Environmental Protection Agency
Region IX
100 California Street
San Francisco, California 94111
October 30, 1973
Technical Support Document for the:
San Francisco Bay Area
San Joaquin Valley and
Sacramento Valley
Intrastate Air Quality Control Regions
Transportation.Control Plan Final Promulgation
As Published In The
November 12, 1973 Federal Register
-------
I. Introduction:
The errata list in the following section II, is for the
purpose of making .the results of an EPA contracted air
quality implementation plan study, consistent with the
EPA transportation control plan, signed on October 30,
1973 by Acting Administrator John Quarles, and published
in the November 12, 1973 Federal Register.
In addition, two EPA Appendices are added to the amended
contractor's report for further clarification.
II. The EPA generated errata list is in reference to the
attached EPA contracted report:
Air Quality Implementation Plan Development
For Critical California Regions
Summary Report
- July 1973
Contract No. 68-02-0048
Prepared by
Transportation and Environmental Operations of
TRW, Inc.
. One Space '.Park
Redondo Beach, Calif.
The errata are as follows:
Ref. Pg. 13, Item 8.; This paragraph should be changed
to read as follows:
8* Pre-1966 Retrofit Device and 1966-1970 Retrofit
Device.The California Air Resources Board has
accredited two devices for reducing hydrocarbon and
oxides of nitrogen emissions from 1955-1965 vehi-
cles. These devices have thus far been required only
in the South Coast, San Diego, and San Francisco Air
Basins. The devices are essentially a vacuum spark
advance disconnect (VSAD) with a thermal override
switch to prevent overheating, or an electronic
ignition system. The State of California's 1966-70
retrofit device program since October 1, 1973, has
been on a change of ownership basis, and is now
scheduled for a mandatory phase to be initiated in
1975. Devices approved for this program include
VSAD-override, and exhaust gas recirculation systems.
-------
-2-
Ref. Pg. n, Item 9.; The following sentence should be
added to this paragraph:
While modified ground operations are not a part of
the EPA promulgation, it is estimated that the effects
of recent reductions in aircraft flights and the imple-
mentation of some modified ground operations initi-
ated as a result of fuel shortages, are approximated
by the TRW aircraft emissions inventory baseline
estimates.
Ref. Pg. 14 the heading - Phase II Measures (If Demon-
strably Warranted);This heading should be changed to
read as follows:
Phase II Measures (Measures 1., 3., and 4. were not
enacted by EPA, while measure 5, and a modified form
of measure 2, were enacted by EPA.)
Ref. Pg. 14, Item 2; The following sentences should be
added to this paragraph for clafirication:
The motorcycle operation ban discussed in this sec-
tion was not enacted. A regulation requiring the
control of new motorcycles starting with 1976 models
has been promulgated instead. The effects of this
change on the emission inventory are detailed in the
next section of these errata.
Ref. Pgs. 31, 44, 56, 63, and 70; The following changes
should be made:
On the above referenced pages, the strategies and
emissions reductions under the category "Projected
Reductions from Additional Optimistic Measures"
should all be disregarded, as none of the strategies
have .been promulgated by EPA. The EPA, instead of
banning the use of motorcycles during the smog sea-
son, is requiring that new motorcycles, starting
with 1976 models, must meet certain emission stand-
ards. It is estimated by EPA that this regulation
will result in the following reactive hydrocarbon
emission reductions in tons/day.
-------
-3-
1975 1977 1980
See Pg. 31: -0 -6 -15
See Pg. 44: -0 -2.0 -5.5
See Pg. 56: -0 - .3 -1.0
See Pg. 63: -0 - .6 -1.4
See Pg. 70: -0 - .8 -2.1
-------
EPA - APPENDIX A
Methodology for Determining the Base Year Oxidant Level
INTRODUCTION
This paper discusses a method for selecting the maxi-
mum values used in the calculation of emission reduction
requirements.
The methodology described in this paper is neither new
nor original. Dr. R. I. Larsen, Meteorology Laboratory,
NERC, Research Triangle Park, outlined such a technique in
1967 and has published numerous papers since that time ex-
plaining the use of his model in the establishment of
standards and in relating air quality measurements to such
standards (Reference 1, 2, and 3).
The rationale for selecting this method is outlined
and some of the advantages and shortcomings are covered.
A comparison of actual measured values with model calculations
is provided.
BACKGROUND
The development of a control strategy to achieve a
National Ambient Air Quality Standard is frequently based
on the premise that the concentration of a man-made
-------
- 2 -
pollutant In the ambient air is linearly related to the rate at which
the pollutant is emitted in the atmosphere. .
This assumption permits the use of a simple proportional (or
rollback) model to determine emission reduction requirements. Such
a model states that:
,,nn\
(100)
( current air quality) - (air quality standard)
(current air quality) - (background) - =
Current air quality is defined as the maximum measured concentration.
The development of the transportation control strategies did not
rely totally upon the rollback model. A non-linear relationship between
oxidant levels and hydrocarbon emissions developed by Schuck (See
Appendix B) was also employed. In some areas data was not available
for the verification of such a non-linear model and the simple proportional
relationship had to be applied.
Regardless of which of these models was used, the selection of an
appropriate maximum concentration was a critical factor in determining
the emission reduction requirements.
-------
- 3 -
There are several methods that can be used to
determine the maximum value needed for these"roll-back"
calculations. Among such methods are:
a. Diffusion modeling
b. Selection of a maximum value from a base year
c. Choosing the highest value over a number of years
d. Determining a maximum value by statistical analysis
Diffusion modeling, where validated models can be ap-
plied, probably represents the best method for determining
both the concentration and the location of high pollutant
levels. Unfortunately, a model with the required accuracy
is not yet available for determining specific oxidant
concentrations.
The selection of a value from a base year, where the
year is usually selected as the year of the latest emission
inventory, has the advantage of being most closely related
to the emission data. It also provides a convenient base
for comparing data at different locations. However, high
concentrations of oxidant occur under certain, as yet not
fully quantified, meteorological conditions and different
sets of these conditions may apply to the production of high
levels at different locations. Since meteorological para-
meters do not necessarily follow an annual cycle, the adverse
-------
- 4 -
conditions producing high levels may not always occur every
year at any given location. The data indicate that maximum
levels at a particular monitoring station may vary from year
to year by as much as a factor of two. High values within a
given region do not always occur at the same site and maxi-
mum concentrations selected from all stations within a region
may also vary considerably, although not usually by as much
as they do at a single location.
Extreme values can occur either because of unusual
meteorological conditions or because sane abnormal periods
would not necessarily be expected to occur every year but
perhaps only once in 5 or 10 years. Thus, the selection of
such an extreme value could require overly stringent control
measures. Conversely, abnormally low values could also be
selected if the data record is short.
A statistical analysis of data collected over a period
of years tends to smooth out the variations due to the
meteorology and to local anomolies. Such an analysis can
also provide a prediction with a specified probability of
occurrence and the extreme or outlying values can be weighed.
This paper compares the results obtained by applying a
particular statistical method to the calculation of maximum
oxidant levels with the actual measured maximum concentrations
at selected stations from data collected over the past three
years.
-------
THE ANALYSIS
Selection of Technique:
The objective of the analysis was to find an oxldant level (con-
centration) that represented the highest level expected to be achieved
with a frequency of one hour per year. The rationale for this objective
Is the National Ambient Air Quality Standards for oxidant: 160 ug/m3
(0.08 ppm) - maximum 1 hour concentration not to be exceeded more than
once per year.
Although there are a number of statistical methods that could be
applied, a technique described in the Office of Air Programs publica-
tion No. AP-89, "A Mathematical Model for Relating Air Quality Measure-
ments to Air Quality Standards" by R. I. Larsen, November, 1971, seemed
to best fit the objective. This model is based on the assumption that
the air quality data fit a log-normal distribution. There is some
disagreement about whether or not this is an appropriate assumption.
For example, Mitchiner & Brewer (5) have suggested the use of a
'double-exponential1 distribution. This is a widely known extreme
value technique. Their analysis, however, was limited to data collected
in three summer months and used only the maximum daily hour data. A
report by Mosher, Fisher, and Brunelle (6) indicates peak oxidant con-
centrations of 0.50 ppm or greater have occurred in Los Angeles County
In all months of the year except January and February. The selection of
only certain months could, therefore, tend to bias the results. Addi-
tionally, extreme value techniques seem most applicable to the selection
of an absolute maximum concentration and not necessarily to the
-------
- 6 -
concentration expected to occur once per year. However, a comparison
of the values calculated by the Mitcniner-Brewer method indicate.that
they do not differ greatly from Larsen's method, at least at the one
station covered in their analysis, even though a different data set was
used.
Larsen (7) analyzed all oxidant data for all California stations
for the period 1963-1967 and presented the cumulative frequency distri-
butions and a calculated maximum concentration for each station. The
tables in that publication were used in conjunction with later available
measured data to determine the location (or areas) of the highest con-
centrations. Stations within those areas were then selected for further
analysis. An attempt was made to obtain a three-year period of record
for each station. It was felt that the period should be comparable to
the latest emission inventory data available (in most cases this was 1970
data) and also should contain a sufficiently long period to help over-
come the problem of meteorological variability. A period of 5 to 10
years would have been desirable, but because of the changing patterns of
emissions and changing vehicular emission factors, it was felt that a
period longer than three years would tend to introduce more emission
variability than the meteorological variability that would be factored
out. Data for 1972 were not available so the period January, 1969, through
December, 1971, was selected. Unfortunately, there were many gaps in the
record and data was not available for some of the desired stations.
-------
- 7 -
Fourteen stations were finally selected for analysis and cum-
mulative frequency distributions for the selected stations for the
three-year period were then obtained. The data were analyzed according
to Larsen (4). A sample of the frequency distribution used is shown in
Figure 1.
CALCULATION OF MAXIMUM CONCENTRATIONS
The frequency distribution as given in Figure 1 is plotted on a
logarithmic probability graph as indicated in Figure 2. If the data
were perfectly log-normally distributed, all points on the graph would
be on a straight line. As can be seen in Figure 2, this is not the
case. However, the points in the frequency ranges from 10% to .01% do
appear to closely approximate a straight line. Since these are the
frequencies of most concern when considering very high values, only
those points are considered. To find the value that would be expected
once a year, Larsen (4) suggested using the .01 and the .10 frequency
points and extrapolating the line connecting these points to the
desired once per year frequency point. This was done for each location
for which frequency distributions were available. The extrapolation can
be done either graphically or mathematically. The mathematical method
is as follows:
-------
- 8 -
The desired frequency using this log-normal distribution is
obtained from
f = r-0.4 (100$)
n
where: r = the rank of the desired concentration if all
the concentrations were ordered from one
through the number of possible samples within
a selected time period
f = the frequency of occurrence in percent
n = total number of samples
FOR EXAMPLE: To find the frequency corresponding to the highest
one-hour average in a year, all of the 8760 one-hour averages in.a year
would be listed in order from 1 (the highest) to 8760 (the lowest). The
rank order, r, then is equal to 1, n, or the total number of samples, is
8760, and
f =1-0.4 (100$) = 0.
Next, the extrapolation of the data to this desired frequency, using
the two known concentration vs. frequency points, is as follows:
The equation of a straight line passing through two known points
and Xpyp is:
y - vi = x - xi
Y2~ vi X2~ xl
this can be rearranged so that
y = vi + Y2" J- (x~xi)
x2- x-L
-------
9
In this case the x and y without subscripts are the intercepts of
the unknown point on this line.
Where we are using the log-normal distribution, the y intercepts are
logarithms and the x intercepts are in terms of standard deviations from
the median. In a normal distribution each frequency can be located as a
distance (standard deviations) from the center of the profile (median).
If the y intercepts are logarithms, then the equation for the
straight line becomes:
In y = In y-
(x2-x1)
The concentration at an unknown point Tx,y' is then equal to the
anti-logarithm of
In y2
(In y + 71 (x-x,) )
( Y Y I
\ "O"^"! /
or to put it in another form: -^ y
concentration at y = exp [ In yp+ ^1 (x-x-^) ]
where 'expf indicates that 'e', the base of natural logarithms, is
raised to the power in the brackets. 'ef is approximately equal to
2.71828.
Following Larsen's suggestion (8), the two known points at the .01
and the .10 percentile levels are used to define the straight line we
wish to extend. Prom a statistical table, such as is given in
Reference 3 on Page 30 , the x intercepts at these percentile points
-------
- 10 -
can be determined. In the case of a log-normal profile, the .01
percentile point is 3.72 standard deviations from the median; the
.10 percentile point is 3.09 standard deviations; and the unknown point
at .00685$ is 3.81 standard deviations from the median. The y intercepts
are the concentrations at each of these percentile points. These x and y
values are then substituted into the above straight line equation and
the unknown concentration at the .00685$ frequency is determined.
To illustrate the procedure, the data from Figure 1 have been
rep lotted on Figure 3, and the points that are used below have been
labeled.
frequency concentration standard deviations
y(ppm) _ x _
.00685 to be determined(y) 3.8l (x)
.01 .27 (y-]) 3.72 (x,)
.10 .23 (yp 3.09 (x£)
Substituting these values into the straight line equation:
concentration at y = exp [ In. 27 + ^ (.2J} (3.81-3.72) ]
(3.09-3.72)
= exp [ -1.30933 + (-0.1603*0 (0.09) ]
(-0.63)
= exp [ -1.286424]
concentration at y = 0.28 ppm
From the example, a concentration of 0.28 ppm would then be the
highest concentration expected to be reached (or exceeded) once each
year.
These maximum concentrations were calculated for each of the
selected stations within each Air Quality Control Region. The results
are listed in Table 1.
-------
- 11 -
TABLE I. Hourly average concentrations for selected frequencies of
occurence.
LOCATION
South Coast AQCR
Riverside
Azusa
Pasadena
San Diego AQCR
San Diego (8th & E)
El Cajon
Sacramento Valley AQCR
Creekside
Chico
San Joaquin Valley AQCR
Fresno (So. Cedar)
San Francisco Bay AQCR
Livermore
San Leandro
Fremont
Percent of time given concentration
equaled or exceeded
0.10$ 0.01$ 0.00685$(Annual Maximum)
0.34 0.56 0.60
0.42 0.51 0.52
0.39 0.51 0.53
0.16 0.23 0.24
0.27 0.30 0.30
0.18 0.24 0.25
0.14 0.15 0.15
0.20 0.25 0.26
0.24
0.19
0.22
0.32
0.2?
0.27
0.33
0.28
0.28
Data used in this Table were hourly averages for the period of
January, 1969, to December, 1971.
-------
- 12 -
It should be noted that these calculated concentrations are not
necessarily the highest values to be expected. It is quite possible
that this value could be nearly twice as high on an unusually "smoggy"
day. Based on this analysis, however, such very "smoggy" days would
not normally occur every year.
COMPARISON WITH 'MEASURED MAXIMA
The calculated maximum values were compared with the actual maxi-
mum values that have been reported within each of the Air Quality Control
Regions since 1969. These values are shown in Table 2. In all cases
the calculated maximum concentration is within .03 ppm of the actual
measured maximum, even though an additional year of measured data was
considered and the high value for the region may have been reported at
a station other than one included in the calculations.
TABLE 2. Comparison of measured and calculated highest hour average
oxidant concentrations calculated.
Calculated Measured
AQCR Maximum Station Maximum
South Coast
San Joaquin
San Diego
S.P. Bay Area
Sacramento
.60
.26
.30
.33
.25
Riverside
Fresno
El Cajon
Livermore
Creekside
.62
.24
.32
.36
.28
Station
Riverside
Modesto
Escondido
San Leandro
Creekside
Year
1970
1972
1972
1971
1972
-------
- 13 -
The number of occurrences of concentrations in excess of the cal-
culated maximum within each Mr Quality Control Region was also tabulated.
For comparison, the daily maximum hourly averages from 1969-1972 were
used. The calculated maximum was equaled or exceeded three times in the
San Francisco Bay Area, once in 1969 and at two separate locations on
the same day in 1971. In the South Coast Basin the calculated concentra-
tion was exceeded once. In the San Diego Area twice, once each in 1971
and in 1972 in Sacramento once and once in the San Joaquin Air Quality
Control Region.
Again, it should be noted that the calculated value represents a
level that is expected to be reached or exceeded once per year and that
the analysis does not attempt to predict the highest possible concentra-
tion. Thus, the occurrence of a concentration greater than the predicted
value tends to verify the procedure if no other concentration measured
during the year was equal to or greater than the calculated maximum.
EVALUATION OF METHOD
The fact that the calculated values are close to the actual measured
concentrations and that the values have been reached or exceeded only
once in a given year, would tend to indicate that reliability of Larsen's
technique. There are, however, some obvious shortcomings to the analysis
presented here. A full three-year period of record was not available from
all of the air monitoring stations within each basin, nor from each of
-------
- 14 -
the stations listed In Table 1. The shorter the period of record that
Is available, the less reliable are the calculated values. To improve
the reliability, additional data should be analyzed and a larger sample
from each Air Quality Control Region should be selected.
Also, it was assumed that the stations selected represented the highest
concentrations within the given Air Quality Control Region. This is not
necessarily a valid assumption. Although only limited data is available,
newly established monitoring sites appear to be recording higher values
than some of the listed stations. For example, data from Escondido was
used to develop the strategy in the San Diego Air Quality Control Region.
The station was established in mid-1972 and the .32 ppm oxidant measured
there represents the highest concentration within the San Diego metro-
politan area in recent years. Agencies are usually continually expanding
their networks to include new areas of high concentrations, and additional
analyses should be performed as new data become available.
The calculations were based on the data measured during the years
1969-1971. They reflect only the emissions during that period of time.
Assuming no changes in emission patterns or emission controls at the
source, these values could be used to predict future air quality. However,
none of the areas considered are static with respect to growth, or to
the numbers and ages of motor vehicles in operation, or even with respect
to the numbers of and outputs from stationary sources. Some care should
be exercised in attempting to relate the concentrations to emissions in
areas of rapid growth.
-------
- 15 -
The oxidant data do not exactly fit a long-normal distribution and
the degree of fit varies at different locations. Thus, use of this
method may result in more reliable results in some areas than in others.
Also, the calculated maximum is quite sensitive to the selection of the
percentile points used in the calculations especially where the log-normal
fit is poor. Larsen (4) has suggested the use of the concentrations at
the .01 and .1 percent frequencies as being most representative of the
distribution of the higher concentrations. In some instances it appears
that the point at the .01 percentile fits the overall log-normal
distribution least well. The problem is particularly evident when a
short period of record is used. In most of the data examined, use of
the ol and the 1 or the 10 percentile points would result in higher
maximum levels than when the .01 percentile is included. This would
indicate that for some reason, probably meteorological, the maximum
possible values are not achieved. In other cases the .01 percentile
value seems too high. A study of the individual days could perhaps
provide an answer to the reasons why some of the high values seem out
of line.
The calculation of the maximum value is quite simple, but it does
require the preparation of cumulative frequency distributions. These
distributions are best processed by computer because of the large amounts
of data required. Once they are available, several other analyses can be
performed (see Larsen, 3)- Additionally, a comparison of the different
yearly and three-year distributions suggests a possible method for trend
analysis
-------
- 16 -
SUMMARY
Air quality data for a number of California air monitoring stations
were reviewed and analyzed according to a method suggested by Larsen.
The objective of the analysis was to determine a maximum oxidant concen-
tration for certain Air Quality Control Regions that could be related to
the available emission data and used to determine the emission reductions
needed to achieve the National Ambient Air Quality Standards.
Because of the variability of concentrations from year to year, at
least a three-year period of record would appear to be required for
analysis. This limits the selection of maximum concentrations to these
stations where data have been collected over that long a period and
could eliminate areas where higher concentrations are possible.
Although values obtained in this analysis compare favorably with
measured conentrations, other statistical approaches may provide equally
meaningful solutions and should be compared with this method. The method
outlined in this paper, however, is relatively simple and well documented
and is applicable to all pollutants. The use of some statistical approach
is certainly less arbitrary than the selection of one particular measured
concentration.
ACKNOWLEDGEMENTS
The author is indebted to Dr. R. I. Larsen of the Meteorology
Laboratory for his assistance and Mr. Don Worley of the Data Systems
Division for providing the necessary frequency distributions.
-------
- 17 -
. REFERENCES
(1) LARSEN, R. I. (1967) "Determining Reduced Emission Goals Needed to
Achieve Air Quality Goals - A Hypothetical Case," YAPCA 17, pp. 823-829
(2) LARSEN, R. I. (1969) "A New Mathematical Model of Air Pollutant
Concentration, Averaging Time and Frequency," YAPCA 19, pp. 24-30
(3) LARSEN, R. I. (197D "A Mathematical Model for Relating Air Quality
Measurements to Air Quality Standards," Publication AP-89, U. S.
Environmental Protection Agency, Research Triangle Park, NC 27711
(4) LARSEN, R. !. (1973) "An Air Quality Data Analysis System for Interrelating
1 i
Effects Standards, and'Needed Source Reductions," presented at WMO-WHO
Technical:Conference on; Observation and Measurement of Atmosphere Pollution,
!
Helsinki, Finland, July 30 to August 4, 1973 '
(5) MTTCHINERi T. L., and BREWER, J. W., "A Comment on the Method Used by
EPA to Calculate Required Reductions 'in Emissions,"'University of California
atDavis, unpublished
(6) MOSHER, T. C., FISHER, E. L., and BRUNELLE, M. F. (1972) "Ozone Alerts in
Los Angeles County,". Air Pollution Control District, County of Los Angeles, CA
(7) LARSEN, R, I.' (197D; "Air Pollution Concentrations as a Function of
Averaging Time and Frequency
(8) LARSEN, R. I. (personal communication, 1973)
-------
OX ID ANT
(PPMJ:
JANUARY
FRTGlNG'TTM-r'ATJ
LOS ANGELES, S.
It 1960TO
PERCENT
AVERAGING
5
10
15
30
1
3
3
8
"1?
1
2
4
" 7
14
1
f
3
6
1
2
~ 3
6
TIME MEAN
MIN 000.
000.
000.
000.
HOUR 0.03
0.03
6-9 AM 0.01
0.03
0.03
DAY 0.03
0.03
0.03
0.03
0.03
MONTH 0.03
0.03
0.03
0.03
YEAR ""0.03
0.03
0.03"
000.
MAX
000.
000.
000.
000.
0.33
0.27
0.05
0.21
0.12
0.10
0.09
0.07
0.06
0.06
0.05
0.05
0.04
0.03
0.03
0.03
"O.~03
000.
0.001
MIN PERCENT
000.
000.
000.
000.
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.03
0703"
000.
000
000
000
obo
069
072
073
073
073
074
074
075
075
075
075
075
075
075
075
050
100
000
. 0700
. 0.00
. 0.00
. 0.00
. 0.33
. 0.27
. 0.05
* 0.21
. 0.12
. 0.10
.0.09
. 0.07
. 0.06
. 0.06
. 0.05
. 0.05
.""0.04
. 0.03
."0.03
. 0.03
. '0.03
. 0.00
DEC.
SAN
PEDRO
ST.
STATION
001
3i r 1971 " "
OF TIME CONCENTRATION IS
EQUALED OR EXCEEDED
99.9
0.01
~o"7bo"
0.00
0.00
0.00
0.27
0.27
0.05
0.21
0.12
0.10
0.09
0.07
' 0.06
0.06
0.05
0.05
0.04
0.03
"0.03
0.03
0.03"
0.00
0.1
"0700"
0.00
0.00
o.otf
0.23
0.21
0.04
0.18
"o'.io""
0.09
0.09
0.07
0.06
0.06
0.05
0.05
0.04
0.03
"b;03
0.03
'0.03~
0.00
1
o'.oo
0.00
0.00
0.00
0.16
0.15
0.03
0.13
0.08
0.07
0.07
0.06
0.06
0.06
0.05
0.05
0.04
0.03
0.03
0.03
0.03
0.00
10
0700
0.00
~o.ob"
o.oo
0.07
0.07
0.02
0.07
0.05
0.05
0.05
0.05
"0.04
0.04
0.04
0.04
"0.04
0.03
0.03
0.03
0.03
0.00
20
O.OO"
0.00
0700
0.00
0.04
0.04
b.oi
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.03
0.00
30
0.00
0.00
0.00
0.00
0.03
0.03
0.01
0.03
0.03
0.04
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.00
40
"0.00
0.00
0.00
0.00
0.02
0.02
0.01
0.02
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Q.03-
0.00
50
"O.Ob
0.00
0.00
0.00
0.01
0.01
0.01
0.02
0.02"
0.03
0.03
0.03
0.03
0.03
"0.03
0.03
0.03
0.03
0.03
0.03
'0.03
0.00
60
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.03"
0.03
0.02
0.03
0.03
0.03
O."03
0.00
70
0.00
0.00
O.Ob
0.00
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.03
0.00
80
"0.00
0.00
"0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.03
0.00
90
0.00"
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.01
0.02
0.02
0.02
0.03
0.03
0.00
99
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.03
0.03
0.00
99.999
99.99
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.03
0.03
0.00
OiOO
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.03
0.03
0.00
. .
-
99.99
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.03
0.03
0.00
-- --
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
b.o
0.0
0.0
0.0
0.0
u.u
0.0
0.0
0.0
O.Q
0.0
b.o
0.0
Figure 1. Cumulative frequency.distribution of pxldant concentrations for Los_Angeles, Downtown station.
-------
.10
.00
.07
.Cfc
.04
.01
\.c
-' !M
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION IX
-------
10
0.9
0.8
0.7
0.6
5.81 5.72. 509 235
1.20
CW2 0
£ 0.4
0*
o.\
"i i
Ml
pirifei
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION IX
-------
EPA - Appendix B
Estimated Impact of VMT Reduction Measures
A. San Francisco Bay Area
Principal transportation data source for the San Francisco
Bay Area is the work of the Bay Area Transportation Study Com-
mission (BATSC), predecessor agency to the Metropolitan Trans-
portation Commission (MTC) which now coordinates transportation
planning activities in the area. The mode choice model used by
TRW-in assessing effectiveness of VMT controls is a model deve-
loped by BATSC in 1968 from 1965 travel data. To use the model,
TRW-took three .steps: 1) BATSC data was reaggregated and mode
chaipef./gu3ave:s--mqdified. to reflect evaluation requirements; ^2)
made several basic assumptions in equating travel time only;
3) derived relationships for converting increases in transit
ridership to VMT reduction.
Modified BATSC Model -
The BATSC model relates the time advantage of auto or
transit travel with percent of total persons using transit.
The pattern of home based work and related business trips,
referred to as "work trips" approximates rush hour travel.
The remaining travel is grouped as "other trips" and closely
-------
-2-
approximates typical midday travel patterns. (For detailed
description see "Page 12.) In almost every way this model is
similar to the Voorhees model, with slight modifications in
the parameters to more accurately reflect conditions in the
Bay Area.
Non-Work Trips to San Francisco*
Figure 4-7 shows mode choice relationships for non-work
trips to San Francisco CBD and non-CBD locations. Note the
steep slope to the CBD oriented curve, while a much flatter
curve applies to -travel to outlying San Francisco. Current
midday transit use to downtown San Francisco is about 20 per-
cent. Midday transit use to non-CBD locations ranges between
14 percent for San Francisco origins to 10 percent for origins
outside the city.
* The following 32 pages are primarily excerpts from the
TRW study of July 1973 for the San Francisco Bay Intra-
state AQCR.
-------
-3-
Non-Work Trips Destined Outside San Francisco
Figure 4-8 illustrates mode choice behavior for non-work
trips to Oakland-Berkeley and to the remainder of Bay Area
(outside San Francisco, Oakland, and Berkeley). Only the
Oakland-Berkeley curve indicates potential diversion from
auto to transit given changes in relative auto/transit travel
time. Even here though, only modest increases in transit
ridership appear possible.
Equating Travel Cost and Time - The Concept of Impedance
To analyze the effects of change in toll, parking costs,
transit fare or other costs associated with auto or transit
travel, several assumptions were made to equate costs to travel
time. The sum of travel time and cost equivalent, expressed as
impedance (Equation 4.1), reflects the total effort involved in
making any given trip. Relationships similar to Equations (4.1)
and (4.2) have been developed by DeLeuw, Gather and Company,
Alan B. Voorhees and Associates, Inc., and also others for
application elsewhere. Specific assumptions for time penalties,
income, and other variables are hypothesized for general analy-
tical purposes only in the absence of data applicable to the
Bay Area. Assumptions are believed valid for general order of
magnitude comparisons.
-------
-4-
CO
D-
o
co
0
60
40
20
To San Francisco CBD
To Elsewhere in San Francisco
0 1.0 2.0 3.0 4.0 5.0
TRAVEL TIME RATIO (Transit Time/Auto Time)
Figure 4-7. Mode Choice for Non-Work Travel
to San Francisco
CO
CO
Q.
o
CO
60
40
Q.
fe 20
i
UJ
o
Jo Oakland-Berkeley
iTo Elsewhere in Alameda, Contra Costa,
San Mateo, Marin, Santa Clara, Solano,
Sonoma & Mapa Counties
0 1.0 2.0 3.0 4.0 5.0
TRAVEL TIME RATIO (Transit time/auto time)
Figure 4-8. Mode Choice for Non-Work Travel
to Locations Outside San Francisco
-------
Assume I
+ F1
(4.1)
where
W
W
F
i
Assume
where
I
W,
0
T
P
i
Impedance (net time effort + money)
required for transit trip.
Walking time x 2 (penalty perceived)
Wait time x 3 (penalty perceived)
Ride time
Transit fare
Factor relating fare to time
(based on average hourly wage),
here assumed constant at 1 hour per $3.
= d] +
+ (0 + T + P)i
Impedance (net time effort + money)
required for auto trip.
Driving time under free flow conditions
Stop and go drivinq time x 2
(penalty perceived)
Time spent at stand still x 3
(penalty perceived)
Walk from parking to destination
Out of pocket vehicle operating expense;
here assumed to be 5^/mile
Toll
Parking cost
1 hour per $3
(4.2)
-------
-6-
CO
O-
HH
cc
90
80
70
60
50
40
o
00
OL
UJ
O.
u. 30
o
Ul
o
cc.
LU
a.
20
10
0
0.5 1.0 1.5 2.0 2.5 3.0
RATIO OF TRAVEL IMPEDANCE -- AUTO/TRANSIT
Figure 4-9. Bay Area Mode Choice Reflecting
Travel Cost and Time
Source: Postulated from BATSC
Mode Split Data
-------
-7-
Figure 4-9 shows the Bay Area mode choice curve postulated from the
foregoing equations and the BATSC travel data. Points on the various
BATSC mode choice curves were converted to impedance, knowing transit
fares, wait times, tolls, parking costs, etc.
Sensitivity of Impedance Variables to Control
Table 4-7 describes potential control over the various impedance
variables. Lower transit fares and increased driving time, vehicle
operating costs, and tolls can be applied with a minimum of administrative
problems; their effect is immediate. Improved transit services require
response time to become effective and offer greatest potential when
applied in conjunction with other controls. Parking controls may be
effective but pose administrative problems because of the multiplicity
of ownership and special circumstances likely to arise. Some loss of
effectiveness in parking control is likely as wives drop husbands off at
work, drivers circle the block while riders shop, etc. The effectiveness
of measures to increase walking time for auto users is difficult to pro-
ject, with a likely side effect being to merely add a short transit
shuttle trip to the end of a long auto trip.
Using the Concept of Impedance - 4 Examples
Figure 4-10 compares transit and auto impedances for four Bay Area
work trip situations. Each round trip originates in Concord. For analysis
purposes BART was considered open under the Bay to San Francisco and
frequent local bus service (10 minute headways) was assumed available in
central Contra Costa County. Impedance for a local trip to Walnut Creek
involving a round trip of 10 miles consists primarily of travel time --
riding, driving, waiting and walking. Reducing transit fare would have
little effect on the ratio of transit/auto impedances. Doubling the
price of gasoline, which would increase mileage costs by 50 percent,
also would have little effect. Significant diversion from auto to transit
wfjld appear to require direct limitations on auto usage, e.g., gas
rationing, restricted car ownership. Significant short term increases in
bicycling and walking or elimination of unnecessary trips appear to require
similar vehicle controls for effectiveness.
-------
-8-
0)
o
c
O)
10
4 Hr
Walnut
Creek
Port of
Oakland
San Francisco
CBD
Mileage
.50
3 Hr
2 Hr
1 Hr
MONEY
Figure 4-10. Impedance for Round Trip from Concord, Work Purpose
-------
-a-
Table 4-7. Control of Impedance Parameters
Impedance
Parameter
Potential
for Control
W, (walk time) Low
W2 (transit wait) Medium
W3 (transit ride) Medium
F (transit fare) High
i (income) Very low
d, d0 d0 (driving High
1 z 3 time)
0
T
P
(vehicle
operation)
(toll)
(parking)
High
High
Medium
Example(s) of Control Aimed
at Decreasing In/Ij * _
More bus stops and/or routes;
auto free zones
Improve frequency of service;
Exclusive bus lanes; rail rapid transit
Lower fares
Lower personal income levels
Limit number of traffic lanes in use
Gasoline tax, "smog" tax
Raise bridge tolls, freeway tolls
Limit on-street parking, parking space tax,
restrict new parking construction
See Impedance Equations (4.1) and (4.2).
A round trip to downtown Oakland from Concord involves an
impedance of about 2 hours and 20 minutes, whether by auto or by transit.
About 50 percent transit usage should be expected under such circumstances.
Out of pocket costs fare, mileage, and parking constitute a larger
share of impedance than in the case of the local trip. Eliminating fare
would reduce the transit/auto impedance ratio to about 0.6, a measure
which conceivably could raise transit usage to 65 percent. Traffic
flow restrictions along the freeway, e.g., elimination of lanes through the
Caldecott Tunnel, could increase the competitive advantage of transit
still further.
A round trip to the Port of Oakland represents longer trips to a
non-CBD location, one which typically requires a public transit rider
to transfer one or more times to reach his destination. Direct transit
service offered on a subscription basis would be one way of making the
-------
-10-
transit ride more competitive with the auto. It would be difficult in
this case to better the 1.25 transit/auto impedance ratio, since BART
holds a time advantage due to separate right of way. Eliminating transit
fare could reduce the impedance ratio to 1.0. Inducing constraints on
travel through the Caldecott Tunnel could have a similar effect.
The fourth situation involves travel to the San Francisco CBD.
Impedance associated with this round trip is quite high -- almost three
hours by transit and four hours by auto. The transit/auto impedance
ratio should be about 0.7 when BART opens, meaning about 63 percent
transit usage. For further reduction in auto use, reducing transit fare
would be one of the most effective measures. It would take about a $5
increase in bridge toll or parking fees to obtain the same 0.5 impedance
ratio possibly by reducing transit fare to a flat 25 cents in each direc-
tion. Using a combination of very low transit fare, increased bridge
and restrictions on auto travel across the Bay Bridge, a 0.4 impedance
ratio appears achievable. This translates to roughly 75 percent of work
trips from Concord to San Francisco CBD by transit.
4.1.4.1.3 Converting Percent Transit Usage to VMT Reduction
Mode choice curves provide the basis for estimating trips diverted
from auto to transit for various VMT control strategies. To convert trip
diversion to VMT reduction requires attention to trip length and auto
occupancy, factors which vary with trip purpose and origin/destination
combination. Transit experience indicates more long trips may be diverted
to transit than short trips with various control measures. To not consider
length of trip in assessing VMT reduction would tend to underestimate the
effectiveness of controls.
Fortunately, BATSC 1980 vehicle travel forecasts were available in
a manner which permitted regional VMT to be disaggregated for the same trip
purposes and origin/destination combinations used in projecting percent
transit usage (Appendix C, Tables C-6, C-7, C-8).
-------
-11-
Results of VMT Reduction Analysis
Quantitative analysis found that control strategies directed toward
VMT reduction would be no more than about 15 percent without imposition
of gas pricing or rationing measures (see Pgs. 30-31). The most effective
measure would be controls over traffic toward major generators such as
the San Francisco CBD and major employers in outlying areas. Work travel
over longer distances encountering traffic congestion, toll and parking
costs are the most subject to diversion from auto to transit. Short
local trips are most difficult to divert to transit. In the short time-
frame of the next 2 to 4 years, walking, bicycling or local transit will
only become attractive options to the automobile under more stringent
and direct controls such as gasoline pricing or rationing.
The most desirable strategy would be one which offers incentives
for transit use or carpooling commensurate with restraints imposed on
auto use. Reduced transit fare coupled with increased bridge tolls, for
example, should be publicly more acceptable than imposing a constraint
without offering an attractive option. Major suburban employers may wish
to financially underwrite subscription bus service with parking fees
imposed on employees who drive. The recommended first line of attack
would be reliance on pricing mechanisms which make transit more attractive
while making driving less advantageous.
It is-recommended that restrictions on auto flow and parking be
viewed primarily as back up measures to pricing mechanisms and improved
transit. As increased toll/reduced fare begins to take effect, exclusive
bus lanes, bus and carpool lanes, or other traffic constraints could be
imposed to keep congestion at current levels and insure maximum effective-
ness of the pricing mechanism. Similarly, constraints on auto parking
could be imposed, e.g., prohibition of additional off-street spaces, to
make certain parking costs stay at or above their current levels.
New major transit improvements, BART in particular, will likely
bring temporary reductions in traffic congestion along freeways parallel
to BART and a tendency for parking lot operators to lower parking rates
to attract greater usage.
-------
-12-
ESTIMATING MODE CHOICE AND VMT REDUCTION
To perform mode choice analyses required by the project within the
time constraints, a certain amount of regrouping of existing data had to
be done. Trip purposes and traffic zones were aggregated to conserve
unnecessary effort. Meanwhile, some disaggregation and experimentation
was necessary to get origin/destination patterns grouped in a manner
suitable for assessing VMT reduction. Each procedure in part necessitated
checking and some modification of BATSC mode choice curves.
1. TRIP PURPOSES
BATSC mode choice curves were developed for three trip purposes.
To simplify analyses, the three purposes were reduced to two:
a) Work trips, including BATSC home based work and related
business trips; these trips approximate rush hour travel patterns.
b) Non-work trips, including BATSC home based all other except
work and school trips and non-home based trips; these trips
approximate off-peak or midday travel patterns.
2. ZONES AND SUPERDISTRICTS
BATSC data were tabulated into 291 zones. For analytical purposes,
the zones were grouped into 30 larger geographic units, referred to by
BATSC as "superdistricts."
3. ORIGIN/DESTINATION COMBINATIONS
BATSC aggregated zonal data into six origin (production)/destination
(attraction) combinations as a basis for determining mode choice curves
from 1965 data (Figures C-l, C-2, and C-3).
o From low density residential to regional CBD
o From low density residential to other
o From medium density residential to regional CBD
o From medium density residential to other
o From high density residential to regional CBD
o From high density residential to other
-------
70
c
to
JD
l/>
CX
O
CO
S-
OJ
CL
e
at
o
Q)
O.
o
o:
(/>
z
o;
P CODES:
1 Low Residential Density
2 Medium Residential Density
3 High Residential Density
A CODES
1 = Regional CBD
2 - Other
60
50 1 1
40
30
20
10
CO
I
1.0 2.0 3.0 4.0
TRAVEL TIME RATIO (Transit time/auto time)
5.0
Figure C-l. Bay Area Transportation Study Modal Split Model
Home Based Work and Related Business Trips
-------
^r 7o
M
C
2
>, 60
jQ
Q.
^ 50
0»
Q.
40
30
g 20
z 10
fC.
P CODES:
1 Low Residential Density
2 Medium Residential Density
3 High Residential Density
A CODES
1 = Regional CBD
2 = Other
1.0 2.0 3.0 4.0
TRAVEL TIME RATIO (Transit time/auto time)
5.0
Figure C-2. Bay Area Transportation Study Model Split Model
Home Based All Other Except Work and School Trips
-------
70
(X
r-
i-
0*
CL
O»
u
O)
CJ
O
DC
60
50
40
^ 30
20
10
P CODES
1 Low Residential Density
2 Medium Residential Density
3 High Residential Density
A CODES
1 = Regional CBD
2 = Other
1.0 2.0 3.0
TRAVEL TIME RATIO (Transit time/auto time)
U1
1
Figure C-3. Bay Area Transportation Study Modal Split Model
Non-Home Based Trips
-------
BATSC defined low density as zones with less than 10 dwelling units
per net residential acre, medium density as 10-30 dwelling units per acre
and high density as over 30 dwelling units per acre. Only San Francisco,
about 50 percent of Oakland and Berkeley and 10 percent of northern San
Mateo County and Richmond fall under the definition of medium or high
density. This residential density distinction, when applied by superdistrict,
separates trips originating in San Francisco from trips originating in
remaining low density areas. To avoid losing the distinction between lower
and higher density travel characteristics in the East Bay (due to
aggregation) trips in and out of Oakland-Berkeley were treated separately
from other Alameda-Contra Costa County travel.
The BATSC regional CBD designation included San Francisco and Oakland
central business districts, a grouping which underestimates transit usage to
the San Francisco CBD and overestimates transit use to downtown Oakland.
To test strategies related to control of travel to downtown San Francisco,
it was important to separate out the San Francisco CBD. Oakland CBD was
aggregated with other Oakland-Berkeley travel.
4. TRIP LENGTH
Intradistrict trips were assumed to be local, whereas interdistrict
trips were considered intercity trips.
5. MODIFICATION OF MODE CHOICE CURVES
Once 1965 travel data had been aggregated by work and non-work
purposes, by intradistrict and interdistrict orientation, and by
destination San Francisco CBD, elsewhere in San Francisco, Oakland-
Berkeley, and elsewhere in the Bay Area (Tables C-l and C-2) -- mode
choice curves were developed according to the following procedures:
f District to district travel times via transit and automobile
were obtained from computer tapes of BATSC 1965 data.
Distinction was made between peak hour and midday travel
times. A single zone in each district was selected for
computing district to district travel times (Tables C-3,
C-4, C-5).
0 Ratios of district to district auto/transit travel times
were plotted against percent of person trips by transit.
Separate plots were developed for (1) trips destined to
San Francisco CBD, (2^ trips destined to the remainder of
-------
Table C-l. 1965 Trip Origins and Destinations, Work Purpose
Daily Person Trips
Daily Transit Trips
Percent Trips by Transit
Destination
San Francisco CBD
Elsewhere in
San Francisco
Oakland-Berkeley
Remainder Alameda
County
Remainder Contra
Costa County
San Mateo County
Marin County
Santa Clara County
Solano County
Sonoma County
Napa County
TOTAL
Source: 1965 BATSC
Origin Origin in Origin
Within San Francisco Elsewhere
District (Mncludes (Other Total
(Local Oakland- Intercity Person
Travel) Berkeley) Travel) Trips
.
107,100
220,800
134,400
134,500
64,100
39,300
261 .800
58,700
60,600
22,400
1,103,700
Trip Table,
241 .5001
89 ,400
75,300*
37,700*
17,100*
29,400
2,800
0
0
0
0
493,200
Productions
186,800
83,800
146,300
38,700
52,300
93,800
19,300
276,400
19,100
20,800
6,100
943,200
(origins)
428,300
280,300
442,400
210,800
203,900
187,300
61 ,400
538,200
77,800
81 ,400
28,500
Origin
Within
District
(Local
Travel )
_
20,600
30,600
2,200
2,500
1,300
500
3,200
1,200
200
0
Origin in Origin . Origin Origin in Origin
San Francisco Elsewhere Within San Francisco Elsewhere
^Includes (Other Total District (*Includes (Other
Oakland- Intercity Transit (Local Oakland- Intercity Total
Berkeley) Travel) Trips Travel) Berkeley) Travel) Trips
125.8001
26,500
17,500*
4,100*
1,100*
5,300
600
0
0
0
0
2,540,300 62,300 180,800
and Attractions (Destinations)
69,100
6,600
4,300
100
500
2,700
400
3,200
900
0
0
88,100
194,900
53,700
52,400
6,400
4,100
9,300
1,500
6,400
2,100
200
0
331 ,000
521
19 30
14 23*
2 11*
2 6*
1 18
3 20
1
6
1 .
0
5 37
37
8
3
1
1
3
2
1
5
0
0
9
45
19
12
3
2
5
2
1
3
0
0
13
1
Includes all trips originating in San Francisco destined to the CBD
-------
Table C-2. 1965 Trip Origins and Destinations, Non-Work Purpose
Daily Person Trips
Daily Transit Trips
Percent Trips by Transit
Destination
San Francisco CBD
Elsewhere in
San Francisco
Oakland-Berkeley
Remainder Alameda
County
Remainder Contra
Costa County
San Mateo County
Marin County
Santa Clara County
Solano County
Sonoma County
Naoa County
Origin
Within
District
(Local
Travel )
_
620,900
1,018,700
807,600
922,300
822,000
256,600
1,494,300
268,900
331 ,300
109.900
Origin in Origin
San Francisco Elsewhere
("Includes (Other Total
Oakland- Intercity Person
Berkeley) Travel) Trips
534, 2001
247,400
140,300
66,700
47,700
55,700
12,100
0
0
0
0
63,400
106,100
149,500
65,200
102,900
180,300
56,400
1,147,600
35,400
69,800
22,400
597,600
974,700
1,308,500
939,500
1,072,900
1,058,000
325,100
2,641,900
304,300
401,100
132,300
Origin
Within
District
(Local
Travel)
60,200
56,000
31 ,200
57,800
53,000
14,400
26,300
5,100
31 ,200
6,900
Origin in Origin Origin Origin in Origin
San Francisco Elsewhere Within San Francisco Elsewhere
("Includes (Other Total District ("Includes (Other
Oakland- Intercity Transit (Local Oakland- Intercity Total
Berkeley) Travel) Trips Travel) Berkeley) Travel) Trips
105.7001
60,000
12,200
1,700
2,300
4,400
1,000
0
0
0
0
12,100
10,300
12,200
1,800
3,900
8,700
1,300
10,900
1,200
8,100
2JOO
117,800
130,500
80,400
34,700
64 ,000
66,100
16,700
37,200
6,300
39,300
9^000
201
10 24
5 9
4 3
6 5
6 8
6 8
2
2
9
6
19
10
8
3
4
5
2
1
3
12
10
20
13
6
4
6
6
5
1
2
10
7
1
M
CO
1
TOTAL
6,652,500 1,104,100 1,999,000 9,755,900 342,100 187,300
72,600 602,000
17
Source: 1965 BATSC Trip Table, Productions (Origins) and Attractions (Destinations)
Includes all trips originating in San Francisco destined to the CBD
-------
-19-
Table C-3. Trips to San Francisco CBS (Superdistrict 1), 1965, Work Purpose
Origin Auto/Transit Total Total Percent Trips
(Superdistrict) Travel Time Person Trips Transit Trips by Transit
1
2
3
4
5
6
7
16
17
18
15
14
20
30
29
6.08/4.46
17.6/13.1
33.0/18.4
33.0/23.4
53.5/33.1
63.7/42.5
78.6/60
49.5/34.5
57.5/41.3
67.9/37.4
74.8/51.7
116.4/65.2
72.5/57.6
53.2/35.8
56.1/47.6
= 1.40
= 1.35
= 1.80
= 1.40
= 1.60
= 1.50
= 1.30
= 1.40
= 1.40
= 1.80
= 1.45
= 1.80
= 1.25
= 1.50
= 1.20
75,500
83,800
71 ,700
30,600
35,300
13,400
11,800
25,700
12,500
8,075
6,910
780
9,850
14,600
12,300
23,700
49,900
35,300
16,900
10,600
6,500
6,400
12,300
7,370
3,260
2,820
170
4,700
2,920
3,930
31
60
49
55
30
49
54
48
59
37
43
22
48
20
32
-------
-20-
Table C-4. Trips to Oakland (Superdistrict 16), 1965, Work Purpose
Origin Auto/Transit Total Total Percent Trips
(Superdistrict) Travel Time Person Trips Transit Trips by Transit
1
2
3
4
5
13
14
15
16
17
18
19
20
22
29
42.9/32.5 =
52.8/35.4 =
53.3/32.1 =
59.9/42.9 =
Poor Connect
No Transit
89.8/45.0 =
37.1/33..4 =
8.9/5.25 =
34.3/26.0 =
58.2/25.1 =
No Transit
46.2/37.2 =
Poor Connect
Poor Connect
1.30
1.45
1.70
1.40
2.0
1.1*
1.7
1.3
2.3
1.2
1,730
3,900
3,340
2,060
1,150
1,590
6,810
52,200
168,500
30,400
15,500
2,560
13,700
1,000
1,015
1,170
515
230
250
-
-
265
3,710
25,600
5,960
990
-
1,050
-
68
13
69
12
0
0
4
7
15
20
6
0
8
0
0
''Manual adjustment to 1.3
-------
-21-
Table C-5. Trips to Berkeley (Superdistrict 17), 19659 Work Purpose
Origin Auto/Transit Total Total Percent Trips
(Superdistrict) Travel Time Person Trips Transit Trips by Transit
1-4
15
16
17
18
19
20
68
83
34
7
53
119
97
.3/42
.5/41
.3/26
.7/4.
.5/22
.9/39
.9/27
.1
.2
6
.3
.9
.2
= 1
= 2
= 1
= 1
= 2
= 3
= 3
.60
.00
.30
.70
.40
.00
.60
3
6
29
52
24
3
10
,860
,330
,900
,300
,200
,200
,200
340
280
4,900
5,040
1,550
280
150
9
4
16
10
6
9
1
-------
-22-
San Francisco, (3) trips destined to Oakland-Berkeley,
and trips destined to other areas. Work trip (peak hour)
values were plotted separately from non-work trip (midday)
values (Figures C-4 and C-5)?
Curves were developed to fit district to district travel
data, first trying to utilize zone to zone mode choice
curves derived by BATSC, and as a second choice interpolating
between curves previously developed by BATSC. Mode choice
curves generally take the same shape from one geographic area
to another, making extrapolation from limited data less risky
than one might otherwise expect. Mode split curve assumptions
behind curves shown in Chapter 4 are identified below and are
documented in Tables C-6 through C-9.
Work Trips (Interdistrict and Intradistrict):
1. To San Francisco CBD from San Francisco BATSC Curve PA = 21
2. To San Francisco CBD from Elsewhere Adjusted BATSC PA = 11
Curve
3. To Remainder San Francisco from San Francisco -- Interpolated
BATSC Curves PA = 32 and PA = 22
4. To Remainder San Francisco from Elsewhere Interpolated
BATSC Curves PA = 22 and PA = 12
5. To Oakland-Berkeley from San Francisco-Oakland-Berkeley
Developed Curve to Fit BATSC Data
6. To Oakland-Berkeley from Elsewhere -- Developed Curve to Fit
BATSC Data
7. To Remainder Alameda County from San Francisco-Oakland-Berkeley
Developed Curve to Fit BATSC Data
8. To Remainder Alameda County from Elsewhere BATSC Curve PA = 12
9. To Contra Costa County from San Francisco-Oakland-Berkeley --
Same as No. 7
10. To Contra Costa County from Elsewhere -- BATSC Curve PA = 12
11. To Marin County from San Francisco BATSC Curve PA = 22
12. To Marin County from Elsewhere BATSC Curve PA = 12
13. To San Mateo County from San Francisco BATSC Curve PA = 22
14. To San Mateo County from Elsewhere ~ BATSC Curve PA = 12
15. To Santa Clara County BATSC Curve PA = 12
16. To Solano County -- BATSC Curve PA = 21
17. To Sonoma County BATSC Curve PA = 21
18. To Napa County BATSC Curve PA = 21
Non-work Trips (Interdistrict and Intradistrict):
1. To San Francisco from all Locations BATSC NBO Curve PA = 11,21
2. To Remainder of San Francisco from All Locations Curve to Fit
BATSC Data
3. To Oakland-Berkeley from Everywhere Curve to Fit BATSC Data
4. To Other Locations BATSC HBO Curve PA = 12
-------
-23-
70
60
50
to
>-
CO
CO
Q.
O
co
oa
LU
o.
u.
o
o
cc
40
30
20
10
San Francisco Origin
(BATSC PA=21 Curve)
Origin Outside
San Francisco
(BATSC PA=11 Curve)
Poorer
Coverage
and Tie
to CBD
29
I
1.0 1-5
TRAVEL TIME RATIO - AUTO/TRANSIT
Figure C-4. Percent Person Trips by Transit Versus Auto-Transit
Travel Time Ratio -- Example: Work Trips to San Francisco CBD
-------
50
>-
t->
40
30
From Oakland-Berkeley - San Francisco
(Used BATSC Curves PA=12 & PA=11)
20
From Outside Oakland-Berkeley-San Francisco
(Used BATSC Curves PA=12 & PA-22)
i
N)
1.0 1.5
TRAVEL TIME RATIO - TRANSIT/AUTO
B150»QAK
14
I
2.0
Figure C-5. Percent Person Trips by Transit Versus Auto-Transit
Travel Time Ratio -- Example: Work Trips to Oakland-Berkeley
-------
Table C-6. 1980 Trip Origins and Destinations, Work Purpose
Daily Person Trips
Daily Transit Trips
Percent Trips by Transit
Destination
San Francisco CBD
Elsewhere in
San Francisco
Oakland-Berkeley
Remainder Alameda
County
Remainder Contra
Costa County
San Mateo County
Marin County
Santa Clara County
Solano County
Sonoma County
Napa County
TOTAL
Source: 1980 BATSC
Origin
. Within
District
(Local
Travel )
108,000
234,800
199,300
185,800
184,300
77,300
369,600
91 ,400
103,200
38,300
1,595,000
Trip Table,
Origin in Origin
San Francisco Elsewhere
(Mncludes (Other Total
Oakland- Intercity Person
Berkeley) Travel) Trips
262, OOO1
116,300
108,800*
58,100*
39,200*
57,900
7,300
0
0
0
0
649,600 1
X-2 Transit
211,200
135,400
254,500
117,300
122,800
197,600
32,600
533,200
38,500
42,200
25,400
,710,700
Network ,
473,200
359,700
598,100
374,700
347,800
439,800
117,200
902.800
132,900
145,400
63,700
3,955,300
Productions
Origin
Within
District
(Local
Travel)
24,400
38,300
3,900
6.500
6.700
1,500
7,500
4,200
3,800
1,900
98,700
(Origins)
Origin in Origin
San Francisco Elsewhere
(Mncludes (Other
Oakland- Intercity
Berkeley) Travel)
161.3001
33,700
24,500*
7,300*
5,700*
15,300
1,800
0
0
0
0
107,900
12.2002
28,300
3,700
4,500
9,600
1,200
18,000
2,100
2,300
1,100
Origin
Within
Total District
Transit (Local
Trips Travel)
269,200
70.300 23
91.100 16
14,900 2
16,700 3
31 .600 4
4.500 2
25,500 2
6,300 5
6,100 4
3,000 5
249,600 190,900 539,200 6
and Attractions (Destinations).
Origin in Origin
San Francisco Elsewhere
(Mncludes (Other
Oakland- Intercity Total
Berkeley) Travel) Trips
621
29
22*
13*
15*
26
24
-
-
-
38
51
092
n
3
4
5
4
3
6
5
5
n
57
20
15
4
5
1
^ X
4 >
3
5
4
5
14
Includes all trips from San Francisco destined to the CBD.
2
Hand adjustment to compensate for inclusion of Southern Crossing in BATSC 1980 assumptions.
-------
Table C-7. 1980 Trip Origins and Destinations, Non-Work Purposes
Dally Person Trips
Dally Transit Trips
Percent Trips by Transit
Destination
San Francisco CBD
Elsewhere in
San Francisco
Oakland-Berkeley
Remainder Alameda
County
Remainder Contra
Costa County
San Mateo County
Marin County
Santa Clara County
SoVano County
Sonoma County
Napa County
Origin Origin 1n Origin
Within San Francisco Elsewhere
District (*Includes (Other Total
(Local Oakland- Intercity Person
Travel) Berkeley) Travel) Trips
47,000
998,300
907,600
936,400
946,000
411,000
1,573,300
346,200
422,300
152,500
471 .5001
348,200
237,200*
116.500*
115.200*
142.400
42.700
0
0
0
0
101.800
183.100
356.100
217,000
275,600
342,900
102,100
963,200
60,100
69.000
59,600
573,300
578,300
1 ,591 ,600
1,241,100
1.327,200
1 ,431 ,300
555,800
2,536,500
406,300
491 ,300
212,100
Origin
Within
District
(Local
Travel)
22.600
40.000
6.200
9.800
10,100
3,200
13.900
2.800
6.300
1.900
Origin In Origin Origin
San Francisco Elsewhere Within
(*Includes (Other Total District
Oakland- Intercity Transit (Local
Berkeley) Travel) Trips Travel)
158.0001
34.700
16,300*
3,600*
4,100*
8.900
1.700
0
0
0
0
23.000
6.8002
12.800
2.400
3.700
5,900
1,200
10.500
700
1,700
900
181,000
64.100
69.100
12,200
17,600
24.900
6,100
24,400
3,500
8,000
2,800
48
4
1
1
1
1
1
1
1
1
Origin 1n Origin
San Francisco Elsewhere
(*Includes (Other
Oakland- Intercity Total
Berkeley) Travel) Trips
341
10
7*
3*
4*
2
4
-
-
-
-
22
42
4
1
1
2
1
1
1
2
1
32
11
4
1
1
2
1
1
1
2
1
TOTAL
6,740,600 1,473,700 2,730.500 10,944,800 116,800 227,300
69,600 413,700
Source: 1980 BATSC Trip Table, X-2 Transit Network, productions (Origins) and Attractions (Destinations)
Includes all trips from San Francisco destined to the CBD
2
Hand adjustment to compensate for inclusion of Southern Crossing in BATSC 1980 Assumptions
15
-------
Table C-8: Percentage Breakdown of Bay Area VMT9 1980
Work Purpose VMT
Non-Work Purpose VMT
Destination
San Francisco CBD
Elsewhere in
San Francisco
Oakland-Berkeley
Remainder Alameda
County
Remainder Contra
Costa County
San Mateo County
Marin County
Santa Clara County
Solano County
Sonoma County
Napa County
TOTAL
Source: 1980 BATSC
Origin
Within
District
(Local
Travel )
0.2
0.7
0.8
0.8
0.7
0.4
1.7
0.5
0.3
0.1
6.2
travel data
Origin in
San Francisco
(*Includes
Oakland-
Berkeley)
0.41
0.4
0.9*
0.7*
0.5*
0.7
0.1
0
0
0
0
3.7
, zone to zone
Origin
Elsewhere
(Other
Intercity
Travel)
2.9
3.2
4.7
2.7
2.4
4.2
0.7
8.5
0.9
1.0
0.5
31.7
vehicle trips
Total
Work
VMT
3.3
3.8
6.3
4.2
3.7
5.6
1.2
10.2
1.4
1.3
0.6
41.6
multiplied
Origin
Within
District
(Local
Travel )
.
0.6
2.1
1.8
2.0
2.0
1.2
3.8
0.9
0.8
0.3
Origin in Origin
San Francisco Elsewhere
(Includes (Other
Oakland- Intercity
Berkeley) Travel)
0.61
1.1
1.5*
1.2*
1.3*
1.2
0.3
0
0
0
0
15.5 7.2
by travel distance and
1.9
3.2
4.3
3.3
3.4
4.5
3.4
8.7
1.0
1.1
0.9
35.7
sunned.
Total
Non-work
VMT
2.5
4.9
7.9
6.3
6.7
7.7
4.9
12.5
1.9
1.9
1.2
58.4
Total
All Purpose
VMT
5.8
8.7
14.2
10.5
10.4
13.3
6.1
22.7
3.3
3.2
1.8
100.0
Includes all trips from San Francisco destined to the CBD
-------
Table C-9. Work Trips by Transit for Selected Control Options
lesti nation
S.F. C°,0
Rer"ainr1er S.F.
Oak-Berk.
Ren. Ala. Co.
Contra Costa Co.
Marin Co.
San Mateo Co.
Santa Clara Co.
Solano Co.
Sonona Co.
:ana Co.
Origin
S.F.
Elsewhere,
Ren. S.F.Z
S.F.
Elsewhere
Oak-Berk. 2
S.F.. Oak., Berk.
Elsewhere
Ren-Alaneda2
S.F., Oak.. Berk.
Elsewhere
Contra Costa2
S.F., Oak., Berk.
Elsewhere
Marin Co.z
S.F.
Elsewhere
S.M. Co.2
S.F.
Elsewhere
S.C. Co.2
Elsewhere
Solano Co.2
Elsewhere
Sonona Co.2
Elsewhere
Ma pa Co.2
Elsewhere
Total Ray Area Work Trips
A
Transit ,
Inprovenents
I by
Transit
1980
62
51
23
29
9
16
22
11
2
13
3
3
15
4
2
24
4
4
26
5
2
3
5
6
4
5
5
5
13.fi
Increase
From
1965
10
14
4
-1
1
2
-1
8
0
2
2
1
9
3
1
4
2
3
8
2
1
2
-1
1
3
5
5
5
1.0
B
Imposition of SOt Toll
to S.F.
" by
Transit
1980
62
53
23
29
10
16
22
11
2
13
3
3
15
4
2
25
4
4
26
5
2
3
5
6
4
5
5
5
13.7
% Increase
Over A
.
2
.
1
.
.
.
-
_
.
.
.
.
1
_
_
.
.
.
.
.
_
.
.
-
0.1
* VTH
Reduction
0
0.12
0
0
0.03
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.15
c
Imposition of $2 Toll
to S.F.
% by
Transit
1980
62
57
23
29
12
16
24
11
2
15
3
3
17
4
2
28
4
4
30
5
2
3
5
6
4
5
5
S
14.2
I Increase
Over A
_
6
_
3
2
.
2
.
_
2
.
.
4
_
4
.
.
.
-
.
.
.
-
0.6
% VTM
Reduction
0
0.35
0
0
0.10
0
0.03
n
0
0.01
0
n
0.01
0
0
0.01
0
0
0.04
0
0
0
0
0
0
0
0
0
0.55
0
Free Fare
by
Transit
1980
64
74
26
31
16
18
25
20
2
15
10
3
17
10
2
27
10
4
28
10
2
10
5
10
4
10
5
10
18.4
% Increase
Over A
2
23
3
2
7
2
3
9
0
2
7
0
2
7
10
3
7
0 -';
2
5
0
7
0
4
.
5
.
5
4.7
t VTH
Reduction
0.02
O.R9
0.01
0.01
0.30
0.01
0.04
0.22
0
0.01
0.17
0
0.01
0.14
0
0
0.07
0
0.01
0.20
0
0.21
0
0.04
0
0.05
0
0.05
2.25
E
Traffic Restrictions
Transit/Auto = 1.0
X by
Transit
1980
62'
51
23
36
13
16
24
11
2
13
5
3
15
5
2
26
5
4
26
5
2
5
5
6
4
5
5
5
14.4
% Increase
Over A
-
-
-
7
4
.
2
.
.
.
2
.
-
1
-
2
1
-
-
-
-
2
-
-
-
-
-
-
0.7
r, VTTI
Reduction
0
0
0
0.04
0.12
0
0.03
0
0
0
0
0
0.05
0
0
0.02
0
0
0.01
0
0
0
0
0
0
0
0
0
0.27
F
Improved Local Trarsit
Trans it/ Auto = 1.4
f, by
Transit
1980
62
51
37
29
9
17
22
11
7
13
7
7
15
6
7
24
6
7
26
6
7
6
6
6
6
6
6
6
14.0
* Increase
'Over A
-
-
14
-
1
-
..
5
-
4
4
-
2
5
-
2
3
-
1
S
3
1
-
2
1
1
1
0.4
", VTM
Reductio
0
0
0.07
0
0
O.ni
0
0
0.04
0
0.04
0.03
0
0."5
0.02
0
0.01
0.02
0
0.04
0.03
0.25
0
0
0.01
0.01
0 '
0
0.63
Assures BART fully operational and Santa Clare County Transit District operating with 200 new buses.
2Local (Intradistrict) trips).
-------
-29-
Results of VMT Reduction Analysis
Quantitative analysis found that control strategies
directed toward VMT reduction would be no more than about
15 percent without imposition of gas pricing or rationing
measures (see Pgs. 30-31) . The most effective measure
would be controls over traffic toward major generators such
as the San Francisco CBD and major employers in outlying
areas. Work travel over longer distances encountering traf-
fic congestion, toll and parking costs are the most subject
to diversion from auto to transit. Short local trips are
most difficult to divert to transit. In the short time-
frame of the next 2 to 4 years, walking, bicycling or local
transit will only become attractive options to the automobile
under more stringent and direct controls such as gasoline
pricing and rationing.
The most desirable strategy would be one which offers
incentives for transit use or carpooling commensurate with
restraints imposed on auto use. Reduced transit fare coupled
with increased bridge tolls, for example, should be publicly
more acceptable than imposing a constraint without offering
an attractive option. Major suburban employers may wish to
financially underwrite subscription bus service with parking
fees imposed on employees who drive. The recommended first
line of attack would be reliance on pricing mechanisms which
make transit more attractive while making driving less advan-
tageous.
-------
-30-
Table 4-12. Summary of Impacts for VMT Reduction Strategies
Strategy Description
Approximate Percent
VMT Reduction
1. Intercept autos entering
San Francisco
o 50 cent added toll
- commute traffic only
- 24 hour basis
0.2
0.5
o $2 added toll
- commute traffic only 0.6
- 24 hour basis 1.4
o $10 added toll
- commute traffic only 3.0
- 24 hour basis 6.8
o Physical constraints
(bus and carpool lanes)
- half lanes reserved 3.0
o Reduce transit fares
- commute traffic only 2.0
- 24 hour basis 4.5
2. Suburban employer parking
restrictions/subscription
bus and carpooling
o 50 percent reduction in 2.5
auto use, firms 1000+
employees
o 75 percent reduction in 3.8
auto use, firms 1000+
employees
3. Moratoriums on development
o Outside transit service 0.4
area
o Major generators outside 0.7
established centers
Hypothetical
Implementing
Agency
MTC
MTC
MTC
EPA/MTC
MTC
EPA/MTC
MTC
-------
-31-
4. Traffic disincentives/transit
preferential treatment
o Excluding entrance to
San Francisco 0.2 MTC
5. Free transit fare
o Local Service 0.2 MTC
o Intercity, excluding 2.0
entering San Francisco
6. Improved local transit 0.7 MTC
7. Gas taxing/pricing
o 20 percent price increase 8-12 MTC
8. Gas rationing
o 80 percent current level 13-17 EPA
Maximum attainable without 15
gas pricing or rationing
-------
-32-
It is recommended that restrictions on auto flow and
parking be viewed primarily as back-up measures to pricing
mechanisms and improved transit. As increased toll/reduced
fare begins to take effect, exclusive bus lanes, bus and car-
pool lanes, or other traffic constraints could be imposed to
keep congestion at current levels and insure maximum effec-
tiveness of the pricing mechanism. Similarly, constraints
on auto parking could be imposed, e.g., prohibition of addi-
tional off-street spaces, to make certain parking costs stay
at or above their current levels.
New major transit improvements, BART in particular, will
likely bring temporary reductions in traffic congestion along
freeways parallel to BART and a tendency for parking lot oper-
ators to lower parking rates to attract greater usage. Some
evidence of reduced congestion has been observed on approaches
to the Caldecott Tunnel, and several parking lot operators in
the Oakland CBD have lowered rates, both presumably as a result
of BART opening to Oakland. Reduced freeway traffic and re-
duced parking demand is likely to be temporary only, unless
constraints are applied to keep driving costs at their present
level. To keep driving impedance high, constraints should be
employed concurrently with BART opening, including reservation
of bus and carpool lanes on freeways and arterials parallel to
BART (e.g., Bay Bridge, U.S. 101, 1280, Caldecott Tunnel, Mar-
ket Street in San Francisco, one-way street system from freeway
-------
-33-
off-ramps) and selected metering, narrowing or closure of free-
way off-ramps (e.g., 1-280 off-ramp at San Jose Avenue).
Restrictions that increase congestion above present levels are
likely to be perceived as restraint without regress.
Locally Implemented Strategies
The public.hearing testimony, written comments, and the
meetings with the State and local transportation officials
(San Francisco "Task Force") indicated a wide variety of mea-
sures that could be best carried out on the local or State
level. In the San Francisco Bay Area, such measures as those
being investigated and implemented through the Metropolitan
Transportation Commission (MTC) are of particular interest to
the Environmental Protection Agency, particularly with regard
to their effect on air quality. At the public hearings on the
EPA transportation control plan proposal, testimony given by
an official representing the MTC shows that 10 to 15 percent
VMT reduction will occur as a result of the MTC plan, and that
an additional 5 percent VMT reduction could occur with an en-
larged MTC program. In regulations promulgated by this rule-
making, the EPA is requiring the State to submit an analysis
of the status of each element of the MTC Regional Transporta-
tion Plan and to substantiate its effect on air quality. It
is the Administrator's understanding that such an analysis is
currently underway by the California Department of Transporta-
tion (Cal/Trans) and will be submitted to EPA.
-------
-34-
Locally implemented strategies that the MTC and State
are investigating include in part: bridge toll fare in-
creases, exclusive bus and carpool lanes, ramp metering,
increased bicycle facilities, improved local transit,
anticipated Bay Area Rapid Transit expansion, increased
ferry service to the North Bay, fringe parking facilities,
and reduced transit fares.
Table A shows the estimated range of impact of the MTC
plan and the EPA promulgated VMT control measures.
-------
-35-
TABLE A
SF BAY AREA - VMT REDUCTION MEASURES
Estimated percent
Reduction in Daily
Vehicle Miles
Traveled - DVMT
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
i
(0
c
/H
>v
^ c
i 1 (li *O
(0 3
C C -P
o o w
H -H
Promul- u «j g
gated Proposed M i
Control Measures by EPA by Local £J §, ^
s w
Exclusive Bus/Carpool Lanes X X
Bus/Carpool Matching X X
Parking Supply Management X X
Mass Transit Incentives X X
for Employees
\
3
2-4**
1-2**
2.5-3.8
Parking Surcharge X X * 4-10
*
VMT/Air Quality Improvement X
Monitoring Program °
Gasoline Limitations X -u ***
c
Mass Transit Improvements/ X *"
local and regional bus
and BART speed, frequency,
service efficiency, comfort
Bicycle Networks and Facilities X
Expansion of BART X
Close CBD entrance X
\
< .7-(3**)
/
See next page
-------
-36-
TABLE A (Continued)
**
Estimated percent
Reduction in Daily
Vehicle Miles
Traveled - DVMT
H
fO 1
C (0 C
O -P nj
H M H
(^ O PU 12
Promul-
D O< On
gated Proposed c o IT?
Control Measures by EPA by Local ^
12.
13.
14.
Ferry Service
Fringe Parking
c
M
to North Bay X
J (o -H sc 3
-i M -P CM -P
3 B W W
Facilities X on
Pedestrian Access X
LI
en
i-t
15.
16.
*
Reduced Transit Fares X .(,
i
i
Bridge Toll Fare Increases X ]
3 .2-4.5
^
U .5-6.8
Cn
17.
Staggered Work
Hours and X <
u
i
and Work Weeks w
18.
19.
Ramp Metering
Tax Penalties
* X
High-Powered X
Vehicles, Gas Tax,
\
3
8-12
/
Should result in VMT reduction where few or no parallel streets
exist to absorb diverted traffic. Restricted highway access must
be accompanied with an increase in transit capacity.
Estimate by EPA, Region IX
*** Amount necessary to meet air quality standards in 1977.
**** An estimate given in MTC testimony at an EPA hearing;
See Table 4-12 also.
-------
737-
B. San Joaquin
In evaluating the impact of VMT reduction measures for the
San Joaquin AQCR, TRW commented that within the three areas of
San Joaquin County with Stockton its major city, Fresno County
(cities of Fresno, Parlier and Five Points) and Kern County
(Bakersfield Kern's main city) there was 65% of the AQCR's
population. Unable to conduct a study for Modesto and Visalia,
TRW believed that the benefits derived from the strategies
developed for the three counties would benefit the Basin-wide
area including Modesto and Visalia.
Public Transit -
The low density land use patterns in the Basin make public
transit at present inefficient. The Origin-Destination survey
showed that less than 1.6 percent of all person trips in the
Stockton urbanized area were made by public transit. No
attempt was made to develop a mode split model or some other
forecasting means for transit trips in this area. However, a
future study will explore the potential for expanded transit
service. If ridership could be tripled, as it has in areas
similar to Stockton, and transit could be made more convenient,
it is estimated that reduction in VMT from expanded transit
could be .8 percent in Stockton and .3 for San Joaquin County.
-------
-38-
Similar calculations were done for the other two counties,
Kern and Fresno, reflecting the estimates shown in Table E.
Carpooling -
TRW assessed that while cities of Stockton, Fresno, and
Bakersfield had low potential for carpooling, their central
areas, however, had greater potential with employment centers
providing the common pool to draw on with incentive systems,
matching service and public information. Even so, the esti-
mates, as shown in Table B are quite low.
Parking Control -
No parking surveys were conducted in Fresno or Bakers-
field, but a 1969 survey provided information in Stockton.
Under an assumption that there would be no further increases
in the supply of parking spaces, 1975 and 1980 deficiencies
were projected. A control measure of limiting construction
of additional long-term parking spaces along with increased
long-term parking rates should help to somewhat decrease
exclusive use of private automobiles for work trips to the
Basin's CBDs. The cities are considering a proposal to make
short-term parking free in the downtown area to reduce the
amoung of supervision necessary. Increased enforcement would
be required to prevent meter feeding in short-term spaces by
-------
-39-
all-day parkers. Estimates of the effect of the parking con-
trols with additional car pooling of work trips and increased
use of transit would only reduce VMT up to .4 percent in the
entire Basin. (See Table B)
Future investigations -
EPA has promulgated regulations calling for studies in
the Stockton, Fresno, Bakersfield, and Modesto areas. The
California Department of Transportation is conducting a study
to determine the potential for public transit in the Fresno
area. Included will be the examination of the potential to
establish preferential bus/carpool treatment on a north-south
corridor between the Fresno CBD and northern residential
suburbs.
-------
-40-
TABLE B
SAN JOAQUIN VALLEY - VMT REDUCTION MEASURES
Estimated percent
Reduction in Daily
Vehicle
Miles
Traveled - DVMT
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Promul-
gated
Control Measures by EPA
Priority Treatment for Mass X
Transit
Bus/Carpool Matching X
Parking Supply Management* X
Mass Transit Incentives* X
for Employees
VMT/Air Quality Improvement X
Monitoring Program
Gasoline Limitations X
Mass Transit Improvements
Bicycle Networks and Facilities
Traffic Flow Improvements
Fringe Parking Park and Ride
EPA-
TRW
O
U
C
H
3 0
C1 U
ra
o o
*} C
Proposed w
by Local ro £J
CO tn
(0-5)
Local
Measures
.
o
u
W
rt
O rH
u w
H
c a
M n)
Q) -P
« CO
(.1-3) (0-2)
.1 .2 .
.1 .6 .
***
X
X
X
1 .1
2 .1
0-2**
0-1**
0-1**
0-2**
* Includes matching
** Estimate by EPA, Region IX
*** Amount necessary to meet air quality standards in 1977.
-------
-41-
Locally Implemented Strategies
The public hearing testimony, written comments, and the
San Joaquin Valley Task Force indicated a wide variety of
measures that could best be carried out on a local or state
level. The EPA plan notes these measures as being applicable
toward attainment of the oxidant standard and will encourage
their implementation, although the reductions are not credited
as part of the EPA plan. Furthermore, EPA will evaluate the
progress and success in the implementation of these and similar
measures and, if found necessary, promulgate additional measures
to supplement them. Many of these measures will likely be con-
tained in the anticipated revisions to the State Implementation
Plan as a result of the Gal/Trans task force findings.
In the meeting between EPA and the transportation task
force officials, it was noted that the cities of Fresno and
i
Stockton were in the process of upgrading and expanding their
mass transit bus system. The City of Bakersfield has recently
taken over operation of a mass transit bus system, and has
experienced encouraging initial success in its effort to in-
crease ridership. The California Department of Transportation
vehicle miles traveled reduction plan to be submitted to
EPA in October, is expected to document the progress made
by the present mass transit programs in the Valley, and
-------
-42-
allow EPA to monitor and evaluate the need for and progress
of future public mass transit program implementation, in-
cluding the present and the estimated future financial com-
mitment to mass transit.
Local agencies and private citizens gave written and
verbal testimony on the desirability of improved bicycle
networks. Due to the level terrain, many trips in the San
Joaquin Valley could be taken by bicycle. It is anticipated
that improvements of this type will be made in the next
several years.
C. Sacramento Valley
VMT reductions measures were evaluated by TRW using data
from the State Division of Highways. Most of the region's
pollution problem centers around the urban areas and the CBD
of Sacramento.
Public transit improvements -
A 1 percent VMT reduction is estimated by TRW for expan-
sion of efficient mass transit in the City of Sacramento.
Estimates are based on an average occupancy of 1.5 persons per
vehicle and an average trip length of 5.5 miles.
-------
-43-
Increased car pooling -
In the Sacramento study area, the most likely
possibilities for car pooling are the central
city area (RAD 190) and the fringe areas to the
south and east of the CBD (SubRADs 1802, 1702,
1701 and 1601). This area contained 42.7 per-
cent of total employment in 1968, and it at-
tracted 72,980 person home-to-work trips. An
increase in employment of 28 percent has been
projected for this area by 1980. Assuming sim-
ilar increases in work trips, the total vehicle
miles of travel generated by these home based
work trips would be:
(72,980 x 2)/1.20 x 1.28 x 7.60 = 591,625 VMT
where:
auto occupancy = 1.20 persons per car
projected growth = 28 percent
projected average = 7.60 miles
trip length (1980)
-------
-44-
By full cooperation and support of all major
employers in the area, it should be feasible to
increase the work trip auto occupancy from 1.20
persons per car to 1.40. This will result in a
reduction of 82,800 daily VMT, or 0.64 percent of
projected travel in Sacramento urban area and
0.43 percent of the total study area travel.
c. Parking Control Measures - Parking control measures
can be used to either discourage the use of private
vehicles or to increase the efficiency of their
usage. This can be accomplished by either limiting
the number of parking spaces or by controlling
their use by pricing mechanisms. The measure is
most effective in the central business district.
Increased long-term parking rates combined with car pooling
should help to somewhat decrease exclusive use of private auto-
mobiles for work trips to the central business district.
Estimates with parking control are that car pooling will
increase, raising work trip auto occupancy to 1.45 persons
per car. This will result in a .1 percent total VMT reduction
in the area. An additional .1 percent VMT reduction will
result from work trips being made by public transit.
-------
-45-
Table C illustrates the measures proposed by EPA and
local task groups for the area. The measures reflect testi-
mony received at the August 10, 1973 public hearing in
Sacramento and written comments received on the plan.
Additionally, various State and/or locally implemented
controls are noted as part of the total strategy to achieve
the standard by 1977. EPA will carefully assess the reduc-
tion in VMT and improvements in air quality obtained from
the various strategies and control regulations, and recom-
mend additional strategies as needed between now and the
1977 legal attainment date.
The proposal to achieve a bus/carpool computer matching
and promotion system by March of 1974 has been modified to
require such a system be initially established at McClellan
Air Force Base in Sacramento. Upon evaluation, the system
wii-ii"be''expahded'-:ty6 covet 'majdV prfva%e emplo'yets^ and1- '. ... -;>,<
government agencies."'i:2l'i'--'i - 'J"-j r - - :i.'
A "Mass Transit Priority and Planning" Regulation for a
four-cbunty portion1 of the 'Sad'ramerito" Valley ;Reg£o'n> 'has: ±>een
added to the'control plan/1- 'The Sacr'amentO' Regional. Area .
Planriirig Commission (SRAPG) -July-1972 report "Transit Plan ..
and Program", states that the''Sacramento City"" J" Street
bus and traffic situation justified -bus Apriority treatment.--
-------
-46-
TABLE C
y'
SACRAMENTO VALLEY - VMT REDUCTION MEASURES
Estimated percent
Reduction in Daily
Vehicle Miles
Traveled - DVMT
fC
EH
1.
2.
3.
4.
Control Measures
Exclusive Bus/Carpool Lanes
Bus/Carpool Matching
Parking Supply Management
Mass Transit Incentives
Promul- H
gated Proposed o ^
by EPA by Local 5 £
X X 2-8
X
X
X
bH >\
<"§
p . [ 1
w w
3-10
.4-4
.2-1
.6
for Employees
5. VMT/Air Quality Improvement
Monitoring Program
6. Gasoline Limitations
7. Mass Transit Improvements
8. Bicycle Networks and Facilities
9. Fringe Parking Park and Ride
10. Traffic Flow Improvements*
X
X
X
X
***
1-2
0-2**
1-2**
0-1**
* Although traffic flow improvements may decrease the amount of
emissions by easing congestion, and may provide some time savings
for bus and carpool by selective use of ramp metering, the in-
creased capacity may awaken the latent demand for additional VMT,
and thus be counterproductive to the objective of reducing VMT.
** Estimate by EPA, Region IX
*** Amount necessary to meet air quality standards in 1977.
-------
-47-
In addition, SRAPC reports and discussions at the meetings
between EPA and the Task Force indicate that there is a near
term potential for mass transit priority treatment (e.g.
U.S. 99 through Southern Sacramento). As a result, EPA has
promulgated final regulations calling for specific action
in the "J" Street situation, and a general action outline
plan for mass transit priority treatment on streets and
freeways.
Locally Implemented Strategies
The public hearing testimony, written comments, and the
Sacramento Task Force indicated a wide variety of measures
that could best be carried out on the local or State level.
The EPA plan notes these measures as being applicable toward
attainment of the oxidant standard and will encourage their
implementation, although the reductions are not credited as
part of the EPA plan. Furthermore, EPA will evaluate the
progress and success in the implementation of these and sim-
ilar measures and if found necessary, promulgate additional
measures to supplement them. Many of these measures will
likely be contained in the anticipated revisions to the State
Implementation Plan as a result of the Cal/Trans task force
findings.
-------
-48-
Major improvements to the mass transit system in Sacra-
mento have occurred during the past year and are expected to
be expanded as additional funds become available for these
purposes.
The meeting between EPA and the transportation task force
officials confirmed that the Sacramento Regional .Transit Dis-
trict had initiated and is in the process of implementing an
innovative and aggressive transit program. It is expected
that the State of California Department of Transportation
vehicle miles traveled reduction plan to be submitted to EPA
in October will further define this program, and will allow
EPA to monitor and evaluate the progress of future public
mass transit program implementation, including the present
and estimated future financial commitment to mass transit.
The Sacramento Regional Transit District initiated a 25C
flat fare recently, and bus ridership has dramatically increased
since that time. Express bus routes to outlying areas of the
metropolitan area were established in the past year and sev-
eral more will be in operation by the end of 1977. Fringe
parking lots have been established and are being expanded
along these routes as well as at major regional shopping
centers. There are plans to place bicycle protection faci-
-------
-49-
lities at these fringe parking lots to allow nearby residents
to bicycle to the bus stops. Bikeways are very extensive in
the Sacramento area due to the flat terrain. Ridership incen-
tive programs are being planned through an extensive public
information program. Dial-A-Bus and subscription bus service
will also be carefully examined. An area being seriously
examined for applicability of special bus service is the route
between the University of California campus at Davis and cer-
tain residential areas in the city of Sacramento, 18 miles
away.
Various traffic flow improvement programs are also under-
way or proposed for the next 4 years in the Sacramento Region.
These include a synchronized traffic signal system for major
arterials, easing of traffic bottlenecks, and additional ramp
metering.
A final measure that will likely be the subject of
further consideration and experimentation in the next few
years will be the variation of the work week to four 10-hour
days, and other combinations to determine the effects of the
variations on air quality and VMT. Governmental agencies
will likely be early candidates for this program.
------- |