EPA-460/3-74-020-C
OCTOBER 1974
IMPACT OF FUTURE USE
OF ELECTRIC CARS
IN THE LOS ANGELES REGION:
VOLUME III - TASK REPORTS
ON IMPACT AND USAGE
ANALYSES
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Contr6l
Alternative Automotive Power Systems Division
Ann Arbor, Michigan 48105
-------
EPA-460/3-74-020-C
IMPACT OF FUTURE USE
OF ELECTRIC CARS
IN THE LOS ANGELES REGION:
VOLUME III - TASK REPORTS
ON IMPACT AND USAGE
ANALYSES
Prepared by
W.F. Hamilton, J.A. Cattani, J.C. Eisenhut,
F.J. Markovich, J.R. Martinez, R.A. Nordsieck, andA.R. Sjovold
General Research Corporation
P.O. Box 3587
Santa Barbara, California 93105
Contract No. 68-01-2103
EPA Project Officer: C.E.Pax
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control Programs
Alternative Automotive Power Systems Development Division
Ann Arbor, Michigan 48105
October 1974
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - as supplies permit - from the Air
Pollution Technical Information Center, Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
This report was furnished to the U.S. Environmental Protection Agency
by General Research Corporation in fulfillment of Contract No. 68-01-2103
and has been reviewed and approved for publication by the Environmental
Protection Agency . Approval does not signify that the contents necessarily
reflect the views and policies oi the agency . The material presented in this
report may be based on an extrapolation of the "State-of-the-art." Each
assumption must be carefully analyzed by the reader to assure that it is
acceptable for his purpose. Results and conclusions should be viewed
correspondingly. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
Publication No. EPA-460/3-74-020-C
n
-------
INTRODUCTION
This report Is published in three volumes:
Volume 1, Executive Summary and Technical Report
Volume 2, Task Reports on Electric Car Characteristics
and Baseline Projections
Volume 3, Task Reports on Impact and Usage Analyses
Volume 1 is a comprehensive account of the effects that electric
cars would have on the air quality, energy use, and economy of the Los
Angeles region in 1980-2000. Volumes 2 and 3 contain ten individual
reports documenting the analyses on which Volume 1 is based. These
reports detail the methods, data, assumptions, calculations, and results
of the study tasks, and were originally published at the conclusion of
each task.
Task reports in Volume 2 project future characteristics of electric
cars and of the Los Angeles region in which they would be used, as follows:
1. D. Friedman and J. Andon (Minicars, Inc.) and W. F. Hamilton,
Characterization of Battery-Electric Cars for 1980-2000
Postulates electric vehicle performance requirements, projects
representative future battery characteristics, calculates urban
driving range versus total car weight, and estimates energy
and material requirements for selected driving ranges.
2. G. M. Houser, Population Projections for the Los Angeles Region.
1980-2000
Projects population of California's South Coast Air Basin, which
includes greater Los Angeles, by county and age group.
-------
3. W. F. Hamilton and G. M. Houser, Transportation Projections
for the Los Angeles Region, 1980-2000
Projects Los Angeles freeway and transit networks, auto
population, auto usage, auto size and age distributions, and
average fuel consumption.
4. J. Eisenhut, Economic Projections for the Los Angeles Region,
1980-2000
Projects employment and income for the South Coast Air Basin,
and the payroll and employment of businesses involved in
production, distribution, and maintenance of automobiles and
parts.
5. A. R. Sjovoid, Electric Energy Projections for the Los Angeles
Region. 1980-2000
Summarizes the US energy situation as forecast in recent
studies, and in this context projects electric energy produc-
tion and consumption in the South Coast Air Basin, noting
energy available for electric car recharging and its basic
sources.
Task reports in Volume 3 project impacts due to various levels of
electric car use and investigate possible future levels of use, as follows:
6. J. R. Martinez and R. A. Nordsieck, An Approach to the Analysis
of the Air Quality Impact of Electric Vehicles
Selects the "DIFKIN" computer model and linear rollback as means
for analyzing future air quality in the South Coast Air Basin,
designates important cases for investigation, and details
required methodology.
7. J. R. Martinez and R. A. Nordsieck, Air Quality Impacts of
Electric Cars in Los Angeles
Forecasts stationary and vehicular pollutant emissions in
spatial and temporal detail, with and without electric cars,
and calculates consequent air quality levels relative to
Federal standards.
8. A. R. Sjovoid, Parametric Energy, Resource, and Noise Impacts
of Electric Cars in Los Angeles
As a function of percentage electric car use, forecasts total
energy consumption and petroleum consumption in the South Coast
Air Basin through the year 2000; compares annual consumption
-------
and rolling Inventory of key electric car materials with past
and projected US production, consumption, and reserves;
analyzes possible reductions of community noise from electric
car use.
9. J. C. Eisenhut, J. A. Cattani, and F. J. Markovich, Parametric
Economic Impacts of Electric Cars in Los Angeles
Projects life cycle costs of alternative electric cars in
comparison with conventional cars; analyzes and projects changes
in employment and payroll in industry segments impacted by
electric cars, including service stations, battery manufactur-
ing, auto parts and repairs, and auto sales; considers overall
regional and national economic impacts of electric cars.
10. W. F. Hamilton, Usage of Electric Cars in the Los Angeles
Region. 1980-2000
Analyzes 1967 data to determine distributions of daily driving
range in Los Angeles and the applicability of limited-range
electric cars; reviews market trends and estimates the potential
free-market sales of electric cars in the South Coast Air
Basin; hypothesizes particular levels of electric car use for
impact evaluations; and considers relative economic incentives
likely to be required to obtain these usages.
-------
TASK REPORT 6
AN APPROACH TO^THE ANALYSIS OF THE AIR
QUALITY .IMPACT OF ELECTRIC VEHICLES
J.R. Martinez
R.A. Nordsieck
-------
ABSTRACT
As a part of an overall evaluation of electric vehicles as a trans-
portation alternative, a plan is outlined for quantifying the air quality
changes which could be expected to accompany various levels of replacement
of conventional automobiles by electrics. The study will focus on California's
South Coast Air Basin in the years 1980, 1990, and 2000. The fraction of
all vehicle miles which are traveled by electrics, itself a function of
such variables as vehicle performance, availability of recharging power,
and resource limitations, is selected as the independent variable for the
parametric analysis. Baseline pollution levels (ozone, oxides of nitrogen,
hydrocarbons, carbon monoxide, particulates, and sulfur dioxide) in the
study years will be predicted using the DIFKIN photochemical-diffusion
model and the rollback technique. Parallel applications of the DIFKIN
model and the rollback technique will predict changes in worst case air
quality with varying electric vehicle use in the study area, with parti-
cular emphasis on known high pollution locations. Predictions of more
typical pollution levels will be based on concentration probability dis-
tributions compiled from existing LAAPCD records.
The worst case conditions will be characterized by meteorology
measured on a typical high oxidant day in September 1969. A series of
time-phased trajectories will be simulated to arrive at selected sites
which have regularly exhibited high pollutant concentrations. The effects
of increased NO emissions due to added power plant demand for battery
X
recharging will be evaluated similarly using air trajectories starting
at selected power plant locations. Emissions from vehicular sources are
to be calculated from network flow model projections of vehicle miles
travelled and estimates of average vehicle emission factors based on
vehicle mix, emission control rules, and deterioration effects.
-------
Extrapolations of air pollution control district emissions inventories
distributed geographically through land use projections will provide a
model of stationary source emissions.
ii
-------
CONTENTS
SECTION PAGE
ABSTRACT i
1 INTRODUCTION 6-1
2 OBJECTIVES AND METHODOLOGY 6-3
2.1 Objectives 6-3
2.2 Methods of Analysis 6-3
3 SIMULATION OF AIR QUALITY USING THE GRC MODEL 6-10
3.1 Overview of the GRC Model 6-10
3.2 Considerations in Air Quality Modeling 6-11
3.3 Selection of Critical Sites 6-14
3.4 Identification of Worst-Case Meteorology 6-20
3.5 Selection of Trajectories 6-21
4 FACTORS AFFECTING AIR QUALITY IMPACT OF ELECTRIC
VEHICLES 6-23
4.1 Effect of Electric Cars on Emissions from
Stationary Sources 6-23
4.2 Public Policies and Electric Vehicle Use 6-26
5 FORECASTING OF POLLUTANT EMISSIONS 6-31
5.1 Overview 6-31
5.2 Vehicular Emissions 6-32
5.3 Stationary Source Emissions 6-36
6 AIR QUALITY IMPACT MATRIX 6-43
7 ANALYSIS AND DATA DISPLAYS 6-46
APPENDIX DERIVATION OF AIR QUALITY STATISTICS 6-51
REFERENCES 6-59
iii
-------
iv
-------
ILLUSTRATIONS
NO. PAGE
2.1 Flow Chart of Major Steps in Air Quality Impact Analysis 6-9
3.1 Schematic GRC Photochemical Diffusion Model for Air
Quality Simulation 6-10
3.2 Map of South Coast Air Basin Showing Critical Sites and
a Typical Autumn Afternoon Flow Pattern 6-18
3.3 Trajectory for September 29, 1969, Which Arrives at
Anaheim at 1300 6-22
5.1 Baseline Projected Peak Demand for Electricity Generated
by Oil Fired Plants in the South Coast Air Basin 6-38
5.2 Projected Baseline Diurnal Power Demand on Oil Fired
Power Plants in the Los Angeles Area for Peak Demand
Month 6-39
7.1 Typical Plot of One Species Concentration Versus Electric
Car Fraction 6-48
7.2 Plot of Normalized Pollutant Concentration Versus Electric
Car Use for a Particular Year 6-48
-------
vi
-------
TABLES
NO. PAGE
1.1 Motor Vehicle Emissions in the Los Angeles Basin in 1970 6-1
3.1 August-October Statistics for 1969 and 1970 6-13
3.2 Selected Locations in the Los Angeles Air Quality Control
Region and Most Significant Pollutants at Each Site 6-17
5.1 Oil Fired Power Plant Emission Factors, Pounds per
Thousand Gallons 6-39
6.1 Matrix of Air Quality Impact Cases 6-43
7.1 National Ambient Air Quality Standards 6-47
vii
-------
viii
-------
1 INTRODUCTION
Electric cars have been proposed as a means of alleviating the air
pollution problems that afflict urban areas. As one step in evaluating
this proposal, this work describes the approach GRC will follow in eval-
uating the air quality changes likely to occur if significant numbers of
electric vehicles replace conventional cars in the South Coast Air Basin
(SCAB) of California, which contains Los Angeles County and outlying
areas.
The Los Angeles region is probably the best known, and certainly
the most studied example of the interactions among air pollution, motor
vehicles, and urbanization. With an area of 1250 square miles, it con-
tains more than 7 million inhabitants and more than 4 million motor
vehicles. An idea of the relative contribution of motor vehicles to Los
Angeles air pollution can be obtained from Table 1.1 which shows the emis-
sions due to motor vehicles in tons per day and as percent of total emis-
sions; motor vehicles are by far the largest emitters of hydrocarbons and
nitrogen oxides, the two pollutants responsible for the formation of
photochemical oxidants. Stringent new emission controls on motor vehicles
are expected to reduce the daily emissions considerably, both in tonnage
and as a fraction of the total. However, even with the new controls the
TABLE 1.1
MOTOR VEHICLE EMISSIONS IN THE LOS ANGELES BASIN IN 19701
_, ,.,,.. Quantity Percent of Total Emissions
Type of Emission , ~, \ ^ < ^ ™
(tons per day) of the Same Type
Carbon Monoxide 8960 98%
Reactive Hydrocarbons 1170 85%
Nitrogen Oxides 755 72%
Particulates 55 42%
Sulfur Dioxide 35 14%
6-1
-------
hydrocarbon emissions have been projected to start increasing once again
around 1985 as projected increases in the car population offset the ef-
fects of controls. A partial conversion to nonpolluting electric vehi-
cles might prevent this upturn. Moreover, the recent extension of dead-
lines for implementing new automobile emission standards makes even more
certain that high air pollution levels will prevail in Los Angeles for
some time to come. Among the many air pollution control proposals which
have been made, the electric car is both a logical and progressive step
forward in the search for an air pollution abatement strategy.
Before adopting electric cars as a component of an air pollution
control strategy, it is, of course, essential that their impact on air
quality be analyzed. Not only do the costs involved demand such an
analysis, but, as experience with pollution control has often shown,
solving one problem can create another. In short, the interactions among
air pollution and its sources must be closely examined; accordingly,
this report presents a program designed specifically to investigate the
impact of electric vehicles on air quality.
The sections that follow outline and discuss the objectives, and
methods of analysis to be used in the subsequent study phase. We describe
our approach to simulating air quality in Los Angeles, methods of fore-
casting emissions from vehicular and stationary sources, the cases to be
considered, and the data displays to be employed. Analysis results are
reported in Task Report 7.
6-2
-------
2 OBJECTIVES AND METHODOLOGY
2.1 OBJECTIVES
The objective of the air quality impact analysis described herein
is a quantitative assessment of the probable concentrations of air pol-
lutants to be found in the Los Angeles region if conventional cars were
replaced with electric vehicles in varying degrees. Evaluations will be
performed for three target years: 1980, 1990, and 2000. For each of
these years we seek answers to the following related questions:
• What would be the air pollution levels if no electric cars
were used?
• What changes in air quality take place as electric car use
increases?
• What is the probable frequency of occurrence of various con-
centrations of pollutants?
As a corollary to the second question, we may wish to determine the
levels of electric car use required to bring air quality levels into com-
pliance with national standards. The methods to be used in answering
these questions are discussed below.
2.2 METHODS OF ANALYSIS
2.2.1 General
This section sketches the processes to be used in performing our
air quality impact analysis. Subsequent sections discuss the topics
introduced here in detail.
Let us begin with a brief explanation of the concepts involved in
the analysis. In essence, our goal is to determine air quality as a
function of electric car use for each of three years: 1980, 1990, 2000.
The first step is to establish baseline air pollution levels prevailing
in those years if no electric cars were used. The techniques to be used
6-3
-------
in forecasting mobile and stationary source emissions for these years
are described in Sec. 5. Electric car use in terms of numbers of vehicle-
miles traveled is then varied parametrically following specified con-
*
straints. Varying electric vehicle use alters the baseline emission
patterns since doing so is presumed to reduce the use of conventional
autos by some commensurate amount. To evaluate air quality these emis-
sion patterns must then be related to pollutant concentrations. Linking
emissions and air quality is thus the heart of the analysis and is dis-
cussed below and in subsequent sections. The results obtained are then
used in conjunction with other data to deduce the statistics of the prob-
able frequency of occurrence of various pollutant concentrations. As
is evident, this procedure will yield answers to the questions posed at
the beginning of this section (Sec. 2.1).
Two different methods will be used to relate source emissions and
pollutant concentrations in the atmosphere. One method employs the air
quality simulation model developed at GRC, and the second utilizes the
so-called rollback formula.
Use of the GRC model is discussed in Sec. 3; the model itself is
described briefly in Sec. 3.1 and extensively in Refs. 2 and 3. In sim-
plest terms, the GRC model simulates the complex interactions among
emissions, weather, and chemistry to predict the concentrations of air
pollutants.
2.2.2 Worst-Case Approach
Recognizing that national air quality standards (cf. Sec. 7) are
specified on a worst-case basis, our analysis will focus on determining
*
Typical factors which affect use levels are market forces, availability
of materials (e.g., lead), availability of electric generating capacity
for recharging batteries, and vehicle characteristics such as range and
speed. A comprehensive discussion of these factors is, of course, be-
yond the scope of this task report.
6-4
-------
worst-case pollution levels. An immediate consequence of worst-case analy-
sis is that we can restrict our attention to those areas of the Los Angeles
basin presently afflicted with severe pollution conditions. Criteria
for identifying such critical locations are discussed in Sec. 3.3.
As we shall see in the next section, worst-case analysis is a simple
matter when using the rollback formula. Using the GRC model, however,
calls for specifying meteorological conditions which are conducive to high
pollutant concentrations. Criteria for specifying the appropriate metero-
logy are described in Sec. 3.4.
2.2.3 The Rollback Equation
In contrast to the GRC model, the rollback model assumes that pol-
lutant concentrations and emissions are proportional. The formula used
is
(PAQ) - (DAQ)
(PAQ) - (B)
where R = fractional reduction in emissions
PAQ = present air quality
DAQ = desired air quality
B = background concentration of pollutant
Thus the rollback formula purports to yield the reduction in emissions
required to achieve the desired air quality, e.g., the air quality stand-
ard, given the current air quality and the background level of the pol-
lutant, i.e., the concentration unavoidably present due to natural causes.
It should be noted that, since the air quality standards are set
on a worst-case basis, a worst-case approach is implicit in using the
rollback formula with the air quality standard as the desired air quality.
Thus, for the present air quality (PAQ) we must use the highest concen-
tration recorded in the region during the period of interest.
6-5
-------
Extreme caution is necessary in using the rollback equation to re-
late emissions and air quality. While the rollback approximation may fit
some cases [e.g., carbon monoxide (CO)], its use for relating emissions
and secondary pollutants such as nitrogen dioxide (NO,,) and ozone (0_) is
highly questionable since these two pollutants are produced by chemical
processes which are highly nonlinear. The use of Eq. 2.1 is also sus-
pect even with highly reactive primary pollutants such as reactive hydro-
carbons (HC) and nitric oxide (NO).
An example of the use of the rollback equation is found in Ref. 4,
which describes the formulation of an abatement strategy for Los Angeles
designed to reduce ozone levels. Reference 5 contains an evaluation of
the ozone changes produced by the cutback in hydrocarbon emissions pre-
dicted by rollback. One important finding of this evaluation is that
under certain circumstances rollback predicts emissions reductions that
are greater than necessary to satisfy the ambient air quality standard
for oxidant. Additional discussion regarding the inception and use of
rollback is found in Ref. 6.
Since our aim is to determine the desired air quality (DAQ) levels
given a present air quality (PAQ), the background level (B), and the
fractional reduction in emissions (R), we must solve for DAQ in
Eq. 2.1, which yields
DAQ = PAQ(1 - R) + RB (2.2)
In using Eq. 2.2 the factor PAQ would be the baseline air quality level
with no electric car use. The factor R would be determined by the
emissions reduction caused by the increasing use of electric vehicles.
One difficulty in using Eq. 2.2 is that the baseline air quality
(PAQ) for 1980, 1990, and 2000 is not readily available. Two approaches
may be followed to circumvent this problem. The first is to predict the
6-6
-------
baseline air quality by means of Eq. 2.2 itself. This would be done by
using for FAQ , the known air quality for a previous year such as 1969,
and estimating R from
R = 1 - =-^~ 5 k = 1980, 1990, 2000
where E are the baseline emissions for the year k . There is an
obvious danger in using this approach, for the gross approximations in-
herent in the rollback formula may yield anomalous air quality levels.
The second approach is to use the values the GRC model predicts as the
baseline air quality levels for 1980, 1990, and 2000. This has the advan-
tage of providing a common starting point for the air quality impact
analysis for both the GRC model and the rollback formula. Which approach
is better will be determined in Task Report 7.
Since the assumptions underlying these two methods are poles apart.,
predicting concentrations with the rollback approach and the GRC model
would yield results which cannot be readily compared. It is therefore
difficult to delineate criteria for comparing the results obtained by
the two methods in advance, and thus we must await the results of the
computations before making any statements about any differences which
occur.
2.2.4 Estimation of Overall Air Quality by Statistical Analysis
Worst-case analysis yields information about pollution levels with
respect to air quality standards. Thus we can tell whether or not the
air quality complies with the standard. Unfortunately, this type of
analysis does not tell us much about the general quality of the air the
rest of the time. A scheme is described below which can provide such
information in the form of frequency of occurrence of various pollution
levels.
6-7
-------
We wish to use the results of the worst-case air quality predic-
tions together with additional statistical data to deduce how often cer-
tain pollution levels may occur. Such information allows us to make
judgments, within some reasonable limits of course, regarding the over-
all quality of the air in the region under study.
The details of the procedure used for deriving statistical informa-
tion from worst-case results are given in the Appendix. The essence of
the process is as follows. We need to find a probability distribution
which passes through the worst-case concentrations that have been pre-
viously computed. In order to do this we must determine the type and
7 8
the parameters of the distribution. The work of Larsen ' may allow us
to assume that the hourly-average concentrations with which we will be
dealing are distributed lognormally. The lognormal distribution is com-
pletely defined by two parameters, of which the worst-case concentration
is one. The second parameter is the standard deviation, which is obtained
7-9
from actual concentration data for various pollutants. Larsen has
tabulated values of the standard deviation for various cities and pollu-
tants; Ref. 9 contains extensive data for California. In addition, we
will process tapes of concentration data from the Los Angeles County Air
Pollution Control District (LAAPCD) in order to obtain updated estimates
of the standard deviation, since Larsen's work stops with 1968. This will
also let us test whether Larsen's results can be used. In any event, the
standard deviation can be estimated independently and this, together with
the worst-case concentrations, completes the definition of the distribu-
tion function. Having the distribution function allows us to determine the
the probability of occurrence of a given pollution level.
Obviously, the above procedure must be used with caution, to say
the least. Interpretation of the results must be guided by the judicious
application of bounding criteria to the derived probabilities. Also, it
is well to mention at this point that part of Larsen's work has recently
come under some justified criticism. However, this has no bearing on
6-8
-------
the process just described since we do not use that part of Larsen's work,.
It should be noted that Larsen's findings about the lognormal distribution
of pollutant concentrations have recently received independent support. *
2.2.5 Summary
The process followed in the analysis of the air quality impact of
electric vehicles is illustrated in Fig. 2.1; the figure is self-
explanatory.
•7 YEAR y / y = 1980; 1990; 2000
/DEFINE LEVEL, I
L, OF ELECTRIC/
CAR USE /
L = 0 DEFINES
THE BASELINE CASE
DETERMINE EMISSIONS
FOR SPECIFIED YEAR
AND LEVEL OF USE OF
ELECTRIC CARS
EMISSIONS
FORECAST
L ^
COMPUTE WORST-CASE
AIR QUALITY LEVELS
GRC
MODEL
ROLLBACK
WORST-CASE
AIR-QUALITY
ESTIMATES
'
OBTAIN STATISTICS
OF OVERALL AIR
QUALITY
ESTIMATE OF
GENERAL AIR
QUALITY
^~
DISPLAY AND
INTERPRET
RESULTS
Figure 2.1. Flow Chart of Major Steps in Air Quality Impact Analysis
6-9
-------
3 SIMULATION OF AIR QUALITY USING THE GRC MODEL
3.1 OVERVIEW OF THE GRC MODEL2'3
The model developed by GRC for simulating photochemical smog employs
a trajectory-oriented, moving-coordinate-system approach. Figure 3.1 is
a diagram of the concepts used in the model. An air parcel moves along
a path determined by the local prevailing wind speed and direction at
various times during the day. The air parcel is divided vertically into
a mesh of equally spaced points. As the air parcel moves over a region
it receives pollutant emissions from the ground. The pollutants within
the air parcel then undergo chemical reactions and diffusion, with the
incident sunlight driving several of the chemical reactions that occur.
At each mesh point, the concentrations of several species of pollutants
are computed: NO, N0_, hydrocarbons, and ozone, to name a few. The out-
put of the program consists of the species concentrations as functions
of time and height.
\\
TIME-DEPENDENTMIXING
AND REACTION IS COMPUTED
FOR AIR PARCEL UP TO THE
MIXING HEIGHT h
SPACE/TIME TRACK
-THROUGH THE SOURCE
GRID IS DERIVED
FROM WIND DATA
-POLLUTANT INFLUXES AT ANY
ELEVATION (INCLUDING THE
GROUND) ARE IMPOSED BYTHE
EMISSION SOURCE FUNCTIONS
Figure 3.1. Schematic GRC Photochemical Diffusion Model for Air Quality
Simulation
6-10
-------
The model thus consists of three parts. One part computes the tra-
jectory followed by the air mass; another determines the emissions that
enter the air parcel as it sweeps over the region; the third part com-
bines emissions, dispersion, and chemistry to calculate the concentra-
tions of the pollutants along the trajectory.
A key component of the model is the chemical submodel. The GRC chemi-
cal model has been extensively documented in Ref. 3. Briefly, it con-
tains 16 reactions and nine chemical species. The photochemical pollu-
tants of interest in air quality standards are represented: nitrogen
oxides, hydrocarbons, and photochemical oxidants, e.g., ozone. Reactions
involving particles are treated in a lumped-parameter fashion. The air
quality simulation model also computes the concentration of species, such
as carbon monoxide, which are regarded as chemically inert.
The model also has provisions for dealing with elevated sources
such as power-plant stacks. Stack heights and locations can be provided.
as inputs, with the emissions inserted at the appropriate height in the
air parcel when the parcel sweeps over the stack.
Summarizing, the GRC model combines chemistry, emissions, and meteor-
ology to simulate the complex interactions of a multicomponent mixture
of pollutant species. The results are the concentrations of the various
pollutants as functions of time and height.
3.2 CONSIDERATIONS IN AIR QUALITY MODELING
In assessing the air quality impact of electric vehicles using the
GRC model, we will first select several critical locations in the Los
Angeles basin which are known to be afflicted by high levels of air pol-
lution. The chosen locations and the criteria used in the selection are
discussed in Sec. 3.3.
6-11
-------
Having designated the target areas (receptors), air trajectories
will then be obtained which arrive at the specified locations at various
times of the day when contaminant levels are expected to be high. For
example, in downtown Los Angeles peak oxidant readings generally occur
around noon, whereas CO maxima coincide with the traffic peak and thus
occur early in the morning. In this manner we obtain the pollutant
levels as functions of time at each selected location and also along the
path of the trajectory.
Several trajectories per receptor will be computed, with each of
the several trajectories per location arriving at a different time. The
total time interval covered is chosen so that it contains the time when
the highest pollution levels are expected. Because of the geographical
dispersion of the receptors, the trajectories will be chosen so that they
cover nearly the whole Los Angeles basin, thus allowing us to obtain pol-
lution levels on a regional scale. In addition, the emission patterns
sensed by the trajectories will contain a wide variety of mobile and
stationary sources. This ensures that the model is sensitive to changes
in source strengths. Section 3.4 contains additional information regard-
ing the determination of the trajectories.
The meteorological data used to compute the trajectories and to
define atmospheric dispersion parameters are data that were taken in Los
Angeles during the Period August-October 1969. During that period a spe-
13
cial monitoring program was performed in Los Angeles which included
very refined inversion base soundings using aircraft at various times of
day. Such data are not available for any other time. Also, our model
3
validation studies were conducted using the same 1969 data and this
tends to strengthen the confidence in the results obtained. Nevertheless,
since we are following a worst-case approach the question may be raised
whether 1969 is indeed representative of the worst year. Examination of
the statistics of pollution concentration in Ref. 1 for the period 1966-
1970 reveals that the frequency of occurrence of high ozone concentrations
6-12
-------
is remarkably stable throughout this period. Furthermore, although four
alerts were called in 1969 and 1970 had nine alerts, the largest number
in five years, during August-October 1970 there were only five alerts
whereas all four 1969 alerts took place during August-October. In fact,
as is shown in Table 3.1, the August-October statistics for 1969 and
1970 are hardly distinguishable. Thus the available data indicate that
this period of 1969 is comparable to that of any other year during 1966-
1970 and it seems satisfactory to use 1969 meteorological data for our
computations.
Under EPA sponsorship we have recently completed an extensive con-
trolled evaluation of the model by comparing predicted and measured pol-
lutant concentrations. The results of the evaluation are reported in
Ref. 3. The testing procedure involved using six days of air quality
data obtained in Los Angeles in 1969. The data for three of the days
TABLE 3.1*
AUGUST-OCTOBER STATISTICS FOR 1969 AND 1970
1969 1970
Number of alerts 4 5
Number of days with eye irritation 65 57
Number of days on which oxidant > 0.1 ppm hourly
**
average 88 84
**
Number of days on which NO- ^0.25 ppm hourly average 41 40
2 **
Number of days on which CO > 10 ppm for 12 hours 50 52
*
Reference 1.
**
California state standard in force in 1970.
*
LAAPCD first alert level for ozone is 0.5 ppm for five minutes.
6-13
-------
were used to adjust the model's parameters and to develop rules for
operating the model in a so-called "hands-off" mode. The remaining three
days were then modeled using the "hands-off" rules. These rules will
also be applied when using the model to perform the air quality impact
analysis. Because of the extensive model validation tests that have
been carried out, we believe that it is not necessary to conduct addi-
tional testing other than the usual shakedown runs required by any model.
As is the case with any model, in order to obtain meaningful results
care must be taken to ensure that the model is not exercised beyond the
limits imposed by the data base used to validate the model. However,
3
we note that the data base used in our validation studies contained a
large assortment of source strengths and pollutant concentrations and
this reduces the chance of operating the model outside its validation
range in the course of the study. In any event, should the occasion
arise, special procedures will be used to prevent the occurrence of such
a situation.
Summarizing, the procedure followed in using the GRC model to eval-
uate the air quality impact of electric vehicles consists of four subtasks:
1. Selection of critical locations.in the Los Angeles basin
2. Identification of worst-case meteorology
3. Determination and selection of air trajectories
4. Computation of pollutant concentrations
The first three of these subtasks are discussed below.
3.3 SELECTION OF CRITICAL SITES
We have chosen to focus our attention on a few selected sites in
the region because of our interest in worst-case air quality conditions,
a desire to be realistic in our assessments, and in the interest of
economy of operation. However, let us note that concentrating on key
locations does not prevent us from estimating air quality on a regional
6-14
-------
scale. The reason is that the selected sites are the terminal points,
the receptors, of the air trajectories and that pollution levels are com-
puted along the trajectory's path. The criteria, described below, used
in the selection of the receptors ensure that the trajectories will
thoroughly cover the region of interest.
The criteria used in selecting the key sites are:
• Prevalence of high pollution levels
• Geographical dispersion
• The position of the site with respect to the prevailing flow
patterns
In order to satisfy the first criterion, air quality monitoring
data provided by local Air Pollution Control Districts were used to guide
the site selection process. Because of the peculiarities of the geographi-
cal distribution of the various types of source and the different chemical
characteristics of the pollutants of interest, the application of the
first criterion will require that the selected sites be matched to spe-
cific pollutants. Thus all the sites cannot be expected to be equally
suitable for determining worst-case levels for all pollutants.
Oxidant levels are of special concern in site selection since re-
cent abatement strategies proposed for Los Angeles by EPA have been aimed
4
at reducing oxidant concentrations. Hence some sites must be selected
on the basis of their high oxidant readings. This requirement implies
that some sites will be located on the eastern part of the Los Angeles
basin since the highest oxidant concentrations are known to occur there.
In contrast to oxidant, high levels of nitrogen oxides (NO ), CO,
X
and HC are more likely to occur close to the sources and we must turn
to the densely populated central basin to select the appropriate sites.
Large concentrations of sulfur dioxide (SO-) will also be found in the
6-15
-------
vicinity of sources. In Los Angeles the highest S09 levels occur near
the coast where the power plants and oil refineries are concentrated.
The sites should be geographically dispersed in order to obtain
pollution estimates on a regional scale. However, the first and third
criteria may impose restrictions on the geographical dispersion. Wide
spatial separation between sites is also desirable because the air in
the trajectories is thus exposed to a large variety of source
configurations.
The Los Angeles basin possesses a distinct pattern of air flows
during the smog season. In general, the air flows from west to east
during the day and reverses its flow at night, thus creating a so-called
"sloshing" action. Accordingly, the daytime flow of air makes the pol-
lutant concentrations in the eastern basin be functions of the emissions
produced in the western and central parts of the region. This is espe-
cially true of ozone levels because the pollutants react to produce
ozone while on their way to the east basin. In addition, topographical
features such as mountains and canyons contribute to creating distinct
pathways through which air flows preferentially. Because of this guide-
way effect certain downwind locations will bear the brunt of the heavy
emissions produced in such central basin areas as downtown Los Angeles.
14
This phenomenon has also been termed the "pipe reactor" effect.
Recognizing the existence of the flow patterns described above,
the third criterion calls for locating the selected sites on the path of
the prevailing winds. This is consistent with our worst-case approach.
In this manner the computation of secondary pollutants, NO. and 0», is
made sensitive to changes in emissions which occur elsewhere in the
basin. It should be noted that the LAAPCD has located monitoring
stations along the path of the prevailing flows.
6-16
-------
Five sites have been selected following the criteria described
above. Data about the statistics of pollution levels at various sites
were examined and compared in order to ensure that the first criterion
is satisfied. Table 3.2 lists the five sites together with the pollu-
tants which are most significant at each location. Figure 3.2 shows a
map of the region which illustrates the geographical distribution of the
selected sites. Also shown in Fig. 3.2 is a typical afternoon pattern
of wind flows in order to enable the reader to see the relationship be-
tween site location and the flow pattern.
A few comments should be made about the selections. First of all,
we note that only one site has been chosen for estimating SO effects.
Los Angeles has a very stringent program for controlling sulfur emissions
and SQj concentrations become significant at very few places. Long Beach
generally experiences the highest SO^ concentrations. (See also Sec. 4.1
for further remarks on the treatment of SO-.) Because of its coastal
location, Long Beach is not as heavily afflicted with other pollutants
as other points in the region.
TABLE 3.2
SELECTED LOCATIONS IN THE LOS ANGELES AIR QUALITY
CONTROL REGION AND MOST SIGNIFICANT POLLUTANTS AT EACH SITE
Site
Downtown Los Angeles
Azusa
Long Beach
Riverside
Anaheim
X
X
NO
x
X
X
X
X
Pollutant
HC CO Particulates
X
X
X
X
SO,
X
6-17
-------
I
I-1
oo
Figure 3.2. Map of South Coast Air Basin Showing Critical Sites and
a Typical Autumn Afternoon Flow Pattern
-------
Four of the five stations, Long Beach excepted, experience either
high or medium levels of ozone. Riverside is well known for the peak
hourly average of 0.62 ppm which occurred in 1970 and which formed the
4
basis of the EPA abatement strategy already mentioned. Azusa consis-
tently records ozone levels which exceed the standard of 0.08 ppm hourly
average and about 30 percent of the daily maximum hourly readings are in
14
excess of 0.20 ppm. Anaheim also is afflicted by high oxidant levels
during the smog season. Downtown Los Angeles, on the other hand, regis-
ters mostly medium and low levels of ozone, with 0.19 ppm daily maximum
hourly average concentration exceeded less than 10 percent of the time.
In fact, ozone levels in downtown Los Angeles have been shown to be fol-
14
lowing a trend toward lower values over the last six years. Thus Down-
town Los Angeles is not important as far as oxidant is concerned. How-
ever, because of the large number of sources present in the vicinity of
Downtown Los Angeles, we can expect this location to have high concen-
trations of NO , HC, CO, and particulate matter.
X.
Azusa is located on the eastern edge of the basin, at the end of
the "reactor pipeline." By Los Angeles standards, the population density
around Azusa is relatively low, so we do not expect to see high concen-
trations of CO and HC. Particulate matter concentrations tend to be high
because of the formation of aerosols by chemical processes. Levels of
N0« also are high because the air masses arriving at Azusa have had
enough time to perform the chemical conversion of NO to NO^. However,
analysis of concentration data for Azusa has shown that NO levels are
J x
not as high as in downtown. This is not surprising inasmuch as downtown
NO consists mostly of NO whereas at Azusa NO. is the main constituent.
x 2
Since NO- is the more toxic component of NO , we shall be interested in
£• X
NO at Azusa.
x
Riverside, like Azusa, is also the recipient of pollutants trans-
ported from Los Angeles and thus is at the end of another "reactor pipe-
line." Accordingly, it has a high incidence of large oxidant levels as
6-19
-------
well as NO . Anaheim has high concentrations of ozone and NO and, in
X X
addition, consistently records high readings of particulate matter.
3.4 IDENTIFICATION OF WORST-CASE METEOROLOGY
In general, the meteorological conditions which lead to high con-
taminant concentrations consist of a hot day with low wind speeds and a
strong temperature inversion. The emissions are trapped by the inver-
sion and the poor dilution capability of the atmosphere thus results in
high pollution levels. When such conditions exist the atmosphere is
said to be stable.
The worst case occurs when the atmosphere remains stable for several
days in a row. Very high pollutant concentrations then occur due to the
extended accumulation of contaminants. The longest episode of this kind
occurred in Los Angeles in July 1955, when high ozone levels persisted
for twelve days. As might be expected, single-day episodes are far
more common. For example, from 1956 to 1971 there were 213 days when the
maximum hourly ozone level equaled or exceeded 0.40 ppm. Of these 213
days, 103 were single-day episodes.
We wish to be realistic in the choice of meteorological conditions
for modeling purposes. The proposed approach is to use the meteorology
that prevailed during a single-day episode since the atmospheric condi-
tions on such a day are typical of the worst that may be encountered.
In addition, we require that weather and aerometric data of high quality
be available in order to provide reliable inputs to the model.
The day of September 29, 1969, satisfies the conditions mentioned
above. Accordingly, it has been selected as the prototype day for the
worst-case predictions of pollutant levels. Heavy eye irritation was
recorded in Los Angeles on September 29, and an ozone first-level alert .
was called on this day (see p. 6-13 for a definition of alert levels).
Moreover, and not coincidentally, extensive records of meteorological
6-20
-------
data are available for this day as a result of a special experimental
13
program carried out by Scott Research Laboratories. These data include
frequent aircraft soundings which delineate the height of the inversion
layer. For this day, the profiles of atmospheric temperature indicate a
very stable atmosphere during the hours 0600-0800 PST, with subsequent
changes to stable and neutral conditions as the day warms up. The inver-
sion was ground-based in the early morning hours and subsequently lifted
to about 1000 feet above sea level at noon and thereafter as the result
of solar heating. Finally, it should be noted that precisely because of
the properties already mentioned, September 29 was also used as one of
the days for which our model was calibrated during the model validation
3
tests. The experience with September 29, 1969 accumulated in the vali-
dation process will facilitate our modeling task and enhance confidence
in the results.
3.5 SELECTION OF TRAJECTORIES
In selecting the trajectories, the following guidelines have been
used. First, the trajectories must fan out in such a manner as to cover
a large territory. This allows us to obtain regional-scale estimates of
pollution levels. Second, the arrival times at the destination must en-
compass the expected time of occurrence of the maximum concentration of
the pollutant of interest. Third, the trajectory must be long enough so
that it will have a high ratio of mass received from emission sources to
the initial mass. This criterion ensures that the computed concentrations
will be sensitive to changes in emissions, a most important requirement in
our work. Fourth, we want to have trajectories which have very low ini-
tial concentrations of the various species, especially HC, NO, and N0_.
This calls for starting the trajectories during the early morning hours,
preferably before dawn, when pollution levels tend to be low. It is
apparent that the third and fourth guidelines are related since having a
small initial mass of pollutants will enhance the sensitivity to source
emissions. We also note that since the starting point of a trajectory is
generally not going to be in the vicinity of a monitoring station, the
6-21
-------
initial conditions are likely to be inaccurately known. Hence it is im-
portant to reduce as much as possible the model's sensitivity to initial
concentrations.
Meteorological data measured at various monitoring stations on
September 29, 1969, will be used to obtain the trajectories which will
be computed using the trajectory-generation program which is an integral
part of the GRC simulation model. To illustrate, one of the trajectories
to be used in the study is shown in Fig. 3.3. The trajectory begins over
the ocean at 0100 and arrives at Anaheim at 1300, its nodes being spaced
at hourly intervals. This trajectory is receptor-oriented since it
arrives at a given point (Anaheim) at a designated time (1300). In addi-
tion to receptor-oriented trajectories, we will use some trajectories which
originate at a specified point in order to test the effects of increased
emissions of power plants; such trajectories are said to be source-oriented.
All the trajectories used in the study will be shown in the companion re-
port which contains the results of the air quality predictions.
0100
• 1300
ANAHEIM
Figure 3.3. Trajectory for September 29, 1969, Which Arrives at
Anaheim at 1300
6-22
-------
4 FACTORS AFFECTING AIR QUALITY IMPACT OF ELECTRIC VEHICLES
In this section we discuss the treatment of several factors which
are specially significant in determining the air quality impact of elec-
tric vehicles.
4.1 EFFECT OF ELECTRIC CARS ON EMISSIONS FROM STATIONARY SOURCES
With respect to stationary sources, the major impact of electric
vehicles will be on power plant emissions by virtue of the increased
power demand created by battery recharging. Several questions arise in
this context:
• How much new generation will be required for various levels
of electric car use?
• Will nuclear or fossil-fueled plants be used to satisfy the
increased demand?
• To the extent that present and/or future fossil-fueled plants
are used to provide power for electric cars, what effect
will this have on emissions and hence air quality?
The answer to the first question is beyond the scope of this report;
this information will be developed in a separate study task. Neverthe-
less, proceeding on the assumption that some amount of new generation
will be required, we can discuss qualitatively the likely air quality
impacts associated with the second and third questions.
The second question goes straight to the heart of the matter. For
it is clear that air pollution from stationary sources will be profoundly
affected by a decision in favor of either type of plant: nuclear or
fossil-fueled. Obviously, any answer to this question involves judgments
which may, and probably will, be modified by future events and should
be judged in this light.
Discussions with public officials and perusal of the literature
reveal the following scenario. Over the short term, until about 1990,
6-23
-------
nuclear units are not expected to provide a large fraction of the total
generating capacity in Los Angeles. For example, a study by Stanford
Research Institute for all of Southern California, not just Los Angeles,
projects nuclear generating capacity at 11 percent of the total in 1980,
35 percent in 1990, and 71 percent in 2000. It is expected that the ex-
pansion of conventional thermal generating capacity will be accomplished
by the addition of topping units, mostly gas turbines, to existing plants.
Thus conventional thermal units, i.e., gas- and oil-fired, will be the
mainstay in power generation for 1980 and 1990, and will still be very
much in evidence in 2000. A separate task of the study will provide de-
tailed information about fossil-fueled plant locations and capacities.
Going on to the third question, it is apparent from the previous
discussion that a major effect of using electric vehicles would be to
shift emissions from mobile sources to power plants. It should be noted
that the character of the emissions will change since power plants emit
mostly NO and S00 and only very minimal amounts of hydrocarbons. Thus
X Z
there would be a marked overall reduction in hydrocarbon emissions and
a lesser degree of improvement in NO emissions; there could be an in-
X
crease in S0_ emissions unless special measures are taken to prevent it.
Stricter controls must also be placed on the emission of particulates
from power plants, lest we see an increase in such emissions. It should
be noted, however, that the published inventory for Los Angeles shows
that power plants account for only 4 percent of particulate emissions.
The reduction in hydrocarbon emissions is likely to be very important in
terms of controlling the formation of photochemical oxidants. The net
effect of the shift to power plants could be an overall reduction in
emissions since stationary sources are generally easier to control than
cars.
Because of a shortage of low-sulphur fuels, i.e., oil and gas, the .
prognosis for the next several years is for an increase in the use of
high-sulphur fuels for power generation in the Los Angeles Basin. This
6-24
-------
implies, of course, increased S0« emissions unless more stringent emis-
sion controls are placed on the stationary sources. However, it is ex-
pected that this problem is of a short-term nature, lasting only until
1976 or 1977. By 1980, the expectation is that control technology to-
gether with the use of low sulphur oil may keep S0_ emissions per unit
of generated power to low levels.
However, increases in power generation may result in an overall
increase in total SO- emissions from power plants. Emissions of nitro-
gen oxides from power plants may still be substantial by 1980 and beyond.
Hence, localized effects at receptors downwind of the power plants must
be investigated. Since, as is noted in Sec. 5, expansions in generating
capacity are expected to consist of additions to present plants within
the L.A. Basin, we can use air trajectories which sweep over some known
plants to determine the incremental effects on air quality.
A way to meet the increased power demand is to recharge electric
vehicles using existing generating capacity, with increased power genera-
tion during off-peak periods. This has the effect of leveling the peaks
and valleys of the power plant operating curve, mainly by raising the
level of the valleys. As a result, nighttime emissions will be increased.
When stagnant conditions prevail at night, a condition which we have
assumed to exist in the course of our worst-case analysis, the increased
NO emissions will tend to accumulate over the region since chemical
X
activity is very low at night. The early-morning pollutant load of nitro-
gen oxides is thus increased and this will affect the formation of photo-
chemical oxidants later in the day. However, whether more or less oxi-
dant is formed is open to question since the reduced hydrocarbon load
may tend to balance out the increase in NO . In fact, low HC/NO ratios
X X
in the atmosphere inhibit the formation of oxidants. Such effects will
be included in the simulation, especially since we have made an effort
to select trajectories which originate at night.
6-25
-------
As a final note, we wish to point out that many of the judgments
and speculations mentioned above regarding the availability of fuels
need to be placed in the context of the current energy crisis. A thorough
evaluation of the fuel situation is being conducted separately and its
results, when available, may indicate the need to modify some of our
statements and proposed approaches.
4.2 PUBLIC POLICIES AND ELECTRIC VEHICLE USE
The implementation of certain policies by legislative or executive
mandate may produce significant changes in the air quality panorama of
the future. For example, policies may be promulgated that will encour-
age the use of electric vehicles at the expense of conventional autos.
Conversely, other policies may point in the direction of improved air
quality with conventional cars and this could have the effect of reduc-
ing to insignificance the contribution of electrics towards cleaner air.
In this section we examine several policies which have been pro-
24
posed by EPA to reduce air pollution in Los Angeles. Variations of
these policies have been suggested for application to many other cities
across the country. The intent of our review is to evaluate the effect
of the policy vis-a-vis electric car use and to explain how such eventua-
lities are treated in our study.
Broadly speaking, the proposed policies fall in the category of
transportation controls whose aim is to reduce the number of automobile
miles traveled. The main provisions of the policies are:
1. Parking restrictions in the central business district.
2. Incentives for the use of busses and car pools.
In addition to the transportation controls, the policies aim to reduce
emissions by
1. Establishing an annual inspection system for private cars
and light trucks to ensure that emission control devices
work properly.
6-26
-------
2. Retrofitting 1966-1974 model cars with oxidizing catalytic
converters to reduce HC and CO emissions.
The plans also envision providing incentives for the eventual construc-
tion of a mass transit system for the area. Notably absent from the first
phase of the policies is the requirement for gasoline rationing which had
4
been previously advanced. However, the EPA announcement states that
gasoline rationing may be put in effect by 1977 if expected improvements
in air quality do not materialize by then.
.The transportation controls for Los Angeles, without gas rationing,
are expected to reduce the number of vehicle-miles traveled from a minimum
of 14 percent to a maximum of 43 percent.
Several of the proposed transportation controls have been the sub-
ject of a recent study at GRC. Two of the strategies studied which
are most relevant to this discussion are a strategy which would reduce
vehicle-miles traveled by 30 percent uniformly across the L.A. Basin and .
another which would reduce the number of vehicle-miles by 90 percent, in
a 60-square-mile area around the central business district. The first
case clearly falls within the range of the expected reductions in vehicle-
miles. The second case ties in with the proposed parking restrictions in
the central business district.
The results obtained for the 30-percent reduction indicate marginal
improvements in air quality. However, we note that greater leverage might
be achieved with a 30-percent reduction in vehicle-miles in the presence
of more stringent stationary source controls; this might well be the case
in 1980 and beyond. Thus, such a policy could have the effect of pro-
moting enthusiasm for electric cars as a way of reducing the inconvenience
to people who wish to move about at will. Admittedly, other considera-
tions such as reduction of congestion may work in favor of an absolute
reduction in the number of cars on the road, but this is beyond our ken
at this time.
6-27
-------
The strategy which would reduce all traffic by 90 percent in the
downtown area can have significant beneficial air quality effects, but
only in localized areas and under special circumstances. This is not
surprising upon recognizing that the area of 60 square miles is imbedded
in a larger region of 1250 square miles. Thus the sphere of influence
of the central business district is likely to be small. Furthermore,
only those localities which lie downwind of the central business district
will be affected and the effect will thus be randomly distributed since
it is a function of the wind direction and speed, themselves random
quantities. In sum, such a strategy could have some beneficial results,
but the effects will be spread over a relatively small area.
It is possible that the imposition of any kind of transportation
controls on conventional cars, such as those mentioned above, may induce
people to gravitate toward electrics. This would, of course, increase
the level of use of electric cars with a concomitant improvement in air
quality. One way to quantify this is to vary the level of use of electric
cars within the 14-to-43 percent expected reduction in the number of
vehicle-miles traveled by conventional cars.
Establishment of a mass transit system in Los Angeles would have
the effect of reducing the number of individual cars on the road, thereby
cutting pollution. However, the timing for implementation is unclear at
*
this time inasmuch as financing difficulties abound. Furthermore, it is
an open question whether, in the absence of other inducements or restric-
tions, significant effects will materialize. In this respect, discussions
with personnel of the Southern California Rapid Transit District regard-
ing a proposed multibillion dollar mass transit system indicate that even
under the most optimistic assumptions the most that can be expected is a
10% reduction in the number of vehicle-miles traveled by 1990. Unless a
greater impact is achieved, it appears that a mass transit system for
After this was written, Los Angeles residents rejected a 1-percent in-
crease in the sales tax earmarked for mass transit at an election in
November 1974, thereby essentially killing the program.
6-28
-------
Los Angeles will not by itself achieve significant gains from an air pol-
lution standpoint.
Another kind of public policy which will have a profound effect on
the air quality impact of electric cars is the policy which defines and
enforces emission standards for heat-engine-powered cars. Even though
strict standards have been imposed on future cars, uncertainty exists
regarding enforcement and even the standards themselves. One reason for
this state of affairs is that basic scientific knowledge is lacking in
several important areas. Auto manufacturers have been granted exten-
sions of the implementation deadlines and the possibility exists
that the NO emission standard may be relaxed. However, should the pro-
mise of significant emissions reductions materialize, then by 1980 and
beyond we would have a situation in which electrics would replace low-
pollution cars. The beneficial impact of electrics from an air quality
standpoint would then be diminished. Quantifying such an effect involves
manipulating the age distribution of automobiles in establishing the base-
line emission factors. For a given year, the lower bound is determined
by assuming that all the cars are current-year models. This would pro-
duce the lowest possible level of automotive emissions. Such an assump-
tion is admittedly unrealistic and other perturbations of the age distri-
bution may yield more credible results. The section on forecasting base-
line emissions (Sec. 5) describes the procedures that are followed in
establishing bounds for such eventualities.
Another situation in which electrics would compete with relatively
low-pollution conventional vehicles arises out of the requirements for
an annual inspection/maintenance and catalytic converter retrofit program
25
for 1966-1974 vehicles. A recent study by J. Horowitz of EPA claims
reductions of 8 percent in HC emissions and 10 percent in CO emissions
for the total car population by instituting an inspection/maintenance
program for pre-1975 vehicles. These reductions would apply until about
1980 and would decrease to about 2-to-3 percent by 1985 because of the
6-29
-------
turnover in the car population. Much higher claims are made by the same
26
author for the retrofit program of installing oxidizing catalytic con-
verters combined with an annual inspection/maintenance schedule. In this
case potential reductions range from 30 to 50 percent by 1980 and decrease
to 10 to 20 percent by 1985, depending on the pollutant. The retrofit
proposal thus offers the possibility of substantial gains, but at a
relatively high cost to the consumer. For purposes of our study we will
estimate realistic bounds for the expected reductions in emissions from
conventional cars. We then plan to consider such a scenario explicitly
as a special case.
6-30
-------
5 FORECASTING OF POLLUTANT EMISSIONS
5.1 OVERVIEW
The task of predicting pollutant emission patterns for our analy-
sis of potential electric car impact in the years 1980, 1990, and 2000
may be divided naturally into two parts. First, we must forecast the
baseline emission patterns for each of the three target years, i.e., the
expected emissions inventories in the absence of electric vehicle usage
on any significant scale. (Uncertainties in these baseline projections,
due in large part to unpredictable shifts in governmental policy, may
suggest establishment of multiple baseline scenarios for one target year
to demonstrate the extent to which such variations can affect the results
of the analysis.) Then we must assess how the introduction of electric
vehicles alters the baseline pattern in each target year. Recalling
from Sec. 2 the two analysis methods described, we see that two levels
of detail will be required in our emissions forecasts, a gross predic-
tion of daily emissions on a regional basis to support the rollback
analysis, and a detailed distribution of pollutant source contributions
over the region on an hourly time scale as needed by the GRC chemical-
diffusion model.
As will be seen in the following sections, data are available which
allow reasonably detailed emission forecasts for HC, CO, and NO , the
primary pollutants modeled in the GRC code. However, the chemistry of
S0~ and particulates in the urban atmosphere is, as yet, poorly under-
stood. S0« is diminished with time under the action of processes known
to include photo-oxidation. Identification of these processes is one of
the objectives of the current EPA Regional Air Pollution Study (RAPS)
program being carried out in the St. Louis area. Particulates, on the
other hand, are generated in photochemical processes involving both back-
ground chemical species and particles, and anthropogenic pollutants such
as S0? and reactive hydrocarbons. Thus the particulate burden is not
simply the sum of its natural and man-made components. It has been esti-
mated that the Los Angeles aerosol is about one-third natural background,
6-31
-------
one-third anthropogenic primary pollutants, and one-third chemically
formed particulates generated by the presence of the man-made primary
27
species. Unfortunately, urban aerosol chemistry is even less well
understood than SO^ chemistry. We are therefore limited to considering
S0? and particulate impacts via the rollback method (wherein chemical
activity is ignored), and hence, we will only require predictions of SCL
and particulate emissions on an aggregated scale (e.g., tons per day
emitted in the region).
5.2 VEHICULAR EMISSIONS
As configured for the Los Angeles region, the emissions
module of the GRC code aggregates HC, CO, and NO emissions in 2-mile-
X
by-2-mile squares. Vehicular emissions in each grid square are calcu-
lated as functions of average gram-per-mile emission factors for the
vehicle mix, including light and heavy duty vehicles, daily surface
street and freeway vehicle miles traveled (VMT), times of day (with
separate time distributions for surface streets and freeways), and vehi-
cle average speed, itself a function of location and time. In the ab-
sence of emission control strategies which alter the character of traf-
fic peaking, such as staggering of working hours, we will not need to
adjust the time distributions. In order to forecast mobile source emis-
sions, we will need predictions of future traffic patterns which deter-
mine VMT and average speed distributions, and estimates of average
vehicle emission factors for the years 1980, 1990, and 2000.
5.2.1 Vehicle Miles Traveled and Average Speed Distributions
We have at our disposal several methods for estimating future traf-
fic patterns in the Los Angeles area. The simplest approach involves
scaling our current VMT distributions (from Ref. 28, circa 1968) by the
appropriate ratio of future to 1968 vehicle registrations in the area.
This method has obvious drawbacks, primarily that it does not reflect
future highway or mass transit construction, nor could it predict changes
in vehicle speed distributions resulting from the increased VMT. Never-
theless, it is very easy to calculate and would provide a basis for
6-32
-------
judging the reasonableness of VMT results obtained by other methods. We
are also obtaining from the cognizant county agencies overall predictions
of expected vehicle populations. With some engineering judgment, these
data may be used to update our 1968 VMT distributions incorporating knowl-
edge of planned freeway additions. Finally, in connection with our work
for the California Department of Transportation, we are obtaining from the Los
Angeles Regional Transportation Study (LARTS) detailed projections of expected
i traffic in the Los Angeles area separated into freeway and surface streets
VMT by grid square. The LARTS projections are being made for the planned
freeway system as operational in the target years of 1985, 1990, and 2000.
We anticipate that it will probably not be possible to use the LARTS data
directly as their projections of population, per capita vehicle ownership,
and driving habits appear to conflict with our estimates based on past
history and current trends. However, since the LARTS traffic distribu-
tions are calculated from detailed network flow modeling of the planned
street and freeway system, they will provide a good basis for scaling
with predicted total VMT.
Thus our approach for predicting VMT and average speed distributions
is to use the first two techniques mentioned above for the relatively
near term target year of 1980 when shifts in traffic due to new highway
or mass transit construction are not a significant factor, and to use
distributions based on the LARTS projections, but tempered by total VMT
data from the counties and our own population and transportation projec-
tions (see the companion papers, Refs. 29 and 30) for the years 1990 and
2000.
5.2.2 Average Vehicle Emission Factors
The average gram-per-mile pollutant emissions from the vehicle popu-
*
lation in a given year are weighted combinations of emissions from light-
*
For purposes of Federal and State emission control standards, vehicles
weighing 6000 pounds or less are defined as "light duty"; those 6001
pounds and over in gross weight are classified as "heavy duty" vehicles.
6-33
-------
and heavy-duty vehicles of various model years and ages, which are driven
varying numbers of miles in their average use. Since 1966, the pollu-
tant emissions from vehicles sold in California have been subject to
State, and subsequently, Federal limitations by model year. Compliance
with standards is determined by emissions collected from the vehicle dur-
ing operation over a prescribed speed-time schedule referred to as a
driving cycle. It is these driving cycle emission factors which serve
as the common basis permitting meaningful combinations of emissions over
a range of vehicle model years. Current standards are based on the Fede-
31
ral Driving Cycle (FDC) which was designed to simulate average vehicle
operation during peak traffic hours within 7 miles of the central busi-
ness district. The resulting simulated trip covers 7.47 miles at an
average speed of 19.6 mph. Empirical speed correction techniques will
be applied to predict emissions at other speeds under both peak and off-
32 33
peak traffic conditions. '
Given a calendar year for which average vehicle emission factors
are required, the model year emission factors for all prior years are
summed after being weighted by three factors; (1) a deterioration factor
which accounts for losses in effectiveness of emission control systems
with vehicle mileage, (2) the fraction of the daily VMT contributed by
vehicles according to their ages and use, and (3) a speed correction fac-
tor for speeds other than the driving cycle average speed. In general,
all of these factors will be functions of vehicle model year. Of course,
in practice it is not necessary to consider all model years prior to the
calendar year of interest, as after a certain age is reached the mileage
contributions become negligible.
Reference 33, compiled by GRC under the sponsorship of the California
Department of Transportation, details the above described calculations
for various mixes of light and heavy duty vehicles sold in California,
for target calendar years from 1967 to 2000, providing us with baseline
vehicle emission factors of HC, CO, and NO for 1980, 1990, and 2000.
X
6-34
-------
These baseline emissions are predicated on compliance with the current
schedule of emission standards applicable to vehicles sold in California,
and include the effects of the interim California standards for 1975 set
34
by EPA on April 11, 1973 and the one year delay of the 1976 NO emis-
35 x
sion standard granted by EPA on July 30, 1973. It is noted in Ref. 33
that after 1990 heavy duty vehicles will become the dominant contributors
to vehicular emissions even if they satisfy presently contemplated emis-
sion standards, suggesting that it may be desirable to postulate as rea-
sonable the application of some further tightening of heavy duty vehicle
emission standards after 1975. Otherwise, vehicular emissions in 1990
and 2000 will be practically insensitive to reductions in passenger car
emissions.
38
In other calculations, we have assessed the possible impact on NO
emissions of California's program to retrofit NO controls on 1966 through
X
1970 model year cars, and determined that by 1980 the change in average
emissions for the vehicle mix will already be negligible.
Expected trends in vehicle sales and use indicate that the vehicle
age and use distribution will not change noticeably from that employed in
Ref. 33, unless drastic public policy forces such as an emissions tax are
instituted. Such a policy could result in lower average emissions in
1980 due to vehicle population shifts towards newer cars, but by 1990 the
vehicle mix will consist entirely of vehicles with the final '76-'77 or
later controls without any additional government coercion.
Although they are not often discussed, Ref. 1 shows that S07 emis-
sions from vehicular sources in Los Angeles are 35 tons per day. This
is comparable to S0_ emissions from power plants (also 35 tons per day)
and petroleum-related industries (55 tons per day). We will employ gram-
per-mile S0? and particulate emission factors from Ref. 36 to calculate
daily emissions of those pollutants for the planned rollback analysis.
Subsequent work did not assume that heavy duty vehicle emissions would be
controlled more stringently than required by present laws.
6-35
-------
The manner in which our characterization of the baseline vehicular
emissions will be altered by the introduction of electric vehicles will
depend on their role in the vehicle mix. Electric cars will, of course,
emit no hydrocarbons, nitrogen oxides, or sulfur oxides. Hence, as far
as these emissions are concerned, the use of electric cars having limited
top speed capability can be simulated by general reductions of surface
street VMT in proportion to the electrics' penetration of the small car
market (taking into account the fraction of vehicle trips which have no
freeway segment). Evaluating the air quality effect of an advanced elec-
tric capable of most or all of the urban trips now performed by conven-
tional cars would involve recalculation of the average vehicle emission
factors for the target calendar years after appropriate adjustment of the
emission factors for previous model years in which electrics were sold.
Emissions of particulates due to tire wear, brake linings, etc., will be
assumed to be the same as for conventional autos. Particulate emissions
in exhaust fumes will, of course, be eliminated in electric cars and
appropriate adjustments will be made to account for this.
5.3 STATIONARY SOURCE EMISSIONS
The GRC photochemical-diffusion model employs separate specifica-
tions of stationary point source emissions from power plants and oil re-
fineries, and aggregates emissions from distributed stationary sources
(e.g., gasoline marketing, dry cleaning, solvent coating, etc.) by grid
square. Each of the three stationary source types has its own diurnal
time distribution which determines the fraction of total daily emissions
occurring in each hour of the day. In this section we will outline the
procedures to be employed in predicting daily emissions of HC, NO , S09,
X fc-
and particulates from these three source categories, and their spatial
and temporal distributions. Stationary source emissions of CO may be
safely neglected since they contribute about 0.1% of the total daily CO
burden in the L.A. Basin.
6-36
-------
5.3.1 Power Plants
37
In a companion report, we have developed baseline forecasts of electric
power demand in the Los Angeles area which are compatible with our fundamen-
tal assumptions regarding population growth. Reference 37 supports three
assumptions which are of importance to our estimation of future power
plant emissions in Los Angeles and environs:
1. Except for those already under construction, no new power
plants (nuclear or thermal) will be located in the SCAB. Any
additional power generated in the region will be produced by
uprating existing units, by retrofiting existing units with
*
combined cycle capability, and by construction of a few
extra units at established thermal plants.
2. Natural gas is not expected to be available in any signifi-
cant amount for power generation, as the available supply
will probably be allocated to meet increasing residential
demands. Hence, the thermal plants in the region are expected
to be almost entirely oil fired after 1980. (Currently,
increased residential demand for natural gas forces power
plants and industry to use low sulfur fuel oil during the
winter months. Residential demand from April to November
is generally low enough to permit power plants and industry
to use natural gas.)
3. While the availability of low sulfur oil (^0.5 percent by
weight) may be a problem in the mid-1970s, the economics and
technology of desulfurization indicate that low sulfur oil or a
non-petroleum equivalent should be in ready supply by 1980
and for the remainder of the forecast period.
By the year 2000, nuclear stations sited outside the basin are expected
to provide the major fraction of electrical generation capacity, meeting
*
Combined cycle refers to systems employing a combination of the Brayton
and Rankine cycles, i.e., a direct combustion turbine (like a jet engine
turbine) is employed both to drive a generator directly and to produce
hot exhaust gases which are used to boost the output of the basic steam
turbine unit.
6-37
-------
most of the continuous basic demand together with some hydroelectric
sources. Early in the forecast period, fossil fueled plants (some coal
fired plants outside the SCAB and the oil fired facilities within the
region) will be providing some of the base load power and all of the
peaking capacity. Figure 5.1 shows our baseline projection for the peak
demand to be met by the generating facilities that really concern us
here, the oil fired plants sited in the SCAB. As the availability of
nuclear power increases, less and less of the oil fueled plant capacity
will be used to satisfy the base load. Figure 5.2 plots the resulting
daily power cycle in the peak demand month (August) for the years 1980,
1990, and 2000. These curves provide the baseline time distribution
functions needed for our detailed forecasts of power plant emissions.
Using data obtained from Southern California Edison (SCE), the Los Angeles
Department of Water and Power (LADWP) and the cities of Glendale, Pasadena,
and Burbank to predict individual site capacities, we will distribute
the peak demand (Fig. 5.1) geographically in proportion to those capacities.
20
15
5 10
BASELINE PROJECTION:
NO ELECTRIC CARS
1970
1980
1990
2000
YEAR
Figure 5.1. Baseline Projected Peak Demand for Electricity Generated
by Oil Fired Plants in the South Coast Air Basin
6-38
-------
20
15
o 10
S
O£
LU
g
Q_
3 5
1980
1990
2000
BASELINE PROJECTION:
NO ELECTRIC CARS
0600
1200
TIME
1800
2400
Figure 5.2. Projected Baseline Diurnal Power Demand on Oil Fired Power
Plants in the Los Angeles Area for Peak Demand Month (August)
Reference 36 gives the emission factors shown in Table 5.1 for oil
fueled power plants. Hydrocarbon emissions from power plants contribute
less than 0.2 percent of the HC burden and hence are considered negligible.
Thus, all that remains is to translate megawatts of power generated
into gallons of oil burned, and we have completed our baseline forecast
TABLE 5.1
OIL FIRED POWER PLANT EMISSION FACTORS,
POUNDS PER THOUSAND GALLONS
36
NO,
SO,
Steam Boiler 105 157(S)
Combustion Turbine 120 142(S)
(S = percent by weight of sulfur in the oil)
Particulates
8
8.4
6-39
-------
of power plant emissions. Of SCE's current generating capacity in the
Los Angeles region, only 10 percent comes from gas turbine generators,
so that the overall operating efficiency of their systems is dominated
by that of the older steam plants. On the average, about 36 percent of
the fuel energy input is realized as electricity output. According to
SCE forecasts, this efficiency is not expected to improve noticeably as
new capacity comes on line. Therefore we plan to assume that all oil-
fired units are of the steam type for purposes of emissions calculations.
Using a 36 percent conversion efficiency and a fuel energy content of
145,000 Btu/gal we find that the average fuel consumption is 65.4 gal/MW-hr.
It should be noted that these emission factors do not include the effects
of any emission control efforts, e.g., off-stoichiometric or two-stage
combustion, flue gas recirculation, low excess air, etc. Where calcula-
tion of the uncontrolled emissions yields a result which would exceed a
Federal, State or local emission standard, the emissions will be assumed
to just meet the standard, reflecting the application of some unspecified
control technique.
When we have completed our baseline air quality calculations, we
will need to modify the diurnal power demand curves of Fig. 5.2 to incor-
porate the added demand associated with recharging of electric car
batteries. Most of the added demand will take the form of a continuous
load starting at 10:00 or 11:00 pm and ending at 6:00 or 7:00 am. A small
fraction of the total recharging load will be spread uniformly over the
remainder of the day. The magnitude of the total load will depend on the
number and types of the electric vehicles in each scenario, but a pre-
37
liminary calculation shows that sufficient off-peak generation capacity
will exist to service over five million cars (out of a projected vehicle
population of 6.7 million cars) in 1980 and more in the later years.
5.3.2 Oil Refineries
On a regional basis, emissions from oil refineries are relatively
small, but since refineries are concentrated sources their emissions
OQ
can be important on the local scale. The 1968 data base used in the
GRC code simply prorates total oil refinery emissions as reported by
6-40
-------
the Los Angeles County APCD to individual refinery locations according
to their crude oil capacities. Current uncertainties as to future oil
supplies in the US make it difficult to forecast oil refinery activities
with confidence, but we will take the view that any real energy shortage
which can be eased by the expansion of refining capacity will probably
lead to relaxations of environmental restraints opposing that expansion.
Therefore, reasoning that it is national demand which will be the prime
consideration in determining gasoline and fuel oil production, we will
scale oil refinery emissions in the SCAB with the projected national popu-
lation to arrive at an emissions forecast for 1980, 1990 and 2000. Emis-
sions to be treated in this fashion include reactive HC, NO , S0_ and
X. £-
particulates.
5.3.3 Area Sources
Included in the distributed source category are activities ranging
from industrial and commercial operations to residential living. Within
the LAAPCD, emissions from these sources contribute from 9 to 42 percent of
total emissions (except for CO) depending on the pollutant. However,
because they are spread over large areas, these emissions are relatively
unimportant on the local level. We will use the regional population as
a basis for scaling total area source emissions to 1980, 1990, and 2000
levels. For the detailed analysis, we intend to use general land-use
plans from the various counties (Los Angeles, Orange, Riverside, and
San Bernardino) as a guide for projecting the distribution of area source
emissions. We will correlate our present emissions distribution with
local zoning using three-to-five zone types (e.g., industrial, residen-
tial-commercial, residential), and then use these emissions/zoning rela-
tionships to construct future emission distributions from planning maps.
Only one category of distributed source emissions is directly af-
fected by the infusion of electric cars into the vehicle mix, that being
the petroleum marketing area. In 1970, hydrocarbon emissions from vari-
ous vehicle and underground tank filling operations contributed about
6-41
-------
13 percent of all HC emissions from stationary sources. This amounted to
less than 4 percent of the total HC burden in 1970, but as baseline ve-
hicular emissions reach their minimum levels, around 1990, the importance
of these emissions will be roughly doubled. Therefore in scenarios where
the electric car fraction of the vehicle mix is large enough to reduce
gasoline sales significantly from their baseline levels (say 20 percent
or more) we will have to adjust area source HC emissions accordingly.
6-42
-------
6 AIR QUALITY IMPACT MATRIX
In previous sections we discussed the relationship between electric
vehicles and air quality and outlined an approach designed to quantify
this interdependence. The heart of the method consists of using an air
quality simulation model to relate pollutant emissions and air quality.
This section describes the actual cases considered in the simulation.
The cases considered are shown in Table 6.1. In selecting the cases
we have gone beyond a strict parametric analysis (Cases I-III) in an at-
tempt to evaluate the effects of certain public policies on the air qual-
ity impact of electric cars. The various cases also serve to check the
sensitivity of the results to the assumptions inherent in the analysis.
It is inevitable that a study of this kind will always raise questions
about a multitude of situations that are worthy of consideration. The
TABLE 6.1
MATRIX OF AIR QUALITY IMPACT CASES
(Numbers shown in table are the priorities assigned to the various cases.)
Year
Case 1980 1990 2000
I Baseline 111
II Low Level of Electric Car Use 221
III High Level of Electric Car Use 112
IV High Level of Power Plant Emissions 3
V Effectiveness of EPA Transportation Controls 3
VI Low-Pollution Conventional Cars 4
VII Delayed Emission Controls for Conventional
Cars 4 4
Cases finally considered in the analysis were Cases I, III, and VII. The
emissions range covered by these cases encompasses the remaining cases
and it was deemed unnecessary to carry those out.
6-43
-------
cases included in Table 6.1 are the result of our posing our own "What
if...?" questions and then selecting what appear to be the most important
cases. The scenarios have been matched to the time period when we have
judged that the maximum impact would be felt. The numbers shown in Table
6.1 represent the priority assigned to each case.
Referring to Table 6.1, we note that the first three cases form
the basis of the parametric analysis'. Case I, the baseline case, assumes
that no electric cars are in use. The forecasting methodology used to
establish emissions levels for Case I has been described in Sec. 5. Cases
II and III are perturbations of the baseline case in that the baseline
emissions are altered by the introduction of certain low and high levels
of electric vehicle use. The specific use levels will be determined in
*
a separate task of the study. Note that in 1980 and 1990 Case II gets
a lower priority than Case III, and that this is reversed in 2000. The
reason for this is that if the high level of electrics in 1980 and 1990
fails to yield significant gains in air quality, then it is probably un-
necessary to consider Case II since the gains, if any, would be even
smaller. In 2000 the situation is reversed in that the "low" level of
electrics may be such as to improve air quality to within ambient stand-
ards and thus an even higher level of electrics would be known to meet
standards by implication.
Case IV considers the possibility that future power plant emissions
of S09 and NO may be higher than expected. This could come about if
tL X.
there are delays in the importation of natural gas, unexpected difficul-
ties with coal gasification, or a possible lack of effectiveness of S0_
and NO emission control technology. As indicated in Table 6.1, we would
expect the major impact of such eventualities to occur in 1980.
Case V considers the effectiveness of the transportation controls
24
for Los Angeles recently promulgated by EPA. Without gasoline rationing,
*
See also the footnote on p. 6-4.
6-44
-------
it is expected that the new controls would reduce VMT by a minimum of 14
percent and a maximum of 43 percent. We want to know what effect this
would have on electric vehicle use and on air quality. For example, a
plausible scenario is that these measures may exceed the threshold of in-
convenience of people and that this would act to increase the use of
electrics. Accordingly, we may see significant gains in air quality im-
provements when the EPA measures are combined with electric vehicles.
Case VI has been included in an attempt to answer the question:
What would be the air quality impact of electric vehicles if somehow the
emissions of conventional cars can be made lower than is presently con-
templated? If such an event materializes we would have electrics compet--
ing with low-emission vehicles and the air quality gains due to electrics
could be diminished. For example, such a circumstance might come about
if a stiff emissions tax were placed on all vehicles. This would encour-
age car owners either to eschew cars altogether or to buy newer, and pre-
sumably low-pollution, conventional cars, thereby reducing overall emis-
sion levels. Of course, such a policy might also have the effect of en-
couraging the purchase of electrics. We have judged the maximum impact
of such an eventuality to occur in 1980, since by 1990 and 2000 techno-
logical changes together with a normal turnover of the auto population
would indeed be placing electrics in competition with low-pollution con-
ventional vehicles.
The last case, Case VII, is the converse of Case VI. It has been
included in response to the current efforts to delay the implementation
of vehicular emission controls. In fact, one-year extensions have already
been granted both for CO and HC controls to 1976 and for NO controls to
22 23
1977. ' The attractiveness of electric cars from an air quality stand-
point would of course be enhanced under these circumstances.
6-45
-------
7 ANALYSIS AND DATA DISPLAYS
In this section we describe the analysis of the air quality compu-
tations and introduce the devices we expect to use to display the results.
The Clean Air Act, as amended, required the promulgation of primary
and secondary air quality standards. Primary standards are designed to
protect human health whereas secondary standards are intended to protect
welfare. At the present time standards exist for six major pollutants;
these are shown in Table 7.1. These standards are subject to revision as
new information becomes available; in fact, the NO standard is currently
18 ^
under critical scrutiny. Such scrutiny notwithstanding, air quality is
presently defined in terms of the standards.
Our basic problem is to determine air quality as a function of elec-
tric vehicle use for the three target years. Thus for each of the six
pollutants of interest we will obtain a plot similar to that shown in
*
Fig. 7.1. Plotted on the graph will be the highest concentration of the
pollutant computed at any point in the region of interest. The concen-
trations obtained using the rollback equation will be tabulated and com-
pared with those produced by the GRC model. Since emissions form the
basis for the application of rollback, mobile and stationary source emis-
sions for the various cases will be displayed in tabular form. As is
shown in Fig. 7.1, by superimposing the air quality standard on the graph
we can determine the level of use which attains the standard.
It is useful to display the information on Fig. 7.1 in a different
*
way as indicated in Fig. 7.2. In Fig. 7.2 we plot the normalized concen-
tration of each pollutant versus the level of electric car use for the
year y , where the normalized concentration is defined as the pollutant
concentration divided by its air quality standard. Thus the normalized
Later analysis showed that the perturbations on air quality induced by
electric cars are relatively small. Accordingly, we departed from a
strict parametric approach and considered only upper bounds on electric
car use.
6-46
-------
TABLE 7.1
NATIONAL AMBIENT AIR QUALITY STANDARDS
19
Level Not to Exceed
Pollutant
SO,
Primary
80 Mg/m (0.03 ppm)
3
365 Mg/m (0.14 ppm)
3
75 Mg/m (c)
260 Mg/m3 (b)
10 mg/m (9 ppm)
3
40 mg/m (35 ppm)
160 ug/m3 (0.08 ppm)
160 ug/m3 (0.24 ppm)
3
100 pg/m (0.05 ppm)
Secondary
3
60 Mg/m (0.02 ppm)
3
260 Mg/m (0.1 ppm)
3
60 Mg/m
150 yg/m3
Same
Same
Same
Same
(a)
(b)
(c)
(b)
(d)
(e)
(e)
(f)
(a)
Particulate Matter
Carbon Monoxide
Photochemical Oxidants
Hydrocarbons
Nitrogen Oxides
(a) Annual arithmetic mean.
(b) Maximum 24-hr concentration not to be exceeded more than once a
year.
(c) Annual geometric mean.
(d) Maximum eight-hour concentration not to be exceeded more than once
a year.
(d) Maximum one-hour concentration not to be exceeded more than once a
year.
(f) Maximum three-hour concentration (6-9 am) not to be exceeded more
than once a year.
6-47
-------
o
Q.
AIR QUALITY
STANDARD
1980
1990
2000
PERCENT MILES TRAVELED BY ELECTRIC VEHICLES
Figure 7.1. Typical Plot of One Species Concentration Versus
Electric Car Fraction
et
o;
o
Q.
Q
cC
s:
o
YEAR y
AIR QUALITY
STANDARD SATISFIED
PERCENT MILES TRAVELED BY ELECTRIC VEHICLES
Figure 7.2. Plot of Normalized Pollutant Concentration Versus Electric
Car use for a Particular Year. Each Curve Corresponds to
one Pollutant
6-48
-------
concentration is unity when the air quality standard is achieved. Figure
7.2 allows us to compare at a glance the varying degree of effectiveness
which the use of electric cars has on the six pollutants. In addition,
because the concentration is normalized, we can read directly from the
ordinate the fractional change in air quality that occurs when the use
of electrics increases. If we let H. , i = l, 2 6 , be the
level of electric car use which meets the standard for the ith pollutant,
then, clearly, L = max(£.. , _, ..., £,) is the use level which attains
all the standards for the year y . The value of L is also placed in
evidence in Fig. 7.2. The values of L obtained with the GRC model and
y
with rollback will also be compared.
The second part of the air quality impact analysis consists of ob-
taining estimates of overall air quality for various levels of electric
car use. A parameter which is useful in this respect is the fraction of
the time which we can expect the air quality to satisfy the standards.
Another useful indicator is the median pollutant concentration, i.e.,
the concentration which is exceeded half the time. The two indicators
will be displayed in tabular form. However, we caution that the useful-
ness of these parameters will be determined by the size of their respec-
tive confidence intervals.
6-49
-------
6-50
-------
APPENDIX
DERIVATION OF AIR QUALITY STATISTICS
A.I INTRODUCTION
As mentioned in the text, it is desired to obtain statistical infor-
mation about the general quality of the air from the predicted worst-case
pollutant concentrations. In particular, Sec. 6 of the text described
two indicators which appear to be especially useful in this respect.
The discussion that follows describes the methodology used to determine
these indicators.
The first step in deriving air quality statistics is to define the
functional form of the probability of occurrence of pollutant concentra-
7 8
tions. Larsen ' has shown that pollutant concentrations are generally
distributed lognormally. We note, however, that we will process Los
Angeles air quality data to test whether the lognormality hypothesis holds
and shall use the proper distribution function to obtain the desired air
quality predictions. For completeness, Sec. A.2 and A.3, respectively,
describe properties of the lognormal distribution and the estimation of
its parameters. Section A. 4 then explains the method used to determine
the desired statistics, including a discussion of the sources of error
associated with the derived quantities.
20 21
A.2 THE LOGNORMAL DISTRIBUTION '
Let X be a random variable whose distribution is gaussian with
2
mean y and variance a and define a new random variable Y by
A A.
Y = exp(X) (A.I)
6-51
-------
Then Y is said to be lognormally distributed. The probability density
function (pdf) of Y is readily obtained from the gaussian pdf and is
given by
p(y) -- - - exp[- | (log
a v/2if
y - O/a 5 (7 > 0) (A. 2)
Thus, as defined here, the lognormal distribution is completely specified
*
by two parameters. Using the properties of the moment-generating func-
tion of the gaussian pdf we can ascertain that the mean and variance
of Y are given by
(A-3)
x)lexp(a)
X'
(A. 4)
2 r 2
PY|exp(ax) -
In our case, the pollutant concentration corresponds to Y and it
is convenient to reverse the process by defining X = log Y . The gaus-
sian distribution is then defined in terms of the parameters of the log-
2
normal function by solving Eqs. A. 3 and A. 4 for u and a as functions
„ XX
of y and O , viz. ,
1 22
log V - !og(l + S/^ (A. 5)
ox = iog(i + C^/PY) (A. 6)
*
It is also possible to generalize the definition of the lognormal pdf
to include three parameters, but this is not pertinent for our purposes.
See Refs. 20 and 21 for a more detailed explanation.
6-52
-------
although these equations may not necessarily be the best estimators of
the mean and variance.
Let us now define a unit normal variable U by the linear transfor-
mation
U = (X - MX)/OX (A.7)
The pdf of U is gaussian with zero mean and unit variance. Making
use of Eq. A.I we can express Y as a function of U as shown below
log Y = yx + 0XU (A.8)
Clearly, Eq. A.8 defines a straight line with intercept yi and slope
A
a . Since a probability is associated with each value of U , then for
A
graphical purposes Eq. A.8 also defines the cumulative distribution func-
tion of log Y . Note that when U = 0 then P[U <_ 0] = 0.5 . Hence
the median value, i.e., the 50th percentile, of Y is given by
median Y = exp(y ) (A.9)
A
Thus the "intercept" of the line is actually located at the 50th percentile
Consider a sample of n values of Y , and let those values be de-
fined by y , i = 1 n . Then the sample geometric mean is defined
by
n \l/n
TTy, (A. 10)
1 /
6-53
-------
However, we have that
n
1 ^ J = nu (A. 11)
where m is the sample mean of the gaussian variable. Hence, from Eqs,
A.10 and A.11
m = exp(iO (A. 12)
If the actual mean of the gaussian distribution is equal to the sample
mean then y = m^ and w
lognormal distribution by
mean then y = m^ and we can formally define the geometric mean of the
y = exp(y ) (A.13)
g x
Comparing Eqs. A.9 and A.13 it can be seen that the median of Y is
equal to the geometric mean of the distribution. Moreover, from Eqs.
A.3 and A.13 we can see that
yy > y (A.14)
hence the percentile associated with the mean of Y is greater than that
associated with its geometric mean, the difference in percentile being a
function of a .
A.
We have now defined some of the relevant properties of the lognor-
mal distribution and its relationship with the gaussian function. It
remains to use these results in a setting related to air quality. This
brings us to the task of estimating the parameters which specify the
distribution using the worst-case predictions from our simulations.
6-54
-------
A. 3 ESTIMATION OF PARAMETERS
A.3.1 General
Two parameters suffice to specify the density function of pollutant
concentrations: y and a . Eq. A.8 tells us that this amounts to
X X
specifying the intercept and the slope of a straight line. We note, how-
ever, that the air quality simulations do not attempt to predict these
two quantities. Instead, the simulations predict worst-case pollutant
concentrations. Since any point on the line qualifies as the intercept,
we need not worry about estimating y . We can thus use the predicted
A
worst-case concentration as the intercept provided we assign a probability
of occurrence to it.
Estimating the slope is another matter. The simulations will not
help us to obtain statistically significant estimates of the slope because,,
for economic reasons, the size of the sample available for estimation is
too small. An independent means of estimating the slope is thus necessary
and we shall use experimental data obtained by the Los Angeles Air Pollu-
tion Control District for this purpose.
A.3.2 Use of Worst-Case Predictions
The worst-case predictions of the air quality model provide a pol-
lutant level to be located on the line defined by Eq. A. 8. The question
is: what probability of occurrence shall we assign to the predicted
worst-case concentration?
One way to resolve this question is to analyze aerometric data to
determine two pieces of information:
1. Given a predicted concentration, what is the historical
frequency of occurrence of that concentration?
2. Do the data show a time trend for this frequency of
occurrence?
6-55
-------
This information places the prediction in the context of historical air
pollution patterns and, by extrapolating the time trend, we can in turn
put the prediction in a future setting. The uncertainties associated
with the latter course are well recognized, however, and error bounds
will be obtained for the projected probability.
An alternative, simpler method is to assign an a priori probability.
To illustrate, the assigned probability could be based on considerations
such as the fact that the meteorological conditions being used, i.e., a
day in which a pollution episode occurs, have had a frequency of occur-
rence of 213/5840 (cf. p. 21). Other similar criteria may also be used
to establish the desired probability. Which of the two alternative
methods is best will be determined in the course of the analysis.
A.3.3 Estimating the Slope
Estimating the slope of the line is synonymous with determining
2
the variance, a . The estimate will be obtained independently from
A
the model's predictions using the aerometric data of Los Angeles. The
2
estimation of a is a standard problem in statistics and many methods
A
exist for this purpose (cf. Ref. 21, Ch. 5). The maximum likelihood
estimator
-,
' x)
is especially convenient since it allows us to obtain exact confidence
2
intervals for a . The confidence interval is given by
A.
where X-, and x9 a^e suitable percentage points for the chi-square
distribution with (n - 1) degrees of freedom. For example, for a 90%
6-56
-------
2 2
confidence interval we would have P(a < Xi) = 0.05 and P(a < x9) = 0.95 ,
22 21
where a = (n - l)s /a . Aitchison and Brown have shown that the maxi-
X
mum likelihood estimator is also very reliable with respect to other
estimators.
2
It is also necessary to determine any time trend which a may
A
follow and we will test the data for the presence of any such trend.
A.4 DERIVATION OF AIR QUALITY INDICATORS
A.4.1 Method
The two indicators sought are:
1. The fraction of the time that air quality standards are
exceeded
2. The median pollutant concentration
These two quantities are to be obtained for each pollutant for the seve-
ral scenarios considered in the study. Their determination is a simple
matter once the probability density function has been specified. By using
log/probability paper we can plot a straight line in accordance with Eq.
A.8 and the desired quantities can then be read off the graph. The graph-
ical approach is also advantageous for displaying the error bounds.
A.4.2 Causes of Error
The accuracy of the indicators is determined by a combination of
the following factors:
1. Accuracy of the model's predictions.
2. Errors in estimating the frequency of occurrence of worst-
case concentration.
3. Errors in estimating the standard deviation of the distribu-
tion function.
4. Uncertainties in extrapolating time trends.
6-57
-------
One of our main concerns is the way the various errors combine to
affect the final result. Another item of interest is the sensitivity of
the result to the various errors. Since the combination of errors can
become complex, our preference is to report the results in the form of
intervals to which, hopefully, a statistical confidence level can be
assigned. By introducing perturbations in the various quantities in-
volved we can obtain measures of sensitivity as well as assist in deter-
mining the size of the interval. If warranted, it may be useful to deter-
mine sensitivity coefficients for each of the quantities involved.
6-58
-------
REFERENCES
1. Los Angeles Air Pollution Control District, Profile of Air Pollution
Control, 1971.
2. A. Q. Eschenroeder, J. R. Martinez, "Concepts and Applications of
Photochemical Smog Models," in Advances in Chemistry, Series 113,
American Chemical Society, 1972.
3. A. Q. Eschenroeder, J. R. Martinez, and R. A. Nordsieck, Evaluation
of a Diffusion Model for Photochemical Smog Simulation, General
Research Corporation CR-1-273, October 1972.
4. Environmental Protection Agency—Region IX, Technical Support Docu-
ment for the Metropolitan Los Angeles Intrastate Air Quality
Control Region, January 15, 1973.
5. A. Q. Eschenroeder, Comments on California Air Quality Standards:
Transportation Control Strategy, General Research Corporation
IM-1741, April 1973.
6. D. S. Earth, "Federal Motor Vehicle Emission Goals for CO, HC, and
NOX Based on Desired Air Quality Levels," J. Air Poll. Control
Assoc., Vol. 20, No. 8, pp. 519-523, August 1970.
7. R. I. Larsen, "A New Mathematical Model of Air Pollutant Concen-
tration Averaging Time and Frequency," J. APCA, Vol. 19, No. 1,
pp. 24-30, January 1968.
8. R. I. Larsen, A Mathematical Model for Relating Air Quality Measure-
ments to Air Quality Standards, U.S. Environmental Protection Agency,
Office of Air Programs, Publication No. AP-89, November 1971.
9. R. I. Larsen, private communication, Air Pollutant Concentrations
as a Function of Averaging Time and Frequency, 1962-1968, unpublished
tables of California measurements, April 1971.
10. N. R. Patel, "Comment on a New Mathematical Model of Air Pollution
Concentration," JAPCA, Vol. 23, No. 4, April 1973, pp. 291-292:
11. J. B. Knox, R. I. Pollack, An Investigation of the Frequency Dis-
tributions of Surface Air-Pollutant Concentrations, Lawrence
Livermore Laboratory, UCRL-74063, October 30, 1972.
6-59
-------
REFERENCES (Cont.)
12. H. Shoji, T. Tsukatani, "Statistical Model of Air Pollutant
Concentration and Its Application to the Air Quality Standards,"
Atmos. Env.. Vol. 7, pp. 485-501 (1973).
13. Scott Research Laboratories, 1969 Atmospheric Reaction Studies in
The Los Angeles Basin, Vols. I-IV, February 16, 1970.
14. G. C. Tiao, G.E.P. Box, W. J. Hamming, Analysis of Los Angeles
Photochemical Smog Data: A Statistical Overview, Dept. of Statis-
tics, University of Wisconsin, Madison, Technical Report No. 331,
April 1973.
15. J. C. Mosher, E. L. Fisher, M. F. Brunelle, Ozone Alerts in Los
Angeles County - 1956-1971, Los Angeles Air Pollution Control
District, September 1972.
16. Stanford Research Institute, Meeting California's Energy Require-
ments . 1975-2000, May 1973, p. 368.
17. J. R. Martinez, R. A. Nordsieck, A. Q. Eschenroeder, Impacts of
Transportation Control Strategies on Los Angeles Air Quality,
General Research Corporation CR-4-273, May 1973.
18. Federal Register, Vol. 38, No. 110, June 8, 1973.
19. Federal Register, Vol. 36, No. 84, April 30, 1971.
20. N. L. Johnson and S. Kotz, Continuous Univariate Distributions - 1,
Houghton Mifflin Co., 1970, pp. 112-136.
21. J. Aitchison and J.A.C. Brown, The Lognormal Distribution, Cambridge
University Press, 1963.
22. Federal Register, Vol. 38, No. 80, April 26, 1973.
23. Federal Register, Vol. 38, No. 161, August 21, 1973.
24. Federal Register, Vol. 38, No. 217, November 12, 1973.
25. J. Horowitz, "Inspection and Maintenance for Reducing Automobile
Emissions," JAPCA, Vol. 23, No. 4, April 1973, 273-276.
26. J. Horowitz, "The Effectiveness and Cost of Retrofit for Reducing
Automobile Emissions," JAPCA, Vol. 23, No. 5, May 1973, 395-418.
6-60
-------
REFERENCES (Cont.)
27. G. M. Hidy and S. K. Friedlander, "The Nature of the Los Angeles
Aerosol," 2nd Clean Air Congress, IUAPPA, Washington, D.C.,
December 1970.
28. P.J.W.Roberts, P. M. Roth, and C. L. Nelson, "Contaminant Emissions
in the Los Angeles Basin—Their Sources, Rates, and Distribution,"
Appendix A of Development of a Simulation Model for Estimating
Ground Level Concentrations of Photochemical Pollutants, Systems
Applications, Inc., Report 71SAI-6, March 1971.
*
29. G. Houser, Population Projection for the Los Angeles Region, 1980-
2000, General Research Corporation RM-1842, November 1973.
**
30. W. Hamilton and G. Houser, Transportation Projection for the Los
Angeles Region, 1980-2000. General Research Corporation RM-1858
November 1973.
31. Federal Register. November 10, 1970, Vol. 35, No. 219, Part II,
p. 17311.
32. A. H. Rose, Jr., R. Smith, W. F. McMichael, and R. E. Kruse,
"Comparison of Auto Exhaust Emissions in Two Major Cities," Journal
of the Air Pollution Control Association. Vol. 15, No. 8, August 1965.
33. R. A. Nordsieck, Estimates of Pollutant Emission Factors for
California Motor^ Vehicles; 1967-2000, General Research Corporation
RM-1849 (in publication).
34. Federal Register, April 26, 1973, Vol. 38, No. 80, p. 10317.
35. Federal Register, August 21, 1973, Vol. 38, No. 161, Part I,
p. 22474.
36. Compilation of Air Pollutant Emission Factors (Second Edition),
U.S. Environmental Protection Agency, AP-42.
37. A. R. Sjovoid, Electric Energy Projections for the Los Angeles
Region, 1980-2000, General Research Corporation RM-1859 Nov. 1973
38. J. R. Martinez, Testimony on the NOX Retrofit Program at a Public
Hearing of the California Air Resources Board, Los Angeles,
California, October 30, 1973.
*
Task Report 2.
**
Task Report 3.
Task Report 5.
6-61
-------
6-62
-------
TASK REPORT 7
AIR QUALITY IMPACTS OF ELECTRIC CARS
IN LOS ANGELES
J.R. Martinez
R.A. Nordsieck
-------
ABSTRACT
Using electric vehicles has been proposed as a means of alleviating
air pollution from automotive sources. This work investigates the changes
in air quality which would occur in Los Angeles if various fractions of
the conventional car population were replaced by electrics during the
period 1980-2000. Primary pollutant emissions of both reactive and
inert species are estimated for the study area, in the absence of electric
cars, and under maximum electric vehicle mileage fractions of 20, 80, and
100 percent in 1980, 1990, and 2000. Emissions tradeoffs from vehicular
to power plant sources are explicitly considered. Because of emissions
shifts to power plants and the increased leverage of stationary sources
in later years, only carbon monoxide shows the full benefit of electric
car use. Hydrocarbon emissions are reduced 7 to 19 percent, particulates
4 to 21 percent, nitric oxide 1 to 12 percent, while sulfur dioxide in-
creases 7 to 14 percent.
Air quality levels during the study period with and without the use
of electrics are assessed by two methods: (1) assuming direct proportion-
ality with emissions (linear rollback), and (2) simulation in the photo-
chemical smog model (DIFKIN) developed at General Research Corporation,
which includes chemical reactions and atmospheric transport effects.
DIFKIN predicts maximum reductions of about 20 percent in ozone and
nitrogen dioxide levels at the higher electric car fractions. Federal
air quality standards for carbon monoxide and nitrogen dioxide are met
during the study years under presently contemplated emission control
schedules without electric cars, but ozone and sulfur dioxide standards
are not. Ozone levels continue to exceed the Federal standards in spite
of reductions approaching 20 percent due to electric cars. Increased
power plant activity with the electric car recharging demand tends to
increase sulfur dioxide levels, already above the statutory limits.
-------
ii
-------
CONTENTS
SECTION PAGE
ABSTRACT i
1 INTRODUCTION 7-1
2 FORECASTS OF POLLUTANT EMISSIONS 7-2
2.1 General 7-2
2.2 Impact of Public Policies 7-4
2.3 Vehicular Emission Factors 7-5
2.4 Stationary Emissions 7-7
2.5 Projections of Vehicle Miles Traveled 7-8
2.6 Baseline Pollutant Emissions 7-8
3 AIR QUALITY IMPACT OF ELECTRIC CARS 7-14
3.1 General 7-14
3.2 Changes in Emissions Due to Electric Cars 7-14
3.3 Air Quality and Electric Car Use 7-28
4 CONCLUSIONS 7-41
APPENDIX A POLLUTANT EMISSIONS ESTIMATES AND PROJECTIONS FOR
THE LOS ANGELES REGION 7-45
APPENDIX B TRAJECTORIES OF AIR MASSES USED WITH DIFKIN AIR
QUALITY MODEL 7-95
REFERENCES 7-113
iii
-------
iv
-------
ILLUSTRATIONS
NO. PAGE
2.1 Emissions Model Study Area 7-3
3.1 Projected Diurnal Power Demand on Oil Fired Power Plants
in the Los Angeles Area 7-16
3.2 Plot of Nitric Oxide Emissions in Los Angeles and
Environs Assuming No Delays in Auto Emission Controls 7-21
3.3 Plot of Reactive Hydrocarbon Emissions in Los Angeles
and Environs Assuming No Delays in Auto Emission Controls 7-22
3.4 Carbon Monoxide Emission in Los Angeles and Environs
Assuming No Delays in Auto Emission Controls 7-23
3.5 Sulfur Dioxide Emissions in Los Angeles and Environs 7-24
3.6 Particulate Emissions in Los Angeles and Environs 7-25
3.7 Total Nitric Oxide Emissions in Los Angeles and Environs
Under Delayed Implementation of Auto Emission Controls 7-26
A.I Emissions Model Study Area, Southern California 7-46
A.2 VMT Study Areas in Southern California < 7-50
A.3 Estimated Total VMT Growth Factors for the South Coast Air
Basin 7-51
A.4 Geographical Distribution of Freeway VMT in 1970, Thousands
of Miles 7-53
A.5 Geographical Distribution of Surface Street VMT in 1970,
Thousands of Miles 7-54
A.6 Geographical Distribution of Freeway VMT in 1980, Thousands
of Miles 7-55
A.7 Geographical Distribution of Surface Street VMT in 1980,
Thousands of Miles 7-56
-------
ILLUSTRATIONS (Cont.)
NO. PAGE
A.8 Geographical Distribution of Freeway VMT in 2000, Thousands
of Miles 7-57
A.9 Geographical Distribution of Surface Street VMT in 2000,
Thousands of Miles 7-58
A.10 Los Angeles Traffic/Time Distributions 7-59
A.11 Variation of Total HC Emissions with Traffic Speed 7-63
A.12 Variation of CO Emissions with Traffic Speed 7-64
A.13 Variation of NO Emissions with Traffic Speed 7-65
x *
A.14 Baseline Projected Peak Demand for Electricity Generated
by Oil Fired Plants in the South Coast Air Basin 7-69
A.15 Projected Baseline Diurnal Power Demand on Oil Fired Power
Plants in the South Coast Air Basin 7-69
A.16 Growth of Distributed Source NO Emissions 7-84
x
A.17 Geographical Distribution of Stationary Source NO
Emissions in 1971, kg/hr X 7-92
A.18 Geographical Distribution of Stationary Source Reactive HC
Emissions in 1971, kg/hr 7~93
B.I Trajectory of Air Mass Arriving at Anaheim at 1200 PDT 7-96
B.2 Trajectory of Air Mass Arriving at Anaheim at 1300 PDT 7-97
B.3 Trajectory of Air Mass Arriving at Anaheim at 1400 PDT 7-98
B.4 Trajectory of Air Mass Arriving at Azusa at 1300 PDT 7-99
B.5 Trajectory of Air Mass Arriving at Azusa at 1400 PDT 7-100
B.6 Trajectory of Air Mass Arriving at Azusa at 1500 PDT 7-101
B.7 Trajectory of Air Mass which goes by Riverside and Starts in
Orange County at 1000 PDT 7-102
vi
-------
Illustrations (Cont.)
NO. PAGE
B.8 Trajectory of Air Mass which goes by Riverside and Starts in
Orange County at 1100 PDT 7-103
B.9 Trajectory of Air Mass which goes by Riverside and Starts in
Orange County at 1200 PDT 7-104
B.10 Trajectory of Air Mass Departing from El Segundo Power
Plant at 0100 PDT 7-105
B.ll Trajectory of Air Mass Departing from Redondo Beach Power
Plant at 0100 PDT 7-106
vii
-------
viii
-------
TABLES
NO. PAGE
2.1 California Exhaust Emission Standards for Passenger Cars,
grams per mile 7-5
2.2 Baseline Projected Average Electrical Power Generation by
Fossil-Fueled Units Within the Study Area 7-8
2.3 Projected Daily Vehicle Miles Traveled in the Study Area 7-8
2.4 1980 Baseline Pollutant Emissions for Los Angeles and Environs 7-9
2.5 1990 Baseline Pollutant Emissions for Los Angeles and Environs 7-9
2.6 2000 Baseline Pollutant Emissions for Los Angeles and Environs 7-10
2.7 1980 Baseline Vehicular Emissions for Los Angeles and Environs
with and without Delays in Implementing Auto Emission Controls 7-10
2.8 Ratio of Vehicular Emissions to Total Emissions for 1980-2000
Without Electric Cars 7-12
2.9 Fractional Contribution of Heavy Duty Vehicles to Vehicular
and Total Baseline Emissions for Los Angeles and Environs 7-13
3.1 Upper Bounds of Electric Car Use 7-15
3.2 Ratio of In-Basin Electric Energy Demand with Electric Cars
to Baseline Energy Demand 7-17
3.3 1980 Emissions with 20 Percent Electric Cars in Los Angeles
and Environs 7-17
3.4 1990 Emissions with 80 Percent Electric Cars in Los Angeles
and Environs 7-18
3.5 2000 Emissions with 100 Percent Electric Cars in
Los Angeles and Environs 7-18
3.6 1980 Emissions with 20 Percent Electric Cars and Delayed
Implementation of Exhaust Emission Controls for Conven-
tional Cars in Los Angeles and Environs 7-19
ix
-------
Tables (Cont.)
NO. ; PAGE
3.7 Ratio of Total Emissions with Electric Cars to Baseline
Emissions 7-19
3.8 Ratios of HC and NO Emissions with Electric Cars to Baseline
Emissions for the Air Trajectories Used in the DIFKIN Air
Quality Model 7-28
3.9 Baseline Values of Maximum Hourly Average Ozone in Vicinity
of Riverside 7-29
3.10 Ratio of Maximum Hourly Average Ozone with Electric Cars to
Baseline Ozone Level 7-30
3.11 Maximum Hourly NO- for Baseline Cases 7-33
3.12 Ratio of Maximum Hourly NO- Concentration with Electric Cars
to Baseline NO- Concentration 7-35
3.13 Increases in Nitric Oxide Emissions Caused by Electric Car
Battery Recharge in Power Plants in Path of Air Trajectories 7-36
3.14 Baseline Maximum Pollutant Concentrations Estimated by Linear
Rollback 7-38
3.15 Ratio of Pollutant Concentration with Electric Cars to Base-
line Concentration 7-39
A.I Diurnal Variations of Source Activities 7-60
A.2 Average Freeway Flow Speeds and Direction-Volume Ratios for
SCAB Grid 7-61
A.3 Exhaust Emission Standards for Light-Duty Vehicles 7-66
A.A Estimated Daily Vehicular Emissions of SO- and Particulates
in Study Area 7-67
A.5 Future Power Plant Time Functions for Oil Fired Units in the
South Coast 7-70
A.6 Power Plant Data 7-71
A.7 Power Plant Capacity Schedule 7-72
-------
Tables (Cont.)
NO. PAGE
A.8 1971 Emissions From Power Plants in the Study Area 7-72
A.9 Power Generation Emission Factors 7-74
A.10 Projected Pollutant Emissions from Power Plants in the
Study Area 7-75
A.11 Geographic Distribution of NO Emissions from Power Plants 7-76
A.12 Pollutant Emissions from Petroleum Refineries in the Study
Area 7-77
A.13 Petroleum Refinery Data 7-78
A.14 Geographic Distribution of NO and Reactive HC Emissions
from Petroleum Refineries 7-79
A.15 Distributed Source Emissions of S0? and Particulates, 1971 7-80
A.16 HC and NO Emissions from Distributed Sources in Los Angeles
County, X 1971 7-81
A.17 HC and NO Emissions from Distributed Sources in Orange
County, X 1971 7-82
A.18 HC and NO Emissions from Distributed Sources in San
Bernardino County, 1971 7-82
A.19 Reactive HC Emissions Reductions Under the SIP 7-84
A.20 Hypothetical Reductions in Reactive Hydrocarbon Emissions for
1971 Under Full Application of the SIP 7-86
A.21 Land Use Areas in the Study Region 7-88
A.22 Study Area Emissions of NO and Reactive HC from Stationary
Sources x 7-89
A.23 Land-Use Emission Factors in the Study Area, 1971 7-91
C.I Study Area Emissions for 1970-71 7-107
C.2 1990 Baseline Study Area Emissions with Delayed Implementa-
tion of Exhaust Emission Controls for Conventional Cars 7-108
xi
-------
Tables (Cont.)
NO. ; PAGE
C.3 2000 Baseline Study Area Emissions with Delayed Implementa-
tion of Exhaust Emission Controls for Conventional Cars 7-108
C.4 1990 Study Area Emissions with Delayed Implementation of
Exhaust Emission Controls for Conventional Cars and 80%
Electric Cars 7-109
C.5 2000 Study Area Emissions with Delayed Implementation of
Exhaust Emission Controls for Conventional Cars and 100%
Electric Cars 7-109
xii
-------
1 INTRODUCTION
This report describes the results of a study to determine the
impact on air quality caused by using various amounts of electric cars
rather than conventional cars entirely in future years in the Los Angeles
basin. An extensive description of the approach followed to determine
the air quality impact of electric cars is found in Task Report 6.
The study focuses on the years 1980, 1990, and 2000. For each of
these years, projections of automotive and stationary emissions of air
pollutants were determined under the assumption that no electric cars
would be in use. These emissions were then adjusted according to the
number of electric vehicles on the road which was determined from other
studies.
The projected emissions were related to air quality using two
different methods. One method uses the linear rollback formula, where
air quality is presumed to vary linearly with emissions. The features
and shortcomings of this approach are discussed in Refs. 1-3. The second
method employs an air quality simulation model (DIFKIN) developed at
General Research Corporation (GRC) under the sponsorship of the Environ-
4 5
mental Protection Agency (EPA). ' In essence, DIFKIN mathematically
simulates atmospheric diffusion and chemistry to compute the concentrations
of air pollutants such as ozone and nitrogen oxides using primary pollutant
emissions of nitric oxide and reactive hydrocarbons as inputs.
The projected pollutant emissions for the years considered are
discussed in Sec. 2. The air quality impacts of electric vehicles are
described in Sec. 3, and Sec. 4 discusses the implications of the study.
7-1
-------
2 FORECASTS OF POLLUTANT EMISSIONS
2.1 GENERAL
Baseline forecasts of pollutant emissions were made for 1980, 1990,
and 2000. These baseline projections assume that no electric cars would
be in use in any of the three target years. Basically, the projections
consist of estimates of emissions of reactive hydrocarbons (HC), nitric
oxide (NO), carbon monoxide (CO), sulfur dioxide (S0?), and particulate
matter. Each of these is associated with various land use and transpor-
tation patterns in the impacted area. Accordingly, the forecasts are
based on regional transportation and land use plans prepared by govern-
mental agencies in southern California. These various plans have been
collected and summarized in Task Reports 2 and 3. '
The emissions estimates take two forms. One form consists of an
aggregated areawide source inventory. In this case, all pollutant emis-
sions of the same type, e.g., CO, are added together regardless of their
geographical distribution. The second form preserves the geographical
distribution of the sources. To do this, the study area (shown in Fig.
2.1) has been partitioned into a grid of 2-mile-by-2-mile squares, and
all the sources within each square have been averaged over the area of
the square. For each square, we have stationary and mobile source emis-
sions, the latter being dependent on the number of vehicle miles traveled
(VMT) in the road segments within the square and the former on the land
use mix, e.g., industrial, residential, park, etc. Traffic information
was obtained from the California Department of Transportation (CALTRANS),
District 7, and the Los Angeles Regional Transportation Study (LARTS).
Land use distribution was based on projections by the Southern California
Association of Governments (SCAG). Appendix A describes in detail the
data sources and methods used to estimate and extrapolate the inventory
of pollutant emissions for the study area.
7-2
-------
8| |10
20| |22| i24i |26i |28i |30| |32| |34| |36i i38| |40i i42| |44|
10' M21 '14' '16' H81 '20l '221 '24' '261 128' '30' '32' !34l 136' '38'
Figure 2.1. Emissions Model Study Area
i
u>
-------
2.2 IMPACT OF PUBLIC POLICIES
To determine baseline emission levels, it is necessary to assess
the direction of public policies which affect future emissions by insti-
tuting various abatement strategies. Two sets of such policies, both
related to the Clean Air Amendments of 1970 and subsequent further amend-
ments, provide the main thrust for emission control efforts. The first
policy is the federal legislation which sets maximum levels of emissions
a
from automobiles. The second is the federal legislation which requires
9
each state to submit a plan designed to control emissions. These con-
trols are needed in order to meet the national ambient air quality stand-
ards within each air quality control region (AQCR) in the nation. The
Los Angeles area is one of these AQCRs. The plan submitted by a state is
known as the state implementation plan (SIP) and is aimed at controlling
stationary and mobile source emissions. Where vehicle emission controls
alone are not expected to achieve the goals on schedule, land use and
transportation controls are mandated. Examples of the latter include
vehicle inspection programs, parking restrictions, and mass transit pro-
grams. Stationary emissions are controlled, for example, by regulating
types of fuel that can be burned and by reducing evaporative losses at
gas stations, to name but two proposed schemes. The law also imposes
emission performance standards on all new stationary sources based on
available technology.
Two scenarios emerge from these general public policy guidelines.
The main scenario used in our baseline emission forecasts for all three
target years is that auto emission controls are implemented as mandated
by the law as amended up until August 1973, when interim standards for
nitrogen oxides (NO ) emissions were promulgated by the Environmental
8X
Protection Agency. Further delays have since been proposed in the
implementation of auto emission controls; hence, this scenario can be
considered optimistic.
The second scenario considers the effect of the proposals that would
delay for several more years the installation of auto emission controls
7-4
-------
and thus is the more pessimistic scenario. These subsequent amendments
would retain 1975 interim auto emission standards for HC and CO through
the 1977 model year. In addition, we have adopted the more pessimistic
interpretation of the proposed amendments regarding NO emissions: NO
X X
emission standards would remain at the 1974 level of 2 grams per mile
forever. The second scenario has been targeted only for 1980 because the
main impact of the delays will be felt that year and we shall designate it
as case 1980d.
Regarding the SIP, under both scenarios we have assumed that the
abatement strategies proposed by California for stationary sources will
be as effective as forecast in the working documents developed by the
state. A detailed breakdown of the assumed reductions may be found in
Appendix A, Sec. A.6.
2.3 VEHICULAR EMISSION FACTORS
Table 2.1 shows the passenger car exhaust emission standards assumed
to prevail in California for the period 1972-1977. Our baseline forecasts
of automotive emissions use measured emissions data through the 1971 model
year and the standards in Table 2.1 from 1972 on. No further changes are
TABLE 2.1
CALIFORNIA EXHAUST EMISSION STANDARDS FOR PASSENGER CARS, grams per mile
1972
1973
1974
1975
1976
1977
HC
3.2
3.2
3.2
0.9
0.41
0.41
CO
39
39
39
9.0
3.4
3.4
NO
X
3.0
3.0
2.0
2.0
2.0
0.40
7-5
-------
shown after 1977 because the law currently does not require additional
controls after that year. Already included in the table is the effect
of one-year delays in implementing the final HC and CO controls (delayed
from 1975 to 1976) and the final NO controls (delayed from 1976 to
X
1977), the delays having been promulgated in April and July 1973 for the
former and in August 1973 for the latter.
Under the amendments to the Clean Air Act proposed in March of
1974 (see Ref. 10), the HC and CO standards shown in Table 2.1 would re-
main at the 1975 level through 1977. The values shown for 1977 would then
come into effect in 1978. For NO , the more pessimistic interpretation
X
of the proposed amendments is that the NO standard will remain at
2.0 grams per mile forever.
Finally, we note that the actual total vehicular emissions in a
given year depend on the vehicle age distribution, the distribution of
vehicular speeds, the amount of heavy-duty vehicles in the population,
as well as on the emissions from vehicles in each model year. Thus com-
puting vehicular emissions is a rather involved task. The effective
vehicular emission factors for California for various years have been
computed in Ref. 11. In computing these emission factors we have included
the effect of deteriorating emission controls as a function of vehicle age.
Information about the deterioration factors used for the various model years
is also contained in Ref. 11. The emission factors used in our two scenarios
have been determined following the methods established by Nordsieck and
assume that heavy-duty vehicles contribute 5 percent of the VMT (see
Appendix A, Sec. A.3). A private communication from the California Depart-
ment of Transportation (CALTRANS) indicates that the fraction of VMT
contributed by heavy-duty vehicles in Los Angeles ranges between 5 and 10
percent. The lower figure was used in order to be consistent with other
work performed by CALTRANS. Using a 10-percent contribution of VMT from
heavy-duty vehicles would increase baseline total vehicular emissions in
1980 by 18 percent for HC, 14 percent for NO , and 27 percent for CO.
X
7-6
-------
The percentage of vehicular and total emissions contributed by the 5-percent
heavy-duty vehicle component is discussed in Sec. 2.6. Reference 11 con-
tains tables of vehicular emission factors with various fractions of heavy
duty vehicles in the populations, ranging from zero to 100 percent.
2.4 STATIONARY EMISSIONS
Stationary emissions are composed of HC, NO , particulates, and S0_.
X £•
Stationary sources of CO can be neglected since they contribute only about
12
0.1 percent of the total daily CO burden in the area. Our main concerns
are the respective contributions of power plants, oil refineries, and
sources such as residential and industrial areas. The SIP impacts most
heavily on HC emissions from area sources, the expectation being a 66-percent
reduction in HC emissions from area sources from 1969 to 1980. Because of
lack of data after 1980, we have assumed that no further changes in total
HC emissions will take place. Power plants are discussed below; baseline
estimates of total emissions for all sources are given in Sec. 2.6; and
details of the emissions distributions are found in Appendix A.
2.4.1 Power Plants
Power plant emissions consist of NO , SO-, and particulates; HC
X £
emissions are negligible and the NO is essentially NO. Of primary
X
interest in this area are the expected changes in electric power demand
which is supplied by fossil-fueled units. Projections of the fractions
of electrical power demand supplied by several kinds of units, e.g.,
13
nuclear, fossil-fueled, hydro, are found in Task Report 5. Table 2.2
shows the projected changes of in-basin fossil-fueled electric power genera-
tion without electric cars. From Table 2.2 it can be seen that this type
of power generation increases slightly from 1980 to 1990 and decreases
thereafter. The decrease is due to the increasing reliance on power
from remote sources as well as on nuclear-powered units. Since power
plant emissions are directly proportional to the generated power, they
will follow the general trend indicated in Table 2.2. The estimation of
these emissions is discussed in Appendix A.
7-7
-------
TABLE 2.2
BASELINE PROJECTED AVERAGE ELECTRICAL POWER
GENERATION BY FOSSIL-FUELED UNITS WITHIN THE STUDY AREA
1980
8,450
(megawatts)
1990
8,700
2000
7,350
2.5 PROJECTIONS OF VEHICLE MILES TRAVELED
The vehicular emissions are a direct function of the miles traveled
by each vehicle. Table 2.3 shows the daily total of VMT in the study
area for the three years of interest. The increases in VMT are of course
related to expected increases in population as well as to changes in
driving patterns. The traffic data base and interpolation methods used
to synthesize these VMT data are described in Appendix A.
2.6 BASELINE POLLUTANT EMISSIONS
Tables 2.4-2.6 contain the estimated values in tons/day of the pol-
lutant emissions from the various sources for 1980, 1990 and 2000. In
addition, Table 2.7 shows for 1980 the estimated changes in vehicular
emissions caused by the newly suggested delays in the enforcement of auto
emission controls which were proposed in March 1974 and which were discus-
sed in Sees. 2.2 and 2.3. We emphasize that the hydrocarbon emissions
consist of reactive hydrocarbons. Under our definition of reactivity,
only benzene, acetylene, and C..-C. paraffins are considered to be
TABLE 2.3
PROJECTED DAILY VEHICLE MILES TRAVELED IN THE STUDY AREA
(Thousands of Miles)
Freeways
Surface Streets
Total
1980
64,933
84,109
149,042
1990
73,408
88,625
162,033
2000
82,251
93,311
175,526
7-8
-------
TABLE 2.4
1980 BASELINE POLLUTANT EMISSIONS FOR LOS ANGELES AND ENVIRONS
(Tons/Day)
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
NO
181.5
106.0
113.3
48.8
449.6
HC
160.3
160.7
—
33.7
354.7
CO
2202.8
—
—
—
2202.8
so2
22.4
195.3
362.0
60.6
640.3
Particulates
91.2
96.9
12.6
11.0
211.7
TABLE 2.5
1990 BASELINE POLLUTANT EMISSIONS FOR LOS ANGELES AND ENVIRONS
(Tons/Day)
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
NO
75.5
140.8
116.2
53.7
386.2
HC
72.3
160.7
—
37.1
270.1
CO
1084.2
—
—
—
1084.2
so2
24.4
213.1
372.0
66.7
676.2
Particulates
99.2
105.7
12.9
12.2
230.0
nonreactive. This is more stringent than the criteria used by the Los
Angeles Air Pollution Control District, which define benzene, acetylene,
C..-C, paraffins, and several other members of the paraffin and cyclo-
paraffin family as nonreactive< The stricter definition of reactivity
used in this report has been documented in Ref. 14.
The vehicular emissions of NO, HC, and CO shown on the tables re-
flect the distribution of mileage traveled by vehicles of various ages.
7-9
-------
TABLE 2.6
2000 BASELINE POLLUTANT EMISSIONS FOR LOS ANGELES AND ENVIRONS
(Tons/Day)
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
NO
80.6
181.1
98.1
57.6
417.4
HC
77.1
160.7
—
39.8
277.6
CO
1152.1
—
—
—
1152.1
so2
26.4
227.7
315.0
71.5
640.6
Particulates
107.5
112.9
10.9
13.0
244.3
TABLE 2.7
1980 BASELINE VEHICULAR EMISSIONS FOR LOS ANGELES AND ENVIRONS
WITH AND WITHOUT DELAYS IN IMPLEMENTING AUTO EMISSION CONTROLS
(Tons/Day)
With Delay
Without Delay
NO
300.6
181.5
HC
168.5
160.3
CO
2328.2
2202.8
These figures also include the effect of the various speeds which normally
prevail throughout the region as well as a 5-percent component of heavy duty
vehicles such as buses and trucks. It is noted that these emissions de-
crease from 1980 to 1990 in spite of the increase in VMT. This is due
to the constant addition of new cars with stringent emission controls.
The situation is reversed from 1990 to 2000 because during this decade
the increase in VMT outstrips the controls. Nevertheless, the vehicular
emissions by 2000 are still relatively small compared to 1970.
The SO- and particulate emissions for vehicles were computed by
applying emission factors of 0.13 g/mi for S09 and 0.54 g/mi for
7-10
-------
particulates (including particulates from the exhaust and tire wear) for
light duty vehicles and 0.26 g/mi and 0.85 g/mi for SO- and particulates,
respectively, for heavy duty vehicles which were obtained from Ref. 15.
These emissions increase throughout the study period in proportion to VMT
because so far no legislation is contemplated to control them.
Tables 2.4-2.6 reflect the fact that hydrocarbon emissions from
stationary area sources are expected to remain constant through the study
period. On the other hand, the NO from the same sources is expected to
increase by 70 percent from 1980 to 2000.
The power plant emissions increase from 1980 to 1990 and decline
thereafter in accordance with the planned power generating schedule shown
on Table 2.2. The stringent NO , S0_, and particulate emission controls
X ^
used in Los Angeles have been taken into account in estimating the future
emissions from power plants. Finally, emissions from oil refineries are
expected to increase due to a projected expansion of refining capacity in
1980-2000.
Details of the techniques and assumptions employed in projecting
future power plant and oil refinery emissions may be found in Appendix A.
Referring to Table 2.7, it can be seen that the proposed delays in imple-
menting HC and CO controls cause only very slight increases in total emis-
sions in 1980: 5 percent for HC and 5.7 percent for CO. We recall that
these delays are only through 1977, hence by 1980 we will have had two years
of the stricter controls. By contrast, the NO emissions with the delay have
been increased substantially, about 66 percent. This is due to the fact
that the NO emission standard has been assumed frozen at 2.0 g/mi
X
rather than decreasing to 0.4 g/mi as originally shceduled.
Table 2.8 shows the fraction of the total emissions of NO and HC
which is due to vehicular sources. Note that the vehicular contribution
to NO emissions decreases throughout the period. For HC, the contribu-
tion from vehicles decreases from 1980 to 1990 and increases slightly by
7-11
-------
TABLE 2.8
RATIO OF VEHICULAR EMISSIONS TO TOTAL EMISSIONS
FOR 1980-2000 WITHOUT ELECTRIC CARS
NO
HC
1980
0.40
0.45
1990
0.20
0.27
2000
0.19
0.28
2000. It is also noteworthy that the vehicular contributions are rela-
tively small by 1990 and 2000. Thus, on an aggregated areawide basis,
stationary emissions of NO and HC loom large in the future total emissions
picture of the Los Angeles area. This is significant because the
electric cars will be operating on a rather small portion of the total
emissions and their leverage on air quality will be attenuated accordingly.
It should be mentioned that the stricter definition of hydrocarbon reacti-
vity causes the stationary hydrocarbons to assume a greater importance
relative to mobile source emissions than they have been assigned in the
past. For example, the fraction of total reactive hydrocarbon emissions
attributed to vehicles in 1970 is 85 percent under the less strict defini-
tion of reactivity and 65 percent under the stricter reactivity criteria.
Using the less strict reactivity definition the ratios shown in Table 2.8
for HC would change to 0.73, 0.55, and 0.56 for 1980, 1990, and 2000,
respectively. Thus on an aggregated basis the leverage exerted on HC by
electric cars is affected by the definition of reactivity. However, in
many locations in the basin the emissions are dominated by vehicular
sources and in those cases the impact of introducing electric cars can
be significant.
Finally, some remarks are in order about the differential effects
on emissions and air quality due to heavy duty vehicles. Table 2.9
shows the fractional contribution of heavy duty vehicles to vehicular and
total baseline pollutant emissions under the assumption that heavy duty
vehicles account for 5 percent of total VMT.
7-12
-------
TABLE 2.9
FRACTIONAL CONTRIBUTION OF HEAVY DUTY VEHICLES TO VEHICULAR AND TOTAL
BASELINE EMISSIONS FOR LOS ANGELES AND ENVIRONS
NO
Vehicular
1980
1990
2000
1980d
0.
0.
0.
0.
18
31
31
11
Total
0.07
0.06
0.06
0.06
HC
Veh
0.23
0.33
0.33
0.22
Tot
0.10
0.09
0.09
0.10
Veh
0.31
0.67
0.67
0.29
CO
Tot
0.31
0.67
0.67
0.29
Veh
0.09
0.09
0.09
0.09
S02
Tot
0.003
0.003
0.004
0.003
Particulates
Veh
0.08
0.08
0.08
0.08
Tot
0.03
0.03
0.03
0.03
It should be mentioned in connection with Table 2.9 that present laws do
not envision placing stricter controls on heavy duty vehicles as is being
done to automobiles. The relatively uncontrolled trucks and buses thus
contribute a large share of future vehicular emissions.
From Table 2.9 it is apparent that trucks and buses have a major
impact on CO emissions in 1980 and beyond. This is the case even though
in our calculations heavy duty vehicles account for only 5 percent of
the VMT. Such a large component of CO emissions from heavy duty vehicles
will naturally reduce the impact that electric cars can have since they
are unlikely to supplant heavy duty vehicles.
The effect of heavy duty vehicles on NO and HC emissions is moderate,
but tends to further erode any gains that may be attributed to electric
cars. Heavy duty vehicles affect only slightly the vehicular and total
emissions of S0_ and particulates.
7-13
-------
3 AIR QUALITY IMPACT OF ELECTRIC CARS
3.1 GENERAL
Our approach for determining air quality levels of photochemical
pollutants consists of selecting various sites in the South Coast Air
Basin which are presently afflicted with heavy air pollution. The sites
selected are shown in Fig. 2.1. Meteorological conditions for a day
(Sept. 29, 1969) in which an air pollution episode occurred were used to
compute worst case pollutant concentrations at the selected locations
using the DIFKIN air quality model. The pollutant levels were calculated
by following the path of air masses which traverse the air basin and
arrive at the selected sites at various times of the day. The air trajec-
tories used in the computation are found in Appendix B. We also obtained
estimates of pollutant levels by means of the linear rollback method.
(See Ref. 1 for additional explanations regarding our methodology.)
The first step in the analysis is to calculate baseline air quality
levels which may prevail in 1980, 1990, and 2000 without electric cars.
Introducing electric vehicles perturbs emissions by reducing vehicular
contributions and increasing the emissions from power plants; new air
quality levels are then computed and compared with the baseline levels.
The following sections discuss the components of the analysis.
3.2 CHANGES IN EMISSIONS DUE TO ELECTRIC CARS
The changes in total emissions caused by electric cars have been
computed on the basis of upper bounds on electric car use. Table 3.1
shows these upper bounds for the various years. The criteria used to
determine these upper limits are diverse and are found in Ref. 16. The
use level of electric cars can be varied parametrically up to the upper
bound of Table 3.1 and the emissions for these various use levels can be
approximated by interpolating linearly between the baseline and upper
bound emissions.
7-14
-------
TABLE 3.1
UPPER BOUNDS OF ELECTRIC CAR USE
Cars, Millions
Daily Mileage, Millions
Percent of Total Daily Mileage
Percent Electric Cars
1980
1.2
28
19
20
1990
5.4
123
76
80
2000
7.6
167
95
100
In introducing electric cars into the vehicle population, we have
assumed that electrics replace conventional cars on a one-for-one basis
with regard to VMT. Thus, for example, if electrics replace 20 percent
of the cars in the population, the VMT due to conventional cars is reduced
by 20 percent also. Since vehicular emissions are directly proportion to
VMT, the emissions due to cars will also be reduced by the same percentage.
It should be noted that such one-for-one VMT replacement is only
an approximation, especially in 1980 when electrics have a limited range
and would be replacing second cars which account for a smaller fraction
of the total VMT than their numbers indicate. Thus, as Hamilton has
shown, in 1980 replacing 17 percent of the conventional cars in the popula-
tion affects only 11 percent of the VMT. For 1990 and 2000, the greater
range of the electrics makes the one-for-one approximation more valid. A
further complication is introduced by the presence of heavy-duty vehicles,
assumed to contribute 5 percent of the VMT. In our analysis we have
assumed that electrics do not replace heavy duty vehicles. Certainly,
the one-for-one VMT reduction will yield air quality levels which represent
the best we could possibly expect by substituting electrics for conventional
vehicles. We can then interpolate between the baseline and upper limit
air quality levels to estimate the effect of smaller changes in VMT.
Linear interpolation should yield accurate estimates since, as we shall
see, the perturbations in air quality levels are not large.
7-15
-------
Using electric cars instead of conventional cars will reduce vehicular
emissions, but will increase emissions from power plants. The latter
effect is due to the recharging of batteries. This has been assumed to
occur continuously throughout the day, with 90 percent of the total re-
charging load occurring between 11:00 p.m. and 9:00 a.m. and the remaining
10 percent spread uniformly over the period 9:00 a.m.-11:00 p.m. Figure 3.1
shows the resulting diurnal variation of power demand on the oil-fueled
power plants in the study area under the upper bound electric car use
for each year of interest. Table 3.2 shows the ratio of average electric
power demand which would occur with electric cars to the baseline power
demand shown in Table 2.2.
It has been determined in Ref. 13 that in 1980 and 1990, there will
be sufficient electric power generating capacity in the South Coast Air
Basin to accommodate the extra demand and that much of it will be borne
20
rO
cn
O)
-o
c
-------
TABLE 3.2
RATIO OF IN-BASIN ELECTRIC ENERGY DEMAND WITH ELECTRIC CARS TO BASELINE
ENERGY DEMAND
1980 1990 2000
1.14 1.31 1.07
1 13
by oil-fired units. ' By the year 2000, the power demand created by
the postulated large number of electrics will be partly satisfied by
power plants located outside the South Coast Air Basin. As a result,
the ratio shown in Table 3.2 for the year 2000 is lower than that for
1990, despite the fact that the number of electric vehicles .has increased
40%, as can be seen in Table 3.1. Since power plants contribute negli-
gible amounts to the hydrocarbon burden in the basin, we expect to see
reductions in total hydrocarbon emissions, but this will be accompanied
by increases in NO and SO- emissions.
Tables 3.3-3.6 illustrate the total basinwide emissions which would
prevail with the use of electric cars. Table 3.7 shows the ratio of the
TABLE 3.3
1980 EMISSIONS WITH 20 PERCENT ELECTRIC CARS IN LOS ANGELES AND ENVIRONS
(Tons/Day)
Vehicular
Stationary Area
Sources
Power Plants
Oil Refineries
TOTAL
NO
151.7
106.0
129.2
48.8
435.7
HC
135.6
160.7
—
33.7
330.0
CO
1898.8
—
—
—
1898.8
so2
18.3
195.3
412.7
60.6
686.7
Particu-
lates
80.6
96.9
14.4
11.0
202.9
7-17
-------
TABLE 3.4
1990 EMISSIONS WITH 80 PERCENT ELECTRIC CARS IN LOS ANGELES AND ENVIRONS
(Tons/Day)
Vehicular
Stationary Area
Sources
Power Plants
Oil Refineries
TOTAL
NO
33.8
140.8
152.2
53.7
380.5
HC CO
33.6 798.0
160.7
—
37.1
231.4 798.0
so2
6.7
213.1
487.3
66.7
773.8
Particu-
lates
53.0
105.7
16.9
12.2
187.8
TABLE 3.5
2000 EMISSIONS WITH 100 PERCENT ELECTRIC CARS
IN LOS ANGELES AND ENVIRONS
(Tons/Day)
Vehicular
Stationary Area
Sources
Power Plants
Oil Refineries
TOTAL
NO
25.0
181.1
105.0
57.6
368.7
HC CO
25.4 771.9
160.7
—
39.8
225.9 771.9
so2
2.5
227.7
337.1
71.5
638.8
Particu-
lates
45.0
112.9
11.7
13.0
182.6
total emissions with electric cars to the total baseline emissions, the
latter obtained from Tables 2.4-2.7.
The data in Tables 3.3-3.5 are graphically illustrated in Figs. 3.2-
3.6. Figure 3.7 depicts the data for NO corresponding to Table 3.6 with
auto emission controls delayed since only NO is significantly affected.
7-18
-------
In all the graphs we have included data for 1970 to complete the emissions
picture. In addition, in Fig. 3.7 we have Included the effect of the
higher automotive NO emissions to 1990 and 2000, even though no air
. quality calculations were carried out for these cases. The emissions
data for these additional cases as well as for 1970 are found in
Tables C.1-C.5 in Appendix C.
TABLE 3.6
1980 EMISSIONS WITH 20 PERCENT ELECTRIC CARS AND DELAYED IMPLEMENTATION OF
EXHAUST EMISSION CONTROLS FOR CONVENTIONAL CARS IN LOS ANGELES AND ENVIRONS
Vehicular
Stationary Area
Sources
Power Plants
Oil Refineries
TOTAL
NO
247.1
106.0
129.2
48.8
531.1
HC
142.2
160.7
—
33.7
336.6
CO S02
1997.6 18.3
195.3
412.7
60.6
1997.6 686.7
Particu-
lates
80.6
96.9
14.4
11.0
202.9
TABLE 3.7
RATIO OF TOTAL EMISSIONS WITH ELECTRIC CARS TO BASELINE EMISSIONS
1980
1990
2000
1980d
Percent
Electric
Cars
20
80
100
20
NO
0.97
0.99
0.88
0.93
HC
0.93
0.86
0.81
0.93
CO
0.86
0.74
0.67
0.86
so2
1.07
1.14
1.07
1.07
Particulates
0.96
0.79
0.80
0.96
7-19
-------
Table 3.7 indicates that for NO, even massive amounts of electric
cars on the road do not reduce the total emissions significantly, at best
12 percent. This is partly due to the increase in NO emissions from power
plants, partly because of the increasing importance of emissions from
other stationary sources, as is evident from Table 2.8 and Fig. 3.2, and
also to the effect that heavy duty vehicles have on these emissions.
As shown in Table 3.7, the use of electric cars yields reductions
in hydrocarbon emissions ranging from 7 to 19 percent. Here we see again
the impact of the heavy duty vehicles, especially in the case of 100 percent
electrics where trucks and buses contribute 9 percent of total hydrocarbon
emissions.
For particulates, the emission reductions due to electric cars
range from 4 to 21 percent. In this case there is not only the effect
of heavy duty vehicles at work, but also the fact that about one-third
of the particulate emissions from cars is due to tire wear (Ref. 15).
Thus the vehicular component of particulate emissions has a relatively large
part which is unaffected by using electric cars.
Since CO is practically produced by vehicles alone, and is so con-
sidered in our study, the basinwide CO emissions are directly affected
by the number of electric cars on the road. However, as indicated by
Table 2.9, heavy duty vehicles account for between 29 to 67 percent of
all CO emissions in 1980 and beyond. This is due to the stringent emission
controls required for cars in contrast to the lack of attention which
present laws pay to the control of heavy duty vehicle emissions. Conse-
quently, large fractions of CO emissions are not affected by electric
cars and the reductions induced by electrics are at best 33 percent and
at worst 14 percent.
For S0?, the picture is different. The increase in SO- emissions
from power plants can be substantial, up to 14 percent. Thus the use of
7-20
-------
500
_
g.
oo
5
EMISSIONS WITH NO ELECTRIC CAR USE
EMISSIONS WITH UPPER-BOUND ELECTRIC CAR USE
(20% IN 1980, 80% IN 1990, 100% IN 2000)
1970
1980 1990
YEAR
2000
Figure 3.2. Plot of Nitric Oxide Emissions in Los Angeles and
Environs Assuming No Delays in Auto Emission Controls
7-21
-------
EMISSIONS WITH NO ELECTRIC CAR USE
EMISSIONS WITH UPPER-BOUND ELECTRIC CAR USE
(20% IN 1980, 80% IN 1990, 100% IN 2000)
1000
c
o
Or:
•a:
cj
o
a:
o
o
u-i
•z.
o
1970
1980
1990
2000
YEAR
Figure 3.3. Plot of Reactive Hydrocarbon Emissions in Los Angeles
and Environs Assuming No Delays in Auto Emission Controls
7-22
-------
10,000
8000
>>
O3
TJ
_
cu
o.
o
z
o
o
ca
a:
6000
4000
2000
EMISSIONS WITH NO ELECTRIC CAR USE
EMISSIONS WITH UPPER-BOUND ELECTRIC CAR USE
(20% IN 1980, 80% IN 1990, 100% IN 2000)
1970
1980
1990
2000
YEAR
Figure 3.4. Carbon Monoxide Emission in Los Angeles and Environs
Assuming No Delays in Auto Emission Controls
7-23
-------
c
o
X
o
IX
I
1/1
U-
o
1/1
o
1/1
500
POWER PLANTS
•EMISSIONS WITH NO ELECTRIC CAR USE
'EMISSIONS WITH UPPER-BOUND ELECTRIC CAR USE
(20% IN 1980, 80% IN 1990, 100% IN 2000)
VEHICULAR
0
1970
1980
1990
2000
Figure 3.5. Sulfur Dioxide Emissions in Los Angeles and Environs
7-24
-------
300
QJ
Q.
C
o
200
ce
<£
CL.
l/l
1/1
100
EMISSIONS WITH NO ELECTRIC CAR USE
EMISSIONS WITH UPPER-BOUND ELECTRIC CAR USE
(2m. IN 1980, 80% IN 1990, 100% IN 2000)
TOTAL
VEHICULAR
POWER PLANTS
0
1970
I
1980 . 1990
YEAR
2000
Figure 3.6. Particulate Emissions in Los Angeles and Environs
7-25
-------
EMISSIONS WITH NO ELECTRIC CAR USE
EMISSIONS WITH UPPER-BOUND ELECTRIC CAR USE
(20% IN 1980, 80% IN 1990, 100% IN 2000)
TOTAL
500
1970
1980
1990
2000
YEAR
Figure 3.7. Total Nitric Oxide Emissions in Los Angeles and Environs
Under Delayed Implementation of Auto Emission Controls
7-26
-------
electric vehicles can increase atmospheric levels of S0_ pollution.
Consequently, it may be necessary to place more stringent controls on
SO- emissions from power plants in order not to make matters worse.
Finally, the last row of Table 3.7 shows that for 1980, delaying
the implementation of HC and CO controls for two years produces no per-
ceptible changes from the case where controls are not delayed. For NO, on
the other hand, the effect of the electric cars on the total emissions
is greater in the delayed-control case because vehicular emissions play
a larger role, thus giving electrics more leverage. This is clearly
illustrated in Fig. 3.7. However, comparing Tables 3.3 and 3.6 we note
that the absolute level of NO emissions is still almost 100 tons per day
greater if the NO emission standard is frozen at 2.0 grams per mile
instead of lowering it to 0.40 grams per mile as originally planned, even
though electrics have a greater effect on a percentage basis.
Thus far, the discussion has focused on aggregated basinwide emis-
sions. However, as might be expected, emissions have a great degree of
spatial inhomogeneity. Hence the ratios shown in Table 3.7 do not
necessarily represent the effects that might occur as air masses sweep
over different parts of the region, thus receiving emissions from a
diverse mix of sources. We have calculated the ratios of NO and HC emis-
sions with electric cars to baseline emissions for each of the air
trajectories used in the photochemical smog modeling task. (See Appendix B
for maps of the air trajectories.) Table 3.8 shows the range spanned by
these ratios. Table 3.8 indicates that in some instances in 1980 the ratio
for total emissions approaches the ratio for vehicular sources. This is
due, of course, to the air mass moving over areas of heavy traffic. In
later years the maximum reductions fall far short of the vehicular ratios.
We note that the increases in NO emissions can sometimes be very
high. This is accounted for by the presence of power plants along the
route of the air mass, the increased emissions being due to the recharging
of batteries. The excess NO will actually cause ozone concentrations to
7-27
-------
TABLE 3.8
RATIOS OF HC AND NO EMISSIONS WITH ELECTRIC CARS TO BASELINE EMISSIONS
FOR THE AIR TRAJECTORIES USED IN THE DIFKIN AIR QUALITY MODEL
Range of Ratios for Total
Vehicular Emissions (Vehicular + Stationary)
1980
1990
2000
1980d
NO
0.84
0.45
0.31
0.82
HC
0.85
0.46
0.33
0.84
NO
0.88-1.4
0.72-3.1
0.68-3.2
0.85-1.3
HC
0.89-0.97
0.72-0.96
0.64-0.95
0.89-0.97
be reduced by virtue of the fast chemical reaction NO + 0~ -> N0» + 02 ;
thus when NO concentrations are high, the ozone is low and vice versa in
a homogeneously mixed gas. Hence because of the excess NO, the ozone
levels will be low in locations downwind of power plants.
3.3 AIR QUALITY AND ELECTRIC CAR USE
In this section we discuss the changes in pollutant concentration
that may occur as a result of the use of electric cars. The DIFKIN air
quality model was used for calculating concentrations of secondary pollu-
tants N0? and ozone. Linear rollback was also used for ozone and N0~,
and exclusively for CO, SO-, and particulates. Appendix D contains a
brief description of the rollback formula and its use.
3.3.1 Effect of Electric Cars on Ozone
Table 3.9 contains the maximum hourly average ozone concentration
for the baseline case obtained using the DIFKIN model and the rollback
formula for the four cases of interest. The year 1970 was used as the
reference year for the rollback calculation; the peak hourly ozone that year
was 62 pphm (parts per hundred million) at Riverside. In 1970, the ambient
air quality standard of 8 pphm hourly average not to be exceeded more than
once a year was violated more than 1500 hours at Riverside. For the same
year, the standard also was violated for more than 1500 hours at Azusa and
7-28
-------
Pasadena. Downtown Los Angeles exceeded the standard for slightly over
600 hours and the San Fernando Valley localities of Burbank and Reseda
showed violations for about 1100 hours. The peak hourly ozone values
ranged from 33 pphm downtown to 58 pphm at Azusa. Thus in 1970, Riverside
was afflicted with the worst ozone pollution, with Azusa a close second.
The maximum ozone concentrations computed with the DIFKIN model also occur
in the vicinity of Riverside and are shown in Table 3.9. The second
highest computed ozone values occurred at Azusa and they were 11, 10, 9,
TABLE 3.9
BASELINE VALUES OF MAXIMUM HOURLY AVERAGE OZONE IN VICINITY
OF RIVERSIDE (PARTS PER HUNDRED MILLION)
1980
1990
2000
1980d
DIFKIN Model
16
15
15
12
Rollback
16
13
13
17
and 7 for 1980, 1990, 2000, and 1980d, respectively. Maximum ozone levels
computed elsewhere in the region were below the air quality standard and
ranged from 3 to 7 pphm, the upper level occurring in the southeastern
part of the basin and the low level occurring near Anaheim. The RMS
error in the ozone concentrations computed using the DIFKIN model is
2 pphm.
The predicted concentrations shown in Table 3.9 show that the rollback
results are remarkably close to DIFKIN model estimates after rounding to
the nearest integer. The most significant difference between the two
methods of prediction is that, compared with the 1980 level, rollback
7-29
-------
predicts a higher value of ozone for the case 1980d, whereas the DIFKIN
model predicts a lower value. The reason is that rollback, being based
solely on hydrocarbon emissions, does not consider NO emissions and their
potential for depressing ozone levels. Since hydrocarbon emissions are
somewhat greater in the case 1980d than in 1980, the rollback formula
predicts a slight increase in ozone. We believe that the DIFKIN model
considers the effects of NO inhibition of ozone production whereas roll-
back does not. To be consistent, we will use the baseline concentrations
obtained with the DIFKIN model to determine the effect of the electric
cars. We note that in all cases shown in Table 3.9 the oxidant standard
of 8 pphm is being violated. It is noteworthy that the state's own pro-
jections forecast a maximum hourly oxidant level of 12 pphm by 1980 using
the linear rollback method. The discrepancy between their estimate
and ours is partly due to the different method of accounting for the
reactivity of the hydrocarbons.
Table 3.10 shows the effect of the use of electric cars on the
ozone level previously shown in Table 3.9. The results shown for the
DIFKIN model use the DIFKIN model's predictions in both the numerator
and denominator of the fraction. The rollback results were obtained by
applying the linear rollback formula to the baseline concentration obtained
with the DIFKIN model. From Table 3.10 it can be seen that electric cars
TABLE 3.10
RATIO OF MAXIMUM HOURLY AVERAGE OZONE
WITH ELECTRIC CARS TO BASELINE OZONE LEVEL
1980
1990
2000
1980d
Percent
Electric Cars
20
80
100
20
DIFKIN Model
0.99
0.90
0.87
1.04
Rollback
0.94
0.88
0.84
0.94
7-30
-------
reduce the ozone level, but that the gains are relatively small compared
with the reductions in vehicular emissions as reflected by linear roll-
back results. For the case 1980d, the DIFKIN model predicts a slight
increase in peak hourly ozone when electric cars are in use. This is due
to the differential removal of NO emissions along the trajectory of the
air mass. In this case more NO than hydrocarbon was removed by the use
of electric vehicles and this increased the ozone level. By contrast,
for 1980 the two species were removed in almost equal amounts by using
electrics and this caused the ozone to dip slightly. For 1990 and 2000,
the substitution of electrics for conventional cars eliminated more
hydrocarbon than NO and the ozone was reduced further. Note that in
every case the rollback prediction overestimates the amount of the reduc-
tion in ozone level, although the difference is not large except for
1980d, where the rollback prediction is in the opposite direction from
that of the DIFKIN model. In this instance we again see the deficiency
in neglecting NO as a factor in ozone predictions using linear rollback.
Also, the application of rollback uses the basinwide emissions rather
than the emissions along the trajectory's path; this accounts for part
of the discrepancy. It is noteworthy, however, that the rollback estimates
shown in Table 3.10 are so close to the DIFKIN model's, although the
difference in sign for the case 1980d can lead to seriously misleading
conclusions. The closeness of the two estimates may be explained by the
fact that we are working with very small perturbations in emissions; in
this situation the difference between linear assumptions and the non-
linear character of the physical phenomena tends to be minimized. The two
estimates are useful as upper and lower bounds on the possible effect of
electric cars.
7-31
-------
For Azusa, although no tables are shown in the text, the results
obtained using the DIFKIN model indicate that the use of electric cars
has imperceptible effects in 1980 and 1990. In the year .2000, the peak
hourly ozone is reduced to 89 percent of its baseline value. The rollback
estimates for Azusa essentially match those shown on Table 3.10 and thus
overstate the effect of the electric cars at this receptor location. As
was the case with Riverside, at Azusa we also obtained an increase in ozone
concentration caused by the electric cars in the case 1980d; this time
the increase is from 7 to 8 pphm or 14 percent. By contrast, rollback
predicts a decrease as it did before.
Our results for other points in the region indicate that the effect
of electric cars on ozone levels ranges from imperceptible to 50% reduc-
tions. We caution, however, that these large fractional reductions stem
from the presence of low baseline concentrations. Hence small absolute
differences can yield large percentage reductions and can be misleading.
Since the ozone levels at these other points are already below the stand-
ard, further reductions are only of interest in establishing a margin of
safety or in case the standard is lowered in the future, a move which is
not currently contemplated.
We note that in no case are the reductions obtained with the elec-
tric cars sufficient to bring the ozone level into compliance with the
air quality standard of 8 pphm in the vicinity of Riverside. The standard
is predicted to be violated at Azusa as late as 1980 and 1990 even with
electric cars. Thus other strategies besides electric cars are probably
necessary to meet the standard. The electric cars will help, however,
to reduce the number of times the standard is violated. Unfortunately,
we cannot estimate the changes that would occur in the frequency of vio-
lations in the vicinity of Riverside because sufficient statistical air
quality data for Riverside were not available to us. For Azusa the effect
of the electrics in 1980 and 1990 is essentially imperceptible and thus
no change in the frequency of violations can be expected. For the
7-32
-------
baseline case, in 1980 and 1990 we estimate that the oxidant standard at
Azusa would be exceeded about 0.7 percent of the time, approximately 60
hours a year, compared with a frequency of violations ranging from 13 per-
cent (vLlOO hrs) to 20 percent (^1750 hrs) during the period 1965-1972.
For the year 2000 the electric cars would reduce the frequency of
violations from approximately 0.2 percent (VL8 hours) to 0.1 percent
hours).
Finally, we reiterate that, as our discussion has shown, the ozone
concentration at other points in the air basin will be affected in diverse
ways because of the spatial inhomogeneity of the emissions.
3.3.2 Effect of Electric Cars on N00
The results obtained for NO with the DIFKIN model and with roll-
back for the baseline cases are shown in Table 3.11. The location where
the maximum occurs is included because it varies for different years. As
was done for ozone, the rollback calculations use 1970 as the reference
year. In 1970 the maximum hourly N0» at Anaheim was 34 pphm and at Azusa
it was 43 pphm. The corresponding annual average NO in 1970 was 4.9 pphm
at Anaheim and 6.3 pphm at Azusa. Since the air quality standard for NO^
is 5 pphm annual average concentration, in 1970 the standard was barely
met at Anaheim but exceeded at Azusa. The maximum hourly average N00
TABLE 3.11
MAXIMUM HOURLY N02 FOR BASELINE CASES
(pphm)
Year
1980
1990
2000
1980d
Location
Azusa
Anaheim
Anaheim
Azusa
DIFKIN Model
12
10
11
13
Rollback
14
13
14
17
7-33
-------
shown in Table 3.11 under the rollback heading was obtained by taking into
consideration the probability of occurrence of the maximum hourly NO- con-
centration which was actually observed. It can be seen in Table 3.11 that
the rollback calculation yields concentrations greater than those obtained
with the DIFKIN model. One possible reason for this difference is that
rollback assumes the full conversion of NO to N0?, a condition unlikely
to be fulfilled if the hydrocarbon levels are low. Since future hydro-
carbon emissions are greatly curtailed by various means of control, it is
questionable whether all of the NO will be converted to N0?, hence the
discrepancy. The difference between the two predictions shown in Table 3.11
ranges from 16 to 30 percent but, in contrast with the ozone shown in
Table 3.9, the baseline NO- predictions of DIFKIN and rollback are in
phase in all cases. We should mention that the baseline rollback predic-
tions indicate that the NO- air quality standard would be met in all but
the 1980d case. By contrast, the NO- standard would be satisfied in all
cases if we accept the DIFKIN model's predictions for the baseline case.
Finally, we note that the maximum baseline NO- levels for 1980, 1990 and
2000 computed by the DIFKIN model at other points in the basin were in
the range 5-10 pphm for all the cases.
The effect of the electric cars on the NO- concentrations is given
in Table 3.12, where the figures shown are ratios of the maximum hourly
N0? with electric cars to the baseline NO- level. Because the background
level is less than 1 pphm, it is inconsequential which of the two
-
predicted baseline concentrations is used in the denominator for the roll-
back calculation. The reductions in concentration obtained by using
electric cars which are implied by the ratios in Table 3.12 indicate that
the NO- air quality standard will be satisfied in all cases, including
the 1980d case with either prediction. The important point to note is
that the NO. levels do not increase, in spite of the fact that NO emis-
sions from power plants increase. The most likely explanation for this
effect is provided by the fact that the use of electrics reduces hydro-
carbon emissions, thereby lowering hydrocarbon concentrations and inhibiting
7-34
-------
the conversion of NO to NCL. As might be expected, the rollback ratios
match those shown for NO in Table 3.7.
TABLE 3.12
RATIO OF MAXIMUM HOURLY N02 CONCENTRATION WITH ELECTRIC CARS TO BASELINE
N02 CONCENTRATION
Percent
Year
1980
1990
2000
1980d
Electric Cars
20
80
100
20
DIFKIN Model
0.92
0.86
0.88
0.92
Rollback
0.97
0.99
0.88
0.93
We can also see in Table 3.12 that, compared with the DIFKIN model,
the rollback predictions tend to underestimate the effect of the electric
cars on N0?. This is tantamount to saying that predicted N0_ levels would
be higher if the rollback forecasts are accepted, which was the same situa-
tion we saw in Table 3.11. Again, this can be partly explained by the
failure of rollback to consider the control of hydrocarbons. It can be
seen that the two ratios for the year 2000 are equal. For this year, the
accuracy of the linear approximation is likely improved by the smallness
of the perturbations in the emissions induced by the electric cars. For
the case 1980d, the apparent 1 percent difference is probably not significant,
The closeness of the two estimates for 1980d is likely due to the fact
that the electric vehicles induce reductions in NO and hydrocarbon emissions
of equal proportions, 93 percent, and thus the accuracy of the rollback
method may be improved. We believe that the DIFKIN and rollback estimates
provide bounds for the possible effects that using electric cars can have
on N0? levels.
7-35
-------
3.3.3 Effect of Electric Cars on Power Plant Emissions
Several air trajectories used in the air quality simulations con-
tained power plants along their path. Our interest focuses on the
emissions of nitric oxide from these plants since these greatly affect
the formation of photochemical oxidants in general and ozone in particular.
Table 3.13 shows the range spanned by the increases in nitric oxide emis-
sions due to electric cars recharging for the air trajectories with power
plants. These increases in nitric oxide coupled with decreases in hydro-
18
carbon emissions along the trajectory inhibit photooxidation and result
in depressed ozone levels downwind of the power plant. These computed
ozone concentrations are in the range 3-5 pphm, well below the standard.
A related effect is that the decrease in hydrocarbon emissions also con-
tributed to retarding the conversion of NO to N0?, thus resulting in
generally lower N0_ concentrations or, if increased, only slight increases
insufficient to augment ozone production through the photochemical cycle.
However, we note that it is possible that for longer space and time scales
than we considered, the ultimate N00 and ozone peaks may be higher.
TABLE 3.13
INCREASES IN NITRIC OXIDE EMISSIONS CAUSED BY ELECTRIC CAR BATTERY RECHARGE
IN POWER PLANTS IN PATH OF AIR TRAJECTORIES
Year Percent Increase
1980 5%-40%
1990 28%-205%
2000 33%-209%
1980d 2%-30%
7-36
-------
Still unresolved, however, is whether the increased SO emissions
from power plants will act through synergistic or antagonistic mechanisms
to enhance or suppress the production of ozone in the areas downwind of
the power plants. Laboratory studies available in the literature indicate
that SO- can inhibit or enhance oxidant formation depending on the concen-
tration of water vapor and the kind of hydrocarbon which participates in
19 20
the reaction. * These same studies also show that the maximum NO- con-
centration is always lowered by the presence of SO-. However, the chemical
mechanism for this effect is not well known. Basic theoretical knowledge
is still insufficient in this area to enable us to resolve the problem
of the effect of the increased SO- emissions from power plants on oxidant
formation. It seems reasonable to conclude, however, that N0~ concentra-
tions will be generally lower at locations of the order 10-20 miles down-
wind of the power plant by virtue of the action of SO-.
3.3.4 Effect of Electric Cars on SO-, CO, and Particulates
These calculations were done using rollback exclusively with 1970
as the reference year. In 1970 the highest annual mean value of SO- was
2.6 pphm, the maximum hourly CO was 54 ppm (parts per million), and par-
3
ticulates were 357 yg/m , maximum 24-hr average. These concentrations
occurred at Lennox, Reseda, and Anaheim, respectively, and were obtained
from Ref. 20. The estimated baseline concentrations of these pollutants
are given in Table 3.14. The SO- and particulate levels contained in
Table 3.14 exceed the air quality standards and are the result of the
projected baseline emission increases previously shown in Tables 2.4-2.6.
It should be noted that in 1970 the SO- air quality standard of 3
pphm annual arithmetic mean was not exceeded in the Los Angeles Basin, in
spite of a high hourly maximum of 53 pphm which occurred at Lennox. For
CO, on the other hand, the standard of 35 ppm hourly average was violated
several times, but only once at the downtown monitoring station. By 1980
and beyond the CO standard would probably not be exceeded if the rollback
calculations are even only moderately accurate and the exhaust emission
controls are implemented on schedule.
7-37
-------
TABLE 3.14
BASELINE MAXIMUM POLLUTANT CONCENTRATIONS ESTIMATED BY LINEAR ROLLBACK
1980
1990
2000
1980d
CO (ppm)
13
7
7
14
**
S02 (pphm)
3.9
4.1
3.9
3.9
3 t
Particulates (yg/m )
402
437
464
402
Hourly average.
**
Annual mean.
24-hr average.
With electric cars in use, the effect is to reduce CO and particu-
lates but to increase the SO- annual mean, as shown in Table 3.15. The
ratios shown in Table 3.15 match those of Table 3.7, as might be expected
in view of the relatively low background concentrations of these pollutants.
Since, as shown in Table 3.15, electric cars cause SO- to increase,
the air quality standard for SO- would continue to be exceeded, assuming
that the rollback calculations are accurate. In Los Angeles, the power
plants tend to be concentrated along the coast and thus the receptors
most heavily affected by SO- pollution are coastal communities downwind
of the power plants. The data show, for example, that in 1970 locations
such as Lennox and Long Beach had annual mean concentrations of 2.6 and
2.4 pphm respectively but that inland areas such as Azusa, Burbank, and
Pasadena had mean concentrations of 1.8, 1.7, and 1.4 pphm, respectively.
Thus areas such as Lennox and Long Beach will be hit hardest by SO- pol-
lution which results from the use of electric cars.
The daily maximum value of particulate loading in 1970 occurred at
3
Anaheim and was recorded at 357 yg/m (Ref. 20). Loadings in the range
3
of 300 yg/m also occurred in Riverside. The ambient standard of 260
7-38
-------
TABLE 3.15
RATIO OF POLLUTANT CONCENTRATION WITH
ELECTRIC CARS TO BASELINE CONCENTRATION
1980
1990
2000
1980d
Percent
Electric Cars
20
80
100
20
CO
0.86
0.74
0.67
0.86
so2
1.07
1.14
1.07
1.07
Particulates
0.96
0.79
0.80
0.96
yg/m , 24-hr average, which is not to be exceeded more than once a year,
was violated twice at Anaheim and once at Riverside in 1970. The aerosol
in Los Angeles has been estimated to be composed of one-third natural
background particles, one-third anthropogenic pollutants directly emitted
into the atmosphere, and one-third chemically generated species produced
21
by the presence of the anthropogenic sources. This complex situation
raises obvious reservations about the accuracy of the rollback formula
when applied to particulate matter. With these reservations in mind,
application of the rollback formula shows that the increase in primary
particulate emissions that would occur in 1980-2000, shown in Tables 2.4-
2.6, over the 1970 emissions would result in the higher particulate load-
ings indicated in Table 3.14. Thus the air quality standard for particu-
lates would continue to be exceeded without electric cars. Using electric
cars would reduce these maxima by the fractions shown in Table 3.15, but
this is still not sufficient to satisfy the standard. However, the fact
that photochemical activity would be reduced by the presence of electric
cars implies that photochemical aerosol formation would decrease. At the
same time, the increase in S0_ emissions points to an increase in sulfate-
based aerosols. What the net effect is has not been considered in the
7-39
-------
estimates. Since the rollback formula does not consider these complex
interactions, and we know of no other presently available model which
allows us to obtain a better estimate, we believe that the values quoted
in Tables 3.14 and 3.15 are more representative of the component of the
aerosols due directly to emissions rather than of the photochemically
induced aerosols.
7-40
-------
4 CONCLUSIONS
The baseline air quality estimates suggest that CO and NO,, air
quality standards will probably be satisfied in the Los Angeles area in
the period 1980-2000 under presently contemplated emission control sche-
dules for mobile and stationary sources. Recently proposed additional
delays in the implementation of CO, NO , and HC controls for automobiles
X
have only slight effects on air quality by 1980. For ozone, S0», and
particulates, the baseline estimates indicate that the ambient air quality
standards will likely be exceeded during 1980-2000, provided that present
standards (or stricter ones) are in existence then.
In general, the substitution of electrics for conventional cars
results in improvements in air quality for ozone, N0?, CO, and particu-
lates. For ozone and particulates, however, the improvements due to elec-
tric cars are insufficient to cause the ambient air quality standards to
be met, although the frequency of the violations would be reduced. For
NO. and CO, the improvements in air quality by electric car use are rela-
tive to atmospheric levels which are likely to satisfy the standards even
without electric vehicles. Thus these beneficial results can be considered
to increase the margin of safety for protecting the public health and
welfare.
Because of the need to increase power generation for battery recharg-
ing, SO- levels will probably increase and result in violations of the
current air quality standard for S0_. This result for SO- was obtained
under the assumption that emission controls for power plants continue to
use present technology. Consequently, any new developments which lead to
more effective control of SO- emissions will modify these results.
The results of the study suggest that electric cars, even in large
numbers, cannot by themselves solve the air quality problems of the Los
Angeles region.. In the case of SO-, they can actually make matters worse
and cause standards to be exceeded where they would not be otherwise.
7-41
-------
Part of the reason for the relatively low leverage of the electrics on
air quality is that legislated automotive emission controls, if imple-
mented on schedule, will decrease the vehicular contribution of the total
air pollution burden (mobile + stationary). Hence the emissions reductions
induced by electrics operate on an ever smaller fraction of total emissions.
We should note that in the case of hydrocarbons, this effect is influenced
by the fact that the fraction of hydrocarbon emissions from stationary
sources considered to be reactive is based on a strict definition of
reactivity and this tends to emphasize the contribution from stationary
sources. Moreover, since emissions from heavy duty vehicles are relatively
uncontrolled and no plans currently exist to reduce them further, the emis-
sions from trucks and buses assume greater importance as those from
passenger cars undergo substantial reductions. This further erodes the
impact which electric vehicles might have on emissions and air quality.
Spatial inhomogeneities in the distribution of sources, however, can
produce significant improvements in localized areas. It seems appropriate,
therefore, to consider the use of electric cars as one component of a wider
strategy designed to improve the quality of the air, and thereby the
quality of life, in the Los Angeles region.
We close with a word of caution regarding the interpretation of
the results presented in this work since many uncertainties exist in the
emissions data as well as in the air quality modeling procedures used.
Our results have shown the air quality impact of electric cars to be
small as regards photochemical oxidants and carbon monoxide. These
results are influenced by two factors: the diminishing importance of
automotive emissions with respect to stationary emissions (except for
carbon monoxide); the large effect of emissions from relatively uncontrolled
heavy duty vehicles. In the case of reactive hydrocarbons, the apparent
lack of importance of automotive contributions is tied directly to the
attribution of high reactivity to stationary hydrocarbon emissions, a
subject which is now being debated by various governmental agencies.
Should the reactivity of hydrocarbons from stationary sources be lower,
7-42
-------
the automotive contribution would increase, thus possibly increasing the
impact of electrics. The leverage of electrics on air quality would be
similarly increased if tighter emission controls are placed on heavy duty
vehicles. However, it should be realized that further controls on heavy
duty vehicles and on the reactivity of stationary hydrocarbons would
result in better air quality without electric cars. Thus even though in
these circumstances the relative impact of electrics would increase, their
absolute impact on air quality would be marginal since they would be
operating in a relatively clean environment.
7-43
-------
7-44
-------
APPENDIX A
POLLUTANT EMISSIONS ESTIMATES AND PROJECTIONS FOR THE LOS ANGELES REGION
A.I INTRODUCTION
*
This appendix documents a pollutant source emissions model for most
of the South Coast Air Basin (SCAB) as assembled by GRC for use in air
quality impact studies sponsored by the US Environmental Protection Agency
(EPA) and the California Department of Transportation (CALTRANS). Since
this emissions model is intended to provide input data for the GRC photo-
23
chemical smog model (DIFKIN) it must include specifications for the
geographical and temporal distribution of pollutant emissions in the study
area. As will be seen, detailed emission forecasts are based on available
data for hydrocarbons (HC), carbon monoxide (CO), and oxides of nitrogen
(NO ), the primary pollutants modeled in the DIFKIN code. The data
X
available concerning S0» and particulate emissions are considerably less
refined, but are included here to permit estimation of current and future
regional emissions on a daily basis. Figure A.I shows the region encom-
**
passed in this source model. Within this 90-by-52-mile rectangle,
emissions are aggregated in 2-by-2-mile squares and modulated by hourly
time factors. All source locations cited in this appendix are measured
with respect to the lower left-hand (southwest) corner of this rectangle,
with x measured east and y measured north from that origin. Individual
grid squares are identified by their I, J indices shown in Fig. A.I.
The center of the rectangle (point 45.0, 26.0) is geographically located at
it
The portions of the SCAB in Santa Barbara and Ventura Counties, as well
as some sparsely settled areas of Orange, Riverside and San Bernardino
Counties are not currently included in the source emissions study area.
A similar emissions model developed for the SCAB portion of Santa Barbara
County is described in the body of this report.
**
This region, referred to in this appendix as the "study area," is termed
"Los Angeles and environs" elsewhere in the air quality analysis to
distinguish it from the entire South Coast Air Basin.
7-45
-------
20i |22| |24| |26i i28i |30| |32| |34i |36i i38| |40i |42: l44i
26
81 '10' '12' M41 M61 M8
Figure A.I. Emissions Model Study Area, Southern California
(Los Angeles and Environs)
-------
33° 55' 46.0" North latitude ) r£m(.OT. nf *..„,,„ *roa
1110 cot ir in TT ^i j^.j i Center of btudy Area
117 52' 15.1 West longitude )
which lies at the center of the northern edge of square 23, 13.
Motor vehicle emissions are calculated as the product of daily vehi-
cle miles traveled (VMT) in each grid square times average vehicle emis-
sions factors per mile of operation. Because vehicle emission factors
vary considerably between stop-and-go and cruise-type driving, we separate
the VMT in each square into freeway and surface street mileages. These
VMT distributions and the methods used to derive them are described in
Sec. A.2 of this appendix. Pollutant emission factors for an average
unit of the California vehicle population were obtained using the method
detailed in Ref. 11. Section A.3 outlines the assumptions and effects
included in this method and presents graphs of the resulting vehicle
emission factors for 1970, 1980, 1990, and 2000.
The GRC emissions model for the SCAB employs separate specifications
of stationary point source emissions from power plants and oil refineries,
and aggregates emissions from distributed stationary sources (e.g., gaso-
line marketing, dry cleaning, organic solvents uses, etc.) by grid square.
Each of the three stationary source types has its own diurnal time distri-
bution which determines the fraction of total daily emissions occurring
in each hour of the day. Sections A.4, A.5, and A.6 describe the proce-
dures employed in predicting daily emissions of HC and NO from these
X
three source categories, and their spatial and temporal distributions.
Stationary source emissions of CO may be safely neglected since they con-
1?
tribute about 0.1 percent of the total daily CO burden in the L.A. Basin.
Public Policy in the area of pollution control plays an important
role in the prediction of future emissions inventories. In the South
Coast Air Basin, Federal, State, and County regulations all impact the
forecasting process. While the EPA has generally preempted the area of
motor vehicle emission control, it has on occasion yielded in favor of
7-47
-------
more stringent controls promulgated by the California State Air Resources
Board (ARB). The schedule of Federal and State standards limiting pollu-
tant emissions from new motor vehicles in California is tabulated in Sec.
A.3. Rules set up by the Los Angeles County Air Pollution Control District
(LACAPCD) prescribe limits on emissions from existing and newly constructed
power generating units. These regulations, which served to guide our pro-
jections of power plant emissions in future years, are included in Sec. A.4.
Finally, as required by the Federal Clean Air Amendments of 1970, Califor-
nia has proposed a set of emission control regulations designed to reduce
significantly the emissions from mobile and stationary sources. This
plan, known as the State Implementation Plan (SIP), has its major impact
in the area of hydrocarbon emissions from stationary sources. We have
employed Revision 4 of the SIP here, and our assumptions regarding its
application and effectiveness are outlined in Sec. A.6.
Following conventional practice, we have reported all NO emissions
in this Appendix as N0?, and, where necessary, have assumed that our data
sources have done likewise. Since, in fact, essentially all emissions of
NO are nitric oxide (NO), the NO emissions reported here may be converted
X X
to NO by multiplying by 30/46.
A.2 ESTIMATION AND PROJECTION OF VMT AND AVERAGE SPEED DISTRIBUTIONS
Our VMT distributions are based on traffic data from two sources,
24
the VMT data collected by Systems Applications, Inc. (SAI) in 1968 as
part of their development of an emissions model for the Los Angeles Basin,
and VMT data for the SCAB at a future population of 16.1 million people,
supplied by the California Department of Transportation, District 07.
These latter data were derived from the results of a network flow simula-
tion performed by the Los Angeles Regional Transportation Study (LARTS)
by allocating network link mileages or portions thereof to grid squares
and multiplying by the corresponding average daily traffic on each link as
predicted by the network flow model. The network simulated by LARTS
included all planned freeways in the region for which construction funds
7-48
-------
have been authorized, such as the Century Freeway (Rte 105) and the Long
Beach Freeway Extension (Rte 7).
As can be seen in Fig. A.2, the LARTS study area encompasses most
of the SCAB with the exception of the portion in Santa Barbara County and
the sparsely populated eastern end of Riverside County. The inner square
(50-by-50 miles) is the area included in the SAI model. The grid selected
for our current emissions model extends the SAI study area east to include
the Riverside-San Bernardino area as shown by the shading in Fig. A.2.
In companion reports to this one, Hamilton and Houser ' have made
projections of transportation usage and population in the SCAB for the
years 1980-2000. The results of these studies indicate that the LARTS
vehicle population and usage estimates leading to the VMT data mentioned
above must now be considered high. To obtain estimates of VMT distribu-
tions for future years and populations, we have chosen to interpolate be-
tween the data reported by SAI (c. 1968) and the VMT derived from the
LARTS simulations, using total SCAB VMT reported in Ref. 7 as the scaling
parameter.
Before the interpolation process could begin, however, it was neces-
sary to resolve two data base problems: first, the existing 1968 data
had to be transferred from the SAI grid to the CALTRANS grid, which is
unfortunately shifted one mile in each coordinate, and second, no 1968
VMT data existed for the portions of the expanded study area outside the
SAI grid. The data transfer method used allocates the mileages in each
SAI grid square to the four surrounding new grid squares in proportion
to the CALTRANS mileages in the new squares. (Freeways not in existence
in 1968 are temporarily deleted from the CALTRANS VMT grid during the
mileage allocation process.) This method preserves total VMT and avoids
putting mileage into squares where none should exist. Separate data trans-
fers were performed for freeway and surface-street VMT. The remaining
grid squares in the expanded study area were filled with 1968 VMT by
7-49
-------
I
Ul
o
SAN LUISOBISPO
• County Boundaries
. South Coast Air Basin
Boundary Line
PACIFIC OCEAN
Figure A.2. VMT Study Areas in Southern California
-------
scaling down the CALTRANS VMT (without future freeways) in proportion to
total SCAB VMT. As was mentioned earlier, the CALTRANS VMT was derived
from a LARTS model run for a SCAB population of 16.1 million. Associat-
ing this population with the year 2000 (under Department of Finance
Series D-150 projections for the SCAB) permits us to use per capita vehicle
ownership and average annual mileage data from Ref. 7 to estimate the cor-
responding total annual SCAB VMT at 120.5 billion miles. Comparing this
figure with a 1968 total annual VMT of 45.9 billion in the same region,
we concluded that the CALTRANS VMT should be divided by 2.62 to arrive
at equivalent 1968 mileages.
Figure A.3 plots our best estimate of SCAB VMT growth from 1968 to
2000 as derived in Ref. 7. With the 1968 distributions of daily freeway
and surface street VMT filled in on the expanded grid, VMT distributions
1.8
1.6
ce.
o
t—
o
2 1.4
§
C3
1.2
1.0
I
1970
1968
BASED ON REF. 7
1980
1990
2000
YEAR, y
Figure A.3. Estimated Total VMT Growth Factors for the South Coast Air Basin
7-51
-------
for intermediate years are calculated on a square-by-square basis for each
category as
M = M
W W
~
2.62 - 1
- M ")
V
where y = desired target year
G(y) = VMT Growth Factor for year y from Fig. A.3
M-j^ = 1968 VMT
M_ = CALTRANS VMT (SCAB population = 16.1 x 1Q6)
M = VMT for year y
This formula scales down the VMT in each square from the 16.1 million
population value to a value corresponding to the growth curve in Fig. A.3.
Figures A. 4 through A.9 show the resulting maps of daily freeway and sur-
face street VMT for the years 1970, 1980, and 2000.
In developing their vehicular emissions model for the Los Angeles
24
Basin, SAI was able to show that only small local errors (<11 percent in the
daytime hours) resulted from the use of a single average time distribution
each for freeway and surface street traffic volumes. In the absence of
new temporal distribution data for either the SAI or extended grids, we
are employing those derived by SAI and plotted here in Fig. A.10. Table
A.I lists the hourly traffic volumes on freeways and surface streets ex-
pressed as fractions of the average daily VMT occurring in each one-hour
time period.
The vehicular emissions model in Ref. 24 also included relatively
detailed data on average freeway traffic speeds for each direction of
flow in each grid square on an hourly basis during the morning traffic
peak, 6:00-10:00 a.m. Lacking information of comparable geographic detail
for the expanded portions of the grid, or for future years, we elected to
use average freeway traffic speeds in each direction over the whole grid
7-52
-------
1
II 1? 19
JO 21 22 ?3 24 ?5 ?6 z' ?8 >0 30 11 32 3)
2*
24
9 0 n ft
3T 36 3* 3« 74 53 41 44 24 Ro
14
13
12
11
IF
<1 0 •"• « 0 189 123 109 641l21210r>aaoOOOo0003Qr'4')OOQOO -• 0 0 141 0 1?* 13u 1UO 1? 95 S4 0 r> 0 0 0 0 0
21 ??3 ?4* 346 *43 494 Q 300 n 252 151 0 60 90 So n 0 0 fl 0 0
66 o ;> n 0 794 234 345 502 116 465 192 5' 44 61 |T3 o 32 54 59 g
0-0 i i) 0 363 000 2U3 423 273 490 *8 Ofl o 0 0 41 30 0 0 0 60 26 0 1?« 113 105 63 0 46 T Q u 0 0 0 79 0 0 44 0 0 0
0 0 n 0 0 35J 0 0 0 0 98 1111 5al 346 5'5 31» 374 329 519 344 249 tig 17J 151 120 '« 0 0 0 J» 93 156 79 70 67 68 59 69 17| 103 96 107 MOO
0 0 0 ' 1 0 ?15 762 324 394 365 686 438 835 856 197 7Q 125 409 311 30 0 Q Q 0 55 20 128 105 98 69 111 155 72 63 b 0 0 59 87 0 0 0 13 75 33
0 0 ft n 29 0 217 217 o 0 313 0 0 361 417 0 108 104 0 8t 70 97 31 47 1 0 61 0 0 0 0 45 0 0 63 70 «4 191 0 0 0 0 0 0 IT
2«1 0 0 36T 1T6 521 155 363 128
35 110 147 134 18]
»1 125 89 122 66 9; 66 35 o 0 12
24T 134 284 4»? lea
i) 00000 35 0000
•I0040A347000
400
30 68
lie 170 o o
0 166 8» 0 28 56 121 44
0 0 69 220 289 121 * ;1
00006 II 0000 12 00
OOU001060001300
0000004400 U 00
t?00
Figure A.4. Geographical Distribution of Freeway VMT in 1970, Thousands of Miles
-------
-J
I
Ui
11 1?
T«06ET VHP SURFACE ST»ttT .HI
to 21 it 13 24 25 2» 27 26 2» 3° '1
35 36 37 3* 34
13
2?
21
20
19
1*
IT
It
IS
14
13
It
n 56
1
10000
I
V 0 0 0
0 ? 9 0
t 1 1 13
15 0 10 5 I 0
31 20 0
114 IOSolo013000000002noOn
17 79 1A I6| ?44 7*4 165 103 52 ^? 4/00422000000000000
3? 153 13P 200 ?31 317 ?M 211 14n 17 50 139 0*01200000120040
67 2*2 269 357 279 339 2A1 141 1*4 6] 6 47 156 51160000000301000
M 201 206 26* B2 359 443 340 344 35] 1611 20 90 33 229 126 15 90001200000
50 25 33 46 06 >93 189 177 2|4 64 06 44] 167 41 39i 355 J69 ?S5 )(,« 99 21 30 51 ;6 ?4 0 n 4
I 6 0 0 0 ?t «0 79 |4S 117 |7B 306 144 106 257 11.5 200 264 |7o 9t 7o 26* 199 134 72 32 45 42 02 »3 27 '11 3 30 41 4| 46 140 95 52 34 12 2 0
9 37 .1 0 0 43 116 247 621 7ol 32» 146 123 113 305 349 27» 254 133 1*0 !'« 310 191 40 21 76 93 54 76 56 3» 10 27 32 76 56 60 52 91 62 10 14 20 0 1
0 14 19 10 77 297 59J 621 503 526 691 796 MS 165 103 115 146 265 206 176 231 106 51 29 29 90 96 01 00 67 16 3 12 >0 31 16 57 36 64 56 16 37 74 (6 13
19 40 67 130 271 477 355 261 456 412 370 727 363 264 266 149 03 77 36 173 219 62 44 35 55 19 53 70 42 24 32 24 17 0 0 13 42 21 26 1 0 16 31 12 30
0 0 0 0 150 279 354 227 297 520 369 536 296 17) 266 250 253 74 0 93 «2 67 55 2* 7 0 9 26 14 1 12 0 39 59 36 61 62 20 2 0 I 0 12 5 1
0 0 i 0 0 66 205 312 440 376 405 353 355 314 272 267 232 277 47 62 14 52 34 2« 0 2 15 14 2 0 12 15 7 16 44 26 115 73 12 0 0 1 9 75
0 0 0 C 0 Y 9{ 322 373 372 360 166 410 253 321 3(15 144 164 259 172 116 47 2* 0 4 4 o 0 * 4 e 36 16 36 69 94 61 20 1 6 12' 0 12 2 0-
0 0 0 0 0 0 50 273 347 35o 235 275 342 227 315 37Q 313 236 224 97 157 214 62 65 4 (, 0 ft 4 0 17 20 0 56 90 27 11 21 7 >7 17 3 6 4 0
0 0 0 0 0 0 25 244 3T4 402 236 205 295 216 2*6 216 147 U3 }73 103 1*5 6| 126 72 56 I « 5 15 13 42 41 1] 16 22 10 11 21 7 21 0 0 e 1 0
0 0 n 0 0 0 0 107 366 274 196 77 93 400 35* 29; 60 124 303 162 412 245 157 63 32 4 52 In 5 n 03' 2 0 4 o 1 4 6 6 15 5 4 1 0
0 0 " 1 0 0 0 6* 460 300 145 66 115 264 1»1 29| 176 179 t96 2*2 294 320 •» 01 20 6 24 o 0 0 0 2 11 4 2 1 1 2 1 14 u 5 2 6 *
II 0 n ft 0 0 0 43 239 174 353 205 106 437 299 215 120 144 197 252 19] 2«7 135 20* »1 9 o o o n 0 t 1 2 0 0 0 0 0 3 16 6 I 1 1
0 0 ft 0 0 0 14 49 123 53 170 1|| 122 34* 29o 293 103 63 157 211 255 273 19] 174 61 5 14 7 0 0 0 0 0 6 1 0 0 0 0 4 20 19 1 0 0
0 0 " 0 0 0 0 22 21 65 212 16 0 0 0 56 41 56 169 174 165 231 272 170 InT 14 3 7 6 6 0 0 0 0 2 2 12 0 0 6 0 19 27 23 20
0 0 n 0 0 0 0 0 0 15 US 0 0 0 0 0 22 66 133 134 109 117 105 63 43 16 M 0 1 2 0 0 1 0 0 0 14 15 7 1 4 14 0 * 4
0 0 fl 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15 51 225 144 169 104 22 26 19 26 3 0 4 2 0 0 0 0 0 6 19 1* n 5 10 4 1 0
0 0
-------
10
o p n o
1 0 0 0
1)! «*i 0 0
IT
16
IS
104 0 0 0 C "?4 ?*3 401 537 1 "»9 <.-«; ?(><* I4R 1ft4 7ft 1 31*
0 0 0 0 0*13 0 0 '0 235 4*11 ?»4 «1?5 74 101 0
00000 ?3? *?5 33' 404 3"* <><.S 4.'»5 **M PR* ?1* 109
o n n o o o o 3ftj ?on o 41? o n i»« i?*. t?*
1R 0 ft 0 Tl 31
•>t>'\
TO ?7 Q 0 0 7ft 73 0 0 I* ?^ **" 0 7 *e Q 0-0
")9 !?^ 77 0 5ft 1 0 0 <* " P **° ft () QC ft n <)
0 ft 46 11 * 1 9* 9T "7 «1 B3 7? 73 ?1* 1 7" 1 19 11? 1 «• 0 0
13 104 94 1 17 ID] RS 73 0 0 0 7S ) 1 1 0 0 1 '£ 9? *0
o ft o o o
o o o o n
ooooo
o o n o o
0 3* 495 1B7 41? 19P 190 475 ?41 ft?? 14? 1» 0 0 *» Pll* 0 0 0 0 0 53 0 0 *A fl OIH11I? *P 4ft ft 71
0 0 36B P ?*5 0 0 3*7 13? 5?1 ?1« ,17S 1*,? Q o 0 11« 0 0 0 0 *5 1B9 ?4H ??? ?B1 173 IS9 0 0 0 0 n fc*
0 0 10 ?^4 *>afl *S 77 7<) " Bt |RS 74" l»c •>*4 17^ jfcr lot t4* 1*? 13B 1?1 1?4 65 0 0 |5 S4 n ft P 0 0 ft 0
n n n
n o o n o o
0000000
90 " 0 G « *". 0
•^i
Ul
Figure A.6. Geographical Distribution "of Freeway VMT in.1980, Thousands of Miles
-------
I
Ui
2345
11 1?
SUPFftCE STREET VMT, KM1
?S 26 ?7 2« 29 30 31 3? 31 3* IS
)152126197?363?5266214|4<> 16 *"? 143 0 1? 0 P 1 1 0 P 0 P 2 1 0 0 0 0 6 1 0 0 0 0 P 1 P 0 2"* 1° II IT *
23 72 ?91 26B 361 ?B1 !»iv ?76 11 ?14 ?65 BO 351 413 335 340 3*1 1 fcS ->\ tfll 17 ?40 IIP 16 1' n n 0 1 3 ft f, fl Q 0 6 0 0 0 0 0,1 I 116 ??B 0 n 1R
75 25 n 46 9Q 19* |93 ]f»0 21<> *a QB &=.« 1H| 4S 414 1E.P P*-* 75* l*« J 01 ?? 40 54 ?fl 10 0 0 5 21 P 1 9 1> 11 72 ?0 IS 27 4* 107 Si i •» 10
? 6 0 0 8 ?2 44 flT 151 |49 IQfl 1?9 15? 19ft ?76 170 ?02 ?ft*> 1*° «4 70 264 195 113 74 42 56 51 100 52 33 9 14 4 17 SI SO 57 171 117 64 41 15
?7 42 0 0 0 42 126 27* 653 731 14* 1*6 13? 116 11* 166 7P7 ?6? 11* ] 41 175 ^10 191 42 23 95 115 66 97 69 44 1? 31 39 94 69 74 61 1 11 76 13 1" ?4
0 16 22 23 8ft 310 622 664 535 562 74fl 869 ?7* 166 t 8P 1«9 if.? ?7B ?i* )9| 740 197 54 1| 13 116 119 99 108 82 22 1 IS 25 1ft 70 70 44 79 72 19 4S q]
28 59 B4 16? 281 419 376 279 4H6 43ft 3«6 779 377 265 291 1*0 91 87 4(1 1B6 ??o 71 SO 17 63 11 66 36 52 30 39 29 21 0 0 16 52 26 3? 1 P 1« IB
0 0 0 0 165 28? 365 239 25? 546 1<*3 569 3QA 171 ?R9 268 ?74 81 n 104 9B. 97 60 3ft 10 0 1 ? 32 1 ft 6 14 10 48 7? 45 74 76 34 1 0 1 ft IS
0 0 ft 0 C 92 199 316 456 394 413 356 381 3?1 27fl ?79 PS1 ? 1 95 315 373 368 .349 185 457 ?4fl 321 114 IS? 1 8Q ?*fl 189 130 60 3? P 4 4 0 0 11 5 9 45 21 47 »4 1 IS 75 ?5 1 ft 15 « 15
13 0 0 0 0 0 9 64 251 346 315 7^7 294 355 7?6 ll« 1*0 -^4] ?c;7 ?r;? 1PS]71?17 qi B? 4 0 ft 0 6 0 21 ?S n 71111 11 14 ?6 9 11 ?P 1 7
1? 0 9 C 0 0 027 251 371 191 215 214 101 210 11 "> 2?P 161 12* I*" 1 12 2!5 «4 13? *4 74 ? Q 4 14 13 50 SO 13 21 ?B 13 |3 ?6 P ?S o " ft
11 0 0 0 0 0 0 0107 183 274 198 ftl 99 4] 7 1SB 794 65 13« It n 164 429 ?511547941 P61 P 4 0 Qt.fi ? ft 4 0 1 4 7 918 f- S
10 00000 0072 450 303 148 71 120 271 189 2AB 187 IftP iqp 255 101 126 lft4P24fll6?7 0 0 t 0 211 4 1 1 I 1 I17]4 *- ?
0 0 0 0 P 0 143 23* 379 3^3 ?g6 185 415 799 PI 7 11« ]S7 ?fti ?5f> 1t>9-»l?)47?ini01 17 0 0 C ft 0 0 1 2 P ft 0 ft n 4 2? *. 1
0 0 0 0 0 0 1* 4B131 S4|7i|ftB12S 362 291 ?9? 107 9r 1*? ?15 768 ?94 199 176 66 B)o 8 0 0 0 0 n 7 1 ft 0 0 P c?4?i 1
ft 0 0 0 0 0 0 0 0 17 1 '1 P 0 P ft 0 21 6Q 140 140 11712B 205 71 6? 47 45 0 1 2 0 C 1 P 0 01«IQ n 1 Sf ft
0 0 0 0 0 P 0 ft 0 0 0 0 0 ft P P *> " Pl*-1«7114l44!4ln 3 Qn M7 64 1*) j n p r\ i *, o t. 741 \? . 1
Figure A.7. Geographical Distribution of Surface Street VMT in 1980,
Thousands of Miles
1 0
1 0
Tin
-------
1 ? 3 <> 5 6 7 B 9 10 II 12
7.0 ?1 2? ?3 74 ?S 26 ?7 28
30 31 3? 33 34 IS
41 44 45
I
Ln
10«'*'3S30174MOOO')OPOftO''Onpnpn()oPCOOOOfl3M'1'>oooflO'. f*00
f sil f>t* ??£ ?7Q SI 9 ?S? P7 ]4| ||A 0 0 0 0 n 0 n n 0 0 0 1 0 0 ft 0 •" 9 C " t> » n P v? HP P P f P P <" " n 0
?3 C P 0 0 0 313 0 IB 3 170 173 0 4 A ?S| l3**flOP'>'*?lPOOflP9''P
22 SS 119 391 436 Sftfl 570 0 309 0 245 1*R 0 119 171 l?n OOP OOOOOpCOO 0000000 S7 7? 0*.7rt|10'l'>OC
)73 0 0 0 f' B7fl 313 -3? Sflfl 159 s?S 7AO 174 ?3P 414 ?94 0 17)* IS* i*n P P 0 0 0 0 0 0 1* 62 60 SB «• |j»* «,« *7 7? 4ft 1 IT | P= *,? o n 0 0
0 0 0 0134 449 0 0 240 307 o? 3?5 1 1 fl 136 ??* P 0 0 0 94 700 193 147 }*7 0 0 97 96 37 0 0 0103 3? 0 0 0 34 120 0 9 *9 0 P O"1
0 0 0 0 O^ll 0 0 -P ?99 535 313 S9| ««; 141 0 0 o 'I 5? 0 n 0 o? 19 0 ?1* 1 *"* 175 ) 05 0 T7 11 0 0 0 0 013S 0 0 7* ft 0 0
00000 50? 0 0 0 0111 14Q? *Q4 4^? fe7| 375 1^7 i?o ^71 1^7 n? ?qi ?4Q ?1? 17S 21S P 0 0 M 155 ?6S 133 lift 113 113 99 9Q ?9? (74 »f>2 1B» 1*7 0 0
00000 269 90S 360 42? 435 lOSf. 4S7 923 941 ?61 1B9 7|1 S?6 41? 49 0 0 0 0 **4? 1*6 197 J54 14? 1?A 159 247 11 t P C 041 13 C 0 P 0 «* ->S*II (• n R
00 0 n 0 P ft f P *» P 0 « 0 pi 54 191 4AA ?ia Q ?£,<; r> n p
2
Figure A.8. Geographical Distribution of Freeway VMT in 2.000, Thousands of Miles
-------
Ul
oo
II I?
sinrrf »«7.
7» 71 77 73 7* 7* 7» 77 78 79
31 V .I!* 3* -»-V >• 3T V 1»
37 76 IS
7 77 1 1ft I ft S?
7* 911 0 9 8* 71 7|l 2* 12 9 55*111* 9 9 09999 • 79 49 0 9* • • 0 • 9
75 7» 86 •» 155 752 27» 717 II' »* 94 »J 9 9 71 II S 0 » • t 1 * • • « 9 0 0 • » 0 9 9 1 9
7* S* 159 116 191 7** 3*5 779 277 l6*18S71S291499*a9>99**99904199999
4] 798 767 368 7«. 335 765 133 145 75 ft *9 |** 7« 7* 9 9 I 9 9 9 9 o o 9 9 • t « 9 9 9 9 9 9
179 397 724 261 T* 336 *ll 376 3*1 3»T ITS 77 198 ** 754 l«9 |ft II II 9 9 3 * 9 9 9 * 9 o 9 90 • • •
21 119 »6 33 *8 93 797 7»1 195 779 77 111 496 ?05 51 46* 1*1 759 75* |7| |)| 7* *K 6| 33 47 II ft 7 ?s 9 2 17 1* IS 19 71 79 3* 67 1*6 72 |* 17
29 3 6 9 9 8 73 57 191 167 17* 73* 373 164 273 317 177 797 ?7o 166 91 7| 76* 188 139 Ml 61 7« 79 117 71 *6 12 14 S 59 A9 64 7ft 736 169 A7 SA 79
61 S3 9 9 9 «? l*» 326 717 74S 3"! 145 1*4 177 337 144 Ml 777 I»J 1*7 174 194 l»o »(. 77 131 IS7 41 137 4S 69 16 *6 S3 I7K 45 191 96 155 19* 17 7* 33
IS 9 79 » 32 194 335 679 7»« 547 633 >3ft 1997 743 169 147 776 |4g 391 >34 716 757 71* 54 3« 34 168 167 135 1*9 117 39 5 21 3* 52 77 45 60 197 411 76 67 17*
17 *6 "0 117 297 39S S94 *1* 31* 5*6 *97 **2 ««6 *03 267 393 l«l 194 4* *7 710 7*4 03 69 *l 91 56 49 117 71 *0 S* *0 20 I 9 77 71 36 ** 7 9 26 52
16 9 9 9 9 193 291 381 761 79* 54* *93 676 334 175 241 204 316 47 1 175 110 116 64 37 16 9 16 ** 75 4 14 I* 65 48 61 191 19* »7 * I 2 9 71
15 99999 193 195 314 *8? *18 *I7 371 *70 336 786 301 7"8 37* 61 13* 75 45 *fl *7 0 5 7* 75 S I 14 7* 1| 26 73 ** 197 177 79 0 0 7 15
1* 9 9 9 9 9 * 92 296 369 375 3*9 142 516 267 333 331 167 700 '41 714 15ft ft* *! 0 * * 9 I IS 7 12 61 2fl 6* 115 157 102 3* 1 11 20 n 70
13 9 9 0 9 9 9 66 7*9 3*8 377 206 316 309 2*0 331 39? 376 74? 317 177 197 797 115 i 1* 5 0 0 0 9 0 79 3* 0 47 157 *5 14 36 If *S 7« 5 10
12 0 9 9 9 9 9 39 2*6 376 379 777 232 375 217 37* 753 193 155 270 139 75* 91 1*11 106 107 * 0 6 19 17 69 69 1ft 7ft 3ft 11 Ift 35 II 3* 1 « 0
11 0009999 198 367 263 193 ft6 109 4?9 351) 79ft 75 167 373 16A *60 751 161 99 66 IS B3 11 6 9 9 66 3 9 6 1 I 6 9 13 25 o 7
19 9999f9979 *31 319 15? 77 129 7R5 195 7HI 70* 70* 7A7 759 171 !** ||3 4ft M 17 17 0 9 5 0 3 1ft 6 * I I 7 7 73 14 n ?
0999001 *? 734 31*4 357 711 18* *3I> 794 770 15A 1R1 71* 769 713 35f, 173 756 13!3490C10073l'000'i5iAQ7
9 9 9 9 9 0 15 *6 1*6 56 161 10? l?ft 391 747 ?Aft ||* |0* 1T| 77* 79* 335 715 17ft 75 1* 75 11 9 0 0 0 9 In I 0 9 0 0 7 32 II 7
53 7*
15 O
0 9
3 9
9 5
*3 27
19 51
• 7
11 «
3 9
7 1
7 9
7 0
9 14
I 3
0 I
76 76 |5* 151 133 151 7*7 9? 109 104
' ) 0 0 I 0 9 0757611 1
176 753 151 701
71 15 0 9 0 0
*1 I 0 •> * 1
Figure A.9. Geographical Distribution of Surface Street VMT in 2000,
Thousands of Miles
-------
IU
a: 8
8
3:
ce
o.
O
£ 6
s
u. ^
O
I—
LxJ
o:
S 2
0
24
l-n
NO
EEWAYS
N-FREEWAYS
—
-
_
r
,
J
\|
|
. nr ^
1 1 1 1
00 0400 0800 1200 1600 2000 24
TIME OF DAY
Figure A.10. Los Angeles Traffic/Time Distributions (Adapted from Roberts,
Roth, and Nelson, Ref. 17)
as obtained for each hour from the SAI data for their grid in 1968. These
average speeds were obtained as VMT-weighted averages for each hour be-
tween 6:00 and 10:00 a.m. Table A.2 shows the temporal distribution of
average speeds and direction-volume ratios thus obtained, and subsequently
used for all freeways in all years.
A.3 AVERAGE VEHICLE EMISSION FACTORS
The average gram-per-mile pollutant emissions from the vehicle popu-
*
lation in a given year are weighted combinations of emissions from light-
and heavy-duty vehicles of various model years and ages, which are driven
varying numbers of miles in their average use depending on age. Since
1963, the pollutant emissions from vehicles sold in California have been
For purposes of Federal and State emission control standards, vehicles
weighing 6,000 pounds or less are defined as "light-duty;" those 6,001
pounds and over in gross weight are classified as "heavy-duty" vehicles.
7-59
-------
TABLE A.I
DIURNAL VARIATIONS OF SOURCE ACTIVITIES (1974)
(Fraction of Daily Total Assignable to a 1-hour Period)
Local
(Midnight)
A.M.
Noon
P.M.
(Midnight)
Time
2400-100
100-200
200-300
300-400
400-500
500-600
600-700
700-800
800-900
900-1000
1000-1100
1100-1200
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1800
1800-1900
1900-2000
2000-2100
2100-2200
2200-2300
2300-2400
Freeways
0.00776
0.00776
0.00776
0.00776
0.00776
0.0178
0.0591
0.0768
0.0648
0.0536
0.0494
0.0494
0.0494
0.0494
0.0569
0.0746
0.0746
0.0746
0.0598
0.0302
0.0302
0.0302
0.0302
0.0302
Surface
Streets
0.00677
0.00677
0.00677
0.00677
0.00677
0.00677
0.0293
0.0651
0.0651
0.0502
0.0502
0.06088
0.06088
0.06088
0.06088
0.06088
0.0820
0.0820
0.0540
0.0540
0.03077
0.03077
0.03077
0.03077
Power
Plants
0.02756
0.01911
0.01695
0.01484
0.01381
0.01484
0.01695
0.02334
0.03709
0.04451
0.05095
0.05404
0.05616
0.06043
0.06146
0.06387
0.06043
0.05940
0.05724
0.05306
0.05404
0.05616
0.04771
0.03606
Oil Refineries and
Distributed Sources
0.01666 JT
0.01666 |
0.01666 10%
0.01666
0.01666
0.01666 }
0.06667 J
0.06667
0.06667
0.06667
0.06667
0.06667 80%
0.06667
0.06667
0.06667
0.06667
0.06667
0.06667 '
0.01666
0.01666
0.01666 10%
0.01666
0.01666
0.01666 '
7-60
-------
TABLE A.2
AVERAGE FREEWAY FLOW SPEEDS AND DIRECTION-VOLUME RATIOS FOR SCAB GRID
Local Time
0000-0500
0600
0700
0800
0900
1000-2400
Average Speed
in Slow
Direction,
mph
60.0
57.7
44.3
42.9
52.6
60.0
Average Speed
in Fast
Direction,
mph
60.0
59.8
59.3
58.9
59.7
60.0
Traffic Volume Ratio
Slow Direction
Fast Direction
1.00
1.48
1.32
1.32
1.18
1.00
subject to State, and subsequently, Federal limitations by model years.
Compliance with standards is determined by emissions collected from the
vehicle during operation over a prescribed speed-time schedule referred
to as a driving cycle. It is these driving .cycle emission factors which
serve as the common basis permitting meaningful combinations of emissions
over a range of vehicle model years. Current standards are based on the
25
Federal Driving Cycle (FDC) which was designed to simulate average
vehicle operation during peak traffic hours within 7 miles of the central
business district. The resulting simulated trip covers 7.47 miles at an
average speed of 19.6 mph. Empirical speed correction techniques are
used to predict emissions at average trip speeds other than this under
11 O £
both cruise and stop-and-go traffic conditions. '
^Reference 11, compiled by GRC under the sponsorship of the California
Department of Transportation, details the calculations required to arrive
at vehicle emission factors for a prescribed mix of light- and heavy-duty
California vehicles in a given calendar year. Basically, the method
involves summing model-year FDC emission factors for all years prior to
the target year, weighting them by (1) a deterioration factor which
7-61
-------
accounts for the loss in effectiveness of emission control systems with
vehicle mileage, (2) the fraction of the daily VMT contributed by vehicles
according to their ages and use, and (3) a speed correction factor for
speeds other than the driving cycle average speed. In general, all of
these factors will be functions of vehicle model year. As a practical
matter, it is not necessary to consider all model years prior to the
calendar year of interest, since after a certain age is reached, the
mileage contributions become negligible.
The fraction of heavy-duty vehicles found in the urban mix is sub-
ject to considerable variation with locale. For the SCAB emissions model,
we have assumed that on the average 5 percent of all VMT are contributed
by heavy-duty vehicles. Reference 11 provides for the calculation of both
cruise and stop-and-go emission factors over the range of speeds from 10
to 60 mph. Of course, in practice traffic does not flow smoothly at
10 mph nor can it maintain an average speed of 60 mph under stop-and-go
27
conditions. The Highway Capacity Manual indicates that the transition
from stable traffic flow to unstable flow occurs rather abruptly at speeds
immediately below 40 mph for multi-lane uninterrupted highways. Since
we are only accounting for speed variations on freeways in our traffic
model, we have assumed for the purposes of computing emission factors
that all traffic at speeds of 40 mph and greater flows smoothly warranting
use of cruise mode emissions. Freeway traffic averaging less than 40 mph
would have emissions evaluated by the stop-and-go (sometimes called
"average route speed") scaling method. It should be noted from Table A.2
that the assumption of geographically uniform speed distributions, made
during the grid extension process, effectively put all freeway traffic in
the cruise category. On the other hand, because the traffic model used
does not include speed distributions for surface street traffic, emissions
from surface street VMT are all assessed at a single speed under stop-and-
go conditions. The speed chosen is 19.6 mph, the average speed of the
FDC.
7-62
-------
Figures A.11, A.12, and A.13 show the currently modeled speed varia-
tions of total HC, CO, and NO emission factors for the years 1970, 1980,
X.
1990, and 2000 as obtained from Ref. 3. As noted in Ref. 11, 70.4 percent:
of vehicular HC emissions are treated as reactive based on earlier analyses
of automobile exhaust. Unfortunately, data is lacking on the reactive
fraction of exhaust emissions from vehicles equipped with advanced emis-
sion control systems such as catalytic converters. Should measurements
of the reactive HC fraction from these devices prove to vary significantly
from the assumed 70.4 percent, projections of future RHC contributions from
vehicular sources would need to be reevaluated. Vehicular NO emissions
, "
are assumed to be 100 percent nitric oxide (NO). The vehicular emissions
shown here are predicated on compliance with the current schedule of
emission standards applicable to vehicles sold in California, and include
the effects of the interim California Standards for 1975 set by EPA on
28
April 11, 1973 and the one-year delay of the 1976 NO emission standard
29 x
granted by EPA on July 30, 1973. Table A.3 contains the resulting schedule
20 r-
5
S
DATA FROM REF. 11
5% HEAVY DUTY VEHICLE MIX
FEDERAL DRIVING CYCLE
AVERAGE SPEED = 19.6 mph
CRUISE
Figure A.11. Variation of Total HC Emissions with Traffic Speed
7-63
-------
160
140
120
100
e
03
S_
O1
o;
o
80
~ 60
o
o
40
20
DATA FROM REF. 11
5% HEAVY DUTY VEHICLE MIX
1970!
FEDERAL DRIVING CYCLE
AVERAGE SPEED =19.6 mph
CRUISE
10
20 30
SPEED, mph
Figure A.12. Variation of CO Emissions with Traffic Speed
7-64
-------
01
0.
1
en
if)
GO
X
o
DATA FROM REF. 11
5% HEAVY DUTY VEHICLE MIX
STOP & GO
FEDERAL DRIVING CYCLE
AVERAGE SPEED =19.6 mph
1970
1980
10
20
30
SPEED, mph
40
50
60
Figure A.13. Variation of NOX Emissions with Traffic Speed
7-65
-------
TABLE A.3
EXHAUST EMISSION STANDARDS FOR LIGHT-DUTY VEHICLES
Emission Standards, grains per mile
1972
1973
1974
1975
1976
1977
Federal
California
Federal
California
Federal
California
Federal
California
Federal
California
Federal
California
HC
3.4
*
3.2
3.4
3.2
3.4
3.2
1.5
0.9
0.41
1.0
0.41
1.0
CO
39
39
39
39
39
39
15
9.0
3.4
9.0
3.4
9.0
NO
X
3.0
3.2
3.0
3.0
3.0
2.0
3.1
2.0
2.0
2.0
0.40
2.0
Underlining denotes requirements for new California cars wherever dif-
ferences arise between Federal and California standards.
of standards compiled in Ref. 11. (Emissions for vehicles prior to 1972
are based on measured data.)
Reference 15 gives the following light-duty vehicle emission factors
for SOp and particulates independent of vehicle model year and traffic
speed; 0.2 g/mi of S0_, 0.38 g/mi of exhaust particulates, and 0.2 g/mi of
particulates from tire wear. For heavy-duty vehicles, the emission factors
are 0.26 g/mi of S02, 0.65 g/mi of exhaust particulates, and 0.2 g/mi of
particulates from tire wear. These factors may be used in conjunction
with estimates of regional VMT to calculate vehicular contributions to
the daily SO- and particulate burden. Table A.4 shows the resulting
estimated daily contributions for the study area assuming heavy-duty
vehicles account for 5 percent of the VMT.
7-66
-------
TABLE A.4
ESTIMATED DAILY VEHICULAR EMISSIONS OF S02 AND PARTICULATES IN STUDY AREA
Year
1970
1980
1990
2000
Daily
VMT (x 10°)
134.3
149.0
162.0
175.6
Emission
so2
20.2
22.4
24.4
26.4
is, Tons /Day
Particulates
82.2
91.2
99.2
107.5
A.4 POWER PLANT EMISSIONS
,13
In a companion report, Sjovoid has developed forecasts of electric
power demand in the SCAB which are compatible with our fundamental assump-
tions regarding population growth. Reference 13 supports three assump-
tions which are of importance to our estimation of future power plant
emissions in the study area:
1. Except for those already under construction, no new power
plants (nuclear or thermal) will be located in the SCAB. Any
additional power generated in the region will be produced by
uprating existing units, by retrofiting existing units with
combined cycle capability, and by construction of a few
extra units at established thermal plants.
2. Natural gas is not expected to be available in any signifi-
cant amount for power generation, as the available supply
will probably be allocated to meet increasing residential
demands. Hence, the thermal plants in the SCAB are expected
to be almost entirely oil fired after 1980. (Currently,
Combined cycle refers to systems employing a combination of the Brayton
and Rankine cycles, i.e., a*direct combustion turbine (like a jet engine
turbine) is employed both to drive a generator directly and to produce
hot exhaust gases which are used to boost the output of the basic steam
turbine unit.
7-67
-------
increased residential demand for natural gas forces power
plants and industry to use low-sulfur fuel oil during the
winter months. Residential demand from April to November is
generally low enough to permit power plants and industry to
use natural gas.)
3. While the availability of low-sulfur oil (_<0.5 percent by weight)
may be a problem in the mid-1970s, the economic and techno-
logy of desulfurization indicate that low-sulfur oil or a
non-petroleum equivalent should be in ready supply from 1980
on.
By the year 2000, nuclear stations sited outside the basin are expected
to provide the major fraction of electrical generation capacity, meeting
most of the continuous basic demand together with some hydroelectric
sources. In the 1980s, fossil fueled plants (some coal fired plants out-
side the SCAB and the oil fired facilities within the region) will be
providing some of the base load power and all of the peaking capacity.
Figure A. 14 shows a projection of the peak demand to be met by the generat-
ing facilities that really concern us here, the oil fired plants sited in
*
the SCAB. As the availability of nuclear power increases, less and less
of the oil-fueled plant capacity will be used to satisfy the base load.
Figure A.15 plots the resulting daily power cycle in the peak demand
month (August) for the years 1974-1980, 1990, and 2000. These curves
provide the time distribution functions needed for detailed forecasts of
power plant emissions. Table A.I contains data for the 1974-1980 time
distribution converted into hourly fractions as needed to allocate total
daily emissions. Similar tabulations for the projected 1990 and 2000
time functions are given in Table A.5.
The locations (in source grid coordinates), stack heights, and capa-
cities of the 16 oil-fueled power plants in our study area are listed in
Four of these thermal plants (Mandalay, Cool Water, Ormond Beach, and
Ellwood) lie outside our study area and hence, are not included in our
current emissions model.
7-68
-------
15
10
_, 5
S
1970
FROM REF. 13
1980 1990
YEAR
2000
Figure A.14.
Baseline Projected Peak Demand for Electricity Generated
by Oil Fired Plants in the South Coast Air Basin
20
15
a 10
S
Q£
LU
g
a.
3 5
1974 - 1980
,990 FROM REF. 13
2000
0600
1200.
TIME
1800
2400
Figure A.15. Projected Baseline Diurnal Power Demand on Oil Fired Power
Plants in the South Coast Air Basin
7-69
-------
TABLE A.5
FUTURE POWER PLANT TIME FUNCTIONS FOR OIL FIRED UNITS IN THE SOUTH COAST
AIR BASIN
(Fraction of Daily Total Assignable to a One-Hour Period)
Local Time
1990
2000
2400-0100
0100-0200
0200-0300
0300-0400
0400-0500
0500-0600
0600-0700
0700-0800
0800-0900
0900-1000
1000-1100
1100-1200
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1800
1800-1900
1900-2000
2000-2100
2100-2200
2200-2300
2300-2400
0.01984
0.00661
0.00331
0
0
0
0.00331
0.01322
0.02975
0.04624
0.05615
0.06119
0.06444
0.07101
0.07273
0.07618
0.07101
0.06947
0.06612
0.05946
0.06119
0.06444
0.05127
0.03306
0
0
0
0
0
0
0
0
0.01133
0.03966
0.05666
0.06516
0.07082
0.08215
0.08499
0.09065
0.08215
0.07932
0.07365
0.06232
0.06516
0.07082
0.04816
0.01700
7-70
-------
Table A.6. Note that the capacities shown are for 1973, but three sites:
Huntington Beach, Long Beach, and Scattergood are scheduled for capacity
increases between 1973 and 1979. Based on data provided by Southern
California Edison (SCE) and the Los Angeles Department of Water and Power
(LADWP), Table A.7 shows the expected capacities of those plants, the
total added power capacity, and the total study area capacity for each
year through 1978 after which no additions are anticipated.
TABLE A. 6
POWER PLANT DATA
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Los Alamitos
El Segundo
Redondo Beach
Huntington Beach
Long Beach
Harbor
Haynes
Scattergood
Valley
Pasadena
Bur bank
Glendale
Etiwanda
Highgrove
San Bernardino
Garden State
x,
mi
31.8
13.2
15.0
38.8
25.0
22.4
32.1
11.9
15.1
29.1
19.6
21.7
66.7
77.9
83.0
54.7
y»
mi
14.9
24.7
20.5
6.4
13.9
15.0
14.4
28.3
47.7
39.5
43.1
41.5
36.9
32.5
36.3
33.8
I
16
7
8
20
13
12
17
6
8
15
10
11
34
39
42
28
, Average Stack
Height, ft
8
13
11
4
7
8
8
15
24
20
22
21
19
17
19
17
214
222
214
211
220
250
240
310
510
60
60
88
188
85
130
50
1973
Capacity,
MW
2071
1020
1602
880*
*
212
525
1593
*
358
517
230
174
153
1025
154
126
12
See Table A.7 for schedule of increased capacity.
7-71
-------
TABLE A.7
POWER PLANT CAPACITY SCHEDULE
Number
4
5
8
Plant
Hunt ing ton Beach
Long Beach
Scattergood
"New" Power
Study Area Total
1973
880
212
358
0
10,652
1974
880
212
667
309
10,961
Capaci
1975
1,252
465
667
934
11,586
ty, MW
1976
1,624
465
667
1,306
11,958
1977
1,960
465
667
1,642
12,294
1978
2,296
465
667
1,978
12,630
On the assumption that the emissions characteristics of the power
plants in the L.A. area are reasonably uniform, and that power demand is
allocated to plants in proportion to their capacities, our model distributes
power plant emissions in the same way. Table A.8 summarizes the total
daily emissions of NO , SO-, and particulates reported by the air pollu-
X 12 30 31
tion control districts of Los Angeles, Orange, and San Bernardino
Counties for 1971 for the power plants in the study area. (Hydrocarbon
emissions from power plants contribute less than 0.2 percent of the total
HC burden and hence are considered negligible.)
TABLE A.8
1971 EMISSIONS FROM POWER PLANTS IN THE STUDY AREA
Emissions, Tons/Day
County
Los Angeles
Orange
San Bernardino
Totals
NO
X
100
11
9.4
120.4
so2
160
0
14.6
174.6
Particulates
5
0
2.5
7.5
7-72
-------
The emissions reported by the LACAPCD were generated on 28 percent
natural gas and 72 percent low-fulfur oil averaging 0.45 percent sulfur
by weight. We have assumed that the remainder of the plants used
100 percent low-sulfur oil at 0.5 percent sulfur content, the legal limit.
Thus the assumed shift to 100 percent low-sulfur oil for all SCAB power
plants by 1980 will result in increased emissions of SO, and particulates
15
from existing facilities. EPA emission factors indicate that the
effect on NO emissions should be negligible. To provide a basis for
X.
scaling SO™ and particulate emissions, we computed equivalent 1971
emissions as if only 0.5 percent low-sulfur oil had been used. Again
using the EPA emission factors as a guide we note that S0» emissions
using natural gas are negligible compared to those from oil-fired boilers.
However, particulate emissions using gas are 26.8 percent of their oil-pro-
duced counterpart. The hypothetical SO- emissions are computed as follows
(see Table A.8):
For Los Angeles County
160 tons/day/O.
0.72
and adding the contribution from San Bernardino County (which we assumed
to be using 0.5 percent low-sulfur oil already) we obtain the equivalent total
emissions of
246.9 + 14.6 = 261.5 tons/day of SO
To calculate the corresponding particulate emissions, we first determine
that of the 5 tons per day contributed by Los Angeles County, 4.53 tons came
from the burning of fuel oil and 0.47 tons from burning natural gas. Then
we proceed as outlined above for S0_, except that the difference in sul-
fur content of the oil does not affect particulate emissions. Hence,
4'53.t°ns/day + 2.5 = 8.79 tons/day of particulates
U • / Z
7-73
-------
Emissions from new power generating facilities coming on line in
the L.A. region are limited by Rule 67 to 140 Ib/hr of NO , 200 Ib/hr of
32 x
S0_, and 10 Ib/hr of particulates. The added power generating capacity
shown in Table A.7 amounts to a total of 1,978 MW from 17 separate addi-
tions. If we assume that Rule 67 applies "per addition" operated at its
capacity, then we can calculate emission factors per megawatt-hour delivered
by the new facilities. Table A.9 compares the resulting power generation
emission factors for existing capacity in 1971 (meeting an average demand
of 5,590 MW) and for new and existing facilities using low-sulfur oil
exclusively.
Finally, using the schedules of existing and planned capacity in
the study area (Tables A.6 and A.7), the projected demand on that generat-
ing capacity (see Figs. A.14 and A.15), and the power generation emission
factors in Table A.9, we can project total power plant emissions of NO ,
X
S0_, and particulates for future years. Assuming that power demand is
distributed such that all units operate at the same fraction of their
total capacities, Table A.10 lists the expected schedule of pollutant
emissions.
TABLE A.9
POWER GENERATION EMISSION FACTORS
NO
X
so2
Particulates
1971
Emissions,
Tons /Day
120.4
174.6
262 (oil)
7.5
8.79 (oil)
Existing Power
Emission Factor,
kg/MW'hr
0.814
1.183
1.1772
0.0507
0.0594
Maximum Emissions
from Added
Power, Tons /Day
28.6
40.8
2.04
"New" Power
Emission Factor,
kg/MW'hr
0.547
0.780
0.0390
7-74
-------
TABLE A.10
PROJECTED POLLUTANT EMISSIONS FROM POWER PLANTS IN THE STUDY AREA
Year
1971
1980
1990
2000
Total
Capacity,
MW
10,652
12,630
12,630
12,630
Average
Demand ,
MW
5,590
8,480
8,700
7,350
Em
*
NO
X
120.4
173.6
178.1
150.5
issions, '
so2
175
362
372
315
"ons/Day
Particulates
7.5
12.6
12.9
10.9
To convert to nitric oxide, multiply by the factor 30/46.
Geographic distribution of the NO emissions in Table A.10 in pro-
portion to power plant capacity provides average daily NO emissions at
X
each site as required by the GRC and SAI smog models. Table A. 11 dis-
plays these data.
A.5 PETROLEUM REFINERY EMISSIONS
On a SCAB regional basis, emissions from petroleum refineries are
relatively small, but since refineries are concentrated sources, their
24
emissions can be important on the local scale. The SAI data base pre-
viously used in the DIFKLN code simply prorates total petroleum refinery
emissions of NO and reactive HC as reported by the Los Angeles County
X
APCD to individual refinery locations according to their crude oil capa-
cities. Since no additional refineries are in the expanded study area,
we have retained the SAI refinery data, transferring locations onto the
new grid, adding estimates of current and future total emissions of S0«
and particulates from petroleum refinery operations, and updating the
inventory emissions of NO and HC relative to the original SAI model.
24
7-75
-------
TABLE A.11
GEOGRAPHIC DISTRIBUTION OF NO EMISSIONS FROM POWER PLANTS
x
1. Los Aland. tos
2. El Segundo
3. Redondo Beach
4. Huntington Beach
5 . Long Beach
6 . Harbor
7 . Haynes
8. Scattergood
9. Valley
10. Pasadena
11. Burbank
12. Glendale
13. Etiwanda
14 . Highgrove
15. San Bernardino
16. Garden State
I
16
7
8
20
13
12
17
6
8
15
10
11
34
39
42
28
J
8
13
11
4
7
8
8
15
24
20
22
21
19
17
19
17
1971
885
436
684
376
91
224
681
153
221
98
74
65
438
66
54
5
Emission
1980
1076
530
833
1193
242
273 .
828
347
296
120
90
80
533
80
65
6
s, kg/hr
1990
1104
544
854
1224
248
280
849
356
276
123
93
82
546
82
67
6
2000
933
459
722
1034
209
236
718
300
233
104
78
69
462
69
57
5'
The pollutant emissions from petroleum refineries listed by the
12
LACAPCD for 1970 are 68.4 tons per day of NO , 61.6 tons per day of
X
total hydrocarbons, 55.4 tons per day of SO,,, and 10.1 tons per day of
particulates. As recommended by SAI in Ref. 14 we have retained 50 percent
(30.8 tons per day) of the total HC emissions from petroleum refineries as
reactive enough to participate in the formation of photochemical smog.
The recommended reactivity definition includes as reactive all hydrocar-
bons except methane, ethane, propane, benzene, and acetylene.
7-76
-------
Current uncertainties as to future oil supplies in the US make it
difficult to forecast petroleum refinery activities with confidence, but
we have taken the view that any real energy shortage which can be eased
by the expansion of refining capacity will probably lead to relaxations
of environmental restraints opposing that expansion. Therefore, reasoning
that it is national demand which will be the prime consideration in deter-
mining gasoline and fuel oil production, we have scaled oil refinery emis-
sions in the SCAB with the projected national population to arrive at an
emissions forecast for 1980, 1990, and 2000.
In keeping with our assumptions concerning population growth in the
study area, we extrapolated the national population using Bureau of the
33
Census Series E projections. Table A.12 gives the schedule of pollutant
emissions from petroleum refineries in the study area.
Table A.13 reprints the petroleum refinery data contained in Table
A.23 of Ref. 24, translating the refinery locations into the new grid coor-
dinate system. The geographical emissions distributions for NO and reac-
X
tive HC given in Table A.14 were obtained by allocating the total daily
TABLE A.12
POLLUTANT EMISSIONS FROM PETROLEUM REFINERIES IN THE STUDY AREA
Year
1970
1980
1990
2000
us 33
Population
(xlO6)
204.88
224.13
246.64
264.44
Growth
Factor
1.0
1.094
1.204
1.291
Emissions, tons/day
*
NO
X
68.4
74.8
82.4
88.3
Reactive
HC
30.8
33.7
37.1
39.8
so2
55.4
60.6
66.7
71.5
Particulates
10.1
11.0
12.2
13.0
Multiply by 30/46 to convert to nitric oxide.
7-77
-------
TABLE A.13
PETROLEUM REFINERY DATA
24
1. Atlantic Richfield Company
2. Douglas Oil Company of California
3. Edington Oil Refineries, Inc.
4. Fletcher Oil and Refining Company
5. Carson Oil Company, Inc.
6. Gulf Oil Corp.
7 . Lunday-Thagard Oil Company
8. MacMillan Ring-Free Oil Company
9. Mobil Oil Corp.
10. Power ine Oil Company
11. Shell Oil Company
12. Shell Oil Company
13. Standard Oil of California
14. Texaco, Inc.
15. Union Oil Company of California
x,
mi
24.2
29.2
28.3
21.6
21.3
34.7
28.1
28.0
18.1
33.9
23.9
23.6
14.0
24.1
23.7
y,
mi
17.3
23.9
21.9
18.2
19.8
24.7
27.5
16.4
20.7
26.3
20.0
17.8
24.6
16.3
15.6
I
13
15
15
11
11
18
15
15
10
17
12
12
8
13
12
J
9
12
11
10
10
13 .
14
9
11
14
11
9
13
9
8
., Crude Capacity
(10 Barrels/Stream Day)
173
26
16
13
10
51
3
10
130
30
44
44
210
61
107
emissions in Table A.12 to each refinery in proportion to its crude
capacity as listed in Table A.13 and combining emissions where two re-
fineries are located in the same grid square. These emissions are
temporally distributed as shown in Table A.I
A.6 AREA SOURCE EMISSIONS
The distributed source category includes industrial, commercial,
and residential activities such as fuel combustion (for temporary power
generation, power-driven machinery and space heating), petroleum market-
ing (e.g., gasoline distributors and service stations), and the use of
7-78
-------
TABLE A.14
GEOGRAPHIC DISTRIBUTION OF NO AND REACTIVE HC
x
EMISSIONS FROM PETROLEUM REFINERIES
I . D . Number
1. and 14.
2.
3.
4. and 5.
6.
7.
8.
9.
10.
11.
12.
13.
15.
Loca
I
13
15
15
11
18
15
15
10
17
12
12
8
12
it ion
J
9
12
11
10
13
14
9
11
14
11
9
13
8
Emissions, kg/hr
197
NO
X
652
72
45
64
142
8
28
362
84
123
123
585
298
0
RHC
294
33
20
29
64
4
13
163
38
55
55
263
134
19£
NO
X
713
79
49
70
155
9
31
396
92
135
135
640
326
0
RHC
322
36
22
32
70
4
14
178
42
60
60
288
147
195
NO
X
785
87
54
77
171
10
34
436
101
148
148
704
359
10
RHC
354
40
24
35
77
5
16
196
46
66
66
317
161
200
NO
X
842
93
58
83
183
10
36
467
108
159
159
755
385
0
RHC
380
43
26
37
83
5
17
210
49
71
71
340
173
organic solvents (for dry cleaning, degreasing, and surface coating).
Data for area source emissions were obtained from the source inventories
compiled by the air pollution control districts of Los Angeles, Orange
31
and San Bernardino Counties. At the time this emissions model was being
assembled, we were unable to obtain an emissions inventory for Riverside
County. Area source emissions in that portion of the study area were
synthesized by a technique using land-use emission factors, to be
described presently.
Average daily S0? and particulate emissions from distributed sources
in the study area are aggregated in Table A.15. (Aircraft emissions are
7-79
-------
TABLE A.15
DISTRIBUTED SOURCE EMISSIONS OF S02 AND PARTICULATES, 1971
a. SO- Emissions, Tons/Day
County
Chemical Industry
Metallurgical Industry
Mineral Industry
Other Industry
Petroleum Production
Commercial
Residential
Aircraft and Railroad
Total
Los Angeles
117
2.8
4
123.8
Orange
2.0
4.6
4.0
o.i
0.1
1.2
12.0
San
Bernardino
26.2
4.1
0.9
31.2
*
Riverside
9.9
1.1
11.0
Total
117
38.9
2.0
4.6
4.0
5.3
0.1
6.1
178.0
b. Particulate Emissions, Tons/Day
County
Chemical Industry
Metallurgical Industry
Mineral Industry
Other Industry
Petroleum Production
Commercial
Residential
Aircraft and Railroad
Total
Los Angeles
11
11
6
6.4
8.1
5
13
60.5
Orange
0.2
0.4
2.3
1.2
0.4
0.4
4.3
9.2
San
Bernardino
4.5
6.9
1.6
0.9
13.9
*
Riverside
1.7
2.6
0.4
4.7
Total
11.2
17.6
17.8
7.6
0.0
10.5
5.4
18.2
88.3
Riverside County emissions estimated from San Bernardino County emissions based
on the ratio of land areas devoted to industrial or commercial use (see Table A.21).
7-80
-------
only included here as a convenience in calculating total regional emissions
of S00 and particulates.) Daily area source emissions of NO and total
£ X
hydrocarbons (THC) abstracted from the district inventories12»30»31 are
listed in Tables A.16, A.17, and A.18. The reactive hydrocarbon (RHC)
14
emissions shown in these tables are based on a reassessment by SAI of
the participation of individual hydrocarbons in the smog formation process.
TABLE A.16
HC AND NO EMISSIONS FROM DISTRIBUTED SOURCES IN LOS ANGELES COUNTY, 1971
X
(From Ref. 12)
Emissions, Tons/Day
Chemical Industry
Metallurgical Industry
Mineral Industry
Petroleum Marketing
Underground Tanks (Service Stations)
Automobile Tanks (Service Stations)
Surface Coating (Protective, Architectural)
Other (Distribution Transfers)
Other Industries
Surface Coating
Degreasing
Other Solvent Use
Miscellaneous
Commercial Activity
Residential
Total
THC
107
2.3
1
30.3
56.8
20
22.7
110
95
35
0.4
55.1
67
602.6
RHC
0
1.7
1.0
30.3
56.8
15
22.7
82.5
71.3
26.2
0.3
41.3
50.3
399.4
NO
X
10
13.6
8
10.5
2.4
23.1
24
91.6
7-81
-------
TABLE A.17
HC AND NO EMISSIONS FROM DISTRIBUTED SOURCES IN ORANGE COUNTY, 1971
X
(From Ref. 30)
Chemical Industry
Mineral Industry
Petroleum Marketing
Other Industries
Organic Solvents Use
Incineration
Domestic and Commercial
Ships and Railroads
Total
Em:
THC
18
31.5
0.2
0.1
49.8
Lssions, Tons/Da
RHC
18
23.6
0.2
0.1
41.9
y
NO
X
0.3
0.7
5.5
0.2
5.2
0.2
12.1
TABLE A.18
HC AND NO EMISSIONS FROM DISTRIBUTED SOURCES
X
IN SAN BERNARDINO COUNTY, 1971
(From Ref. 31)
Emissions, Tons/Day
Petroleum Marketing
Organic Solvents Use
Industrial and Commercial
Aircraft and Railroads
Total
THC
10
8.2
0.7
0.7
19.6
RHC
10
6.2
0.4
0.7
17.3
NO
X
12.4
0.8
13.2
7-82
-------
We concur with the SAI recommendations that, in the long run, all hydro-
carbons except methane, ethane, propane, benzene and acetylene should be
considered "reactive." Using this definition, SAI estimated that HC emis-
sions from petroleum marketing operations are 100 percent reactive, and that
HC emissions from the various solvent uses (degreasing, surface coating,
dry cleaning, etc.) are 75 percent reactive.
Prediction of future distributed source emissions in the study area
requires that we account for two competing effects on the sources involved,
(1) the growth of emissions with increased population and the associated
expansion of industry, commerce, and housing, and (2) the opposing effects
of the California State Implementation Plan (SIP) for the South Coast Air
Basin which seeks to reduce pollutant emissions and bring the area more
nearly into compliance with the national ambient air quality standards.
Prior to promulgation of the SIP, the California State Air Resources Board
(ARB) published estimates of future NO and HC emissions in the SCAB in its
34 X
1970 annual report. These estimates showed no measurable growth in sta-
tionary source emissions of HC. NO emissions from distributed sources
J x
were seen to follow the growth shown in Fig. A.16. We have assumed that
total area source emissions of S0« and particulates scale with local (SCAB)
population. The emissions controls on stationary sources specified in the
SIP will have a major impact on future emissions of HC, but negligible
effect on other pollutants. Stationary source controls will take the
form of requirements for vapor recovery devices to greatly reduce evapora-
tive emissions from gasoline transfer operations, substitution of less
reactive organic solvents for those currently used in degreasing and dry
cleaning operations, and institution of additional control measures affect-
ing surface coating and dry cleaning operations. We have assumed that
the controls proposed in the SIP will be accepted by EPA and will have
reached full application by 1980. These stationary source controls are
expected to effect the reductions in reactive HC emissions shown in Table
*
A.19.
*
Based on emissions reduction estimates in a working paper by N.A. Moyer
of the California State Air Resources Board, December 1973.
7-83
-------
2000
Figure A.16. Growth of Distributed Source NOX Emissions
TABLE A.19
REACTIVE HC EMISSIONS REDUCTIONS UNDER THE SIP
Source Type
Petroleum Marketing
Degreasing
Surface Coating
Dry Cleaning
Miscellaneous
1975
30%
80%
20%
95%
^40%
Expected
1977
55%
90%
30%
95%
^50%
Reductions
1980
(Full Application)
90%
100%
35%
95%
^55%
7-84
-------
To determine the overall effect of these reductions on RHC emissions:
from stationary sources, we have evaluated the hypothetical application
of the final SIP reductions in Table A.19 to the relevant portions of the
study area emissions inventory for 1971. These calculations, outlined in
Table A.20, show that at full application the SIP should reduce RHC emis-
sions from distributed stationary sources to approximately 34 percent of
their anticipated, uncontrolled levels. Hence, this factor was applied
in all projections of stationary source inventories for 1980 and later.
Estimates of the corresponding factors for intermediate years (1975-1980)
may be made by recalculating the righthand portion of Table A.20 using the
appropriate percentage reductions from Table A.19.
It was originally our intention to use maps of existing land-use in
the study area as a basis for geographical distribution of current sta-
tionary source emissions. Unfortunately, the only regional land use map
available was a future land-use plan published by the Southern California
Association of Governments (SCAG) called the SCAG 90 Land Use Plan for
the Southern California Metropolitan Area. This map shows major areas
or concentrations of six types of land-use in a region essentially coter-
minous with the LARTS Study Area shown in Fig. A.2, covering approximately
11,300 square miles of land area. The land use categories employed by
SCAG are
0 Agriculture, Open Space and Vacant Land
1 Low Density Urban (1000 to 8000 persons per square mile)
2 Medium Density Urban (8000 to 20,000 persons per square mile)
3 Major Public Facilities and Military
4 Urban Center
5 Industrial
where the order and type indices are ours, added for ease of identifica-
tion. To organize the geographical distribution of these areas in terms
of our source grid of 2-by-2-mile squares, we overlaid the grid on the
land-use map and employed a coded number to identify the fraction of each
7-85
-------
00
TABLE A.20
HYPOTHETICAL REDUCTIONS IN REACTIVE HYDROCARBON EMISSIONS FOR
1971 UNDER FULL APPLICATION OF THE SIP
Source Type
Petroleum Marketing
Degr easing
Surface Coating
Dry Cleaning
Other
Stationary Sources
Unaffected by SIP
Controls
Total
Emissions, tons/day
Estimated Prt
12
Los Angeles
109.8
71.3
172
18.8
26.5
1.0
399.4
i-SIP Emiss:
30
Orange
18
1.5
7.5
1.9
12.8
0.2
41.9
.ons by County
San n,
Bernardino
10
6.6
0.7
17.3
(See Note)
*
Riverside
5
3.3
0.3
8.6
Study
Area
-Total
142.8
72.8
179.5
20.7
49.2
2.2
467.2
Estimated
SIP
Reduction
128.5
72.8
62.8
19.7
27.1
...
310.9
Post SIP
Emissions
14.3
0
116.7
1.0
22.1
2.2
156.3
NOTE: The category totals shown here include contributions from some of the broader source categories
listed in Tables A.16, A.17, and A.18 (e.g., Commercial and Residential) as itemized in the. respec-
tive references.
Total emissions for Riverside County portion of study area were estimated using land-use emission factors
calculated for San Bernardino County. The various source categories were assumed to contribute in the
same proportions as in San Bernardino (see Tables A.22 and A.23).
-------
square devoted to each land-use within ±5 percent. Thus a typical square
might be 0.6 type 1,. 0.1 type 2, and 0.3 type 5.
The emissions sources listed in Tables A.16, A.17, and A.18 can rea-
sonably be divided into industrial, commercial, and residential categories.
However, there is no simple one-to-one correspondence between these cate-
gories and the land-use types listed above. Only the allocation of indus-
trial emissions is relatively straightforward. We observe that the coarse-
ness of both our fractional allocation scheme and of the map itself pre-
clude individual representation of "islands" of commercial land use less
than about 0.2 square miles (M.3 acres), yet there are obviously many such
neighborhood commercial establishments within most low- and medium-density
urban areas. Probably the most important case in point is the corner gaso-
line service station. The only land-use category on the SCAG map which
is clearly commercial in nature is the urban center, which is often found
attached to an area of type 2, medium-density urban land-use. We also
noted that no inventory emissions are readily identified as type 3. Most
type 3 areas are associated with public and military airports and hence
might properly be used to distribute aircraft emissions. However, in terms
of the stationary source emissions being considered here, military facili-
ties on sizeable pieces of land more nearly resemble residential or
commercial/residential civilian areas. Having made these preliminary
observations, the following ad hoc assumptions were made concerning allo-
cation of inventory emissions to land use types:
1. No emissions are allocated to type 0 areas.
2. Type 4 areas are combined with type 2 areas into an "urban-
commercial" category.
3. Type 3 areas receive the same final allocation as type 1
areas.
4. "Residential" source emissions are spread over areas of type
1 through 4 with types 2 and 4 (urban-commercial) receiving
double weighting compared to 1 and 3.
7-87
-------
5. "Commercial" source emissions are allocated 50% to the urban-
commercial areas and 50% to the type 1 and 3 areas.
6. "Industrial" source emissions are only allocated to type 5
areas.
Table A.21 shows the total land areas in each land-use category and county
within the study area.
In identifying the emissions in the inventories of Tables A.16, A.17,
and A.18 as residential, commercial or industrial, the following assump-
tions were made:
1. Los Angeles—All petroleum marketing emissions with the excep-
tion of those associated with distribution transfers were put
in the residential category.
2. Orange—Reactive HC emissions, dominated by petroleum market-
ing and organic solvents use, were split roughly half-and-
half between the residential and industrial categories.
TABLE A.21
LAND USE AREAS IN THE STUDY REGION
(Units Are 2-by-2-mile Squares)
County
Los Angeles
Orange
San Bernardino
Riverside
Low Density
Urban (1)
190.8
84.8
A5.3
27.6
Urban-Commercial
(2 and 4)
28.5
4.6
1.1
0.3
Public, Military
Facilities (3)
2.2
4.1
3.0
4.0
Industrial
(5)
41.1
11.3
9.3
3.5
Total
Populated
Area
262.6
104.8
58.7
35.4
Study Area
348.5
34.5
13.3
65.2
461.5
7-88
-------
3. San Bernardino—Reactive HC emissions from petroleum marketing
and organic solvents use were split roughly half-and-half
between the residential and industrial categories. Industrial
and commercial RHC was split 50:50. Industrial and commercial
NO was allocated to give emissions per unit area similar to
X
Los Angeles County in the industrial and urban-commercial
areas.
Table A.22 shows the resulting allocation of these inventories by source
type.
The ad hoc rules listed above for allocation of the residential,
commercial, and industrial emissions in Table A.22 to the land-use areas
in Table A.21 are implemented by the following equations:
1 " LA1
Qr\ — ~
2 c A
4- 9 f A 4- A ^ 4- A ? f A 4- A M
T ^ V. A_ T A . ) f A.J ^ <. A^ "3 ' 1
2T1 , "2 1
•4-2 fA 4-A"i+A 9^A +AM
f ^^2 f A^; t- A.J ^^2 T A^; i
V.A.i^
(A. 2)
TABLE A.22
STUDY AREA EMISSIONS OF NO AND REACTIVE HC FROM STATIONARY SOURCES
x
Emissions, Tons/Day
County
Los Angeles
Orange
San Bernardino
Riverside
Residential
NO
• x
24.0
5.2
3.4
2.37
RHC
152.4
20.0
8.0
5.1
Commercial
NO
x
23.1
—
1.0
0.27
RHC
41.3
—
0.2
0.1
Industrial
NO
x
44.5
6.9
8.8
3.3
RHC
205.7
21.9
9.1
3.4
Total
NO
x
91.6
12.1
13.2
5.94
RHC
399.4
41.9
17.3
8.6
Calculated from land-use emission factors for San Bernardino. See Table A.23.
7-89
-------
Q3 = QI (A. 3)
Q4 = Q2 (A.4)
T5
-^ (A.5)
A5
where A. = Area devoted to land-use type i (see Table A.21)
C = Conversion factor from tons/day to kg/hr
T. = Pollutant emissions from type j sources, tons/day (see
Table A.22)
with j = 1 = residential
j = 2 = commercial
j = 5 = industrial
Q. = Land-use emission factor for type i areas, kg/hr/unit
area
Application of these equations for each pollutant and county yields the
land-use emission factors for 1971 given in Table A.23.
Assuming that land-use emissions in Riverside County should be
similar to those in San Bernardino County, we were able to reconstruct
an estimated emissions inventory for Riverside County (see Table A.22)
by solving Eqs. A.I through A.5 for T.. through T,. using the Q's for
San Bernardino and the A's for Riverside.
Finally, the land-use emission factors in Table A.23 were used with
the land use breakdown in each grid square to assemble a single average
daily emission factor for NO and RHC in each square. Figures A.17 and
X
A.18 show the resulting geographical distributions of NO and reactive
X
HC emissions from stationary sources in the study area for 1971. These
7-90
-------
TABLE A.23
LAND-USE EMISSION FACTORS IN THE STUDY AREA, 1971
(Units are kg/hr per 2-by-2-mile Square)
Pollutant
NO
X
Reactive HC
County
Los Angeles
Orange
San Bernardino
*
Riverside
Los Angeles
Orange
San Bernardino
*
Riverside
Low Density
Urban (1)
5.9
2.0
2.9
2.9
27.1
7.7
6.1
6.1
Urban-Commerc ial
(2 and 4)
22.6
4.0
22.3
22.3
73.5
15.4
15.4
15.4
Public, Military
Facilities (3)
5.9
2.0
2.9
2.9
27.1
7.7
6.1
6.1
Industrial
(5)
40.9
23.1
35.8
35.8
189.2
73.3
37.0
37.0
Lacking an emissions inventory for Riverside, we have assumed that the land-use emission factors
for San Bernardino County (SCAB portion) are also applicable in Riverside County.
i
VO
-------
13 14 15 16 17 le 1» 20 21 It 23 24 *» « « « « 10 31 « "
16 37 3« 39
1)2125622211100000000000000000(10000000000000000
113)l(,t6VBA45ZOOOOOOOOQOOOOOC30POOOObOOOftOOOOOO
ll5*A?132o20271S204S431000000000COODt}00000001000000
22 S13684U69bl32010S3S6441 0000000000000000040221 1 0 00
Il0i:i5>56»**2'22051535917103200
0 00 0 n 4 6 13 19 23 21 9 19 5 13 9 9 25 6 2 6 6 13 6 4 5 13 " 6 3 3 12 7 32 10 3 3 10 6 9 29 12 4 1 1
„ 0 , z j t 11 15 15 16 IB 2> 23 ZB 14 It 5 15 13 9 9 t 9 4 4 7 15 16 23 t3 15 16 6 9 22 18 2 11 15 9 9 2 4 6 5 1
I 1 3 1ft 21 IS IB 1» 16 21 26 30 6 8 9 2 2 I 20 9 4 2 4 12 5 4 10 7 6 16 14 14 IB 0 0 I 15 1 20 0 0 I I
0 0 0 7 9 14 B 5 B 16 19 41 37 30 23 5 4 0 4 13 1* 20 22 2 ° 0~ 1 12 0 0 0 0 4 6 2 3 7 15 0 0 0 0 0 0
15 0 0 0 0 0 11 20 6 11 15 13 16 IB 9 23 Z3 9 16 6 2 3 1 3 I 1 0 1 1 1 2 0 0 0 I 9 9 2 2 10 1 0 0 0 0 0
14 OOOC806 14 1» 11 13 9 16 15 16 6 16 23 6651000000000010163321 12210
13 OOOOflP1334914139139686276991J46l°OOnc0120932200Z33ZO
12 OOOOC04l6136131614a20B6Z>34642Z<2<000011545S0000133100
11 OOI>OCn07253037B379209a624410171r'll°°00231000001Z2000
10 0000001141*13132020616125222652(221 000000400000000000
On dfi 004 16 6 16 la 23 13 20 94262Z5 15 10 ZlOOC 0000000000000008
000002654 16 30 37 22 96226ZZ2ZZZ219(>0000000000000000
OHOOOC3>«lS2912aO|9644222T221000000000000000000
OOCP00001 12 400000042262 19 3200POOOOOOOOOOOOOOOO
5 l)01-to(100000«00000042Z152ll6Z?lll"00(>000000000000
onoOluODOOOOOOOOOlZZZB IB 2224*1000000000000000
1 oOnI??(1i*OG"t000090000000110<)l>??000900000000000
Figure A.17.
Geographical Distribution of Stationary Source NOX Emissions
in 1971, kg/hr
-------
I
10
II 12 13 1*
20 II »
STATIONARY SOURCC tM!SS!ONS or NCR (KG/HR)
(4 25 26 27 » " 30 31 12 33 »
W 36
I
VO
Co
2* 0 « S |1 24 « 97 II 5 5 3
23 5 60 4« ?7 ?T 27 43 38 33 16
I* 16 T6 76 60 76 B5 43 76 38 0 0 22 It
23 9 64 27 32 22 53 92 «2 12* 70 t 0 19 2« 16
22 24 S3 27 32 16 44 27 36 27 60 85 16 22 14 24 27 19 16 I 0 0 • 0 • 0 0 0 0 0 0 0 0 0 0 0 0 5 0 4 5 > 1 « I 0
21 3 5 0 3 1» II 27 24 27 11 22 127 27 24 24 66 64 IB 71 57 9 61 21 16 3 1 8 14 3 4 2 I 0 1 0 2 14 S 5 6 6 6 4 I •
20 0 0 0 0 0 3 14 11 22 31 36 10a 43 M 29 36 27 34 32 9 26 60 IT fT 27 27 2T 2T 9 7 9 4 0 7 2 7 6 T 11 1' 12 * 4 I *
14 0 0 0 0 0 It 27 46 64 71 69 36 87 24 60 36 43 111 27 11 27 17 U fT 16 24 60 43 ' 6 6 14 8 14 11 6 6 12 » 12 10 11 6 > 1
l> 0 0 B 14 27 41 96 92 99 60 73 73 119 M 6t {4 70~ 60 4] 41 (7 M It It 12 68 6« 10B IS 16 IB t 1 21 20 4 12 IT 10 10
17 1 9 5 14 39 to 64 71 6t 62 M t7 117 17 32 41 B B B tt 41 It B I* 97 IB 16 12 B t IB IS IS IB 0 0 1 19 2 4 0 0 1 1 4
16 0 0 0 0 22 36 97 IB 22 12 61 7i iBt 171 ]4i lol 2» 1« B It 40 it
II 0 0 0 0 0 0 »2 157 41 37 60 41 6j 4] J7 1J 27 It4 JT 41 2* 41 14 fl • 0 0 0 0 0 0 2 9 0 12 S 4 9 1
12 0 0 0 0 0 0 It 76 60 27 60 76 97 U t< U <7 10B 1S7 tl 11 B 7 • B 4 t 0 0 0 2 17 9 7 T 0 1
0 0 tB 111 141 171 IB 17] 41 tt 16 32 {7 B 14 14 14 94 94 29 9 3 1 0 0 0 4 9 1 0 0 0 0 0 1
0 0 IS 57 to 60 92 92 27 76 97 24 B t B 22 17 B 2T B 4 1 0 0 0 0 0 0 6 0 0 0 B 0 0
0 0 1> 76 27 76 BO 101 91 02 41 It B 21 B B It 4T 14 6 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0
0 11 »T 22 It 71 111 170 78 16 27 t B 22 B B B B B B B 4 0 0 00 0 0 0 0 0 0 0 0 0
0 0 14 19 27 70 112 97 IB 0 1 2t 21 14 IS B B B 29 t 6 3 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0005 47 It 000002 12 BT (IB 40 10 7?u 000000000 000
0000000000000 12 BB 47 ttT21»l> " '1000000000
000000000000000 10 »4B2T>>834*72000000000
1
Figure A.18. Geographical Distribution of Stationary Source Reactive HC
Emissions in 1971, kg/hr
-------
average daily emissions are modulated by the time distribution function
given in Table A.I which allocates 80 percent of the daily total uniformly
over the daytime hours (6:00 a.m. to 6:00 p.m. local time) and the re-
mainder to the hours from 6:00 p.m. to 6:00 a.m.
7-94
-------
APPENDIX B
TRAJECTORIES OF AIR MASSES USED WITH DIFKIN AIR QUALITY MODEL
The air trajectories shown in this appendix were used in the air
quality computations performed with the DIFKIN air quality model. These
trajectories were determined using meteorological data for September 29,
1969, a day in which an air pollution episode occurred in the Los Angeles
Basin. See Ref. 1 for a description of the criteria used to obtain these
trajectories.
7-95
-------
'PASADENA
• AZUSA
• HOLLYWOOD
• LOS ANGELES
WHITTIER
ANAHEIM
1200
Figure B.I. Trajectory of Air Mass Arriving at Anaheim at 1200 PDT
(Trajectory nodes are separated by one-hour intervals)
7-96
-------
• LENNOX
WHITTIER
0100
1300
ANAHEIM
Figure B.2. Trajectory of Air Mass Arriving at Anaheim at 1300 PDT
(Trajectory nodes are separated by one-hour intervals)
7-97
-------
>PASADENA
• AZUSA
• HOLLYWOOD
• LOS ANGELES
• WHITTIER
ANAHEIM
1400
Figure B.3. Trajectory of Air Mass Arriving at Anaheim at 1400 PDT
(Trajectory nodes are separated by one-hour intervals)
7-98
-------
• PASADENA
• HOLLYWOOD
• LOS ANGELES
0100
• LENNOX
AZUSA
1300
'WHITTIER
ANAHEIM.
LONG BEACH
Figure B.4. Trajectory of Air Mass Arriving at Azusa at 1300 PDT
(Trajectory nodes are separated by one-hour intervals)
7-99
-------
• PASADENA
• HOLLYWOOD
' LOS ANGELES
LENNOX
AZUSA
1400
0100
WHITTIER
ANAHEIM (
Figure B.5. Trajectory of Air Mass Arriving at Azusa at 1400 PDT
(Trajectory nodes are separated by one-hour intervals)
7-100
-------
PASADENA
• HOLLYWOOD
0100
WHITTIER
' AZUSA
1500
ANAHEIM.
Figure B.6. Trajectory of Air Mass Arriving at Azusa at 1500 PDT
(Trajectory nodes are separated by one-hour intervals)
7-101
-------
I
I-1
o
LOS ANGELES
1800
LENNOX
WHITTIER
1000
\
Figure B.7. Trajectory of Air Mass which goes by Riverside and Starts
±n Orange County at 1000 PDT (Trajectory nodes are separated
by one-hour intervals)
-------
>LOS ANGELES
1900
• LENNOX
• WHITTIER
1100
Figure B.8. Trajectory of Air Mass which goes by Riverside and Starts
in Orange County at 1100 PDT (Trajectory nodes are separated
by one-hour intervals)
-------
o
-c-
LOS ANGELES
LENNOX
1900
Figure B.9. Trajectory of Air Mass which goes by Riverside and Starts
in Orange County at 1200 PDT (Trajectory nodes are separated
by one-hour intervals)
-------
• PASADENA
•AZUSA
• HOLLYWOOD
• LOS ANGELES
EL SEGUNDO
POWER PLANT
0100
• LENNOX
•WHITTIER
1500
• ANAHEIM
Figure B.10.
Trajectory of Air Mass Departing from El Segundo Power
Plant at 0100 PDT (Trajectory nodes are separated by one-hour
intervals)
7-105
-------
.PASADENA
• AZUSA
.HOLLYWOOD
• LOS ANGELES
WHITTIER
1500
REDONDO BEACH
POWER PLANT
Figure B.ll.
Trajectory of Air Mass Departing from Redondo Beach Power
Plant at 0100 PDT (Trajectory nodes are separated by one-hour
intervals)
7-106
-------
APPENDIX C
ADDITIONAL EMISSIONS ESTIMATES FOR STUDY AREA
This appendix contains the estimated emissions for 1970 which were
used as the reference for the rollback calculations. Also presented are
emissions estimates with and without electric cars for 1990 and 2000 for
the scenario which assumes additional delays in the implementation of auto
emission controls. Only the figures for NO, HC, and CO are given for
these cases since the others are the same as shown in Tables 2.5 and 2.6
for the baseline cases and Tables 3.4 and 3.5 for the cases with electric
cars. These correspond to the emissions for the case 1980d and have been
depicted in Fig. 3.7 for nitric oxide. No air quality calculations were
carried out for these cases.
TABLE C.I
STUDY AREA EMISSIONS FOR 1970-71
(Tons/Day)
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
NO
HC
CO
SO,
Particulates
455
77.5
78.6
44.6
655.7
918.4
472.3
—
30.8
1421.5
9719.7
—
—
—
9719.7
20.2
178
175
55.4
428.6
82.2
88.3
7.5
10.1
188.1
7-107
-------
TABLE C.2
1990 BASELINE STUDY AREA EMISSIONS WITH DELAYED IMPLEMENTATION
OF EXHAUST EMISSION CONTROLS FOR CONVENTIONAL CARS
(Tons/Day)
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
NO
280.7
140.8
116.2
53.7
591.4
HC
73.5
160.7
—
37.1
271.3
CO
1103.4
—
—
—
1103.4
See Table 2.5.
TABLE C.3
2000 BASELINE STUDY AREA EMISSIONS WITH DELAYED IMPLEMENTATION
OF EXHAUST EMISSION CONTROLS FOR CONVENTIONAL CARS
(Tons/Day)
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
NO
305.7
181.1
98.1
57.6
642.5
HC
78.5
160.7
—
39.8
279.0
CO
1166.9
—
—
—
1166.9
See Table 2.6.
7-108
-------
TABLE C.4
1990 STUDY AREA EMISSIONS WITH DELAYED IMPLEMENTATION OF
EXHAUST EMISSION CONTROLS FOR CONVENTIONAL CARS AND 80% ELECTRIC CARS
(Tons/Day)
NO
HC
CO
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
74.9
140.8
152.2
53.7
421.6
34.1
160.7
37.1
231.9
803.3
803.3
See Table 3.4.
TABLE C.5
2000 STUDY AREA EMISSIONS WITH DELAYED IMPLEMENTATION OF
EXHAUST EMISSION CONTROLS FOR CONVENTIONAL CARS AND 100% ELECTRIC CARS
(Tons/Day)
NO
HC
CO
Vehicular
Stationary Area Sources
Power Plants
Oil Refineries
Total
25.5
181.1
105.0
57.6
369.2
25.9
160.7
39.8
226.4
770.3
770.3
See Table 3.5.
7-109
-------
7-110
-------
APPENDIX D
THE LINEAR ROLLBACK FORMULA2
The rollback formula assumes that pollutant concentrations and emis-
sions are proportional and is given by
(PAQ) - (DAQ)
(PAQ) - (B)
where R = fractional reduction in emissions
PAQ = present air quality
DAQ = desired air quality
B = background concentration of pollutant
Thus the rollback formula purports to yield the reduction in emissions
required to achieve the desired air quality, e.g., the air quality stand-
ard, given the current air quality and the background level of the pollu-
tant, i.e., the concentration unavoidably present due to natural causes.
It should be noted that, since the air quality standards are set on
a worst-case basis, a worst-case approach is implicit'in using the roll-
back formula with the air quality standard as the desired air quality.
Thus, for the present air quality (PAQ) we must use the highest concen-
tration recorded in the region during the period of interest.
Caution is necessary in using the rollback equation to relate emis-
sions and air quality. While the rollback approximation may fit some
cases [e.g., carbon monoxide (CO)], its use for relating emissions and
secondary pollutants such as nitrogen dioxide (N0_) and ozone (0~) is
questionable because these two pollutants are produced by chemical pro-
cesses which are highly nonlinear. The use of Eq. D.I is also suspect
7-111
-------
even with highly reactive primary pollutants such as reactive hydrocarbons
(HC) and nitric oxide (NO).
Since our aim is to determine the desired air quality (DAQ) levels
given a current air quality (PAQ), the background level (B), and the frac-
tional reduction in emissions (R), we must solve for DAQ in Eq. D.I,
which yields
DAQ = PAQ(1 - R) + RB (D.2)
In using Eq. D.2, the factor PAQ would be the baseline air quality level
with no electric car use. The factor R would be determined by the emis-
sions reduction caused by the increasing use of electric vehicles.
One difficulty in using Eq. D.2 is that the baseline air quality
(PAQ) for 1980, 1990, and 2000 is not readily available. Two approaches
may be followed to circumvent this problem. The first is to predict the
baseline air quality by means of Eq. A.2 itself. This would be done by
using for PAQ the known air quality for a previous year such as 1970,
and estimating R from
R - 1 - ~— ; k = 1980, 1990, 2000
where E, are the baseline emissions for the year k . The second ap-
tc
proach is to use the values the GRC model predicts as the baseline air
quality levels for 1980, 1990, and 2000. This has the advantage of pro-
viding a common starting point for the air quality impact analysis for
both the GRC model and the rollback formula. Both approaches have been
used and compared in the text.
7-112
-------
REFERENCES
1. J.R. Martinez, R.A. Nordsieck, An Approach to the Analysis of the
Air Quality Impact of Electric Vehicles, General Research Corpora-
tion RM-1831, February 1974, (also Task Report 6).
2. D.S. Earth, "Federal Motor Vehicle Emission Goals for CO, HC, and
NO Based on Desired Air Quality Levels," J. Air Poll. Control
X -------
Assoc.. Vol. 20, No. 8, pp. 519-523, August 1970.
3. A.Q. Eschenroeder, Comments on California Air Quality Standards;
Transportation Control Strategy, General Research Corporation
IM-1741, April 1973.
4. A.Q. Eschenroeder, J.R. Martinez, "Concepts and Applications of
Photochemical Smog Models," in Advances in Chemistry, Series 113,
American Chemical Society, 1972.
5. A.Q. Eschenroeder, J.R. Martinez, and R.A. Nordsieck, Evaluation of
a Diffusion Model for Photochemical Smog Simulation, General Research
Corporation CR-1-273, October 1972.
6. G. Houser, Population Projection for the Los Angeles Region, 1980-
2000, General Research Corporation RM-1842, November 1973, (also
Task Report 2).
7. W. Hamilton, G. Houser, Transportation Projections for the Los
Angeles Region. 1980-2000, General Research Corporation RM-1858,
November 1973, (also Task Report 3).
8. Federal Register. Vol. 38, No. 80, April 26, 1973; No. 124, June
28, 1973; No. 126, July 2, 1973; No. 161, August 21, 1973.
9. Federal Register. Vol. 36, No. 67, April 7, 1971; Vol. 38, No. 14,
January 22, 1973; No. 110, June 8, 1973; No. 120, June 22, 1973;
No. 126, July 2, 1973.
10. Russell E. Train, Administrator, U.S. Environmental Protection
Agency, Legislative Proposals to Amend the Clean Air Act of 1970,
March 22, 1974.
11. R.A. Nordsieck, Estimates of Pollutant Emission Factors for California
Motor Vehicles: 1967-2000, General Research Corporation RM-1849,
October 1974.
7-113
-------
REFERENCES (Cont.)
12. Los Angeles Air Pollution Control District, Profile of Air Pollu-
tion Control, 1971.
13. A.R. Sjovold, Electric Energy Projections for the Los Angeles
Region, 1980-2000. General Research Corporation RM-1859, November
1973, (also Task Report 5).
14. P.J.W. Roberts, M. Liu, S.D. Reynolds, and P.M. Roth, "Extensions
and Modifications of a Contaminant Emissions Model and Inventory
for Los Angeles," Appendix A of Further Development and Validation
of Simulation Model for Estimating Ground Level Concentrations of
Photochemical Pollutants, Systems Applications, Inc. Report R73-15,
January 1973.
15. Compilation of Air Pollutant Emission Factors (Second Edition), U.S.
Environmental Protection Agency, AP-42, April, 1973.
16. W.F. Hamilton, Usage of Electric Cars in the Los Angeles Region,
1980-2000, General Research Corporation RM-1891, April 1974, (also
Task Report 10).
17. The State of California Implementation Plan for Achieving and Main-
taining the National Ambient Air Quality Standards, Revision 4, South
Coast Air Basin Plan, December 31, 1973.
18. W.A. Glasson and C.S. Tuesday, "Inhibition of Atmospheric Photo-
oxidation of Hydrocarbons by Nitric Oxide," Environmental Science
and Technology, Vol. 4, No. 1, January 1970, pp. 37-44.
19. W.E. Wilson, A. Levy, "A Study of Sulfur Dioxide in Photochemical
Smog I. Effect of S0_ and Water Vapor Concentration in the 1-Butene/
NO /SO System," J. Air Pollution Control Assoc., Vol. 20, No. 6,
X ^
June 1970, pp. 385-390.
20. W.E. Wilson, A. Levy, and D.B. Wimmer, "A Study of Sulfur Dioxide
in Photochemical Smog II. Effect of Sulfur Dioxide on Oxidant Forma-
tion in Photochemical Smog," J. Air Pollution Control Assoc., Vol.
22, No. 1, January 1972, pp. 27-32.
21. Ten-Year Summary of California Air Quality Data, 1963-1972, State of
California Air Resources Board, January 1974.
22. G.M. Hidy and S.K. Friedlander, "The Nature of the Los Angeles
Aerosol," 2nd Clean Air Congress, IUAPPA, Washington, D.C., December
1970.
7-114
-------
REFERENCES (Cont.)
23. J.R. Martinez, R.A. Nordsieck, and M.A. Hirschberg, User's Guide to
Diffusion/Kinetics (DIFKIN) Code, General Research Corporation
CR-2-273/1, December 1973.
24. P.J.W. Roberts, P.M. Roth, and C.L. Nelson, "Contaminant Emissions
in the Los Angeles Basin—Their Sources, Rates, and Distribution,"
Appendix A of Development of a Simulation Model for Estimating
Ground Level Concentrations of Photochemical Pollutants, Systems
Applications, Inc., Report 71SAI-6, March 1971.
25. Federal Register. November 10, 1970, Vol. 35, No. 219, Part II,
p. 17311.
26. A.H. Rose, Jr., R. Smith, W.F. McMichael, and R.E. Kruse, "Compari-
son of Auto Exhaust Emissions in Two Major Cities," Journal of the
Air Pollution Control Association, Vol. 15, No. 8, August 1965;
27. Highway Capacity Manual 1965, Highway Research Board Special Report
87, National Academy of Sciences, National Research Council Publica-
tion 1328, Washington, D.C., 1965.
28. Federal Register. April 26, 1973, Vol. 38, No. 80, p. 10317.
29. Federal Register, August 21, 1973, Vol. 38, No. 161, Part I,
p. 22474.
30. Emissions Inventory for 1971 Calendar Year, County of Orange Air
Pollution Control District, May 1972.
31. 1971 Annual Report, San Bernardino County Air Pollution Control
District.
32. Rules and Regulations, County of Los Angeles Air Pollution Control
District.
33. Statistical Abstract of the United States, 1973, U.S. Department
of Commerce.
34. Air Pollution Control in California; 1970 Annual Report, State of
California Air Resources Board, January 1971.
7-115
-------
7-116
-------
TASK REPORT 8
PARAMETRIC ENERGY, RESOURCE, AND NOISE IMPACTS
OF ELECTRIC CARS IN LOS ANGELES
A.R. Sjovoid
-------
ABSTRACT
Widescale use of electric cars in the South Coast Air Basin (SCAB)
as alternatives to conventional autos can reduce the consumption of primary
energy that otherwise might be used in auto travel. The degree of reduction
depends strongly on future improvements in gasoline fuel economies for
conventional autos. The battery recharge energy requirements for a 100%
SCAB electric car population can potentially be accommodated within the
forecast capacity of the SCAB electrical utilities if the electric cars
are recharged during the normal off-peak period. In this mode prior to
2000 recharging would be associated with the use of oil-fired steam-electric
plants in the basin on high demand days. During low-demand periods
significant amounts of recharge energy would come from coal, gas, and,
to a lesser extent, nuclear generation such that over the year substantial
savings in petroleum might be obtained.
For large SCAB electric car populations, significant additions to
total U.S. demands of some materials will occur, depending on the battery
technology employed. Critical material demands in this respect are
nickel for nickel-zinc batteries, and lithium and graphite for lithium-
sulfur batteries.
For a largescale nationwide implementation, critical material de-
mands to be considered are for lead, antimony, nickel, titanium metal,
lithium, and graphite. It is assumed that all of the metals will be
highly recycled.
Electric cars also possess the potential of decreasing community
noise levels arising from urban traffic. Potential reductions in noise
intrusions appear more significant than reductions in general background
noise levels.
-------
ii
-------
CONTENTS
SECTION
1
2
3
4
ABSTRACT
SUMMARY
INTRODUCTION
ENERGY IMPACTS
2.1 Primary Energy
2.2 Primary Energy Consumption
2.3 Gasoline Consumption
2.4 Power Plant Generation
2.5 Petroleum Savings
RESOURCE IMPACTS
3.1 SCAB Electric Car Resource Impacts
3.2 US and World Reserves of Electric Car
Materials
3.3 Resource Implications of a National
Implementation
3.4 Price Trends of Mineral Resources
3.5 Other Resource Considerations
POTENTIAL IMPACT OF SCAB ELECTRIC CARS ON URBAN
NOISE
4.1 Urban Noise Environments
4.2 Comparison of Electric and Conventional Auto
Noise Generation
4.3 Effect of Electric Car Use on Community
Noise
4.4 Electric Generation Power Plant Noise
PAGE
i
ix
8-1
8-4
8-4
8-16
8-21
8-22
8-26
8-29
8-34
8-58
8-60
8-64
8-66
8-68
8-68
8-70
8-78
8-79
REFERENCES . 8-81
111
-------
iv
-------
ILLUSTRATIONS
NO. PAGE
2.1 Comparison of Alternative Uses of Coal for Automotive
Transportation 8-14
2.2 Decrease in Energy Consumption for Private Auto Travel
Due to Electric Car Usage Relative to Average Gasoline
Car 8-17
2.3 Primary Energy Consumption for Total Auto Travel 8-19
2.4 Private Auto Gasoline Consumption 8-21
2.5 Electric Power Demand Profiles 8-23
2.6 Off-Peak Power Generation for Different Battery Techno-
logies and Introduction Scenarios 8-25
2.7 Petroleum Savings as a Function of Electric Car Usage 8-27
3.1 Annual Battery Sales in SCAB 8-32
3.2 Annual Battery Recycle in SCAB 8-32
3.3 Lead Requirement as Function of Usage in SCAB 8-35
3.4 Lead Requirement as Function of Time 8-36
3.5 Antimony Requirement as Function of Usage 8-41
3.6 Antimony Requirements as Function of Time 8-42
3.7 Nickel Requirement as Function of Usage in SCAB 8-44
3.8 Nickel Requirement as Function of Time 8-45
3.9 Lithium Requirement as Function of Usage in SCAB 8-50
3.10 Lithium Requirement as Function of Usage 8-52
3.11 Lithium Requirement as Function of Time 8-52
-------
ILLUSTRATIONS (Cont.)
NO. . PAGE
3.12 Graphite Requirement as Function of Usage in SCAB 8-53
3.13 Graphite Requirement as Function of Time 8-54
3.14 Electric Car Lead Inventories 8-61
3.15 Electric Car Antimony Inventories 8-61
3.16 Electric Car Lithium Inventories 8-62
3.17 Electric Car Graphite Inventories 8-62
3.18 Electric Car Nickel Inventories 8-63
3.19 Electric Car Zinc Inventories 8-63
3.20 Electric Car Titanium Inventories 8-64
4.1 Comparison of Electric Car and Conventional Car Engine
Noise 8-75
vi
-------
TABLES
NO. PAGE
2.1 Average Energy Consumption Rates Assumed for Gasoline-
Powered Autos 8-9
2.2 Average Primary Energy Consumption Rates Assumed for
Electric Cars 8-11
2.3 Projected SCAB Electric Car Populations for Several
Postulated Car Sales 8-19
3.1 Relationship of SCAB Electric Car Requirements to US
Lead Demand 8-38
3.2 Relationship of SCAB Electric Car Requirements to US
Lead Supply 8-39
3.3 Annual Requirement for New Zinc to Support SCAB Electric
Car Usage 8-46
3.4 Annual Requirement for New Titanium to Support Electric
Car Usage in SCAB 8-48
3.5 Summary of Resource Impacts 8-56
3.6 Estimated US and World Reserves 8-59
3.7 Price Trends of Materials Important to Electric Cars 8-65
4.1 A-Weighted Outdoor Noise Levels 8-69
4.2 Exterior Noise Levels for Various Conditions and 1970
Vehicle Types 8-71
4.3 Comparison of Conventional Auto and Electric Car Motor
Noise Levels, Cruise Conditions 8-77
vii
-------
viii
-------
SUMMARY
i
Potential impacts of electric car use on energy consumption,
material supplies and demands, and community noise are analyzed. The
primary impacts are related parametrically to electric car use only in
the South Coast Air Basin of Califronia (Los Angeles region) and cover
the years 1980 to 2000.
Electric cars as alternatives to conventional autos have the
potential of reducing the consumption of primary energy. For both
alternatives a significant source of primary energy prior to the year
2000 will be crude oil, with electrics being somewhat more efficient
in converting crude oil to vehicle-miles than the average conventional
auto of the future. The degree of reduction depends on future battery
technologies for the electrics, and improvements in gasoline mileage
for conventional cars. For 100 percent electric car population in the
year 2000, utilizing advanced battery technology, primary energy consumption
for automotive travel would be two-thirds of the 1974 level. Conventional-
auto gasoline mileages would have to more than double to obtain the same
reduction. Further reductions are possible with electric cars if future
power plant efficiencies should rise above the existing 36 percent for
Southern California Edison (SCE) oil fired plants.
The battery-charging requirements for a 100 percent electric car
population can be met with the existing and forecast unused off-peak
generating capacity. No new power plants need be constructed for this
purpose in the South Coast Air Basin beyond the additions already planned
by utilites in the Basin.
High levels of electric car usage in SCAB can cause significant
increases in total US demands for at least some of the materials associated
with any given battery alternative. Estimated impacts on materials
ix
-------
resources are summarized on the next page. Existing technology with lead-
acid batteries will significantly increase demands for lead and antimony,
even at high recycle rates. Nickel-zinc batteries will increase demands
significantly for nickel and much less so for zinc. Zinc-chlorine
batteries will impact zinc demands even less, but they impose significant
demands on titanium metal supplies. However, titanium metal is a very
small fraction of total titanium demands, so that electric car demands
are not significant in terms of total titanium demand. Lithium-sulfur
batteries will cause significant increases in demands for lithium and
graphite.
The assessments assume that the following materials will be highly
recycled if employed in electric car batteries: lead, antimony, nickel,
zinc, metallic lithium, and metallic titanium. Of these materials, the
US is self-sufficient only in lithium. However, world reserves of all
the other materials appear adequate to support 100 percent electric car usage
in SCAB. World reserves are inadequate to support a nationwide electric
car implementation utilizing lead-acid batteries. They are most ample
to support nationwide implementations with lithium-sulfur batteries, and
appear adequate to support nationwide implementations with either zinc-
chlorine or nickel-zinc batteries.
Electric cars should prove to be quieter than the conventional
autos they replace. Large-scale usage of electric cars should cause
perceptible decreases in background traffic noise levels, and significant
reductions in peak auto noise intrusions.
x
-------
SUMMARY OF RESOURCE IMPACTS
Ouantitits, Thousands of Tons per Year
Battery Type Material
Lead-Acid Lead
An t imor.y
Kickel-Zinc Nickel
Zinc
Zinc- Zinc
Chlorine
Titanium
(metal )
Chlorine
Lithium- Lithium
Sul£ur Graphite"
Sulfur
I'S Primary
Production
1968 2000 (Range)
354
1.9
15
529
305
(0)
8,400 26
2.9
3.0
11,000 28
520-1,120
2.5-4.8
36-52
786-1,500
670-1,610
(0)
,400-43,900
9.4-14.4
4.0-4.7
,000-45,000
US rrimarv Penand
1968
88C
21.1
160
1,406
440
(13)
8,400
2.6
60
10,000
2000 (Range)
1,300-2,800
28-52
382-550
2,040-4,000
960-2,160
(62-234)
26,400-43,900
8.7-13.1
80-135
26,000-41,500
*
Equilibrium.
Annual Electric
Car Requirement
for New Material
with 2-Year Bat-
terv Life and 90%
Recycle (SCAB)
22
0.6
29
20
12
9.5
130~
3.2
43*
130*
Usage
Conditions
(SCAB Level
and Year)
17%, 1980
17%, 1980
46%, 1990
46%, 1990
100%, 2000
100%, 2000
100%, 2000
100%, 2000
100%, 2000
100%, 2000
Potential
Problem
((Electric Car Requirement as
Percent of National Primary
Demand for Given Usage Year)
SCAB
Implementation
1.7%
2.1
7.8
0.8
' 0.4
0.6
0.4
29tf
40
0.4
**
Nat ionwide
Implementation
33%
43
157
16
8
4.4.
12' '
8
587"
800
8
Equilibrium conditions fcr the fraction of electric cars in the total population are assumed; that is, the electric car population fraction
is relatively constant and not undergoing any long-term buildup.
The scaling factor to be applied to the "Annual Requirements" column for nationwide electric-car use is approximately 20.
No recycle assumed.
Electric car requirements for these materials would impact on production capacity for metallic forms.
-------
xii
-------
1 INTRODUCTION
Electric cars replacing conventional autos have been suggested as
possible solutions to pressing urban problems such as air pollution and
1—3
the growing rate of energy consumption. Several analyses have
pointed to the potential for electric cars, especially with improved
battery technology, to alleviate our energy supply problems. However,
the ultimate impacts of electric cars on energy supply, air pollution,
and economic resources are not readily deduced. They will depend on
future conditions and the ways in which electric cars come into general
use.
The objective of this paper is to estimate the likely impacts of
electric car sales and use on energy consumption, material resources, and
urban noise. The analyses are part of a comprehensive impact study of
electric cars, focusing on the South Coast Air Basin of Southern California
and covering the years 1980 to 2000. This paper depends on several previous
tasks of this study that produced forecasts of population, energy,
*
transportation, electric car technology, and future economic conditions.
**
It also depends strongly on a companion study task which developed
alternative scenarios governing the introduction and operation of electric
cars in the South Coast Air Basin, with corresponding estimates of sales
and use. Additional work, largely based on Ref. 10, gave a qualitative
assessment of all potential electric car impacts, and thereby provided
an additional basis for the more quantitative assessments presented in
this paper.
The impact assessments presented here focus strongly on a system of
battery-powered electric cars which are used daily and recharged primarily
at home during the night. No new electric power plants, beyond those
Task Reports 1-5 (also Refs. 4-8)
**Task Report 10 (also Ref. 9).
8-1
-------
already forecast, are likely to be required; the recharging during off-
peak periods may utilize generation from several types of power plants
(e.g., coal, gas, nuclear) but will rely strongly on local oil-burning
power plants, especially during peak demand days. The electric cars
are expected initially to fulfill a "second car" role, utilizing lead-
acid batteries. Future battery developments may center on several
alternatives—nickel-zinc, zinc-chlorine, or lithium-sulfur—which pro-
mise to provide from three to ten times the specific energy storage of
lead-acid batteries.
Although cars with lead-acid batteries may find early acceptance
by a small fraction of the auto buying public, their limited ranges
will likely prevent them from ever achieving wide acceptance. Consequently,
large-scale use of electric cars must rely on the later development of
better batteries. Likely scenarios therefore involve mixes of battery
technology within the electric car population. Furthermore, it is not
certain that advanced batteries can directly replace earlier ones. For
example, the configuration of a high-temperature lithium-sulfur battery
system is likely to be quite different from that of a lead-acid battery
system.
As a matter of convenience in this paper, and in general accord
with its parametric character, we have avoided the ticklish question of
just how and when one steps from one battery technology into the next.
Instead we first establish the electric car energy and resource requirements,
as functions of the level of electric car population and usage, for a
given electric car technology. These parametric relationships are then
used to assess the energy and resource impacts of each of the electric
car usage scenarios developed in Ref. 9. In the calculation of these
impacts, it is assumed that only "pure" systems (no mix of electric
car technologies) are employed in each specific time frame. This leads
to clearly-defined and useful results, but admittedly may be less than
realistic. It seems preferable, however, to basing impact calculations
8-2
-------
on detailed schedules of technological and marketing progress, which must
be arbitrary in nature because they cannot be accurately forecast.
The assessment of electric car impacts in this paper begins in
Sec. 2, with an analysis of the impact on energy supply and demand in
the South Coast Air Basin (SCAB). The analysis includes effects on total
primary energy, auto transportation energy, electric energy, and gasoline
demand. Section 3 treats impacts on resources, primarily the materials
to support the alternative battery technologies. Section 4 deals with
the impact of electric car usage on vehicle-generated noise and its
contribution to urban noise.
Impacts on air quality and the economy are treated in companion
study tasks. Social and cultural impacts are intimately interwoven with
the considerations and evaluations involved in the development of
scenarios for electric car sales and usage, and are therefore treated
in the study task reported in Ref. 9.
8-3
-------
2 ENERGY IMPACTS
The source of primary energy for energizing electric cars in the
South Coast Air Basin has already been identified. ' Under all demand
conditions in 1980, off-peak generation for recharging electric cars will
come from oil-fired plants in the basin. In 1990, off-peak generation
during high demand seasons will be predominantly from oil, while low-demand
seasons will make available some off-peak generation from coal and gas
fired plants. In 2000, off-peak generation will come from several sources
during both high and low-demand seasons, with oil-fired plants the last
to be used at high levels of electric car usage. Oil-based generation
will be an important, though declining source, of recharge energy throughout
the forecast period. Present and future autos, some of which would be
displaced by electric cars, would depend on the same source, oil; indeed
the same refineries produce all the required products—distillate and
residual fuel oils for power plants, and high-octane gasolines for internal
combustion engines. Consequently, the analysis of the impact on primary
energy identifies the changes in required crude oil supply and traces the
path of energy flow through the refinery to each end use.
2.1 PRIMARY ENERGY
2.1.1 Refinery Operations
A modern refinery represents a complex process which is capable of
adjusting the relative outputs of gasolines, fuel oils, and other
products over a wide range to suit demand. The simplest (and least
energy-consuming) process is the separation of the crude oil into its
natural constituents by fractional distillation. Unfortunately, the
yields of the various fractions depend on the nature of the source crude;
fractionation usually does not yield sufficient automotive gasoline.
However, simple fractionation would undoubtedly provide ample product
for power plant fuel oil, if that were the primary product needed. To
increase the amount of high-octane gasoline, processes such as catalytic
reforming and cracking, thermal cracking, polymerization, and alkylation
8-4
-------
are added (along with an increased energy input).
According to data reported by Stanford Research Institute, refineries
in California produced 42 percent of their total product as motor gasoline
in 1970. Based on SRl's data, the total energy output of the refineries
as useful products was 88 percent of the total energy input (crude plus
external energy). Thus the refineries were running at 0.88 overall thermal
12
efficiency. Another study indicates that the average difference for US
refineries between energy input and energy output is 11 percent, for an
overall thermal efficiency of 0.89.
It is difficult to allocate the energy used in refining among the
products, although it is the relatively high demand for automotive gasoline
which establishes the need for the more energy-consuming processes. In
13
one study, refinery energy input was allocated as a constant percentage
to all end products. That is, the efficiency was taken to be the same
for power-plant fuel as for gasoline. However, the same study used a
quite different allocation in dealing with refinery pollutant emissions,
indicating that 82 percent by weight of refinery air emissions are associated
with the catalytic cracking processes, and only 0.3 percent with vacuum
distillation. Emissions, of course, do not precisely correspond to energy
consumption; nonetheless, the data strongly suggest that the catalytic
cracking process, which is important for increasing gasoline yield, is
the most significant energy consumer in the refinery.
As we have indicated in a previous study task, a more equitable
allocation of energy consumption might be based on the requirements for
single-purpose refineries. It is feasible with modern hydrocrackers to
obtain a 90 percent yield of high-octane gasolines; however, we know of
no refineries that operate on this basis, and consequently cannot estimate
the overall thermal efficiency of such an operation. If the entire
refining energy loss is allocated to gasoline production, then gasoline
refining is only 75 percent thermally efficient (at a product fraction
8-5
-------
of 42 percent and an overall efficiency of 0.89). However, it is not
reasonable to assume that the other refinery products are obtained free
of any energy losses. Therefore, we have assumed for comparative purposes
that it takes roughly 10 percent more energy to refine gasoline than
power-plant fuel oil. Consequently, at a-gasoline product fraction of
42 percent and an overall refinery efficiency of 0.89, the corresponding
product efficiencies in refining would be 0.84 for gasoline and 0.93
1-3
for power plant fuels. Other studies of the comparative efficiencies
of electric cars and conventional cars have assumed gasoline refining
efficiencies of 0.74 to 0.87.
In the comparisons of overall energy consumption rates between
electric and conventional cars that follow, only the relative refining
efficiency of 0.9 is used (gasoline refining efficiency divided by fuel
oil refining efficiency). Thus in absolute terms the estimates of
energy consumption for both cars are low by about 7 percent in accounting
for total refinery energy losses. Also, the energy consumed in extracting
the oil and transporting it to the refinery has not been included in our
comparisons, since these energy losses are common to the use of either
vehicle.
2.1.2 Energy Used in Transporting Fuels
For refined fuels, the energy used in transporting them to gasoline
jobbers or power plants may be appreciable if distances are long. In the
South Coast Air Basin, however, there are many refineries and hence distances
are short. It is assumed in this paper that both gasoline and power plant
14
fuel are transported by truck. A study by Hirst shows that average
inter-city freight hauling uses 2,340 Btu per ton-mile. In-city hauling
may have a slightly greater energy consumption rate. However, assuming
the inter-city energy consumption rate, a 50-mile trip (probably a reasonable
estimate for SCAB) would require 117,000 Btu to carry a tone of gasoline,
which contains about 42 million Btu. Thus such a short trip consumes less
than 0.3 percent of the energy that is delivered. Because of the proximity
8-6
-------
of refineries to both power plants and ultimate gasoline sellers in SCAB,
energy transport costs are small and the differences are smaller yet.
Thus this energy cost is assumed to be insignificant for comparative
purposes.
2.1.3 Energy Conversion Efficiencies
There have been several other investigations that compared electric
car energy consumption with that of conventional autos. Each analysis
proceeds from certain assumptions regarding the stepwise efficiencies in
the conversion of primary energy to the ultimate delivery of road miles
under some assumed driving scenario.
In one of these studies, a comparison under steady speeds showed
electrics to be 1.5 times more efficient in consumption of primary energy.
However, the battery recharge efficiency was not included and the study
assumed vehicles of roughly comparable weight, whereas the 1980 electric.
car of our study is substantially heavier than the compact car it is
assumed to replace.
2 3
The other two studies ' also had varying assumptions regarding
the efficiencies in each step of the energy flow and the scenarios for
electric car use, with conclusions that electrics could perform anywhere
between equal and twice the efficiency of conventional gasoline-powered
autos.
In this analysis we will rely on the baseline projections developed
in Ref. 6 and the electric car configurations described in Ref. 7 to guide
the calculation of comparative energy consumption rates for electric and
conventional cars. Significant conditions established for the baseline
projections are:
• Gasoline-powered vehicles, in response to economic and/or
legislative forces, will achieve improved gasoline mileage
(for the average new car) of 15.5 mpg, 22.0 mpg, and 25.0
8-7
-------
mpg by 1980, 1990, and 2000 respectively. This will yield
average mileages (for all cars in the population) of 13.75
mpg, 20.0 mpg, and 24.0 mpg respectively.
The present average gasoline mileage is taken as 12.5 mpg.
These gasoline mileages are appropriate in characterizing
all trip-making within SCAB.
In addition, we will assume for the parametric analysis of impacts
that the electric cars will replace 1C cars with the above stated average
gasoline mileages. This assumption will provide for accurate assessments
at high electric car usage levels. However, at low levels of electric
car usage, especially in the introductory period (early 1980s), electric
cars will likely be chosen as alternatives to small compact or sub-compact
cars which achieve gasoline mileages in the range of 20 to 30 mpg. Later
versions of electric cars with more advanced battery technology, may be
chosen as alternatives to a wider range of 1C cars than just sub-compacts
or compacts; especially so if conventional cars are to achieve the improved
gasoline mileages we have postulated. One approach will undoubtedly
rely on lowering the weights of the heavier, full sized 1C autos; thus
future electric cars will be closer in size and weight to a greater number
of future 1C cars.
In accord with these assumptions, energy consumption rates for
converting primary energy to road miles for conventional auto use in
SCAB are shown in Table 2.1.
Energy consumption rates for electric cars will depend on the thermal
efficiency of power-plant energy conversion, electric power transmission
losses, battery recharge and discharge efficiencies, and electric motor
and power train efficiencies. In Ref. 5 it was shown that the average
efficiency of oil-fired power, plants in the South Coast Air Basin is
presently about 36 percent (substantially higher than the 1972 national
8-8
-------
TABLE 2.1
AVERAGE ENERGY CONSUMPTION RATES ASSUMED FOR GASOLINE-POWERED
AUTOS
Present
1980
1990
2000
11,000
10,000
6,875
5,729
Btu/mi
Btu/mi
Btu/mi
Btu/mi
*
Includes 10 percent energy surcharge for refining gasoline relative to
power plant fuels, and gasoline energy content of 125,000 Btu/gal.
average of 32 percent) and is expected to remain at or slightly
above that level essentially through the period of interest for this
study. Also indicated in Ref. 5 is the very likely prospect that oil-
fired plants would be used most frequently throughout the period 1980 to
2000 to generate the necessary power during off-peak hours to recharge
the electric cars. However, by the year 2000 it may be necessary to
provide some of the off-peak power for electric car recharge by firing
out-of-basin coal-fired plants. This is expected to be a minor fraction
of the required recharge energy for significant levels of electric car
usage, and the coal-fired plants will not be significantly different in
conversion efficiency from the 36 percent assumed for in-basin plants.
Consequently, it will be assumed that conversion of fuel oils to electric.
energy for electric car recharge will be accomplished at 36 percent
efficiency or 9500 Btu/KWH.
Detailed planning data of the Southern California Edison Co. (SCE)
indicates that 36 percent thermal efficiency will be achieved out to the
year 1983. Addition of combined cycle capability with its higher effi-
ciency to existing generation plants within the SCAB may cause the
overall average efficiency of thermal generation to increase slightly.
8-9
-------
Beyond 1983 there may be some possibility to further upgrade the
system with technological improvements to combined cycle systems. After
1985 it is expected that combined cycle systems will achieve 48 percent
thermal efficiency, which is significantly better than the 40 percent
presently postulated for combined cycle systems. Thus an improvement of
this degree could cause the SCE in-basin generation efficiency to rise
significantly above the existing 36 percent level. Even further improvement
may be possible if the magneto-hydrodynamic (MHD) topping cycle technology
is successfully developed. MHD topped power plants may achieve efficiencies
as high as 55 to 60 percent, but the availability of this technology is
forecast for 1995 and beyond. Consequently, our assumption of 36 percent
efficiency throughout the forecast period must be regarded as conservative
in regard to estimating electric car performance.
Transmission of electric energy may incur significant losses,
depending on the distance of transmission as well as other transmission
operations such as transforming voltage levels. These losses are difficult
to allocate to specific end uses of electric energy; therefore we will
use the SCE average loss rate of 9 percent.
Battery discharge and recharge, as well as the conversion of electric
energy to road mileage in electric cars, is a strong function of the
detailed manner in which these operations are accomplished. In a companion
study task concerned with electric car technology, efficiencies for
each of these operations are analyzed for typical expected charging and
driving conditions. Based on a standard metropolitan driving cycle, near-
term four-passenger electric cars with lead-acid batteries are expected
to consume 0.79 KWH/mi, electric energy measured at the input to the
battery charger. Similarly, intermediate-technology cars with nickel-
zinc batteries are expected to consume 0.51 KWH/mi and advanced-technology
electric cars with zinc-chlorine or lithium-sulfur batteries are expected
to consume 0.41 and 0.45 KWH/mi, respectively. A smaller two-passenger
car in all cases will consume less energy; however, the four-passenger
8-10
-------
configurations are better able to provide the accommodations of the
conventional autos they most likely will replace. We would therefore
not expect the two-passenger configuration to achieve as wide an
acceptance among car buyers, and consequently we have chosen to base
our comparisons on the four-passenger configurations.
The overall energy consumption rates for electric cars when recharged
by in-basin thermal power plants are listed in Table 2.2 along with
"break-even" gasoline mileages for conventional gasoline powered cars.
TABLE 2.2
AVERAGE PRIMARY ENERGY CONSUMPTION RATES ASSUMED FOR ELECTRIC CARS
(Parentheses are break-even gasoline mileages including refinery energy)
Battery Technology 1980 1990 2000
Lead-Acid (16.7) 8,250 Btu/mi
Nickel-Zinc (25.8) 5,330 Btu/mi
Zinc-Chlorine (32.2) 4,280 Btu/mi 4,280 Btu/mi
Lithium-Sulfur (29.4) 4,710 Btu/mi
Assumes 9 percent electric power transmission loss and 9,500 Btu/KWH
conversion rate at the power plant.
2.1.4 Conversion of Non-Petroleum Energy Sources for Automotive Transportation
It was shown previously (in Task Report 5) that beginning sometime
around 1990 coal-fired plants outside the basin would be utilized to an
increasing degree to recharge electric cars during off-peak hours. Pre-
sently these plants are conventional Rankine cycle plants of relatively
recent construction. By the year 2000 it is expected that in addition
to coal and oil some nuclear power generation would be utilized during
8-11
-------
off-peak hours for recharging electric cars. Thus electric cars have
the potential, even in the Los Angeles region, to shift some transportation
to a non-petroleum source of primary energy consumption.
Much attention is being paid the possibilities of expanding our
utilization of coal in order to decrease our dependence on imported oil.
Although the baseline energy forecast for SCAB does not reflect a large
usage of coal, the present interest in western U.S. coal reserves may
presage an unexpected expansion in the use of this source for SCAB energy
demand, and electric cars may become an important element in that demand.
However, other approaches to using coal as a primary source of auto-
motive energy are also being investigated. One very promising method
involves the conversion of coal to synthetic crude ("Syncrude") from
which motor gasoline is refined. This process is well suited to the sub-
bituminous western coals, and furthermore the syncrude process involves
the removal of unwanted sulfur.
A comparison of two alternative approaches to the use of coal as
a primary source of automotive energy has been made. One alternative
considers the path: coal-to-syncrude-to-gasoline-to-internal combustion
powered car. The other considers the path: coal-to-low BTU gas-to-combined
cycle power plant-to-electric car. Both alternatives are clean processes
insofar as sulfur removal is concerned. The comparison is made on the
basis of overall thermal efficiency in converting coal energy to road
miles. Large scale production of syncrude from coal is expected to be
available beginning in 1982, or about the same time that combined cycle,
electric power plants utilizing low-BTU gas from coal are expected to be
commercially feasible. This latter combination is presently being researched
under the title of advanced power cycles (APC). Although the combined
cycle portion of APC research is concentrated on a staged combination of
Brayton and Rankine thermodynamic cycles, there is also the possibility
that, later on, the magneto-hydrodynamic (MHD) topping cycle may become
8-12
-------
availble to replace the first stage or Brayton cycle.
We have calculated the specific primary energy consumption—BTUs
of coal per road mile of transportation—for each alternative path and
the results are shown in Fig. 2.1.
The alternative converting coal to gasoline is depicted by three
boxes, labeled as to the assumed gasoline mileages that could be obtained
on the metropolitan driving cycle. The top of the box is the energy
consumption rate if the total syncrude output is refined to gasoline, a
process expected to be 0.65 efficient (coal-to-motor gasoline). If the
refining of coal syncrude is split between motor gasoline and distillate
fuels, overall efficiency is expected to be 0.70. The three boxes are
cast in a time sequence generally reflecting our basic assumptions concerning
improvement in future gasoline mileages obtained. However, there is no
need to constrain the comparison in this way, and the reader is free to
consider 30 mpg in 1982 if this is deemed a more appropriate comparison.
The coal-to-electric car energy consumption rates are somewhat more
time constrained, in that the efficiencies of APC systems are forecast
to increase as technology allows. The band depicting the expected range
for APC-based systems reflects alternative forecasts of APC efficiencies. '
We also show the effect of replacing the Brayton topping cycle of the APC
with MHD topping which is assumed to produce even higher efficiences.
The APC with MHD option is combined with each of three battery technologies
which convert electricity to road miles utilizing the electric car config-
urations already presented. (The Li-S battery technology, which lies
between the Ni-Zn and Zn-Cl in efficiency, was left out only to provide
clarity in the figure). The circled point in 1974 is for reference only,
showing the combination of a modern coal fired plant (Rankine cycle only
13
at a heat rate of 9,000 BTU/KWH) with a lead-acid battery, electric
car.
8-13
-------
oo
10
o
I-H
0.
o
o
Q.
O
Q_
00
0
MODERN COAL FIRED
POWER PLANT
Pb-ACID BATTERY
D
COAL/SYNCRUDE/GASOLINE/20 MPG
25 MPG
30 MPG
MHD TOPPING, LEAD-ACID
MHD TOPPING, Ni-Zn
MHD TOPPING, Zn-Cl
ADVANCED POWER CYCLE (APC) ESTIMATES: NPC (1972), EPA (1972)
_ MHD TOPPING ESTIMATES: NPC (1972)
COAL/SYNCRUDE/GASOLINE ESTIMATES: EPA (1974)16
I
I
1980
1990
2000
APC
MHD
= COAL/GAS/BRAYTON/RANKINE
TOPPING = COAL/MHD/RANKINE
Figure 2.1. Comparison of Alternative Uses of Coal for
Automotive Transportation
-------
The coal syncrude conversion to motor gasoline is attractive because
it promises to preserve the beneficial features and investments in our
present mode of transportation. However, this is accomplished at some
loss in overall thermal efficiency, primarily associated with the pro-
duction of syncrude from coal. This is to be contrasted with the expected
improvements in thermal electric generation efficiencies, which tends to
enhance the possibilities of electric cars as the comparison in Fig. 2.1
clearly shows, especially with expected future battery technologies.
There are other constraining factors which can affect this comparison
to a slight degree. First, no transport energy costs are included in the
calculations. In the case of the coal syncrude process, the estimates are
based on the syncrude plant 'located near the source of the coal. The
refinery could be co-located in which case the transport of gasoline to
the market is the major energy transportation cost. As previously men-
tioned, the estimates are based solely on conversion of western sub-
bituminous coals which, for a national source of motor gasoline, may
entail significant energy transportation costs. The APC alternative
depends on conversion of coal to low-BTU gas which is uneconomical to
transport over large distances. Thus the APC systems imply that electric
power plants must also be somewhat near the source of coal. However, it
is not known how well the coal gas conversion process is suited to other
sources of coal (e.g., eastern anthracite and bituminous coals). Thus
there may be inherently more flexibility in coal gas conversion to
avoid energy transportation costs.
8-15
-------
2.2 PRIMARY ENERGY CONSUMPTION
In the previous subsection the comparative energy consumption rates
of electric and conventional autos have been analyzed and reduced to
single values for energy consumption per mile representative of average
trip-making conditions for SCAB. An analysis of likely levels of total
future travel in SCAB shows that the combination of expected auto popu-
lations and average miles driven per year for each car will result in
the following totals for private automobile travel in SCAB:
1980 60.8 billion vehicle-miles
1990 72.0 billion vehicle-miles
2000 83.3 billion vehicle-miles
These are the values forecast without the introduction of electric cars.
However, it is anticipated that successful application of future battery
technology will lead to electric cars of sufficient range and performance
not to affect trip-making behavior. Consequently, it is assumed that
total vehicle-miles within SCAB will be the same with or without electric
cars. Only in the case of the initial introduction of electric cars with
near-term lead-acid technology is the range constraint likely to affect
trip-making behvaior. However, even in this case it will be assumed that,
although the lead-acid based electric cars will be driven fewer miles
per year than the average auto, they will be second cars and routine
adjustment of car-trip selections in multi-car households will compensate
for the lowered electric-car mileage.
Based on the forecast SCAB automobile travel and the energy consump-
tion rates for conventional autos (Table 2.1) and electric cars (Table
*
2.2), the change in overall primary energy consumption as a function c
electric car usage was calculated. Figure 2.2 presents the calculated
For electric cars, the energy consumption rates used are based on the
assumption that lead-acid technology is appropraite for 1980, nickel-
zinc for 1990, and an average of zinc-chlorine and lithium-sulfur for
2000.
8-16
-------
30
20
cc:
LU
CL
510
20
40
60
80
1990
2000
1980
100
150 x 10
12
100
50
ENERGY CONSUMPTION, Btu/mi
1980 1990 2000
ELECTRIC
CAR
GASOLINE
CAR
8250 5330 4495
10,000 6875 5720
I
20 40 60 80
ELECTRIC CAR USAGE, PERCENT OF TOTAL VEHICLE MILES
100
Figure 2.2.
Decrease in Energy Consumption for Private Auto Travel Due
to Electric Car Usage Relative to Average Gasoline Car
8-17
-------
changes, in terms of absolute energy and in terms of percentage of total
energy consumed in auto travel, as a function of the percentage of total
vehicle-miles in SCAB accounted for by electric cars. Curves are pre-
sented for each of the years 1980, 1990, and 2000 which show that, at
high levels of electric car usage, and even with significant improvements
in gasoline-auto efficiencies, electric cars can provide significant
savings in total energy used for auto travel when compared to the average
gasoline powered car.
The parametric results of Fig. 2.2 were used to compute the energy
impacts of several alternative scenarios, developed in Ref. 9, for
electric car sales and use in the South Coast Air Basin. The range of
possibilities is bounded on the upper end by a rate of electric car
sales, which was taken as 80 percent of the new car sales in each year
from 1978 through 2000. The lower bound is an estimate of the rate of
sales based on the likely competitive position of electric cars in
"free market" conditions. (Because the "free market" case results in such
low electric car usages and consequently trivial or uninteresting impacts,
an alternative rate at a constant 17 percent of market sales was postulated.)
In between these bounds are two scenarios which assume progressively
increasing market shares throughout the period of interest. The "high
rate" of sales begins at 17 percent of the market and progressively
increases to 75 percent by the year 2000. The "intermediate rate" is
one-half the high rate. The resulting electric car percentages in the
SCAB auto populations for each time period are listed in Table 2.3.
The impact of electric car sales and use on SCAB primary energy
consumption for auto travel has been calculated and the results are
presented in Fig. 2.3. Only the upper bound case for electric cars
has been presented and compared with the energy consumption based on
the assumed schedule for improved conventional automobile mileages.
The primary energy consumption rates for other scenarios were not depicted
as the differences between the baseline and upper bound case are not large.
8-18
-------
TABLE 2.3
PROJECTED SCAB ELECTRIC CAR POPULATIONS FOR SEVERAL POSTULATED CAR SALES
SCENARIOS
Percent Electric Cars in Auto Population
Scenario Rate of Sales 1980 1990 2000
Free Market 0.25 2.6 7.8
(Constant 17%) 4.8 15.5 17
Intermediate 2.4 15.5 30
High 4.8 31 60
Upper Bound 22.5 73 80
1.0 x 10I5[-
ZD
OVERALL ENERGY CONSUMPTION RATES,
Btu/mi
1980 1990 2000
CAR0™11 825° 533° 4495
^OLINE 10,000 6875 5720
(IMPROVED)
^OL1NE 11,000 11,000 11,000
(UNIMPROVED)
BASELINE (NO IMPROVEMENT IN
^,- GASOLINE MILEAGE OVER EXISTS
^ — 12.5mpg)
, fAlV^. -""*'
o.t
G 0.6
0.4
0.2
BASELINE (IMPROVED GASOLINE
MILEAGE)
WITH UPPER BOUND RATE OF
INTRODUCTION OF ELECTRIC CARS
1970
1980
YEAR
1990
2000
Figure 2.3. Primary Energy Consumption for Total Auto Travel
8-19
-------
It is significant to note that auto travel energy consumption without
electrics, and with no improvements in mileage, is projected to grow at
an annual rate of 1.6 percent per year as indicated by the dashed line
in Fig. 2.2. This is in contrast to the forecast made by Stanford
Research Institute for energy consumption in the transportation sector
of Southern California, which they projected to grow at 3.2 percent per
11
year.
The curve depicting the baseline condition with improved gasoline
mileage shows a large impact on automotive energy consumption from lowering
gasoline consumption. It may be argued that these are only goals to be
achieved and hence are quite likely to overestimate the degree of improve-
ment that may occur. Even so, the electric cars have the potential to
further improve somewhat on this situation. In this respect there is an
interesting observation that can be made concerning the conditions which
are likely to create an incentive to achieve improved gasoline mileages.
If in the indefinite future there is a continuing high-level concern
with the efficiency with which we consume energy in travel activities, the
introduction of a relatively efficient electric car in competition with
conventional autos may provide the necessary incentive. Conversely,
without efficient competing alternatives, the rate at which we approach
our goals is likely to be much slower.
On the other hand, if the prospects for future electric generating
plant efficiencies of 50 percent and higher are realized, then the
advantages of electric cars over even the postulated more efficient
year-2000 conventional autos will be further enhanced, and overall
energy consumption in transportation would be lower than that indicated
in Fig. 2.3 for a given mix of electric and conventional cars. This
association of electric car overall energy consumption rates with the
higher efficiencies achieved with combined cycle power plants (assumed
to use fuel oil) is correct, providing that future utility policy is to
provide base loads from nuclear and out-of-basin coal fired plants.
8-20
-------
It is conceivable that future economic conditions in combination with
significant changes in the hourly demand profiles may require a reordering
of power plant generation priorities, although we consider it unlikely
as long as some oil (the most costly fuel source for the indefinite future)
is required.
2.3 GASOLINE CONSUMPTION
The same basic assumptions used in the calculation of primary
energy consumption were used as the basis for calculating the impact
of electric cars on SCAB gasoline consumption. The results of these
calculations are presented in Fig. 2.4, which shows that for all but
the "free market rate" scenario, electric car use would have a significant
impact on gasoline sales in the basin. At the upper bound rate of electric
car sales, corresponding use in the year 2000 will require about one-fifth
the existing gasoline sales to fuel the remaining conventional autos.
5 x 10
ASO
BASELINE (IMPROVED
GASOLINE MILEAGE)
WITH FREE MARKET RATE
OF ELECTRIC CAR SALES
WITH INTERMEDIATE RATE
WITH HIGH RATE
WITH UPPER SOUND RATE
1970
1980
1990
2000
YEAR
Figure 2.4. Private Auto Gasoline Consumption
8-21
-------
2.4 POWER PLANT GENERATION
It is assumed that recharging of electric cars will routinely
make use of the available generating capacity during off-peak hours.
Off-peak recharging can virtually be guaranteed by designing special
features in home recharger equipment. For example, it is quite feasible
to design special sensor circuits in the recharger that can be activated
by modulating the power delivery with a unique signal. The modulated
signal would be sent from the generating station to a sector (or sectors)
of the service area whenever the total load approached the peak capacity
of the station. Receipt of the signal at the rechargers in a selected
sector would cause these rechargers to shut off.
For each electric car scenario, the daily vehicle-miles driven by
electric cars were calculated from annual totals, assuming equal distri-
bution over 365 days. Allowing for a 9 percent transmission loss, daily
vehicle-miles were converted to daily requirements for recharge energy
for both existing and advanced battery technologies (0.79 KWH/mi and
0.45 KWH/mi, respectively). Also, the available off-peak energy as a
function of peak demand was calculated for the years 1980, 1990, and 2000
based on the forecasts presented in Ref. 5 for both peak day and average
weekday conditions. Figure 2.5 shows the assumed hourly profiles of
electrical demand for two levels of SCAB electric car usage (shown as
*
a percentage) for each time period based on peak day conditions. The
shaded regions in the figures indicate the portions of total generation
that are used to recharge electric cars in the normal off-peak period.
The electrical energy generation available for electric car recharge
during the off-peak period as a function of the level of off-peak demand ,
The particular levels of electric car usage shown in Fig. 2.5 were chosen
earlier to bound the parametric study of air pollution undertaken in a
companion study task.
8-22
-------
30rx 10J
20
10
ELECTRIC CAR USAGE, PERCENT
1980 (LEAD-ACID BATTERY TECHNOLOGY)
:00
6:00
12:00
6:00
A.M.
P.M.
12:00
40
30
20
10
10J
ELECTRIC CAR USAGE, PERCENT
1990 (FUTURE BATTERY TECHNOLOGY)
:00
6:00
12:00
A.M. I P.M.
6:00
12:00
50 rx 10
I 30 J
2000 (FUTURE BATTERY TECHNOLOGY
10 -
:00
12:00
Figure 2.5. Electric Power Demand Profiles
8-23
-------
was calculated for each time period for both peak day and average week
day conditions. These amounts of recharge energy were then converted to
electric car usage, assuming appropriate car efficiencies for each time
period. The results of these calculations are depicted in Fig. 2.6,
where off-peak demand is shown on the ordinate in terms of the percent of
peak capacity for each of the three time periods and SCAB electric car
usage is presented as the percent of all SCAB vehicle miles driven by
electrics. Cross-plotted on Fig. 2.6 are the requirements to meet electric
car usage for each of three scenarios governing the rate of electric car
introduction to the general auto population (the low usage scenarios show
minor impact on off-peak demand and are not depicted in the figure).
It is expected that the off-peak demand shown for the years 1980,
1990, and 2000 would require the excess generation to occur typically
between 10 p.m. and 9 a.m., as shown in Fig. 2.5 (1980 off-peak demand
will probably be between 11 p.m. and 8 a.m.). Undoubtedly, some of the
recharging requirements will be met outside this time interval, with
some occurring during the peak hour. Assuming that 10 percent of the
recharge requirement might be uniformly spread over the complementary
period (9 a.m. to 10 p.m.), we estimate that peak demands will be increased
by less than 2 percent over what is otherwise forecast.
SCAB power plants need only operate at 84 percent of peak demand
during the off-peak period on peak days in the year 1990 to meet the
requirements for the advanced battery systems at the upper bound rate of
electric car population buildup. Lead-acid battery systems at these
high utilization rates in the years 1990 through 2000 would overtax the
off-peak capacity of the basin's power plants on peak days. The situation
for average weekday conditions is quite different. At the upper bound
rate of electric car population buildup, off-peak demand for electric
car recharge maximizes at 72.5 percent of the yearly peak demand. And
based on these average weekday conditions, substantial usage of electric
cars based on lead-acid batteries could be supported (72 percent usage
8-24
-------
si
s
90
80
70
60
50
40
1980, Pb-ACID
1990. Nl-Zn
UPPER BOUND RATE
2000, AV6.LJ.-S, I
r!990
F7T7T1980 BREAKPOINT
OIL
mifl
NO OIL
0 20 40 60 80 100
SCAB ELECTRICAL CAR USAGE, PERCENT OF TOTAL DAILY VEHICLE MILES
a) Peak Day Conditions
1980, Pb-ac1d
UPPER BOUND RATE
1990, N1-Zn
2000, AVG
W¥»/>»»> '980 BREAKPOINT, OIL/NON-OIL
0 20 40 60 80 100
SCAB ELECTRIC CAR USAGE, PERCENT OF TOTAL DAILY VEHICLE MILES
b) Average Weekday Conditions
figure 2.6. Off-Peak Power Generation for Different Battery Technologies
and Introduction Scenarios
8-25
-------
at 85 percent of yearly peak demand in 1980). However, such usage levels
could not be accommodated on peak-days.
We have also indicated (see curves, Fig. 2.6) the levels—in percent
of yearly peak demand—for each time period above which the primary source
of energy for electric generation would be fuel oil. Below these levels
coal, nuclear and hydroelectric sources would be utilized. In general
additional off-peak generation during peak day conditions would be met
by fuel oil for the 1980 and 1990 time periods, but by 2000 substantial
amounts of off-peak generation from non-oil sources could be utilized.
For average weekday conditions oil would be used exclusively for addi-
tional off-peak generation only during the 1980 time period. By 1990
a 22 percent electric car usage could be supported with non-oil sources
of electric generation, reaching almost a 90 percent usage level in 2000
before oil sources would be required.
2.5 PETROLEUM SAVINGS
Large scale usage of electric cars may produce significant reductions
in the amount of petroleum used as a source of primary energy. First
because they may be slightly more efficient than the average gasoline
car they can effect savings in petroleum use. Second, because power
plants under average weekday conditions in the future will be able to
utilize non-petroleum sources to make recharge energy available for
electric cars, much of the daily vehicle mileage displaced from gasoline
autos by electric cars will produce proportionate savings in petroleum
required. These two effects have been incorporated in a calculation of
the average petroleum saved as a function of electric car usage and the
results are presented in Fig. 2.7. The savings, shown as the ordinate,
are expressed in percent of the amount that would have been otherwise
used for the baseline forecast with all gasoline powered cars and with
the improved schedule of average gasoline mileages. Electric car recharge
requirements and the supply of electric recharge energy are based on the
generation capabilities for average weekday conditions as shown in Fig. 2.6.
8-26
-------
UJ
d
Zn-Cl, 2000
X
x/Li-S, 2000
Ni-Zn, 2000
Zn-Cl, 1990
- Li-S, 1990
| |_J ^
' Ni-Zn, 1990
20 40 6C
PERCENT ELECTRIC CAR USAGE
80
100
Figure 2.7. Petroleum Savings As a Function of Electric Car Usage (SCAB)
However, each battery technology has been kept separate in these calcula-
tions as indicated in the figure.
In general, for a given year and battery technology, savings are
directly proportional to usage until a point is reached where in the
lowest demand period in the year oil must be used during off-peak to
recharge electric cars. At higher average usage rates there are a
greater fraction of days on which off-peak demands must be met by oil
sources. Finally with still higher usage levels the peak day condition
is reached where all additional off-peak demands draw from oil sources
that would have been used to power the gasoline cars displaced by electrics.
At this point the savings are proportional to the differences in overall
conversion efficiencies (Btu/mi), which for the case of lead-acid electric
cars in 1990 and beyond favor gasoline cars (down sloping curve), and
in the cases of electric cars with future battery technologies favor
electrics over gasoline cars. In every case the curves represent an
8-27
-------
upper bound which are quite close to curves that would be developed from
actual distribution functions of hourly demands instead of representative
low, average, and peak demand characterizations. Incorporation of an
actual distribution function would produce a smooth curve, lying slightly
under the bounding line segments, but tangent to the first and third
segments.
Figure 2.7 clearly shows that for 1990 and 2000 significant savings
in petroleum are potentially possible with conversion to electric cars
with future battery technologies. Only small savings are possible with
lead-acid based cars in 1980 and in 1990 at low usage rates.
8-28
-------
3 RESOURCE IMPACTS
The most significant demands on resources associated with the pro-
8 10
duction of electric cars have already been identified. ' In general
the materials required for several of the proposed battery systems pose
the most serious resource impacts. These include lead, antimony, nickel,
zinc, graphite, titanium, and lithium. Also, motors and other electrical
gear may place significant demands on copper supplies.
The impacts of SCAB electric car requirements for these materials
on traditional and forecast supply and demand relationships for the entire
United States have been quantified as a function of the level of electric
car use in terms of the percent of total SCAB vehicle-miles driven by
electrics. The inventories of materials for each battery technology and
alternative electric-car configuration are presented in Ref. 7 and form
the bases for the impact assessments that follow. However, consistent
with the energy impact assessments presented in the previous section,
only the four-passenger configurations in Ref. 7 have been included for
study. Basic information on supply and demand relationships has been
18
drawn largely from the 1970 edition of the Bureau of Mines compilations
of mineral resource data.
The annual material requirement for a given battery component
(plate, terminal, electrolyte, etc.) or other unique electric car compo-
nent depends on several factors. First, each new car sold must be sup-
plied with the component. Second, depending on the life of the compo-
nent and the past history of car sales, components in older cars may be
frequently scrapped; total component sales must take this factor into
account.
For materials that are scarce, there will be strong incentives to
salvage them. The percentage of scrap recycled as secondary materials
production depends strongly on price incentives, reflecting primarily the
scarcity. At present, a high percentage of battery lead is recycled.
8-29
-------
19
1969 figures published by the Battery Council International (BCI) show
that 520,913 tons of battery lead were reclaimed while demand for battery
lead was 582,546 tons. Thus lead reclaimed was approximately 90 percent
of lead demanded. If these values were to represent steady-state condi-
tions it would be reasonable to assume a 90 percent effective recycle
reate. However, a BCI spokesman indicated that 85 percent was probably
a more representative value. No other materials, with the exception of
*
antimony, have demonstrated the rate of recycle that lead has. However,
the employment of heretofore new materials in battery systems on a wide
scale may provide the basis for the development of similar recycle
industries as well as stimulate existing recycle operations.
For example, there is the prospect, with relatively high levels of
electric car usage, for development of retail battery service centers,
where electric car batteries are exchanged routinely. The fact that an
electric car battery has a salvage value of say one or two hundred dol-
lars, rather than the two or three dollars for today's accessory battery,
means that few electric car batteries will ever be lost in the shuffle.
Also such battery centers will ease the collection problem by providing
conspicuous collection and exchange points, thereby aiding in the collec-
tion of all other storage batteries. Furthermore, because electric car
batteries are expensive but necessarily replaceable, there may be incen-
tive towards designing standardized as well as reusable battery cases.
At present, the cases from scrapped batteries pose significant solid
waste problems, especially since environmental sanctions prohibit burn-
ing them in ma'ny places. Used electrolyte from scrapped batteries also
poses a waste disposal problem, but it is difficult to speculate what
opportunities may be afforded for mitigating this problem.
The requirements for scarce or critical materials to support vari-
ous levels of electric car usage in SCAB have been related parametrically
*
Most of the recycled antimony is in the form of antimonial lead, used
primarily in storage batteries.
8-30
-------
to the amount of material supplied from recycle, and to average battery
life. The material requirement is calculated in terms of the annual
amount of new material required. Recycled material is expressed as the
annual amount (in percent) of the total material requirement supplied
from reclaimed sources. This percentage should not be confused with a
strict definition of recycle efficiency, which defines the fraction re-
claimed of that potentially available. For example, in cases where elec-
tric car populations are rapidly building after their initial introduc-
tion, a recycle efficiency of say 85 percent may provide only 60 percent
of the total material requirement. Consequently, simple parametric
representations of resource impacts may tend to underestimate the signi-
ficance of the effect on resources due to the rate of buildup of the elec-
tric car population. Figures 3.1 and 3.2 illustrate these effects by
showing the demand for new batteries, and the available recycled batteries,
for the range of scenarios governing the sales and use of electric cars
within SCAB developed in Ref. 9 and previously summarized in Sec. 2.2.
Figures 3.1 and 3.2 depict, as examples only, the cases for two-year
battery life and an 85 percent recycle efficiency (defined as the per-
centage recycled of those potentially available for recycle).
In general, the results of the parametric impact analysis on materials
requirements include two ranges of battery life, two years and three years,
and three levels of material supplied from reclaimed sources, 85, 90, and.
95 percent. The choice of battery lifetimes of two and three years are
based primarily on experience and expectations with lead-acid batteries.
Consequently, these choices may not be representative of future batteries
where lifetimes of five or more years may be feasible with, for example,
nickel-zinc batteries. In cases where future technology may produce
longer battery lifetimes, the parametric results that follow will under-
estimate the amount of materials supplied from reclaimed sources and over-
estimate the requirement for new materials. In a like manner, the range
in the fraction of material supplied from reclaimed sources may not ade-
quately span the possibilities with some of the future batteries. However,
8-31
-------
TWO-YEAR BATTERY LIFE
Figure 3.1. Annual Battery Sales in SCAB
3.Or .< 10°
THO-YEAR BATTERY LIFE
85',. RECYCLE EFFICIENCY
INTERMEDIATE
RATE
1980
2000
Figure 3.2. Annual Battery Recycle in SCAB
8-32
-------
it is extremely difficult to predict such possibilities. The low end
of the range of recycle, 85 percent, is representative of present
experience with lead-acid batteries. It is expected that the dollar
value of a single electric car battery system should cause a signifi-
cant improvement in this situation which we allow for by the higher recycle
levels of 90 percent and 95 percent. However, in some cases where a
single battery may contain raw materials approaching $1,000 in value,
such as may be the case with nickel-zinc, it is difficult to imagine
any significant losses from the used battery inventory. In such cases
the parametric results would produce conservative estimates of the new
material required.
To provide further insight into the effect of the rate of buildup
of electric car population on resource requirements, additional curves
for each mineral are presented, showing the electric car requirement
for new materials as a function of time over the period 1980 to 2000
9
for each of several selected scenarios.
In the analyses of resource impacts that follow, consideration is
given primarily to existing lead-acid battery technology and to foresee-
able future battery technologies (nickel-zinc, zinc-chlorine, and lithium-
sulfur cells). Provision of lead-acid batteries for electric cars on
any significant scale for SCAB may pose problems in the supply of lead
and antimony. Large-scale use of nickel-zinc batteries within SCAB may
cause significant perturbations in the national supplies of these two
metals; other incidental battery materials will be of minor consequence.
Zinc-chlorine batteries may also perturb zinc supplies, but to a lesser
degree than nickel-zinc batteries at the same scale of employment. Zinc-
chlorine batteries also require potentially significant amounts of
titanium, but there should be no difficulty in providing chlorine.
Lithium-sulfur batteries, when provided for large numbers of electric
cars, will impact supplies of lithium and graphite most significantly,
while sulfur supplies should be virtually unperturbed.
8-33
-------
3.1 SCAB ELECTRIC CAR RESOURCE IMPACTS
3.1.1 Lead
Figures 3.3a and 3.3b show the annual requirements for primary
production (new material) of lead as a function of electric car usage
in SCAB. Three sets of curves are shown representing percentages of 85,
90, and 95 percent of material supplied from reclaimed resources. Each
set is further broken down to representations for each time period of
interest, 1980, 1990, and 2000. Also shown is the Bureau of Mines (BOM)
18
forecast for US primary lead demand to the year 2000, and the BOM
estimate of 354,000 tons for the 1968 US production of primary lead from
*
domestic mine sources. Total US primary lead demand in 1968 was 898,000
tons, the difference being made up largely of imports. Because a high
level of battery recycling takes place, battery demand is not the most
significant item in total primary demand. Thus Fig. 3.3 shows that
electric car usages greater than 20 percent or 30 percent in the SCAB
begin to require appreciably greater proportions of primary lead for
batteries than in the past, which may significantly perturb forecasts of
primary lead demand. However, higher utilizations of electric cars are
not generally consistent with the constraints (primarily limited range)
imposed by lead-acid battery technology. Consequently, a reasonable
level of utilization of electric cars in SCAB should not unduly disturb
future lead supply and demand relationships. Nonetheless, since total
US lead supply depends heavily on foreign sources, there may be some
concern regarding the impact of any additional demands.
To demonstrate the effect of the rate of buildup of the electric
car population, Fig. 3.4 depicts the annual requirement for primary
lead over time for each of three scenarios governing the introduction
of electric cars. The case labeled "upper bound rate of sales" places
*
Mineral production is variously quoted in short, long, and metric tons.
In this paper, short tons (2000 pounds) are used throughout.
8-34
-------
1.2 rx 10
1.0
0.8
3 0.6
0.4
0.2
LEGEND:
ST. JOE MINERALS
^7777777777. '
US BOM
SSSfSrtSSSSf. ,
FORECAST US
1980 PRODUCTION •„,„„//,////////.
1990
1980
DOMESTIC PRIMARY PRODUCTION,
US BUREAU OF MINES FORECAST
RANGE, YEAR 2000
7777777! 1968 PRODUCTION
1.2 x 10V
1.0
'0.8
0.4
0.2
SUPPLY FROM
RECLAIMED SOURCES
85?
20 40 60 80
IC CAR USAGE IN SCAB, PERCENT
(a) Two-Year Battery Life
ST. JOE MINERALS
DOMESTIC PRIMARY PRODUCTION,
US BOM FORECAST RANGE,
YEAR 2000
US BOM
FORECAST US
1980 PRODUCTION
71968 PRODUCTION
SUPPLIED FROM
RECLAIMED SOURCES
20 40 60 80
ELECTRIC CAR USAGE IN SCAB, PERCENT
(b) Three-Year Battery Life
Figure 3.3. Lead Requirement as Function of Usage in SCAB
8-35
-------
1.2 x 10C
1.0
0.6
0.4
0.2
2 yr BATTERY LIFE
3 yr BATTERY LIFE
85% RECYCLE EFFICIENCY
FORECAST OF US
PRIMARY PRODUCTION
BUREAU OF MINES
SCAB ELECTRIC CAR
REQUIREMENTS
17% OF MARKET
1970 1980 1990 2000
YEAR
Figure 3.4. Lead Requirements as Function of Time
significant demands on primary lead supplies from the start until about
1985, when the electric car inventory is sufficiently large that reclaimed
lead becomes a significant source of supply. The "high rate" case builds
electric car inventories more slowly, and consequently primary lead
demand builds more slowly. However, by the year 2000 the "high rate"
case begins to surpass in primary lead requirement the "maximum rate"
case, even though its electric car population is less.
Another factor in the lead supply and demand relationship deserves
mention. The single most important consumptive use of lead is for anti-
knock gasoline additives. The BOM forecast, based essentially on 1968
and earlier data, has allowed for a significant expansion of this use
20
for lead; however, more recent appraisals, consistent with current
environmental concerns, project a significant decline in this use of lead.
It is of interest to examine what effects the combination of electric
8-36
-------
cars with the elimination of lead additives might have on forecasts of
lead supply and demand. Table 3.1 has been prepared to summarize lead
demand forecasts with and without electric cars.
The first part of Table 3.1 shows the expectations without electric
cars. The data for the years 1968 and 2000 are from the BOM, while 1974
20
and 1980 data are taken from analyses by St. Joe Minerals. The year-
2000 forecast of 3,650,000 tons total demand assumes 900,000 tons used
for lead additives. We assume that the alternative zero level for lead
additives in 2000 reflects an increasing environmental concern. Recent
data reported by the Battery Council International (BCI) show lead demands
for additives down in 1971 from the levels in 1969 and 1970. Furthermore,
as shown in Table 3.1, St. Joe Minerals expects lead demand for additives
to be down by 68 percent in 1980, consistent with existing EPA regulations
to reduce lead in gasoline from existing levels of 2.2 g/gal to 0.5 g/gal
by 1979.
The second half of Table 3.1 shows the effect of SCAB electric car
requirements on lead demand forecasts, for both storage battery and total
demands. We assume that the alternative low forecasts for lead additives
are appropriate in estimating total demand. Comparison of the two cases—
with and without electric cars—shows that the assumptions governing lead
additives are crucial to the overall forecasts of lead demand. The
year-2000 demand without electrics but with expanded use of lead additives
is within 15 percent of total demand with electrics in the SCAB and no
additives for the maximum rate of electric car sales.
Consideration of the fact that battery lead is recyclable, while
gasoline lead is not, makes this comparison more significant. Table 3.2
summarizes the effect on lead supply for the two corresponding cases,
with and without electric cars. The calculations assume that 85 percent
of storage battery lead will be recycled as secondary production,
leaving all other lead uses to be made up by primary production, imports,
8-37
-------
TABLE 3.1
RELATIONSHIP OF SCAB ELECTRIC CAR REQUIREMENTS TO US LEAD DEMAND
Lead Demand, Thousands of Tons
* ** ** *
1968 1974 1980 2000
No Electrics
Total US Demand 1449 1515 « J2740
Storage Batteries 500 750 930 1750
(on ( o
Lead Additives to Gasoline 262 260 j '" < "
Other 690 505 445 990
With Electrics in SCAB
Storage Battery Demand (2-year
life)
Upper Bound Rate of
Sales 1349 3147
High Rate 1020 2878
17% of Market 1020 2053
Total Demand
Upper Bound Rate of
Sales 1884 4137
High Rate 1575 3868
17% of Market 1555 3043
Bureau of Mines, base forecast. Range of total demand in year 2000 is
2,520-to-4,140-thousand tons.
**
St. Joe Minerals Corp.
With electrics, it is assumed that low projections of lead additives
are appropriate.
8-38
-------
TABLE 3.2
RELATIONSHIP OF SCAB ELECTRIC CAR REQUIREMENTS TO US LEAD SUPPLY
Lead Supply, Thousands of Tons
***** *
1968 1974 1980 2000
No Electrics
Total US Supply 1636 1670 1620 3640
Primary Production (Including ,„, 7in 7,-n I 520
US Refining of Imported Ores) \1120
Imports 338 150 120
Stockpile 260 185 50_
Subtotal 1085 1045 920 (2140)f
Secondary Production 551 625 700 (1500)
With Electrics in SCAB (85% Recycling.
2-year Life for Batteries)
Secondary Production
Upper Bound Rate of
Sales 856 2525
High Rate 733 2290
17% of Market 733 1717
Remaining to be Supplied by
Primary Production, Imports, and
Stockpile (to be Compared with
Subtotals Above)
Upper Bound Rate 1028 1612
High Rate 842 1578
17% of Market 822 1326
*
Bureau of Mines, base forecast.
**
St. Joe Minerals Corp.
Assumes 85% of storage battery demand for secondary production, leaving
2140 to be supplied from primary production, imports, and stockpile.
Assumes total demand forecast with electrics, Table 3.1.
8-39
-------
and stockpile. Comparison of required primary lead supplies (including
imports) with and without electric cars shows that requirements for the
introduction of lead-acid batteries on a large scale in the SCAB are
within very reasonable bounds and generally well below the BOM year-2000
forecast. However, this situation further underscores the importance
of the assumptions governing the lead additive forecasts in the BOM
estimates.
3.1.2 Antimony
Present lead-acid batteries have about 2.5 percent of antimony
alloyed with the lead to improve battery performance. Storage batteries
accounted for 48 percent of total US antimony demand in 1968; conse-
quently antimony, like lead, has a high recycle rate. Figure 3.5 shows
the SCAB electric car requirements for primary antimony as a function
of electric car usage. Figure 3.6 presents electric car requirements
for primary antimony as a function of time over the period 1980 to 2000
for several electric car sales scenarios. Also shown for reference is
the US 1968 demand and domestic production for primary antimony, which
indicates that under present conditions the US is heavily dependent on
foreign sources of antimony.
Except for this greater reliance on foreign sources, the expecta-
tions for future antimony demands parallel closely those for lead.
Antimony reserves in the US are small, estimated in 1968 by the BOM
to be only 110,000 short tons. World reserves appear adequate for
electric car use in SCAB, but half of them are in China. The greatest
US stock of antimony is contained in antimonial lead in use, which is
highly recyclable. This stock of antimony was estimated in 1968 to be
181,000 tons, which at a high recycle rate constitutes the most signi-
ficant US reserve.
A technological change is taking place which may ultimately elimi-
nate lead-antimony alloys in battery grids. Other, more plentiful
8-40
-------
8 x 10° r-
U)
c
o
YEAR 2000
FORECAST
OF DOMESTIC
PRODUCTION
_1968 US
7777777?
PRODUCTION
20
40 60
ELECTRIC CAR USAGE, PERCENT
80
8 x 10
3
(a) Two-Year Battery Life
c
o
OL
Q.
BOM YEAR 2000
FORECAST OF
DOMESTIC PRODUCTION
80
85% SUPPLIED
FROM RECLAIMED
SOURCES
85%
90%
100
40 60
ELECTRIC CAR USAGE, PERCENT
(b) Three-Year Battery Life
Figure 3.5. Antimony Requirement as Function of Usage
8-41
-------
35 x 10V
i
t~
20
15
10
1968 US PRIMARY DEMAND
2 yr BATTERY LIFE
3 yr BATTERY LIFE
85% RECYCLE EFFICIENCY
UPPER BOUND RATE OF SALES IN SCAB
1968 US PRODUCTION
(FORECAST DOMESTIC
IPRIHARY PRODUCTION, BOM
=•172! OF MARKET
1970
1980
YEAR
1990
2000
Figure 3.6. Antimony Requirements as Function of Time
materials, such as calcium, may be used in place of antimony to provide
maintenance-free batteries. While promising to alleviate US dependence
on foreign sources for antimony, the new battery requirements may pose
short-term technological and eocnomic problems for the lead recycling
industry.
3.1.3 Nickel
Nickel-zinc batteries may provide the earliest available replace-
ment for lead-acid batteries, leading to significant improvements in
overall electric car performance. The nickel-zinc battery envisioned
for a four-passenger electric car will weigh just over 1,000 pounds
(total battery system weight), of which 362 pounds are nickel metal and
328 pounds are zinc oxide (representing 263 pounds of zinc metal). For
this battery configuration it appears that nickel supplies will be the
most critical. The annual requirement for primary nickel has been
calculated as a function of the percent of electric car usage in SCAB,
8-42
-------
and the results are presented in Fig. 3.7. Three sets of curves are
shown, representing three different levels of the amount of nickel that
may be supplied by recycling old batteries. For each level, the curve
set breaks out the relationship for the years 1980, 1990, and 2000.
Also shown are the 1968 levels of US primary production and demand, which
indicate that we rely on external sources almost completely. The level
of demand in the year 2000 forecast by the Bureau of Mines (BOM) is also
shown for reference. As Fig. 3.7 indicates, the forecast of US nickel
demand will allow for high levels of electric car usage in SCAB, especially
at high recycle rates. However, high levels of electric car use nation-
wide (see Sec. 3.3) would very likely cause the BOM forecasts to be
exceeded, even for high recycle rates.
Figure 3.8 presents the requirement to supply primary nickel as
a function of the rate of buildup in electric car populations over the
period of interest, assuming of course that the nickel-zinc batteries
could be commerically available in 1980. The annual requirement is
shown for three scenarios governing the rate of sales, assuming 85 per-
cent recycle efficiency of the battery nickel and two values for average
battery life. Also shown is the BOM forecast for primary demand and
production. With the possible exception of the curve representing the
"upper bound rate of sales," the figure shows that the buildup of
electric car populations within SCAB can be achieved without an undue
perturbation of expected future demands for nickel in the US.
Metallic nickel is relatively expensive, but world supplies are
deemed sufficient to provide for the worldwide forecast demands without
requiring too much of a price increase. Despite its high price, less
than 20 percent of total nickel supply is from secondary sources. Use
of scrap metal is most easily facilitated when the intended second use
is the same as first. At present this causes difficulty in using old
scrap, because it must be carefully sorted. In the case of nickel-zinc
batteries, where the nickel is not alloyed and can be promptly reclaimed,
8-43
-------
0.6 x 106i-
0.5
0.4
c
o
a:
Q-
0.3
0.2
0.1
TWO-YEAR BATTERY LIFE
NI-ZN BATTERY TECHNOLOGY
BOM
YEAR 2000
FORECAST OF US
PRIMARY DEMAND
YEAR 1980
FORECAST OF US PRIMARY DEMAND
LEGEND:
?SS8i-
1980
1968 US DEMAND
SUPPLIED BY
RECLAIMED SOURCES
85%
1968 US PRODUCTION
20 40 60 80
ELECTRIC CAR USAGE IN SCAB, PERCENT
Figure 3.7. Nickel Requirement as Function of Usage In SCAB
8-44
-------
0.6 x 106
0.5
0.4
0.2
0.1
2 yr BATTERY LITE
3 yr BATTERY LIFE
UPPER BOUND RATE OF SALES IN SCAB
FORECAST OF US
PRIMARY PRODUCTION
MARKET
1970
1980
1990
2000
YEAR
Figure 3.8. Nickel Requirement as Function of Time
high recycle efficiencies should be economically attainable.
3.1.4 Zinc
Two battery technologies requiring zinc are considered, nickel-
zinc and zinc-chlorine. The former will require about four times as
much zinc as the latter. Calculations of the requirement for primary
zinc for 100 percent electric car usage in SCAB show that the require-
ment will probably cause only a very small effect on the forecast supply
and demand relationship for the United States. The results of these
calculations for both cell technologies are summarized in Table 3.3,
along with the Bureau of Mines forecast for US domestic primary demand
and production. The electric car requirements shown in the table are
for 100 percent electric car usage within SCAB at two levels of supply
from reclaimed sources, 85 percent and 95 percent. The SCAB estimates
have also been scaled to estimates for 100 percent electric car usage
within the US.
8-45
-------
TABLE 3.3
ANNUAL REQUIREMENT FOR NEW ZINC TO SUPPORT SCAB ELECTRIC CAR USAGE
(100 percent Electric Car Usage in SCAB )
Zinc Demand, Thousands of Tons
Nickel-Zinc Batteries, 2 yr Life
85% Supplied from
Reclaimed Sources
95% Supplied from
Reclaimed Sources
1980
58
(814)
19
(271)
2000
75
(1050)
25
(350)
Zinc-Chlorine Batteries, 2 yr Life
85% Supplied from
Reclaimed Sources
95% Supplied from
Reclaimed Sources
14
(196)
4.7
(66)
18
(255)
6.1
(85)
US Zinc Availability
Primary Demand
Primary Production
1968
1406
529
1980
/1600
\2300
) 600
I 870
2000
/2040
14000
( 786
(1500
Total US in parentheses.
8-46
-------
Comparison of SCAB requirements for primary zinc under any of the
conditions represented in Table 3.3 with US primary demand for zinc
indicates that there should be little effect on the BOM forecasts. Even
zinc requirements for a 100 percent electric car usage on a national
level may be accommodated, especially with the lesser demands inherent
in the zinc-chlorine cell.
Under 1968 conditions, reclaimed zinc provided for only 15 percent
of the total demand. This is partly due to the fact that significant
portions of the zinc supply are provided for end uses in which the zinc
is not easily recoverable, e.g., zinc galvanizing. Battery zinc, however,
should prove to be highly recyclable, given sufficient economic incentives.
Since the BOM expects zinc prices to double by the year 2000 (in 1968
dollars) in order to bring supplies into equilibrium with demands, the
incentive to recycle zinc should improve.
Presently the US depends to a significant degree on foreign sources
of zinc, as implied by Table 3.3, and the BOM expects this situation to
continue to the same degree. Zinc and lead are generally co-products of
the same extraction process. Consequently future expectations for the
supplies and demands of these materials must take into account this inter-
dependence and the possibility of large-scale use of electric cars employ-
ing first lead-acid batteries with a later shift to one of the zinc-based
batteries.
3.1.5 Titanium
For the zinc-chlorine battery configuration adopted in Ref. 7,
approximately 50 pounds of titanium per battery is necessary for electrode
and heat exchanger construction for the four-passenger electric car
configuration. Based on this requirement, the annual quantity of primary
titanium metal to support various levels of electric car usage in SCAB
was calculated. The results of these calculations are summarized in
Table 3.4 for the conditions of 100 percent electric car usage for each
8-47
-------
TABLE 3.4
ANNUAL REQUIREMENT FOR NEW TITANIUM TO SUPPORT ELECTRIC CAR USAGE IN SCAB
(100 percent Electric Car Usage in SCAB)
Titanium Demand, Thousands of Tons
1980
1990
2000
Two-Year Battery Life
85% Recycle
90% Recycle
95% Recycle
11.0 12.8
7.3 8.5
3.7 4.3
14.3
9.5
4.8
US Titanium Supply and Demand,
Thousands of Tons
1968
2000
Titanium Primary Demand as Compounds
As Metal
Titanium Primary Production as
Compounds
As Metal
440
13
305
0
960-2160
62-234
670-1610
0
of the years 1980, 1990, and 2000, for three assumed levels of supply
from reclaimed sources, and for a two-year battery life. Also shown
are the 1968 estimates of US total and primary demands for both titanium
metal and compounds, as well as the year-2000 forecast for these same
demand categories.
A comparison of electric car requirements with 1968 titanium
demands shows that the production of metal would have to be increased
significantly. However, existing demand for metal is such a small
fraction of total titanium demand that overall titanium supplies should
be little perturbed. The forecast shows that titanium metal demand
8-48
-------
should expand faster than total demand, owing primarily to its importance
in many advanced-technology systems. Should such an expansion of titanium
metal demand occur, other advanced systems employing titanium, such as
the zinc-chlorine cell, should benefit from the concomitant expansion
in the economic supply of the metal.
Titanium "sponge metal" was priced at $1.32 per pound in 1968,
and the BOM forecast expects that this price should remain more or less
constant to the year 2000. However, the electrolytic processing of
titanium compounds to yield sponge metal uses from 6 to 15.5 KWH per
pound, plus another 2 to 2.5 KWH per pound to produce metal ingot from
the sponge. Expected higher costs for energy may cause the finished
metal prices to rise womewhat above the BOM forecasts. In any case,
titanium is expensive enough that there should be ample incentive for
recycling.
Table 3.4 also shows that the US produces virtually no titanium
metal in this country, relying solely on foreign sources. This situation
arises because of the suitability of different ores for processing into
metal. According to the BOM, the most suitable ores are in foreign
countries, although there are some deposits here in the US.
3.1.6 Lithium
Improvements in electric car performance and range depend on the
development of advanced battery systems. Lithium-sulfur batteries have
been proposed as a promising candidate. The four-passenger electric car
configuration described in Ref. 7 utilizes a battery system containing
16.7 pounds of lithium, 68.1 pounds of sulfur, and 22.8 pounds of porous
graphite along with small amounts of other commonly available materials
(steel, aluminum, etc.). SCAB electric car requirements for lithium as
a function of electric car usage are presented in Fig. 3.9, under the
assumption that lithium in batteries will be recycled.
8-49
-------
6 x
o:
Q.
8 x 10-
s_
>>
c
o
<:
LU
QC
«£
20 40 60 80
ELECTRIC CAR USAGE IN SCAB, PERCENT
(a) Two-Year Battery Life
•1968 US PRODUCTION
20 40 60 80
ELECTRIC CAR USAGE IN SCAB, PERCENT
85% SUPPLIED
FROM RECLAIMED
SOURCES
90%
95%
100
LEGEND:
^,2000
1980
85% SUPPLIED
FROM RECLAIMED
SOURCES
90%
95%
100
(b) Three-Year Battery Life
Figure 3.9. Lithium Requirement as Function of Usage in SCAB
8-50
-------
The bulk of lithium used today is consumed in lithium compounds.
As a result lithium (as well as graphite) is virtually unrecycled under
present conditions. Consequently, we show for lithium in Fig. 3.10
the annual amounts required as a function of electric car usage under
the assumption that no material is supplied from reclaimed sources.
However, since the lithium used in electric cars is metallic lithium,
which is very expensive, we would expect its use to initiate a vigorous
recycle effort in metallic lithium, in which case Fig. 3.9 is more
appropriate.
Lithium may become an even more important material beyond 2000 if
the development of fusion power is successful. The less abundant lithium-6
(Li ) isotope will probably be a central material for a fusion process.
In that case, the mineral production and processing to retrieve the Li
isotope would have the very abundant Li isotope as a byproduct, which
would undoubtedly enhance the economic supply of lithium for other uses.
Figure 3.11 presents the growth of SCAB electric car requirements
for lithium over the time period 1980 to 2000 for several selected
scenarios. Two cases in which no recycle is assumed, and one case with
recycle, are shown and compared with forecast US lithium production. It
is noteworthy that, even with recycle, Fig. 3.11 shows that either an
"upper bound" or "high" introductory rate of electric car sales in SCAB
will have a significant impact on the forecast of lithium production.
2.1.7 Graphite
SCAB electric car requirements for graphite (used as electrode
material) have been calculated as a function of electric car usage and
are presented in Fig. 3.12. Recycling is not considered, since graphite
is presently not recycled and its cost is not high. However, SCAB
requirements are significant when compared to the forecast of total US
demand. Furthermore, the US produced only about 5 percent of its 1968
demand, relying on foreign sources for the remainder. There are significant
8-51
-------
3
40 x 10
30
20
g 10
2000
1990
1980
' 2 yr BATTERY
LIFE
3 yr BATTERY
LIFE
YEAR 2000
FORECAST
, j PRODUCTION, US
I
20 40 60 80
ELECTRIC CAR USAGE IN SCAB, PERCENT
100
Figure 3.10. Lithium Requirement as Function of Usage (No Recycle)
30 x 103
22.5
15
7.5
.2 yr BATTERY LIFE
3 yr BATTERY LIFE
1968 US PRODUCTION
W-42707
UPPER BOUND RATE
OF SALES, NO RECYCLE
HIGH RATE, NO RECYCLE
FORECAST OF US
PRODUCTION
UPPER BOUND
_ f RATE, 85% RECYCLE
1980
YEAR
1990
2000
Figure 3.11. Lithium Requirement as Function of Time
8-52
-------
160 x icr
120
80
40
YEAR 2000
FORECAST
DEMAND, US
;1968 US DEMAND
2 yr BATTERY
LIFE
3 yr BATTERY
LIFE
20 40 60 80
ELECTRIC CAR USAGE IN SCAB, PERCENT
Figure 3.12. Graphite Requirement as Function of Usage in SCAB (No Recycle)
graphite reserves in the US, but the particular end uses of graphite
require physical characteristics more readily attainable from foreign
sources, a situation which is likely to persist.
Figure 3.13 presents the graphite requirements as a function of
time over the period 1980 to 2000 for two electric car sales scenarios,
and also relates these requirements to forecast US demands.
3.1.8 Other Materials
SCAB electric car requirements for sulfur (in the lithium-sulfur
cell) and copper (in motors) have been calculated. Based on high utili-
zation rates of electric cars in SCAB, and no recycling of the material,
the requirements for each of these materials represent small fractions
of the existing and forecast demands. SCAB requirements at 80 percent
electric car population are 1.1 percent and 0.46 percent of 1968 demands
for sulfur and copper respectively. We assume that battery sulfur is
8-53
-------
150 x 103r-
100
50
2 yr BATTERY LIFE
3 yr BATTERY LIFE
FORECAST OF
US DEMAND
1968 US DEMAND
1968 US
SSSS PRODUCTION
UPPER BOUND RATE OF SALES
IN SCAB (NO RECYCLE)
HIGH RATE (NO RECYCLE)
1970
1980
YEAR
1990
2000
Figure 3.13. Graphite Requirement as Function of Time
not recycled, and that copper in the motor and car is only recycled
when the car's life expires (approximately 17 years life span, 50 percent
surviving at 11.5 years), which leads to inconsequential levels of
recycled material until after one life span period has passed in the
buildup of the electric car population.
Sulfur supplies are plentiful, and present environmental concerns
to remove sulfur from fossil fuels, and consequently make sulfur available
as a byproduct, promise to add further to supplies.
Although SCAB requirements are an insignificant factor in US copper
supply and demand, we should note that the US depends on the rest of the
world for approximately 25 percent of its total demand for copper. The
US has significant copper reserves of 85.5 million tons, to which must be '
added the approximately 40 million tons of copper in use. Copper is
reclaimed to a significant degree, which lends importance to the in-use
8-54
-------
reserve. World reserves were estimated in 1968 to be slightly more than
three times US reserves. Demand growth in other developed countries
has been substantial and is expected to continue. Consequently, world
supplies cannot be considered plentiful.
Chlorine gas to supply zinc-chlorine batteries should be in plentiful
supply. A 100 percent electric car inventory in SCAB in the year 2000
represents about 270,000 tons of chlorine in use; this amount is only
3 percent of the 1968 US production. Thus even high rates of buildup
in the electric car population using zinc-chlorine batteries would be
an insignificant contributor to total US demands for chlorine.
3.1.9 Summary of Impacts on Mineral Supplies and Demand
In the previous paragraphs the materials requirements for various
levels of SCAB electric car populations and usage have been quantified.
These requirements then are compared to historical and forecast supply
1 8
and demand relationships developed by the BOM. In general, the fore-
cast relationships do not explicitly account for the advent of electric
cars on any significant scale, although the importance of particular
minerals to such a development is noted. Instead the forecasts by the
BOM reflect considered, expert judgments of the effects of known techno-
logies on future growth trends. Consequently, electric car requirements
represent variations from, the underlying considerations in the forecasts
and should be judged as additions to the forecast levels.
In Table 3.5. we summarize the electric car impacts on mineral
resources for the most important materials for each battery technology.
The table presents the present and year-2000 forecast production and
demands, and compares them to the estimated requirements for particular
levels of electric car usage within SCAB, selected to reflect the appro-
priate time period for each battery technology and corresponding sales
expectations. The annual material requirement to support electric car
usage in SCAB is based on a two-year average battery life and a 90
8-55
-------
oo
I
TABLE 3.5
SUMMARY OF RESOURCE IMPACTS
Quantities, Thousands of Tons per Year
US Primary Production
Battery Type
Lead-Acid
Nickel-Zinc
Zinc-
Chlorine
Lithium-
Sulfur
Material
Lead
Antimony
Nickel
Zinc
Zinc
Titanium
(metal)
Chlorine
Lithium
Graphite7
Sulfur
1968
354
1.9
15
529
305
(0)
8,400
2.9
3.0
11,000
2000 (Range)
520-1,120
2.5-4.8
36-52
786-1,500
670-1,610
(0)
26,400-43,900
9.4-14.4
4.0-4.7
28,000-45,000
US Primary Demand
1968
880
21.1
160
1,406
440
(13)
8,400
2.6
60
10,000
2000 (Range)
1,300-2,800
28-52
382-550
2,040-4,000
960-2,160
(62-234)
26,400-43,900
8.7-13.1
80-135
26,000-41,500
*
Equilibrium
Annual Electric
Car Requirement
for New Material
with 2-Year Bat-
tery Life and 90%
Recycle (SCAB)
22
0.6
29
20
12
9.5
130*
3.2
43+
130T
Usage
Conditions
(SCAB Level
and Year)
17%, 1980
17%, 1980
46%, 1990
46%, 1990
100%, 2000
100%, 2000
100%, 2000
100%, 2000
100%, 2000
100%, 2000
Potential
Problem
((Electric Car Requirement as
Percent of National Primary
Demand for Given Usage Year)
SCAB
Implementation
1.7%
2.1
7.8
0.8
0.4
0.6
0.4
29tt
40
0.4
**
Implementation
33%
43
157
16
8
12if
8
5877t
800
8
Equilibrium conditions for the fraction of electric cars in the total population are assumed; that is, the electric car population fraction
is relatively constant and not undergoing any long-term buildup.
The scaling factor to be applied to the "Annual Requirements" column for nationwide electric-car use is approximately 20.
No recycle assumed.
ft
'•Electric car requirements for these materials would impact on production capacity for metallic forms.
-------
percent recycle rate.
In summarizing the impacts, the most significant comparisons to be
made are between the estimated primary demand and the electric car annual
requirement. The column headed "US Primary Production" is presented for
comparison with primary demand, to demonstrate sensitivities to foreign
sources. The impacts are qualitatively assessed in the last columns,
which show an "X" where a SCAB (or nationwide) implementation'at1the
assumed usage level would pose a significant problem for that material's
supply and demand situation. It must be remembered that in general a
rapid buildup in electric car population, say 10 years to achieve the
assumed usage level, can increase the annual requirement for a given
material by a factor of two over those shown in Table 3.5. In those
cases, where assumption of a two-year battery life and a 90 percent
recycle rate may be overly conservative (e.g., nickel-zinc batteries),
the requirement would undoubtedly be constrained by the rate of buildup
in the electric car population.
Assuming that a material requirement which exceeds 20 percent of that
material's projected national primary demand in the year identified for
usage conditions represents a significant impact, potential problems
are identified for SCAB implementations involving only lithium-sulfur.
On this same basis, potential problems for a national implementation
are identified with at least one material for each of the four battery
technologies. However, as noted in the table, the impacts on titanium
and lithium supply and demand are associated with production capacities
of the metallic forms rather than resource availability. For lithium
metal, electric car requirements exceed the total of all other demands,
but lithium reserves are more than adequate. There are large worldwide
reserves of both nickel and graphite, so that it is feasible to consider
even greater expansions in demand than the forecasts shown in the table.
8-57
-------
3.2 US AND WORLD RESERVES OF ELECTRIC CAR MATERIALS
The forecasts of future supply and demand relationships have taken
into account the estimated levels of both US and world reserves. In this
section we briefly review the estimates of these reserves of electric car
materials.
Estimates of US and world reserves of minerals must be viewed with
some caution. By definition, "reserves" refers to the known availa-
bility of resources that can be economically recovered at prevailing or
reasonable future prices. Hence, estimates are quite often conservative
in that they do not adequately allow for new technologies which can make
profitable heretofore uneconomical sources. In this sense, technology
is very likely to be as important as new geological discoveries in enhanc-
ing the economic supplies (reserves) of minerals. Quite often it is the
coincident effects of price increases and new technology that bring about
significant expansions in supply. The supply of nickel is a good case
in point: present prices have provided sufficient incentive to process
the lateritic nickel deposits (a low-grade ore), which are much more
widespread than the sulfide deposits previously relied on. As a result,
21
world nickel reserves have been augmented manyfold. Only a small
amount of the lateritic ores of known quality and profitability are in-
cluded in the Bureau of Mines (BOM) estimates of world reserves. None-
the less, recognizing that in most cases the BOM estimates of reserves
are probably conservative for the reasons stated above, we present in
Table 3.6 the BOM estimates of US and world reserves for those materials
deemed most important to electric car production and use. Also shown
in Table 3.6 is the estimated total amount in use for a 100 percent
SCAB electric car inventory in the year 2000. As indicated in the
table, US reserves of lead, lithium, zinc, and titanium appear adequate;
graphite reserves are marginal; and antimony reserves would leave little
for other uses. US nickel reserves under BOM estimates appear inadequate.
However, world reserves of antimony, graphite, and nickel appear to be
adequate, although their availability may be subject to more uncertainty
than domestic sources.
8-58
-------
TABLE 3.6
ESTIMATED US AND WORLD RESERVES
Quantities, Thousands of Tons
US Reserves World Reserves
Lead 39,000 99,000
Antimony 110 4,000
Lithium 5,254 6,036
Graphite 600 M.O
Nickel 900 >75,000
Zinc 78,000 90,000
Titanium 25,250 160,000
*
SCAB levels scaled by factor of 20.
00
i
Ul
VO
Amounts in Use, 100% SCAB
Electric Car Inventory,
Year 2000
3,550
91
64
106
1,360
Ni-Zn 1,000
Zn-Cl2 240
188
100% National
Electric Car Inventory,
Year 2000*
71,000
1,820
1,280
2,120
27,200
20,000
4,800
3,760
-------
Figures 3.14 to 3.20 show in more detail the total amounts of
critical materials in use in electric cars in SCAB as a function of the
electric car population for the years 1980, 1990, and 2000. Existing
and forecast primary production rates are shown for reference, and also
as an indication of the limitation that material production might enforce
on the rate of buildup of the electric car population in SCAB. These
curves should be scaled by a factor of 20 to provide estimates of amounts
of material in use for a nationwide implementation of electric cars.
3.3 RESOURCE IMPLICATIONS OF A NATIONAL IMPLEMENTATION
The previous discussion has dealt primarily with the impact of
electric car use within the Los Angeles region on the US supply and
demand balances. SCAB population is about 5 percent of the total US;
thus, we can obtain a rough estimate of the resource implications of a
national implementation of electric cars by scaling the previous curves
by a factor of 20. Clearly lead and antimony supplies could support
only a small percentage of electric car use on a national scale, but
this is consistent with the low expectation of an electric car future
based only on lead-acid battery technology. Electric cars using lithium-
sulfur cells will place significant demands on lithium supplies. How-
ever, the availability of large US reserves plus the anticipated develop=
ment of a recycle market in metallic lithium should prove sufficient.
Only the rate of buildup of electric car population based on future
technology may be limited by the rate at which lithium refining and
processing capacity may be expanded. Graphite requirements on a national
scale will further exacerbate our dependence on foreign sources, which
on a physical bases seem sufficient. If worldwide availability of
resources were the constraining factor, it would be feasible, based on
estimated world reserves, to consider national implementations of electric
cars employing either nickel-zinc or zinc-chlorine battery technology.
8-60
-------
4 x 10
20 40 60 80
ELECTRIC CAR POPULATION, PERCENT
SCAB
, YEAR 2000
US PRIMARY
\ PRODUCTION,
I tons/yr
1968 US PRIMARY
PRODUCTION, tons/yr
100
Figure 3.14. Electric Car Lead Inventories (Lead-Acid Batteries)
80 x 10V
CO
Z3
I
40
Of.
s
f* 20
o
LU
j_
I
20 40 60 80
ELECTRIC CAR POPULATION, PERCENT
SCAB
<
o-
100
YEAR 2000 US
FORECAST
1968 PRIMARY US
PRODUCTION,
tons/yr
Figure 3.15. Electric Car Antimony Inventories (Lead-Acid Batteries)
8-61
-------
c
o
oo
ID
50 x 10-
o:
•=£
O
or
20 40 60 80
ELECTRIC CAR POPULATION, PERCENT
2000
1990 SCAB
1980
YEAR 2000
FORECAST
PRODUCTION
1968 PRODUCTION
100
Figure 3.16. Electric Car Lithium Inventories (Li-S Batteries)
160 x
a:
Q-
-------
1.6 x 10&
c
o
1.2
0.8
0.4
1968 ANNUAL US
PRIMARY PRODUCTION
I
20 40 60 80
ELECTRIC CAR POPULATION, PERCENT
100
Figure 3.18. Electric Car Nickel Inventories (Ni-Zn Batteries)
1.0 X 103r-
~ 0.5
1968 ANNUAL US
PRIMARY PRODUCTION
Ni-Zn BATTERIES
SCAB
-C19 BATTERIES 2000 SCAB
1990
1980
20 40 60 80
ELECTRIC CAR POPULATION, PERCENT
Figure 3.19. Electric Car Zinc Inventories (Ni-Zn, Zn-Cl Batteries)
8-63
-------
200 x 103r-
'VIRTUALLY NO METALLIC TITANIUM
PRODUCED IN US IN 1968
20 40 60 80
ELECTRIC CAR POPULATION, PERCENT
100
Figure 3.20. Electric Car Titanium Inventories (Zn-Cl~ Batteries)
3.4 PRICE TRENDS OF MINERAL RESOURCES
The Bureau of Mines has also made forecasts of prices out to the
year 2000 for most of the mineral resources. Price forecasts take into
account expected future demands, both US and worldwide, and the corres-
ponding availability of reserves. The BOM estimates, which are in con-
stant 1968 dollars, are presented in Table 3.7 for each of the materials
discussed in Sec. 3.1. Also shown for reference are some recent commo-
dity market quotations, where available, for these materials.
In general a significant increase in the forecast for a commodity's
price indicates the relative scarcity of that commodity. Thus, according
to the BOM forecasts, antimony, copper, nickel, and zinc are likely to
be relatively scarce well into the future. All the current prices are
well above the levels forecast by the BOM. One reason, of course, is
the substantial inflation that has occurred since 1968. Another is
the unforeseen energy crisis, which has resulted in increasing the costs
8-64
-------
TABLE 3.7
PRICE TRENDS OF MATERIALS IMPORTANT TO ELECTRIC CARS
BOM Estimates,
1968 Cents per Pound
1968
2000
Lead
Antimony
Lithium (Metallic)
13.5 11.5
45.75 60
750 to 1100 750 to 1100
Current Spot Prices
(May-June 1974),
1974 Cents per Pound
21.5-24.5
223
Graphite
(Crystalline)
Sulfur
Copper
Nickel
Titanium (Metal)
Zinc
7.2
2.1
42.2
94
132
13.5
7.2
9.5
1.9
2.0
75
200
132
27
82-87
185
36-40
of the energy associated with the production of any material. Further-
more, because of existing inflation, these prices cannot be taken to
reflect steady equilibrium prices for supply and demand relationships.
For example, the prices of copper "futures" are running well above 50
percent higher than the present spot prices. Also some of these materials
have had significant momentary price rises in their history. In 1970
antimony reached a level of 178 cents per pound for four months; yet in
the following year (1971) the price averaged out at 71.3 cents per
pound.
Price rises in these materials that persist for any length of time
may provide the incentive for establishing secondary supply sources.
8-65
-------
Lead and antimony are already highly recycled and copper moderately so.
Consequently, further price increases in lead and antimony can do little
to improve recycle rates and may instead initiate the utilization of
cheaper, alternative materials where possible. Copper recycling will
undoubtedly benefit from the high prices.
Because of the highly volatile price behavior of many commodities,
it is difficult to estimate what changes might be occurring to affect
the long-term prices. Of those materials that may be significantly
affected by electric car demands, primarily lithium, there is no firm
basis upon which to estimate future prices. For lithium, with its
apparent large reserves in this country, we would expect prices to vary
little from the BOM forecasts in terms of 1968 dollars. However, some
correction may be required for the costs of energy used in the electro-
lytic production of metallic lithium.
3.5 OTHER RESOURCE CONSIDERATIONS
Throughout the discussion of each material's supply and demand
relationship, the critical importance of secondary supplies (recycling)
was stressed, especially for those materials of greatest scarcity. Based
on past trends, the development of significant recycle efforts has depended
most strongly on direct economic incentives, primarily high costs for
processing primary sources. However, it is reasonable to expect other
incentives to reinforce trends to greater recycling, such as greater
difficulties posed in the elimination of scrap as solid waste. It has
previously been mentioned that spent battery cases and used electrolyte
pose just such a situation.
Presently batteries are recycled to a significant degree because
of the scrap value of lead. However, increasing environmental sanctions
are inhibiting the easy disposal of scrap cases by burning, or of used
electrolyte by flushing down the sewer. Thus, these waste products are
beginning to take on nuisance values which conceivably may make it more
8-66
-------
profitable, for example, to recycle battery cases than pay the extra costs
to dispose of them cleanly.
In the case of electric car batteries, which will have significantly
greater costs, the inherent value of the battery case may be appreciable
and so designed to be readily re-used. Battery depreciation may be the
most significant operating cost of an electric car; recycling of the
case as well as the valuable metal may significantly decrease these
depreciation costs. Thus nuisance value as a solid waste problem,
combined with likely high battery depreciation costs, can act in concert
to enhance recycling of materials.
Electrolyte for lead-acid batteries will not be costly and so will
little affect battery depreciation costs. Only its nuisance value and the
ease with which it can be restored to its useful state will dictate whether
battery acid will be recycled.
With regard to the advanced-technology batteries employing unique
features such as refrigeration systems or heating systems, there will
doubtless be some high-value components for which sufficient incentive
to recycle them should develop. In general where estimates of battery
depreciation costs would otherwise be high, there will be a strong tendency
to recycle materials and components. As a consequence, electric cars will
probably pose less of a solid waste problem than conventional autos (e.g.,
as pointed out in Ref. 10, there are apparent existing problems in the
disposal of waste lubricating oils from conventional cars that will not
be characteristic of electric cars).
8-67
-------
4 POTENTIAL IMPACT OF SCAB ELECTRIC CARS ON URBAN NOISE
It has already been suggested that electric cars, by virtue of
having no internal combustion engine, may be quieter vehicles to operate.
The widespread use of a large number of quieter vehicles may have an
effect in reducing urban noise levels. Concern over rising noise levels
in communities is increasing and legislative actions are becoming more
prevalent. Accordingly, we examine in this section the impacts of elec-
tric cars on urban noise levels.
4.1 URBAN NOISE ENVIRONMENTS
The level of urban noise has been found to vary greatly in magni-
tude and character among various locations throughout a community. Eval-
uations of typical urban noise environments show a residual or background
noise level that is low at night and increases in the morning to a level
22
that holds throughout the day and early evening. Superimposed upon the
residual level are many random instances of significant noise intrusions,
the character of which depends on activities peculiar to each neighborhood
or area of the community. In general, aircraft and nearby automobile
operations constitute the most prevalent and significant forms of intru-
sive noise. Noise intrusions are instances that are usually quite easily
characterized as to the source (e.g., sports car accelerating, dog bark-
ing, aircraft taking off, etc.), and are distinct from the general din
generated by continuous traffic and other general human activity.
The effects of noise pollution on human activities depend on the
kind of activity and whether it is being conducted indoors or outdoors.
The effect of the noise may be merely annoying or may greatly interfere
with the activity. A great deal of research has been conducted in the
development of noise scales that attempt to quantify the intrusive,
annoying, or interfering qualities, among others, of noise. No single
scale has been found entirely satisfactory for all aspects. One conven-
ient scale that has been found generally appropriate for overall noise
surveys is the "A-weighted noise scale," which weights the spectral content
of a sound in accord with the average spectral sensitivity of the human ear.
8-68
-------
22
The Environmental Protection Agency has conducted and reported on
surveys of outdoor community noise, utilizing primarily A-weighted sound
measurements. Eighteen outdoor environments were surveyed, ranging from
the Grand Canyon to a third-floor apartment next to a freeway. We have
selected the survey results for five of those locations as representative
environments for electric car studies. Table 4.1 presents a summary of
the results for the five selected locations in terms of the noise levels
that were exceeded 99, 90, 50, 10, and 1 percent of the time during the
daytime hours. The 90 percent level closely represents what may be
interpreted as the residual or background noise level.
The table is arranged in descending order based on background or
residual noise level. For the other levels, the locations follow the
same order, with the exception of the location near a major airport.
The effect of airport operations is manifest in the less frequent intru-
sions of aircraft noise at significantly higher levels relative to the
TABLE 4.1
A-WEIGHTED OUTDOOR NOISE LEVELS
Decibels Above 20 yN/m2 (dBA)
Level Exceeded x% of Time
90% 50% 10%
Third Floor Apartment, Next to Freeway 76 77 80 85 89
Third Floor High-Rise, Downtown Los
Angeles 69 72 77 82 86
Urban Shopping Center 59 62 64 69 72
Urban Residential Area Near Major
Airport 49 53 59 76 93
Urban Residential Area 46 47 52 58 66
Effective residual or background level.
8-69
-------
background. The ordering of the table also correlates well with descend-
ing auto traffic environments, which may be important sources of background
noise levels.
Widescale utilization of electric cars will undoubtedly affect traf-
fic noise in any of the environments typified in Table 4.1. The signifi-
cance of that effect will depend on the electric car's inherent noise
generation relative to the conventional auto it displaces, and the rela-
tive usage it obtains. One effect of electric cars thus may be to alter
the background levels slightly.
Even if it were to be shown that electric cars might reduce back-
ground noise levels, it is not certain that the reduced background would
be an improvement; in many instances there is adaptation to background
noise, and it is the momentary intrusions which are annoying. However,
in situations of high background levels, such as exemplified in the first
location of Table 4.1, reductions in that level will very likely represent
an improvement.
On the other hand, noise intrusions most often cited as interfering
usually entail one single recognizable noise source such as a poorly muf-
fled auto or motorcycle undergoing acceleration on a street very near to
the recipient of the noise. The background may be anywhere from 10 to
40 dB lower in noise level than the intrusive source, as seen by compar-
ing the 90 percent and 1 percent levels in Table 4.1. Electric cars may
have an even greater effect on reducing occurrences of such intrusions.
4.2 COMPARISON OF ELECTRIC AND CONVENTIONAL AUTO NOISE GENERATION
23
R.J. Vargovick of the Ford Motor Company reported on noise mea-
surements for several types of autos under various operational modes.
Tests were conducted to study the relative significance of exhaust, tire-
road interaction, and fan noises in total auto noise. Table 4.2 sum-
marizes the results of those studies, showing the A-weighted sound levels
8-70
-------
TABLE 4.2
I
EXTERIOR NOISE LEVELS FOR VARIOUS CONDITIONS AM) 1970 VEHICLE TYPES
dBA at 25 ft Distance
Mode of Operation
Cruise
Cruise (no fan)
Coast (engine off)
WOT*
WOT (two mufflers)
WOT corrected to 50 ft
High-Power
Sedan
66.5
65.5
65.5
88.3
82.5
82.3
Low-Power
Sedan
67.5
66.0
64.8
80.8
~
7A.8
35 mph
High-Power
Sporty Compact
70.0
71.0
66.0
87.2
—
81.2
Low-Power
Sporty Compact
68.2
68.7
68.2
87.2
—
81.2
High-Power
Sedan
74.5
74.5
74.0
89.5
—
—
Low-Power
Sedan
77.7
75.5
75.7
82.5
—
—
65 mph
High-Power
Sporty Compact
78.7
78.5
76.0
88.7
—
—
Low-Power
Sporty Compact
79.2
78.7
78.0
87.7
—
—
NOTE: Production tires for sedan—H78-15 bias belted; production tires for sporty compact—F70-14 bias belted.
WOT—Wide open throttle.
OO
I
-------
for several modes of operation. The individual contributions of exhaust,
tire-road, and fan noise can be inferred from these data.
Of the four types of autos studied, the low-powered sedan (which
we assume is an adequate representation for a compact sedan) is the
most likely candidate to be displaced by electric cars. The results
for this auto indicate that at 35-mph cruise conditions on a smooth
concrete surface roughly half the noise occurs from tire-road contact
(64.8 dBA coasting). Exhaust plus fan noise contributes the remainder.
At 65-mph cruise conditions, the picture is similar except that the fan
appears to be a greater contributor. However, in instances of "wide
open throttle" operation, exhaust noise easily dominates over tire-road
and fan noises at either speed.
It is difficult, of course, to deduce the relative contributions
of each of these sources to general traffic noise, where at any given
instant there is a mixture of cruising operation and various levels of
acceleration (i.e., more open throttle settings). Also, tire-road
noise depends strongly on the road surface as well as several other
variables. Typical ranges in road surface roughness can introduce up
to 7 or 8 dB variation in tire-r
is at the low end of the range).
23
to 7 or 8 dB variation in tire-road noise generation (smooth concrete
Under traffic conditions where speeds are likely to be nearer 35
mph, we would expect the general level of traffic noise to be the result
of a large percentage of stop-and-go driving, implying frequent periods
of acceleration. Under these circumstances auto exhaust noise is likely
to dominate over other sources. For example, an average of 10 percent
of the time spent in acceleration at a level halfway between cruise and
wide open throttle would produce an integrated effective noise level
4 dB above the tire-road level (for smooth concrete). On very rough
roads, tire-road noise may dominate over modest acceleration noise con-
tributions. However, wide open throttle operations when they occur will
probably dominate over all other auto noise sources.
8-72
-------
At 65 mph, which imples cruise conditions with little opportunity
for acceleration, we would expect little contribution to aggregate noise
levels from instances of wide-open throttle operation. Thus the noise
levels shown for 65-mph cruise conditions, for a representative mix of
the autos in Table 4.2, may fairly characterize traffic noise for uncon-
gested freeway travel.
As previously stated, electric cars will have no exhaust noise
contributions; however, apart from normal tire-road and drive-train
noise sources which will remain, their electric motors may also be a sig-
nificant noise source. In general, an electric motor can be a source of
a variety of noises ranging from purely mechanical sources such as bearing
noise, unbalances, and friction, to airflow sources associated with the
moving parts (which may include small cooling fans), to magnetic sources
25
arising from magnetostrictive effects. Each of these noise sources is
susceptible in various degrees to mitigation through careful design and
manufacture. The relative contributions of each source to total noise
generation are not easily established and depend strongly on the type
of motor and its application.
The requirements for a relatively light and efficient motor with
the right torque-speed characteristics for electric car propulsion are
sufficiently unique that it is difficult to speculate on what may be its
significant noise sources. The equipment noise specifications that were
met in developing the Bay Area Rapid Transit (BART) system indicate that
the cooling system of the electric traction motors was the most signifi-
r\r
cant noise source. In particular, the cooling fan and its associated
air flow paths required the greatest effort in designing for quiet opera-
tion. No mention was made in Ref. 26 of any particular difficulty with
noise due to magnetic sources in the motors, and it was concluded that
BART-car wayside noise levels are lower than those of current trucks and
busses.
27
In another study of train system noises, an estimating relation-
8-73
-------
ship was developed to predict the noise generation from electric traction
motors. The relationship is given by the following expression by which
sound power levels (PWL) in octave bands can be estimated as a function
of motor horsepower (P) and speed (V):
PWL = Br + 10 log1QP/Pr + 15 log1QV/Vr
where the B are the reference power levels in each octave band for the
reference conditions (P , V ) of motor horsepower and speed. For P = 150 hp
and V = 1400 rpm , the B for each octave band are given as follows:
Octave Band Center Frequency, Hz
31 63 125 250 500 1000 2000 4000 8000
Sound Power Level (B ),
dB above 10~12 watts 93 94 98 102 103 103 102 96 89
The estimating equation is based on noise measurements of more than ninety
bare motors (i.e., without noise quieting enclosures) ranging in power
between 1 and 4,000 hp and in speed between 450 and 3,600 rpm. The measure-
ments included all component sources of traction motor noise: magnetic,
mechanical, and aerodynamic. The reference condition produces an A-weighted
sound pressure level at 25 feet of 81.6 dBA. These levels may be signifi-
cantly reduced when the motor is installed as part of a vehicle with
attention paid to noise quieting.
Traction motors used in electric cars will not be operated at as
high a power level as the reference motor characterized above. The four-
passenger electric car configuration selected for study has a peak power
capability of 85 hp ; however, for typical driving conditions it will
draw only a fraction of this power. For example, at a steady cruising
speed of 35 mph, it is estimated that only 8 hp are required on a level
road. On the other hand, during periods of acceleration considerably
more power will be required.
8-74
-------
Operating motor speeds for different electric car configurations
may vary widely, since motor designers are relatively free to juggle this
parameter in optimizing motor design. In one study of electric vehicle
28
systems, characteristic rated speed was allowed to vary between 6000
29-31
and 12,000 rpm. Other electric-car designs have considered motors
with rated speeds in the 3,000-to-4,000-rpm range. Rated motor speeds will
be achieved only at top speed or when changing gears while accelerating;
otherwise motor speeds will generally be much lower.
Traction motor noise for electric cars undergoing acceleration was
calculated using the above described estimating relationship with design
motor speed as a parameter. Figure 4.1 presents the results of those cal-
culations and compares the electric motor noise levels with the values
reported for conventional autos at wide open throttle presented in Table
4.2. It is assumed that maximum acceleration levels will most frequently
3 ^
5
90
80
70
60
50
CONVENTIONAL CARS,
WIDE OPEN THROTTLE
RANGE BETWEEN LOW T
POWER AND HIGH §
POWER SEDANS
ELECTRIC CAR, MAXIMUM
ACCELERATION (3000 rpm,
RATED MOTOR SPEED)
*ASSUMES ENGINE NOISE AT
WIDE OPEN THROTTLE IS
INDEPENDENT OF VEHICLE
SPEED
10
20 30
VEHICLE SPEED, mph
40
Figure 4.1. Comparison of Electric Car and Conventional Car Engine Noise
(Maximum Acceleration at Low Speeds)
8-75
-------
occur during traffic or low speed operations. Thus, according to the
curves in Fig. 4.1, motor noise levels for either electric cars or conven-
tional cars will dominate over tire-road noises (Table 4.2 indicates that
tire-road noise at 35 mph for a low-power sedan is approximately 65 dBA).
Expected noise quieting for installed electric car motors will further
improve the comparison between the noisier conventional autos at wide
open throttle and electric cars. If, for example, noise quieting should
result in a 10 dBA drop in the electric car motor noise levels, motor
noise under maximum acceleration conditions will continue to dominate over
tire-road noises (except in the cases of very rough roads) and the elec-
tric car total noise level would be decreased proportionately. By this
comparison, electric cars should aid materially in reducing the incidence
of noise intrusions by accelerating automobiles.
In order to draw further comparisons of electric car noise genera-
tion with that of the conventional auto, we used the above estimating
relationship to evaluate electric car noise at typical 35-mph and 65-mph
cruise conditions. At 65 mph the four-passenger configuration requires
about 35 hp, for which it was assumed that vehicle speed corresponds to
rated motor speed. At 35 mph, required cruise power is 8 hp and motor
speed is approximately 54 percent of rated speed. Table 4.3 summarizes cal-
ulations of expected A-weighted noise levels at 25-foot distances for these
conditions, with rated motor speed parameterized over a range of 3,000 to
12,000 rpm. The equivalent value for the low-power sedan of Table 4.2
is presented for comparison.
At the lower speeds, expected quieting for the installed electric
car motors should lower the motor noise contribution to where tire-road
noise, even on smooth surfaces, dominates. Under cruise conditions the
electric car will then contribute as much to urban noise as conventional
low-power sedans. Only in the case of the highest electric motor speed
does it appear that motor noise, allowing for quieting, may become the
most significant source. If the estimating relationship is accurate in
8-76
-------
TABLE 4.3
COMPARISON OF CONVENTIONAL AUTO AND ELECTRIC
CAR MOTOR NOISE LEVELS, CRUISE CONDITIONS
2
dBA at 25-ft Distance re 20 yN/m
35 mph 65 mph
Rated Motor
Speed, rpm*
3,000
6,000
12,000
Electric Car
Without Noise
Quieting
69.9
74.4
78.9*
Conventional
Low-Power
Sedan**
67.5
Electric Car Conventional
Without Noise Low-Power
Quieting Sedan**
80.3 77.7
84.8*
*
89.3
Note that the estimating relationship was based on data for motor speeds
in the range 450 to 3,600 rpm; consequently, the calculations with
greater speeds must be viewed with some caution.
**
Columns 1 and 6 in Table 4.2; total noise, cruise condition.
the sensitivity it accords to traction motor speed, these noise levels
suggest that low rated motor speed should be selected as a design cri-
terion. However, operation at cruise speeds on rough roads would probably
increase tire-road contributions to a level that would dominate total noise
generation from either type of vehicle. The estimates in Table 4.3 at
high electric motor speeds must be viewed with caution since the relation-
ship was developed from motor data that spanned the speed range 450 to
3,600 rpm. Also a comparison of predicted sound pressure levels based
on the estimating relationship with actual sideline measurements of a
BART car showed agreement only within 5 to 10 dB for octave-band levels
27
over the sound spectrum. Otl
ing relationship is not known.
27
over the sound spectrum. Otherwise the goodness of fit of the estimat-
It is noteworthy to report some of the data on auto noise genera-
tion resulting from the 1972 Urban Vehicle Design Competition (UVDC).
8-77
-------
Comparison of noise measurements taken at 25 feet sideline distance of
the Cornell electric car and 35 other experimental cars employing variously
fueled internal combustion engines showed the electric to be 12 dB quieter
in terms of overall sound pressure level (the electric was measured at
o p
64 dB while the average of the 35 other vehicles was 76.3 dB). The mea-
surements were taken at a cruise condition of 30 mph. These values are
not directly scalable to A-weighted noise levels, since the spectral con-
tents of the sound levels are not known and are probably quite variable
over the range of experimental vehicles investigated. Nonetheless, the
figures tend to indicate that electrics under cruise conditions will be
significantly quieter than conventionally powered internal combustion
cars.
4.3 EFFECT OF ELECTRIC CAR USE ON COMMUNITY NOISE
Based on the foregoing appraisal of likely electric car noise genera-
tion, some general observations and conclusions are offered concerning
the effects wide usage of electric cars may have on community noise.
With regard to general background traffic noise, the electric car
is likely to have some impact only through a reduction of the contribu-
tions arising from start-and-stop operation in traffic. Operation on
surface streets with traffic regulation involves much stop and go, with
corresponding accelerations and decelerations, under which conditions the
electric car would be expected to be significantly quieter.
The total effect on background noise further depends on the propor-
tion of electric cars in the vehicle population. The reduction in the
contribution to background noise, on an integrated basis involving cruise
and acceleration conditions, due to electrics replacing conventional cars
is not large and will probably be significant only if there is a large
population of electric cars. Also there may be other sources that domi-
nate background noise levels in some instances which will remain unaffected
by electric car use, e.g., significant levels of diesel truck operations
8-78
-------
in some neighborhoods. All these considerations combine to create a
situation in which it is difficult to quantify the likely effect of elec-
tric cars on background community noise levels.
Within the category of traffic noise intrusions, significant sources
are: motorcycles, variations in traffic flow, automobile exhaust noise,
and diesel trucks. Surveys on noise annoyance factors indicate that
often these intrusions are believed to be the result of improper operation
of the vehicles, such as gunning engines, accelerating at high speeds,
33
etc. They tend to be much more noticeable at night and involve vehicle
operations near the receiver to the extent that the noise level incident
at the receiver rises 20-to-30 dB or more above the background.
Electric cars do not have the characteristics to produce noise intru-
sions of this sort, as indicated by the comparison shown in Fig. 4.1.
However, it is difficult to speculate to what degree purchasers of elec-
tric cars would be previous owners of the type of cars generally thought
to be involved in most noise intrusions. Surveys have shown that most
often the annoyed person thought the source, when identified as an auto-
mobile, to be a sports car. In general, electric cars are not likely to
appeal to those purchasers seeking the characteristics offered by a
sports car. However, in those cases of noise intrusions from inadvertent:
auto operations, we would expect electric cars to have a significant
effect in reducing the number of noise intrusion incidents.
4.4 ELECTRIC GENERATION POWER PLANT NOISE
34
A 1971 survey by the EPA on noise generation from industrial plants
included studies of the effect of electric power plant noise on community
noise levels. The electric power plant that was surveyed employed five steam
turbine generators of 100 MW capacity and one gas turbine generator. The
plant was 3,000 to 4,000 feet from the nearest residential neighborhood. Noise
measurements were taken at a variety of stations ranging from just out-
side the plant building to several places throughout the nearby neighbor-
hood. Day and night measurements were made, and it was concluded that
8-79
-------
the power plant was not the primary source for the background noise
levels measured anywhere within the nearby residential neighborhood,
although it may have been a significant contributor. At the 10 percent
exceedance level, the average values in the nearby community were 61 dBA
during the day and 60 dBA during the night. The average background
noise levels were around 53 dBA for either night or day. These values
suggest a representative urban neighborhood (see Table 4.1). This
conclusion held for both day and night conditions. Therefore, based
on this limited sample, it is unlikely that the additional generating
requirements for recharging large numbers of electric cars at night
will significantly affect nearby community noise levels.
8-80
-------
REFERENCES
1. N. Mapham, Conservation of Petroleum Resources by the Use of Electric
Cars, Society of Automotive Engineers 740171, March 1974.
2. P. A. Nelson et al., ANL High Energy Batteries for Electric Vehicles,
Third International Electric Vehicle Symposium, Washington, February
19-20, 1974.
3. D. P. Gummer and Kuszczynski, "Lost Power," Environment, Vol. 14,
No. 3, April 1972.
4. G. Houser, Population Projection for the Los Angeles Region, 1980-
2000, General Research Corporation RM-1842 (Also Task Report 2).
5. A. R. Sjovold, Electric Energy Projections for the Los Angeles
Region, 1980-2000, General Research Corporation RM-1859, November
1973 (also Task Report 5).
6. W. F. Hamilton and G. Houser, Transportation Projections for the
Los Angeles Region, 1980-2000, General Research Corporation RM-1858,
November 1973 (also Task Report 3).
7. D. Friedman, J. Andon, and W. Hamilton, The Characterization of
Battery Electric Vehicles for 1980-2000, Minicars, Inc.,
January 1974 (also Task Report 1).
8. J. Eisenhut, Economic Projections for the Los Angeles Region,
1980-2000, General Research Corporation RM-1860, February 1974
(also Task Report 4).
9. W. F. Hamilton, Usage of Electric Cars in the Los Angeles Region,
1980-2000, General Research Corporation RM-1891, April 1974
(also Task Report 10).
10. A. R. Sjovold, Identification of Electric Car Impacts, General
Research Corporation RM-1864, November 1973.
11. Meeting California's Energy Requirements, 1974-2000, Stanford
Research Institute, May 1973.
12. Patterns, of Energy Consumption in the United States, Office of
Science and Technology, January 1972.
8-81
-------
References (cont.)
13. Energy and the Environment, Electric Power, Council on Environmental
Quality, August 1973.
14. E. Hirst, Energy Consumption for Transportation in the US, Oak
Ridge National Laboratory ORNL-NSF-EP-15, March 1972.
15. System Forecasts, 1973-1995, Southern California Edison Company,
February 1973.
16. Feasibility Study of Alternative Fuels for Automotive Transportation,
3 Vols, U.S. Environmental Protection Agency, EPA-460/3-74-009,
a,b, and c, June 1974.
17. William D. Ruckelshaus, EPA Statement before the Committee on
Interior and Insular Affairs, U.S. Senate, hearing on Advanced
Power Cycles, Serial No. 92-21, February 8, 1972.
18. Mineral Facts and Problems, 1970 Edition, Bureau of Mines Bulletin
650.
19. The Storage Battery Manufacturing Industry, 1971-1972 Yearbook,
Battery Council International.
20. S. J. Kostman and F. E. Van Voris, The Problems, Opportunities and
Outlook for the Domestic Lead and Zinc Industry, St. Joe Minerals
Corporation, March 17, 1974.
21. D. B. Brooks and P. W. Andrews, "Mineral Resources, Economic Growth,
and World Population," Science, Vol. 185, p. 4145, July 1974.
22. Community Noise, US Environmental Protection Agency NTID300.3,
December 1971.
23. R. J. Vargovick, Noise Source Definition—Exterior Passenger
Vehicle Noise, Socity of Automotive Engineers 720274, January 1972.
24. W. A. Leasure, Jr., "Automobile Tire Noise: A Review of the Open
Literature," National Conference on Noise Control Engineering,
Institute of Noise Control Engineering, October 15-17, 1973.
25. C. M. Harris, et al., Handbook of Noise Control, McGraw-Hill, 1957.
26. G. P. Wilson, "Noise Performance Achieved by the San Francisco
Bay Area Rapid Transit District," National Conference on Noise
Control Engineering, Institute of Noise Control Engineering,
October 15-17, 1973.
8-82
-------
References (Cont.)
27. A Study of the Magnitude of Transportation Noise Generation and
Potential Abatement, Volume V, "Train System Noise," Serendipity
Incorporated NTIS PB-203186, November 1970.
28. L. E. Unnewehr, Electric Vehicle Systems Study, Ford Motor Company.
29. R. S. McKee and B. Borisoff, Sundancer: A Test Bed Electric
Vehicle, Society of Automotive Engineers 720188, January 1972.
30. W. E. Goldman, The Design and Development of a Third Generation
Electric Highway Vehicle, Society of Automotive Engineers 720110,
January 1972.
31. J. J. Gumbleton et al., Special Purpose Urban Cars, Society of
Automotive Engineers 690461, May 1969.
32. S. L. Rosson, "Cornell's Electric Car," Engineering; Cornell
Quarterly, Vol. 8, No. 4, Winter 1974.
33. A Study of Annoyance from Motor Vehicle Noise, Bolt Beranek and
Newman, Inc. Report 2112, June 1971.
34. Noise from Industrial Plants, US Environmental Protection Agency
NTID300.2, December 1971.
8-83
-------
8-84
-------
TASK REPORT 9
PARAMETRIC ECONOMIC IMPACTS
OF ELECTRIC CARS IN LOS ANGELES
J.C. Elsenhut
J.A. Cattani
F.J. Markovich
-------
ABSTRACT
This paper estimates the economic impacts which would accrue to the
Los Angeles region in the event electric cars are substituted for internal
combusion engine cars. The average 1990 internal combustion car in the
Los Angeles region is estimated to have a life-cycle cost of 15.5 cents a
mile (in 1973 dollars). Electric car life-cycle cost per mile would range
from 23 percent less to 38 percent more, depending on the type of battery
employed. Reductions in regional employment opportunities would range from
0.2 to 1.2 percent, also depending on battery type.
-------
ii
-------
CONTENTS
SECTION . PAGE
ABSTRACT i
1 INTRODUCTION AND SUMMARY 9-1
1.1 Introduction 9-1
1.2 Summary 9-1
2 IMPACTS ON TRANSPORTATION CONSUMERS: A COMPARISON
OF VEHICLE COSTS 9-4
2.1 Internal Combustion Engine Vehicle Costs 9-4
2.2 Electric Vehicle Costs 9-9
3 IMPACTS ON TRANSPORTATION SUPPLIES AND SERVICES 9-23
3.1 Major Impacts 9-23
3.2 Moderate Impacts 9-33
3.3 Minor Impacts 9-46
4 SECONDARY IMPACTS 9-55
REFERENCES 9-61
iii
-------
iv
-------
TABLES
NO. PAGE
2.1 Internal Combustion Engine Automobile Operating Cost at
10,000 Miles a Year for 10 Years 9-7
2.2 Summary of Internal Combustion Vehicle Operating Cost per
Mile at 10,000 Miles/Year for 10 Years in 1973 Dollars 9-9
2.3 Four-Passenger Electric Car Operation Cost with the Lead
Acid Battery for 12 Years in 1973 Dollars 9-10
2.4 Four-Passenger Electric Car Operation Cost at 10,000 Miles
per Year for 12 Years in 1973 Dollars 9-11
2.5 Summary of Life-Cycle Costs 9-21
3.1 Impacts on Battery Manufacturing (SIC 3691) with 100
Percent Electric Car Usage in the SCAB 9-25
3.2 Impacts on Petroleum Distribution (SIC 5092 and 5541) with
100 Percent Electric Car Usage in the SCAB 9-30
3.3 Service Station Sales 9-32
3.4 US Retail Sales of Automotive Lines for 1972 9-36
3.5 Top Selling Automotive Lines 9-37
3.6 Auto Repairs and Service 9-38
3.7 Increases in Employment in Automotive Aftermarket and
Repair Industry Due Exclusively to Use of Lead-Acid
Battery 9-44
3.8 , Increases in Employment in Automotive Aftermarket and
Repair Industry 9-45
3.9 Total Net Impact of Electric Cars on Employment and Pay-
roll in the Automotive Aftermarket and Auto Repair
Industry 9-47
-------
TABLES (Concl.)
NO. PAGE
3.10 Comparison of Prices for Electric Cars with Internal Com-
bustion Engine Cars by Year 9-48
3.11 Recommended Inventories for Car Dealers Service and Repair 9-50
3.12 US Automobile Manufacturing 9-53
3.13 National Impacts on Motor (SIC 3621) and Industrial Con-
trols (SIC 3622) Manufacturing with 100 Percent Electric
Car Usage in Los Angeles 9-54
4.1 Impact of 100 Percent Electric Car Usage on Employment
and Income in the Los Angeles Area 9-56
4.2 Structure of Employment Changes 9-57
4.3 Potential Cumulative Impacts in Los Angeles 9-58
vi
-------
1 INTRODUCTION AND SUMMARY
1.1 INTRODUCTION
This study presents, in parametric form, the impacts imposed on the
Los Angeles economic system by electric car usage. Task Report 4 (also
Ref. 1) projected the levels of economic activity for these enterprises
which were considered susceptible to electric car impacts. The projec-
tions were based upon current trends without electric car usage. This
study, for various levels of electric car usage, quantifies the impacts
that can be expected upon those economic activities.
Section 2 develops the impacts which would accrue to transportation
consumers, primarily the comparative costs of internal combustion engine
automobiles and electric cars. Section 3 shows the impacts on suppliers
of transportation equipment and services, discussed by level of impor-
tance. Section 4 discusses the secondary impacts, specifically, the cost
changes imposed upon disposable personal income.
1.2 SUMMARY
The impacts of conversion from internal combustion engine automobiles
to 'electric cars are shown as they affect the consumer and the Los Angeles
regional economy (the California South Coast Air Basin, or SCAB, to be pre-
cise) . The impact upon the consumer is shown by deriving the user life-
cycle cost of the average internal combustion engine automobile (the 1973
cost of each car class adjusted for the expected market share of that class
in future decades) and the comparable cost of the electric car. The impact
upon the region (secondary impacts) are calculated by determining the em-
ployment shifts and payroll changes which would accrue. These are expected
to result from decreased automobile servicing and increased battery manu-
facturing requirements. They are displayed as a percentage of expected
employment and total personal income levels.
9-1
-------
The impact of a complete conversion to electric car use was examined
for each of four battery types, and, in two cases, best and worst perform-
ance levels. Both primary and secondary impacts are assumed proportional
to the level of conversion. The findings for each battery type are sum-
marized below. All costs are given in 1973 dollars.
Lead-Acid Battery; Best Performance Assumption
The user life cycle cost is 16.0 cents per mile. This is 1 percent
more than the user life cycle cost of the average 1980 internal combustion
engine automobile. The secondary impacts of a complete transition to
electric car use are expected to reduce employment opportunities by 0.75
percent and reduce total personal income by 0.20 percent.
Lead-Acid Battery; Worst Performance Assumption
The user life cycle cost is 21.8 cents per mile. This is 37 percent
more than the cost of the average 1980 internal combustion engine automo-
bile. This is the costliest electric car option. Secondary impacts are
negligible. A 100 percent substitution of electric cars would reduce area
employment opportunities by 0.15 percent and increase total personal income
*
by about the same amount.
Nickel-Zinc Battery
The user life cycle cost is 18 cents per mile. This is 13 percent
more than the user life cycle cost of the average 1990 internal combustion
engine automobile. Secondary impacts include a 0.85 percent reduction in
employment opportunities and a total personal income loss of 0.20 percent.
Zinc-Chlorine Battery
This is the least expensive alternative. The user life cycle cost
of 11.9 cents per mile is 23 percent less than the average 1990 internal
combustion engine automobile user life cycle cost. There are, however,
*
Total income increases while employment decreases because of the loss of
low-salaried employees and an increase of high-salaried employees.
9-2
-------
significant secondary impacts with an employment opportunity loss of 1.2
percent and a total personal income loss of 0.35 percent.
Lithium-Sulfur Battery; Best Performance Assumption
The user life cycle cost is 12.4 cents per mile, 19 percent less
than the comparable (year-2000) internal combustion engine automobile cost,
The secondary impacts are the most significant: an employment opportunity
loss of 1.4 percent and total personal income reduction of 0.35 percent.
Lithium-Sulfur Battery; Worst Performance Assumption
The user life cycle cost is 13.2 cents per mile, about 14 percent
less than the comparable (year-2000) internal combustion engine automobile
costs. The employment opportunity loss is expected to be 1.1 percent and
the total personal income loss 0.3 percent.
9-3
-------
2 IMPACTS ON TRANSPORTATION CONSUMERS; A COMPARISON OF VEHICLE COSTS
The relative costs of electric car ownership are critical in deter-
mining their desirability to the consumer. Section 2.1 develops the com-
parative costs for internal combustion engine and electric cars. The cost
of owning a car is a combination of factors: initial cost, finance costs,
maintenance, taxes, fuel, insurance, repairs, etc. All the cost elements
must be taken into consideration, along with the expected mileage, in order
to have a valid basis for comparing the costs of vehicles. The following
subsections 2.1 and 2.2 develop the costs for the internal combustion en-
gine automobile and the electric car.
2.1 INTERNAL COMBUSTION ENGINE VEHICLE COSTS
These costs are largely taken from 1972 US Department of Transporta-
2
tion (DOT) figures on auto costs. Figure 2.1 shows the assumptions that
underly the data. One DOT assumption that this study changes is the price
of gasoline. DOT assumed about 38 cents per gallon (including 11 cents
tax). All costs other than gasoline were adjusted from 1972 to 1973 dol-
lars using a 10 percent inflation factor. If gasoline were treated in
this manner, the estimated 1973 cost would be 42 cents, unrealistic in
view of recent price developments. The derivation of gasoline costs are
discussed in following paragraphs.
Other additions are made to the DOT costs for pollution devices and
their maintenance and for finance costs. These costs for pollution devices
are taken from an EPA study which derives the additions to vehicle purchase
price, given current legislative requirements for pollution control. In
3
addition, the study estimates yearly maintenance costs for the device.
The DOT figures show the cost of purchase and operation. To account for
total life-cycle costs to the user, the cost of financing the purchase
must be included. This is important because of the variation in the amounts
required for initial purchase.
9-4
-------
Item
Automobile
Description
Repairs and
Maintenance
Replacement
Tires
Accessories
Gasoline
Oil
Insurance
Garaging,
Parking,
and Tolls
Taxes
Standard Size Automobile
1972 model 4-door sedan. Equipped with:
V-8 engine, automatic transmission,
power steering and brakes, air condi-
tioning, tinted glass, radio, clock,
white-wall tires, and body protective
molding.
Compact Size Automobile
1972 model 2 -door sedan. Equipped with:
6 cylinder engine, automatic transmis-
sion, power steering, radio, and body
protective molding.
Subcompact Size Automobile
1972 model 2 -door sedan. Equipped with:
standard equipment plus radio and body
protective molding.
Includes routine maintenance such as lubrications, repacking wheel bearings, flushing cooling system, and aiming headlamps;
replacement of minor parts such as spark plugs, fan belts, radiator hoses, distributor cap, fuel filter, and pollution control
filters; minor repairs such as brake jobs, water pump, carburetor overhaul and universal joints; and major repairs such as a
complete "valve job."
Purchase of 7 new regular tires and 4 new snow tires during the lives of the cars was assumed.
Purchase of floor mats the first year, seat covers the sixth year, and miscellaneous items totaling $2.00 per year was assumed.
Consumption rate of 13.60 miles per
gallon was used.
Consumption was associated with gaso-
line consumption at a rate of 1 gallon
of oil for every 186 gallons of
gasoline.
Consumption rate of 15.97 miles per
gallon was used.
Consumption was associated with gaso-
line consumption at a rate of 1 gallon
of oil for every 166 gallons of
gasoline.
Consumption rate of 21.43 miles per
gallon was used.
Consumption was associated with gaso-
line consumption at a rate of 1 gallon
of oil for every 135 gallons of
gasoline.
Coverage includes $50,000 combined public liability ($15,000/$30,000 bodily injury, and $5,000 property damage), $1,000 medical
payments, uninsured motorist coverage, and full comprehensive coverage for the 10-year period. Deductible collision insurance
was assumed for the first 5 years ($100 deductible).
Includes monthly charges of $10.00 for garage rental or indirect cost of the owners garaging facility; plus parking fee average
of $54.00 per year, and toll average of $6.94 per year, both of which were assigned in proportion to annual travel.
Includes Federal excise taxes on tires (10 cents per pound), lubricating oil (6 cents per gallon), and gasoline (4 cents per
gallon); plus the Maryland tax on gasoline (7 cents per gallon), titling tax (4 percent of retail price), and registration
fee ($20.00 for 3,700 pounds or less shipping weight, or $30.00 for vehicles over 3,700 pounds).
VO
Ui
Figure 2.1. Automobile Operating Costs, Bases for Estimates
-------
The interest paid over the 10-year life of the car is calculated
with the following standard interest formula.
Annual Payment
where
(1 + i)n - 1
P = cost of car
i = 10 percent annual interest rate
n = 10-year lifetime of car
The interest rate of 10 percent is an average between the cost of borrowing
money and the opportunity cost to the consumer who finances his own car.
The annual payment is multiplied by n to show the total payment. When
the initial purchase price is subtracted from the amount of the total pay-
ment the residual amount is the finance cost.
All costs in this section are presented as cost per mile in 1973
dollars. This allows the costs of the two vehicle types to be compared
on an equal basis. Table 2.1 shows the development of the cost per mile
for the internal combustion engine auto.
The cost of gasoline is subject to some uncertainty. The National
4
Petroleum Council Study examined in Ref. 5 predicts a long-term equilib-
rium price of $6.50 for a barrel of petroleum in 1973 dollars. This cor-
responded to a March 1974 level of $5.25 a barrel for controlled (domestic)
oil and an average price of $10.35 for exempt (foreign) oil. The weighted
average price for a barrel of crude used in US gasline production was about
$6 to $7. Oil-company accounting principles dictate that the price of
gasoline is related to the current price of crude rather than to the cost
of inventory stocks. The Oil and Gas Journal noted that the March 1974
price of gasoline in Los Angeles was 47 cents per gallon. Thus, assuming
the price of crude is the dominant factor in gasoline price variances, the
9-6
-------
TABLE 2.1
INTERNAL COMBUSTION ENGINE AUTOMOBILE OPERATING COST
AT 10,000 MILES PER YEAR FOR 10 YEARS*
Standard Compact Subcompact
Costs Excluding Taxes
Amortization
**
Pollution Device
Finance Costs
Repair and Maintenance
**
Maintenance of Pollution Device
Replacement Tires
Accessories
Gasoline
Oil
Insurance
Garaging, Parking, Tolls, etc.
Subtotal
Taxes and Fees
State
Gasoline
Registration
Titling
Federal
Gasoline
Oil
Tires
Subtotal
Total
Cents Per Mile
$ A, 817
280
3,200
2,362
650
440
57
2,941f (5,145)ft
156 (250)
1,485
1,990
18,378 (20,676)
442f
330
195
295f
3
34
1,299
$19,677 ($21,975)
19. 7c (22. OC)
$ 2,966
280
2,042
1,963
650
376
57
2,505f (4,385)tt
149 (239)
1,429
1,990
14,407 (16,377)
275f
220
121
250*
3
29
898
$15,305 ($17,275)
15. 3C (17. 3C)
$ 2,270
260
1,591
1,953
600
344
57
l.Se?* (3,266)tf
137 (218)
1,376
1,990
12,445 (13,925)
280f
220
93
1861"
3
23
805
$13,250 ($14,730)
13. 3C (14. 7C)
Based upon Ref. 2. All costs (except gasoline, oil and their taxes) are adjusted to 1973 dollars using
a 10 percent per year inflation factor.
**
From Ref. 9.
Gasoline assumed to cost 50 cents per gallon, including 10 cents tax, in 1973 dollars; oil costs are
adjusted accordingly.
Gasoline assumed to cost 80 cents per gallon, including 10 cents tax, in 1973 dollars; oil costs are
adjusted accordingly.
9-7
-------
projected long-term price of gasoline (including tax) should be about 50
cents per gallon, corresponding to a long-run price of crude at $6.50.
The short-term price of gasoline is far more uncertain with politics
influencing the supply of oil and with the price elasticity of demand some-
what uncertain. Various studies done at Harvard and by the Federal Energy
Office have predicted a possible short run upper bound price of 75 to 80
8
cents per gallon.
Accordingly, the internal combustion engine auto cost was computed
*
for two levels of gasoline cost. Operating costs are calculated for cars
using gasoline at 50 and 80 cents per gallon (this includes 6 cents state
and 4 cents federal gas tax). This 60 percent variance in the price of
gasoline resulted in only a 12 percent variance of the per mile cost of
the standard car.
Table 2.2 summarizes the cost per mile for internal combustion en-
gine cars. The parameters of 10,000 miles a year for 10 years, while taken
from the referenced DOT paper, are consistent with figures developed in
Ref. 10 for Los Angeles. Table 2.2 also has an average cost per mile based
on the auto population by class as projected in Ref. 10. The decreasing
cost from 1980 to 1990 to 2000 is due to the increasing percentage of sub-
compacts (which are less expensive to operate).
In calculating fuel costs, the fuel economy figures from the referenced
DOT paper were used without modification, as shown in Table 2.1. If instead
the reduced fuel consumption rates projected in Ref. 10 were used, the total
costs per mile (for gasoline at 50 cents per gallon) for the average car in
Table 2.2 would be reduced 0.2 cents in 1990 and 0.5 cents in 2000.
Oil was scaled to the price of gasoline.
9-8
-------
TABLE 2.2
SUMMARY OF INTERNAL COMBUSTION VEHICLE OPERATING COST PER
MILE AT 10,000 MILES/YEAR FOR 10 YEARS IN 1973 DOLLARS
** t
Cents per Mile Cents per Mile
Standard 19.7 22.0
Compact 15.3 17.3
Subcompact 13.3 14.7
*
Weighted Average
1980 . 15.9 18.0
1990 15.5 17.5
2000 15.4 17.4
*
Based on SCAB auto population share by class as shown
in Ref. 10 with intermediate and specialty assumed to
cost midway between standard and compact.
**
Gas at 50 cents per gallon
Gas at 80 cents per gallon
2.2 ELECTRIC VEHICLE COSTS
Electric vehicle costs are derived from a variety of sources and
will be explained below by cost item. The cost items are similar to those
in Table 2.1 with four exceptions. Battery depreciation is added to the
amortization account. Batteries are a substantial recurring portion of
electric car costs. The cost of electricity is substituted for gasoline
and oil costs. Pollution control costs are omitted and a road use tax is
retained. As before, costs are given in 1973 dollars at retail prices.
Tables 2.3 and 2.4 show the results of the electric car cost calcu-
lations. Table 2.3 shows costs based upon travel of 6,300 miles per
vehicle per year because Task Report 10 shows that initial penetration of
electric vehicles will be in a market characterized by lower annual mile-
age. DOT expected a 10-year life for the ICE car and Leeds of England and
others experienced a 15-year electric vehicle life. The 12-year life span
was chosen as a compromise.
9-9
-------
TABLE 2.3
FOUR-PASSENGER ELECTRIC CAR OPERATION COST WITH THE
LEAD-ACID BATTERY FOR 12 YEARS IN 1973 DOLLARS
Costs Except Taxes
Amortization
Vehicle (without battery)
Battery and Replacements
Finance Costs
Repairs and Maintenance
Replacement Tires
Accessories
Electricity
Insurance
Parking, Tolls, etc.
Subtotal
Taxes
State
Registration
Title
Federal Tax on Tires
Road Use Tax*
.Subtotal
Total
Cents Per Mile
Best Battery
Performance
at 6,300 mi/yr
S 2,977
3,592
3,805
755
325
69
1,134
1,782
2,388
16,827
264
93
25
525
907
$17,734
23. 5C
Worst Battery
Performance
at 6,300 mi/yr
$ 2,977
7,518
4,005
755
325
69
1,134
1,782
2,388
20,953
264
93
25
525
907
$21,860
28. 9C
Best Battery
Performance
at 10,000 mi/yr
$ 2,977
4,200
3,009
1,200
527
69
1,800
1,782
2,388
17,952
264
93
41
850
1,248
$19,200
16. Oc
Worst Battery
Performance
at 10,000. mi/yr
$ 2,977
10,982
3,225
1,200
527
69
1,800
1,782
2,388
24,950
264
93
41
850
1,248
$26,198
21. 8c
Imposed as a surrogate for gas taxes.
9-10
-------
TABLE 2.4
FOUR-PASSENGER ELECTRIC CAR OPERATION AT 10,000 MILES PER YEAR
FOR 12 YEARS IN 1973 DOLLARS
Costs, Except taxes
Amortization
Vehicle (without battery)
Battery and Replacements
Finance Costs
Repairs and Maintenance
Replacement Tires
Accessories
Electricity
Insurance
Parking, Tolls, etc.
Subtotal
Taxes
State
Registration
Title
Federal Tax on Tires
*
Road Use Tax
Subtotal
Total
Cents Per Mile
Nickel-Zinc
Battery
$ 2,945
6,125
4,462
900
527
69
1,163
1,782
2,388
20,361
264
93
41
850
1,248
$21,609
18. Oc
Zinc-Chlorine
Battery
$ 2,891
993
2,640
900
451
69
935
1,782
2,388
13,049
264
93
34
850
1,241
$14,290
11. 9c
Best Lithium-Sulfur
Battery Performance
$ 2,795
1,440
2,580
900
451
69
1,026
1,782
2,388
13,431
264
93
34
850
1,241 :
$14,672
12. 2c
Worst Lithium-Sulfur
Battery Performance
$ 2,795
2,400
2,628
900
451
69
1,026
1,782
2,388
14,439
264
93
34
850
1,241
$15,680
13. 1C
Imposed as a surrogate for gas taxes.
9-11
-------
2.2.1 Amortization of Car Cost
The initial cost of the vehicle is calculated using data from a
12
Ford Motor Company report, Electric Vehicle System Study. . This report
was utilized because it was the only report available which estimates
costs and prices of electric vehicle systems for a large quantity of units.
The following 1973 price data for a 45-hp electric car was derived from
the Ford report:
Basic Car (without battery) $1,880
Motor and DC Chopper $ 725
That electric car weighs 2,067 pounds (exclusive of batteries, motor, and
chopper) and thus the basic car (without batteries, motor, and chopper)
costs 91 cents per pound.
This cost figure of 91 cents per pound was applied to the four-
passenger electric car of this study to derive costs of $2,252, $2,220,
$2,165, and $2,070 (exclusive of batteries, motor, and chopper) for the
basic lead-acid, nickel-zinc, zinc-chlorine, and lithium-sulfur cars,
respectively. It is reasonable that the basic car cost should vary with
battery weight as the car requires less structural material to support a
lesser battery weight. To this basic cost was added $725 for motor and
chopper. The rationale underlying this was that the drive train for the
four different cars would reflect weight to some degree but that the cost
of the rest of the car would vary with the weight.
2.2.2 Amortization of Battery Cost
The cost of the lead-acid battery requirements for the vehicle life-
time is calculated using the battery performance cruves in Task Report 13.
The curves show best and poorest battery life in years as a function of
available range used. Maximum range is listed as 54 miles based upon the
SAE metropolitan driving cycle. Percent utilization is calculated by di-
viding average daily mileage by 54. The curve gave the following battery
life figures:
9-12
-------
At 6,300 miles per year:
Best: 4.0 years
Poorest: 1.9 years
At 10,000 miles per year:
Best: 3.4 years
Poorest: 1.3 years
The total battery cost was then calculated for 12 years using 80 cents
per pound (or $1,200) for each pack. This figure was derived from current
SLI lead-acid battery retail prices.
The price of the lead-acid battery assumes the turn-in of a battery
of the same size. The valuation of this turn-in is usually about 10 per-
cent of the replacement lead-acid battery cost. Since the initial pur-
chaser of an electric vehicle will have no batteries to turn in, a one-
time assessment will probably be required. If this assessment amounts to
10 percent of the cost of the first lead-acid battery pack, it would add
$120 (or 0.1 cents per mile) to the cost of the vehicle.
Conversely, the cost of the other three electric car battery types
do not include a turn-in allowance, since the assignment of some turn-in
value is based upon many uncertain variables (e.g., demand and cost of
handling and refining) not yet available. This omission is probably most
critical in the nickel-zinc battery, where both the cost and quantity of
the recyclable raw materials and the cost of the battery are high. As-
suming some upper limit of turn-in value, say 25 percent of purchase price,
and applying this factor to the purchase of the second and succeeding
battery packs, it is possible to reduce the vehicle cost per mile by 1.0
cents. A more likely turn-in value is about 10 to 15 percent, which would,
of course, have a lesser impact on per-mile costs.
Table 2.4 shows the most likely costs for the future batteries. An
original Ni-Zn battery cost of $2940 is estimated using data from Task
Report 1 showing a cost of $60 per KWH and a 49 KWH capacity. Assuming
9-13
-------
400 deep discharge cycles and a 144-mile range, the mileage available
*
from the battery is 57,600 miles. The $2,940 cost is then adjusted for
a 12-year, 10,000-mile-per-year car. The zinc-chlorine battery is esti-
mated to cost $600 and have a capacity of 500 deep discharge cycles and
a range of 145 miles. Thus the mileage available is 72,500 miles and the
cost is adjusted upward to provide 120,000 miles of total battery life.
The battery costs for the lithium-sulfur battery are more easily cal-
culated. In Task Report 1 it is estimated that the battery will have a
life of 3 to 5 years. The constraint here is years rather than cycles as
total hot life is estimated to be the lifetime determinant. The lithium-
sulfur battery is assumed to last from 3 to 5 years. The cost per pound
is estimated in Task Report 1 to be $2, or $600 for the pack, once suffi-
**
cient production levels are achieved.
2.2.3 Finance Costs
The finance costs for the electric car (without battery) is derived
using the same formula used in deriving ICE car finance costs. The time
period (n) is 12 years rather than 10, however. The calculations used to
derive the finance cost of the battery are based upon the same formula but
the time period (n) is the expected lifetime of the battery rather than of
the car. The finance cost for one battery is converted to a mileage cost
based upon the expected lifetime of the battery. Finance costs are derived
by applying these mileage costs to the total number of miles expected over
the lifetime of the car.
Obviously, 144 miles is more than the average 29 miles per day which
would be driven to accumulate 10,000 miles a year. However, at less
than 144 miles the battery discharge would be proportionately shallower
and more discharges would be obtained during the life of the battery.
**
All battery cost factors are based upon production levels sufficient
to obtain economics of learning and of scale.
9-14
-------
2.2.4 Repairs and Maintenance
When compared to the internal combustion engine vehicle, the elec-
tric vehicle is projected to require fewer repairs and maintenance actions.
The engine, with all its encumbrances, is replaced by an electric motor,
batteries, and controls. Repair and maintenance costs for the electric
car are estimated on a per mile basis using data from the same Ford
12
report, from DOT costs, and from maintenance data in Sec. 3.2 which
develops the percentage of repair activity that is directly related to
the internal combustion engine.
The total cost is 1.0 cents per mile for the lead-acid battery car
and 0.75 cents for all other battery types. This is allocated on the
basis of 0.2 cents per mile for motor, commutator, and controls, 0.25
cents per mile for lead-acid battery maintenance, and 0.55 cents per mile
for all other maintenance. The 0.55 cents figure was derived from the DOT
maintenance figures, through application of the factor developed in Sec.
3.2 showing that 28 percent of vehicle repairs are germane to the basic
vehicle (mostly tire and suspension related) rather than to the internal
combustion engine propulsion system. This percentage is applied to the
DOT maintenance cost figure of 1.9 cents per mile to derive the electric
car general maintenance cost of 0.55 cents per mile.
At 10,000 miles a year, this is $20 per year for maintenance and
parts for the motor, commutator and controls, $25 per year for lead-acid
battery maintenance, and $55 per year for basic car maintenance. The
brushes in the electric motor should be checked about once every year
and replaced if excessive wear is indicated. This is a relatively minor
task and should require no more than a half hour of labor. The controls
are all electric and should not require any maintenance, though it should
be inspected occasionally. Twenty dollars a year should be adequate for
this activity. Lead-acid battery maintenance consists of adding water to
the cells and cleaning the terminals. The $25 allows ten checkups a year,,
assuming that the 72 or so cells can be checked and the terminals cleaned
9-15
-------
in half an hour at $5 an hour. A lower hourly rate is used here as it
is assumed this work can be done by lesser skilled service station at-
tendants. The $55 would go for other miscellaneous general maintenance
such as lubrication of the suspension, alignment and balancing of the
tires, brakes, light adjustment, body repair, transmission, shocks, and
steering.
The other battery types are not expected to have sufficient main-
tenance requirements to justify any cost. The nickel-zinc battery will
require only the semi-annual addition of some liquid. Energy Development
Associates is designing their zinc-chlorine battery so that maintenance
requires only the infrequent changing of cannisters designed to trap exit
gases and add electrolytes. They predict that this can be done simply by
the vehicle operator every 2 to 3 months. The lithium-sulfur battery must
be maintained at 300°C-400°C (internal temperature) and no battery maintenance
is possible as a drop to any temperature where repairs could be enacted
would permanently disable the battery.
2.2.5 Replacement Tires and Accessories
Replacement tires are costed on a per-mile basis using the DOT figures
for the internal combusion engine vehicle. The replacement tire cost for
the lead-acid battery car at 10,000 miles a year for 12 years is 1.2 times
the tire cost for the standard car (chosen because of the weight similarity
to the lead-acid car), while the cost for the lead-acid battery car used
only 6,300 miles a year is 0.75 times the DOT standard car tire costs. The
zinc-chlorine and lithium-sulfur battery car tire replacement cost is 1.2
times the compact car tire cost. Since accessories were accrued on a year-
ly basis, their cost is simply 1.2 times the DOT cost for subcompact acces-
sories.
*
Labor-saving devices, such as a filler which requires the addition of water
in only one orifice, are being developed but are still 3 to 6 years from
production. A maintenance-free battery with high frequency discharge cap-
ability is also in the development stages but is 5 to 10 years from produc-
tion. The nickel-zinc battery will require the addition of liquid to only
one orifice.
9-16
-------
2.2.6 Electricity
Task Report 1 predicts that the four-passenger lead-acid vehicle
will require 0.79 KWH per mile; the nickel-zinc, 0.51 KWH per mile; the
zinc-chlorine, 0.41 KWH per mile; and the lithium-sulfur vehicle will
travel 0.45 KWH per mile. This estimate accounts for the inefficiencies
of the mechanical components, motor, batteries, and charger, so this is
an estimate of demand at the meter. Accessories have been assumed to be
zero for the electric car in the SCAB.
Electric rates are structured so that beyond certain increments of
electricity used, the marginal rate is less than the average rate. That
is, the more electricity consumed, the lower the cost per KWH. According
to the Edison Electric Institute, average residential use in 1972 was 640
KWH a month. A lead-acid electric car getting 0.79 KWH per mile uses
about 660 KWH per month. Cars using the other battery types require from
340 to 425 KWH per month. In 1973, PG&E charged residential users 1.9
cents per. KWH between 600 and 1200 KWH. Most KWHs demanded for electric
car use will fall into a range of the rate structure where power demanded
is priced at 1.9 cents per KWH. The cost for electricity is then 1.5
cents per mile for the lead-acid battery car, 0.97 cents per mile for the
nickel-zinc battery car,. 0.78 cents per mile for the zinc-chlorine battery
car, and 0.86 cents per mile for the lithium-sulfur battery car.
2.2.7 Insurance, Parking, and Taxes
The cost of insurance, state registration, garaging, parking, tolls,
etc., are those DOT cost categories adjusted for the difference in years
of operation. Insurance costs, as well as state title cost, is based on
the DOT subcompact costs to reflect the lower rates accruing to smaller
cars. To calculate the tire tax, the DOT tire tax rate is applied to the
electric car tire costs.
9-17
-------
2.2.8 Road Use Tax
Historically, the costs of roadways have been borne by the user
through gas tax revenues. If electric vehicles are implemented, that
mode of allocating costs will no longer be possible. We assume that the
public policy of user-borne costs will continue and accordingly have added
a road use tax to the cost of the electric car. This cost was calculated
using the rate of 10 cents per gallon (6 cents state, 4 cents federal)
given the figures for average gas mileage and auto class as a percent of
auto population derived in Ref. 10. This is calculated as $71 per year for
the 1980 car driving 10,000 miles per year at an average of 14.1 miles per
gallon.
Figures 2.2, 2.3 and 2.4 show the cost of the electric vehicles as
a function of mileage. Also plotted on the curves are the projected costs
for the internal combustion engine car for a 10-year life at 10,000 miles
per year. The one upward-bending curve in Fig. 2.2 is a result of rapidly
decreasing battery life as mileage and depth of discharge are increased.
The percent change of life-cycle costs shown in Table 2.5 are based
upon costs from Tables 2.2 through 2.4. The percent impact of conversion
to subcompact internal combustion engine cars is included because it is
expected that conversion to electric cars will initially be in that car
class. The percent change to the average internal combustion engine car
shows the average impact should total conversion occur. In summary, the
consumer who uses an electric car would spend from 25 percent less (zinc-
chlorine battery at 11.4 cents per mile) to 37 percent more (worst lead-
acid at 21.8 cents per mile) on his personal outlays for autos and auto-
related items.
For the past 10 years, the amount spent on autos and auto-related
items has averaged a fairly constant 12.4 percent of total personal ex-
14
penditures. Assuming elasticity of demand for private transportation
is zero, that is, a change in the price does not change the quantity
9-18
-------
40 r-
30
OJ
Q.
OO
o
o
10
20
kWORST BATTERY PERFORMANCE
BEST DATTERY PERFORMANCE
AVERAGE 1980 INTERNAlVCOMBUSTION ENGINE 0 50 50
-------
30
20
s_
(U
Q.
I/O
O
O
10
WORST BATTERY PERFORMANCE
AVERAGE 2000 INTERNAL • COMBUSTION ENGINE @ 50
-------
TABLE 2.5
SUMMARY OF LIFE-CYCLE COSTS
Subcompact Internal Combustion Engine Car (1973) 13,3c
Average Internal Combustion Engine Car (1980) 15.9
-------
demanded, the substitution of electric cars would adjust this expenditure
by the percentage change in operating costs. That is, the percentage of
personal outlays spent on autos could increase 37 to 17.1 percent of total
personal expenditures with the worst lead-acid battery car or decrease 26
to 9.3 percent of total personal expenditures with the zinc-chlorine bat-
tery car.
9-22
-------
3 IMPACTS ON TRANSPORTATION SUPPLIES AND SERVICES
This section discusses the variations in demand for goods and ser-
vices induced by electric car usage. Task Report 4 identified the industry
sectors that were susceptible to the impacts of changing demands placed
upon the system by the manufacture and use of electric cars. This section
presents the sectors by level of impact (major, moderate, or minor) and
then identifies the nature and level of the impacts.
3.1 MAJOR IMPACTS
Two sectors are identified as sustaining major impacts. These are
battery manufacturing, and petroleum distribution. The impacts on each of
these are discussed below.
3.1.1 Battery Manufacturing
*
Starting, lighting, and ignition (SLI) battery manufacturing is
characterized by a half-dozen major firms and numerous smaller ones. The
manufacturing units themselves are small and close to their market. This
situation results because (1) SLI battery products are fairly homogeneous;
that is, the products and manufacturing processes are similar for most
brands, (2) there are no significant economies of scale beyond production
**
of 1,000-2,000 units a day, and (3) it is easier to transport lead than
the finished battery. Due to similarities between the SLI lead-acid bat-
tery and the electric vehicle lead-acid battery, it is reasonable to pre-
dict that electric vehicle battery manufacturing requirements will be met
by an increased number of manufacturing units within the area.
To calculate the extent of the impact in the South Coast Air Basin,
the number and lifetime of batteries required are computed. The average
weight of the two lead-acid electric-vehicle batteries (Delco and ESB)
*
This type of lead-acid battery is used in the internal combustion engine
automobile.
**
Telephone discussions with representatives at the Lead Industry Associa-
tion and ESB Corporation.
9-23
-------
considered in Task Report 1 is 57 pounds, thus a vehicle requiring 1,500
pounds of battery would require the equivalent of 26 of these batteries.
Due to the homogeneity and automation of the lead-acid battery manufac-
turing, and the similarity of the SLI and electric-vehicle batteries,
the manufacture of electric-vehicle batteries should require about the
same capital investment and labor per unit as the SLI battery.
These requirements are estimated by the industry spokesmen to be $1
million and 75 employees for a capacity of the equivalent of 1,000 SLI
batteries per day. These amounts scale directly with production increases.
The lifetime is estimated in Sec. 2 to range from 1.3 to 3.4 years,
while a SLI lead-acid battery lasts about 3 years. Assuming 250 working
days a year, 250,000 SLI batteries are produced by each 75 employees.
The gross impact in Table 3.1 shows the employees who would be displaced
if SLI batteries were no longer produced. This is not equal to the em-
ployment in battery manufacturing (SIC 3691) as this category includes
manufacturing of all types of storage batteries. With the equivalent of
26 SLI batteries required for each car, from 23 to 60 times as many bat-
teries are required yearly for each electric car, varying with the expected
life of the batteries. Table 3.1 gives the impacts on economic activity
in battery manufacturing assuming 100 percent substitution of electric cars
for internal combustion engine cars. The increase from the baseline is
linear for lead-acid batteries and any intermediate impact levels can
easily be determined. Using the same methodology, the increased require-
ments for capital are computed to range from $175 to $475 million, the
range again varying with battery life assumptions. This compares to a
current investment of about $8 million in battery manufacturing in the
SCAB.
The requirements of the other three battery types (nickel-zinc,
zinc-chlorine, and lithium-sulfur) examined in this paper are also pre-
sented in Table 3.1. They are treated in the same manner; their production
9-24
-------
TABLE 3.1
IMPACTS ON BATTERY MANUFACTURING (SIC 3691) WITH 100 PERCENT ELECTRIC CAR USAGE IN THE SCAB
Baseline
Gross Impact from the Loss
of SLI Battery Requirements
Requirements for Best Lead-
Acid Battery
Requirements for Worst Lead-
Acid Battery
Requirements for Nickel-Zinc
Battery
Requirements for Zinc-
Chlorine Battery
Requirements for Best
Lithium- Sulfur Battery
Requirements for Worst
Lithium-Sulfur Battery
1980
Employment
2,235
-588
13,510
35,336
*
Payroll
24,100
-6,340
145,678
381,028
1990
Employment
2,895
-673
15,447
40,401
21,310
7,244
*
Payroll
34,300
-7,974
183,016
478,672
252,481
85,827
2000
Employment
3,800
-760
17,439
45,609
23,752
8,178
5,929
9,853
*
Payroll
45,600
-9,120
209,268
547,308
285,027
98,136
71,148
118,236
Thousands of dollars
vo
Ul
-------
requirements are related to the known requirements for lead-acid batteries
so the requirement may be related to existing levels.
In Table 3.1, the gross impact indicates the level of activity which
will be lost due to the displacement of internal combustion engine cars.
The requirements for each of the battery types indicates the level of ac-
tivity required after full implementation of electric cars using that bat-
tery type. The net impact (the absolute change in employment) is simply
the requirements less the gross impact.
The three other battery types under consideration have not been de-
veloped, much less produced in any quantity. Since there is little published
data on the manufacturing process, representatives of the developer of each
of the battery types were contacted in order to obtain some projections re-
lating the manufacturing requirements to the lead-acid battery.
As was the case with the lead-acid battery, only 100 percent substi-
tution of electric cars in the SCAB is shown in Table 3.1; however, require-
ments for various levels can be readily interpolated. The table shows the
change relative to the baseline condition (i.e., internal combustion engine
cars). If between some time frames there is a switch of battery types, the
change can be determined simply by taking the difference in requirements
for the two battery types in the latest time period. For example, if there
is a change from the best lead-acid battery to the nickel-zinc battery from
1990 to 2000, there are potentially 6,313 more manufacturing jobs in 2000.
Since linearity is assumed (past 5 percent), an equal mix of the two bat-
tery types would halve the potential job gains.
The change in impacts is probably not linear until about 1 to 5 per-
cent market penetration has occurred. This is because, as opposed to lead-
acid battery manufacturing, the processes are proprietary, vary from company
to company, and the companies developing these batteries are located in the
Eastern US. Thus it is likely that, unless the SCAB is the first major
9-26
-------
battery market, initial plant development will be elsewhere with plants
being developed in Los Angeles only after the market is sizeable. It is
unlikely that manufacturing would continue elsewhere after the Los Angeles
market is developed for the same reasons that lead-acid batteries are
manufactured locally. It is easier and less costly to transport the ma-
terials than the batteries. Gould estimates that they would build a plant
with a moderate production capacity after slight market penetration in the
area.
Gould estimates that the nickel-zinc battery they are developing
will sustain a wage bill of about $10 per KWH. When adjusted for esti-
mated overhead rates this equates to a payroll of $430 a battery pack or
$75 a year per car (assuming a battery life of 57,000 miles). This figure
is applied to the expected auto population and adjusted by the expected
*
industry salary to derive the employment requirements shown in Table 3.1.
Energy Research estimates that each manufacturing unit will have a larger
capacity and greater capital requirements than lead-acid manufacturing
units. This indicates a minimum capital investment of $90 million with
100 percent substitution in the SCAB.
The zinc-chlorine battery undergoing development by Energy Develop-
ment is expected to require half the labor per KWH of the lead-acid bat-
tery. Since the KWH capacity of the zinc-chlorine car is about twice
that of the lead-acid car, the battery manufacturing requirements per
battery pack are the same. The requirements for the lead-acid battery
are merely adjusted for frequency of battery replacement to derive the
zinc-chlorine battery manufacturing requirements•
* ...
Energy Research, another developer of the nickel-zinc battery, estimates
their requirements as being about equal to lead-acid requirements per KWH.
Adjusting for KWH capacity and frequency of replacement, their manpower
requirements would be the same as those for best lead-acid battery.
9-27
-------
Energy Development estimates that a production line with the capacity
for 1,600 zinc-chlorine batteries a day will require abour $1.4 million.
This is about 10 percent less per battery than lead-acid manufacturing re-
quirements. Thus, assuming replacement of the battery pack every 7.2
years, the capital investment requirements for 100 percent substitution
in the SCAB would be about $75 million dollars. This is less than the
other alternatives as the developer plans to maximize utilization of al-
ready available "off-the shelf" items (e.g., the pumps and refrigerator)
easily obtained and assembled. This will have some third-order impacts;
however, it is not clear that these will be felt in LA, as those items are
/
already produced throughout the country and are easily transported.
The lithium-sulfur battery is expected to require about 2.5 times
more labor per pound than the lead-acid battery. Since the lithium-sulfur
battery pack is 1/5 the weight of the lead-acid battery pack, each pack
will require 1/2 the employment. The requirements depicted in Table 3.1
are adjusted to account for this and for an expected lifetime of 3 to 5
years. The manufacturing process is expected to be more costly, but is
still too uncertain to be the basis for any capital requirement estimates.
There is no lead mining or refining in the SCAB. Thus there is no
impact in the SCAB, nor is there likely to be as there are no known lead
reserves. There is some potential for increased employment requirements
nationwide. As indicated in Task Report 4, about 13,000 persons are em-
ployed nationally in these activities. Another in this series of reports
indicates that if the utilization of the lead-acid battery in LA reaches
100 percent, the requirements for lead will be about four times the other-
wise projected demand for all lead used in the US. This increase is less
than the projected increase in battery manufacturing requirements because
of the use of recycled lead and the assumed lessened use of leaded gasoline.
9-28
-------
3.1.2 Petroleum Distribution
The impacts on service stations and wholesale petroleum distributors
have been topics of concern in dealing with the impacts of electric cars.
This employment comprises 1.3 percent of the region's total employment and
4.1 percent of its business units. Since these businesses primarily sell
gasoline, they are very susceptible to any reduction in the number of
gasoline engine cars. Table 3.2 summarizes the findings of this section.
The impacts on petroleum distribution are calculated by computing
the percentage of total fuel consumption dependent on automobiles. Of the
92 billion gallons of fuel consumed for highway use in California in 1970,
71 percent was for automobiles. The remaining 29 percent was for motor- '
cycles, trucks, buses, and other commercial uses. We assume that replace-
ment by electric cars would reduce the volume of business of petroleum
distributors by 71 percent and that employment would be reduced by a cor-
responding amount.
Linearity of impact is again assumed. Thus for some level of elec-
tric car use less than 100 percent, the negative impact or the added
requirements can be adjusted by the percentage use. Linearity is assumed
because of the large number of small homogeneous units (service stations)
which would be affected. If there were a small number of large firms,
there would be some discontinuous or non-linear impacts. With the average
service station accounting for about 0.015 percent of the market, each
significant change in the market can bring about a change in the number
of stations and employees.
As with previous tables, the change from lead-acid to other batteries
can be found by taking the difference between the total projected activity
of lead-acid batteries and of the other batteries in the latest time period.
The impacts here are the same for all battery types except for some
additional maintenance manpower requirements imposed by the use of the lead-
acid battery.
9-29
-------
VO
I
U)
o
TABLE 3.2
IMPACTS ON PETROLEUM DISTRIBUTION (SIC 5092 and 5541) WITH 100 PERCENT
ELECTRIC CAR USAGE IN THE SCAB
Petroleum Distribution Baseline
Net Impact Petroleum Distribution
Projected Activity Petroleum
Distribution
Service Station Baseline
Net Impact Service Stations
Projected Activity Service
Stations
Total Projected Activity
Percent Loss
Electric Vehicle Requirements
for Service Stations*
*
Net Impacts Service Stations
Projected Activity Service
Stations*
*
Percent Loss
Total Projected Activity
Percent Loss
19
Employment
3,646
-2,589
1,057
39,400
-35,775
3,625
4,682
,89
7,941
-27,834
11,566
71
12,623
71
30
**
Payroll
45,300
-32,163
13,137
182,300
-165,528
16,771
29,908
87
36,742
-128,786
53,514
71
66,651
71
19
Employment
3,366
-2,390
976
43,090
-39,126
3,964
4,940
89
9,060
-30,066
13,024
69
14,000
70
30
**
Payroll
57,900
-41,109
16,789
208,700
-189,500
19,200
35,989
87
43,881
-145,619
63,081
69
79,870
70
20
Employment
3,420
-2,428
992
46,360
-42,095 -
4,265
5,257
89
10,228
-31,867
14,493
67
15,485
69
00
**
Payroll
64,600
-45,866
18,734
237,100
-215,287
21,813
40,547
87
52,310
-162,477
74,223
67
92,957
69
Lead-acid only.
Thousands of dollars.
-------
The row "Electric Car Requirements for Service Stations" (lead-acid
only) in Table 3.2 is derived from an increase in two service areas re-
quired for lead-acid battery car operation. The first is battery main-
tenance. As explained in Sec. 2, there are no significant manpower
requirements for maintenance of other batteries. Lead-acid battery main-
tenance is estimated to take 5 hours a year (see Sec. 2) so that at 2,000
hours a year one man could service 400 vehicles. This manpower requirement
is adjusted based on the assumption that 50 percent of electric car owners
*
would perform their own battery maintenance. While there is no data to
support this assumption, it seems reasonable in view of the relatively
simple activity required. In addition, most SLI battery maintenance is
now performed in conjunction with internal combustion engine tune-ups.
Since this activity will be eliminated, battery maintenance would require
a special stop at the service station, a less likely occurrence. In ad-
dition, service station business will probably increase by an additional
1.5 percent over other electric vehicle battery types, as the expected
frequency of electric vehicle battery sales is about four times greater
for lead-acid batteries than for other batteries. As explained below,
other electric vehicle batteries account for 0.5 percent of service station
business.
The impacts upon service stations is derived from a summary of ser-
vice station sales ratios. These figures are for the SCAB area service
stations who buy bookkeeping and management services from E. K. Williams
and Company. Stations who buy this service include almost all company-
owned stations (about 80 percent of all stations are company owned) and
many independent stations. The data in Table 3.3 shows the percentage
breakout of the dollar value of sales (including sales of labor).
No similar reduction is made in the maintenance cost, as the operation
is a real cost to the consumer whether he has the work done or does the
work personally.
9-31
-------
TABLE 3.3
SERVICE STATION SALES
Percent of Total
Value of Sales
Electric
Vehicle Related
Gas and Oil
Accessories
Labor
Tires
Batteries
Other
Lubrication
82.2
7.9
4.4
2.9
1.0
0.9
0.7
100%
3.0
1.2
2.9
0.5
0.9
0.7
9.2%
The percentage of electric-car-related sales is 9.2 percent, shown
in the right-hand column. This is derived by eliminating the sales which
would disappear with the introduction of electric cars. It is assumed
that the amount of labor required is proportional to the value of goods
and services provided. The amount of accessories sales which are depend-
ent on internal combustion engines and thus eliminated is 62 percent,
excluding tires and batteries. The derivation of this value is discussed
in Sec. 3.2. Also discussed there is the basis for reducing repair-
oriented labor sales by 72 percent to account for the elimination of in-
ternal combustion engines. The reduction of battery sales from 1.0 to
0.5 percent is based on the fact that the three longer range electric
vehicle batteries have a lifetime about twice that of SLI batteries.
While the electric vehicle batteries would be more expensive and increase
the percentage of sales devoted to batteries, we are concerned with the
Adjustments for greater lead-acid battery sales requirements due to
higher frequency of replacement were made previously.
9-32
-------
number of persons who are supported by this activity. Even if the battery
is more expensive, the number of sales persons varies more with the number
of sales. Further rationale for this is presented in Sec. 3.2.
3.2 MODERATE IMPACTS*
The focus in this subsection is on the economic impact of electric
cars on the automotive "aftermarket" and the entire automotive service in-
dustry in the South Coast Air Basin. The aftermarket includes sales of
tires, batteries and accessories (TEA), as well as wholesale/retail sales
of replacement parts. The baseline projection for total employment in
these industry sectors (SIC 5013 and 5531) for the year 1980 is 24,109;
for 1990, 29,017; and for 2000, 33,972. The projection for the automo-
tive service industry and specifically auto repair shops (SIC 7538 and
7539) for 1980 is 10,584; for 1990, 12,793; and for 2000, 15,200. This
represents 25 percent of area employment in auto-related activity and 0.8
percent of total employment.
In order to assess the magnitude of electric car impact on the auto^
motive aftermarket in terms of total sales, number of firms in the industry,
and total employment, we must determine what percentage of automotive
business or "lines" are directly related to the internal combustion engine,
which are relatively independent of type of engine or motor and which lines
would be introduced or expanded in the advent of a significant electric
car population.
*
The analysis in this section is based on total annual sales of automo-
tive lines, frequency of repair data, estimated labor time for repairs
and adjusted recommended maintenance figures. The relative percentages
of internal-combustion-engine-related and non-internal-combustion-
engine-related sales and repairs derived here were applied to the DOT
estimated cost per mile figures in Sec. 2. The nature of the data uti-
lized in this section precludes arriving at maintenance cost per mile
estimates; however, the relative percentages derived are considered
compatible with existing DOT estimates.
Wholesale Tire Distribution (SIC 5014) and Tire Retreading Shops (SIC
7534) will be examined in Sec. 3.3 under Minor Impacts.
The employment projections are detailed in Tables 4.5, 4.6, and 4.7 of
Task Report 4.
9-33
-------
An evaluation of the breakdown of both the automobile aftermarket
and of auto service and repair is aided by the following definition:
• Tune-up and ignition includes spark plugs, points, rotor and
condensers, distributor caps, spark plug wires, and ignition
coils.
• Cooling system includes belts, hoses and clamps, thermostats
and gaskets, radiator caps, anti-freeze chemicals and pumps.
• Filter service includes air, oil and fuel filters, transmis-
sion filters, PCV valves and charcoal cannister elements.
• Battery/Starting/Charging system includes batteries, cables
and terminals, starter motors, generators, alternators and
voltage regulators.
• Tires and wheels includes tires, tire valves, snow tire studs,
tire truing, O..E.-type replacement wheels, custom wheels,
wheel balance and wheel alignment.
• Brake system includes shoes and pads, wheel cylinders and
kits, spring kits, disc and drum machinery, master cylinders
and fluid.
• Steering and suspension includes shock absorbers, ball joints,
idler arms, pitman arms, wheel bearings and grease seals.
• Exhaust system includes exhaust pipes, mufflers, tail pipes,
brackets and gaskets.
• Fuel system includes gasoline tanks, tank straps, vent pipes
and hoses, gasoline gauges, fuel pumps, and carburetors.
• Transmission includes bushings, rods and seals, springs, gas-
kets, bearings, transmission assemblies, shift rods, automatic
transmission assemblies, pipes and valve body assemblies.
*
This definition is consistent with the repair and maintenance items found
in the DOT Repair and Maintenance category (Fig. 2.1) used in Sec. 2.
9-34
-------
• Body and paint includes cowl and dash assemblies, hardware,
paints, and body panels.
• Engine includes pistons and piston rings, valves, valve
springs, short block, cylinder head gasket, and bearings.
The dollar value and percent of the market of individual sales items
in the US automotive aftermarket for 1972 is shown in Table 3.4. Utilizing
the percentages from Table 3.4, we attempt a further breakdown of automo-
tive lines into those which are directly related to the internal combus-
tion engine and those for which future consumer demand would presumably
be unaffected even at maximum utilization of electric cars. The break-
down detailed in Table 3.5 was defined by assessing relative dependence
of each line on the conventional internal combustion engine. In cases
where there was no clearcut division, an approximate percentage was esti-
mated in consultation with those knowledgeable in the automotive field.
From this breakdown, we conclude that 44 percent of total sales of auto-
motive lines are directly related to the internal combustion engine. The
remaining 56 percent will be unaffected by changes in the means of power-
ing the car.
Before discussing the implications of these observations on total
employment and number of firms in the industry, a similar analysis of the
auto repair industry will enable us to consider the impacts on both sec-
tors simultaneously since the similarities are apparent.
Table 3.6 details the most frequently encountered automobile repairs
performed by independent garages, service stations, car dealers, and those.
repairs done by car owners themselves. Due to the lack of readily avail-
able data on auto repair, as pointed out by the reluctant response to
9-35
-------
TABLE 3.4
US RETAIL SALES OF AUTOMOTIVE LINES FOR 1972
Line
Tires
Batteries
Spark Plugs
Mufflers, Pipes, etc.
Filters (All Types)
Chemicals
Brake Lining and Lined Shoes
Shock Absorbers
Ignition Parts
Fan Belts, Radiator and Heater Hose
Paint and Body Supplies
Remanufactured Units
Motor and Chassis Parts
Anti-Freeze/Coolant
Gasket and Oil Seals
Clutch Assemblies, Parts, etc.
Brake Parts
Front End Parts
Engines (Small) and Parts
Wipers and Blades
Equipment (All Types)
Lamp Bulbs, Flashers and Sealed Beams
Bearings (Anti-Friction)
Tools (Small Hand)
Wire and Cable Products
Grease and Oil
Carburetors and Parts
Bearings (Motor)
Fuel Pump and Parts
Piston Rings
Thermostats
Brass Fittings and Fuel Lines
Wheels and Rims
Tools (Power)
Automatic Transmission and Parts
All Other Lines
Total
SOURCE: Motor Age, January 1974, p.
Sales
$4,700,000,000
1,210,000,000
840,000,000
774,000,000
618,000,000
609,000,000
527,000,000
420,000,000
403,000,000
346,000,000
288,000,000
264,000,000
214,000,000
206,000,000
156,000,000
148,000,000
140,000,000
132,000,000
130,000,000
124,000,000
121,000,000
115,000,000
107,000,000
106,000,000
99,000,000
91,000,000
82,000,000
81,000,000
80,000,000
75,000,000
68,000,000
46,000,000
44,000,000
42,000,000
41,000,000
698,000,000
$14,145,000,000
40-45.
Percent of
Total Sales
33.2
8.6
5.9
5.5
4.6
4.3
3.7
3.0
2.8
2.4
2.0
1.9
1.5
1.5
1.1
1.0
1.0
0.9
0.9
0.9
0.9
0.8
0.8
0.8
0.7
0.6
0.6
0.6
0.6
0.5
0.5
0.3
0.3
0.3
0.3
4.9
99.6*
Does not add to 100% due to rounding error.
9-36
-------
TABLE 3.5
TOP SELLING AUTOMOTIVE LINES
Lines Related to
Internal Combustion Engines
Percent Share
of Market
Lines Not Related to
Internal Combustion Engines
Percent Share
of Market
Batteries
Spark Plugs
Mufflers, Pipes, etc.
Filters
Chemicals (50%)
Ignition Parts
Fan Belts, Radiator and Heater Hose
Remanufactured Units
Motor Parts
Anti-Freeze/Coolant
Gasket and Oil Seals (60%)
Clutch Assemblies, Farts, etc.
Engines (Small) and Parts
Grease and Oil (80%)
Carburetors and Parts
Bearings (Motor)
Fuel Pumps and Parts
Piston Rings
Thermostats
Brass Fittings and Fuel Lines
All Other Lines (50%)
**
Total
8.6
5.9
5.5
4.3
2.15
2.8
2.4
1.9
0.7
1.5
0.66
1.0
0.9
0.48
0.6
0.6
0.6
0.5
0.5
0.3
2.45
44.34
Tires
Chemicals (50%)
Brake Lining and Lined Shoes
Shock Absorbers
Paint and Body Supplies
Chassis Parts
Gasket and Oil Seals (40%)
Brake Parts
Front End Parts
Wipers and Blades
Equipment
Lamp Bulbs, Flashers and Sealed Beams
Bearings (Anti-Friction)
Tools (Small Hand)
Wire and Cable Products
Grease and Oil (20%)
Wheels and Rims
Tools (Power)
Automatic Transmission and Parts
All Other Lines (50%)
**
Total
33.2
2.15
3.7
3.0
2.0
0.7
0.44
1.0
0.9
0.9
0.9
0.8
0.8
0.8
0.7
0.6
0.3
0.3
0.3
2.45
55.94
SOURCE: Derived from Table 3.4. Relative percentages were determined in consultation with industry
representatives.
Starting-lighting-ignition (SLI) batteries are included with the internal-combustion-engine-related lines
in conformance with the structure of the analysis. An initial gross impact reflecting losses due to the
internal combustion engine will be contrasted with the net impact of the electric car with each of the
four proposed electric vehicle batteries. The proposed lead-acid electric vehicle battery will also be
manufactured to different specifications in terms of size, weight, and expected life and the cost per
battery will be significantly higher.
Does not add to 100% due to rounding.
9-37
-------
TABLE 3.6
AUTO REPAIRS AND SERVICE
*
Tune-Up
*
Periodic Maintenance
Brakes Relined
*
Exhaust System
Wheels Balanced
Front Wheels Aligned
Headlight Beams Aimed
*
Cooling System
Carburetor Overhauled or Replaced
Generator Repaired
Voltage Regulator Adjusted or
Replaced*
Body Repair
*
Fuel Pump Replaced
Automatic Transmission Repaired
*
Major Engine Overhaul
*
Fuel Pump Repaired
Shock Absorbers Replaced
*
Standard Transmission Repaired
Power Steering Repaired
Standard Steering Repaired
Total
Percent Cars
Serviced/Year
66
160
20
20
16
16
11
11
7
7
7
7
6
5
5
4
3
2
1
1
375
Labor
Time
2.1
0.5
2.0
0.5
0.8
1.2
0.3
0.35
1.2
1.1
0.45
0.4
0.35
0.8
14.5
0.35
1.0
0.7
0.5
0.9
Adjusted
Percentage
33.4
19.2
9.6
2.4
3.1
4.6
0.7
0.9
2.0
1.9
0.8
"0.7
0.5
0.9
17.5
0.3
0.7
0.3
0.1
0.2
**
98.8
SOURCE: Automotive Service Digest, Ref. 21, various Auto Repair Associa-
tions, and other industry sources.
**
Repairs directly related to internal combustion engine automobiles with
the exception of Periodic Maintenance which is approximately 60% inter-
nal-combustion-engine-related and 40% non-engine-related.
Does not add to 100% due to rounding.
9-38
-------
*
recent federal legislation, a number of sources were used to estimate
the percentage of cars serviced per year. The following information was
18
included: data on frequency of repair; recommended replacement inter-
19
vals for selected auto.components; frequency and price of three major
20
engine related services performed routinely by garages; and recommended
21
periodic service and maintenance for average driving. In the case of
maintenance and service, the recommended intervals were adjusted to more
18
accurately reflect the longer intervals reported in the frequency data.
Frequency alone does not account for the actual percentage of a re-
pair shop's business dependent on different categories of repairs. To
address this problem, the percentage of cars serviced per year was weighted
21
by the estimated labor time, in hours, required to perform the repair.
The adjusted percentage reflects the relative composition of an auto repair
shop's business.
The average number of repairs or service items performed per car per
year is 3.8. If more than one repair is performed in a given trip to the
garage, the weighting factor may be overstated since the time necessary to
perform a second repair decreases if the two repairs are related to the
same system. Since we have no accurate way of assessing how often in fact
this happens, we merely alert the reader to the consideration.
On the basis of the adjusted percentages in Table 3.6, we are able
to assess internal combustion engine related repairs as 72 percent of the
auto repair business and those repairs not dependent on the internal com-
bustion engine as 28 percent.
*
Title II of the Motor Vehicle Information and Cost Savings Act of 1972
requires that the National Highway Traffic Safety Administration (NHTSA)
develop ways of finding answers to questions regarding the cost of diag-
nosis and repair of mechanical or electrical malfunctions and body re-
pair; and issue regulations on how this information shall be distributed
to consumers by February 1975.
9-39
-------
Comparing the auto supply sector with the auto repair sector, we
observe that the percentage of sales related to the internal combustion
engine is 44 percent; repairs, 72 percent. This difference is partially
explained by the fact that 41 percent of auto supply sales are tires and
batteries. Since installation of these items is included in the sales
sector, they are not reflected in Table 3.6 as repairs. A recalculation
of relative percentages of total sales of automotive lines excluding tires
and batteries shows that 62 percent of sales are related to the internal
combustion engine and 38 percent are non-related.
3.2.1 Impact on the Automotive Aftermarket
The specific impact of maximum utilization of electric cars on the
automotive aftermarket will be reflected mainly in the 44 percent of total
sales related to internal combustion engines and detailed in Table 3.4.
The gross impact on employment and number of firms in the industry evaluated
in terms of the total replacement of the internal combustion engine is es-
timated at a loss of 10,608 jobs in 1980; an additional 678 or a total of
11,286 jobs in 1990; and 3,662 additional jobs in 2000, or a total loss of
*
14,948. The number of firms can be expected to decrease by approximately
44 percent. We would expect the smaller ones to be the most vulnerable
because of their comparative lack of flexibility and diversity.
3.2.2 Impact on the Auto Repair Industry
The impact on the auto repair industry will be even greater since
72 percent of repair activity is related to the internal combustion engine.
Total employment in the industry would be expected to decrease from 10,584
to 2,964 in 1980, a decrease of 7,620 jobs. In 1990 an additional 1,590
jobs would disappear for a total loss of 9,210. An additional 1,734 jobs
would be lost in 2000 for a total decrease of 10,944 jobs at maximum uti-
lization. The total number of firms could be expected to decrease by 72
percent and those remaining would be faced by significant inventory changes.
*
Calculated on the basis of projections from Task Report 4.
9-40
-------
3.2.3 Impacts on Employment Due to Proposed Batteries
The net impact on employment of the use of electric cars is assessed
by estimating the increases in employment in the automotive aftermarket and
auto repair industry due to each of the four proposed batteries. By sub-
tracting this increase from the employment loss determined through the
replacement of the internal combustion engine, we are able to arrive at an
estimated net loss for each of the four battery operated cars. The assump-
tion implicit in the analysis is that employment per sale of battery will
remain constant. There is the possibility of some differentiation in brand
for any given battery design. It is also reasonable to assume that the
markup on the given electric-vehicle batteries will be a lesser percentage
of the total cost than the markup on SLI batteries. Since the electric-
vehicle batteries are much more expensive, the potential profit to the
dealer is higher even with a lesser percentage markup. These excess pro-
fits will encourage entry into the market by new firms. Selling costs
will be driven up by the increased advertising expenses and additional re-
quired manpower in sales. In the worst case, employment in battery sales,
which is now 8.6 percent of total auto supply sales, would increase on the
average of say 50 percent per battery for an additional impact on battery
sales employment of 4.3 percent, bringing the total to 12.9 percent of auto
supply sales. Thus the impact on area employment would be an increase of
0.08 percent, which is considered minimal.
In assessing the impact of the lead-acid electric-vehicle battery,
we consider both the "best case" and the "worst case." The best case has
a life expectancy of 3.4 years compared to the current SLI battery life
of 3 years. Since the best-case electric-vehicle lead-acid battery needs
to be replaced approximately 12 percent less frequently than the current
SLI battery, the decrease in total number of sales is expected to be ap-
proximately 12 percent. The worst case has a life expectancy of 1.3 years
and total number of sales may be forecast to increase 2.3 times when com-
pared with the present internal combustion engine battery.
9-41
-------
SLI battery sales are responsible for 8.6 percent of total after-
market sales, or an estimated 2074 jobs in 1980. The decrease in total
employment in Auto Supply Stores and Wholesale Parts Distribution (SIC
5531 and 5013) due to reduced SLI battery sales may be expected to be
about 12 percent of those jobs originally forecast, or 249 jobs in 1980;
an additional 49 jobs may be lost in 1990 for a total decrease of 299
from the forecast 2495; an additional 51 jobs are forecast to be lost in
the year 2000, bringing the total loss of employment in battery sales to
350 out of the 2,922 projected. These estimates reflect 100 percent uti-
lization of electric cars. The number of jobs lost is expected to decrease
linearly at lesser percentages of utilization.
The "worst case" lead-acid electric-vehicle battery can be expected
to add 2,697 jobs or increase total employment in battery sales to a total
of 4,770 jobs in 1980 since the number of battery sales will increase by a
factor of 2.3. An additional increase of 968 jobs may be expected in 1990
for a total of 5,738 and 982 more jobs are forecast for 2000, increasing .
the 2,922 forecast to 6,720 at 100 percent utilization of electric cars.
Again, we point out the expected linear relationship between the percent-
age of electric car use and the increase in the number of jobs.
Since the new lead-acid electric-vehicle battery pack will be ap-
proximately 60 times heavier and occupy more space than the SLI battery,
the number of batteries kept in stock will be determined by the space
available in a given facility. Given the alternatives of additional in-
vestment in physical plant and increased rate of turnover, the latter
seems more likely to occur.
3.2.4 Impacts on Employment Due to Required Maintenance
The maintenance or replacement of the lead-acid electric-vehicle
battery is responsible for a significant portion of auto-related employ-
ment. This impact, however, has been analyzed in Sec. 3.1 under the
assumption that service stations will be more likely to perform these
9-42
-------
labor-intensive but simple functions since wage rates of service station
attendants are considerably lower than those of auto mechanics.
The maintenance of motor and controls may well be performed by auto
repair shops. Maintenance requirements are estimated at approximately one
hour per year per car, and include checking the controls, motor, and in-
frequently adjusting or replacing the brushes. Total labor required to
perform these tasks is estimated to be 2940 man years in 1980, 3,366 in
1990, and 3,800 in 2000. The same linear relationship is expected to hold
for lesser percentages of electric car usage. The estimated changes in
employment resulting from use of the lead-acid battery are summarized in
Table 3.7.
The nickel-zinc battery is expected to require replacement about
every 5.7 years. This interval exceeds the present SLI battery by 2.7
years and points to a decrease in total number of battery sales of 47 per-
cent due to reduced frequency of replacement. The estimated impacts on
employment are summarized in Table 3.8.
The zinc-chlorine battery is expected to have a replacement interval
of 7.2 years, exceeding the present SLI battery by 4.2 years, and indica-
ting a decrease in total number of battery sales of 58 percent. The esti^
mated impact on employment is summarized in Table 3.8.
The lithium-sulfur battery has an expected life of 3 to 5 years.
With a 3-year life expectancy, the total number of sales would remain un-
changed and total employment in battery sales would be unaffected. The
only increase in employment would result from motor and controller main-
*
tenance. Given a 5-year life, the number of battery sales would decrease
by 40 percent. The estimated impact on employment is summarized in Table
3.8.
An increase of 2,940 in 1980, 3,366 in 1990, and 3,800 in 2000.
9-43
-------
TABLE 3.7
INCREASES IN EMPLOYMENT IN AUTOMOTIVE AFTERMARKET AND REPAIR
INDUSTRY DUE EXCLUSIVELY TO USE OF LEAD-ACID BATTERY*
Best Lead-Acid Battery
Employment
Battery Sales:
Forecast
Change
**
Maintenance
Total
Battery Sales:
Forecast
Change
**
Maintenance
Total
1980
2,073
-249
+2,940
+4,764
Worst Lead-Acid
2,073
+2,697
+2,940
+7,710
1990
2,495
-299
+3,366
+5,562
Battery
2,495
+3,243
+3,366
+9,104
2000
2,922
-350
+3,800
+6,372
2,922
+4,799
+3,800
+10,521
**
Reflects total change for each year independently, based on projections
in Tables 4.5, 4.6, and 4.7 of Task Report 4.
*
Motor and Controller only.
9-44
-------
TABLE 3.8
INCREASES IN EMPLOYMENT IN AUTOMOTIVE AFTERMARKET AND
REPAIR INDUSTRY
EXCLUSIVE USE OF THE NICKEL-ZINC BATTERY
Employment
Battery Sales:
Forecast
Change
Maintenance
Total
1990
2,495
-1,172
+3,366
+4,689
2000
2,922
-1,373
+3,800
+5,349
EXCLUSIVE USE OF THE ZINC-CHLORINE BATTERY
Employment
Battery Sales:
Forecast
Change
Maintenance
Total
1990
2,495
-1,447
+3,366
+4,414
2000
2,922
-1,695
+3,800
+5,027
EXCLUSIVE USE OF THE LITHIUM-SULFUR BATTERY
Employment
Battery Sales:
Forecast
Change
Maintenance
Total
1990
—
—
—
—
2000
2,922
-1,169
+3,800
+5,553
**
Reflects total change for each year independently.
Ic
Motor and controller only.
9-45
-------
The total net impact on employment in the automotive aftermarket
and auto repair industry is summarized in Table 3.9. Since again the
relationships are assumed linear, given mixes of batteries across years
may be interpolated from the table.
From the point of view of national implementation, we would expect
the same relative impacts. Intensity of impact on the aftermarket and
auto repair industry would be expected to vary with concentration of cars,
variations in levels of use, and environmental conditions affecting the
life of the car.
3.3 MINOR IMPACTS
The analysis pursued in this subsection examines the impacts of
electric car utilization on sectors we believe represent negligible ef-
fects, such as auto sales and wholesale distribution; tires and tire
retreading; and automobile, motor, and control manufacturing. While some
adjustments will be necessary in the short run, the impacts are expected
to be minimal in the long run.
Although the impact of electric cars on auto sales will be direct,
it is reasonable to assume that new car dealers will be able to shift
their activities with a minimum of cost and effort. Dealer's inventories
are mainly in cars and physical plant. Since their facility requirements
will remain essentially unchanged with the introduction of electric cars,
no additional capital investment directly related to plant capacity for
the vehicles is foreseen. The constant turnover of cars should enable
the dealer to gradually increase the proportion of electric cars as re-
quired. Dealer employees will need to familiarize themselves with speci-
fications, performance characteristics, and other relevant information
desired by consumers.
Total sales by new car dealers will be affected by the comparative
cost of electric cars to the consumer, vehicle performance in terms of
9-46
-------
TABLE 3.9
TOTAL NET IMPACT OF ELECTRIC CARS ON EMPLOYMENT AND PAYROLL IN THE
AUTOMOTIVE AFTERMARKET AND AUTO REPAIR INDUSTRY
1980
Lead-Acid Battery
"Best Case"
"Worst Case"
Lead-Acid Battery
"Best CAse"
"Worst Case"
Nickel-Zinc Battery
Zinc-Chlorine Battery
Lead-Acid Battery
"Best Case"
"Worst Case"
Nickel-Zinc Battery
Zinc-Chlorine Battery
Lithium-Sulfur Battery
"Best Case"
"Worst Case"
Total
Employment
21,229
24,176
23,509
27,048
22,636
22,361
29,652
33,801
28,629
28,307
28,833
30,002
Number of
Jobs Lost
13,464
10,517
1990
14,935
11,396
15,808
16,083
2000
19,520
15,371"
20,543
20,865
20,339
19,170
Percent of
Originally Forecast
Employment*
39%
30%
39%
30%
41%
42%
40%
31%
42%
42%
41%
39%
Payroll
Decrease*
(1972 $ Millions)
109.4
84.1
139.7
107.5
146.9
150.4
183.3
142.1
192.5
192.5
187.9
178.7
Includes SIC 5013, 5531, 7538 and 7539.
-------
speed, range, frequency and -cost of maintenance relative to the conven-
tional car, changes in personal tastes and preferences and incentives for
electric car use (or disincentives for conventional car use) created by
*
public policy decisions.
The average price of the conventional internal combustion engine
automobile in 1973 dollars, weighted by the projected auto population
&&
share by class, is estimated to be $3,618 in 1989, $3,282 in 1990, and
$3,238 in the year 2000. The electric car, including the initial bat-
tery, is estimated at $4,177 with the lead-acid battery, $5,885 with the
nickel-zinc battery, $3,291 with the zinc-chlorine battery and $3,395
with the lithium-sulfur battery. A comparison of these costs by year is
detailed in Table 3.10.
TABLE 3.10
COMPARISON OF PRICES FOR ELECTRIC CARS
WITH INTERNAL COMBUSTION ENGINE CARS BY YEAR
Lead-Acid Battery Car
Nickel-Zinc Battery Car
Zinc-Chlorine Battery Car
Lithium-Sulfur Battery Car
1980
13% higher
1990
21% higher
44% higher
No change
2000
45% higher
2% higher
5% higher
Including battery.
t
Examples of such incentives or disincentives are found in Ref. 23.
Task Report 3, Fig. 6.6.
Calculated in 1973 dollars and including costs of required emission
control apparatus.
9-48
-------
At a constant level of sales, the dollar value of dealer inventories
will vary with the higher or lower price of the electric car relative to the
conventional. For example, if in 1990 the nickel-zinc battery powered
auto is 55 percent higher than the cost of the conventional car, the
dealer's inventory would also be 55 percent higher. On the other hand,
the higher cost of the electric vehicle may serve as an incentive for con-
sumers to purchase new vehicles at less frequent intervals than the present
24
national average of one out of eight households per year.
The national average of total annual sales per dealer shows an up-
14
ward trend over the last decade. Since the average sales per dealer in
the State of California are among the highest in the nation, we assume
that the state will follow the national trend at a higher level. While
total sales may vary somewhat from this upward trend, it is not apparent
that this will significantly affect the total employment or number of
firms in automobile sales and distribution (SIC 5012, 5511 and 5521) in
the SCAB.
Dealerships will be affected by electric car utilization in the
areas of service and repair. Approximately 15 percent of a dealership's
14
total sales are attributable to service and repair activity. Table
3.11 details recommended inventories for service departments of dealer-
ships. Accepting this breakdown as a proxy for that portion of the
dealer's repair activities which are internal-combustion-engine related,
we observe that 50 percent of parts inventory is related and 44 percent
is unrelated to internal combustion engines. Considerably less of the
dealer's service and repair operation is related to internal combustion
engines (cf. 72 percent for auto repair shops). This may be partially
explained by the fact that the dealer operated shop performs a higher
percentage of body repair work than does the general auto repair shop.
Dealers also tend to emphasize service to newer cars which are under war-
ranty, therefore they engage in less major engine work than auto repair
shops.
9-49
-------
TABLE 3.11
RECOMMENDED INVENTORIES FOR CAR DEALERS SERVICE AND REPAIR
Percent of
Total Inventory
*
Engine Cooling 21%
Electrical
Related to Internal Combustion Engine 10
Unrelated to Internal Combustion Engine 2
Rear End and Transmission Accessories 12
Accessories
Related to Internal Combustion Engine 2
Unrelated to Internal Combustion Engine 2
**
Miscellaneous Mechanics
*
Related to Internal Combustion Engine 23
Unrelated to Internal Combustion Engine 11
Body and Sheet Metal 17
Total 100%
SOURCE: Phone conversation with Mr. Perron of the National Automobile
Dealers Association (NADA), Washington, D.C. Related and unre-
lated relative percentages were estimated by those knowledgeable
in the auto repair field.
*
Repair directly related to internal combustion engines.
**
Includes:
Mechanical, fuel, brakes, exhaust, shocks, steering, and front end.
9-50
-------
Since 15 percent of the dealership's total sales are in the service
and repair category and 56 percent of that is related to internal combus-
tion engines, we would expect a decrease of 8.4 percent in sales and
employment in SIC 5511 and 5521. This impact can be evaluated as an
average of 3,450 jobs lost at maximum utilization of electric cars. This
figure is relatively constant over the three decades since little growth
is projected in this area of employment through 2000. Decrease in total
payroll is estimated to be 33.1 million in 1980; 35.6 million in 1990;
and 38.2 million in 2000.
Impact on tire sales and retreading (SIC 5014 and 7534) will be
slight in terms of employment and number of firms in these two classifi-
*
cations. Present day tires should prove adequate, with minor modifica-
tions in production methods and specifications at most. Electric cars
will require larger, more expensive tires due to the increased weight
imposed by heavier batteries. Electric cars will accelerate less rapidly,
thereby reducing wear on tires. At the same time, however, they will be
heavier than conventional cars due to battery requirements. The increased
weight counteracts the reduced wear from slower acceleration leaving tire
mileage, number of tire sales, and tire retreading unaffected.
The regional impacts on auto manufacturing do not appear to be great.
**
An informal survey of auto manufacturing firms (SIC 3717 ) with an excess
of 250 employees showed that activity is limited mainly to assembly opera-
tions and manufacturing of automobile accessories not related to new or
replacement parts for equipment related to internal combustion engines.
*
Analysis of increase in cost of tires is found in Sec. 2.
**
SIC 3717 formerly included Motor Vehicle and Parts Manufacturing and
Assembly. The classification was changed as follows in 1972:
SIC 3711 Motor Vehicle Manufacturing
SIC 3712 Passenger Car Bodies
SIC 3714 Motor Vehicle Parts and Accessories
Source: County Business Patterns, 1972.
9-51
-------
The firms approached in the survey account for approximately 40 percent of
total employment in auto manufacturing in the Los Angeles area. Also es-
tablished through the survey was the fact that the major engine-related
parts are manufactured outside of the South Coast Air Basin. The parts
are shipped to Los Angeles and autos are assembled in the area by the
major manufacturers. An estimated 80 percent of those cars assembled in
the area remain there. Little impact would be expected on assembly opera-
tions and passenger car body manufacturing since few changes would be
required to convert present internal-combustion-engine designs for use
with electric motors. The effect of large scale utilization of electric
cars on the number of existing firms and employment in the auto manufac-
turing classification will be that normally expected as a result of design
changes and new tooling requirements frequently encountered in present
year-to-year model changes.
Table 3.12 contrasts percentage changes in factory sales with changes
in employment in automobile manufacturing at the national level. Observing
the average change and standard deviation of the two, we notice that em-
ployment has not been extremely sensitive to fluctuations in factory activity.
We might therefore expect a lesser impact on employment as a result of short
term fluctuations inherent in the changing market conditions foreseen with
the implementation of electric cars, than we would expect if a closer re-
lationship had been shown to exist between factory sales and employment in
the past.
The impacts of electric car substitution on motor and controls manu-
facturing is minor in terms of national employment. The increase in activity
is expected to be absorbed in regions already supporting these industry sec-
tors because these two industries are characterized by sizeable levels of
capital investment and less important transportation costs.
The projections for the dollar sales in millions of 1973 dollars for
motor and control manufacturing for the years 1980, 1990 and 2000 are
9-52
-------
TABLE 3.12
US AUTOMOBILE MANUFACTURING
24
1967
1968
1969
1970
1971
1972
*
Employment
765,314
774,246
806,461
714,226
746,929
766,926
Average Change
Standard Deviation
Percent
of Change
1.17%
4.16
-11.44
4.58
2.68
0.23
6.66
Factory Sales
(1972 $1,000)
7,436,764
8,822,158
8,223,715
6,546,817
8,584,592
8,823,928
Percent
of Change
18.63%
-6.78
-20.39
31.13
2.79
5.08
20.35
County Business Patterns.
detailed in Table 3.13 along with the projected values of additional re-
quirements given maximum substitution of electric cars. The percentage
increases, also shown in Table 3.13, are used to derive the national em-
ployment impacts.
The projected requirements are derived by applying the expected
auto sales figures for each of the decade intervals to the manufacturing
cost of the required electrical equipment. These figures of $100 for
the controls and $225 for the motor are taken from Ref. 12. These prices
are presented in 1973 dollars for manufacturing costs in order to be con-
sistent with the value of manufactured equipment from the Department of
25
Commerce Report.
These increases do not appear significant in terms of the national
labor force. In fact, the total projected impacts are considerably smaller
9-53
-------
TABLE 3.13
NATIONAL IMPACTS ON MOTOR (SIC 3621) AND INDUSTRIAL CONTROLS (SIC 3622)
MANUFACTURING WITH 100 PERCENT ELECTRIC CAR USAGE IN LOS ANGELES
Forecasted Equipment Manufacturing Other
Than Electric Vehicle
25
Motors ($ Millions)
Employment
Controls ($ Millions)25
Employment
Projected Additional Electric Vehicle
Requirements
Motor ($ Millions)
Controls ($ Millions)
Percentage Impact
Motors
Controls
Employment Impact
Motors
Controls
1980
3,800
100,000
2,000
50,000
127
56
3.3%
2.8%
3,300
2,800
1990
6,073
159,813
3,650
91,246
144
64
2.4%
1.7%
3,835
1,551
2000
9,705
255,402
6,661
166,517
160
71
1.6%
1.1%
4,086
1,831
then the Commerce Department's projected yearly growth rates. In the past,
3 percent of this national activity has been in Los Angeles. If that trend
continues, the impacts will be practically nil.
9-54
-------
4 SECONDARY IMPACTS
The impacts upon transportation supplies and services are cumula-
tive in their impact upon the South Coast Air Basin. Table 4.1 presents
the change in economic activity which is expected to accrue to the region.
This was derived, for each battery alternative, by summing the net change
in activity expected for battery manufacturing, petroleum distribution,
auto parts sales, auto repair, and auto sales. This impact is also shown
as a percentage of projected baseline activity for employment and total
personal income. Personal income is used, rather than payroll, as it is
*
a better indicator of total regional economic activity.
As was the case in Sec. 3, the impacts are assumed linear. That is,
for any level of electric car substitution less than 100 percent, the
impacts are proportional to the level of substitution. The activity change
as a percent of baseline activity is shown to be fairly constant for each
of the decade intervals.
This table should not be construed to predict any addition to the
unemployment rolls which would result from a conversion to electric cars.
This is a summation of the total dislocation in employment which would
occur over the conversion period. It is reasonable to assume that this
conversion would take place over a period of years and that much of this
dislocation would accrue through normal attrition.
The reader will note that the worst lead-acid battery case shows a
loss of employment and a gain in payroll. This is due to the change in
the mix of employment. There is a loss of employment in the petroleum
distribution sector including very low salaried service station attendants,
and a corresponding gain in high salaried manufacturing employment. Table
4.2 shows this in more detail. The table contains data from projected
1990 activity. The structure of employment impacts varies little with
the assumed decade of conversion. The structure does vary both by battery
type and by employment category. As shown in Table 4.2, there is a
*
Salary income is about 70 percent of total personal income.
9-55
-------
TABLE 4.1
IMPACT OF 100 PERCENT ELECTRIC CAR USAGE ON EMPLOYMENT AND INCOME
IN THE LOS ANGELES AREA
Best Lead-Acid Battery .
Worst Lead-Acid Battery
1
SCAB Employment Baseline
SCAB Personal Income
1
1980
Change in Activity
Employment
-34,415
- 9,642
A
Payroll
-164,111
96,539
4,335,000
69,830,000*
Change as Percent of SCAB Baseline
Employment
-0.79%
-0.22
Personal Income
-0.23%
0.14
Best Lead-Acid Battery
Worst Lead-Acid Battery
Nickel-Zinc Battery
Zinc-Chlorine Battery
SCAB Employment Baseline
SCAB Personal Income
Nickel-Zinc Battery
Zinc-Chlorine Battery
Best Lithium-Sulfur Battery
Worst Lithium-Sulfur Battery
SCAB Employment Baseline
SCAB Personal Income
1
Change i
Employment
-36,067
- 7,574
-40,131
-54,472
4,579,
95,465,
2
Change i
Employment
-45,594
-61,420
-63,143
-58,050
5,025,
135,293,
?90
T Activity
Payroll
-186,986
140,870
-168,602
-338,756
300
300*
300
i Activity
*
Payroll
-299,353
-407,837
-425,225
-368,937
500
DOO*
Change as Per
Employment
-0.74%
-0.17
-0.88
-1.20
Change as Per
Employment
-0.91%
-1.21
-1.25
-1.15
cent of SCAB Baseline
Personal Income
-0.20%
0.15
-0.18
-0.36
cent of SCAB Baseline
Personal Income
-0.22
-0.30
-0.31
-0.27
1972 dollars in thousands.
9-56
-------
TABLE 4.2
STRUCTURE OF EMPLOYMENT CHANGES
*
Average Salary
Employment Impacts
Best Lead-Acid
Worst Lead-Acid
Nickel-Zinc
Zinc-Chloride
Best Lithium-Sulfur
Worst Lithium-Sulfur
Average
Standard Deviation
Auto
Service
9.3
-14,953
-11,396
-15,808
-16,083
-15,634
-14,636
-14,751
1,730
Auto Sales
10.3
-3,450
-3,450
-3,450
-3,450
-3,450
-3,450
-3,450
0
Manufacturing
11.8
14,774
39,728
20,637
6,571
4,579
8,054
15,724
13,176
Petroleum Distribution
5.7
-32,456
-32,456
-41,510
-41,510
-41,510
-41,510
-38,492
4,675
*
1972 $thousands.
considerable gain in higher skilled and higher paid manufacturing employees
with the lead-acid and nickel-zinc battery types. In all cases, there is
a considerable loss of employment in the lower paid and lower skilled ser-
vice station employees, a group that could have difficulty finding alter-
native employment.
In addition to impacts upon consumers and some employment sectors,
the conversion to electric cars has the potential to disturb the economic
equilibrium in the region. This impact is difficult to quantify in a
region as large and complex as Los Angeles. Some general assessments are
possible, however. Recall that consumers expenditures for private trans-
portation vary with electric car types and that there is a consistently
negative input upon regional employment opportunities with electric car
substitution. If the consumer is paying more and there are fewer jobs
being generated, these funds are being diverted. In this case, the funds
9-57
-------
are diverted outside the region to procure the extra materials (lead,
nickel, zinc, etc.) required for battery production. These materials
are not available within the region.
Table 4.3 shows the potential impacts upon the Los Angeles economy
for various electric car scenarios. These impacts were estimated by
calculating the wholesale cost of goods which would be procured outside
the region. The benefits of retail activity accrue to the region irres-
pective of the origin of the product. These wholesale costs were adjusted
to reflect the traditional multiplier effect in order to estimate the
total potential impact.
There are competitive market factors which make it unlikely that
Los Angeles would experience the full extent of those potential impacts.
To experience the full impact, regional activity would have to decrease
TABLE 4.3
POTENTIAL CUMULATIVE IMPACTS IN LOS ANGELES
Change in Regional Payroll
1980 Lead-Acid Battery: Best Case
Lead-Acid Battery: Worst Case
1990 Lead-Acid Battery: Best Case
Lead-Acid Battery: Worst Case
Nickel-Zinc Battery
Zinc-Chlorine Battery
2000 Nickel-Zinc Battery
Zinc-Chlorine Battery
Lithium-Sulfur Battery: Best Case
Lithium-Sulfur Battery: Worst Case
$ Millions
241.4
-3,085.7
6.7
-3,797.3
-2,208.4
2,558.5
-2,568.8
2,812.0
2,530.8
1,922.8
Percent of
Total
0.5%
-7.1
0.01
-8.2
-4.8
5.6
-5.1
5.6
5.1
3.8
9-58
-------
proportional to the increase in materials imports. It is likely that the
region would substitute these materials for other imports (e.g., refrig-
erators), thus keeping their level of economic activity stable. Alter-
natively, if imports increased, that increasingly available portion of the
labor pool would be available at some lower price. This would lower the
cost of exports and lead toward a new equilibrium position for the region.
9-59
-------
9-60
-------
REFERENCES
1. J.C. Eisenhut, Economic Projections for the Los Angeles Region,
1980-2000. General Research Corporation RM-1860, February 1974,
(also Task Report 4).
2. Cost of Operating an Automobile, U.S. Department of Transportation,
April, 1972.
3. The Economic Impact of Pollution Control; A Summary of Recent
Studies, Prepared for the Council on Environmental Quality, Depart-
ment of Commerce, and Environmental Protection Agency, March 1972.
4. U.S. Energy Outlook, National Petroleum Council, December 1972.
5. A. Sjovoid, Electric Energy Predictions for the Los Angeles Region,
1980-2000, General Research Corporation RM-1859, November 1973-,
(also Task Report 5).
6. Wall Street Journal, March 13, 1974
7. U.S. News and World Report, p. 20, March 25, 1974.
8. Wall Street Journal, February 27, 1974.
9. The Economic Cost of Pollution Control, Environmental Protection
Agency, March 1972.
10. W.F. Hamilton and G.M. Houser, Transportation Projections for the
Los Angeles Region. 1980-2000, General Research Corporation RM-1858,
November 1973, (also Task Report 3).
11. Care and Feeding of 4,000 Electric Trucks, G. Pearson, London,
England.
12. L.R. Foote et al, Electric Vehicle Systems Study, Ford Motor Company
Scientific Research Staff, Technial Report SR 73-132, October 25,
1973.
13. D. Friedman and J. Andon, Minicars, Inc., and W.F. Hamilton, Char-
acterization of Battery-Electric Vehicles for 1980-1990, General
Research Corporation RM-1931, January 1974, (also Task Report 1).
9-61
-------
REFERENCES (Cont.)
14. Facts, National Automobile Dealer's Association, Washington, D.C.,
1973.
15. A.R. Sjovoid, Parametric Energy and Resource, and Noise Impacts of
Electric Cars in Los Angeles, General Research Corporation RM-1906,
1974, (also Task Report 8).
16. California Statistical Abstract, Sacramento, California, 1972.
17. "U.S. Service Station Sales and Operating Ratios, 1972," National
Petroleum News Factbook, McGraw-Hill, Inc., 1973.
18. B. Ikert, "Interview with Wallace Wegge, Look's Research Director,"
Automotive Service Digest, November 1959.
19. D.A. Randall and A.P. Glickman, The Great American Auto 'Repair
Robbery, Charterhouse, 1972.
20. Motor Age, January 1974, Chilton Industries, pp. 40-45.
21. A. Till, What You Should Know Before You Have Your Car Repaired,
Shelbourne, 1970.
22. Reference 20, pp. 48-49.
23. W.F. Hamilton, Usage of Electric Cars in the Los Angeles Region,
General Research Corporation RM-1891, 1974 (also Task Report 10).
24. Automobile Facts and Figures, 1973/74, Motor Vehicle Manufacturer's
Association of the United States, Detroit, Michigan, 1974.
25. U.S. Industrial Outlook, 1974, Department of Commerce.
9-62
-------
TASK REPORT 10
USAGE OF ELECTRIC CARS
IN THE LOS ANGELES REGION, 1980-2000
W.F. Hamilton
-------
ABSTRACT
Future usage is estimated for four-passenger battery-electric cars
with ranges between overnight recharges of 55 miles before 1980 and 145
miles thereafter. A new analysis of the Los Angeles 1967 travel survey
provides basic data: distributions of daily driving mileages, and avail-
ability of off-street parking. On this basis, one million second cars
at single-family households, or 17 percent of 1980 Los Angeles area cars,
could be replaced by the 1980 electric car with very little loss of capa-
bility, while the advanced battery cars could replace 46 percent of the
area's cars in 1990 and 74 percent by 2000.
From a review of these estimates, plus current market share data
for conventional subcompacts and results of consumer surveys regarding
electric cars, free-market electric car populations for the Los Angeles
region are subjectively estimated at under 1 percent in 1980, about 5
percent in 1990, and about 10 percent by 2000. Much higher populations
could be supported by potential production capability: near 75 percent
by 1990. Hypothetical intermediate electric car populations ranging from
23 to 46 percent in 1990 are then selected for eventual use in analyzing
potential impacts of electric cars on regional environment, energy, and
economy. Until cost disadvantages of electric cars are eliminated by
improved battery technology, powerful legislative incentives would be
required to bring about these hypothetical usages.
-------
ii
-------
CONTENTS
SECTION
PAGE
1
2
3
4
5
6
7
8
APPENDIX
ABSTRACT
INTRODUCTION
APPROACH TO USAGE ESTIMATION
PATTERNS OF CONVENTIONAL CAR USE
LOS ANGELES SURVEY DATA ANALYSIS
APPLICABILITY OF ELECTRIC CARS
CONSUMER MARKET FOR ELECTRIC CARS
6.1 Performance and Price Competition
6.2 Overall Market Trends
6.3 Current Surveys of Market Potential
6.4 Estimated Market Potential in Los Angeles
UPPER-BOUND RATE OF ELECTRIC CAR INTRODUCTION
SCENARIOS FOR ELECTRIC CAR INTRODUCTION
8.1 Intermediate Levels of Use
8.2 Implementation Measures
8.3 Gasoline and License Fee Surcharges
8.4 Subsidy of Electric Cars
8.5 Direct Travel Restrictions
8.6 Scenarios For Electric Car Introduction
COMPUTER PROCESSING OF LOS ANGELES TRANSPORTATION
DATA
REFERENCES
±
10-1
10-3
10-6
10-11
10-23
10-30
10-31
10-35
10-36
10-44
10-50
10-54
10-54
10-56
10-58
10-64
10-65
10-66
10-69
10-87
iii
-------
iv
-------
ILLUSTRATIONS
NO. ; PAGE
3.1 Distributions of Trip Times 10-8
3.2 Distributions of Daily Mileages for Automobiles, by Annual
Usage Class 10-9
4.1 Adjustments of Surveyed Daily Mileage Distributions 10-20
4.2 Adjusted Distributions of Daily Travel, Los Angeles Region,
1967 10-21
5.1 Probability of Daily Driving Less Than a Given Distance for
Secondary Drivers with Cars 10-24
5.2 Daily Travel by Low-Travel Drivers 10-28
6.1 Shares of the Subcompact Market Relation to Price and
Performance, South Coast Air Basin, 1972 10-33
7.1 Automotive Product Development Phases 10-51
8.1 Alternative Electric Car Population Projections, South
Coast Air Basin 10-57
8.2 Required Surcharges to Establish Subcompact Conventional
Car Cost Disadvantage Relative to Electric Cars 10-62
A.I Processing of Los Angeles 1967 Travel Survey 10-70
-------
vi
-------
TABLES
NO.
3.1
3.2
3.3
A.I
4.2
4.3
4.4
4.5
5.1
5.2
5.3
5.4
6.1
6.2
6.3
6.4
6.5
6.6
6.7
Baseline Auto Travel Projections — South Coast Basin
Electric Car Characterizations
Annual Mileage Versus Car Age
Sample Selection from Los Angeles 1967 Home- Interview Study
Characteristics of Sample Auto Driver Trips
Comparison of Los Angeles 1967 Home- Interview Survey Results
with Independent Controls
Adjusted Sample Averages
Survey Availability of Cars to Drivers
Survey Availability of Parking, Los Angeles Area
Distributions of Housing Units and Cars, LARTS Area
Distribution of Housing Units and Cars, South Coast Air
Basin
Candidates for Electric Car Replacement, South Coast Air
Basin
Sales of 1972 Subcompact Cars
Electric Car Consumer Surveys, April-May, 1973
Survey Factors Reported Important in Auto Choice
Electric Car Description in Survey
Improvements Most Desired in Electric Car
Potential US Electric Vehicle Population
Interest in Electric Car Purchase, March-April 1972
PAGE
10-7
10-7
10-10
10-13
10-14
10-16
10-18
10-22
10-25
10-26
10-27
10-29
10-32
10-37
10-38
10-39
10-39
10-41
10-42
vii
-------
Tables (Cont.)
NO. PAGE
6.8 Estimated National Potential Market for Electric Cars in 1974 10-43
6.9 Estimated National Market Penetration of Electric Cars to
1980 10-44
6.10 Potential Free-Market Sales and Use of Electric Cars, South
Coast Air Basin 10-47
7.1 Projected Maximum Electrification of Cars in South Coast
Air Basin 10-53
8.1 Alternative Projections of Electric Car Market Share,
Population, and Use, South Coast Air Basin 10-56
8.2 Extra Initial Costs of Electric Subcompact Cars 10-59
8.3 Extra Life-Cycle Costs of Electric Subcompact Cars 10-59
8.4 Extra Total Costs of Electric Subcompact 10-60
8.5 Scenario Elements for Electric Car Introduction 10-67
viii
-------
1 INTRODUCTION
The objective of this paper is to estimate likely levels of electric
car use in 1980, 1990, and 2000 in the Los Angeles region. These estimated
levels of use are eventually to be utilized in an overall evaluation of
electric car impacts. The specific region of concern is California's
South Coast Air Basin, which includes greater Los Angeles, and the usage
characteristics required are the numbers and daily mileages of the elec-
tric cars.
Even after the characteristics of future electric cars have been
detailed, estimating their use is difficult. There is little data or
experience to go on, since electric cars have not found much application
since the early years of this century. On the one hand, electric cars
could do much of what conventional cars now actually do. On the other
hand, however, the electric cars are likely to be less than competitive
with the conventional alternative in dollar cost, in performance, in
overall accommodations, and—more important—in daily range. The exist-
ing marketplace offers some clue to the acceptability of cars which are
smaller, or more expensive, or less powerful than most of their competi-
tion, but it tells us nothing about cars with a stringent daily range
limitation.
Were there nothing else involved, prospects for substantial use of
electric cars would be dim indeed. But other factors are very much in-
volved: the undesirable side-effects of the conventional automobile on
the quality of urban life; the prospects that electric cars might con-
tribute much less to pollution of the urban environment; and the poten-
tial independence of electric cars from dwindling petroleum reserves.
These are presently externalities to the market transactions wherein
consumers choose between electric and conventional automobiles, but they
are likely to play an increasingly important role in the marketplace as
public policy and collective action increasingly regulate the character-
istics and the usage of private automobiles. Beyond whatever today's
10-1
-------
market shows, then, about the electric car's potential, there is a real
prospect that public policy will greatly enhance its relative desirability.
The effect of public policy is particularly important in this study
of electric car impact. Without the support of positive public policy,
it is quite conceivable that the most likely level of electric car use
would remain near zero, making impacts trivial. Simply to make this im-
pact study useful, then, it is essential that sufficiently influential
public policy be assumed to justify estimates of substantial usage.
Such policy may not be the most likely future course; but the usage
estimates developed here need not be confined to the most likely possi-
bility.
In the next section, this paper sets forth briefly its overall
approach to estimating reasonable levels of electric car use. In subse-
quent sections, it deals with each step in the approach: the current
patterns of automobile use, the applicability of electric cars to them,
the potential free-market penetration of electric cars, and possible
limitations of production capability. Finally, it discusses alternative
public policies for electric car use, required measures for their imple-
mentation, and elemental scenarios for introducing alternative electric
cars in 1980, 1990, and 2000.
10-2
-------
2 APPROACH TO USAGE ESTIMATION
Because data is relatively scarce for incisive estimation of elec-
tric car usage, a great many uncertainties arise in the estimation process
and its results. It is thus appropriate to develop upper and lower bounds
on usage which can be established relatively firmly, and then to develop
within these bounds different estimates which might be associated with
different public policies for urban transportation.
The most striking feature of the electric car, at least for the
next decade, is its limited range between recharges, which may be on the
order of 50 miles in urban use. Far more than sluggish performance,
limited accommodation, or increased cost, this range limitation distin-
guishes the electric car from competing conventional automobiles.
The effect of this range limitation depends significantly on the
overall institutional system within which electric cars are operated.
At one extreme, all electric cars will be recharged only overnight, and
consequently will be limited to a daily range no more than the range be-
tween recharges. At the other extreme, electric car batteries will be
replaceable in minutes at service stations which will maintain a supply
of recharged battery packs just as conventional service stations maintain
large tanks of fuel for conventional cars. In this situation, motorists'
daily range would be virtually unlimited; the only penalty of the limited
range between recharges would be more frequent stops for "refueling."
Either of these extremes seems technically feasible. The latter,
however, is institutionally complex, requiring not only proper designing
of auto battery packs for easy replacement, but networks of service
stations equipped with the necessary handling equipment and high-power
recharge capability. A single battery pack may weigh 1,500 pounds, for
example, while a hundred battery packs on recharge might easily require
a thousand kilowatts of charging power. More complex still would be the
development of an equitable system for leasing battery packs, each worth
10-3
-------
well over a thousand dollars, for each 50-mile use of electric automobiles,
especially since the conditions of use substantially affect battery life.
I
Because of its simplicity, the home-recharge system of electric
car use is the most reasonable prospect for the introduction of electric
cars. Once electric cars thus became numerous, home recharge would surely
be supplemented by recharge facilities at places of work, and at parking
lots generally; and eventually battery replacement stations and systems
might be expected to develop through successful competition with the
home-recharge system.
The initial arrangement of home recharge might thus give way even-
tually to the more elaborate system of battery replacement—but only so
long as the stringent range limitation remains. Were range between re-
charges significantly increased, the motivation for the battery replace-
ment system would largely disappear.
The introduction of high-performance batteries with a range between
recharges of 150 miles or more is projected in this study before 1990.
In consequence, battery replacement systems seem likely to be obviated
before they have had a chance to grow to fruition.
Thus the usage estimates developed in this report are based on the
home recharge system, that is, on daily range limited to the range be-
tween recharges of the electric cars. If anything, this will tend to
make total usage estimates conservative. In any case, it makes the im-
portance of daily range, and its compatibility with typical travel needs,
crucial in the estimation procedure.
The approach to usage estimation undertaken in this paper proceeds
as follows:
• Analyze usage patterns of conventional automobiles, empha-
sizing the frequency with which given daily ranges are
exceeded, and the probable feasibility of home recharge.
10-4
-------
Determine prospective numbers of automobiles which are
driven and parked in such manner that electric cars could
accomplish essentially all their functions.
Estimate an upper bound on electric car use based on replace-
ment of appropriate conventional cars and possible limita-
tions due to likely lead times in introducing electric cars
into the automobile population.
Estimate a lower bound on electric car use by considering
potential free-market penetration likely, given the rela-
tive performance, accommodations, and cost of future elec-
tric cars relative to competitive conventional automobiles.
Select intermediate levels of electric car use which might
be sought by public policy, outline quantitatively the
measures needed to realize these policy levels of use,
and hypothesize elemental scenarios for introducing alter-
native electric cars to meet the usage levels in 1980,
1990, and 2000.
10-5
-------
3 PATTERNS OF CONVENTIONAL CAR USE
Overall transportation projections for the Los Angeles area were
2
developed elsewhere in this study. Their overall result is reproduced
here as Table 3.1, which shows average individual car usage projected to
future years for the South Coast Air Basin.
Four passenger electric cars characterized for future years in this
3
study are briefly described in Table 3.2. Evidently, the daily range
capability of even the lead-acid battery cars is substantially greater
than average daily usage, which is about 30 miles. Costs and other
aspects of electric cars will be discussed in relation to their probable
use in Sees. 6-8.
Though average usage may be well within electric car range, almost
every automobile is likely to be driven more than the electric car range
on some percentage of days. Only if this percentage is small is the con-
ventional automobile a reasonable candidate for replacement by the elec-
tric car. The probability that conventional cars are driven more than a
given distance in a day thus becomes the critical factor in estimating
the potential or applicability of electric cars in urban travel.
In the last two decades, tremendous quantities of urban travel data
have been gathered, analyzed, and expanded into projections for the future.
A typical result is reproduced as Fig. 3.1, which shows the distribution
of trip times less than a given number of minutes. With the average
speeds of Table 3.1, trip times may be approximately converted into trip
distances, and thus the probability of a work trip longer than a given
distance may be readily derived. Obviously, there are very few trips
surveyed or forecasted in Los Angeles with lengths greater than 50 min-
utes or roughly 25 miles, and thus beyond the daily capability of even
the 1980 electric car. As Table 3.1 shows, however, most cars are driven.
on four or five daily trips. Thus the distribution of individual trip
distances is not by itself a good indicator of electric car applicability.
10-6
-------
TABLE 3.1
BASELINE AUTO TRAVEL PROJECTIONS--
SOUTH COAST BASIN
1980 1990 2000
Daily Vehicle-Miles , millions
Percent on Freeways
Percent on Streets
Daily Miles Per Vehicle
Daily Trips Per Vehicle
Daily Minutes Per Vehicle
Miles Per Trip
Minutes Per Trip
Average Speed , mph
167
39
61
29.3
4.6
53
6.15
11.5
32.0
196
42
58
29.2
4.6
54.7
6.35
11.9
32.0
228
45
55
30.0
4.6
56.2
6.52
12.2
32.0
TABLE 3.2
ELECTRIC CAR CHARACTERIZATIONS
Battery Type Lead-Acid Nickel-Zinc Zir.c-Chlorine Lithium-Sulfur
*
Urban Driving Range (Ref. 1)
Test Weight (Ref. 1), with 450-
Pound Pay load
Initial Cost** (Ref. 4)
**
Total Per-Mile Cost, cents (Ref. 4)
Estimated Availability (Ref. 1)
54 mi
3,975 Ib
$4,177
16c-21.6c
1978
144 mi
3,530 Ib
$7,325
21. 2c
1980
145 mi
2,950 Ib
$3,491
11. 7c
1985
139 mi
2,655 Ib
$3,395
12.2c-13.lC
1990
*
On SAE Metropolitan Driving Cycle.
**
In 1973 dollars, including battery cost. Initial costs and life-cycle per-raile costs (including
finance charges) of a comparable ICE subcompact in 1973 were $2,270 and 13.3C, respectively.
10-7
-------
1001— HOME TO SHOPPING - 1990 __
HOME TO WORK - 1967
TO WORK - 1990
10
20 30 40
TRIP TIME, minutes
Figure 3.1. Distributions of Trip Times (From Ref. 5)
The real question is how often the total of all trip mileages in a day
is likely to exceed the range of the electric car.
It happens that the wealth of urban travel analysis has been almost
entirely focused on the individual vehicle or person trip as a unit of
analysis, rather than daily vehicle or person travel. Distributions of
daily vehicle mileage do not appear in the many urban travel analyses
available in the literature to this study. Even the definitive experi-
mental investigation of individual automobile usage in Los Angeles empha-
sizes distribution of trip distances and nowhere presents distributions
of daily travel distances.
The only source of daily travel distributions located in this
study was developed specifically for estimating the potential market for
electric vehicles. Its results are plotted in Fig. 3.2, which shows
the probability that cars in three different annual usage classes will
10-8
-------
99.8
99.5
99
98
tg 95
60
40
3000 mi/yr
6000 mi/yr
12,000 mi/yr
I
10
20
40 60 80 100
DISTANCE, mi
200
Figure 3.2. Distributions of Daily Mileages for Automobiles, by Annual
Usage Class (From Ref. 7)
will be driven less than a given mileage in a day. This analysis further
estimates that some 15 percent of US cars are in the 3,000-mile-per-year-
or under class, 40 percent are in the 6,000 mpy-or-under class, and 70 per-
cent are in the 12,000 mpy-or-under class.
Though the information of Fig. 3.2 is a great step forward from
trip-based analyses towards the requirements of this study, it is less
than ideal in several respects. First, it is based primarily on average
travel distances observed in Chicago in 1956. Thus it may not be appro-
priate for 1990 Los Angeles. Second, it relies entirely on assumed dis-
tribution forms for urban travel: Poisson distributions were assumed
for the distributions of both daily trip numbers and trip distances, and
only the average values of these distributions were obtained from obser-
vation. Third, the usage classes by miles per year are inappropriate
where market penetration must be estimated, because they are correlated
with car age. As Table 3.3 shows, cars which are driven very little are
generally very old. While a new electric car may be immediately applicable
10-9
-------
TABLE 3.3
ANNUAL MILEAGE VERSUS CAR AGE (from Refs. 8 and 9)
Average Annual Mileage (light-duty vehicles)
Age , years
1
2
3
4
5
6
7
8
9
10
>10
US
15,900
15,000
14,000
13,100
12,200
11,300
10,300
9,400
8,500
7,600
6,700
California
15,000
13,000
11,000
9,600
8,400
7,000
5,300
5,000
4,400
4,200
3,500
to the functions of a ten-year-old conventional car, it is probably not
a reasonable competitor for replacing the old car because it necessarily
carries a new-car price tag. In the longer run, electric cars could
"trickle down" through the system and become competitors for inexpensive
conventional used cars—but only if the electrics can gain a toe-hold
somewhere else in the auto market.
Because information on daily car use was not available in the
literature, it became necessary to reprocess extensive travel survey
data to deduce distributions of daily car use, as opposed to distributions
of individual trip characteristics as developed in past studies.
10-10
-------
4 LOS ANGELES SURVEY DATA ANALYSIS
As noted in Ref. 1, the basic source of Los Angeles transportation
data is the 1967 survey conducted by the Los Angeles Regional Transporta-
tion Study (LARTS). The survey consisted of intensive interviews at a
1% sample of households in the LARTS area during the fall of 1967.
Following up on previous notifications and requests for cooperation
through the mails, the interviewers recorded data about each household,
about its members, and about trips they had taken on the survey day. For
each trip, the addresses of origin and destination, the mode of travel,
the identity of the traveler, the purpose of the trip, and a host of
other descriptions such as time and duration, freeway usage, parking,
and vehicle occupancy were all recorded.
After coding and keypunching for computer processing, the data base
from the interviews became the principal basis for developing analyses,
models, and projections of Los Angeles travel demand. But because the
individual trip, rather than the day's travel by an individual or vehicle,
was the basic analytic unit in this processing, the results are not di-
rectly applicable to the question of electric car range adequacy.
To investigate typical vehicle use in an entire day, the Los Angeles
survey data was reprocessed in this study. Several reels of tape provided
by LARTS described each of almost 200,000 trips detailed in the interviews.
A separate reel of tape contained descriptions of households whose members
made these trips.
A computer program was written to read the trip and housing tapes.
Basically, the program accumulated total mileages traveled during the
survey day for occupants and vehicles of each individual household.
From this basic result, it then compiled distributions of total travel
mileage, so that the fraction of persons or vehicles traveling more than
a given total mileage on the survey day could be determined.
10-11
-------
Ideally, the program might have compiled distributions only for
total vehicle travel on the survey day. Unfortunately, however, the
interviews did not record which vehicle at a multi-vehicle household was
used for each of the trips reported by members of that household. Thus
the program was only able to develop vehicle mileage distributions for
single-car households. The interviews did record which individual of
the household made each trip, which made it possible to develop distri-
butions of daily travel for individual drivers of the households.
The computer program employed an approximate airline distance
rather than the actual over-the-road distance for each reported trip.
Though the interviews obtained addresses of trip origins and destinations,
this level of detail was lost as subsequent coding assigned each address
to one of some 1200 traffic zones into which the study region was divided.
Only the zones of origin and destination appeared on the tapes. Coordi-
nates of zone centroids (centers of gravity of population) were provided
by LARTS; but no detailed representation of the street and highway net-
work could readily be obtained and utilized to determine actual driving
distances. Consequently, simple straight-line distances between zone
centroids were used initially as trip distances, and adjusted upward later.
For trips which began and ended in the same zone, the program em-
ployed an average intrazonal travel distance which had been precomputed
for each zone. This distance was taken to be half the airline distance
from the zone centroid to the centroid of the nearest neighboring zone.
In the processing of the survey data, attention was focused on
those households reporting automobile trip details on the survey day.
In consequence, almost a third of the survey households were not included
in the development of daily mileage distributions.
Table 4.1 details the reasons for omissions of households and trips
in developing the set of households and trips in the sample for which
10-12
-------
TABLE 4.1
SAMPLE SELECTION FROM LOS ANGELES 1967 HOME-INTERVIEW STUDY
Total Survey Households 33,030
Total Survey Trips 195,693
Ommitted Households 10,704
unoccupied 21%
no cars 27%
unused cars 22%
no trip data 30%
Omitted Trips 190
erroneous ID 35
erroneous zone name 155
Sample Households 22,326
Associated Auto Driver Trips 130,800
distributions were calculated. Of almost 11,000 households omitted, the
largest single category indicated auto driver trips on the household data
tape but had no corresponding trip descriptions anywhere on the trip tape.
LARTS personnel suggest that this is at least partly the result of un-
usable trip descriptions given by survey respondents. Somewhat smaller
numbers of households were also omitted for each of three reasons: they
were vacant, had no cars, or reported no use of their cars. Of the omit-
ted households, 87 had descriptive data which were not in the standard
format and were unreadable by the processing program.
Only 190 auto driver trips were omitted from the 130,000 reported
by households in the processed sample. Most of these stated zone names
of origin or destination which were not in the list of zone names actually
10-13
-------
employed in the analysis. A few others had identification numbers which
did not correspond to any of the household identification numbers.
A flow chart and listing of the computer program which processed
the sample of Table 4.1 appear in the appendix.
The overall characteristics of the processed sample are summarized
in Table 4.2. Overnight and external trips—trips beginning or ending
outside the study region—amounted to less than 1 percent of all trips.
Neither were included in daily travel distributions, since in the case of
external trips, no distance could be associated with them, and because
data on such travel is available from the National Transportation Survey
of 1967. Intrazonal trips, though they amounted to 18 percent of all trips,
accounted for only 3 percent of total travel distance. Thus the precise
distance of each intrazonal trip is probably unimportant in total daily
mileages, and the probably inaccuracies in the estimates used for intra-
zonal trip lengths will not significantly degrade results.
TABLE 4.2
CHARACTERISTICS OF SAMPLE AUTO DRIVER TRIPS
Total trip distance 616,889 miles
Intrazonal trip distance 17,815 miles
Total trips 130,800
Intrazonal trips 23,503
Overnight trips 516
External trips 584
"Cars" were defined in the survey processing for this project as both
passenger automobiles and pickup trucks. In Los Angeles, it appears
that most pickup trucks are used in essentially the same manner as
10-14
-------
personal automobiles. The survey asked whether each reported vehicle
was capable of "long-distance commuting"; all but 3 percent of vehicles
were included in this category.
In Los Angeles there are essentially as many cars as drivers. On
the survey day, 88 percent of all drivers reporting trips came from house-
holds with as many or more cars than drivers reporting trips. Thus in
the great majority of cases, driver travel was not constrained by unavail-
ability of a car.
This important point is the key to deriving useful results from a
survey which did not report which vehicle was used for each trip. Essen-
tially, it implies that driver travel and vehicle travel were similar,
since 88 percent of drivers had vehicles available to them. There is no
absolute assurance, of course, that drivers all used available vehicles,
rather than waiting to take turns on a lesser number of preferred
vehicles. Nevertheless, this seems likely to have been the case.
After computer processing of survey trips, a substantial adjustment
was introduced to make results more realistic. Adjustment is necessary
for two reasons: first, because airline distances rather than over-the-
road distances, were developed in the program; and second, because com-
parisons with other data indicate that in the survey itself, respondents
neglected to report a substantial amount of their actual travel.
Evidence of under-reporting is presented in Table 4.3 which shows
the discrepancy between survey results and independent control data with
which the results were compared. The first four characteristics noted
in Table 4.3 are all modestly under-reported, in approximately the same
amount, as might be expected in an origin-destination survey. The cor-
ridor checks and vehicle miles of travel, however, are very much lower
than might be expected. Original screen line crossings—counts of
vehicle movements across two lines bisecting the study area from north
10-15
-------
o
TABLE 4.3
COMPARISON OF LOS ANGELES 1967 HOME-INTERVIEW SURVEY RESULTS WITH INDEPENDENT CONTROLS (from Ref. 5)
Characteristic
Population
Housing Units
Auto Ownership
*
Resident Cordon Check
**
Screenline Crossings
Ventura/L.A. boundary
L.A./San Bdo./Riv. boundary
Corridor Checks
(average of 7 corridors)
Vehicle Miles of Travel
Total
Freeways Only
Discrepancy
-5.7%
-4.0
-8.2
-7.4
-6.4
-6.8
-20.2
-18.6
-16.8
Control
Data from various government agencies
Data from various government agencies
Dept. of Motor Vehicles registration data
External survey
Actual ground count
Actual ground count
National Highway Functional Classification
Study
1967 Annual Traffic Census
**
Drivers entering and leaving the study area were interviewed as they crossed a cordon line sur-
rounding the area, to obtain information on travel by persons non-resident in the area. Data from
resident drivers obtained at the cordon was compared with external trip data from the home inter-
view survey.
*
Discrepancies are after survey adjustment; see text for explanation.
-------
to south—were apparently also very much lower; values in Table 4.3,
according to LARTS personnel, apply to screen line counts after adjust-
ment, even though no adjustment is mentioned in the text. All other dis-
crepancies in Table 4.3 are before adjustments.
The LARTS adjustment of survey results was accomplished by increas-
ing numbers of reported trips by up to 80 percent, according to trip type,
with overall upward adjustment of trip numbers equal to about 30 percent.
The trip types increased most were those judged most likely to be neglected
and under-reported in a survey; work trips, which presumably are unlikely
to be forgotten, were not increased at all.
In processing of the LARTS data tapes for this study, individual
adjustment of trip types was not feasible. Accordingly, the total number
of trips was simply increased by 30 percent. In consequence, mileage dis-
tributions for individuals and drivers were uniformly increased 30 percent.
To account for over-the-road routing rather than airline distances
between zones, total mileages were additionally adjusted upwards by 23
percent. This figure was chosen to make the adjusted average trip length
*
equal to that in LARTS network model runs. Furthermore the 23-percent
adjustment is in reasonable agreement with a simple analysis. If trips
are made between points randomly selected in a rectangular street grid,
the average over-the-road travel distance will be greater than the air-
line distance by a factor of A/TT, a 27-percent adjustment. In the real
world, however, it seems likely that trips will not be uniformly distri-
buted in direction, and that in addition, there will be important
diagonal streets and freeways reducing travel distances which otherwise
would be required in a rectangular grid.
*
As reported in Ref. 2, the home-interview survey is the basis for con-
tinuing computer analyses and forecasting. The principal computer out-
puts are trip production runs, summarizing and projecting trip-making,
and network model runs, showing trip routings and total traffic levels
on roadways which might result.
10-17
-------
The total adjustment applied in this study is thus +60 percent; +23
percent in individual trip distances due to actual rather than airline
routings; and 30 percent in number of trips taken, due to apparent under-
reporting of trips in the survey. After adjustment, the processed sample
of travel in Table 4.1 appears in reasonable agreement with other analyses
and data, as indicated in Table 4.4. The miles per trip agree not only
TABLE 4.4
ADJUSTED SAMPLE AVERAGES
Travel
Characteristic
Adjusted Comparable
Sample Value Source of Comparison
Miles per trip 5.8
Trips per car 4.9
Miles per car 28.6
5.8
5.7
4.6
3.7
27.6
24.9
21.0
Network Model Run
Kalish7
Network Model Run
Kalish7
Trip Production and Network
Model Runs
DOT Highway Statistics -
Kalish7
1969
with the LARTS network model run, but very closely with the average travel
distance used by Kalish in the previous study of daily vehicle usage.
The number of trips per car is moderately higher than that of the network
model run, as might be expected since in the sample, cars which were not
used on the survey day were dismissed from this average. About 7 percent
of cars were in this category; if they were included, the trips per car
after adjustment would be very close to that of the network model run.
The lower value for trips per car reported by Kalish may be explained
10-18
-------
by its derivation from data in another city (Chicago) some years ago
(1967). Overall, the average daily car mileage is reasonably close to
that implied by the LARTS models and DOT statistics. The adjusted sample
differs from the DOT figure because cars which were not used on the survey
day are not included, and because long-distance travel is not included in
the sample. These omissions tend to compensate one another, but not com-
pletely. According to the National Travel Survey of 1967, there were
3.6 annual long-distance or overnight trips per vehicle, on the average,
of about 240 miles each, amounting to 2.4 miles per day which is not in-
cluded in the adjusted sample mileage. The omission of unused cars from
the average, on the other hand, means the adjusted sample is a little
over 2 miles per day higher than it would otherwise be.
The effects of the adjustments of the survey distributions are shown
in Fig. 4.1. The upper curve in this figure shows the cumulative total
number of drivers in the survey who drove less than the indicated mileage
on the survey day, before any adjustment. The lower curve shows the
results of the 23-percent adjustment for road rather than airline mileage,
and the 60 percent total adjustment to compensate for under-reporting in
addition. Also shown in Fig. 4.1 are the daily mileage capabilities for
the particular electric cars of this study. The adjustment is consider-
able: based on the raw data, the four-passenger lead-acid battery car
would be adequate for 93 percent of the daily mileage; but based on the
adjusted data, it would be adequate for only 83 percent.
For comparison, the result of Ref. 7 is included in Fig. 4.1, for
cars driven 12,000 miles per year. Since that study included overnight
and long-distance travel, it is to be expected that it would deviate
increasingly with distance from the other curves of the figure. Near
the lower electric car ranges, however, and in the vicinity of most
daily travel distances, it is in reasonable agreement with the adjusted
distributions of Los Angeles mileages. This may be surprising in view
of the basis of this previous study, which assumed simple mathematical
10-19
-------
99.9
99.8
99.5
99
95
LU
S 90
•z
o
OC.
£ 80
LlJ
;>
s
i 60
o
40
20
^ (REF. 7)
-"I (CARS DRIVEN 12,000
miles/year OR LESS)
TWO- FOUR-
PASSENGER PASSENGER
LEAD-ACID BATTERY
CAR RANGES
ADVANCED BATTERY
CAR RANGES
I I I I I
10
20
40
60
80 100
200
DISTANCE, mi
Figure 4.1. Adjustments of Surveyed Daily Mileage Distributions
forms for the distributions involved in its synthesis, and for distribu-
tion parameters relied on 1956 Chicago data.
The basic adjusted distributions of daily travel distances from
the Los Angeles survey are shown in Fig. 4.2. This figure shows the most
important of several categories of travel for which distributions were
produced. The first of these is for the daily travel distance of drivers
who had cars available to them without interference on the survey day—
drivers, that is, from households reporting at least as many cars as
drivers on the survey day. The other distribution shown is for the daily
travel of a single car in households reporting one car driven by two dri-
vers on the survey day. In such instances, it is to be expected that the
travel desires of two drivers would cause the car to be used more than a
single driver might use it, but less than two separate cars. This is the
case: cars used by two drivers, typically travel 50 to 80 percent fur-
ther in a day than cars used exclusively by single drivers.
10-20
-------
UJ
99
98
95
90
80
£ 60
§
§ 40
o
20
2 DRIVERS,
1 CAR
DRIVERS WITH CARS
TWO- FOUR-
PASSENGER PASSENGER
LEAD-ACID BATTERY
CAR RANGES
ADVANCED BATTERY
CAR RANGES
I
10
20
40 60
DISTANCE, mi
80 100
200
Figure 4.2. Adjusted Distributions of Daily Travel, Los Angeles Region, 1967
Distributions were also developed for other cases, such as indi-
vidual drivers in 1-, 2-, and 3-car households. They are not much dif-
ferent, however, from that for drivers with cars in Fig. 4.2.
If several drivers using a single car were common in Los Angeles,
the daily range requirement for electric cars would be substantially in-
creased and consequently much more difficult and expensive to meet.
Such, however, is not the case. Processing of the 1967 Los Angeles survey
shows that very few drivers did not have cars available on the survey day.
And with increasing automobile ownership rates, unavailability of cars
to drivers will be even less frequent in the future. Table 4.5 shows
that on the survey day 88 percent of drivers had cars available to them.
Thus it seems reasonable to employ distributions of daily travel for these
drivers as representative of what electric cars will be required to do
in the future, when individual cars for drivers will be even more gene-
rally available.
10-21
-------
TABLE 4.5
SURVEY AVAILABILITY OF CARS TO DRIVERS
Los Angeles Area
Percent of area drivers with cars available to them
in 1-car households 26%
in 2-car households 46%
in 3-car households 13%
in 4-car households 4%
in all households 88%
Percent of area drivers without cars available to them
in 1-car households 7%
in multi-car households ... 5%
in all households 12%
In conclusion, it should be noted that a basic assumption is re-
quired to make survey results useful. This assumption is that the distri-
bution of daily mileages for all days in the life of a single typical
car is the same as the distribution of daily mileages for the survey
sample of cars on a single day. It will be implicit in the following
analysis.
10-22
-------
5 APPLICABILITY OF ELECTRIC CARS
The basic limitation on the applicability of electric cars is their
daily range capability. A limited-range car is not really applicable to
the needs of a driver if it frequently cannot go as far as he might wish
during a single day. On the other hand, it is not necessary to insist
that the electric be able to do anything that its gasoline counterpart
might, nor satisfy all a driver's needs on every day of his life. Any
compromise definition of applicability is, of course, arbitrary, but it
seems safe to say that applicability will require adequate range for the
great majority of the driver's travel days.
In the study, capability adequate for 95 percent of driving days
has been adopted as a definition of applicability. Figure 4.2 shows that
under this definition, the "advanced-battery" cars are applicable to the
needs of drivers in general, 98 percent of whom travel less than the cars'
ranges on a typical day. Furthermore, these cars are applicable to the
needs of two drivers sharing a single car at a household. The "lead-acid
battery" cars, on the other hand, are not applicable under this definition
to the travel of the average driver.
Despite its range limitation, the four-passenger lead-acid battery
car is applicable to the role of second car in a two-car household, so
long as the second car is defined to be that car used less on each day.
It may be assumed that in a two-car household, the probability of long-
distance travel by one car on a given day is independent of that for the
other car. In this case, Fig. 5.1 shows the probability that the second
car in a two-car household will be driven less than the indicated range,
or that the second and third cars in a three-car household will be driven
less. On 97 percent of the days, the four-passenger lead-acid battery
car would be capable of the travel demanded of the lesser-used car in the
two-car household. On only 91 percent of the days, however, could two
of these cars perform the functions of both secondary cars in a three-car
household, which falls short of the adopted applicability threshold.
10-23
-------
99.8
99.5
99
98
95
S 90
LU
I 8°
s
3 60f-
40
20 -
10
THREE-CAR
HOUSEHOLD
TWO- FOUR-
PASSENGER PASSENGER
LEAD-ACID BATTERY
CAR RANGES
ADVANCED BATTERY
CAR RANGES
I 1 I I I
20
40 60 80 100
DISTANCE, mi
200
Figure 5.1. Probability of Daily Driving Less Than a Given Distance
for Secondary Drivers With Cars
In practice, of course, applicability of an electric car to a driver's
needs presumes overnight recharge facilities. Unless such facilities can
reasonably be provided, the car cannot be considered applicable even if
its range is adequate. To investigate the provision of overnight charg-
ing, the LARTS 1967 travel survey tapes were also processed to show the
kinds of parking available by household type. Summary results appear in
Table 5.1.
Overall, only 74 percent of cars in 1967 had off-street parking.
Cars parked overnight on the street are obviously poor candidates for
recharge, which requires electric power at levels usually met only from
220-volt outlets. Where the electric car is to be one of several cars
at a household, however, all that is necessary is that the household have
at least one off-street parking space, and a larger number of households—
87 percent—fall in this category. Not every off-street parking space,
however, is equally amenable to recharge; in multi-family residences with
large parking lots, provision of 220-volt, individually metered outlets
for recharging could be a significant problem.
10-24
-------
TABLE 5.1
SURVEY AVAILABILITY OF PARKING, LOS ANGELES AREA
Percent of All Area Households with Off-Street Parking
1-Car Households
Multi-Car Households 49%
Total 87%
Percent of All Area Cars with Off-Street Parking
At 1-Car Households 22%
At Multi-Car Households 52%
Total 74%
Percent of Single-Family Dwellings with Off-Street Parking 89%
Percent of Other Dwellings with Off-Street Parking 83%
Accordingly, the best candidates for electric car recharging are
single-family households with off-street parking. As Table 5.1 shows, about
89 percent of such dwellings have at least one off-street parking space.
The number of automobiles for which the four-passenger 1980 electric
is applicable would thus be the same as the number of single-family house-
holds with two or more cars and off-street parking. To determine this
number, survey results and projections by LARTS may be employed. These
necessary basic data appear in Table 5.2.
To apply the data of Table 5.2, it is necessary to scale the 1990
figures to the population forecast of this study and to interpolate for
specific years of concern. This is done in Table 5.3. The first three
lines follow from Table 5.2, and the population and number of cars already
used in this study are shown in the next two lines. The remaining lines
of the table show the number of single-family housing units with cars,
10-25
-------
TABLE 5.2
DISTRIBUTIONS OF HOUSING UNITS AND CARS, LARTS AREA
1967
Total Housing Units, Thousands 3,077
Single-Family Housing Units, Thousands 1,956
With Cars 1,715 2,437
With Two or More Cars 979 1,409
Total Cars, Thousands 4,322 7,439
Cars at Single-Family Housing Units, Thousands 2,264 4,630
Population, Thousands 9,019 13,446
SOURCE: Los Angeles Regional Transportation Study, Trip Production Model
Runs, Tabs 400260-4, 4005-4-7, 400603.
and with two or more cars; and the total number of cars at single-family
housing units.
According to Table 5.3, 1,140,000 single-family housing units will
have more than one car by 1980 in Los Angeles. If 89 percent of these
have some off-street parking, as indicated for all single-family units in
Table 5.1, then electric cars will be applicable in 1980 at slightly over
one (-million Los Angeles households. Though this implies applicability to
the roles of only 17 percent of all Los Angeles automobiles in that year,
it is still a very large number on an absolute basis, especially since
the standards of applicability involve minimum sacrifice and inconvenience
on the part of the drivers and households.
The 1990 electric car, as noted previously, is applicable to the
daily travel of most drivers. Recharging problems remain, however, so
applicability will still be limited to households with ready recharge
10-26
-------
TABLE 5.3
DISTRIBUTION OF HOUSING UNITS AND CARS, SOUTH COAST AIR BASIN
1970 1980 1990 2000
Persons Per Housing Unit 2.94 2.81 2.77 2.70
Percent of All Housing Units Which Are
Single-Family Housing Units
With Cars 55.0 52.7 50.3 47.9
With Two or More Cars 31.5 30.3 29.1 27.9
Percent of All Cars at Single-Family
Housing Units 53.7 58.0 62.2 66.5
Population, Thousands 9,700 10,600 11,600 12,400
Cars, Thousands 5,060 5,880 6,730 7,600
Single-Family Housing Units, Thousands
With Cars 1,840 1,980 2,110 2,200
With Two or More Cars 1,050 1,140 1,220 1,280
Cars at Single-Family Housing Units, 2,720 3,400 4,190 5,060
Thousands
capability. In this case, single-family households are again the most
promising for recharging. Assuming that 74% of cars at single-family
households have off-street parking, in accord with the area-wide figure
of Table 5.1, electric cars would be applicable in over three million
cases in 1990, to 46 percent of all individual automobiles in the area.
In the longer term, as electric cars come into general use, we may expect
that recharge provisions would be made in the off-street parking provided
by multi-unit buildings. Thus by the year 2000, electric cars might be
applicable at all off-street parking places. If the current 74-percent
rate continues to prevail, this would make electric cars applicable in
5,600,000 cases, to 74 percent of the total car population.
10-27
-------
Once the advanced-battery cars arrive, daily range of electric cars
would be so great that average daily driving of electrics would be essen-
tially the same as average driving of cars without range limitations.
Thus in 1990 and 2000, the fraction of daily vehicle miles by electric
cars would be essentially equal to the fraction of electric cars in the
total car population. In 1980, however, electric cars with limited ranges
will constitute most of the electric car population. They can be assigned
only the less-demanding day-long travel patterns, so the percentage of
electric car vehicle miles in 1980 will be considerably less than that
of cars in the total population.
The average usage of such cars can be approximately deduced by inte-
grating the mileage distributions produced by the computer processing of
the Los Angeles travel survey. The normalized result of this integration
is shown in Fig. 5.2. The 1980 lead-acid electric car is adequate for
100
80
60
40
-------
83 percent of drivers on a typical day. Figure 5.2 shows that the average
travel mileage of that 83 percent of drivers is about 63 percent of average
travel for all drivers.
This result and the previous estimates of applicability are sum-
marized in Table 5.4.
TABLE 5.4
CANDIDATES FOR ELECTRIC CAR REPLACEMENT, SOUTH COAST AIR BASIN
Cars, Thousands
Percent of Total
Daily Vehicle Miles, Millions
Percent of Total
1980
1,011
17
18
11
1990
3,099
46
90
46
2000
5,624
74
169
74
10-29
-------
6 CONSUMER MARKET FOR ELECTRIC CARS
The preceding section has established the broad applicability of
electric cars in Los Angeles, even with lead-acid batteries and prospec-
tive near-term technology. As early as 1980, such cars could replace
over a million automobiles in Los Angeles with very small sacrifice in
mobility.
This section addresses not what electric cars could do, but what the
cars of Table 3.2 are likely to do in a free market operating essentially
as at present—that is, a marketplace in which electric cars have not
been relatively favored either by inducements to their purchase and use,
or by penalties on the purchase and use of competing gasoline cars. In
such a market, it is consumer preferences and perceptions which, applied
to competing electric and conventional automobiles, will decide the sales
and usage of the electric cars.
Unfortunately, conclusive analysis of trends and prospects in the
consumer market for electric cars is not now possible. Almost no data
is available which is directly applicable, since essentially no electric
cars have been sold in either Los Angeles or the nation. Thus there is
no avoiding major qualitative inferences in moving from existing data to
sales and use estimates for the future.
This section first considers the competitive price and performance
situation for future electric cars, which is generally unfavorable. Then
it reviews overall market trends, which appear relatively favorable to
electric automobiles. Next, it turns to relevant results of consumer
surveys, which elicit consumer intentions as the best available alterna-
tive to actual observations of consumer electric-car purchases. Finally,
potential free-market penetration and use estimates for the future are
described.
10-30
-------
6.1 PERFORMANCE AND PRICE COMPETITION
Because energy stored in lead-acid batteries is rather like gasoline
weighing 500 pounds per gallon, electric cars must be carefully designed
for energy efficiency. In consequence, as pointed out previously, elec-
tric cars in the lead-acid battery era are unlikely to be designed for
acceleration and speed performance comparable to those of the conventional
automobile. Moreover, electric cars promise to be more expensive initially
than their conventional counterparts, and to be more expensive in total
per-mile costs so long as lead-acid and nickel-zinc batteries are utilized.
On these topics—the relative importance of performance, first cost,
and overall costs in determining market shares—there is a considerable
body of data for existing cars. It can be reviewed to investigate whether
lower performance or higher price is demonstrably detrimental to sales.
Table 6.1 presents recent performance, cost, and market data for all
subcompact cars holding more than 1 percent of the market in the South
Coast Air Basin of California. Performance is represented by advertised
horsepower per ton of curb weight. Also shown are portions of the total
US subcompact market held by each of the various entries in the table.
In Fig. 6.1, market shares from this table are plotted for the most impor-
tant makes and models versus horsepower per ton and versus price.
If there are any general relationships among market share, perfor-
mance, and price in this data, they are certainly not obvious. Instead,
review of specific figures generally raises questions which are difficult
to answer. Overall, it would be exceedingly difficult to argue from this
data that price and performance as measured here are the major determi-
nants of market share.
i
The Vega and Pinto, for example, are similar automobiles: they cost
about the same, offer about equal horsepower per ton, and are similar in
appearance. Moreover, they were introduced at about the same time, both
10-31
-------
TABLE 6.1
SALES OF 1972 SUBCOMPACT CARS
(From Ref. 1)
Share of Subcompact
Market, Percent
Make
Ford
Volkswagen
Chevrolet
Ford
Mazda
Toyota
Datsun
Chevrolet
Lincoln-Mercury
Datsun
Toyota
Ford
Toyota
Volkswagen
Dodge
American Motors
Toyota
Honda
Buick
Lincoln-Mercury
Volkswagen
Porsche
Toyota
M ,i~i South Coast
Model Air Basin US
Pinto (86 hp)
Beetle/Superbeetle
Vega
Pinto Wagon
RX Series
Corolla
PL-510
Vega Wagon
Capri 2000
1200
Mark II
Pinto (54 hp)
Corona
Fastback, Squareback
Colt
Gremlin
Celica
600
Opel 1900 Series
Capri 2600
411 Sedan and Wagon
914
Corona
15.5
14.6
7.6
7.3
5.0
4.9
3.6
3.5
3.4
3.3
2.7
2.3
2.2
2.0
2.0
1.9
1.8
1.7
1.6
1.5
1.3
1.3
1.1
11.7
16.9
14.4
5.7
2.0
5.1
3.4
2.8
2.9
3.1
2.8
1.7
2.3
2.3
1.6
4.4
1.9
1.0
2.2
1.3
1.5
0.7
1.1
„ * Advertised Horsepower
Price Per Ton
2070
1999/2249
2060
2265
2800-3100
1955-2300
2304-2656
2285
2528
1976-2116
2310-2430
1960
2385-2532
2550-2750
2243-2451
1999-2153
2847
1473-1610
2433-2656
2821
3280
3900
2261
84
47
77
72
83
75-88
90
67
80
87
84
53
89
45
79
78-106
86
48
69
90
62
89
92
Retail delivered price, exclusive of transportation charges, taxes, and accessories.
10-32
-------
15r
(D
S">
O.
PINTO
VW BEETLES
0 VEGA
PINTO WAGON*
MAZDA RX
CAPRI 20.00
VEGA WAGON* ••
DATSUN 1200
COROLLA
DATSUN PL-510
1
1
1
25 50 75
HORSEPOWER PER TON
100
c
Ol
Ol
CL
Q:
OO
10
1
PINTO*
VW BEETLES
VEGA
PINTO WAGON
TOYOTA _CO_RO_LLA
DATSUN L2_00DAIS-UN-^-5-l°
MAZDA RX
VEGA
WAGON
CAPRI 2000
1500 2000
PRICE, dollars
2500
3000
Figure 6.1. Shares of the Subcompact Market Relation to Price and
Performance, South Coast Air Basin, 1972
10-33
-------
through impressive dealer organizations, by two of the nation's (and the
world's) largest corporations. One might consequently expect their sales
to be approximately equal. Yet in the Los Angeles area during 1972, the
Pinto outsold the Vega by a factor of 2 to 1, while nationally the Vega
substantially outsold the Pinto. The determinants of these disparate
market shares are by no means obvious. In any event, it is certain that
in comparison with the electric four-passenger compacts, the Pinto and
Vega appear virtually identical; yet their sales are very different in
one direction in the South Coast Air Basin and in the opposite direction
elsewhere. The necessary conclusion is that the big differences in sales
between these cars must arise in relatively minor and even regional fac-
tors of consumer perception and motivation, since it is certainly not
obvious in price or performance data, in general accommodations and styl-
ing offered, or in marketing muscle and expertise.
Other small mysteries are easy to find in Table 6.1. The Mazda,
renowned for its performance, appears relatively ordinary in terms of
advertised horsepower per ton. Perhaps this figure is no more than a
poor beginning in describing performance. The American Motors Gremlin
is notably less popular in the South Coast Air Basin than in the nation
at large; is there something special about Los Angeles consumers, and if
so, what?
In the context of Table 6.1, the four-passenger electric cars of
this study are not winners in either the price or performance category.
At 48 horsepower per ton, they fall near the bottom-ranked VWs; and at
prices of $3,400-$7,300 (see Table 3.2) they are at or well above the
*
range of prices in Table 6.1.
As discussed in Ref. 1, the electric motor of the characterized electric
cars can operate safely at 2-1/2 times its continuous-duty horsepower
for short periods of acceleration. The figure of 43 hp/ton applies only
to these periods. The motor controller will limit motor power input to
rated levels; otherwise they might be considerably exceeded in full-
throttle driving.
10-34
-------
Nevertheless, the market position of the electric cars is by no
means hopeless on these accounts. In Table 6.1, and in Fig. 6.1, the
Pinto and Volkswagen Beetles stand out as top sellers, in a sales class
by themselves. Their 1972 prices were similar. The Beetles, however,
provided a horsepower per ton little more than half that of the Pinto.
Evidently a substantial market share can be obtained despite low power
and consequent minimal performance. So far as price is concerned, the
Mazda suggests that unique power plants may justify a price premium
approaching 50 percent. Mazda purchasers may be seeking principally
flashing acceleration, but Mazda advertising has emphasized smoothness,
simplicity, and reliability as much as sheer performance from a standing
start. These latter advantages, of course, are commonly expected of
electric propulsion in even greater measure, suggesting that modern elec-
tric cars may command higher prices in the market on that account.
6.2 OVERALL MARKET TRENDS
Much of the literature on the automobile market is concerned with
total automobile sales rather than relative shares for different makes
and models. Whether based on consumer surveys or sales observations,
however, it clearly spells out trends generally favorable to the intro-
duction of electric cars.
Auto sales by class have long shown declines in the market share
of the larger automobiles, with a rapid rise at the subcompact end of the
size, performance, and cost scales. As previous investigators have con-
cluded, it is impractical for battery-electric cars to compete head-on
with full-size automobiles, but it is possible to compete with subcompact
automobiles—or at least to approach their accommodations and performance
(excepting daily range). Thus consumers increasingly have come to prefer
the type of automobile most amenable to electric propulsion.
Leading authorities in consumer research see elaborate related
12
trends in consumer behavior. "Studies conducted ... in the late 1960s
10-35
-------
indicate that the car has increasingly become a means for serving important
ends, rather than the highly prized possession it once was ..." "...
Frequently, two different types of cars were wanted. Long distance family
travel—and probably also prestige, although this was rarely mentioned by
survey respondents—called for the ownership of a large car. But even
prestige obviously did not require that the second car be large ..."
"Interviews revealed that the major problems confronting consumers no
longer concerned the car purchase itself—the make, type, or price of the
car bought. Their concerns had become centered around accessories, and,
more particularly, around problems that arise after the purchase . . .
repairs had become a major problem, not only because of their high cost
and often inconvenient timing, but also because of the difficulty in
getting them done properly".
The simplicity, reliability, and minimal maintenance requirements
4
of electric propulsion should make electric cars increasingly more desir-
able as these trends persist. In addition, the electric car can offer a
unique advantage which the conventional subcompact cannot: through its
novelty and its high social desirability in fighting pollution, it can be
a prestige car despite its modest dimensions.
Particularly in the Los Angeles region, prospects are favorable
for small and novel automobiles. The Los Angeles area has been half a
decade ahead of the nation in adopting the subcompact and the imported
car: in 1972, subcompacts had almost 35 percent of the Los Angeles auto
13
market as compared with little over 20 percent nationally. Moreover,
in Los Angeles the air pollution problem has long been acute and interest
in minimizing pollution is high.
6.3 CURRENT SURVEYS OF MARKET POTENTIAL
There have been two recent surveys of consumer buying inclinations
regarding electric cars. Consumer intentions are not always reliable,
of course, and where the product is not presently in the market or well-
known to the consumer, the hazards are obvious. Nevertheless, 'in the
10-36
-------
absence of sales data in which true consumer preference is revealed, re-
ported intentions are an expedient indication of the sales potential of
electric cars.
Investigators at the University of Wisconsin-Whitewater made a con-
14
sumer survey specifically related to electric car purchase in mid-1973.
The extent and location of consumer interviews are shown in Table 6.2.
The surveys addressed factors of importance to consumers choosing a auto-
mobile, with rank order results as indicated in Table 6.3. These findings
give specific support to general trends and consumer preferences already
noted. In particular, they illustrate relative lack of consumer interest
in capacity, size, acceleration, and speed of automobiles—precisely
those attributes which once were considered to make the "family car"
immune to electrification. Among factors high in rank in Table 6.3, the
electric car can be as safe as other cars of its size, good in mileage or
efficiency, and superb so far as maintenance and pollution are concerned;
but it will be more expensive, especially so long as it must rely on lead-
acid and nickel-zinc batteries.
Furthermore, of course, electric cars will have a severe range limi-
tation until the advanced batteries become available. The consumer survey
plainly brings out the importance of this problem. Table 6.4 presents a
TABLE 6.2
ELECTRIC CAR CONSUMER SURVEYS, APRIL-MAY, 1973 (FROM REF.14)
Place Survey Responses
Milwaukee, Wisconsin 556
Madison, Wisconsin 332
Chicago, Illinois 341
Total 1,229
10-37
-------
TABLE 6.3
SURVEY FACTORS REPORTED IMPORTANT IN AUTO CHOICE (FROM REF.14)
Factor Rank
Safety 1
Cost of Operation 2
Price 3
Mileage 4
Cost of Maintenance 5
Pollution 6
Ease of Maintenance 7
Passenger Capacity 8
Size 9
Acceleration 10
Luggage Capacity 11
Speed 12
description of a possible electric car offered to survey respondents.
Among nine other attributes, under the label "performance" it notes a sub-
stantial range capability at different constant speeds (which tends to be
unrealistically high relative to actual stop-start urban driving). Respon-
dents were asked which of these electric car attributes they would most
like to see changed, with results shown in Table 6.5. Increased range
was mentioned far more frequently than any other improvement; it was fol-
lowed by rechargeability, .which amounts to another means of extending
range. Regardless of the capability of electric cars for accomplishing
daily driving, as noted in Sec. 5 of this paper, consumers may not wish
to purchase cars of limited range.
In the end, it must be recognized that the consumer purchases oppor-
tunity when acquiring an automobile, regardless of actual need for long
daily range. If the consumer feels it important to be able to take long
10-38
-------
TABLE 6.4
ELECTRIC CAR DESCRIPTION IN SURVEY (FROM REF. 14)
1. Meeting 1976 safety standards.
2. Either two- or four-passenger capacity.
3. Baggage capacity of 100-200 pounds (similar to a Pinto or Vega).
4. Heater. Radio.
5. Size: Subcompact or four-seater/station wagon (hatchback)
6. Performance
at 30 mph the range is 95-100 miles
40 75-80
50 60-65
60 50-55
7. Fuel cost of 0.5 cent per mile compared to 2.5 cents per mile for
an internal combustion engine (equivalent to 55 miles per gallon).
8. Batteries recharge fully overnight or to 80 percent of capacity with
a special quick-charge device taking 45 minutes.
9. 1975-76 cost of $2,500 compared to Pinto/Vega cost of $3,000.
TABLE 6.5
IMPROVEMENTS MOST DESIRED IN ELECTRIC CAR (FROM REF. 14)
1. Performance (Range) (38%)
2. Rechargeability (22.4%)
3. Size (11.8%)
4. Passenger Capacity (10.4%)
5. Baggage Capacity (4.6%)
10-39
-------
trips in the automobile, even though he may never do so, then limited-
range electric cars will appear less attractive.
The Wisconsin surveys asked consumers to rate their intention to
buy the electric car of Table 6.4, when available, on a scale from 0 to
10. The mean response was 4.5 for Milwaukee, 4.7 for Chicago, and 5.0
for Madison. Ten to 15 percent of respondents expressed zero intention
to purchase such a car, while 6 to 10 percent expressed the highest
intention.
This considerable interest in buying an electric car must, of course,
be qualified to the extent that the car described in Table 6.4 is imprac-
tical or infeasible. Relative to the lead-acid battery car characterized
in this study (see Table 3.2), the description of Table 6.4 is overoptimis-
tic in the following respects:
• It states a purchase price substantially less, not more, than
the prices of conventional subcompacts.
• It does not mention the high cost of battery depreciation,
but instead cites only a very low fuel cost, with the clear
implication that total operating cost will be less, not more,
than that of conventional subcompacts.
• It does not mention electric car acceleration, which by impli-
cation may be taken as comparable to, not much less than,
the conventional subcompacts.
• It mentions a heater but does not note that it would probably
be considerably less effective than those of conventional
subcompacts, or else penalize range severely.
In the other direction, the description of Table 6.4 does not mention
prospective advantages of electric cars in quietness or in reduced main-
tenance. Nevertheless, on balance it appears that overall consumer in-
tentions to purchase lead-acid battery cars characterized in this study
might be substantially lower than determined in the Wisconsin survey.
10-40
-------
A study of the market potential of electric cars was conducted in
1971 for the Copper Development Association and the Electric Vehicle Coun-
cil of the Electric Energy Association. The basis of the study was an
analysis of electric car applicability which has already been discussed
in Sec. 3. Overall findings of the study are reproduced here in Table 6.6,
indicating the potential population of both trucks and cars, given low and
high estimates of electric car capability. In the case of passenger cars,
the estimates span a potential market range of 50 to 1, and are conse-
quently so broad as to provide little specific guidance here.
In a nationwide survey of consumer interest in an electric car pur-
chase, the Electric Vehicle Council and the Copper Development Association
obtained results shown in Table 6.7, An electric car with range of 150
miles (conditions unstated), speed of 40 mph, and price of $2,000 was
briefly described to a representative sample of over 2,000 interviewees,
who were finally asked whether they would be interested in purchasing such
a car. The results show, as did the Wisconsin survey, a large minority
interested in purchasing the electric car, with interest concentrated among
those having high income, education, and family car ownership, and profes-
sional or managerial occupations.
TABLE 6.6
POTENTIAL US ELECTRIC VEHICLE POPULATION
(From Ref. 7)
Thousands of Vehicles
Light Trucks Passenger Cars
Year Low High Low High
1970 37 534 427 20,000
1975 45 661 523 24,600
10-41
-------
TABLE 6.7
INTEREST IN ELECTRIC CAR PURCHASE, MARCH-APRIL 1972 (FROM REF. 15)
(150 mi range, 40 mph speed, $2,000)
Total US Public
No Car
One Car
Two Cars
High School Incomplete
High School Complete
Some College
Under $5,000 Family Income
$5,000-$6,999
$7,000-$9,999
$10,000-$14,999
$15,000 or Over
Yes
42%
41
39
46
34
41
59
35
38
40
47
52
No
52%
46
54
50
57
56
35
55
57
54
49
44
No Opinion
6%
13
7
4
9
3
6
10
5
6
4
4
The Winconsin investigators proceeded from their consumer survey to
estimates of the consumer market for electric vehicles. They concluded
that the "primary prospect is an urbanite who owns a 'new1 second car,
earns $10,000, and may have some college education . . . there are some
700,000 households in the US that meet these criteria." They also identi-
fied a miscellaneous secondary market among non-owners of cars, rural
households, and multiple electric car households, and a major secondary
market among urban owners of older second cars and urban single car house-
holds. The total potential market and an annual sales potential for an
assumed replacement period are shown for these and industrial/institutional
markets in Table 6.8. For lack of complete data, the annual potential
is necessarily a subjective estimate.
10-42
-------
TABLE 6.8
ESTIMATED NATIONAL POTENTIAL MARKET FOR ELECTRIC CARS IN 1974
(From Ref. 16)
Potential Market, Thousands
Total
Consumers
Primary
Miscellaneous Secondary
Major Secondary
Industrial-Institutional
Fleet
Non-Fleet
Total
Low
500
200
600
210
90
1,600
High I
700
400
1,400
300
210
3,010
Replacement
•eriod, Yrs
2
4
4
3
3
Annual
3 Low
250
50
150
70
30
550
High
350
100
300
100
70
920
To translate market potential into estimated market penetration by
year, percentage penetrations were subjectively estimated and applied by
the Wisconsin investigators. The results are shown in Table 6.9 for the
years 1975 through 1980. Also shown in Table 6.9 are percentage penetra-
tions of an assumed total car market growing at the rate previously pro-
jected in this study for Los Angeles.
By 1980, the Wisconsin estimate expects 1- to 2-percent penetration
of the electric car into the car market. If this prevailed in the Los
Angeles area, it would amount to some 5,000 to 10,000 electric cars sold
annually. As noted earlier, these figures may be high owing to the opti-
mism of the electric car description.
10-43
-------
TABLE 6.9
ESTIMATED NATIONAL MARKET PENETRATION OF ELECTRIC CARS TO 1980
(From Ref. 16)
Annual Potential
Market, Thousands
Year
1975
1976
1977
1978
1978
1980
Low
600
650
700
750
825
925
High
1000
1100
1200
1300
1450
164G
Market Penetration
Percent
1.0
1.5
2.5
5.0
10.0
15.0
Thousands
Low
6
10
15
38
83
139
High
10
17
30
65
145
246
Penetration of Total
Car Market, Percent
Low
0.06
0.1
0.14
0.33
0.70
1.1
High
0.1
0.16
0.27
0.57
1.2
2.0
6.4 ESTIMATED MARKET POTENTIAL IN LOS ANGELES
Despite the many difficulties, this study requires projections of
potential electric car market penetration and use for the Los Angeles
area. In this section, the basic factors involved are summarized and
projections then offered. Because both the circumstances and the electric
cars of this study differ from those of previous work summarized already,
it is inappropriate simply to use the results of others directly.
The electric cars characterized in this study differ from those of
previous consumer surveys and market estimates, at least for the next
decade, primarily in two respects: their costs are substantially higher,
and their range in urban driving is less than that stated or implied pre-
viously. The circumstances surrounding previous projections are not
clearly defined, but in the Wisconsin work at least, conventional car costs
are apparently anticipated to be somewhat higher than electric car costs.
10-44
-------
The baseline projections and forecasts of this study are based on
the explicit assumption that there is not to be a major disruption in
established economic and social patterns. Analyses in the areas of popu-
lation, transportation, energy, and the environment all support this
assumption, in that they bring forth no evidence that there must be some
such major change as gasoline rationing or restriction of electric power
supply and demand. Generally, the lower population growth foreseen in
the study baseline lessens prospective pressure on energy supplies, on
transportation facilities, and on pollution control arrangements. Major
improvements are in store, and are already in progress, in the average
miles per gallon of US automobiles, which will probably reduce the total
annual consumption of gasoline in Los Angeles until near the end of this
century. At the same time, market prices for "new" petroleum are now
higher than those found to be adequate for calling forth maj or increases
in domestic drilling and probable discovery rates, regardless of what
happens internationally. The air pollution which has plagued Los Angeles
more than any other city will be substantially reduced in 15 years, after
sufficient cars meeting 1977 emission standards replace the dirtier cars
now being produced and operated in the Air Basin. Planned electric power
plants will be adequate not only to meet additional growth at a substan-
tial—though reduced—rate, but provide enormous unused capacity in off-
peak hours and seasons.
Despite the 1974 "crisis" in energy supply, in short, despite vex-
ing pollution problems, and despite occasional electricity blackouts,
there is no necessity nor even compelling basis for presuming that such
conditions need persist through the remainder of the century in the Los
Angeles area. In consequence, the baseline projections of this study do
not describe a future in which electric automobiles are assured supremacy
by a succession of natural and economic disasters for the conventional
automobile.
10-45
-------
The electric car has some significant advantages. It promises to
be reasonably efficient and low-polluting, and it is certainly novel, all
of which may well make it a potent prestige and status symbol. These are
psychic benefits, however, and though they may be important, they need
not overshadow the more tangible major benefit of the electric: its pros-
14
pective high reliability and minimum maintenance. Its simplicity, the
long mean times between failure of its motor and controller, and the mini-
mal need for routine maintenance, all promise the electric motorist an
end to service hassles. Quite apart from the status value, an electric
car which might ordinarily require only a simple lubrication and inspec-
tion once a year could have wide appeal.
Allied against these advantages are the higher costs of electric
cars relative to conventional cars and the limited daily range which they
afford. Even if their reduced performance and accommodations are not
deterrent to would-be purchasers, their higher initial and operating
costs are sure to work against them. Moreover, consumers have already
expressed their dissatisfaction with limited range. Both range and price
drawbacks are projected to be reduced with the advent of zinc-chlorine
and lithium-sulfur battery technology after 1985; nevertheless, they will
still be at a substantial handicap. An initial cost disadvantage will
remain, and though the advanced batteries may permit daily ranges far in
excess of those required 99 percent of the time in urban driving, they
will still not permit every trip of which conventional vehicles are
capable. Particularly for family vacations and longer recreational trips,
the range limitation will deter buyers.
The balancing of these advantages and disadvantages must now be com-
pleted subjectively, for lack of additional substantial evidence on con-
sumer behavior.
The initial cost of electric cars characterized in this study, shown
in 1973 dollars in Table 3.2, is substantially higher than the typical
$2,270 cost of the comparable four-passenger internal-combustion subcompact,
10-46
-------
Moreover, the total per-mile cost of the earlier electric cars (with lead-
acid and nickel-zinc batteries) is likely to be 25 percent or more above
the 11.6 cents per mile (in 1973 dollars) of the conventional subcompact.
These combined cost disadvantages are so great that purchase by only a
few percent of potential buyers seems likely: by 1980 we estimate the
lead-acid battery cars will capture only 5 percent of the limited second-
car market (shown in Table 6.10) for which they are reasonable candidates.
The subsequent advent of advanced-battery cars with ranges adequate for
general urban use will considerably expand the potential market, but the
high total per-mile cost of the nickel-zinc battery car will prevent
TABLE 6.10
POTENTIAL FREE-MARKET SALES AND USE OF
ELECTRIC CARS, SOUTH COAST AIR BASIN
1980 1990 2000
Total Car Population, thousands 5,880 6,730 7,600
Total Car Market, thousands per year 570 640 710
Candidates for Electric Car Replacement, percent 17 46 74
Electric Car Market Share
Capture of Candidates, percent 5 10 15
Market Penetration, percent 0.9 4.6 11.1
Annual Sales, thousands 5 29 79
Electric Car Population
Thousands of Cars 15 176 589
Percent of Total 0.26 2.6 7.8
Electric Car Use
Daily Vehicle Miles, millions 0.27 5.1 17.8
Percent of Total 0.16 2.6 7.8
10-47
-------
increases in the capture rate until after 1985. The zinc-chlorine battery
car may then offer competitive overall costs, but its 50-percent initial
cost premium and other limitations lead us to estimate an increase in the
capture rate to only 10 percent by 1990. Still, with the expanded market
this represents a six-fold increase in estimated sales during the decade.
Modest further battery improvement in the 1990s and the further probable
market expansion of that decade lead us to estimate a 15-percent capture
rate, with total sales more than doubling, by the year 2000.
The implications of these estimates are shown in the top six lines
of Table 6.10. The first three lines of this table recapitulate projec-
tions for total cars in the Los Angeles region, total car sales, and
candidates for electrification among them, assembled from Ref. 2 and
from Table 5.4. The next three lines reiterate the capture estimates
just presented, and show their consequences and penetration of the total
car market in the Los Angeles area.
An important assumption in this market share estimation process is
that the candidates for electric car replacement are replaced at the same
rate as automobiles in the total car population. This is much more con-
servative than the assumption of the Wisconsin investigators in Table 6.8,
which anticipates a much more rapid introduction of electric cars. In
the long run, however, and in the overall car population no such sales
rates can be sustained, since in the car population as a whole the
average replacement period is about 11 years—the life of a car.
The remaining four lines of Table 6.10 complete the projection with
electric car population and use figures. The population projections
assume sales growing at a linear rate during years between those called
out in the table, together with the survival rates of Ref. 2, Table 6.4.
The usage figures follow from resultant populations and the mileage pro-
jections of Table 5.4.
10-48
-------
A subjective appraisal such as this is inevitably contentious. The
basic questionable point is, of course, the capture percentage in Table
6.10, which was obtained by weighing primarily the high cost and limited
capabilities of the electric car against its potential reliability, high
status, and environmental benefit. It is certainly possible to argue that:
a much larger fraction of consumers would put cost and performance disad-
vantages second and thus buy the electric cars. Without additional solid
information, however, it is surely impossible to demonstrate that the
assumed capture rates of Table 6.10 are too low. We believe, in fact,
that if anything they are optimistic.
It should be noted that the sales estimates of Table 6.10 simply
presume that electric cars as described in Table 3.2 will be offered in
the Los Angeles market. In retrospect, however, this is less than likely
under free market conditions. Given a 10 percent market penetration in
the year 2000 (plus similar sales elsewhere) manufacturers might well
offer cars like the latter entries of Table 3.2. In earlier years, on
the contrary, it seems quite unlikely that lead-acid and nickel-zinc bat-
tery cars could and would be offered at the prices of Table 3.2. Their
prospective market is simply too small to support the mass-production
economies implicit in these prices. Accordingly, actual sales of electric
cars in Los Angeles will probably be much lower than estimated in Table
6.10, and prices substantially higher than in Table 3.2, until improve-
ments in battery technology open prospects of capturing a significant
market share.
10-49
-------
7 UPPER-BOUND RATE OF ELECTRIC CAR INTRODUCTION
The free market penetration for electric cars represents a probable
lower bound on prospective electric car use. Section 6 argues that this
level of use will be quite low—too low, in fact, to obtain major benefits
from automobile electrification. To capture these benefits, a public
policy of electric car encouragement may be contemplated, as discussed in
Sec. 8, with relative incentives tailored to achieve some higher target
level of use. Before considering possible policy levels of use, it is now
appropriate to establish an upper bound on levels which might reasonably
be obtained.
One important factor limiting electric car use is simply the time
required to start quantity production. Beyond this, it seems clearly un-
reasonable to seek rates of electric car introduction in excess of pro-
jected sales of conventional cars. Such rates imply that somehow or other,
the Los Angeles public will be required either to scrap conventional cars
earlier than normal, with attendant economic losses, or else export these
cars to other regions of the country.
Typical lead times involved in bringing auto modifications to pro-
duction status are shown in Fig. 7.1. This figure was derived by an im-
partial research organization in 1972 and is specifically applicable to
1975-1976 model years. Its contents are based on extensive surveys of
industry schedules and performance and thus offer a reasonable basis for
present estimates. They show that about 48 months is required from the
commencement of project conceptualization to production, with about 26
months required from the conclusion of vehicle preliminary design.
Given business as usual, 48 months or even more might be reasonably
assumed as the minimum lead time required to bring electric cars into
production, since auto electrification, even in its simplest form, is more
complicated than the changes implicit in Fig. 7.1. The object of this
section, however, is to project an upper bound rate of electric car intro-
duction. This necessarily implies a situation of some urgency rather than
mere "business as usual."
10-50
-------
RESEARCH AND ADVANCED h - PRODUCT DEVELOPMENT LEAD TIME
DEVELOPMENT
PRODUCT CONCEPTUALIZATION 170*7777*
CONCEPT DEVELOPMENT/
VEHICLE PRELIMINARY _ _ |— PRODUCTION LEAD TIME
DESIGN Y/////////////////////S////X
CAR PROGRAM APPROVAL
PRODUCTION ENGINEERING/ PRODUCTION
CAR PROTOTYPE TESTING TIME REFERENCE
PARTS PROCUREMENT/TOOL
CONSTRUCTION, INSTALLATION
AND TRYOUT
PILOT ASSEMBLY
PRODUCTION BUILDUP E
i i I I I i i I I I i i i i i I I I i i I I
<=>
Csl
^>-
48 36 24 12
MONTHS TO VEHICLE PRODUCTION
Figure 7.1. Automotive Product Development Phases (From Ref. 17)
For electric car introduction at an urgent pace, the 48-month lead
time was assumed to be reduceable, primarily by shortcuts in conceptuali-
zation and preliminary design. Although the result may be less than opti-
mum, initial electric cars can make extensive use of existing automotive
assemblies, components, and production facilities. Moreover, major auto-
mobile manufacturers have already engaged in pre-production research and
preliminary design for electric cars. The Ford Motor Company, for example,
1 ft
has an ongoing program which includes novel traction motor development.
General Motors built an electric minicar several years ago with particular
19
attention to manufacturability. American Motors is supplying Hornet
cars to Electric Fuel Propulsion, Inc., for addition of electric traction
motors, controllers, and batteries; and EFP expected to sell some 2,500
20
electric cars of this type during 1974. It thus was assumed that a 36-
month lead time to production could be achieved: the 26-month production
lead time of Fig. 7.1, plus 10 months (just under half) of the usual 22-
month conceptualization and preliminary design phases.
10-51
-------
Any decision to proceed urgently with electric car introduction
would require significant governmental actions which are not now in pro-
gress. Such actions could easily require years, if they were ever under-
taken at all—which now appears unlikely. The purpose of this analysis,
however, is not to predict what the government will do, nor how long it
will take, but rather to determine the probable results of a range of pos-
sible actions. In the present instance, possible actions are to introduce
electric cars in Los Angeles at various rates, and we seek to establish
an upper bound on the range of rates possible.
Given an immediate decision to proceed, at the beginning of 1975,
this suggests that the earliest possible quantity production would begin
at the outset of 1978. Once initiated, production was assumed to rise
very rapidly to capacity. Even though production rises immediately to
planned levels, however, there will be only three years for electric car
manufacturing before the end of 1980, the first target year of this study.
In this postulation of an upper bound, it might be reasonable to
assume that electric cars could be manufactured in sufficient quantity to
supply all new car sales in Los Angeles, which annually amount to some
600,000 units. This, however, seems undesirable even for the most urgent
and determined public policy decision to electrify, since a substantial
minority of personal transportation vehicles sold in the Los Angeles
area are so specialized as to be inappropriate for replacement by electric
cars. House cars, pickup trucks, off-road vehicles, and high-performance
sports cars, for example, amount to well over 10 percent of the Los
Angeles personal-vehicle market. Accordintly, if electric cars are sold
only as replacements for new conventional passenger vehicles which would
otherwise be sold, perhaps 80 percent of the projected total market might
be electric.
Table 7.1 projects the results of this assumption. The first two
lines of the table are projections of total car population and total
10-52
-------
TABLE 7.1
PROJECTED MAXIMUM ELECTRIFICATION OF CARS IN SOUTH COAST AIR BASIN
Total Car Population, thousands
Annual Car Sales, thousands
Electric Car Sales, thousands
Electric Car Population, thousands
Electric Cars, percent of total cars
Electric Car Use, percent of daily car mileage
1980
5,880
570
456
1,326
23
14
Year
1990
6,730
640
512
4,913
73
73
2000
7,600
710
568
6,080
80
80
annual car sales in the South Coast Air Basin from Ref. 2. At 80 percent
of the annual total, electric car sales are shown in the third line of
the table. Consequent electric car populations in the fourth line assume
that production and sales begin at the outset of 1978 and that electric
car survival rates are the same as those of conventional cars in the
Basin, as developed in Ref. 2. The usage of electric cars assumes lead-
acid battery cars in 1980 driven 6,500 miles per year, and advanced bat-
tery cars in 1990 and 2000 driven 10,000 miles per year, as are conven-
tional cars. It should be noted here that a mobility loss in the early
1980s is implicit in Table 7.1. As shown in Sec. 6, the lead-acid battery
car can replace, without substantial mobility loss, only about 17 per-
cent of area automobiles, or about one million units in 1980. Thus
under the assumed introduction rate, these limited-performance cars will
have displaced some three hundred thousand automobiles which would have
been driven significantly beyond the limited capability of the lead-
acid electric. With the appearance of advanced battery cars after 1980,
of course, no further mobility loss of this type will arise.
10-53
-------
8 SCENARIOS FOR ELECTRIC CAR INTRODUCTION
, The spectrum of possible electric car use in the Los Angeles region
ranges between the levels of the hands-off free market situation of Sec. 6
and those of the maximum electrification pace of Sec. 7. The maximum
electric car situation described in Sec. 7 would undoubtedly be economically
costly and socially onerous, involving major economic transients and signi-
ficant limitations of area mobility. The free market situation of Sec. 6,
however, promises so low.a level of electric car use that beneficial
impacts could at best be minor on a regional scale, even in the year 2000.
The desirability of intermediate levels of use will be appraised fully in
the concluding phase of this electric car impact study. Meanwhile, it is
worth noting that there are important rationales for intervention in the
free-market situation to raise electric car use above the free-market case
towards the maximum projection. The free market transactions, after all,
now neglect almost entirely such externalities as air pollution, noise,
and other automotive degradation of the urban environment. Moreover, so
long as substantial governmental control is being exercised to hold energy
prices down, the energy conservation issues may also be externalized in
some degree. In the long run, finally, the most important benefit of
electric cars could be to end the dependence of US personal transportation
on imported petroleum and the political machinations of producing nations.
In this section, specific intermediate levels of use are considered
which might be sought as a matter of public policy. Next, various imple-
mentation measures which might produce them are investigated. Then
various dimensions of uncertainty are reviewed. Finally, specific
scenarios for arriving at intermediate usage levels are hypothesized.
8.1 INTERMEDIATE LEVELS OF USE
A "high use" public policy for electric cars in Los Angeles should
probably be tailored to avoid the near-term economic transients and
mobility losses already noted for the maximum usage projection in Sec. 7.
These ends can be accomplished simultaneously if the usage target is held
10-54
-------
to the market potentials developed in Sec. 5. Thus until 1980, the objec-
tive would be limited to capturing the second car market at single family
households with off-street parking. The total population of such cars
is 17 percent of all cars in Los Angeles. In the long run, then 17 per-
cent of area sales is the maximum sustainable rate for the electric cars
capable of no other functions, and presumably it would be unwise to deve-
lop manufacturing capability in excess of the long-term potential. Thus
the initialization of the high-use policy might be production and sales
in 1978 and subsequent years of 17 percent of projected total car sales.
By 1990 and 2000, the more-capable advanced-battery cars could replace
46 percent and 75 percent of area cars. On the same line of reasoning,
production of these cars should be built up during the 1980s to these
fractions of the area new-car sales projections. Relative to the upper-
bound projection, the overall result would be relatively little economic
dislocation and mobility loss, primarily since early emphasis on lead-
acid battery cars would be very much reduced.
A "medium use" policy might reasonably seek a level half-way be-
tween the high use case and the free market case. Since the free market
usage is very low, this is effectively equivalent to seeking half the
usage sought by the high use policy.
Projected market shares, populations, and usages of electric cars
under these policies are shown in Table 8.1, together with the upper and
lower bounds represented by the upper bound case and the free market
case. Again, survival rates of Ref. 2 and linear growth in sales between
years tabulated are assumed.
Alternative electric car populations relative to the total for the
South Coast Air Basin are illustrated in Fig. 8.1. The figure makes
clear that the high-use policy differs most from the upper bound during
the coming decade, and thereafter approaches this bound as the higher
performance electric cars come into service.
10-55
-------
TABLE 8.1
ALTERNATIVE PROJECTIONS OF ELECTRIC CAR MARKET SHARE, POPULATION, AND USE,
SOUTH COAST AIR BASIN
Market Share, percent
Upper Bound
High Use Policy
Moderate Use Policy
Free Market
Total Population, percent
Upper Bound
High Use Policy
Moderate Use Policy
Free Market
Total Vehicle Miles, percent
Upper Bound
High Use Policy
Moderate Use Policy
Free Market
1980
80
17
8.5
0.9
23
4.9
2.5
0.26
14
3.1
1.5
0.16
Year
1990
80
46
23
4.6
73
31
15
2.6
73
31
15
2.6
2000
80
74
37
11.1
80
60
30
7.8
80
60
30
7.8
8.2 IMPLEMENTATION MEASURES
It remains to investigate legislative and regulatory measures which
might suffice to implement the high and medium use policies of Table 8.1
and Fig. 8.1. This is not a policy study, so no lengthy analysis is
appropriate here. Moreover, recent work in this field has yet to
establish quantitative models of the effects of various possible actions
on either car use in general, or electric car use in particular, so the
discussion must ultimately rest on qualitative judgments.
10-56
-------
UPPER BOUND
ELECTRIC CAR POPULATIONS
MEDIUM USE POLICY
1980
YEAR
1990
2000
Figure 8.1. Alternative Electric Car Population Projections, South Coast
Air Basin
Electric car usage may be encouraged either by direct incentives,
or by disincentives applied to the competing conventional cars. In the
interests of air quality and petroleum conservation, a great many tactics
in the latter category have been proposed recently. They include gasoline
rationing, additional gasoline taxes, auto excise taxes, parking restric-
tions and price increase, restrictions on the use of roadways, and
restrictions on the entry of automobiles into particular urban areas.
Gasoline rationing and taxes automatically differentiate in favor of the
electric car; the other measures would favor electric cars only to the
extent that they were applied solely to conventional cars.
Each of these measures has advantages and disadvantages. Generally,
a preferred combination of measures would achieve the desired electric
car-usage primarily through the price system as modified by direct user
taxation, since this maximizes individual freedom of choice. Very large
10-57
-------
direct taxes, however, may be undesirable in their immediate effect on the
distribution of income, since they particularly impact the poor unless
and until compensating changes are introduced in our systems of welfare
and progressive income taxation. Very large taxes may also be impractical
in that they are highly conducive to black markets and a host of both
technological and institutional dodges. Where such taxes arise, various
subsidies or direct regulation of car usage deserve consideration.
8.3 GASOLINE AND LICENSE FEE SURCHARGES
To illustrate the sort of taxation which might be required to produce
high electric car use, surcharges on gaoline and annual licensing will
now be considered. Gasoline taxes are advantageous in that the mechanism
for their collection already exists and functions smoothly. Moreover,
they tend to spread disincentives to gasoline-powered cars proportional
to their use. Problems would certainly develop, of course, in preventing
supplies of untaxed gasoline from entering the South Coast Air Basin, if
local taxes were at very high levels. The annual license fee surcharge
might be simpler to enforce, but has the disadvantage of being insensitive
to the different usages of individual cars.
Estimation of the order of magnitude of surcharges required to en-
courage high electric car use begins at relative car costs as suggested in
Table 3.2. On this basis, we may develop surcharges required to put
gasoline-powered cars at specific cost disadvantages relative to
alternative electric cars. Finally, we may speculate on the magnitude of
these cost disadvantages required to bring about desired levels of electric
car use.
Tables 8.2 and 8.3 elaborate the extra initial and overall operating
A
costs of electric cars shown in Table 3.2. Table 8.2 distinguishes that
part of extra initial cost required to purchase the car battery, which
generally depreciates considerably faster than the car itself. Table
8.3 shows extra total per-mile operating costs of electric cars at two
different gasoline prices, as detailed in Ref. 4.
10-58
-------
TABLE 8.2
EXTRA INITIAL COSTS OF ELECTRIC SUBCOMPACT CARS (from Ref. 4)
i
Extra Initial Cost, 1973 dollars
Car Battery Type
Lead -Ac id
Nickel-Zinc
Zinc-Chlorine
Lithium-Sulfur
*
Relative to conventional
Without Battery
$707
675
621
525
subcompact costing $2270.
With Battery
$1,907
3,615
1,221
1,125
TABLE 8.3
EXTRA LIFE-CYCLE COSTS OF ELECTRIC SUBCOMPACT CARS (from Ref. 4)
*
Extra Per-Mile Cost, 1973 Dollars
Car Battery Type For Gasoline at 50c/gal
Lead-Acid
Best Life
Worst Life
Nickel-Zinc
Zinc-Chlorine
Lithium-Sulfur
Best Life
Worst Life
$0.027
0.085
0.047
-0.014
-0.011
-0.002
For Gasoline at 80c/gal
$0.012
0.070
0.032
-0.029
-0.026
-0.017
**
Relative to conventional subcompact with total life-cycle costs per
mile (including financing) of $0.133 to $0.148 (for gasoline at 50 to
80 cents per gallon, including taxes, and fuel economy of 21.4 mpg).
-------
Given the basic operating cost differentials of Table 8.3 with gaso-
line at 50 cents per gallon, Table 8.4 shows total extra cost per 30-mile
driving day. Since 30 miles is approximately the distance the conventional
2
subcompact car could be driven on a gallon of gas, figures in Table 8.4
approximate the gasoline tax per gallon which would equalize total costs
of electric and conventional cars. Furthermore, since 30 miles is about
the average daily driving distance in Los Angeles, figures in Table 8.4
may also be regarded as the extra daily costs which would have to be
added to conventional subcompact costs for equalization with electric car
costs, as might be accomplished by differential parking fees.
Simple equalization, of course, is probably insufficient to insure
wide use of electric cars. Even if costs were made equal, electric cars
might remain relatively unpopular because of their limited range and
limited performance, despite their advantages in reduced maintenance
TABLE 8.4
EXTRA TOTAL COSTS OF ELECTRIC SUBCOMPACT*
Total Extra Cost Per
Car ^P6 30-Mile Day
Lead-Acid $0.81
Best Life $0.81
Worst Life 2.55
Nickel-Zinc 1.41
Zinc-Chlorine -0.42
Lithium-Sulfur
Best Life -0.33
Worst Life -0.03
See notes to Tables 8.2 and 8.3.
10-60
-------
burdens. Chances are that a substantial cost saving must be arranged for
the electric cars to insure popularity. Figure 8.2 shows surcharges on
subcompact ICE cars which would produce cost disadvantages of 0 to 50 per-
cent relative to electric cars characterized in this study. The surcharges
are stated in dollars per gallon of gasoline and in dollars of additional
annual license fee. If cost disadvantages of 30 to 50 percent should be
necessary to produce the high level of electric car use summarized in
Fig. 8.1, then these surcharges are truly enormous for the lead-acid and
nickel-zinc battery cars: approximately $3 to $5 per gallon of gasoline
or $1,000 to $1,800 of annual fee. For the lithium-sulfur and zinc-
chlorine battery cars, much smaller surcharges would be indicated, assum-
ing the low battery cost goals of developers are achieved, but they would
still be considerable: roughly $1 per gallon of gasoline, or $360 per
year.
We have no data available to pinpoint the cost differential required
to obtain a specific level of electric car sales and usage. In Sec. 6.1,
however, it was pointed out that relatively minor differences in conven-
tional car characteristics apparently account for large differences in
selling prices acceptable to the consumer. This suggests that the required
level of surcharge in Fig. 8.2 would fall towards the high end rather
than the low end of the scale.
Before leaving the dramatic surcharges of Fig. 8.2, we must briefly
review the major uncertainties in electric car costs underlying these
surcharges, as developed in Refs. 4 and 1. In their entirety, they convey
considerably more about ranges of uncertainty and relative levels of
uncertainty than does Fig. 8.2.
In Fig. 8.2, a broad range of surcharge is shown for the lead-acid
electric car, despite its similarity to widely-used and relatively well-
understood electric vehicle technology. In this case, most of the surcharge
is attributable to battery depreciation, which depends in turn on three
uncertain quantities: first, battery energy density (and hence weight
and cost); second, battery initial cost; and third, battery cycle life in
10-61
-------
2000 i-
00
Qi
LU
OO
1500
1000
IS
o:
oc
on
=>
oo
a
LU
a:
a:
a.
o
on
CL.
O_
ct
500
6 i-
o
o
o
to
4
a.
o
ce
a.
NICKEL-ZINC
BATTERY CAR
LITHIUM-SULFUR
BATTERY CAR
LEAD-ACID
•:-| BATTERY CAR
ZINC-CHLORINE
BATTERY CAR
10
20
30
40
50
PERCENT TOTAL COST DISADVANTAGE
FOR SUBCOMPACT CONVENTIONAL CAR
Figure 8.2. Required Surcharges to Establish Subcompact Conventional Car
Cost Disadvantage Relative to Electric Cars
10-62
-------
automotive use. Among these factors, the third is by far the most impor-
tant in its effects on cost uncertainty, followed by the first. Reference
1, for example, indicates a 3-to-l range of uncertainty in cycle life (at
65 percent discharge depth) for the lead-acid electric vehicle battery
characterized for 1980. The lower life figure is that attained by cur-
rently-marketed electric vehicle batteries of similar energy density, so
the upper surcharge area shaded in Fig. 8.2 is relatively certain of
attainment. Considerable improvements over existing technology are impli-
cit, however, in the lower surcharge range, and these are relatively less
certain, despite the long experience we have had with lead-acid battery
technology.
For the nickel-zinc battery car surcharge in Fig. 8.2, no range of
uncertainty is indicated. This, however, is because little experience
and data are at hand: barely enough, in fact, to arrive at the nominal
battery life, depreciation, and surcharge projections subsumed in Fig. 8.2.
On the one hand, "adequate" cycle life has long been the principal obstacle
to nickel-zinc battery development and application. On the other hand,
current laboratory data suggest that lifetimes several times greater than
that implicit in Fig. 8.2 may be achieved. Thus at least as wide a range
of uncertainty should be associated with the nickel-zinc battery car in
Fig. 8.2 as with the lead-acid battery car.
For the other two battery cars of Fig. 8.2, risks and uncertainties
are almost certainly greater still, since even less laboratory data and
achievement than in the nickel-zinc case stands behind their technologies.
The nominal surcharges in Fig. 8.2 for these cases are based on performance
goals established and sought by their respective developers. There is
relatively little basis on which to establish these goals as realistic,
rather than merely wishful thinking. In these cases, moreover, initial
cost may be relatively more important as a source of uncertainty. Both
of these advanced batteries are relatively far from production, so asso-
ciated costs are difficult to establish at present. In both cases, initial
10-63
-------
costs estimated by the developer are quite low, especially on a basis of
energy storage capacity rather than weight. If both initial cost and
cycle life estimates prove low by a factor of 2, the cost advantages of
the respective cars with these batteries (see Table 8.3) could disappear.
In any case, uncertainties abound and the surcharges of Fig. 8.2 seem far
more likely to represent a lower bound rather than an upper bound on
prospective cost differentials.
8.4 SUBSIDY OF ELECTRIC CARS
Gasoline taxes of up to $1 per gallon are certainly conceivable;
they would produce gasoline prices no higher than those actually prevailing
in the Spring of 1974 in Europe and Japan. Taxes above $2 per gallon,
however, as the lead-acid and nickel-zinc battery cars might require to
be competitive with the conventional subcompact, are surely extreme,
implying major disruptions of established travel patterns if imposed.
Recent data indicates, for example, that the price elasticity of gasoline
22
is lower than previously estimated, nearer 0.2 than 0.3. This means
that a 50-cent (100 percent) increase in gasoline price might cause a 20-
percent drop in consumption—and auto usage. But a $2 per gallon added
gasoline tax represents a 400-percent increase from a price of 50 cents,
which extrapolates to an 80-percent decline in auto travel. This extra-
polation is unrealistic, of course, but its size clearly suggests that
whatever the real figure, a 400-percent price increase spells major dis-
ruption of auto travel—and of the owner's budget as well.
In consequence, the imposition of tax penalties on conventional
cars to encourage lead-acid and nickel-zinc battery car use appears un-
desirable. Alternatively, means might be found to subsidize electric
car travel. If the cost to the consumer were reduced from the 15 to 20
cents-per-mile range to 8 to 10 cents-per-mile by subsidy, then the
needed cost advantage to ensure electric car use might be less painfully
achieved. If necessary funds were obtained from general revenues, the
costs to consumers would not be intolerably concentrated in the transpor-
tation sector, with resultant non-optimum suppression of travel demand.
10-64
-------
Various mechanisms for this are at hand: programs of aid to battery
makers and electric car makers, reduction of road user taxes and fees,
elimination of relevant sales taxes, and so on. Particularly where small
numbers of electric cars are envisioned, as in 1980, this subsidy approach
seems less objectional than conventional-car tax disincentives.
8.5 DIRECT TRAVEL RESTRICTIONS
If direct restrictions on roadway use are to encourage electric car
use, they might begin by setting aside freeway and arterial lanes, not only
for busses and car pools as now proposed in Los Angeles, but for electric
cars as well. Under this arrangement, the value of time lost in travel
due to congestion might be expected to approach the surcharges of Fig.
8.2. A study of motorist behavior in 1967 showed that motorists then
23
were willing to pay $2.80 in tolls to save an hour of travel. This sug-
gests that the current value of motorists' time is $4 to $5 per hour, de-
pending largely on whether increases in average hourly income or the con-
sumer price index are taken as the basis of adjustment. From Fig. 3.1,
the average Los Angeles home-to-work trip will require around 15 minutes,
and traverse some 8 miles at the travel speeds of Table 3.1. For the
lead-acid and nickel-zinc battery car, the required surcharge on 8 miles
is 30 to 80 cents, equivalent to 4 to 12 minutes at the $4.00 per hour
rate. Thus equalization of the electric car costs would require lane
restrictions imposed on conventional cars which would increase their
travel time from 25 to 100 percent.
In the first approximation, the costs of roadway restrictions are
equivalent to those of the other means discussed for eliminating the
electric car disadvantage. Other factors, however, make such travel
restrictions undesirable in comparison with direct taxation and subsidy.
If congestion delays are increased to this extent, conventional cars will
be forced into extended stop-and-go driving on both streets and freeways.
Although the result may lend relative encouragement to electric cars, it
will also substantially increase pollutant emissions by the conventional
10-65
-------
cars and at the same time will considerably increase their energy consumption
per mile.
Another regulatory alternative, restrictions on entry of gasoline
cars into certain urban areas at certain times, could potentially
generate equally serious side effects. One would be differential penaliza-
tion of area businesses, to their serious detriment. Another would be
economic inefficiency arising as households acquired more cars than
necessary: electric cars to use in restricted areas at restricted times,
and conventional cars for use at other times and places.
8.6 SCENARIOS FOR ELECTRIC CAR INTRODUCTION
The policy levels of use developed in Sec. 8.1 and displayed in
Fig. 8.1 were constructed with only a preliminary regard for the phasing-
in of consecutive developments in electric car battery technology. This
is, of course, an oversimplification; as Sec. 8.3 details, there are very
large disparities in dollar costs and in uncertainties among the battery
technologies. In consequence, the usage buildups of Fig. 8.1 are
inappropriate themselves as scenarios for electric car introduction, since
high usage in 1980 could be as undesirable as it might be desirable by
the year 1990 or 2000, when the high-cost battery technologies will
hopefully have given way to lower-cost technologies.
Accordingly, schedules for introducing single electric car
technologies must be hypothesized here. This will provide pure, elemental
cases for impact assessment and evaluation at the conclusion of the over-
all study which this usage analysis supports. Only on the basis of these
assessments will a construction of complete scenarios, integrated from
the elemental pieces, be productive.
Overall, the objective here must be to define scenario elements
using pure technologies which may be combined as needed to obtain
moderate and high levels of use as shown in Fig. 8.1. In each case, of
course, these scenario elements must comply with constraints implicit in
10-66
-------
Fig. 8.1; i.e., they must begin sales of electric cars no earlier than
the availability of the corresponding battery technology, and they must
not exceed the market shares appropriate to prospective car applicability.
A set of such scenario elements is presented in Table 8.5. It
includes the most interesting cases for impact evaluation, and complies
fully with the appropriate constraints.
In 1980, only lead-acid battery technology will have been available
for any length of time, and only 5 percent of cars may reasonably be
electrified in the available time. This corresponds to a high-use case
which is itself so modest that consideration and evaluation of a "moderate1
use case would involve trivial electric car use and impacts, so no "mode-
rate" use case is included. The market share required here, deduced from
the auto age distribution of Ref. 2, is within the 17% upper limit of
Sec. 8.1.
TABLE 8.5
SCENARIO ELEMENTS FOR ELECTRIC CAR INTRODUCTION
Target
Year
1980
1990
2000
Use
Policy
High
Moder-
ate
High
Moder-
ate
High
Electric
Car
Population
Target
5%
15%
30%
30%
60%
Battery Type
Lead-acid
Lead-acid
Nickel-zinc
Zinc-chlorine
Nickel-zinc
Zinc-chlorine
Lithium-sulfur
Zinc-chlorine
Lithium-sulfur
Sales Period
1978-1980
1978-1990
1980-1990
1985-1990
1980-1990
1990-2000
1990-2000
1990-2000
1990-2000
Approximate
Average
Market Share
16%
16%
17%
30%
35%
35%
35%
70%
70%
10-67
-------
For 1990, all available technologies are included for moderate use,
but only one is appropriate for high electric car usage. Lead-acid cars
have insufficient range to replace more than 17 percent of area cars and
consequently cannot support the 30 percent high usage level, whereas
zinc-chlorine technology -is only available for five years prior to the
target year for evaluation, and could not reach the high use level with-
out sales in excess of the 46 percent maximum assumed in Sec. 8.1. For
the target year 2000, lead-acid and nickel-zinc technologies are omitted
because the more advanced battery systems may achieve much lower costs
and should then be much preferable. Moreover, the 1990 high-use nickel-
zinc case should reveal impacts quite similar to those which would be
deduced for this battery car in 2000.
A variety of combinations of these scenarios may be constructed
after impact evaluation, since where two technologies are used at one
time, their impacts may be simply added because the impacts are basically
linear with electric car use. The range of possible combinations is very
wide, since any two of the moderate use elements of Table 8.5 in a given
year may be taken simultaneously to develop a high level of use in that
year.
The legacy of electric cars left in a given target year from the
population of the previous target year is so small that it may be
dismissed. Only about 13 percent of area cars are more than 10 years
old. Since each decade in Table 8.5 sees electric car populations double,
the implication is that the legacy would usually be only 6 1/2 percent
of the total number of electric cars under consideration, a figure smal-
ler than other uncertainties involved in impact assessment. Accordingly,
if an overall scenario is built up for electric car introduction from the
elements of Table 8.5, the car population in a later target year will
not substantially depend on that of a previous target year.
10-68
-------
APPENDIX
COMPUTER PROCESSING OF LOS ANGELES TRANSPORTATION DATA
Data from the 1967 home-interview survey on magnetic tapes was
processed by a computer program according to the logic of Fig. A.I. A
listing of the program and its subprograms follows.
Most of the processing is accomplished within the main program. A
separate subprogram, however, was used to process zone-node data.
Identification numbers and centroid coordinates for over a thousand
traffic zones were card-punched in compressed format. The subprogram
read these cards and organized their zone data for easy subsequent access
when it was first called. Subsequently, it referenced the data upon
demand to determine zone-to-zone distances. Two other small subroutines
were employed to decode BCD tape entries for processing, and to place
headings on output pages.
The program listing which follows is largely self-explanatory.
It contains numerous comment cards which explain functions of individual
blocks of code and define most data. Numbers in Fig. A.I are repeated
as statement numbers beginning corresponding blocks of code in the
program listing.
10-69
-------
10 INITIALIZE PROGRAM
30 READ RUN SPECIFICATIONS
60 READ AND ORGANIZE ZONE DATA
100 GET NEXT HOUSEHOLD DATA
120 EXTRACT CAR AND PARKING DATA
140 CLASSIFY HOUSEHOLD
160 UPDATE CAR/PARKING AVAILABILITY ARRAY
-
200 C
300 (
400 /
500 ]
NO MORE
HOUSEHOLDS
GET NEXT HOUSEHOLD DRIVER TRIP DATA
CALCULATE TRIP DISTANCE
ASCRIBE TRIP TO CAR
INCREMENT TRAVEL ARRAYS
NO MORE
DRIVER
TRIPS BY
HOUSEHOLD
600 CLASSIFY DRIVERS AND CARS
700 UPDATE TRAVEL DISTRIBUTIONS
800 UPDATE DISTRIBUTION DESCRIPTORS
900 COMPLETE AND PRINT TRAVEL DISTRIBUTIONS
940 PRINT DISTRIBUTION CONDITIONS
960 PRINT CAR/PARKING DATA
STOP
Figure A.I. Processing of Los Angeles 1967 Travel Survey
10-70
-------
PROGRAM AUTOUSE
c
C COMPUTES DISTRIBUTIONS OF DAILY CAR AND DRIVER TRAVEL FROM
C HOUSEHOLD ANU TRIP DATA TAPES OF THE LARTS 1967 HOME INTERVIEW SURVEY,
C
C HOUSEHOLD DATA TAPE IS ON UNIT 1. TRIP DATA TAPE IS ON UNIT i>.
C HOUSEHOLD DATA IS IN 13-WORD RECORDS, EACH DESCRIBING A SINGLE
C HOUSEHOLD, SORTED ACCORDING TO HOUSEHOLD ID. TRIP DATA IS IN
C PI-CORDS Q^ 720 WORDS, EACH DESCRIBING 36 TRIPS, ORDERED ACCORDING
C HOUSEHOLD ID, EXCEPT A VERY FEW SHORTER RECORDS.
C
C
COMMON KONTPOL(3)
C
C THESE ARE INPUT INTEGFR PRINTOUT CONTROLS. IF NUN-^Pn. CAUSE
c INTERMEDIATE OUTPUT FROM EACH HOUSEHOLD TAPE RECORD, EACH IHIP DATA
C ENTRY, AND FROM INCOMPLETE 0ISTRIbUTIONS, RESPECTIVELY.
C
COMMON NHOMHtC»NTRPRt.C»NTENTRY.NPARl »NPAR2«NGAP.^ ' IGAPrlLE «NTPIPS»
1 NHOMSKP»NT A»TPP»NTRKTRP»NHOMIT»NOUSE,NOCARS«NOTKH-s.NOPLt-'s.
2 NTOMIT,NU/.GNt
C
c COUNTS OF m. CORDS READ FROM HOUSEHOLD TAPE AND IPIP iiu-t • IVIPS
C PROCESSED FRuM CURRENT TRIP RECORD, PARITY ERRORS ON UNITS I ANU ?•
C T"TAL NUMHLK OF GARBLES IN DATA, FLAG SET FOR CU^PtNT r-AKMLt,
c T-MPS PROCESSED FOR CURRENT HOUSEHOLD, NUMBEK OF HOUSEHOLDS
C DIPPED IN ATTEMPT TO MATCH CURRENT TRIP ID. NUMbF'R UK TAX!
C il^IVER TRIPS SMPi:>ED» NUMritH OF TRUCK DRIVER TRIPS Si^IfPtl'i,
C Ni'MdER OF HOUSEHOLDS SKIPPED IN TOTAL, AND NUMHER OF ONLY CftUS
C M'MHEP OF HOUSEHOLDS SKIPPED IN TOTAL, NUMBFK OF SM^PFIJ
C -ir.UStHOLOS wlTH UNUSED CARS, NUMBER OF SKIPPED HOUSEHOLDS WITH
C D-'IVER TRIPS JNUICATtl) flUT NOT FOUND ON TRIP TAPt, NUMHFH
c OF SKIPPED HOUSEHOLDS WITHOUT ANY PERSONS RESIDENT, NUM^R OF
c T^IPS OMITTED FOR FAULTY HOUSEHOLD ID, AND NUMBER OF TRIPS
c OMITTED FOP MAD ZONE NAMES.
c
COMMON Kl,K.;e,K3.K<»»KS»Kll,M2,Kl3«Kl<»,Klb
C
C T"ESt FLAGS ARt. sc.T bY SUBROUTINE DECODER IN t * THAC TlNd NtEOKO
C NUMERICAL DATA FROM CURRENT HOUSEHOLD AND TRIP DATA.
C
COMMON HOME (13) ,HOM I D,NMONTH.NDAY , HOMT YP«NPERS ,NPASSV/, NPMiP «
1 :;TRIPD,NSTOf«LONGDIS,NPKG»MKTVAL.NA267
INTEGER HOMTrP
C
C CURRENT HUUStHOLU DATA ARF STORED IN THESE VARlAriLtS. THt ChRrtKNT
C HOUSEHOLD RECORD, VERBATIM, IS IN HOMEU3). SELECIt" vA*v|Arii.fc'S
C FXTRACTEO f'^OM HOMH13) ARE STORt.0 IN THt REMAININu vAulftnLhS,
c INITIALLY i\ A FORMAT.
c
COMMON HHDtN1(0)
LOGICAL HhCOL(W)
COMMON HHQATAItt.ft)
DIMENSION HHLABL<£,6>
C
C THIS BLOCK C^ VAKIABLKS IS USED IN DEVELOPING GVK^Ml. i>tsCRIPTORS
C OF HOUSEHOLD DATA. HHOtNT HOLDS ONfc COLUMN Of DATA TO bf AUDFu TO
10-71
-------
C APPROPRIATE COLUMNS OF HHDATA. HHCOL DESIGNATES THE APPK)PRIAR
C COLUMNS. HHLAbL CONTAINS HOLLERITH LABELS FOP THE ENTRIES IN
C MHDENT AND FOR THE WOWS OF HHDATA.
C
COMMON TRIPS<20.36)» TRIPIO. PERSON. MODE. FRSWRK, BEGTIM, t'MDTIM,
1 F'WY. ONITE. PKG. NPASS. NORIG, NDEST. TRPTYP
C
C CURRENT TRIP DATA ARE STORED IN THESE VARIABLES. TRIPS HOLDS ALL
C 3* TRIPS RECORDED IN A SINGLE RECORD ON THE TRIP TAPE. THE
C REMAINING VARIABLES APE FILLED IN. INITIALLY IN A FORMAT. WITH
C 0 .CARVELO.IO)
LOGICAL DRlASS 16.10) .CAPASS . TRIPDI S ( 1 O'.^O ) »CONDI S( 1 0 . 9 )
DIMENSION L)ST (tl ) , TRIPNOC.1) ,CONDX(9)
C
C THESE VARIABLES ARE USED IN ACCUMULATING DRIVFR ANU CAP TPIP DATA
C FOR EACH HOUSEHOLD. AND IN DEVELOPING FROM IT TRAVEL OTSTPIBU-
C T10NS FOR ALL HOUSEHOLD DRIVER TRIPS. TPP CONTAINS KEY UhSC^IP-
C TORS OF THE CURRENT TRIP. CONTENTS OF TRP AkE ADDcD TO RO*S OF
C THt DRIVER TKAVtL AND CAR TRAVEL ARRAYS DRIVEL AND CARVEL. KACH
C SUCH ROW ACCUMULATES DATA FOP TRIPS ON THE SURVEY UAY MADE HY 6
C SINGLE DRIVER OR CAR. THE DRIVER CLASS AND CAW CLASS ARRAYS.
C DP I ASS AND CARASS. SHOW WHICH OF VARIOUS FINAL DISTPI htjT in.MS AND
C SnHOlSTRlbUTlONS EACH DRIVER AND CAR BELONG (r>. THff VARIOUS TPJP
C MILEAGE DISTRIBUTIONS AWE ACCUMULATED IN THE COLUMUS OF ARPAv
C OISTDIS. WHILE THE TKIP NUMBER DISTRIBUTIONS ARE ACCUMl.iL*Tt.l) 1 iN
C APRAY TRIPDIS. VARIOUS DESCRIPTIVE CONDITIONS ARF; ACd.'MUI. AI f I)
C I.\ THE COLUMNS OK ARRAY CONDIS. WHICH ARt INCRKMEMEu K OK F. ACH
c HOUSEHOLD ANU ITS TPI&S HY CONDX. DST AUO TRIPNO LEFIMK THK.
C CrLLS OF THfc !'. ISIANCF. AND TRIP igUMbtw DISTR T bUT IONS. AND Ob7
C Sl'PFLIKS kOW LAKF.l S » .f)WVWNAM(lO)
C
C T;>*AL IS A WORK 1N(... STORAGE APRAY USED IN NORMALIZATION ANP
c CUMULATION OF THE DISTRIBUTIONS FOR FINAL PRINTOUT.
C r)^vRNAM(I) IS THt (ALDHAHEFIt) NAMt OF THE I-TH DRlVhR AT A
C HOUSEHOLD.
C
OATA (DST( I ) .l = l.tl ) /
1 0.<^.*.«8.«lG.«l?..l^.«l6..1B..2G.«£?.i3i'.»
2 3?..3<».»36.»3y.«':»0..<»?.»<»<«.«1*b.t'»8.»5U.»Sb..hU.»r>t>..flj.«r,..
3 80..85..90..95..100.•1?S..IbO.•17b.,aOO.,l^v^. /
DATA (TRIPNOI1).1 = 1 .tl ) /
1 0...5.1.'i.?.5,3.'>.4.b,5.5.6.b,7.lD,H.b.^.t,.10.^.11.S.l^.T
? 13.5«l'..5,lb.b.l6.S,17.S,18.b.l9.S,20.b.i)J.b.^.S.27.S,^b.S,
-------
1 CALL HEADI10HAUTO USAGE)
CALL HEAD(10)
DO 14 I»l,30
14 KONTROL(I)=0
NTLNTHs36
C
C RFAD RUN SPECIFICATIONS
C
30 PRINT 32
32 FORMAT(IX 25HRUN CONTROL PARAMETERS - //)
DO 36 1=1.3
"FAD 34. KONTROL(I)
J4 H',RMAT(40H
3b PRINT 34. KONTkOL(I)
C
C *FAD AND ORGANIZE ZONE NAMES AND CENTROIIJ LOCATIONS
C
60 CALL DIST.HOME)
1 HOMID. NMONlH, NUAY. HOMTYP, NPERS. NPASSV, NPKUP,
2 NTRIPD, NSTOR, LONGDIS. NPKG. MKTVAL. NAZ67
106 H>kMAT< A7, 9XAc!, 7XA1, 2XA1, 3XA2. 4XA1. Al, 10^A3»
1 1XA1. Al. 13XA")
IF (KONTROL(l).LT.l) GO TO 120
CALL HEADI3)
PRINT 108.NHOMREC
108 FORMATI31H DATA FRUM
PPINT 112
112 FORMAT!
1x61.
HOUSEHOLD RECORD NO.
bOH
50H
30H
bOH
ID
PASS.
PARKING
VEH.
30H
MONTH
PICK-
MARKET
UPS
VALUE
DAY
DRIVER
ZONE
TRIPS
//)
HOMt
STORIES
/
TYPt
PERSONS
L-DIST
Vfc n
P*"INT 114,
1 HOMID, NMONTh
2 NTRIPD. NSTOR
114 FORMAT! UA8.1X 4XA2.4X 2(5XA1,4X)
1 4XA3.2X <*XA2»<«X 3ISXA1.4X) 3XAt»,3X
NUAY. HOMTYP.
LONG01S. NPKG
NPERS. NPASSV, NPKUP.
MKTVAL. NAZ67
XA?,^»A
-------
CALL DECODER(LONGDISt1»LONGD1S»K3)
CALL DECODER=.IwUE.
IF (HOMTYP.EO.lHh) HHCOL(/)=.TRUE.
\)<> IS? l=i»,7
152 IF (HHCOL(I)) GO TO 160
HHCOL(8)=.TRUE.
C
C UPDATE CAR/PAPKlNG DATA KOR HOUSEHOLD
C
160 Do 166 J=l»6
DO 166 1=1,8
166 I*' GO TO 190
M^INT 17?
172 ^)HMAT(1HO/ 3<»H CUWRENT HOME DATA UlSTWlbUTION - / )
P^INT 174
17-* FORMAT(30A 33<1H-) 14HHOUSEHOLD TYPE 33UH-) // 30X
i SDH ALL I-CAR MULTICAR I-FAM. DUPLEX
2 SOH 3«4PLEX APT OTHtW )
P^INT 176, ((HHLABL(I,J),1 = 1,?),(HHUATA(I,J),1 = 1 ,8) .J=l.n)
176 KORMATUX 2A10» 9X K(F8.0«2X) )
10-74
-------
190 If (NHOMREC.NE.l ) 00 TO 230
C
C ----- GFT NEXT HOUSEHOLD DRIVER TRIP DATA
C
200 If (NTENTRY.LT.NTLNTH.AND.NTENTRY.NE.O) GO TO 220
NTENTRY=0
NTRPREC=NTRPREC»1
210 BUFFER IN (2,0) < TRIPS ( 1 , 1 ) .TRIPS (£0, 36) )
til IF (UNIT, 2) 211,219, 217, 21b
217 ENOTRPS=123<».
(iO TO 600
216 NPAR2=NPAR2*1
219 NTLNTH=LENGTH<2>/20
220 NTENTRY=NTENTRY«1
DhCODE ( 200, 222 » TR I PS (1, NT ENTRY) )
1 TRIPID, PERSON, MODE, FRSWRK. BEGT1M, ENOIIH.
2 FWY. ONITE, PKG, NPASS, NOR1G, NDEbT. TRP1 YP
222 FOHMAKA7, 13XA1, <*6XA1, UXA1, 1XA5, 2XAS, 2XA3, 3x6?, Al,
1 2XA1, 24XA4, 11XA4, 37XA1 )
\r (TRIPID. F.O.MOMID) GO TO 230
DF.CODE <7.22
IF (IULTR.GT.26) GO TO 226
CALL DECODER) IDNM8k,6, IDNMBR,KONTENT )
IF (KONTENT.Nf.l) l->0 TO 2?6
00 To 600
NTOMIT=NT()M1 T«l
If (NTOMIT.LT.SOOO) GO Tu 200
CALL HEAi)(3)
FORMATO2H TKl^ll) IS NOT IN PWOPtR FORM - /|HI) / l"0 IOX ^AI
(.0 TO 900
it (r,ONTPOU (2) .LT . 1 ) GO IT' <:'60
1? (NTPIPb.fJO.O) PRINT 2<«<»
F'JRMAT (1HO/50H 10 PERSON MODE WORK HEM'"
1 SON LNU FWY OVER PARR TOTAL
2 30H ORIGIN DEST TRIP /
3 50H TRIP 1 1 IMf
<» 50H TIME NITK IN vKn
S 30M ZONE ZONE TYPE //)
PPINT 246,
1 TRIPID, PERSON, MORE, FRSwPK, BEGTIM, ENOTIM,
2 FWY, ONITE, Pn(i« I-JPASS« NORIG, NDEST, TRPfYP
F(>RMAT ( 1XA6, IX JlbXAl.^X) 2I3XAS.2X) <»XA3,3X
1 2<»
262 CONTINUE
GO TO 200
O^VSNAM ( 1 u« VK / -'-'f "'^0^l
C''-LL UECODLK(ONl ft ,tr, IGNI TE,K
10-75
-------
IF (KlFCOOER(PKG,l«IPKG.K13>
I1 (K13.NE.1) 1PKG=0
CALL DECODER ( NOR Kit 4 » 10R1G.K14)
IF (K14.NE. 1) 1GAR«L£=1
CUL DECODER (NDEST, 4, 10ESTtK15>
\> (K15.NE.1) 1GASBLE=1
IK ( IGARBLL.NE.O) 00 TO 200
C
C ----- CALCULATE TRIP DISTANCE
C
300 0=0.
IrXT=l
IF ( IORIG.LE. 1000. OP. IUEST.LF. . 1000) GO TO <*0'J
I<-'xT = 0
1' (IGNITE. tO. 0. AND. IGAkBLL. EO. 0) CALL U I S [<1 OR 1 G, I OhST .1) >
IF (U.GE.O) GO TO 400
liO TO 300
C
C ----- A^CWlHt TRIP TO CAR
C
IF (NCAPS.NIf .0) GO TO
M) TO ?00
•* 10 1CAR=1
C
C ----- jMCStMENT TPAVtL ARRAYS
C
SCO T-"P ( 1 ) =0
ThP (*») = 1 .
IK (IORIG.NE.lDf.sT) GO TO
T'-'P(?)=D
T*k(b)=l .
S10 IF (NOAY.LT.?.OP.MUAY.r,T.b) GO TO
T^p( J)=0
TPP(6)=1.
f>aO U ( lONITt .Nt .0) TRP(7) = 1.
Ic (IfXT .NK.O) r-(P(8) = l.
On 5^0 1=1.8
'.)'-' I VtL( I. IORVR)=DR1VEL ( 1 . IORVP) »TRP ( 1 )
IF (NCARS.Nf. .1 ) GO TO i4C
C.ARVtL ( I » ICAk )=CARVtL ( 1 . ICAR )»TRP(1)
T^P(I)=0.
IF CONTROL (i;) .L T.3) (-U TO 200
Kt;PMATb3 J=1.10
P^INT 564. ( J. (DRI VtL ( 1 . J) . 1=1
FORMAT ( 18. t'X H(Fb.ltiiX) )
FORMAT UHO sx BHCAWVEL - )
nn 5b7 J=l«10
'567 HPINT S64. ( J, (CARVfL ( I . J) . 1 = 1
no TO 200
C
C ----- CLASSIFY OKlVfcRS AlMO CARh,
C
10-76
-------
600 IF (NTRIPb.FO.O) GO TO 100
IF ( IGAFMLE.NEi.O) GO TO 830
<)0 604 Il>t.-V«=l . 10
60** If (DRJVtL (4. IUPVW) .Hi .0.) ND»VWS=10KVM
610 Of< 618 J=l .NDRVRS
IVIAS'S< 1. J)=.I»Ut".
0^ 612 I=£:,t,
612 I'^IASSI I.J>=. FALSE.
lr (NDRV*S.('T.lv.CAHb) GO TO 618
IF (NCARS.LT.S) OK 1 AbS •' 638 1 = 1,4
038 CARASS( I • 1 )=. FALSE.
IK (NCAMS.NE.U bO TO 700
IK (NDRVkS.LL.J) CAPASb(NUKVWS» 1 >=.TkUE.
r,
C ----- UPDATE TRAVEL Ul STK I MUT 1 0Mb
C
700 OH 740 J=1.40
UO 720 ID«VH=1 .NOWVKS
IF (DRIVELd flDKV*) .L T.ObT < J) .0«.
1 ONIVEL (l.IuRV*) .OE.OSTI J»l )) GO TC 710
DO 706 1 = 1. b
70ft IF (OK1ASS< I » lOPV^) ) DISTIllSI I i J)=DISTU1S( 1 . J) »1 .
710 IF (DRIVEL (4, IUHVV) .LT.T«IHNO( J) .Ok.
l DRIVEL <4, ID^VW) .of. TWIHNO< J»l > > GO TO 7^0
L'O 716 1 = 1 .ft
716 IF (OR1ASS< I « It)PV*> ) Tn'IPOIbl 1 *J)=TR1PD1S< 1 . J) +1 .
/?0 CONTINUE
IF (NCARS.Nf. . 1 I GO TO 7<«0
IF (CARVtLU . Rflk) .LI .UM U) .OK.
1 CARVEL ( 1 .ICAP) .GL.u^r (J*l I ) GO TO 7.10
f).i 726 1 = 1. t
T?n IF ICAM! .Of. . IKIuNO ' J« 1) ) GO 10 r<*(j
DO 7J6 1=1.4
73o If (CARASSI I » ICAH) ) T^lHL)l->( l*t>. J)=TK1PDIS( 1*6. J) »1 .
Ktd C')i\TlNUE
C ----- u-'DATF. OlSTKIt^OTlON
C
«00 0'~ t)06 J=l.rt
0-, 802 ICLASS=1»6
C>0 802 IDPVK=1 .NDPVKS
IF (O'ftl ASS ( 1CL. ASS. IDKV») ) CONOIS ( ICLASS. J) =CONOIS ( 1CL A>-S. J)
1 «l)f1 S ( 1 CL ASS»t> . J )
1 »CARVEL (J. 1 )
-106 CONTINUE
o-i 80* J= 1 . 10
(J) =0.
10-77
-------
On 808 1 = 1.8
0"IVEL(I»J)=0.
80* CARVEL(I.J)=0.
DO 814 ICLASS=1,6
00 814 IDRVR=1«NDRVRS
614 IF (DRIASSUCLASS.IDRVR) ) CONDIS < ICLASS.9) =CONDIS < ICLASS.9) « 1 .
00 816 ICLASS=1,4
ICAR=1
b!6 IF (CARASS(ICLASS.ICAR) ) CONDIS < 1CLASS»6.9> =C()NOIS < ICL ASS»b.9 >
1*1.
820 IF < IGARBLE.NE.O) NGARBLE=NGARBLE* 1
IGAR«LE=0
If (NGAR8Lt.GT.100) GO TO 900
IF (ENOTRPS.tO.1^34.) GO TO 900
If (KONTROLO) .LT.l) GO TO 100
CALL HEADO)
PRINT 902
PPINT 908, (DST(J) ,DST(J«1> . (OISTDIS ( I , J) ,1=1.10) ,J=1. 40)
CALL HEAOO)
ooiNT 922
DO 827 J=l,40
JJ=J-1
827 PPINT 928, JJ, < TRIPUIS< I , J) »I=1,10)
CALL HEAD(6)
PPINT 944
PPINT 946. (CONDX(J) , (CONUIS(I»J) , 1=1 » 10) , J=l ,9)
PWINT 982. NHOMREC*NTRPREC»NPAR1,NPAR2,NGAKBLE»NTAXTRP.NTRKTPP.
1 NHOMIT,NOUSE,NOCA«S,NOTRIPS,NOPERS,NTOMIT«N070N£
GO TO 100
C
C ----- COMPLETE AND PKINT TRAVEL DISTRIBUTIONS
C
900 . CALL HEAOO)
PPINT 901
PRINT 902
''01 FORMAT (42H PE.P CENT DISTRIBUTIONS OF DAILY TRAVEL - // >
•vo? FORMAT <
1 36h MILEAGE ALL DWIVtPS
2 *»6H OKlVERS WITH CARS BY CARS PER HOUSEHOLD
3 48H ONLY CARS BY NUMBER OF DRIVERS /
4 36H hANGE DRIVERS WITH CARS
t> 48H ONE TWO THREE FOUR
(5 48H ONE TWO THREE ALL //)
DO 906 1 = 1 « 10
TOTAL(I)=0.
DO 904 J=l .40
904 TOTAL* I)=TOTALU)*OISTD1S
PPINT 902
00 916 1=1.10
10-78
-------
DO 916 J=2,40
916 DTSTD1SI I,J)=DlSTDISd»J)*DISTDISdtJ-l )
PRINT 908, (DST(J) »DST(J»1> t (DISTDISd ,J> ,1=1,10) ,J=1.40)
920 CALL HEAD<3)
PRINT 901
PRINT 93?
922 FORMAT)
1 36H NO. OF ALL DRIVERS
2 4t)H OKIVERS WITH CARS BY CARS PEN HOUSEHOLD
3 4BH ONLY CARS RY NUMbER OF DRIVERS /
<* 3fc>H TRIPS DRIVERS WITH CARS
5 48H ONE TWO THREE KOUW
6 «,8H ONE TwO THREE ALL. //)
lid 42h 1 = 1.10
TOTAL ( I)=0.
DO 924 J=1.40
924 TCTALd)=TOTAL< 1 ) « TPIPDI S ( I , J)
DO 926 J=l,40
926 IF (TOTAL ( I) .N£ .0) TR IPU1 S ( 1 » J) =1 DO .MR IPO I S ( I , J) / I 0 T AL < 1 )
00 927 J=l,<.0
JJ=J-1
927 P^INT 9?«. JJ, (T»1PU1S( 1 • J) t 1=1 t 10)
928 FORMAT «U Ib, bX 10 lf'*i. 1 t«»X) >
930 CftLL HEADO)
911
922
1 = 1.10
IIP 936 J=2»4U
v3fj T^IPUISI I< J)=TRIHDIS( I t J)»TRIPL)IS( ItJ-1)
00 937 J=1.40
J.' = J-1
ill P*INT 928. JJ. (TPIPUISI 1 ,J) ,1=1, 10)
c
C ----- pi. INT DISTRIBUTION CONDITIONS
C
^i.0 CALL HEArj<6)
H(.'INT 942
Ui,.-: i-ORMAT(27h 1.1 1 S TR I duT 1 ON CONDITIONS - // I
1 36H ALL DRIVERS
4»ih Ot-lVt^S WITH CARS Bf CARS PEW HOIISFHOU,
4rtn 0\L> CArS ^Y ^,UMb^R OF u^IVKPS /
36*' DRIVERS «ITH CARS
> 4«h O':t T»'0 THREE KOUf
-«dH O.Nt TWO THREE ALL //)
HWINT 9^e>. (CON')A( J) t (CONOlSdt J) t I = it 10) t J=1.9)
FriPMATdX AlO. 11 10) . J- 1 . ^
•.OUi HUM DATA
10-79
-------
980 CALL HEAD(b)
PRINT 983, NHOMREC.NTRPRECfNPARl,NPAR2.NOARHLK.NTAXTKP,NTRKT u P,
1 ^'HOMIT«NOUSE. NOCARSfNOTRIPS«NOPEP.S»NTOMIT«NOZONE
98? FORMAT*
1 10XSOH NUMBER OK HOUSEHOLD RECORDS PROCESSED . I*//
1 10X50H MUMHER Of TRIP RECORDS PROCESSED If//
1 10X50H NUMBER Of PARITY ERRORS IN HOUSEHOLD DATA I*//
1 lOXbOH NUMBER OF PARITY ERRORS IN TRIP DATA Irt//
1 lOXbOH NUMBER OF HOUSEHOLDS ASSOCIATED WITH GARBLED DATA lh//
1 10XSOH NUMBER Of TAXI DRIVER TRIPS SKIPPED 1H//
1 10XSOH NUMBER Of TRUCK DRIVER TRIPS SMPPfl) I8//
1 lOX^OH NUMBER OK HOUSEHOLDS SKIPPED I'.V/
MjMrttP Ot aKIPPED HOUSEHOLDS WITH U'JUSEI! CARS IH//
1 OX50H MUMBtR 01- SKIPPED HOUSEHOLDS WITH Nd CA^'b Ifl//
NljMhEw Of 'DIPPED nOUSEHOLUb WITH NO ItJJP ':*1A i«//
NUMbt> OF SKJPPFU HOUSEHOLDS WITH NT, Ph*V':-;S !(•*//
«S^H NifMnER p)f .>K. (PPKU IKIPS wllf HA;.; Jb !«//
..
-------
SUBROUTINE -DECODER(DATA,NCHAR,NUMBER.KONTENT)
C
C DECODES FIRST NCHAR BCD CHARACTERS OF DATA, RETURNS
C NUMBER AND KONTENT AS FOLLOWS —
C
C NUMBER KONTENT DATA
C
C
C (INT.) 1 ALL CHARACTERS NUMtRIC
C 0 0 ALL CHARACTERS BLANK
C 0-1 ALL CHAR. BLANK, -, OP X
C 03 CHARACTERS OTHER THAN ABOVE
C
C IF NCHAR=1 ANU UATA IS A LETTER, KONTENT IS Stf 10 2 ANO
C NUMBER IS SET TO THE NUMBER OF THE LETTER IN THK AlPHABtT
C
DIMENSION IDIGUO) ,ICHAR(10)
DATA (1010(1),! = !,10)/1,10,100,1000,10000,1000000,1000000,
1 10000000.100000000,100000COOO/
UKCODE <10»<»»OATA) IChAR
<» K"iRMAT < 10R1 >
N=l $NN=NCHAR iKONTENT=l $NUMHER=0 *NULL=0
GO TO 11
10 N=N*1 iNN=NN-l
11 IF (ICHAR(NN).LE.1RZ) GO TO <*0
If ( I CHAR INN) .GT.1RSI) GO TU bO
IK (NULL.EU.O) ,MUMBER=NUMHER»(ICHAR(NN)-iRO)*!DIG(N)
20 JF (NN.GT.l) GO TO 10
IK (NULL.EU.O) RETURN
NMMBER=0 $KONTENT=0
IF (NBLANK.EO.NCHAR) RETURN
KONTENT=-1
IF (NULL.EO.NCHAR) RETURN
30 NUMBER=0 $KONTENT=3 SRETUhN
«0 IF ( ICHAR(NN).NE.1RX) GO TO 70
NDLL=NULL«1
M. TO 20
SO IK (ICHAR(NN).NE.lK ) GO TO 60
N>-LANK=NBLANK*1
NULL=NULL»1
GO TO 20
60 IF (ICHAR(NN).NE.1R-) GO TO 30
WiLL=NULL+l
GO TO 20
70 KONTENT=2 1NUMBER=ICHAR(1)
IK (NCHAR.EO.l) RETURN
GO TO 30
SUBROUTINE DECoDt'P
10-81
-------
SUBROUTINE DlST(ItJ,0)
C
C CALCULATES DISTANCE 0 IN MILES BETWEEN POINTS I AND J.
C READS AND SCALES COORDINATE LIST IF 1*0*
C RETURNS D EQUAL TO -1. IF ZON£ I OR ZONE J IS NOT IN INPUT LIST.
C
DIMENSION NM(7),CX<7),CY<7)«CZ(7)»NC(4,1340)»C<4,1240),NREF<100)
EQUIVALENCE
-------
45 FORMAT (40X 41H«»» TOO MANY INPUT CARDS FOR D1ST »•• )
C
C ESTIMATE 1NTKAZONAL DISTANCES AS ONE-HALF OF CtN TROIU-TO-CENT^O I n
C DISTANCE FROM A ZONE TO ITS NEAREST NEIGHBOR.
46 DO 48 N=1»NPOINT
IF )««2
IF (D.GT.DMIN.OR.M.EQ.N) GO TO 47
MMIN=M
D"IN=D
47 CONTINUE
C<4«N)=SORT(DMlN>/2.
C(4»MMIN)=C(4,N)
46 CONTINUE
C
C PREPARE AND PRINT TABLE OF ZONE NAMES. COOKO 1NATES • A NO 1NTRA20NAL
C DISTANCES. 200 ZONES PEW PAGE IN FOUk COLUMNS OF FIFTY tACH.
50 'MPTS=NPOINT
KPAGE=0
5>1 LINE = 0
KPAGE=KPAGE«I
CALL HEAD(2)
PKINT 5?
52 FORMAT (4(32H AZ67 X POS Y POS INTRA ) /
1 4(32H NAME Ml MI Ml ) // )
C
5b L1NE=LINE+1
Oo 58 KK = 1»<*
K=KK-1
I=LINE«50»K«200"(KPAGF-1 )
MM (KK)=NC ( 1» I )
CX (KK)=C(2»1 )
CY(KK)=C(3.I )
C /(KK)=C(4» I )
K ( I .GT. (NHTS-50) ) GO TO bO
58 CONTINUE
60 Pi^INT 62. < (NM(1) .CX( I) ,CY(I) ,Cl( I) ) .1 = 1.KK)
62 FORMAT (4( iti.jFd. 2) )
IF (I.EO.NPTS.ANU.KK.EQ.l) RETURN
IK (LINE.LT.bO) 0.0 TO Sb
If (LlNE.tU.t>O.ANi).KK.E0.4.AND. I .LT.NPTS) GO TO bl
C
C LOCATF NAMED POINTS IN COORDINATE LIST BY SEQUENTIAL SLAkCn STARTING AT
C KEY NflME LOCATION STORED IN NREF.
100 1^=1/100
JO=J/100
IF ( IR.GE.10t .OR.JR.t.E.100) GO TO 190
M=NREF (IR)
N=NREF(JR)
C
110 II-
-------
IF (M.GT.NMAA) GO TO 190
G^ TO 110
C
120 1^ (NC(lfN) .tO.J) GO TO 150
N=N*1
IF (N.GT.NMAX) GO TO 190
00 TO 1?0
C
C CALCULATE DISTANCE ASSUMING STRAIGHT LINt PATH FROM I TO
IbO IK (M.HQ.N) GO TO 152
ij=SOh(T «C(2«M>-C(2«N))«4»2»(C(3tM)-C(3tN>
Rr-'TUWN
152 U = C(*».M)
PFTUWN
C
190 D=-l.
FND SUHPOUT1NF DIST
10-84
-------
SUBROUTINE HEAO(N)
C
C HEADS PAGE WITH PROGRAM NAME. DATE. TIME. AND PAGE NUMBER. AND THEN
C SKIPS DOWN N LINES. IF N IS BCD. STOKES IT AS PROGRAM NAME INSTEAD.
C
IF (N.LT.56.AND.N.GE.O) GO TO 50
NAME=N
CALL MDATE(UATt)
RFTURN
C
50 KPAGE=KPAGE»1
CALL MTlMt(TIME)
PRINT 55,NAME,DATE.TIME.KPAGE
55 KORMATUH1 8HRKOGRAM A 1 0 . ?1 X Al 0 .30X Al 0 « ?2X4HPAGE It)
00 57 1=1,N
57 PPINT 58
58 FORMAT(1H )
RETURN
tMD SUBROUTINE HEAD
10-85
-------
10-86
-------
REFERENCES
1. D. Friedman, J. Andon, and W. Hamilton, Characterization of Battery-
Electric Cars for 1980-2000, General Research Corporation RM-1931,
August 1974 (also Task Report 1).
2. W. Hamilton and G. Kouser, Transportation Projections for the Los
Angeles Region. 1980-2000, General Research Corporation RM-1858,
November 1973 (also Task Report 3).
3. Reference 1, Summary Table.
4. J. C. Eisenhut, et al., Economic Impacts of Electric Cars in Los
Angeles, General Research Corporation RM-1904, August 1974 (also
Task Report 9).
5. LARTS Base Year Report: 1967 Origin-Destination Survey, Los
Angeles Regional Transportation Study, December 1971.
6. D. H. Kearin, R. L. Lamoureux, A Survey of Average Driving Patterns
in the Los Angeles Region, System Development Corporation
TM-(1)-4119/000/01, February 28, 1969.
7. S. J. Kalish, The Potential Market for On-The-Road Electric Vehicles,
Electric Vehicle Council/Copper Development Association, New York,
May 1971.
8. Compilation of Air Pollutant Emission Factors, second edition,
US Environmental Protection Agency, Office of Air Quality Planning
and Standards Research, Triangle Park, N.C., April 1973.
9. Air Quality Manual, Volume II, "Motor Vehicle Emissions Factors
For Estimates of Highway Impact on Air Quality," Federal Highway
Administration FHWA-RD-72-34, April 1972.
10. G. M. Houser, Population Projections for the Los Angeles Region,
1980-2000, General Research Corporation RM-1842, November 1973
(also Task Report 2}.
11. Automotive News Almanac 1972, R. L. Polk Company.
12. G. Katona, B. Strumpel, E. Zahr, Aspirations and Affluence, McGraw-
Hill Book Co., New York, 1971.
10-87
-------
REFERENCES (Cont.)
13. Reference 2, Figs. 14-15.
14. G. M. Naidu, G. Tesor, G. G. Udell, What Does the Consumer Want in
an Electric Car? Paper 7471, Third International Electric Vehicle
Symposium, Washington, 1974.
15. Interest Among the General Public in a Proposed Electric Car,
Opinion Research Corporation Caravan Surveys, March 1972.
16. G. G. Udell, G. M. Naidu, G. Tesor, Just How Big is the Consumer
Market for Electric Vehicles? Paper 7471, Third International
Electric Vehicle Symposium, Washington, 1974.
17. Final Report: Assessment of Domestic Automotive Industry Production
Lead Time for 1975/76 Model Years, Aerospace Corporation Report ATR-
73-(7321)-l, 15 December 1972.
18. L. E. Unnewehr, Electric Vehicle Systems Study, paper presented at
the Third International Electric Vehicle Symposium, Washington,
1974. '
19. J. J. Gumbleton, D. L. Frank, S. L. Gerslak, A. G. Lucas, Special
Purpose Urban Cars, SAE Paper 690461, May 1969.
20. Wall Street Journal. February 21, 1974, p. 30.
21. Automotive News, Almanac issue for 1959.
22. "Reassessing the Impact of Gasoline Prices," Business Week, July
27, 1974, p. 58.
23. T. C. Thomas, The Value of Time for Passenger Cars; An Experimental
Study of Commuters' Values, Stanford Research Institute, Menlo Park,
May 1967.
10-88
-------
|