six classes of electric vehicles, mainly in
terms of the ability of the projected state of the art in electrochemical power sources to
meet their technical requirements within an acceptable level of cost. The future technology
of motors and controls can be projected with somewhat greater confidence, although cost
estimates are based on large extrapolations and are, at best, very approximate. Some con-
sideration is also given to the problems relating to the refueling of electric vehicles.
The findings of this study indicate the technical areas where most effort is necessary
if the development of electric vehicles is to be accelerated. From these findings, and taking
into account the anticipated progress of existing research and development programs in both
the public and private sectors of the economy, a number of recommendations are made
concerning the emphasis and direction of future effort.
A. TECHNICAL PROSPECTS
The general technical capabilities of the various electrochemical power sources con-
sidered in Section V are summarized in Table 6 on the basis of values that might be attained
when the systems now under development reach maturity. While these generalized figures
may not accurately apply to some individual systems, they will suffice for a first analysis of
vehicle requirements.
1. Family Car
Conventional Lightweight
Battery or Gas Piston Battery or Gas Piston
Requirements: Fuel Cell Hybrid Fuel Cell Hybrid
Energy Density (w-hr/lb) 135 - 87
Power Density (w/lb) 94 92 60 48
The technical requirements for the power source of the family car are the most
demanding of those for all six types of vehicles considered. This is somewhat discouraging,
since vehicles in this general category comprise approximately 80% of all vehicles on U.S.
highways and are thus responsible for a correspondingly large proportion of the air pollution
caused by vehicles. It is apparent from the requirements shown in Table 1 that only the high
temperature alkali metal batteries have any prospects of simultaneously satisfying the
energy density and power density needs of this vehicle, regardless of the kind of construc-
tion. While a number of the metal-air batteries might, with further development, meet the
requirements for energy density in the family car with lightweight construction, they cannot
meet the power density requirement. The same is true for fuel cells, and particularly so for
organic electrolyte batteries.
67
31.lLittIf.3lnr.
-------
TABLE 6
GENERALIZED TECHNICAL PARAMETERS FOR
ELECTROCHEMICAL POWER SOURCES
System
High temperature alkali metal batteries
(rechargeable)
Maximum Energy
Density
(w-hr/lb)
> 100
Maximum Power
Density
(w/lb)
> 100
Metal-air batteries
(rechargeable)
50-80
30-40
Alkali metal batteries with organic
electrolytes (rechargeable)
75-100
15-20
Lead-acid
Special construction
15-20
20-30
60
Nickel-cadmium
Special construction
15-20
75-100
300
Fuel cells
30-40
68
ZD.lUttlo.Dnr.
-------
As shown in Section VI, the power density requirements for a mechanical-electrical
hybrid in a family car are not much lower than for the purely electrochemical power source.
They can, of course, be satisfied by high-temperature batteries and also by the nickel-
cadmium battery, but the costs of the former are uncertain and those of the latter are
known to be too high. Pile-type lead-acid batteries, if their further development is success-
ful, might also be marginally capable of meeting the hybrid power source requirements for
the family car.
Among electrochemical hybrid systems the power density of air-breathing fuel cells
is likely to be too low to allow the hybrid to provide a sufficiently high power density, even
in combination with high-power-density batteries. The high energy density of the lithium-
organic electrolyte battery in hybrid combination with a high rate nickel-cadmium battery
could satisfy the technical requirement; this is unlikely to be a suitable power source,
however, because of the particularly formidable technical problems associated with the
development of rechargeable lithium-organic electrolyte batteries and the unavoidably high
cost structure that this system would entail.
Thus, with the marginal exceptions noted above, it can be concluded that high-
temperature alkali metal batteries are the only electrochemical power sources which appear
technically suitable for use in the family car.
2. Commuter Car
Conventional Lightweight
Battery or Gas Piston Battery or Gas Piston
Requirements: Fuel Cell Hybrid Fuel Cell Hybrid
Energy Density (w-hr/lb) 41 - 28
Power Density (w/lb) 46 38 31 23
Power source requirements for the commuter car are less stringent than for the
family car. Besides being a lighter vehicle and having lower acceleration capability, it has
only one-half the range; as a consequence, the energy density requirement is about one-
third, and the power density requirement is about one-half, those of the family car. High-
temperature alkali metal batteries should be able to meet these requirements with ease.
Metal-air batteries would have adequate energy density capability, but the maximum power
density required in the commuter car with conventional construction is somewhat above the
upper range of the expected capability of devices using air cathodes. With lightweight
construction, the power source requirements should fall easily into the range of capabilities
of metal-air batteries and also of fuel cells.
Among hybrid power sources, a high-power-density battery in combination with a
fuel cell, lithium-organic electrolyte battery or an engine-generator would all qualify. The
power requirements are, however, too high for existing lead-acid batteries, and nickel-
cadmium batteries are unacceptable on the grounds of cost and material availability.
69
2D.littlc.3lnr.
-------
The commuter car thus appears also to require high-temperature alkali metal bat-
teries, but rechargeable metal-air batteries would appear to have prospects of providing an
adequate alternative.
3. Utility Car
Conventional Lightweight
Battery or Gas Piston Battery or Gas Piston
Requirements: Fuel Cell Hybrid Fuel Cell Hybrid
Energy Density (w-hr/lb) 26 - 18
Power Density (w/lb) 40 34 28 20
The requirements for the utility car are the most modest of the series. Considering
first the requirements for lightweight construction, the power source needs can be met by
lead-acid batteries of the type representative of the best of today's technology and com-
fortably within what normal evolution should provide within a few years. With conventional
construction the utility car has a rather high power density requirement, close to the upper
range likely to be achieved by metal-air batteries. High-temperature alkali metal batteries, of
course, easily satisfy all requirements of the utility car.
While a number of hybrid power source systems, both mechanical-electrical and
electrochemical, are technically feasible power sources for the utility car, their complexity
makes them inappropriate for such a small vehicle.
Thus, improved lead-acid batteries would seem to be technically adequate for this
vehicle application; the higher performance metal-air batteries and high-temperature alkali
metal batteries offer potential alternatives whose merit might be determined on the basis of
cost.
4. Delivery Van
Conventional Lightweight
Battery or Gas Piston Battery or Gas Piston
Requirements: Fuel Cell Hybrid Fuel Cell Hybrid
Energy Density (w-hr/lb) 50 - 33
Power Density (w/lb) 55 43 36 25
70
3rthur 2D.1itrir.3nr.
-------
The requirements for the delivery van with conventional construction show the need
for a relatively high power density of 55 w/lb. This is likely to be above the developed
capabilities of the metal-air battery systems and of fuel cells. As in the case of the family
car, it would appear that the delivery van will require high-temperature alkali metal batteries
as its power source. In a hybrid with a gas piston engine, the power density requirement for
the battery falls to 43 w-hr/lb, but this is still attainable only by the high-temperature and
nickel-cadmium batteries.
With lightweight construction, however, the power source requirements for the van
fall into a range likely to be attained by metal-air batteries, and in the gas piston hybrid
vehicle the battery requirements might be accommodated by the lead-acid system.
While high-temperature alkali metal batteries seem in general to be the indicated
power source for the delivery van, it is evident that a much wider range of possibilities
would be opened up through the use of lightweight construction.
5. City Taxi
Conventional Lightweight
Battery or Gas Piston Battery or Gas Piston
Requirements: Fuel Cell Hybrid Fuel Cell Hybrid
Energy Density (w-hr/lb) 96 - 64
Power Density (w/lb) 45 35 30 22
For the city taxi the energy density requirement is high, since it is equivalent to that
of the family car but scaled down in proportion to its somewhat shorter range. The lower
acceleration performance demanded of the city taxi, however, reduces the power density
requirement to less than one-half of that for the family car. With conventional construction,
this reduced power density still seems to be above the expected capability range for the
rechargeable metal-air batteries, and high-temperature alkali metal battery systems again
appear necessary. In a hybrid vehicle with a gas piston engine, the nickel-cadmium battery
would be satisfactory, and improved lead-acid batteries might also meet the power density
requirements.
With lightweight construction, metal-air batteries should be capable of meeting the
power source requirements for the city taxi, although the energy density specification might
be close to the upper limit in some systems. Lightweight construction would also appear to
qualify the lead-acid system for inclusion as the battery in the gasoline-powered hybrid
system.
71
3rtbur Zn.ltittlc.3nr.
-------
6. City Bus
Conventional Lightweight
Battery or Gas Piston Battery or Gas Piston
Requirements: Fuel Cell Hybrid Fuel Cell Hybrid
Energy Density (w-hr/lb) 81 - 55 -
Power Density (w/lb) 36 25 25 15
The city bus with conventional construction has a high energy density requirement,
a consequence of the frequent stops in its driving cycle and the value assumed for the range
(which may be rather high for city driving service). The power density requirement of 36
w/lb is comparatively moderate. High-temperature alkali metal batteries are capable of
meeting both energy density and power density requirements, and the latter is also within
the range of metal-air batteries. A relaxation of the range specifications would thus allow
the latter system to be an acceptable power source for the city bus. With lightweight
construction, the power source requirements should be well within the range of capabilities
of metal-air batteries.
Hybrid systems offer some attractive possibilities for the city bus. For both conven-
tional and lightweight construction and for the three types of prime movers, power density
requirements calculated for the battery are within the range of lead-acid systems. The
hybrid bus system thus appears to be feasible using components from existing technology
which are likely to have an acceptable cost.
B. COST FACTORS
1. Considerations of First Cost
Cost estimates for the classes of electric vehicles under consideration in this study
involve even greater uncertainties than in the evaluation of the technical factors. It is clear
that no vehicle, electric or otherwise, has any prospect of becoming competitive until it is
manufactured and marketed on a scale generally comparable with that of today's conven-
tional vehicles; for passenger cars, the minimum level may be of the order of 100,000 units
annually. Even then, it seems very likely that the electric vehicle will have a significantly
higher first cost. As a broad approximation, fuel costs may be about the same for both
electric and internal-combustion-powered vehicles when allowance is made for the heavy
fuel tax on refined petroleum products.
A higher first cost may be acceptable for commercial vehicles, where the well recog-
nized lower maintenance and longer life of electric vehicles will be a partial offset. But for
commercial and especially for private vehicles it will be necessary to regard the increased
expenditure as a social cost attributable to the necessity of eliminating sources of air
pollution. In these circumstances the desirability of electric vehicles must, of course, be
evaluated in comparison with other potential solutions to this problem.
72
3rthur 2D.lttttIr.Knr.
-------
2. Family Car
Some broad perspectives on the costs of electric vehicles can be obtained by follow-
ing the same general approach as for the technical factors, namely by taking an existing
vehicle and considering which components should be replaced in the conversion from a
conventional to an electric power source. If we take the family car as an example and assign
to it a factory price of $1400, some $500 might be eliminated by removal of the engine,
transmission, gearbox, and other ancillaries directly related to the use of an internal combus-
tion power plant. Replacement of these by electric motors, SCR controllers, speed reducers,
cabling, etc., according to the specifications shown in Table 4 would, on the most optimistic
assumptions, add approximately $650 back to the cost. (This is based upon cost estimates
per vehicle horsepower of $4 for a dc commutator motor, $2 for the controller system and
$1 for cabling and miscellaneous; see Section IV.)
Thus even before we consider the cost of the power source, the factory price of the
electric vehicle would already be some 10% higher than that of a conventionally powered
car. The cost which can be tolerated for the power source is thus a rather subjective
quantity, as there are clearly no savings to demonstrate. In our opinion, an electric family
car might be justifiable if its retail price were not more than approximately $1000 over that
of a corresponding conventionally powered vehicle. Since average distribution costs should
be unchanged, this would translate into a permissible increase of $1000 in the factory price.
Such a differential might be regarded as a premium for longer life and lower main-
tenance, but for private cars, some further incentive — probably in the form of selective
taxation - would almost certainly be necessary. While $1000 is a relatively large sum today,
the period under consideration is in the 1980's at the earliest, due to the time required for
technical development and cost reduction. By then, personal affluence should have in-
creased to an extent that such additional expenditure might be much less significant.
If the allowable weight for the battery in an adequately powered family car with
conventional construction is 900 pounds (Table 1) the allowable factory cost for the battery
must be in the region of $0.95/lb. With lightweight construction the allowable cost per unit
weight of a battery just meeting specifications would be lower than this, for two reasons: 1)
it would be heavier (~1400 pounds), and 2) the use of lightweight nonferrous metals for
structural members and plastic materials for body panels is likely to increase the manufac-
turing costs, leaving less money to be spent on the battery. In this situation, to meet the
cost criteria we have set, a figure of $0.50/lb is probably all that could be tolerated.
These estimates are somewhat clouded by the question of battery replacement, since
it is unlikely that the life of the battery will be as great as that of the rest of the vehicle.
Costs amortized over a five-year period are probably low enough to be written off against
operating costs, but it must be recognized that a five-year lifetime introduces a significant
new technical requirement for the battery.
73
3rthur ZD.lUttlc.lfnr.
-------
3. Commuter Car
For the commuter car the assumed factory price might be $800, with $200 assigned
to engine, transmission, etc. Using the same basis for the costs of electric motors and
controllers as for the family car, we find that the electrical components are comparably but
slightly more expensive than the ones they replace ($210 vs $200). Thus, the cost of the
power source can be equated with the extra retail price which we have assumed can be
borne by the electric vehicle. If this is set at $500, the cost of the power source in the
commuter car with conventional construction must be about $0.75/lb. Such a battery
would have less than one-third the energy density and less than one-half the power density
of the family car's power source. The performance of a battery capable of meeting the
family car's requirements of 135 w-hr/lb and 94 w/lb would support a cost of about
$1.90/lb in the commuter car; at $0.95/lb the first cost of the vehicle would be close to that
of a conventionally powered vehicle.
4. Utility Car
Since no well-established model exists for a utility car constructed from conven-
tional materials, the permissible cost of the necessary battery can only be estimated by
judicious guesswork. Figures of $0.40-0.50/lb and a total cost of under $200 appear to be
appropriate.
5. Commercial Vehicles
Somewhat less stringent cost requirements may hold in the commercial vehicle
category, since - for the van and bus, at least - production volumes are lower and unit
costs relatively higher than is the case for passenger cars. Moreover, since the percentage of
operating time is very much higher for commercial than for private vehicles and since the
longer life and expected lower maintenance costs are better recognized, a higher cost struc-
ture for the battery can be tolerated. Inspection of the data for assignable weight indicates
that $1.00-1.50/lb might be a reasonable estimate of the allowable cost for a power source
which met the minimum technical requirements for the van, taxi, and bus with conventional
construction.
6. Future Cost Structure of Power Sources
The above cost estimates for the vehicle power source, crude as they are, form some
basis for an assessment of the prospects of the various power sources. As a point of compar-
ison, it may be noted that today's lowest cost battery system, lead-acid, has an OEM price
of approximately $0.25/lb in the high-volume application of automobile starter batteries.
This is based upon relatively low-cost raw materials, the most expensive of which is lead at
$0.13-0.16/lb, and a very mature manufacturing technology. For the lower production vol-
ume and longer-lived industrial lead-acid battery, OEM prices are in the range of $0.50-0.60/lb.
Standard types of nickel-cadmium batteries, which are manufactured on a much smaller
scale than lead-acid and use high cost materials (the raw materials costs being approximately
74
3rthur 3l.3Uttlr.3nr.
-------
$0.80/lb for nickel and $3/lb for cadmium), have a representative cost of about $3/lb in the
0.1 to 1.0 kw-hr size range.
Estimates of the potential cost structure of the new types of batteries now under
development are very speculative. Among the high-temperature alkali metal batteries we
believe that the capacitive storage device of SOHIO and the sodium-sulfur battery of Ford
have prospects of achieving a cost level of approximately $1.50/lb in volume produc-
tion; this seems to be a sufficiently encouraging basis for anticipating an ultimate price in
the $ 1.00/lb range.
Metallic lithium, presently at $7/lb and with prospects of its price falling 50% as its
production volume increases, might seem an expensive element for inclusion in a vehicle
battery. However, its low equivalent weight results in a low requirement per battery, and
with a lithium content of only 5-10%, the cost of this element should not be prohibitive.
The CM lithium-chlorine battery seems likely to have a somewhat higher potential cost
structure, less on account of the lithium than on the materials engineering problems asso-
ciated with the handling of chlorine and the temperature (over 100°C higher than the other
systems). The lithium-selenium and lithium-tellurium systems under investigation at the
Argonne National Laboratory offer no prospect of developing into low-cost batteries be-
cause of the cost and the present scarcity of the cathode materials, selenium and tellurium.
The recent work there on a high-temperature lithium-sulfur battery does, however, hold
promise of yielding a low-cost battery.
Most of the rechargeable metal-air batteries now under development use acceptably
low-cost materials (zinc, sodium, iron) as anodes. While air is free, a significant material cost
is involved in the catalysts needed for effective and long-lived air cathodes and for the
structural material of the cathode itself if it is to be used for oxygen evolution on recharge.
Electrically rechargeable metal-air batteries fall into two broad classes: those which
involve some engineering artifice and those which have a relatively conventional construc-
tion. Among the former are the zinc-air batteries of General Atomic and Yardney Electric
and the sodium-air battery of Atomics International. Both the zinc-air batteries are of some
mechanical complexity, and the sodium-air battery may entail an appreciable materials cost
for mercury. While there can be no firm estimates at this stage, the initial costs of these
"engineering artifice" metal-air batteries in quantity production might be in the region of
$3/lb with good prospects of falling to the $1.50-$2.00/lb region as experience with the
systems grows.
Conventionally rechargeable zinc-air batteries, such as those being investigated by
Leesona, ESB, Union Carbide, and McDonnell-Douglas, may cost close to $1.50/lb, since
they would be of simpler construction. However, the fundamental technical problems asso-
ciated with the recharging of the zinc electrode are so great that such batteries may never
attain a cycle life acceptable for vehicle applications.
75
ZD.Tltttkllnr.
-------
The characteristics of mechanically rechargeable metal-air batteries developed by
Leesona in the zinc-air system, and considered conceptually by Atomics International for
sodium-air, are not sufficiently established for conclusions to be drawn about their applic-
ability to vehicles. When their technology matures, first costs are likely to be comparable
with those of other metal-air batteries; but operating costs will depend largely on the
economics of reconstituting spent anodes, probably at centralized service stations, and at
present these appear to be unfavorable. This type of battery would be attractive, however, if
it could be recharged electrically in regular service and mechanically recharged only during
an occasional long trip.
The very approximate cost criteria which we have set for electric vehicles would
appear to eliminate the prospects of the fuel cell, either as the sole power source or in a
hybrid combination with batteries. A fuel cell is inherently a more complex and thus more
costly device than a battery, because it has to accommodate the orderly passage of materials
across its boundaries in addition to heat and electricity. Moreover, there is as yet no
indication that fuel cells whose performance is adequate for vehicular application will not
require precious metal catalysts. Because of these limitations, ambient-temperature fuel cells
seem unlikely to attain an ultimate cost structure of less than S2 to S3/lb. Although
high-temperature fuel cells will probably have lower costs, their power densities appear to be
too low. Any equipment required for pre-treatment of the fuel, such as reforming or
thermal cracking, will add further to the first costs.
Some general comments on the costs of hybrid systems have already been made in
Section VI. If the hybrid is to be considered as a substitute for the conventional vehicle
power plant, it is apparent that the cost of the engine, generator and battery must be
compared with that of the battery or fuel cell of the purely electric vehicle. If we use the
family car as an example and apply the cost criteria for the gasoline piston engine shown in
Figure 12, the engine-generator costs for a 40-60, prime mover-battery hybrid would be
about $500 (allowing about S2/kw for the ac generator rectifier and speed increaser). With
$650 assigned to motors and controllers, offset by only $500 for the replaced engine
transmission, gearbox, etc., the net extra cost is $650 plus that of the battery. If the total
cost is to be kept within $1000 of the cost of a conventionally powered family car, it is
apparent that the cost of the necessary 51-kilowatt battery, weighing not more than 550
pounds, must be $350 or less. Only the high-temperature alkali metal batteries and the
nickel-cadmium system have prospects of this technical capability, and in this situation the
latter is, of course, too expensive by a factor of at least five. The cost criteria for the alkali
metal battery used in a hybrid system are somewhat more demanding than when this
battery is used in a purely electric vehicle, so that when such batteries become available
there would seem to be no incentive for using them in hybrids.
As previously noted, the best prospects for the hybrid vehicle appear to lie with the
commercial vehicles, and particularly the city bus. The calculations of Section VI indicated
that a 5000-pound battery with a power density of 29 w/lb (or less if lightweight construc-
tion is used) would be adequate in a diesel-powered hybrid. It should be possible for a
specially designed lead-acid battery to meet these criteria, using existing technology, for a
cost of about $3000. Such a hybrid-powered bus would undoubtedly have higher first costs
than a conventional bus, but the percentage increase should be substantially less than for the
smaller vehicles.
76
3rthur ZD.HittlrJnr.
-------
To summarize, it seems apparent that electric vehicles are likely to have higher first
costs than conventional vehicles, even in high-volume production and with maturation of
the contributing technologies. Some of this cost differential may be acceptable on the basis
of lower operating costs due to a reduced maintenance requirement and longer life, but
most must be attributed to the social benefits of decreased air pollution. On the assumption
that an additional $1000 might be placed on the retail price of the family car and perhaps
an additional S500 for the commuter car, high-temperature alkali metal batteries appear to
offer the best chances of meeting the cost criteria. Metal-air batteries will probably have a
somewhat higher cost structure and be acceptable only in commercial vehicles. Fuel cells are
likely to be too expensive, and mechanical-electrical hybrid systems appear promising only
in the case of the city bus.
C. REFUELING
It is apparent that the large-scale introduction of electric vehicles would have a
major impact on a number of important sectors of the national economy. This study,
however, is concerned with the technology of the vehicle itself; the effects of the large
increase in the use of power on the electrical generating industry,* the changes in the
demand for materials, the impact on the petroleum industry, and the effect on the myriad
of ancillary suppliers to the vehicle manufacturers are outside its scope. Consideration is
given to the question of vehicle refueling, however, since it has direct bearing on the
selection of the vehicle power source.
Present-day conventional vehicles are refueled by the simple expedient of filling
their tanks with gasoline or diesel oil as appropriate. The whole operation can be carried out
in five to ten minutes. There would be no change in procedure for a mechanical-electrical
hybrid, and an equivalent procedure is possible for a fuel-cell-powered system, although the
nature of the fuel might set some special restrictions. A vehicle powered entirely by bat-
teries, however, clearly needs a totally different method of refueling. This problem has been
considered in terms of the three general options of rapid recharge, slow recharge, and
mechanical replacement.
Rapid recharge would involve the acceptance of a full charge by the battery within a
period often minutes or so, comparable to the time it takes to refuel a conventional vehicle.
A moment's consideration suggests that this mode of recharge is rather impractical. To
recharge the battery in the family car in ten minutes would require a power source with a dc
output of approximately one megawatt (allowing for inefficiencies). While we have not
determined the probable capital costs of a dc rectifying system with this power capability, it
is likely to be excessively high, even when manufactured on a large scale. Moreover, such a
requirement for power would be likely to invoke a demand charge from the supplying utility.
The implications for the vehicle itself are even more severe. Factors such as safety, insula-
tion, and the cost of batteries are likely to hold the voltage of its electrical system to a max-
imum of 500 volts. At this voltage, the wiring in the vehicle would need to carry a current of
about 2000 amperes during recharge, which would require heavy and costly quantities of cop-
per cabling. A further consideration is the large and probably damaging amount of heat gen-
erated in the battery due to the inefficiencies of the charging processes.
'Complete conversion to electric vehicles today would create an additional load averaging about 40% of
present electrical power consumption.
77
Arthur ZD.3Uttle.3nr.
-------
From the standpoint of battery technology, rapid recharging would be feasible only
for systems having the highest power density. The ratio of maximum energy density to
maximum power density, having the units of time, is a good index of a reasonable minimum
recharging period for a battery. For this to be as low as ten minutes, the ratio must be less
than 1:6; only the nickel-cadmium battery and possibly the SOHIO system seem likely to
meet this criterion.
Slow recharge of a battery is considered to mean a charging period of from four to
perhaps eight hours. The obvious time for this is at home overnight, or possibly at metered
outlets when the vehicle is parked during the day. The rectifying equipment could be either
on the vehicle or at the outlet. The advantages to this approach are the converse of the
disadvantages of rapid recharge: slow recharging involves a very much lower investment cost
in recharging equipment, the weight and cost of cables on the vehicle is less, the electrical
efficiency is higher so that energy is conserved, there is less damage to the battery due to
heating, and its lifetime is consequently lengthened. Home recharging also has the great
advantage of making the motorist independent of visits to a service station, at least for
refueling, and it would have the desirable effect of spreading the load on the electrical
generating system largely onto the off-peak hours.
The obvious disadvantage of slow recharge is that the vehicle is unusable tor a
substantial part of the time. This probably would not be a serious difficulty, since the great
majority of trips are undertaken for relatively short distances and on a fairly fixed daily
schedule. Longer trips might depend on exchanging depleted batteries for fully charged ones
at service stations, which raises questions of ownership and of battery standardization.
Battery leasing with a metered use charge, might well be the procedure adopted. Maint-
enance work on batteries, such as occasional electrolyte replacement in alkaline systems,
might be carried out during their residence time in the service stations.
For batteries with air cathodes an alternative to replacing the batteries on extended
trips is the replacement of the anode alone. However, this procedure is certainly more
complex and probably more costly.
It thus appears that the formidable technical difficulties associated with rapid re-
charging make slow recharging the necessary method of refueling, with mechanical substitu-
tion of batteries being used to extend the range on occasional long trips. This arrangement is
less than ideal, but it appears to have enough advantages to make it acceptable.
D. SUMMATION AND RECOMMENDATIONS
The general conclusion to be drawn from this analysis of the requirements for
electric vehicles is that the necessary technology appears attainable, and that the cost of
electric vehicles, while higher than that of conventional vehicles, might still be acceptable. A
very substantial technical effort is called for if the required technologies are to be brought
to maturity. One of the objectives of this study is to identify areas in which more effort
might profitably be made.
78
3rtliur JD.lUttkllnr.
-------
For the electric vehicle to achieve performance broadly comparable to that of
today's internal combustion powered cars, a low-cost, high-temperature alkali metal battery
must be developed. The major virtues of this class of battery are their high energy density,
which is a consequence of the high reactivity of the alkali metals, and their high power
density, due to the rapid rate at which the electrochemical reactions take place at the
elevated temperatures. Since the alkali metals are well above their melting point in these
batteries, many technical problems of anode recharging are avoided. Of the alkali metals,
only lithium and sodium are serious candidates: lithium because it is the lightest, and
sodium because it is the cheapest. Calcium, although having a high melting point (840°C),
might provide an alternative anode material for high-temperature batteries; no work has been
done in this area, however. Among cathode materials the most promising elements are sulfur
and chlorine. It is also possible that phosphorus might be acceptable but its electrochemical
properties are almost completely unexplored.
Bringing the technology of high-temperature alkali metal batteries to maturity will
obviously be no easy task, since very formidable problems in materials selection and safety
are evidently involved. Much engineering inventiveness will be required in the development
of auxiliary heaters and insulation in the systems for temperature control, these complex-
ities being appropriate only for batteries of the size capable of powering vehicles. The public
acceptance of high temperature alkali-metal batteries is also a major question; the presence
of a 900-pound battery containing alkali metals and elements such as chlorine or sulfur at a
temperature possibly as high as 600°C is obviously not a particularly desirable feature for
the family car. Effective, low cost insulating materials are clearly essential, and the design
must make the battery as crash-proof as possible. Groups presently developing batteries in
this class are optimistic that this can be done.
Given the requirement that the performance of a conventional vehicle must be
matched, metal-air batteries appear to be too power-limited for application in the family
car. They are likely to be somewhat more costly on both a weight and performance basis
than the high-temperature systems, although neither technology is advanced enough to
generate firm cost estimates. Metal-air batteries do, however, appear to have reasonable
prospects for the commercial vehicles, particularly if advances are made in the utilization of
lighter structural materials. The metal-air batteries in general have a major advantage in that
they operate at temperatures close to ambient. For these reasons, further development of
the metal-air systems seems fully justified. Moreover, there are prospects that the pattern of
transportation, particularly in the cities, may change sufficiently in the future in the direc-
tion of smaller and less powerful vehicles so that the metal-air battery would provide a
completely adequate power source.
Alkali metal batteries using organic electrolytes share the large advantage of oper-
ation at ambient temperatures and have high energy density, but their power density is low.
The major problems limiting their future potential as a vehicle power source are the ques-
tions of whether they can ever be recharged at practical current densities and whether they
can achieve an adequate cycle life. As in the metal-air batteries, it is conceivable that
mechanical artifices might be introduced into these systems to overcome the electro-
chemical problems, but as yet such approaches have not been investigated. In view of its
79
21.1LittIc.llnr.
-------
relatively less attractive prospects this class of battery does not appear to merit a major
effort to adapt it for use as a vehicle power source. However, it will undoubtedly see
further development in view of its potential applications in other areas.
Prospects for the fuel cell as a vehicular power source are rated as unfavorable partly
because of its power density limitation but mainly because of its inherently greater com-
plexity and cost relative to batteries. There thus seems to be no justification for a fuel cell
development program oriented exclusively toward civilian highway vehicles, although fuel
cells might well find vehicle applications in the military or specialized industrial fields.
Our consideration of the technology of electric motors and controllers shows that
vehicular requirements are certainly attainable and, in fact, can be largely met today. The
major development effort for the future relates to the lowering of costs to a level acceptable
for the vehicle application. The opening up of such a large market for these components
would be a strong incentive for manufacturers to make such cost reductions, but it is
evident that specifically oriented research programs will be needed to devise production
techniques that will yield the extremely low unit costs demanded.
Any formulation of recommendations for additional research and development pro-
grams in the technologies relating to electric vehicles must take into account existing and
planned programs in these areas by other agencies, both public and private. A considerable
budget is already allocated by the government for support of research and development on
electrochemical power sources and the electrical technologies relating to motors and con-
trols. The two agencies most involved are the Department of Defense and NASA.
Within the Defense Department the greatest expenditures are by the Army. Pro-
grams carried out and sponsored by the Engineering Research and Development Laboratory
at Fort Belvoir, Virginia, have been concerned with novel electrical propulsion systems for
military vehicles, with emphasis on gas turbine generators and fuel cells as the power
sources. Support has been given to work on fuel cells using hydrocarbons (either directly in
high-temperature systems or via a reforming or thermal cracking step), fuel cells using
hydrazine, and various development programs concerning electric motor and control sys-
tems, both ac and dc, which might be employed. The Army Tank Automotive Command at
Warren, Michigan is also working on electrical systems development, complementing the
work at Fort Belvoir. At the Electronics Command at Fort Monmouth, N.J., support has
been given to work on fuel cells and on metal-air and lithium-organic electrolyte batteries.
The latter is concerned mainly with primary batteries and is thus of little relevance to
vehicles.
Work supported by the Navy has included programs on lithium-organic electrolyte
primary batteries and on the identification of more effective and lower cost catalysts for
ambient-temperature fuel cells. The Navy is also concerned with the technology of advanced
high speed electric motors, primarily for torpedo application.
The Air Force has supported most of the earlier work on rechargeable lithium-
organic electrolyte batteries; it has also supported work on metal-oxygen batteries (which
are closely related to the metal-air systems), and on fuel cells and rechargeable batteries for
satellite applications. Much of the initial development of static inverter technology was
carried out under Air Force sponsorship.
80
ZD.lUttle.llnr.
-------
NASA's work in electrochemistry and electrochemical engineering includes the spon-
sorship of major fuel cell programs and related support for manned spacecraft power sources
and the development of improved sealed, rechargeable batteries, particularly high rate
nickel-cadmium and silver systems for satellite application. NASA has also supported a good
deal of work on lithium-organic electrolyte primary batteries, some on metal-oxygen sys-
tems, and has also been active in static inverter technology.
Total expenditures by all government agencies for research and development in the
field of electrochemical power sources have been estimated at approximately $10 million
annually, exclusive of the large fuel cell hardware projects relating to the Apollo manned
spacecraft program. Expenditures of a further $5 million or so can probably be attributed to
developments in the motors and controls field that have some bearing on the technology of
electric vehicles. This latter figure includes estimates for that portion of the electric trans-
mission systems developed for high-speed rail transportation which are applicable to self-
propelled electric vehicles. The size of these amounts naturally varies with the definition of
the relevance of the program and fluctuations in contract appropriations from year to year.
It is also important to recognize that most of the federally sponsored programs have
been focused on such different objectives that they have only a very limited contribution to
make to the development of commercially viable electric vehicles. Thus, while NASA fuel
cell programs have demonstrated that efficient and reliable fuel cells with electrical outputs
of several kilowatts are feasible, the details of the technology are not translatable into the
needs of a mass-produced, low-cost power source. Similarly, the work on vehicle systems
being carried out by the Army naturally focuses on military field situations which involve
specific requirements very different from those of the civil environment. When the specific
content of the programs are taken into account, it can be seen that only a small proportion
of the federal expenditures are applicable to the problems of civilian electric vehicle devel-
opment.
An account of private research effort in power source development has been given in
the body of this report. An overall figure can only be a crude estimate, but it would appear
that expenditures by U.S. private industry on novel power source development relating to
electric vehicles total S5-7 million annually. A substantial fraction of this is represented by
the programs of the two major automobile manufacturers, and a proportion of the re-
mainder is invested in anticipation of later support from federal agencies.
The magnitude of the technical problems involved in the development of batteries
capable of meeting vehicle power source requirements and the increasing urgency of air
pollution abatement in the cities both point to the need for an acceleration of technical
effort if electric vehicles are to be an effective solution. If the assumptions about vehicle
performance made in this study are correct, the major effort should clearly be directed
toward extension of the scale of work on high-temperature alkali metal batteries. The
existing work in this area at GM, Ford, SOHIO, and Argonne National Laboratory is valid
and valuable, but considering the magnitude of the technical problems and the massive
implications of the development, a more comprehensive and coordinated program appears
to be needed.
81
31.lLittIc.Dnr.
-------
Such a program might involve basic electrochemical studies of all elements that have
prospects of satisfying both technical and cost criteria for the vehicle battery; these should
include lithium, sodium, and possibly calcium, magnesium, and aluminum as anodes and
sulfur, chlorine, and possibly phosphorus as cathodes. The prospects of air cathodes in high
temperature systems should also be evaluated, and for completeness some attention should
be given to the potentialities as cathodes of the halides and sulfides of the transition metals.
An important part of the program would be the characterization of the molten salt and
ceramic systems which would be the electrolytes of these batteries. Other work should focus
on the materials selection and engineering problems of the battery. Noncorrodable con-
ductive materials for current collectors and suitable inorganic materials for separators and
supporting structures, sealants, and insulating materials are all needed. Materials selection
and engineering design optimization are usually specific to a particular battery system, so
that the initial work should be concentrated as far as possible on developments having the
greatest general application. The technology of low-cost thermal control systems, heaters,
and insulation is another important area.
The other major technical area deserving additional support is that of rechargeable
metal-air batteries. While their projected performance is not adequate for the most
demanding of the vehicle applications, it may well be acceptable for the types of vehicles
which may evolve from new patterns of transportation within the city. The technical
problems of directly rechargeable metal-air batteries will respond, if at all, to increased
emphasis on the basic electrochemistry of the systems. The engineering solutions such as
those being worked on at General Atomic and Atomics International can, however, be
supported at the development stage, since their difficulties are of a less fundamental nature.
Further work on the problem of developing low-cost catalysts for air cathodes is desirable
and such work would also contribute to fuel cell technology.
In the area of motors and controls the development of low-cost manufacturing
techniques for components and systems should be encouraged; the technology is already in
existence, but the chief emphasis has been on specialized, low-volume applications where
cost is not important. While the prospect of large commercial markets will certainly stim-
ulate private research expenditures in this field, some judicious seeding of development
funds could bring the technologies more sharply into focus for electric vehicle applications.
In summary the technology of electric vehicles appears to be attainable, quite
possibly at an acceptable cost. But this is unlikely to come about in less than ten to fifteen
years and then only if substantial efforts are made at the present stage in the planning and
execution of programs - particularly in the field of power source development — which are
specifically oriented toward the performance and cost parameters demanded by the vehicle
application. The technical risks of this endeavor are too great for the investment to be borne
by private capital alone, particularly since the major objective is a technology desirable for
its social good rather than for its profit potential.
82
Arthur ZD.ltttle.llnr.
-------
APPENDIX A
NOTES ON ELECTRIC VEHICLE DEVELOPMENT IN THE UNITED STATES
While electric automobiles represented a significant proportion of all highway
vehicles in the early 1900's, they had been almost completely displaced by the 1920's, due
to improvements in the internal combustion engine. Apart from some occasional attempts
to market small specialized electric cars, the only continuing commercial activity in recent
years, relating to highway vehicles, has been the manufacture of electric delivery trucks by
the Battronic Truck Corp., a company jointly owned by the Boyertown Auto Body Works,
ESB, Inc., and Smith's Delivery Vehicles Limited, a major British manufacturer.
Gould-National Batteries, Inc., manufactured a line of electric trucks in the mid-I950's, but
discontinued the activity due to lack of demand. Cushman Motors, Inc., manufactured a
dozen small, three-wheel vehicles powered by lead-acid batteries for the Post Office
Department in 1959, and four larger vehicles were purchased in 1962 from a division of
ESB. Inc. At about the same time, National Union Electric Corporation converted some
50-60 Renault Dauphines to electric drive, the product being called the "Henney Kilowatt."
Electric utility companies bought many of them, and several have been the subject of
further conversions described below. While not highway vehicles, electric golf carts are, of
course, a flourishing product, and personnel carriers for use at airports, resorts, and large
industrial plants are becoming commonplace.
Over the last few years there has been considerable activity in experimental electric
vehicle development, mainly in anticipation of their much wider use in the future. This
Appendix contains notes on some of these activities.
A. AMERICAN MOTORS CORPORATION - GULTON INDUSTRIES
AMC and Gulton have announced a joint program to develop a small electric car
making use of a hybrid combination of Gulton's lithium-nickel fluoride batteries and
high-power-density bipolar nickel-cadmium batteries (a concept described in Section V).
The vehicle will seat three persons and weigh about 1100 pounds. The power sources will be
two 75-pound lithium batteries for energy storage and two 25-pound nickel-cadmium
batteries to provide adequate power. With regenerative braking, a 150-mile range is
projected. The limiting feature to this development would appear to be the state of the art
in lithium batteries, which have as yet not demonstrated a convincing degree of
rechargeability in practical battery configurations nor an adequate shelf life.
83
3rthur ZD.1Uttle.Ilnr.
-------
B. ELECTRIC FUEL PROPULSION, INC.
This small company is manufacturing electric cars by conversion of the
conventionally powered Renault-10. The power source is eight 12-volt, 205 amp-hour
(5-hour rate) lead-acid batteries weighing approximately 1600 pounds. The batteries contain
a cobalt additive which is claimed to reduce grid corrosion and thus give relatively longer life
at high charging rates. The car is heavy-4040 pounds curb weight compared with 1735
pounds for the original Renault-10-but the performance specifications are quite impressive:
0-40 mph in 12 seconds, a maximum speed of 60 mph and a range of 70-120 miles under
city driving conditions. The car has a regenerative braking system. A substantial number
(now approaching fifty) of these cars have been purchased or are on order by electric
utilities and other companies interested in the electric propulsion field. Their evaluation
should provide some valuable data on the operating costs of highway vehicles powered by
lead-acid batteries.
C. ESB--BATTRONIC
As mentioned above, ESB, Inc., participates in the electric highway vehicle field via
its part-ownership of Battronic Truck Corp. The principal product of the latter company is a
delivery truck having a loaded gross weight of 9500 pounds. The load and battery weight
combined are 5000 pounds and can be mutually adjusted to suit the route requirements
within load limits of 2000 to 3000 pounds. The batteries used are 84-volt lead-acid systems
with capacities of either 425 or 340 amp-hours. The range of the vehicles is about 40 miles
with a top speed of 25 mph.
ESB also operates a converted Renault Dauphine of the Henney Kilowatt type. This
vehicle is powered by twelve 6-volt golf cart batteries rated at 140 amp-hours (5-hour rate),
which give it a maximum speed of 40 mph, a range of 25 to 35 miles, and an acceleration of
from 0 to 29 mph in 3 seconds.
D. FORD MOTOR COMPANY
Published reports of Ford's activities with experimental vehicles are confined to its
work with the "Comuta," a small vehicle built by English Ford, one of which has been
shipped to Ford's Dearborn headquarters. Comments on the Comuta are included in
Appendix B.
E. GENERAL ELECTRIC
In October 1967, General Electric announced that it had built an experimental car
having a maximum speed of 55 mph on level ground and a range of 40-50 miles under
normal stop-and-go driving. The vehicle, which had a magnesium frame and a fiberglass-poly-
ester body, weighed 2300 pounds and could carry two adults and two children. A novel
feature was the use of lead-acid batteries as the main energy source with nickel-cadmium
84
-------
batteries in parallel to provide power for acceleration and hill climbing. The combined
weight of batteries in the hybrid power source was about 40% of the vehicle's curb weight.
F. GENERAL MOTORS
General Motors Corporation has constructed two full-size electric vehicles primarily
for use as test beds-the Electrovair II, a converted Corvair powered by silver-zinc batteries,
and the Electrovan, a converted GMC Handivan powered by a hydrogen-oxygen fuel cell.
Both vehicles have substantially the same motor and control system, a novel, lightweight,
high-speed (13,000 rpm) ac squirrel cage induction motor. This motor, which weighs 130
pounds and is capable of delivering a maximum of 1.3 horsepower per pound, receives
power from a thyristor inverter. The present weight of the controls, however, tends to
eliminate the weight advantage of the motor system used; the electrical system has a total
weight of 550 pounds. The voltage of the system, 530 volts, is the highest used in any
electrochemically powered vehicle.
Both vehicles were designed to have substantially the same performance as their
conventionally powered counterparts. In the case of the Electrovair II, this meant an
acceleration of 0 to 60 mph in 16 seconds and a top speed of 80 mph. The Electrovair's
silver-zinc batteries, provided by Yardney Electric, weigh about 650 pounds and give a range
of 40-80 miles under normal driving conditions. The hydrogen-oxygen fuel cell in the
Electrovan was manufactured by Union Carbide and operates on cryogenically stored fuel
and oxidant. It has a continuous power rating of 32 kw. A favorable characteristic of fuel
cells is their capability of responding to temporary overload; the fuel cell in the Electrovan
has an overload capability of up to 160 kw. Both vehicles are much heavier than their con-
ventionally powered counterparts: 3400 pounds versus 2600 pounds for the Electrovair and
7100 pounds versus 3250 pounds for the Electrovan.
G. GOULD-NATIONAL BATTERIES, INC.
In cooperation with North Star Electric, a Minnesota electric utility, Gould-National
has recently modified two Henney Kilowatt Renault Dauphines with improved battery and
control systems. The original twelve 6-volt, 180 amp-hour golf cart batteries were replaced
with eighteen 12-volt, 80 amp-hour automotive starter batteries of newly improved design,
having a polypropylene case, internal connectors, glass mat separators, and improved plate
construction. These batteries give 20 watt-hours per pound at the 20-hour rate. Using a GE
solid-state control system, the modified Henney Kilowatts have a range of 65 miles
compared with 35-40 miles for the original models under comparable conditions.
H. ROWAN CONTROLLERS, INC.
The activities of this Westminster, Maryland, based company in the electric vehicle
field are covered in the section on Italian activity in Appendix B.
85
Arthur
-------
I. U. S. ARMY ENGINEER RESEARCH AND DEVELOPMENT LABORATORIES
As part of the U. S. Army's continuing development programs on various approaches
to electrical propulsion in vehicles (see Section VII), USAERDL has converted an M-37
truck to fuel cell power. The power source was a hydrazine-air fuel cell developed by
Monsanto and constructed in 5-kw modules. Four modules were used in the initial
conversion to give a 20-kw rating, but the design allows for later use of an 8-module, 40-kw
fuel cell.
The curb weight of the converted truck was 6000 pounds, of which 700 pounds was
taken up by a conventional dc series motor, 525 pounds by the 20-kw fuel cell system, and
300 pounds by the controller. All of these weights are believed subject to further reduction.
The 20-kw power source permitted a maximum speed of 47 mph and steady climbing up a
20% grade. When the 40-kw system is installed, accelerative performance will be fully
comparable with trucks having the standard 94-hp gasoline engine.
J. WEST PENN POWER COMPANY
West Penn Power (a unit of Allegheny Power System Inc.) is an electric utility
headquartered at Greensburg, Pa. It has constructed a small, open electric car for
demonstration purposes. The chassis and suspension were taken from a Volkswagen sedan,
and an aluminum body was built onto it.The power source is an assembly of six 12-volt, 205
amp-hour lead-acid batteries, and the motor is a conventional 7.1-hp dc traction motor of
the type used in the Henney Kilowatt. The control system designed by West Penn uses
thyristors. The overall curb weight is 2160 pounds. With this system a maximum speed of
50 mph has been attained, and under favorable driving conditions the range is as much as 50
miles.
K. WESTINGHOUSE
In 1967 Westinghouse announced the availability of the "Markette," an electric
passenger vehicle for highway use, capable of carrying two passengers. The vehicle was
derived from the golf carts and personnel carriers which Westinghouse has manufactured
since its acquisition in 1964 of Electric Marketeer of Redlands, California. It was intended
as an auxiliary vehicle for special uses where its quietness, economy, and smooth
performance would offset its limited speed and acceleration. With a curb weight of 1730
pounds, the Markette carried 792 pounds of lead-acid batteries (twelve 6-volt, 217 amp-
hour), and its two 4.5-hp dc series motors provided a top speed of 25 mph in 12 seconds
from a standing start. The range at one stop per mile was about 50 miles. The vehicle was
designed from existing state of the art components and was priced at "under $2000."
Optimal solid-state controls to give it greater operating economy would cost about another
$500.
86
21.HittIc.Ilnr.
-------
The Markette has recently been temporarily withdrawn from production because it
does not comply with all of the new federal safety requirements. This points up a significant
difficulty for developers of electric vehicles, since changes in specifications can be very
costly for vehicles made in small quantities.
L YARDNEY ELECTRIC
The Yardney Electric Company is operating a Henney Kilowatt in which silver-zinc
batteries have been substituted for lead-acid. The higher energy density and power density
of the silver batteries give much improved performance. In the original conversion four
batteries were used, each having fourteen 85 amp-hour cells. Their total weight was 240
pounds in contrast to the 700 pounds of lead-acid batteries which they replaced. Maximum
speed was 55 mph, acceleration 0-30 mph in 5 seconds, and a steady-speed range of 80
miles. Yardney is now planning to reconvert the vehicle for use with batteries of the same
type but improved design (the original were intended for service in military aircraft), with
which it hopes to improve performance still further.
M. ACTIVITIES OF INDIVIDUALS
Besides these activities of industry and government there are also instances of
conversion of conventionally powered vehicles to electric drive by individuals. Among
these are the conversion of an MG roadster by Mr. William Roden of San Diego, California,
and of a Karmann Ghia by Dr. Hugo Myers of Los Angeles. Mr. John Hoke of Washington,
D. C. has constructed his own electric car based on a conventional small car chassis. AM
three of these vehicles are in regular use.
87
21.ltittlr.3lnr.
-------
APPENDIX B
NOTES ON FOREIGN ACTIVITY
While interest in electric vehicles is less widespread in Europe and Japan than in the
United States, many countries, particularly the United Kingdom, are quite active in vehicle
design, control development, and investigation of new types of electrochemical power
sources. The motivation of this interest is generally concern about air pollution, although
the use of small electric vehicles to relieve traffic congestion and parking problems in the
cities is equally important. In Europe, petroleum-based fuels are taxed more heavily than in
the United States, and, as a result, there is proportionately greater incentive to use electric
vehicles. In the materials handling field, for example, well over half of European lift trucks
are electric. Electric propulsion is thus at a higher level of acceptance than in the United
States.
The following are some notes on the more prominent of the foreign activities.
A. UNITED KINGDOM
Activity in electric highway vehicle development in the United Kingdom takes place
at two levels. The first involves the manufacture of electric trucks for specialized purposes,
primarily milk delivery, and is a well-established commercial undertaking. The second
concerns the development of a small car, generally a two seater for use as a commuter
vehicle in urban areas. In addition, some research is being done on novel types of electro-
chemical power sources.
The use of electric trucks on United Kingdom highways has increased substantially
over the last 15 years; by the end of 1967 there were approximately 45,000 in service.
About 85% of them are for domestic milk delivery. The most prominent manufacturers are
Austin Crompton Parkinson (ACPEV), Smith's Delivery Vehicles, Stanley Engineering, and
W & E Vehicles. A typical truck having a carrying capacity of about 3000 pounds is shown
in Figure B-l. Power sources for these vehicles are typically 72-volt, 300 amp-hr lead-acid
batteries, which give a range of 20-30 miles with 200 stops and a maximum speed of about
15 mph. Their robustness, ease of maintenance, and low-operating costs make them very
popular with dairies and others engaged in door-to-door delivery.
Hybrid electric trucks have recently appeared in the United Kingdom. Some new
models being offered by Austin Crompton Parkinson achieve extended range through use of
a small propane-fed engine generator.
Electric car development, which is still at the prototype stage, has focused on small
vehicles, usually two seaters, to be powered by lead-acid batteries and to be priced at around
$800. A photograph of the "Scamp", a typical vehicle of this type, is shown in Figure B-2.
These vehicles generally have a maximum speed of about 40 mph and a range of 30-40 miles
89
ZD.^tttkllnr.
-------
Courtesy of W&E Vehicles
FIGURE B-1 BRITISH ELECTRICALLY DRIVEN MILK TRUCK
Courtesy of Scottish Aviation, Ltd.
FIGURE B-2 THE SCAMP: A DEVELOPMENTAL VEHICLE
90
-------
under non-stop driving conditions. Table B-l lists specifications for some of them. The two
Electricity Council Minis and the Peel Trident are conversions from conventionally powered
cars, but the Scamp (of which a dozen were built) and the Ford Comuta were designed as
electrics. All are two-seaters except the Trident, which is essentially a single-seater.
Besides the cars listed in Table B-l, design, mock-up, and prototype construction is
being undertaken by Tube Investments, Hamblinetta (Flamenco), Carter Engineering (Carter
Coaster), and Telearchics Ltd. (Winn City Car). These cars are being designed to have
lightweight plastic bodies.
Most research on power sources has concentrated on fuel cells. Shell Oil Company
carries out its research in England and has successfully demonstrated a 5-kw fuel cell system
operating on reformed methanol with air as the oxidant. Chloride Electrical Storage, the
largest British battery manufacturer, has shown a 2-kw hydrogen-oxygen cell fed by cylinder
gas; it is used to power an in-plant tractor truck. Energy Conversion Ltd. is also active in
fuel cell research.
In the battery field there is considerable interest in improved lead-acid batteries with
thinner plates and higher energy and power density. This work is being carried out by
battery companies such as Chloride and Oldham and at the laboratories of the Electricity
Council. In the metal-air battery field, Joseph Lucas Ltd. and Crompton-Parkinson (Division
of Hawker Siddeley) have cooperative programs with General Atomic and Leesona, respec-
tively. British Motors has recently announced its interest in developing a vehicle to be
powered by a zinc-air battery of the Leesona type in its subsidiary ACPEV, jointly owned
with Crompton-Parkinson. No comment was, however, made on the prospective economic
feasibility of such a vehicle.
Work on solid-state control systems is going on at Lansing Bagnall, Ltd. and at
ACPEV. The initial application for this development is, of course, in the electric truck-
materials handling area. Programs aimed at improving the characteristics of motors are
being carried out at Bristol University and the laboratories of the Electricity Council at
Capenhurst.
B. FRANCE
The major French activity in electric vehicles is a government-coordinated program
directed toward the development of a fuel-cell-powered car by 1971. Present plans call for
the power source to be a hydrogen-oxygen fuel cell with a rated power of 25 kw.
Various aspects of this development program are being conducted by specialized
contractors. The government-controlled automobile company, Renault, will build the
vehicle, and companies such as CSF and C. F. Thomson-Houston are developing motors and
control systems.
91
ZD.lUttle.3lnr.
-------
Electricity Council
AEI Mini-1
Electricity Council
Telearchics Mini-2
TABLE B-1
SPECIFICATIONS Of SOME BRITISH EXPERIMENTAL ELECTRIC VEHICLES
Curb
Weight
(Ib)
Motor hp
(Continuous
Rating)
2,499 12hp
2,378 2 at 3 hp
Controls
Thyristor
Carbon pile
Batteries
96 v, 66 amp-hr
(1-hr rate)
829 Ib
64 v, 110 amp-hr
(1-hr rate)
Acceleration
0-30 in 12
seconds
0-20 in 5
seconds
Maximum Speed
(mph)
41
40
Range
(miles)
30 non-stop at
37 mph
28 non-stop at
40 mph
823 Ib
Peel Engineering
Trident
600 24 volt
5hp
Carbon pile 24 v, 153 amp-hr 0-30 in 7
seconds
35
Scottish Aviation
Scamp
1,000 2at2.7hp Carbon pile
48 v, 105 amp-hr 0-30 in 10.5
(5-hr rate) seconds
400 Ib
35
30 non-stop
Ford Comuta
1,200 2 24 volt Thyristor 48 v, 85 amp-hr 0-30 in 14
5 hp (1-hrrate) seconds
> 30
40 at 25 mph
£
-------
A major fuel cell program is under way at the Institut Franchise du Petrole
supported by development work on carbon-based electrodes at Societ^ le Carbone Lorraine
and on sintered-metal-based electrodes at Ugine-Carbone. Other types of fuel cells are also
under investigation: there are programs on a Bacon-type fuel cell at ONIA (Office National
Industrial d'Azote) and on hydrazine fuel cells at Alsthom. This latter company has
recently announced a joint vehicle development program with the automobile manufacturer,
Peugeot. Metal-air batteries are under investigation at CIPEL.
An interesting example of electric vehicle application in France is the use of a
battery-powered bus between the airport and the city of Marseilles. This is a full-size vehicle
weighing 36,000 pounds and capable of carrying 26 people seated plus 54 standing. It is
powered by a 96-volt, 750 amp-hour lead-acid battery that provides a range of 50
kilometers.
C. GERMANY
A small but significant number of electric delivery trucks were in use on German
highways prior to 1950; the system of license fees was then changed so that the fees were
based largely on weight, and this virtually eliminated the vehicles. In recent years there has
been a revival of interest, although on a much smaller scale than in the United States and the
United Kingdom. However, self-propelled rail cars powered by lead-acid batteries are in
substantial use on the German national railroads. They were introduced after World War II,
and about 300 are now in service.
Most of the work in Germany on novel power sources has been focused upon fuel
cells. Both Siemens-Schuckertwerke and Varta have large groups working on hydrogen-
oxygen fuel cell development, and Bosch has sponsored work on methanol fuel cells at the
Battelle Institute in Frankfurt. The group at Varta has an interesting concept for
regenerative braking. The energy from the generator is used to electrolyze water, providing
hydrogen and oxygen which are stored and used subsequently in the fuel cell.
93
ZD.TUttlrJnr.
-------
D. ITALY
Italian participation in electric car development is confined to the design and con-
struction of small experimental vehicles with relatively conventional motor and control
systems, powered by lead-acid batteries.One of the more prominent of these vehicles, shown
in Figure B-3, was built by Ghia S.p.A. and De Tomaso Automobili, subsidiaries of the
Rowan Controller Company of Westminster, Maryland. This four/five seat vehicle has a curb
weight of 1,300 pounds and is powered by lead-acid golf cart type batteries. The maximum
speed is 40 mph and the range 50-100 miles, depending on driving conditions. A relatively
inexpensive solid-state controller is used which operates by switching the field current of the
two 8-kw dc compound motors specially designed for this application by the General
Electric Company. The controller circuit is also utilized for regenerative braking.
A very small electric vehicle called the "Urbanina" of unconventional design and for
use in congested downtown locations has also been exhibited. This car weighs only 800
pounds. The passenger compartment holds two people and is in the form of a rotating
turret. A range of 50 miles and top speed of 38 mph have been reported. Other Italian
electric cars involve conversions of conventional Fiat 500's.
E. SWEDEN
Urban concentration in Sweden is much less than in other developed countries, and
the comparatively large distances between populated centers make electric vehicles less
immediately attractive; nevertheless, interest in their development is strong. The most
notable Swedish activity is in fuel cell development: ASEA has constructed the world's
largest fuel cell system, which uses cracked ammonia and has a power rating of 240 kw. The
immediate application was as a power source for submarines, but smaller modules of the
system have been used to power experimental lift trucks, and electric highway vehicles are
also under consideration.
In a recent demonstration co-sponsored by the local Stockholm electrical utility,
ASEA exhibited an electrically powered SAAB. This vehicle contained a 120-volt nickel-
cadmium battery weighing 1,000 pounds. It had a range of 25 miles, falling to about 15
miles in city traffic. The top speed was 43 mph.
F. BELGIUM
The most significant work on electric vehicles in Belgium is being carried out by
Ateliers de Constructions Electriques Charleroi (ACEC), which has developed a full sized
diesel-electric bus. This vehicle has twin electric motors built into its back wheels. While
built largely for experimental purposes, the bus spends most of its time in local service in
the town of Charleroi. In the present model no auxiliary traction battery is used; electricity
is supplied by the diesel-powered generators as needed.
94
3rthur 2D.lLittle.3nr.
-------
01
Courtesy of Rowan Controller Corporation.
FIGURE B-3 PROTOTYPE ELECTRIC CAR BUILT IN ITALY
BY SUBSIDIARIES OF ROWAN CONTROLLER CORPORATION
-------
G. JAPAN
Japanese development of electric vehicles has been focused mainly on the conversion
of small, conventionally powered cars to electric propulsion. The work has been prompted
by the country's increasing concern about air pollution caused by heavy traffic in the cities.
Three of these program are collaborative efforts involving a number of manufacturers and
are organized by three utility companies: Tokyo Electric, Kansai Electric, and Chubu
Electric. The Kansai Electric and Tokyo Electric groups used standard lead-acid batteries in
conversions of 800 cc and 360 cc vehicles, respectively, while Chubu Electric Battery
Company developed a special 80-volt, 400 amp-hour lead-acid battery for its conversion of a
1, 500-cc vehicle. All these vehicles are reported to have maximum speeds of the order of 40
mph and a range of about 50 miles at steady driving. All have thyristor controls and
regenerative braking.
An interesting recent development is the announcement by Tokyo Shibaura Electric
Company (Toshiba) of a small car powered by a high-speed brushless dc motor. Rated at 27
hp at 20,000 rpm, the motor gave the vehicle a maximum speed of 62 mph.
In addition to the above activity, the city of Osaka is reported to be contemplating
purchase of a fleet of small electric buses for downtown use.
96
ZD.llittklliu
------- |