EPA-600/5-76-003
                                             May 1976
CASE STUDIES OF TRANSIT ENERGY AND AIR POLLUTION IMPACTS
                           by

                     James P. Curry
                De Leuw, Gather & Company
                 Washington, D.C.  20036
                 Contract No.  68-01-2475
                     Project Officer

                    Steven E.  Plotkin
         Office of Energy, Minerals and Industry
                 Washington, D.C.   20460
          U.S.  ENVIRONMENTAL  PROTECTION AGENCY
           OFFICE OF RESEARCH AND DEVELOPMENT
         OFFICE OF ENERGY,  MINERALS  AND INDUSTRY
                 WASHINGTON,  D.C.   20460

-------
                             DISCLAIMER
This report has been reviewed by the Environmental Protection Agency
and approved for publication.  Approval does not signify that the
contents necessarily reflect the views and policies of the Environ-
mental  Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation of use.
                                 n

-------
                         FOREWORD

     The continuing crisis in the supply of petroleum, the resulting
call for strong measures to reduce fuel consumption in the transpor-
tation sector, and the potential effects such measures might have on
the environment have caused the Office of Research and Development
(ORD) , U.S. Environmental Protection Agency to undertake studies of
the effects that transportation-related energy conservation measures
will have on petroleum demand and air quality.

     This study examines the changes in fuel consumption and emission
of air pollutants caused by the introduction of new transit services.
A complementary study of the effects of policy measures aimed directly
at the automobile (gasoline taxes, rationing, new car mileage regula-
tions and excise taxes) is scheduled to be published at about the same
time as this one.

     The principal investigator of this study was Mr. James P. Curry
of De Leuw, Gather & Company, 1201 Connecticut Avenue, N.W.,
Washington, D.C.  Mr. Edgar A.  Gonzalez of De Leuw, Gather was
responsible for model development and analyses as well as case
studies data collection.  Mr. William E. Piske of TRW Environmental
Services assisted with study methodology development and directed
air quality inputs preparation.  Mr. Charles Scardino, also of TRW
Environmental  Sciences, was responsible for air quality inputs and
assisted with  case studies data collection.  Mr.  John L. Grain
of Bigelow-Crain Associates provided data regarding the San
Bernardino busway and Orange County projects, in  addition to
providing key  study review inputs.  Ms. Christine L. Nelson
of De Leuw, Gather prepared all report illustrations.  The EPA
project officer for the study was Mr. Steven E. Plotkin of the
Office of Energy, Minerals and Industry.

     A major conclusion of this report reflects what thoughtful
proponents of mass transit expansion have always  known.. .that expansion
of service without extremely careful attention to latent demand, system
competitiveness with alternate transportation modes, and selection of
appropriate equipment, schedules and operating methods can lead to
a less rather than more efficient transportation  system.  The theoretical
edge in efficiency and environmental attractiveness of mass transit
over the automobile is enormous: the actual edge  is, on the average,
much less but  still significant.  In a particular case, the advantage
may evaporate, as shown in this report.  We conclude by this that,
as much as general support for mass transit is admirable from an
environmental  standpoint, we must judge each case by its merits...
certainly not  a unique idea.
                      --, ..^        -v.

                        Stephen J. Gage
                 Deputy Assistant Administrator
               for Energy, Minerals and Industry
             U.S. Environmental Protection Agency
                                   iii

-------
                             ABSTRACT
     This report summarizes  an analysis  of the  energy  consumption  and
air pollution impacts of eight case studies of  new  or  improved  transit
services.  The case studies  include (a)  areawide  bus service  improvement
programs involving route extensions, increased  frequencies, new lines,
demand responsive service, and fare reductions; (b) new  corridor
exclusive busway service on  the Shirley  Highway and San  Bernardino
Freeway; and (c) new rail  transit service in the  Philadelphia-Lindenwold
corridor.  Probabilistic models were developed  for  each  of these three
service improvement scenarios  to account for key  travel  demand  and
transportation system factors  affecting  energy  consumption and  air
pollution impact levels.  Results showed that low patronage response
to areawide bus improvements as well as  diversion from prior  bus service,
carpools, etc. and extensive auto access (park-and-ride, kiss-and-ride)
to corridor systems reduce expected energy and  air  pollution  gains and"
may, under certain conditions  found in four case  studies, result in
possible energy use increases.  Additionally, it  was found that auto
use for corridor system access may worsen air quality  conditions in
suburban areas in the vicinity of corridor transit  terminal locations.

-------
                            TABLE OF CONTENTS

Chapter                                                             Page
  1.  STUDY SUMMARY AND MAJOR FINDINGS 	  1
      1.1   Introduction	1
      1.2  Selection of Case Studies	4
      1.3  Methodology 	  5
      1.4  Study Findings	12
           1.4.1  Finding No. 1	12
           1.4.2  Finding No. 2	19
           1.4.3  Finding No. 3	25
      1.5  Further Research	29

  2.  STUDY METHODOLOGY	31
      2.1   Introduction	31
           2.1.1  Selection of Case Studies	32
           2.1.2  Model Development	33
      2.2  Key Factors Affecting  Energy Consumption and Air
           Pollution Impacts 	  38
           2.2.1  Auto and Transit Unit Air Pollutant Emissions
                  Characteristics	39
                  2.2.1.1  Diverted Auto Cold Starts 	  40
                  2.2.1.2  Diverted Auto Driving  Cycles	43
                  2.2.1.3  Bus Air Pollutant Emissions 	  43
                  2.2.1.4  Rail Transit Air Pollutant Emissions.  .  .  46
           2.2.2  Auto and Transit Unit Energy Consumption
                  Characteristics	46
                  2.2.2.1  Improved Auto Efficiency	49
                  2.2.2.2  Bus Energy Consumption Characteristics.  .  50
                  2.2.2.3  Rail Transit Energy Consumption  	  53
           2.2.3  Ridership Response to New Transit Services  ....  60

-------
                       TABLE OF CONTENTS (Continued)

Chapter                                                              Page
                  2.2.3.1   Short-Term Changes in Travel  Behavior. .  .   61
                  2.2.3.2  Long-Term Changes in Travel  Behavior ...   63

  3.  EXPANDED REGIONAL BUS SERVICE
      3.1  Introduction	66
           3.1.1   Methodology	69
      3.2  Atlanta	72
           3.2.1   New Ridership	73
           3.2.2  Diversion of Automobile Trips 	   74
           3.2.3  Net Energy Consumption Savings	74
           3.2.4  Net Air Pollutant Emissions Reduction 	   78
      3.3  Washington	78
           3.3.1   Net Energy Consumption Savings	80
           3.3.2  Net Air Pollutant Emissions Reduction 	   85
      3.4  San Diego	89
           3.4.1   New Ridership	91
           3.4.2  Net Energy Consumption Savings	92
           3.4.3  Net Air Pollutant Emissions Reduction 	   96
      3.5  Orange County	104
           3.5.1   New Ridership	105
           3.5.2  Net Energy Consumption Savings	105
           3.5.3  Net Air Pollutant Emissions Reduction 	   105

  4.  NEW CORRIDOR EXPRESS BUS SERVICE	110
      4.1  Introduction	110
           4.1.1   Methodology	Ill
      4.2  Shirley Highway Busway 	  ...   112
           4.2.1  Diversion from Auto Driving  	  	   115
           4.2.2  Busway Access Mode	117
           4.2.3  Busway Influence on Residence Location	117

-------
                      TABLE OF CONTENTS (Continued)

Chapter                                                              Page
           4.2.4  Busway Passenger Load Factors	   120
           4.2.5  Net Energy Consumption Savings 	   120
           4.2.6  Net Air Pollutant Emissions Reduction	122
      4.3  San Bernardino Busway	125
           4.3.1  Diversion from Auto Driving	127
           4.3.2  Busway Access Mode	129
           4.3.3  Use of Auto Left at Home	129
           4.3.4  Busway Influence on Residence  Location  Choice.  .  .   131
           4.3.5  Net Energy Consumption Savings 	   133
           4.3.6  Net Air Pollutant Emission Reduction 	   139

  5.  NEW CORRIDOR RAIL TRANSIT SERVICE	14.1
      5.1  Introduction	143
           5.1.1  Methodology	144
      5.2  Bay Area Rapid Transit	144
           5.2.1  Corridor Mode Choice and Diversion to BART ....   145
           5.2.2  Previous Travel Mode	148
           5.2.3  Mode of BART Access	148
      5.3  Lindenwold Rapid Transit	151
           5.3.1  Corridor Mode Choice and Diversion to Lindenwold
                  Line	153
           5.3.2  Mode of Access to Lindenwold Line	153
           5.3.3  Access and Line Haul Trip Lengths	153
           5.3.4  Net Energy Consumption Savings 	  156
           5.3.5  Net Air Pollutant Emissions Reduction	156
      5.4  Rail Transit Air Pollutant Emissions  	    •  160
  APPENDIX A.  Methodology Description and Input Data	153
  APPENDIX B.  Selected References 	  182
                                   vn

-------
                              LIST OF FIGURES

Number                                                               Page
 1.1   Generalized Busway Model  	   11
 1.2  Fuel  Consumption Reduction,  Washington,  D.C	   14
 1.3  Reduction in Fuel  Consumption,  San  Diego Action  Plan  	   15
 1.4  Reduction in Fuel  Consumption,  Orange County,  California  ...   17
 1.5  Reduction in Fuel  Consumption,  San  Diego Fare  Reduction.  ...   18
 1.6  Reduction in Fuel  Consumption,  Lindenwold High Speed  Line.  .  .   20
 1.7  Lindenwold High Speed Line Before and After Peak Period Energy
      Consumption	  .   21
 1.8  Shirley Busway Before and After Energy Consumption  	   23
 1.9  San Bernardino Busway Before and After Energy  Consumption..  .  .   24
 1.10 San Bernardino Busway, Distribution of CO Changes	   26

 2.1   CO Emission Rates  Versus  Auto Trip  Length	   42
 2.2  Auto Gasoline Consumption (Versus Trip Length) 	   48
 2.3  Bus Fuel  Consumption and  Average Route Speed	   51
 2.4  Bus Fuel  Consumption and  Bus Stop Frequency	   52

 3.1   Trend of Average Fare	   67
 3.2  Trend of Vehicle Miles	   67
 3.3  Trend of New Motor Buses	   67
 3.4  Trend of Revenue Passengers	   68
 3.5  Fuel  Consumption Reduction,  Atlanta, Georgia  	   77
 3.6  CO Emissions Reduction, Atlanta, Georgia 	   79
 3.7  Fuel  Consumption Reduction,  Washington,  D. C	   81
 3.8  Effect of Auto Diversion Rate in Fuel Consumption Reduction,
      Washington, D. C	   82
 3.9  Break-even Values, Washington,  D. C	   83
 3.10 Effect of Auto Diversion Rate in Fuel Consumption Reduction,
      Washington, D. C.  (Improved  auto fuel consumption)  	   84

-------
                         LIST OF FIGURES (Continued)

Number                                                              Page
 3.11  Reduction in CO Emissions.   Washington, D.C	86
 3.12  Effect of Auto Diversion Rate in Reduction of CO Emissions
       Washington, D.C	87
 3.13  Air Quality Index Readings.   Washington, D.C.  July, 1974.  .  .  90
 3.14  Reduction in Fuel Consumption.  San Diego Fare Reduction.  .   .  93
 3.15  Reduction in Fuel Consumption.  San Diego Fare Reduction.  .   .  94
 3.16  Reduction in Fuel Consumption.  San Diego Fare Reduction.  .   .
       (Improved auto fuel  consumption)	95
 3.17  Reduction in Fuel Consumption.  San Diego Action Plan..  ...  97
 3.18  Reduction in Fuel Consumption.  San Diego Action Plan ....  93
 3.19  Reduction in Fuel Consumption.  San Diego Action Plan, (Im-
       proved auto fuel  consumption	99
 3.20  Reduction in CO Emissions.   San Diego  Fare Reduction	100
 3.21  Reduction in CO Emissions.   San Diego  Fare Reduction	101
 3.22  Reduction in CO Emissions.   San Diego  Action Plan	102
 3.23  Reduction in CO Emissions.   San Diego  Action Plan	103
 3.24  Reduction in Fuel Consumption.  Orange County, California,
       (25 percent auto  trip diversion)	106
 3.25  Reduction in Fuel Consumption, Orange  County,  California,
       (50 percent auto  trip diversion)	107
 3.26  Reduction in Fuel Consumption, Orange  County,  California.  .  .  108
 3.27  Reduction in CO Emissions.   Orange County California	109

 4.1    Generalized Busway Model	114
 4.2    Shirley Highway Busway.   Washington 	  116
 4.3    Shirley Busway Daily Fuel  Savings 	  121
 4.4    Shirley Busway Before and After Energy Consumption	123
 4.5    Shirley Busway Daily CO Emissions Reduction 	 124
 4.6    San Bernardino Busway.   Los  Angeles	126
 4.7    San Bernardino Busway Daily  Fuel Savings	134
 4.8    San Bernardino Busway.   Before and After Energy Consumption  . 135
 4.9    Approximate Volume - Speed  Relationship 	 137
 4.10  Auto Fuel Consumption	138
 4.11  San Bernardino Busway Daily  CO Emissions Reduction	140
 4.12  San Bernardino Busway.   Distribution of CO Changes	142

                                   ix

-------
                         LIST OF FIGURES (Continued)

Number                                                              Page
 5.1   Reduction in Fuel  Consumption.   Lindenwold High Speed Line.  .   157
 5.2   Reduction in Fuel  Consumption,  Lindenwold 	   158
 5.3   Lindenwold High Speed Line.   Before and After  Peak Period
       Energy Consumption	159
 5.4   Reduction in CO Emissions.  Lindenwold High Speed Line.  ...   161
 A.I   Typical Beta Distribution	165
 A.2   CO Emissions Reduction.  Atlanta, Ga	   174

-------
                               LIST OF TABLES

Number                                                              Page
 1.1   Typical Energy Consumption and Air Pollutant Emission Rates .   6
 1.2   Selected Data on Corridor Transit Case Studies	7
 1.3   Factors Affecting Energy Consumption and Air Pollutant
       Emission Changes Due to Expanded Regional Bus Service ....   8
 1.4   Factors Affecting Energy Consumption and Air Pollutant
       Emission Changes for Mew Corridor Transit Service 	   9
 2.1   Factors Affecting Energy Consumption and Air Pollutant
       Emission Changes Due to Expanded Regional Bus Service ....   35
 2.2   Factors Affecting Energy Consumption and Air Pollutant
       Emission Changes for New Express Busway Service 	   36
 2.3   Factors Affecting Energy Consumption and Air Pollutant
       Emission Changes for New Rail Transit Service 	   37
 2.4   Comparison of Driving Cycle and Federal test Procedure.   CO
       Emission Factors	44
 2.5   Average 1975 CO Emission Factors with Driving Cycle
       Correction	45
 2.6   Bus Air Pollutant Emission Factors	47
 2.7   Comparison of Estimated Energy Use Components 	   55
 2.8   Total Automobile Energy Use, 1971	59
 2.9   Driver Trips in Multi-Car Households	65
 3.1   Selected Operating Data for Case Studies	70
 3.2   Factors Affecting Energy Consumption and Air Pollutant and
       Emission Changes Due to Expanded Regional Bus
       Service	71
 3.3   Previous Travel Mode for Atlanta New Bus Riders 	   75
 3.4   Weekday Distribution of Previous Automobile Drivers  	   76
 3.5   Summary of 1975 Transportation Control Strategy for Washing-
       ton Metropolitan Region 	   88

 4.1   Factors Affecting Energy Consumption and Air Pollutant
       Emission Changes for New Express Busway Service 	  113
                                   XI

-------
                         LIST OF TABLES (Continued)

Number                                                              Page
 4.2   Previous Travel  Mode of Shirley Busway Riders,  AM Peak Period 118
 4.3   Mode of Access for Shirley Busway Riders,  AM  Peak Period.  .  . 119
 4.4   Previous Travel  Mode of San Bernardino Busway Users  	 128
 4.5   Mode of Access for San Bernardino Busway Users	130
 4.6   Use of Commuting Car Left at Home	132

 5.1   Factors Affecting Energy Consumption  and Air  Pollutant
       Emission Changes for New Rail  Transit Service 	 145
 5.2   San Francisco East Bay Mode Split AM  Peak  Period	147
 5.3   San Francisco East Bay Mode Split All  Day, Both Directions.  . 149
 5.4   Previous Travel  Mode of BART Passengers, 1973	150
 5.5   Mode of BART Access, 1973	152
 5.6   Previous Travel  Mode of Lindenwold Passengers,  1970  	 154
 5.7   Estimated Mode of Access for Lindenwold Line  Users,  1970.  .  . 155
 5.8   Estimated 1975 Contribution of Rail Transit Operations to
       Annual Regional  Air Pollutant Emissions 	 162

 A.I   Values Used in AREAWIDE Computer Model.  Atlanta, Ga.  .... 169
 A.2   Values Used in AREAWIDE Computer Model.  Washington,  D.C.  .  . 170
 A.3   Values Used in AREAWIDE Computer Model.  San  Diego Fare
       Reduction	•  • 171
 A.4   Values Used in AREAWIDE Computer Model.  San  Diego Action
       Plan	  . 172
 A.5   Values Used in AREAWIDE Computer Model.  Orange County.  ,.  .  . 173
 A.6   Values Used in AREAWIDE Computer Model.  San  Diego Fare
       Reduction Trip Length Parametric Analysis	 175
 A.7   Values Used in AREAWIDE Computer Model.  San  Diego Action
       Plan.   Trip Length Parametic Analysis	 176
 A.8   Values Used in BUSWAY Computer Model.   Shirley  Busway 	 179
 A.9   Values Used in BUSWAY Computer Model.   San Bernardino Busway. 180
 A.10  Values Used in CORAIL Computer Model.   Lindenwold Rapid Rail. 181
                                   XII

-------
                              CHAPTER  1
                STUDY SUMMARY AND MAJOR FINDINGS

1.1  INTRODUCTION
     Current policy considerations regarding energy conservation  and  pro-
duction independence, as well as requirements  of the 1970 Clean Air Act
regarding attainment of ambient air quality standards have both pointed  to
the development of improved public transportation services as  a key program
element.  Such expansions of transit service hold the promise  of  reductions
in both air pollutant emissions and transportation energy consumption.
Comparison of typical auto and transit air pollution and  energy consumption
characteristics indicates the relative attractiveness of  transit  ,  On a
seat-miles basis, transit is two to four times more energy efficient  than
the automobile and many times less polluting.
     The actual effectiveness of new public transportation services in pro-
viding major pollution reduction and energy savings without the introduction
of major auto-use restraint or crisis conditions is for the most  part not
                2
well  established .  Much of the reported research has dealt with  variations
in the energy use and air pollutant emissions  characteristics  of  alternative
transportation systems under assumed or  empirically 'average1  conditions.
 For example--Healy,  T.  J.   Energy Use  of Public  Transit  Systems.  Prepared
 for the California Department of Transportation, August  1974.  Lansing, N. F.
 and H. R.  Ross.   Energy Consumption  by Transit Mode.   Prepared for the
 Southern California  Association of Governments,  March  1974.  The MITRE
 Corporation.  Transportation  Energy and Environmental Issues.  February, 1972.
2
 TRW/De Leuw,  Gather  & Company.   Travel  Impacts of  Fuel Shortage and Price
 Increase Conditions.  Prepared  for the U.S.  Environmental Protection Agency,
 December 1974.

-------
It is the general finding of this study that the use of average conditions
without full attention to the actual  characteristics of new transit trips
has distorted analysis of transit energy and air pollution impacts.  Under
many conditions found in case studies carried out in this project, new or
improved transit services may have distinctly negative energy and air
pollution impacts.  More specifically, the following factors may be signifi-
cant determinants of transit energy consumption and air pollutant emissions
impacts:
     • new transit ridership diverted from carpooling or other transit
       service, from trips made in small autos, and being trips not
       previously taken;
     • the number and length of trips made by auto to access new
       corridor transit service;
     • the length of former auto trips made on new transit services-
       are long or short auto trips being diverted to transit;
     • 'cold start1 air pollutant emissions effects associated with
       corridor transit access trips made by auto;
     • the effects of improved auto traffic flow due to trips
       diverted to new transit service;
     • effects of current trend towards smaller autos with improved
       energy efficiency;
     • use of a former commuting auto for other trips by other
       household members; and
     • indirect energy consumption for transit or highway system
       construction and operation.

-------
Despite transit's apparent attractiveness for energy conservation and air
quality, study results indicate that transit's limited ability to capture
new riders, especially former automobile users, and the continued use of
autos for corridor transit system access following its development without
effective collection-distribution service can substantially reduce (and in
some cases eliminate) expected energy use or overall air pollution savings.
Futhermore, current trends towards lighter automobiles with greater fuel
efficiency serve to reduce transit's fuel saving potential  since diverted
auto trip energy is lowered.   A comprehensive approach to transit develop-
ment to maximize its energy and air pollution reduction effectiveness is
required if full advantages are to be realized.
     This report describes results of a research study undertaken for the
Environmental  Protection Agency in which the net energy consumption and air
pollutant emissions  changes  associated with eight case studies of new or
improved transit services have been examined.  The case studies are repre-
sentative of three new transit service scenarios:
     .  Bus service improvements on a region-wide basis consisting of
       route extensions, increased operating frequencies, new lines,
       reduced fare programs, and demand-responsive service;
     .  Corridor express bus service via exclusive bus lanes; and
     .  Corridor commuter rail transit service utilizing modern
       rail transit equipment and performance standards.
'  Only limited analysis was carried out for pollutants other than carbon
  monoxide and hydrocarbons.

-------
The case studies involving recent examples of the introduction of new or
improved service for each scenario were chosen to provide a data base re-
garding changes in trip-making behavior following transit service changes.
     It should be cautioned that case study results have been developed
using available data for key factors which, in many instances, has been
incomplete or limited.  Parametric analyses and techniques for recognizing
the degree of uncertainty associated with study variables were invoked to
accommodate incomplete data.  Furthermore, study results do not have long-
range applicability particularly with regard to potential transit impacts
on land use and urban development patterns, which may have direct and sub-
stantial impacts on regional energy use and air quality.
1.2  SELECTION OF CASE STUDIES
     Eight case studies were selected to provide inputs for study analyses.
They were chosen on the basis of data availability and to provide as wide a
range as possible of new or improved transit service examples.  Case studies
span a spectrum of service levels, characterized by modal shares from a few
percent of total regional or corridor travel to nearly one-half of all
trips during peak periods.
     Five case studies involved expanded regional bus service:
                                Fare
                              Reduction
Atlanta (1972)
Washington (1974-75)
  Improved
Headways and
    Route
  Extensions
      X
      X
 Demand
Actuated
 Service

-------
Fare
Reduction
X
Improved
Headways and
Route
Extensions
X
X
Demand
Actuated
Service

San Diego (1973)
San Diego (1975-)
Orange County,
  California (1974-75)                                           X
Selected data describing before and after conditions for each of the five
cases are summarized in Table 1.1.
     Two case studies involved new corridor busway—one operating on the
Shirley Highway in suburban Washington, D. C. and the second via the San
Bernardino Freeway in the Los Angeles area.  Finally, the Lindenwold high-
speed commuter rail linewas reviewed as a corridor rail transit case study.
Preliminary Bay Area Rapid Transit (BART) data was also compiled for compari-
son but no energy or air pollution analysis was carried out.  Selected data
for each of the three corridor transit case studies is presented in Table 1.2.
1.3  METHODOLOGY
     Each of the case studies was analyzed using simple models which incor-
porated all  factors affecting energy consumption and air pollution impacts
for which reasonable data could be found and applied.  Table 1.3 summarizes
relevant factors for expanded regional bus service, and the availability of
data items for each of the case studies.  Data for most key factors was not
generally available, and was inferred from what was available or from
secondary sources for analysis purposes.  In Table 1.4, relevant factors
for new corridor transit service studies are listed with available data

-------
              CO
       
       co  S-  C

       CQ U_  O
                                        If)
                                                     LO
                                                     CM
             LO
             CM
                                                                  O
                                                                  LO
                                                                  to
                                                                  CM
           Q. CO
          ••- T3
       >>.«=  c.
       r—  CO  (O
       •r-  S-  CO
       03  0)  3
       Q TD  O

          S •»->
                                        CO
                                        CM
                                              in
                                                     CM
                                                     CO
                                                                  CM
       CO
       CO
           CO  CO

       ^51
       •r-  S  tO
       
                           CO S_  C
                           »•-  =
       r— -r-  (O

       (O     3
       O  CO  O
           3 ^:
          CQ 4-»
                                        LO
                                        CO
ro
co
                                                     LD
CM
CO
                                                     CM
                                                            LO
                                                            ro
                           a»  a*—
                           •»->  o  to
                           «J -r-  C
                           E 4J  O
                           •^  (^ "r-
                           X r— r—
                           O  3 •—
                           L.  Q.T-
                           Q. O  E
                           CS.Q. —•
                           O
                            •

                           CO
                                              LO
                                 to

                                  (U
                                  co
                                  to
                                 o
                          CO

                    •*•"•*   CJ1
                    CM    i—

                    cr>
                    i—     C
                    •—    o
                                        
-------
                                    CU
                                   -O
                                    c
                                         
                                                                                                CM
CQ
       cc:
       o;
       o
       t—
           cu
                           c     ex
                           CU     >—
                                                        ai
                                                               O
                                                              O
      .c     cu
       in    -4->
       S-     S-
       CU     CU
      -o     >
                    in
                    in
                    cu
                    u
                    o
                                                                                   cu
                                                                                   4->
                                                                                   in
^

o

 to
 cu
Q.

 o>
 c

 c

 o

-------






n
•
Q)
3
f







2
Q
h
a
§
COIMSU
IS
DC
UJ
UJ

i
S
U.
u.
<

(0
DC
0
\
UJ
Q

in
lu
n
1 CHANI
•MISSION
UJ
h
I
J

0
a
oc
5
a
<


UJ
u
>
a
LU
(0
§
OIMAL Bl
(3
UJ
DC
a
UJ
Q
z
<
2
X
UJ
0
h

•
•
•
•











LU
CJ
>
QL
LU


U_
O
CO
1
CO
3:
UJ
z



•


















CHANGE
LU
rr

u_
sna jo
13A31

•
•
•
•
















n
RIDERSHl
T


•
•
•
•
•

















FACTOR
Q
O















CI3
Z
E INCLUD
iRPPOLS
Q 7^


i £
£ "-
IPUS TRA
VERSION
fT, °
ct
a.
sna
















LU
f-
cc
U 1



(/i
o
1—
o
cr
LU
03

z-
•















t-
•z.
LU
_J



1-
O
Q
LU
H
(T
LU

"











CO
Q_
^-i


O
h-
Q
UJ
h-



•-1
O
Q
U_
O
s:






•


CO
Q-
C£
}~
O

gf

Q
F DIVERT
0


o
h-
O
_J
O

III
CSL
•








•-•

s:
r>
to
7-
C~)
0
GASOLINE

t—

^
0
h-
Q
LU
rr
LU

Q
•














00
NOISSIW3



^
o
a
LU
h-
LU

"
•















Z.
o
h-
a.
1D


O
t )
_J
LU
ID
U_
h-
CO

CD
•



















0
to
CO
LU
1—
fO

OQ
NPISd3Aia
didi oinv













7
n

CJ
•n
o
LUI
a:
o



s:
LLI
O
O
LTi
CT>


2!
















o
^
Q
LU
Qi
CO
O


CO
s:
o
X
un
cn


2:
















CO
CD
2!


to

LU
U,
LH
cn
r-H
H-
III
•z.










o
t-
ID
^

Q

O
o:
Q_
s:
x
i—
•—.


CO
CD
Z.
CO
-J
LU
U_
H.
LU
Z
indino

-------
                   Table 1.4
FACTORS AFFECTING ENERGY CONSUMPTION
 AND AIR  POLLUTANT EMISSION CHANGES
   FOR NEW  CORRIDOR TRANSIT SERVICE
-
a: cc ££
CCIjJ 0
O co «
Q
0
to
ct
UJ
>
Q
O
1-
<
EC to
£ l/)a
g«s
U- ct



lil
Cl
CO
CO
o
o
<£


s:
a: to
1- 0
I <
§5
_j
h-
r3
Q-
O
o

NEW RIDERSH1P
PEAK PERIOD LOAD FACTOR
OFF-PEAK LOAD FACTOR
PREVIOUS TRAVEL MODE
INCLUDING DIVERSION FROM CARPOOLS
LEilGTH OF LIME
CORRIDOR TRIBUTARY AREA
STATION SPACING
RAIL TRANSIT UNIT ENERGY CONSUMPTION
RAIL TRANSIT UNIT EMISSIONS
NUMBER OF AUTOS DIVERTED
DIVERTED AUTO TRIP LENGTH
TIME OF DAY OF DIVERTED AUTOS
REGIONAL LOCATION OF DIVERTED AUTO TRIPS
DIVERTED AUTO UNIT GASOLINE CONSUMPTION
DIVERTED AUTO UNIT EMISSIONS
NUMBER OF TRIPS DIVERTED FROM
FORMER BUS SERVICE
FORMER BUS SERVICE LOAD FACTOR
BUS UNIT FUEL CONSUMPTION
BUS UNIT EMISSIONS
NUMBER OF PARK-AND- RIDE PASSENGERS
NUMBER OF KISS~AND- RIDE PASSENGERS
NUMBER OF FEEDER BUS PASSENGERS
UNIT FUEL CONSUMPTION
FOR AUTO ACCESS TRIPS
UNIT EMISSIONS FOR AUTO ACCESS TRIPS
FEEDER BUS UNIT FUEL CONSUMPTION
FEEDER BUS UNIT EMISSIONS
ACCESS TRIP LENGTH
RESIDENCE LOCATION CHANGE
REDUCED AUTO OWNERSHIP
REDUCED TRAFFIC CONGESTION
OTHER USE OF AUTO
NET 1975 CO EMISSIONS REDUCTION
NET 1975 HC EMISSIONS REDUCTION
NET 1975 FUEL SAVINGS
NET FUEL SAVINGS WITH IMPROVED AUTOS
•
•
•
•
•
•



•

___•_,
•


•
•


•
•
•





•







•


•
•
•



•

•
•


•



•
•
•





•


•







•





•

•



•
•


•
•
•







•





•


•
•
•



•
•

•


•



•
•
•




•


•












•





•
•


•
•



•
•
•
•










-------
summaries noted.  Again, data for several  key factors was missing and was
approximated for study analysis.
     A key aspect of the modeling procedures was the capability to handle
important factors as random variables with associated probability functions.
In this manner, the known variations of certain factors due to local  condi-
tions or operating procedures as well as the uncertainty involved in  specify-
ing values for other factors where there is little or conflicting available
information was accommodated, and in turn, directly reflected in estimates
of output planning variables .  For example, unit fuel  consumption rates and
pollutant emission rates were treated as random input variables from  proba-
bility distribution functions defined by lower bound, modal, and upper bound
values which accounted for the range of empirical and other published data.
Similarly, data regarding the former trip characteristics of new transit
service riders, which may have a substantial impact on energy consumption
results, was incompletely reported.  For study analysis purposes, available
data was used to define appropriate probability distribution functions for
application to case studies where specific data items were not reported.
     For the bus and rail transit corridor models, simplified corridor con-
figuration and operating policy assumptions were required.  As shown  in
Figure 1.1, the model assumed a user-defined rectangular corridor configura-
tion with (a) busway passenger demand uniformly distributed within the
specified service area which means that approximately the same number of
buses enter the busway at each entry point, and (b) buses assumed to  circulate
through a portion of the service area for passenger collection before entering
   i, R. Y. and I. F. Kan.  "A Decision Aid to Transportation Planners."
 Paper presented at 41st Annual Meeting of the Operations Research Society
 of America, New Orleans.
                                      10

-------
                    I/I

    BUSWAY MODEL
II
a
x
Ul
a

8
              rvtce Area
     11

-------
the busway.  Average bus load factors were taken from Shirley busway re-
ported experience, and individual route frequencies determined by applying
the average load factor at each busway entry point.  Neighborhood bus routes
were assumed to provide complete one-quarter mile coverage over a rectangular
grid street network in the defined service area.  Similar assumptions were
invoked for the corridor rail transit model.
1.4  STUDY FINDINGS
     Detailed analysis of travel demand and transportation system characteris-
tics of eight new transit service case studies has demonstrated three princi-
pal findings with general application for transit planning and implementation.
     • New ridership response to areawide bus service improvement programs
       including route extensions, improved service frequencies,  and new
       bus lines, and demand responsive service may reduce system load
       factors resulting in net energy use increases;
     • Energy consumption savings for corridor transit service improve-
       ments are significantly reduced due to diversion to new transit
       service from other transit and carpooling,  auto use for corridor
       system access, and corridor transit system energy requirements;
     » Air quality in suburban areas may be worsened by auto trip making
       involving cold starts for corridor transit system access.
1.4.1   FINDING NO.  1
          New ridership response to areawide bus service improvement
          programs including route extensions, improved service fre-
          quencies, new bus lines, and demand-actuated service may
          reduce average system load factors resulting in net energy
          use increases.
                                     12

-------
     For case studies involving increased bus-miles of service in San Diego,
Washington, and Orange County, greater energy consumption was estimated to
result from the introduction of expanded service due to the low patronage
response and the low proportion of new transit users diverted from prior
auto trips.
     In developing the energy consumption estimates, incomplete data regarding
diverted auto trip characteristics was available as already noted.  Data
regarding prior mode of travel collected following the Atlanta fare reduc-
tion program was used as input for all areawide bus improvement case studies.
In Atlanta, approximately 42 percent of new bus riders were former auto
drivers, i.e., less than one-half of the new bus trips represented elimina-
tion of an auto trip from Atlanta's streets.  Furthermore, Atlanta survey
data indicated that approximately 21 percent of new bus trips were trips not
previously made.  New transit trips not previously made do not contribute
to any reductions in energy use or air pollution levels although they reflect
increased mobility for certain population segments.
     The output net energy consumption probability distribution function for
the Washington bus service improvement program is shown in Figure 1.2,
demonstrating approximately 56 percent probability of energy use increase
conditions.  This was one of the three case studies which generated increased
energy use.  Parametric studies of net energy savings over a range of diverted
auto trip characteristics—percent of new transit users diverted from auto
driving, diverted auto trip length—were conducted to support study findings.
Figure 1.3 illustrates estimated energy consumption savings resulting from
                                      13

-------
                                                                           - 0
01
r  *
z
0
H
u
D
Q
UJ
DC

PTIOIM
IMSUIVI
8

Ul
LL









•
U
•
D
m
0
h , , .

0 0 B> (D
2 r" o d
I
1



























1 1

a
j
u
2r*
2
z Q
2
DUCTION II
iAND QAUL
i
0
DAILY REI
CTHOUS


- 01
1
- I*
                                                =10
                                         14

-------
                                                                                       in
                                                                                       in
                        \
                           \
                             \
                                \
                                     \
                                       \
                                          \
                                            \
                                                - o
                                                - 01
                                                                         \
(0
                                        01
                                         I
ID
i
t AVd
                                    SIMO1~IVQ  QNVSnOHJ.)

                                                 MI
                                        15

-------
the 1975 San Diego Action Plan bus service improvements program.   If new
transit service attracted auto trips having average trip lengths  greater
than approximately eight miles, fuel consumption savings would be generated.
For shorter trip lengths, increased overall fuel usage would be expected.
Similar results are plotted in Figure 1.4 for the introduction of "dial-a-ride"
service in Orange County, showing estimated net energy consumption changes
for both 25 and 50 percent of new transit riders diverted from auto driving
over a range of diverted auto trip lengths.
     In each of these three case studies, low ridership response  to new
services represents a key reason for negative energy consumption  results.
For example, a sixteen percent increase in bus miles operated in  the Washington,
D. C.  area generated only a four percent ridership increase.  Assuming that
the new transit riders are traveling the same distance on the average as old
riders; the system average load factor has decreased.  In fact, unless new
riders are traveling four times as far or longer, the system load factor
will be reduced following the service improvements.
     Two case studies involved bus service improvement programs including
fare reductions.  In both instances, ridership response resulted  in increased
system load factors and significant fuel savings.  Figure 1.5 demonstrates
the range of expected fuel savings for the San Diego fare reduction program,
showing actual savings for diverted auto trip lengths as short as approxi-
mately two miles.
                                       16

-------
                                                                  0
                                                                  r
                                                                     0)
                                                                     «
                                                                     i
                                                                     w
                                                                     I
                                                                     0
                                                                     111
                                                                     a
                                                                  to
                                                                     D
                                                                     5
                                                                     s
                                                                     a
                                                                     ui
                                                                  q  5
                                                                 q
                                                                 ai
o
to
                               tu
                           CAVO H3ed  SIMOJ.I
SNOISSIIAI3  O3  Nl  SMOIJLOna3»d  IMV31AI
                  17

-------
            \
              \
                   \
Q
a
it
u
j  u
u  a
3  111
u.  g

Z*
2  a
o  a
iu
a  to
                                           (U
                            (Aep jad*uo||«B puoenoL|j.J

                                land NI
                           18

-------
1.4.2  FINDING NO. 2
          Energy consumption savings for corridor transit service
          improvements are significantly reduced due to diversion
          to new transit service from other transit and carpooling,
          auto use for corridor system access, and corridor transit
          system energy requirements.
     Analysis of the Lindenwold high-speed rail transit line indicated only
small net energy savings with probability of actual increased energy consump-
tion.  The output probability distribution function derived from model analysis
of the Lindenwold line is plotted in Figure 1.6.  Lindenwold rider survey
results revealed that only twenty-eight percent of Lindenwold riders were
diverted from corridor auto driving in comparison with approximately 40
percent shifting from former bus and commuter rail passenger service.  Further-
more, nearly 90 percent of Lindenwold users reach the system by auto, either
by driving and parking at a station or by being dropped off.  Both factors
have contributed to the net energy impact results.  Figure 1.7 summarizes
before and after energy consumption components in the Lindenwold corridor,
reflecting actual corridor trip characteristics.   'Before' energy consumption
includes estimated former auto and transit consumption for corridor trips
now using the Lindenwold line.  'After' energy consumption includes estimated
energy use for system access by auto, for Lindenwold rail transit propulsion
and vehicle accessories, and for limited feeder bus service, resulting in an
overall net energy savings equivalent to 800 gallons of refined gasoline per
peak period.  This amount is substantially lower than would be expected from
examination of individual modal energy efficiencies under average load
factor conditions as have been frequently reported.
                                       19

-------
ID 2
^ g
4
UMPTI
CONS
J
HI
U.

2
EDUCTION
a







LU
J
D
Ul
UJ
a
8
(9
I
a
^^M
0
Ul
Q
2
j




. 	 . — i —
_i —
r- 0








	 	 — 	 	 — 	 '
___ 	 	 — ' — • 	





._ 	 1 	 •
— . 	 1— — 	 	 '

-------
  0
UJ
3
i
0
u.
u >
Q 0


ii
j u


S 0

n G
(fl Q,



i <
T m
A a
Q ~
  t
z <
111
        a
        ut

        u.
        4
            Q

            o
            o
            0
                              a oc
         I
         u.
         UJ
         0
       o
       o
       o
              o

              8
              CO
                 21

-------
     For the Shirley Highway and San Bernardino busway case studies, sig-
nificant energy consumption savings have been generated, although as sum-
marized in Figures 1.8 and 1.9, potential  savings have been reduced through
diversion from previous bus service and carpooling, increased bus fuel
consumption, and auto use for busway system as was found for the Lindenwold
rail transit line.  For the San Bernardino busway, nearly 80 percent of
new corridor bus riders were former auto driver commuters since there was
little transit service previously operated in the corridor.  This has pro-
vided tremendous leverage for generating net energy consumption and air
pollution reductions.  In the Shirley Highway corridor, only 40 percent of
the new riders represent former auto drivers and approximately one-third have
shifted from other bus service, somewhat similar to trip characteristics
found in the Lindenwold corridor.  System auto access levels also vary
between the San Bernardino and Shirley busways reflecting different operating
practices.  Approximately 72 percent of San Bernardino riders reach the system
by auto while two-thirds of Shirley busway users walk to catch corridor
service buses circulating in neighborhoods prior to entering the busway.  In
the latter case, the energy use of circulating buses is a significant factor.
No comparisons of energy consumption for different levels of auto and walk
access (with bus circulation through service area neighborhoods) were con-
ducted.
     Figure 1.8 also includes estimated after energy consumption by autos
no longer used for commuting in the San Bernardino corridor but left at home
and used by other household members.  Survey responses from San Bernardino
busway users indicated that approximately 25 percent of autos formerly used
for commuting were no longer used on a regular basis, genrating opportunities
                                      22

-------
             I
•B
r K.
  IL
  D
  Ul

  gz
  ul 0

  in P
  ID
    w
  (0 Z
  3 o
  CD 0


  i§

  (0 Ul
Ul


IL
Ul
a
o
u.
ui
o
                        10

                        a
       o

       I
        o

        §
        CO
             (O
           o
           o
           o
§
                23

-------
0)
UI
oc
0
u.
UI
ID
1
^9
SAN BERNARDINO BUS
SUMPTION
-ER
•9 \L
Q "
VJ
ANO AFTER ENERGY C
1
a.
u
a
(0
0
0
0
10
n





nd ride
d off
- — Commute auto
left at home
" a
i °

a D HB"B**i" y
1 ' '
I










BUS TOTAL
ii
D
^


I
I
IU
U.
111
(0

0 0
§ §
00 (0


i '
o o
o o
o o
^ N






AUTO BUS TOTAL
               t Ava/siMomvaj isioixcJiAinsiMoa nan=j
                               24

-------
for reducing household auto ownership.  Of the remainder, about 15 percent
were reported being utilized for trips by other household members and
approximately 60 percent were used for trips to and from busway terminal
locations.  Both latter uses of autos formerly used for commuting might
permit replacement of the commuting autos for ones in poor repair or defer-
ment of needed repairs or replacement, generating potential economic benefits
in either instance for new busway users.
     Analysis of the San Bernardino corridor traffic diversion characteris-
tics indicated a potential secondary impact due to reduced traffic congestion
approximately equal to the energy savings generated by diversion to transit
use due to reduced corridor traffic congestion.  No data was available to
validate this second-order impact for the San Bernardino corridor although
available traffic volume data for both the Lindenwold and BART  corridors
showed  that similar traffic  impacts  did  not  materialize as  expected.
1.4.3  FINDING No.  3
          Air Quality in suburban areas may be worsened by auto trip
          making involving cold starts for system access.
     Figure 1.10 illustrates the distribution of carbon monoxide (CO)
emissions changes in the San Bernardino Freeway corridor due to busway
service introduction, showing a significant increase in CO emissions occur-
ing in the vicinity of the El Monte terminal.  Regional air pollution
levels are often as high or higher in suburban areas as in central  portions
of the region, and  the localized impact noted for the San Bernardino corridor
should not be minimized because of its suburban location.
                                     25

-------
0
                                   26

-------
     For the reported study, automobile air pollutant emission rates were
estimated to account for the specific characteristics of auto trips diverted
to transit.  For example, a corridor busway would divert auto trips having
a driving cycle composed of suburban, corridor, and urban segments which may
be quite different than the cycle assumed for the Federal Test Procedures (FTP)
upon which the Environmental Protection Agency emissions estimation procedures
is based.  Also, pollution effects associated with cold vehicle starts have
greater significance when considering trips diverted to transit and recent
research findings regarding cold start emissions are applicable.
     A cold start may be defined as the starting of a vehicle after it has
been inoperative for some period of time such that the engine is  below
operational temperatures.  The combination of inefficient combustion, due
to the engine being below its operational temperatures, and the rich fuel
to air ratio caused by the choke leads to excess carbon monoxide  (CO) and
hydrocarbon (HC) emissions.  The Federal Test Procedure for motor vehicle
emissions testing requires that a vehicle be left standing for a  period of
at least twelve hours before it is tested as a cold start vehicle.  The
difference between emissions generated in approximately four minutes after
starting under these conditions and emissions for warmed-up running conditions
are designated as being due to a cold start.  Recent research  indicates
that the standard procedure may be overly strict, and that a cold start
probably occurs after only six or eight hours of non-use at normal ambient
temperature ranges.
 Argonne National Laboratory.  Handbook of Air Pollutant Emissions from
 Transportation Systems.  Prepared for Illinois Institute of Environmental
 Quality, December, 1973.
                                      27

-------
     Recent studies  using travel data for the Washington and Pittsburgh
metropolitan areas assumed that cold starts may be associated with all
trips that originated at home or work.  Applying this criterion which
reflects recent cold start research findings, it was estimated that approx-
imately 60 percent of all daily trips involve a cold start whereas the  FTP
assumes that cold starts are associated with only 43 percent of urban area
trips.  On this basis, cold start emissions related to trip volumes but not
to trip lengths or speeds amount to approximately one-quarter of total  carbon
monoxide (CO) emissions and about 15 percent of total hydrocarbon (HC)
emissions.
     These results have particular significance when considering transit
trip characteristics.  First, it is expected that a high proportion of
auto trips diverted to new transit service would include elimination of
cold start emissions.  To illustrate this rationale, 1972 survey data for
Washington, D. C. bus trips showed that 69 percent of daily trips were  for
work purposes.  Assuming that all work trips and that about 40 percent  of
the remainder involve cold starts, an estimated 81 percent of all bus trips
if made by auto would involve cold starts.  Similar data for other metro-
politan areas show that between 70 and 90 percent of all diverted auto
trips would have cold starts.  Consequently, transit trips may represent
a potentially higher payoff for air quality improvement due to elimination
of work trips involving cold starts at both the home and work trips ends.
Second, use of an automobile for transit system access trips may lessen
 Horowitz, J. L. and L. M. Pernela.  "Comparison of Automobile Emissions
 According to Trip Type in Two Metropolitan Areas."  Presented at 54th
 Annual Meeting of the Transportation Research Board, January 1975.
                                      28

-------
potential reductions in CO and HC emissions emitted in the vicinity of the
point of transit system access.  San Bernardino case study results demon-
strated that this does occur and may serve to worsen air quality conditions
in suburban areas although overall corridor improvements are obtained.
1.5  FURTHER RESEARCH
     In carrying out the reported analysis, only limited data regarding a
number of factors which have considerable influence on energy consumption
and air quality results was found.  For example, data regarding prior mode
of travel of new transit riders was only available for one case study of
areawide bus service improvements (i.e., Atlanta).  For other case studies,
the Atlanta data was employed to define a probability distribution function
and additional parametric analysis was also carried out to bound analysis
results.  As a second example, trip length data was only reported for the
Lindenwold Line and partially for the Shirley Busway case studies and esti-
mates based on corridor/area characteristics were used for other case studies,
again to define input probability distributions.  Both these data items are
key determinants of energy and air pollution impacts but only limited data
was found for analysis purposes.  In conclusion, additional  research studies
are required for several  areas of importance for complete assessment of
energy and air pollution impacts of new transit service projects:
     • identification of the characteristics of new transit  riders
       especially prior mode of travel and trip length;
     • ridership response to short-range transit improvement programs
       involving route extensions, new lines, public information,
       improved headways, and demand responsive service;
                                     29

-------
• corridor system access mode choice characteristics  and  determinants;
• household auto ownership and trip generation  impacts  following  new
  transit service introduction specifically with  regard to  latent
  trip demand, auto ownership reduction,  and use  of former  commute
  autos for other trips;
• corridor traffic congestion impacts following new transit service
  introduction offering substantial energy and  air quality  improve-
  ment potential;
• second-order residence location and other land  use  impacts of new
  corridor transit service; and
• comparison of transit and highway system indirect energy  utilization
  for construction and operations.
                                30

-------
                                 CHAPTER  2

                            STUDY METHODOLOGY

2.1  INTRODUCTION

     The methodology employed  for carrying  out study analyses  considered
factors affecting the potential  energy consumption  savings  and air  pollutant
emissions reduction of new or  improved transit services  of  two general  types:
          •  energy consumption  and air pollutant  emissions characteristics
             of transit vehicles and automobiles under varying operating
             conditions; and

          •  changes in both short- and long-term  travel  behavior in
             response to the increased level  of transit  service.

     Three new transit service scenerios were selected for  analysis:
          •  bus service improvements on a  region-wide basis consisting of
             route extensions, increased operating  frequencies, new lines,
             reduced fare programs,  and demand-actuated  service;
          •  corridor express  bus service via exclusive  bus lanes;  and
          •  corridor commuter rail  transit service utilizing  modern rail
             transit equipment and  performance standards.
Case studies involving recent  examples of the introduction  of  new or improved
service for each scenario were chosen to provide a  data  base regarding  changes in
trip-making behavior following transit service changes.
                                       31

-------
2.1.1  SELECTION OF CASE STUDIES

     A total of nine case studies were selected to provide inputs for study
analyses.  They were chosen on the basis of data availability, and to provide
as wide a range as possible of new or improved transit service examples.  Case
studies span a wide spectrum of service levels, characterized by model shares
from a few percent of total regional or corridor travel to nearly one-half of
all trips during perk periods.

     Five case studies involved expanded regional  bus service of three
types:
Fare
Reduction
X
Improved
Headways and
Route
Extensions
X
X
Demand
Actuated
Service

    Atlanta (1972)
    Washington (1974-75)
    San Diego (1973)
    San Diego (1975-   )
    Orange County,
    California (1974-75)
Two case studies involved new corridor busway -- one operating on the Shirley
Highway in suburban Washington, D. C. and the second via the San Bernardino
Freeway in the Los Angeles area.  Finally, the Lindenwold high-speed commuter
rail  line and the Bay Area Rapid Transit (BART) were reviewed as corridor rail
transit case studies, although no detailed analysis was carried out for BART
due to the lack of available data for full system operations.
     Case study data will generally not address more than directly-measured
model shift changes following new transit service introduction, although
                                      32

-------
considerable research has been reported describing impacts of the Lindenwold rail
transit service introduction, and major research effort is currently being
applied to measure BART systems impacts.  In the course of this project, it has
been important to recognize (1) inconsistencies among case studies in view of
data limitations; and (2) means of handling inconsistencies which may include:
          t  ignoring them;
          •  inferring estimates from available information; or

          •  carrying out parametric analyses to bound potential impact
             levels.

     The shortcomings of a case study approach have been documented on previous
occasions.   However, they remain the principal means of learning about travel
behavior characteristics and interrelationships with transportation system
attributes and land development patterns.   At the 1972 Williamsburg Conference
on Urban Travel Demand Forecasting, one of the key recommendations was to
extend efforts in analyzing "before" and "after" data for transportation system
changes.2
2.1.2  MODEL DEVELOPMENT
     Three general computer models — one  each for studying (a) regional bus
service improvements, (b) new express corridor busway service,  and (c) new
      Charles River Associates, Inc.   Measurement of the Effects of Transportation
Changes.   Prepared for the Urban Mass Transportation Administration.   August, 1972.
Boyce, D.  C.   "Notes on the Methodology of Urban Transportation Impact Analysis" in
Impact of the BART System on the San  Francisco Metropolitan Region, Highway Research
Board Special Report III.  1972.
     2Brand,  D.  and M. L. Manheim, eds.  Urban Travel  Demand Forecasting,
Transportation Research Board Special Report, 143.   1973.
                                         33

-------
corridor rail transit service -- were developed and applied to analyze the case
studies of improved transit service.

     As summarized in Tables 2.1, 2.2, and 2.3, the models were designed
to accommodate key factors affecting energy consumption and air quality impacts
in one or more of six ways:
          t  as a fixed input value.

          •  as an input random variable -- in this case, lower bound, modal,
             and upper bound values must be specified.   Then, values for the
             variable are sampled from an approximate beta distribution function
             defined according to lower bound, modal, and upper bound inputs.1

          •  as an output random variable— by repeatedly sampling values for input
             random variables, probability functions for output variables (which
             are a function of input variables) may be derived.

          •  parametrically where a range of fixed input values are specified
             to study output changes.

          t  according to default relationships -- the capability to generate
             a probability function for selected variables expected as input,
             but not available, on the basis of other input data items was
             incorporated.  For example, diverted auto trip length was generated
             as a function of urban area population size in studying regional
                                      2
             bus service improvements.
        i, R. Y. and I. F. Kan.  "A Decision Aid to Transportation Planners."
Paper presented at 41st Annual Meeting of the Operations Research Society of
America, New Orleans.
     <•>
      Alan M. Voorhees & Associates.   "Factors and Trends in Trip Lengths,"
National Cooperative Highway Research Program Report No. 48.  1968.
                                         34

-------
        o

        CO
                           \


03
 u.

1-0
CO CO

O LU
»— t :>

LU O
a:
a.

sna
«

«
-




















a
LU
H-
ee
UJ


a
o

?
u.
o

ce
UJ

i
-

-
-

•
















:c
h-
(i

LU

a.
i —
CT

«*
C3
LU
1—
ce
UJ
o










-





(/>
a.
Of


o

s>

1=1
LU
1—
ce.
LU

o
Lt-
0

§

u.
o

LU
h-









-

CO
a.
rr


O
1—



t—
£
S"

o



§
(— 1
h-

f^
— '
-J

^
O

ts
LU
o:
-

«
•

-






Z


Q_

5
O
CJ

LU
^

O

s

K-
z
0
1 —
<
a
LU
h-
CC
LU
O
-

•
-

-














CO
z
o
CO
CO

LU

t-
z
0
n

0
LU
1 —

LU
a
-

•
-

-
















TT


1—
a.
=3
CO
z
o
	 I
LU
75
U-
1—

Z
=J

m



•

-






















CO
z
0
CO
CO
T
LU
J—

Z


1/1
CO

dm oinv

-















_,
f^

1—
u
s
LU


CO
z
o

CO
5-
LU
0

in
rv.
en


1—
LU
Z

-















y.
o

h-
U
g
LU
ce

CO
z
o

CO
CO
LU
(_,
=
CO
r>«
O>


1—
LU
Z

-
























S
Z

s
LO
— J
UJ
u.
m
^.
a^


H-
LU
Z

-












CO
C5
1-
3
a
LU

o
a.
2;


^
"— •
•*
i

5
Wl
—I
UJ

u.

h-
UJ
Z

indino
                                                  35

-------
             TABLE 2.2
FACTORS AFFECTING ENERGY CONSUMPTION
 AND AIR POLLUTANT 011SSION CHANGES
   FOR NEW EXPRESS BUSWAY SERVICE
A
BUS
RIDERSHIP
CORRIDOR
SERVICE
DESCRIPTION
AUTO
DIVERSION
t/i
=
co in
Of
CC LU
UJ Q
5 5
s
UJ
Q
to
t/>
UJ
o
3
C/}
c: s o
z a: <
o uj a.
_J 1— Z
1—
=)
o_
I—
NEW RIDERSHIP
PEAK PERIOD
OFF-PEAK LOAD FACTOR
PREVIOUS TRAVEL MODE INCLUDING DIVERSION
FROM CARPOOLS
LENGTH OF BUSWAY
CORRIDOR TRIBUTARY AREA
ENTRY/EXIT SPACING
BUS UNIT FUEL CONSUMPTION
BUS UNIT CO EMISSIONS
NUMBER OF AUTOS DIVERTED
DIVERTED AUTO TRIP LENGTH
TIME OF DAY OF DIVERTED. AUTOS
REGIONAL LOCATION OF DIVERTED AUTO TRIPS
DIVERTED AUTO UNIT GASOLINE CONSUMPTION
DIVERTED AUTO UNIT EMISSIONS
NUMBER OF TRIPS DIVERTED FROM
FORMER BUS SERVICE
FORMER BUS SERVICE LOAD FACTOR
BUS UNIT FUEL CONSUMPTION
BUS UNIT EMISSIONS
NUMBER OF PARK-AND-RIDE PASSENGERS
NUMBER OF KISS-AND-RIDE PASSENGERS
NUMBER OF FEEDER BUS PASSENGERS
UNIT FUEL CONSUMPTION FOR AUTO ACCESS TRIPS
UNIT EMISSIONS FOR AUTO ACCESS TRIPS
FEEDER BUS UNIT FUEL CONSUMPTION
FEEDER BUS UNIT EMISSIONS
RESIDENCE LOCATION CHANGE
REDUCED AUTO OWNERSHIP
REDUCED TRAFFIC CONGESTION
OTHER USE OF AUTO
NET 1975 HC EMISSIONS REDUCTION
NET 1975 FUEL SAVINGS
NET FUEL SAVINGS WITH IMPROVED AUTOS
NET 1975 CO EMISSIONS REDUCTION
1
1
1
1
I
1
1


1

I
1


1
1


•
•
•




1








•


i
i
•
i


•

•
•


i



t
•
•




i


•











•
•




i
i

t
i
•



i
•
•
i








i



i
i
•


•





i



•
•
i













i





•
i
•
•


•
•
i
i
•
•
•
i
i
•
i
i
i





















•
•







•
•








































•
•
•
•
                      36

-------
             TABLE 2.3
FACTORS AFFECTING ENERGY CONSUMPTION
 AND AIR POLLUTANT EMISSION CHANGES
    FOR NEW RAIL TRANSIT SERVICE
TRANSIT
RIDERSHIP
CORRIDOR
SERVICE
DESCRIPTION
AUTO DIVERSION
O£ c/>
UJ Cg
X 1/1 UJ
a: = o
O 03 i—
u. a:
LJJ
0
s
in
V)
LU
O
o
f
z
ae 1/1
UJ 1—
h- 
-------
          •  output values to "break-even" with regard to net energy consumption
             or air pollutant emissions.

     A key aspect of the modeling procedures was the capacity to handle factors
as random variables with associated probability functions.   In this manner, the
known variations of certain factors due to local conditions or operating procedures
as well  as the uncertainty involved in specifying values for other factors where
there is little or conflicting available  information may be accommodated, and in
turn, directly reflecting in estimates of output planning variables.   Input
variables were characterized by an approximate beta probability distribution
function permitting specification of a lower bound, modal,  and upper bound value for
the variable (instead of mean and estimated variance, for example).*  The models
as developed are generally applicable for regional  or corridor sketch planning
purposes, and this capability to handle variables as ranges or according to
default  conditions is a powerful asset for generating first approximation results
in a quick manner.  The results of this approach are fully displayed in
subsequent chapters.

     The remainder of this chapter presents an overview of factors incorporated
in the study methodology.  Additional  information primarily relating to changes
in travel characteristics due to new or improved transit service introduction
for each of the case studies follows in Chapters 3-5.
2.2  KEY FACTORS AFFECTING ENERGY CONSUMPTION AND AIR POLLUTION IMPACTS

     Case studies data and project analyses were directed towards identification
and assessment of factors affecting net energy consumption  savings and air
pollutant emissions changes associated with the introduction of new or improved
      Hertz, D. B.   "The Risk Analysis in Capital  Investment"  in Harvard Business
Review. Volume 42,  January-February,  1964.
                                         38

-------
transit services including:
          •  auto and transit unit air pollutant emissions characteristics;

          •  auto and transit unit energy consumption characteristics;

          •  ridership response to new transit services including system
             load factor levels;

          •  previous mode of travel of new transit service users;

          •  use of auto for transit system access; and

          •  second-order impacts not directly considered in this study
             including the use of autos formerly employed for commuting
             for other household trips, and possible reductions in auto
             traffic congestion due to diversion to transit.
2.2.1  AUTO AND TRANSIT UNIT AIR POLLUTANT EMISSIONS CHARACTERISTICS

     Automobile emission rates for CO and HC were estimated to account for
specific characteristics of auto trips diverted to transit.  For example, a
corridor busway would divert auto trips having a driving cycle composed of
suburban, corridor, and urban segments which may be quite different than the
cycle assumed for the Federal Test Procedures (FTP) upon which the Environmental
Protection Agency emissions estimation procedure is based.   More specifically,
unit auto emissions factors were developed with special attention to two refine-
ments:
     United States Environmental Protection Agency.   Compilation of Air
Pollutant Emission Factors.   Second Edition, AP-42, September, 1973.
                                       39

-------
          •  application of recent research findings regarding cold start
             emissions; and

          •  use of more representative driving cycles instead of the FTP.
2.2.1.1  Diverted Auto Cold Starts
     A cold start may be defined as the starting of a vehicle after it has been
inoperative for some period of time such that the engine is below operational
temperatures.  The combination of inefficient combustion, due to the engine
being below its operational temperatures, and the rich fuel to air ratio caused
by the choke leads to excess carbon monoxide (CO) and hydrocarbon (HC) emissions.

     The Federal Test Procedure for motor vehicle emissions testing requires
that a vehicle be left standing for a period of at least twelve hours before
it is tested as a cold start vehicle.  The difference between emissions generated
in approximately four minutes after starting under these conditions and emissions
for warmed-up running conditions are designated as being due to a cold start.
Recent research by the Angonne National Laboratory  indicates that the standard
procedure may be overly strict, and that a cold start probably occurs after only
six or eight hours of non-use at normal ambient temperature ranges.
                                           o
     Recent studies by Horowitz and Pernela  using travel data for the Washington
and Pittsburgh metropolitan areas assumed that cold starts may be associated with
all trips that originated at home or work.  Applying this criterion which reflects
      Argonne National  Laboratory.   Handbook of Air Pollutant Emissions from
Transportation Systems.  Prepared for Illinois Institute for Environmental
Quality, December, 1973.
      Horowitz, J. L. and L. M. Pernela.  "Comparison of Automobile Emissions
According to Trip Type in Two Metropolitan Areas."  Presented at 54th Annual
Meeting of the Transportation Research Board, January, 1975.   Also, Horowitz, J. L.
and L. M. Pernela.  "An Analysis of Urban Area Automobile Emissions According to
Trip Type," in Transportation Research Record No.  492, 1974.
                                        40

-------
recent cold start research findings, it was estimated that approximately 60
percent of all daily trips involve a cold start whereas the FTP assumes that cold
starts are associated with only 43 percent of urban area trips.  On this basis,
cold start emissions related to trip volumes but not to trip lengths or speeds
amount to approximately one-quarter of total carbon monoxide (CO) emissions and
about 15 percent of total hydrocarbon (HC) emissions.

     These results have particular significance when considering transit trip
characteristics.  First, it is expected that a high proportion of auto trips
diverted to new transit service would include elimination of cold start emissions.
To illustrate this rationale, 1972 survey data for Washington, D. C. bus trips showed
that 69 percent of daily trips were for work purposes.  Assuming that all  work trips
and that about 40 percent of the remainder involve cold starts, an estimated 81
percent of all bus trips if made by auto would involve cold starts.   Similar data
for other metropolitan areas shows that between 70 and 90 percent of all diverted
auto trips would have cold starts.  Consequently, transit trips may represent
a potentially higher payoff for air quality improvement due to elimination of
work trips involving cold starts at both the home and work trip ends.   Second,
use of an automobile for transit system access trips amy lesses potential  reductions
in CO and HC emissions due to the occurrence of cold start and HC evaporative emissions
emitted in the vicinity of the point of transit system access.   In Figure  2.1,
emission factors are plotted as a function of trip length for a suburban trip
including a cold start.   For trip lengths shorter than 2-3 miles, the  cold start
may have a pronounced effect.   Results of case study analyses will  illustrate
that this does occur and may serve to worsen air quality conditions  in suburban
areas although overall  corridor improvements are obtained.

     Cold start emission factors for 1975 vehicle fleet characteristics were
developed using Argonne  National  Laboratory estimates of cold start  emissions

                                          41

-------
                                                                _  w
                                                                     (0
                                                                  O  UI
                                                                     J
                                                                     I

                                          in

                                          a
                                          5

                                          o
                                                               h  «o
                                  ui

                                  a
                                  1
                                                                     ui
                                                                     (9

                                                                     2
                                                                     UI
                                                               -  (U
0

-------
(grams/vehicle) for new cars for each model  year from 1960 to 1975.   These
unit values were subsequently weighted by the percent of each model  year making
up the 1975 national  fleet with correction for deterioration of control  equipment.1

2.2.1.2  Diverted Auto Driving Cycles
     The Federal Test Procedure employs a driving cycle lasting 23 minutes and
covering 7.5 miles.  It includes multiple stops with cruising portions resulting
in an average speed of about 20 miles per hour.  To more accurately capture diverted
auto trip characteristics, auto driving cycles were developed for suburban, corridor,
and urban travel with variations involving average speeds and number of stops.
Emission factors for component pieces of each cycle were obtained from A Study of
                                                 2
Emissions From Light Duty Vehicles in Six Cities.

     A comparison with equivalent Environmental Protection Agency CO emission
factors (corrected for zero percent cold starts and average speed) indicates
that correcting for different driving cycle conditions within metropolitan areas
can make a difference of up to 30-40 percent (see Table 2.4).  Using EPA speed
and fleet composition adjustment factors and the set of ratios tabulated in
Table 2.4, representative 1975 CO emissions factors were developed for each driving
cycle.  These estimates are summarized in Table 2.5.  For HC emissions estimation,
average EPA unit factors for exhaust emissions were utilized and driving cycle
corrections were not made.
2.2.1.3  Bus Air Pollutant Emissions
     Data on the air pollutant emission rates of diesel-powered buses is not
      Calspan Corporation.   Automobile Exhaust Emission Surveillance, A Summary.
Prepared for the Environmental Protection Agency, May, 1973.
     n
      Automotive Environmental Systems, Inc.  A Study of Emissions From Light
Duty Vehicles in Six Cities.  Prepared for the Environmental Protection Agency,
March, 1973.
                                         43

-------
                             TABLE 2.4

                   COMPARISON  OF DRIVING CYCLE
         AND FEDERAL TEST  PROCEDURE CO EMISSION FACTORS
Driving
Cycle
Suburban


Corridor


Urban


Average
Speed
(mph)
17.6
13.0
9.7
43.9
31.0
23.5
11.3
7.8
5.9
Average
Emissions
(g/mi)
63.1
96.3
129.4
29.5
46.4
65.1
90.9
138.1
185.4
FTp(a)
Emissions
(g/mi)
59.9
76.5
99.4
27.7
37.1
46.9
87.0
119.4
151.0
Ratio of
CO Emissions
Factors
1.1
1.3
1.3
1.1
1.3
1.4
1.0
1.2
1.2
Note:   'a'Corrected for zero percent cold  starts and average speeds.
Source:  TRW Environmental  Services
                                   44

-------
                    TABLE 2.5

       AVERAGE 1975  CO EMISSION FACTORS
          WITH DRIVING CYCLE  CORRECTION

Driving Cycle                         Grams/Mi 1e
   Suburban                            47-103

   Corridor                            27 -  55
   Urban                               45 -  128
   Source:  TRW Environmental  Services
                       45

-------
widely available.  For this study, the values listed in Table 2.6 were identified
and used to define ranges for suburban, corridor, and urban operations for model
input.

     Due to the nature of the diesel engine, different parameters are involved
in producing emissions.  Diesel engines run at higher efficiencies and temperatures
utilizing more oxygen than do light duty gasoline engines.   Therefore, they
produce less CO and HC, and more NO  on a per mile basis than do light duty
                                   s\
vehicles.  One of the major parameters effecting these emissions is the fuel
injection system.  In 1970, a new injector (called a "N" injector) was introduced
which served to reduce the amount of fuel entering the cylinder.  These injectors
were adopted by older diesel engines in order to conserve fuel.   As a consequence
of this fuel conservation measure, diesel engine emissions  were  also reduced.
No analysis of emissions of either nitrogen or sulfur compounds were carried
out as part of this study.

2.2.1.4  Rail  Transit Air Pollutant Emissions
     Electric vehicles have no direct emissions.  Their contribution to air
pollution stems from additional loads placed upon the power plants serving the
system.  Furthermore, power plants do not contribute significantly to CO, or HC
emissions totals in urban areas.   However, power plants may be a major source of
particulates (TSP), NOX, and sulfur dioxide  (S02) pollutants depending on how the plant
is fueled.  Analysis relating to the potential air pollution impacts of BART and
the Lindenwold rapid transit line are summarized in Chapter 5 of this report.

2.2.2  AUTO AND TRANSIT UNIT ENERGY CONSUMPTION CHARACTERISTICS
     Direct automobile gasoline consumption characteristics have been well  researched
and reported.   Figure 2.2 shows the range of auto gasoline  consumption rates employed
                                         46

-------
                               TABLE 2.6


                BUS AIR POLLUTANT  EMISSION FACTORS

                          (GRAMS PER MILE)
               Source
EPA Heavy Duty Vehicle  (5 mpg)(a)

Argonne Heavy Duty Vehicle  (3 mpg)'b'

Environmental Protection Agency(c)
        Arterial
        Downtown

Department of Transportation

        Express Bus
        Other
(d)
                   Carbon Monoxicte

                        20.4

                        32.5
                        15.0
                        28.8
                        10.5
                        10.9
Hydrocarbons

    3.4

    n/a
    3.8
    7.2
   11.7
   14.7
References:  (a)   U.  S.  Environmental Protection Agency.  Compilation of
                 Air Pollutant  Emission Factors, Second Edition, AP-42.
                 September,  1973.

            (b)   Argonne  National Laboratory.  Handbook of Air Pollutant
                 Emissions from Transportation Systems.  Prepared for
                 Illinois Institute for Environmental Quality, December, 1973.

            (c)   Communication  with Mr. D. Syskowski citing preliminary data
                 for 1971  model  year coach supplied by General Motors
                 Corporation.

            (d)   U.  S.  Department of Transportation.  Characteristics of
                 Urban  Transportation Systems.  May, 1974.
                                      47

-------
01
oi
z
g

a


(0
2
0
U
Ul
2
j
0
(0
     19
     0

     5
o .5
                                                                                  D
                                                                                  Z
                                                                                  a
                                                                                  S
                C3-IIIAI/SVQ3
                                        48

-------
for study simulation analyses.  Data prepared by Hirst^ for standard and compact
autos using Federal Highway Administration data summaries is plotted for comparison.
     For this study, special attention was directed towards three aspects which
appeared not to have received thorough consideration in previous research
studies:
          a  Effect of improved auto efficiency on potential energy savings
             due to diversion of auto trips to transit;

          •  Identification of key factors affecting transit system energy
             consumption characteristics; and
          t  Comparison of indirect energy use of auto and transit systems.
2.2.2.1  Improved Auto Efficiency
     Several recent studies have concluded that means for improving automobile
fuel consumption efficiency offers substantial short-term (via carpooling) and
longer-term (via propulsion system and auto design improvements) potential for
transportation energy conservation.  In view of current auto dependency amounting
to approximately 98 percent of total urban passenger travel  by auto, the basis
for significant reductions in energy use through improved efficiency is apparent.
     Whether or not this strategy is able to realize its potential  effectiveness
without additional  means of controlling dispersed land development trends, which
have historically accompanied increasing auto use, is a companion issue which
has not been fully addressed to date.  Recently-completed research studies at
Northwestern University and Princeton University suggest that additional land use
controls are required if long run energy consumption is to be reduced to within
           ,  E.  Direct and Indirect Energy Requirements for Automobiles.
Oak Ridge National  Laboratory, February, 1974.
                                       49

-------
national supply limitations projected for future years.   For the time being,
national policy appears limited to goals related to improved auto efficiency
as the primary thrust for energy conservation in the transportation sector.  On-
going major research programs regarding national energy source development by
the Federal Energy Administration, Environmental Protection Agency, and others
should provide final directions and guidance in this regard.

     To evaluate the impact of imrpoved auto efficiency on the relative energy
attractiveness of new transit system developments, the net energy consumption
change associated with each case study was also estimated assuming that auto
gasoline consumption rates improved by forty percent.   As will be reported in
subsequent chapters, net energy consumption reductions are considerably smaller
and consumption increase conditions more pronounced.

2.2.2.2  Bus Energy Consumption Characteristics
     In Figure 2.3, average bus fuel consumption recorded for seven bus lines
operated in San Francisco and Los Angeles is summarized.   Note that express bus
service operating with few stops on free-flowing freeway lanes results in
significantly lower fuel consumption than local bus service operating at slower
average speeds with regularly-spaced stops.  For planning purposes, use of
average bus fuel consumption data which primarily reflects local service operations
may provide significantly high estimates of actual fuel  consumption rates for
new express service operations.  Model analysis carried out as part of this study
incorporated unit consumption values which accounted for different operating
conditions.

     Average route speed is highly dependent on bus stop spacing policy.  In
Figure 2.4, the same bus fuel consumption data is plotted as a function of average
number of stops per mile.  This data illustrates that express bus operations may
                                        50

-------
PI
oi
   D
   111
   UJ
   a
2 ID

0 u
3 a
in
z w
OP
    CD
                                                    a ~
                                                    in i
                                                      a
                                                   .0
                                                      UI
                                                      a
                                                      CO

                                                      UI
                                                      0
                                                      a

                                                    o H!
                                                    tu a
                                                      ui
           iq      «      n

              C31HAI/ 1VEJJ
                                    tu
                            51

-------
tf )" '-'''f'lv*
^* fj • -l-r -
wil . fi
H 0 •*•:•;?- .
tt H Vi:»
§••• »-«5T?' -
EC • ?^e
D U.
(B a
n 0
8 S J
::{ U) «»"-
ill "^ ;!
=! 3 '£T--,,
• ^MB •. -* -'''''
"* in
u. u
ai 9
rn a •*• +

r
Ul
r i
Dt
Ul
- (0 tt
(0
a
0
-UJ «
u.
0
a
Ul
-« i
i

- 01

in
n
          cu
52

-------
result in significantly improved fuel consumption in comparison with local bus
service operations and overall system average values.
2.2.2.3  Rail Transit Energy Consumption
     As a starting point, six components of total rail transit energy consumption
may be distinguished:
          •  propulsion energy required to move vehicles over the guideway;

          •  vehicle accessories, energy for heating, air conditioning, and
             lighting;

          •  station energy required to provide station heating, lighting,
             information, and parking as well as administrative operations;
          •  maintenance energy associated with the upkeep and repair of
             vehicles and other system equipment and facilities;
          •  construction energy to build the entire system including guideways,
             terminals, vehicles,  and administrative and maintenance facilities; and

          •  impact energy due to  trans it-related development not part of the
             system itself.
     To simplify, transportation energy use may be considered as consisting of
direct and indirect components.   For rail transit, direct energy consists of
vehicle propulsion and accessory energy consumption while indirect energy includes
station,  maintenance, construction, and impact energy components.
     A pair of recent studies have analyzed in some detail  the direct and
indirect energy consumption characteristics of BART and of the Lindenwold rail
transit line.  Healy has derived estimates of five energy use components including
                                         53

-------
construction energy but not impact energy.1  Boyce and Noyelle have estimated
the first four energy components listed above for the Lindenwold under a
recently-completed study for the Federal  Energy Administration.2  Both results
are summarized in Table 2.7 for comparison.

     These are substantial inefficiencies associated with electrical power
generation (from petroleum, fossil fuel,  and other energy sources) and
distribution.  Estimates of point-of-use/source energy vary between 0.30 and
0.35.3  In other words, a unit of source  fuel with an energy equivalent of
100 BUT's will contribute 30-35 BTU's of  electrical  energy for vehicle propulsion
or other purposes.  There are also energy losses associated with the refining
of source petroleum fuels for gasoline and diesel fuel, amounting to source/
point-of-use ratio of approximately 0.9.   For this study, all  energy analyses
are expressed in terms of equivalent gallons of refined petroleum fuels.
Electrical energy use has been converted  to equivalent gallons with appropriate
factors employed to reflect energy losses.

     Direct rail transit energy requirements are dependent on  a number of
factors relating to system design and performance characteristics including
the following:
          •  vehicle weight
          t  vehicle performance
      Healy, T. J.  "Energy Requirements of the Bay Area Rapid Transit (BART)
System."  Prepared for the State of California Department of Transportation,
November, 1973.
     2Boyce, D. C. and T. Moyelle.  "The Energy Consumption of the Philadelphia-
Lindenwold High-Speed Line and Other Modes of Transportation in the South Jersey
Corridor."  Preliminary Draft Report prepared for the Federal  Energy Administration,
January, 1975.
     3Healy, op. cit.  Also Lansing, N. F. and H. R. Ross.  Energy Consumption by
Transit Mode.  Prepared for the Southern California Association of Governments,
March, 1974
                                         54

-------
                                    O

                                    0)
                                <+-  C
                                O  U-l
                                 c  o
                                   =3
                                0.  Q.
                                    O

                                   Q.
                             O
                             O
                                O
                                CM
                                                             CO
            .a

             T3

             "o
              s

              O)
                                                          CM
                                                                                       O

                                                                                       CM
                                                                                              CO
            UJ
            O
            Q_
CM
uj
_:
CQ

UJ
             CO
             u_
             o
             CD
             00

+*
Q.

O
•
1 —
1
Lf)
•
O
^
•
CM
1
CO
«
•""
CTi
•
r**
i
o
•
ID
                                                                                                  Q.
                                                                                                  o
              •!-
                                        o 
                                        o  «-r-
                                        o
                                        ns
                                           •i-  O
                                        (U +->  O
                                       i— _C
                                        O 01  S_
                                                                     01
                                                                     ro
                                                                     +J
                                                                     (O
                                                        HI
                                                        Ol
                                                        O)
                                                     OJ
                                                     O)
                                                     (0
                                                     s_
                                                     o
                                                    4->
                                                     (Si

                                                     d)

                                                    13
                                                                               Ol
                                                                  Ol
          0)
          (J
          c
          IB
          c
          Ol
                                                                                                        >,    OJ
                                                                                                        r—     O
                                                                                                        >
                                                                                                        a>     o
                                                                                                        3=    00
                                                                    <

                                                                    O
                                                                                        (O
                                                                                        •M
                                                                              cu
                                                                              o
                                                                              s.
                                                                              rs
                                                                              o
                                                                              CO
                                                       55

-------
          •  propulsion control system

          •  station spacing
          •  vehicle loading

          t  vehicle accessories
In-depth discussion of each factor is available in a number of engineering
references and planning analysis reports.1  One example may serve to illustrate
tradeoffs involved in system design.   If transit cruise speeds were lowered
from 80 miles per hour, which modern rail transit vehicles are designed to attain,
to 50 miles per hour, source energy requirements per car-mile would be reduced
by over one-half.  This change would cause the average line haul speed to be
reduced from 39 mph to 33 mph, or for a ten-mile journey, a difference of nearly
three minutes in travel time.  This represents about a 15 percent change in
overall line haul travel time, which would generate an expected patronage loss
of approximately six percent.^  However, it would also be necessary to increase
the vehicle fleet size by 10-15 percent to maintain equivalent line haul service
frequencies and seated passenger capacity which would add to total energy
utilization.  In summary, the example performance change would reduce total energy
consumption by approximately 35-40 percent with an estimated six percent ridership
decrease and 10-15 percent increase in the transit vehicle fleet size.
     For rail transit systems, indirect energy use includes energy for stations,
maintenance, system construction and related items.
          example, see Smylie, J. S.  "Energy Consumption of Alternative Transport
Modes -- Bus, Light Rail, Conventional Rail, Group Rapid Transit, and Personal Rapid
Transit."  Prepared for the Third Intersociety Conference on Transportation,
Atlanta, July, 1975.
     ^Applying a direct elasticity value of -0.39 reported by Charles River
Associates, An Evaluation of Free Transit Service (1968) using Boston data.
                                         56

-------
Stations and Maintenance
     For BART system stations and maintenance energy categories, Healy reports
estimated energy consumption equal to approximately 40 percent of direct energy
utilization for vehicle propulsion, accessories, and live storage.  This value
is based on full system operations.  Boyce and Noyelle estimate that station
and maintenance energy for the Lindenwold rail transit line amount to an
additional 25-30 percent of direct energy consumption, comparable with Healy1s
data for BART although no allowance has been made for live vehicle storage.
These relationships may be viewed as approximations for planning purposes.  Energy
consumption characteristics for stations, maintenance and related non-vehicle
system facilities are dependent on a variety of design factors and not necessarily
proportional to direct energy requirements to any significant degree.

Construction
     Healy has also derived an estimate for BART construction energy, determining
that it is approximately equal to source energy requirements for vehicle pro-
pulsion, accessories and live storage over a 50 year period.  For the assumed
50 year lifespan:
                Direct Energy                    1.0 x 1014 BTU
                Indirect Energy
                   stations, maintenance         0.4 x 1014 BTU
                   construction                  1.1 x 1014 BTU
The estimate was developed by applying 1963 energy/GNP dollar coefficients developed
by Herendeen  for each sector of the economy to the dollar amount of materials/
services in 'each sector required for BART construction.  Based on studies of the
energy required to manufacture unit quantities of materials used for guideway
and vehicle construction, Fels has estimated construction energy for BART guideway
 Herendeen, R. A., An Energy Input-Output Matrix for the United States, 1963.
 University of Illinois, Urbana. 1973.
                                        57

-------
and cars to be approximately 25 million KWH per single guideway mile and 1.2
million KWH per car respectively.1  Applying these estimates for full system
development over a 50 year period for comparison with Healy's data results
in a construction energy total of approximately 0.4 - 0.5 x 1014 BTU or
about one-half of Healy's estimated total.  Thus, two independent approaches
to estimating transit system construction energy result in two different
findings—one placing construction energy approximately equal to direct system
energy utilization and the second about one half of this amount.
     Hirst has developed estimates of direct and indirect energy requirements
for automobiles using updated,energy/GNP dollar coefficients by Herendeen2,
the same approach employed by Healy for estimating BART total energy compo-
nents.  From Table 2.8 which summarizes Hirst's findings, indirect automobile
energy equals approximately 60 percent of direct automobile energy for pro-
pulsion and vehicle accessories.   Of this total amount, construction energy
is equivalent to about 22 percent of direct energy or from two to four times
less than for rail  transit construction.   In Summary,
                                     Auto3            Rail  Transit3
     Direct Energy                   100                  100
     Indirect Energy
     Maintenance, System Facilities   40                  25-40
     Construction                     22                  50-100
 Fels, N.F.  Comparative Energy Costs  of  Urban Transportation  Systems.   Trans-
 portation Program Report,  Princeton  University.   September,  1974.
2
 Hirst, E. Direct and Indirect Energy Requirements for  Automobiles.
 Expressed as percentage of direct energy consumption for  vehicle propulsion
 and accessories.
                                     58

-------













r-.
CD


A
UJ
to
"^

00 O
CM UJ
CO UJ
-Cl UJ

LU CQ

r— CM

4-» "~O
O •—
1— 	 •
0>
CO
13 ^^^*

>
C CO
III *- -^
UJ ***^





4_)
c
 o -^
S- i- 3
CO M_ 4J
C M- CO
UJ CO 	
O
o
CO
o

^"
S. >
(O "^

o
a













c
0
•r-
4->
O
C
3
l_l_

O CO CM i—






O O O O
000 O
LO O CO r—
A A
CT> CM


O O O O
O O O O
CTii— «5f i—
A A
CT> CM







,~.
(O
*— O
§o
^^
CM «

1 O CO 1






co cr> CM
o o o
o o o
1 . .
1 O O O






CO
n™
rt3
to
c
o
•r" 4-)
4J O) S-
a. c o
ai s •— a.
C 3 C to
•r- to T- C
•— C M- (0
O O (U S-
CO O S- 4J i—
fO *^
C5 O

CO •—






O O
0 0
CM CM
A



O O
0 0
CO ro
A
r^









O 0
o o
O 0
M A
W3 O^
IO CM






CO O^
C\J O
0 0
• •
o o







3
«c

CM r-






O O
0 0
^r CM




o o
0 0
*f CM











O 0
o o
0 0
n •«
r^ 0
CM VO






VjD «3"
r- O
0 0
* •
O O


A
CO
(J
g*
ra
c
(U
+J
c
• fH
its
E

A
to co
i. -p
•r- i. CO
(t3 T} Cl)
a. a. s-
co ••-
0= H-

oo






o
o
*3"




o
o
^J-











o
o
0
A
LO
CM






fs,^
»—
o
•
o














CO
o
c

S.
3

c
H-l

co






o
o
LO




o
0
LO











o
o
o
A
^^
CM






00

o
•
o





A
O)
c
•r—
O^
trs
s*
(Q
o>

n
o> to
C i—

j^ O
S. 4->
(T3
a.
0
UD O






0 O
O O
0 CO
A A
i— LO


0 0
o o
i~~ r^*
A A
r— <£>
i^—








o
o
0
A
CO







*sf CO
i— CM
O r-
• •
o o








n*""**
>» c
rO O

TC7 ^>
0> 0

f S-
>^_^J _^J
CO CO
CO C i—
CO O 03
X O 4J
fO O
1— 1—















































•
co


3 •
4-> Q.
CO

3 40
4J tO
CO i.
•r*
— ^X

*«^X
• .
CO

(U i-
4-> 3
0 O
2: to
59

-------
consider the implications of this difference if, in fact, the report estimates
of  indirect energy use are correct in their relative magnitude.  Using Lansing
and Ross' estimates of average source energy consumption per passenger mile
for transit and auto modes :

                                     Btu per passenger mile
                                     Auto             Rail Transit
     Direct Energy                   8,360              5,240
     Indirect Energy
     Maintenance,System Facilities   3,340              1,300-2,100
     Construction                    1,840              2.620-5,240
                       TOTAL        13,540              9,160-12,580

At the upper bound level  for rail transit indirect energy elements, the
difference between transit and auto energy consumption on a passenger-mile
basis using average vehicle load factors is not large.  While this is a sig-
nificant conclusion especially when considered in conjunction with our other
findings regarding total  transit system energy consumption (including, for
example, access requirements), it should be cautioned that the reported
methodologies have employed aggregate data summaries and varying approaches
for lifetime amortization.  Further research is needed to verify the magnitude
of indirect energy use differences and clarify available results.
2.2.3  RIDERSHIP RESPONSE TO NEW TRANSIT SERVICES
     The case studies examined in this project involved a variety of new
transit service types, each offering a different degree of attractiveness
 Lansing and Ross'.  Energy Consumption by Transit Mode.   Using 0.25 vehicle
 load factor.
                                     60

-------
for competition with auto travel.  Estimating ridership response to new transit
services is a difficult task and becomes particularly difficult for bus service
improvement programs involving route extensions, increased service frequencies,
new lines, public information and marketing, and new equipment such as for
case studies in Atlanta, Washington, San Diego, and Orange County .   There is
considerable data regarding observed ridership elasticities to fare changes
and, additionally, much on-going and recent research concerning new corridor
transit services offering high levels of service.  However, relationships
describing potential ridership of areawide bus service improvement programs
are not well established.  For this study, actual ridership response data
was utilized although it was necessary to rely on default relationships based
on limited case study data for new rider characteristics data including
former mode of travel.  As will be reported in later chapters, further
research is required to establish the characteristics of trips diverted to
transit to fully assess transit's energy consumption and air pollutant
emissions impacts.
2.2.3.1  Short-Term Changes in Travel Behavior
     It has generally been observed that ridership response to new transit
service will stabilize within three months of its initial  introduction.  Within
this period, individuals may elect to use it under one or more of three con-
ditions.
 Holland, D. K.  A Review of Reports Relating to the Effect of Fare and
 Service Changes in Metropolitan Public Transportation Systems.   Prepared
 for the Federal Highway Administration, June, 1974.
                                      61

-------
Make a Trip Not Previously Made
    These trips do not contribute to any reductions in energy use or air
pollution levels,  "induced" or "latent" demand for transit service is an item
not well understood.  Household trip-making has generally been treated as being
independent of transportation system characteristics,  and little research has
been applied to studying latent travel demand characteristics.

Shift From Using an Automobile
    This group of new service users will generate air pollution and energy
consumption reductions to the extent that the amount of automobile travel is
reduced or eliminated (minus the increases due to new transit services).  The
degree of impact may be increased if the reductions in auto trip-making occur
in heavily-travelled corridors during peak periods, resulting in improved auto
travel speeds and less traffic congestion.  On the other hand, net energy
consumption savings and air pollutant emissions reductions may be lessened
to the extent that:
         •  automobiles not used as a result of shifting to transit are
            employed for other trips by other household members; and
         •  new transit users may make an automobile trip to reach the
            new transit system, either to park and ride or as an auto passenger
            to be dropped off.
Urban area transit trips are generally shorter on the  average than auto trips,
reflecting a combination of generally poorer transit service levels in outlying
areas, greater opportunities and propensity for auto travel in suburban
areas, and other factors.  The degree to which new transit services continue
                                      62

-------
to attract relatively shorter trips will influence potential energy consumption
savings and air pollutant emissions reduction amounts.

Shift From Other Modes
   Passengers diverted from other transit modes, from traveling as an auto
passenger, and in some cases perhaps from bicycle and walk modes, may provide
some overall reductions in system energy utilization and air pollution to the
extent, if any, that new transit services are less energy and pollution
intensive than former modes.  It appears that diversion of persons traveling
as auto passengers is the most critical element of this issue.  Furthermore,
shifts from carpooling to new services recently or currently being implemented
must be interpreted in view of existing low auto occupancy and carpooling
activities.  Increased attention to the potential benefits of large-scale
carpooling utilization suggests that the issue of competition between transit
and carpooling as alternatives to the automobile cannot be answered from
analysis of available experience.

2.2.3.2  Long-Term Changes in Travel Behavior
   Over a longer period of time, the introduction of new transit services
may generate changes in household trip-making characteristics with substantial
energy consumption and air quality implications.  These changes are related to:
        •  reduced automobile ownership; and
        •  shifts in residence locations and other land uses.
Major decisions of this type are dependent on many social, economic,  and
environmental  concerns in addition to accessibility and it becomes difficult
                                      63

-------
to distinguish what should be attributed to development of new transit service
and what would have occurred in any case.

     The decision to replace or purchase an automobile should be influenced
by the availability of improved transit services.   A reduction in automobile
ownership levels due to new transit system development and availability is able
to generate substantial savings in total household transportation costs,  and
provide a key basis for new system justification.   For purposes of this study,
auto ownership reductions are especially significant as the longer range basis
for achieving energy consumption and air quality goals.  Table 2.9 shows  the
variation in daily household auto trips as a function of household auto owner-
ship from available data summaries.  While this tabulation does not account for
other key household characteristics (e.g., size, number of employed persons,
income) which affect both trip generation  and auto ownership, it does suggest the
potential for reductions in vehicle miles  traveled accompanying reduced auto
ownership levels.  To date there has been  little research into measuring  how
increased transit accessibility influences household auto ownership.   A recent
study by Dunphy using the 1968 home interview data collected by the Metropolitan
Washington Council of Governments identified statistically significant relation-
ships between transit accessibility and household auto ownership.^  Further
analysis of this important area is required.
     •'•Dunphy, R. T.  "Transit Accessibility as a Determinant of Automobile
Ownership."  Paper presented at the 52nd Annual  Meeting of the Highway Research
Board, January, 1973.
                                       64

-------
en

CSJ

UJ

ca

o: ,—
<_3 • 	
GO
O
O • — -
oo i —
1 — O^
1 1 1 , —

Q_


O
<_>
GO . — -
(_) to
2T CT^
ce -^
u_
;>-
z <:
• CTl

CO •* — -
<£.






o — -
CC LO
LJ CT)
i — i , —

O









CO


CXL
h—


LU


Q

LU
CJ


LU

^£












CO
ce

o
1 1
o

LU
CO
•z.

, — CO
OJ eg csj to ro «*
coroco ^cOCM CTirotNJ





OOO 'shCMiJ- OOOVO
r^^. r*- t**x cxi r— LO ci ^O r*^
• •« . . . ..,

, —







oicncn cxiuaro r^tDLO
Ch OlO^CTi LOCO^

CJCXICSJ LOCXICNJ CQOJCSJ









^
ro i-

o
LO *O tO
$*- Q. C CL T3
 <3J QJ ^ ^J CD
S-O_Lu -r-CUQ_ -F-D-O.

CD CD Q 1) CD Q  cn 01 cn CD en
CDrtJfO i — fOfO ' — rt3fO

•r- (D QJ 4-J (D QJ 40  ^r
3 ^
O 0
£! !_
CJ3

(O C
U ro
w 1 JU
"O O S-
"o +J
>
S- en s
fO O- r-
u -i- cn
1 S- •«-
0) 4-> 3T
c
O 4- OJ
0 S-
to 3
 to
en -o
"3 C t/1 c
s- o n o
CD U .— • -r-
> CD Q. 4->
fT3 LO rB
-a c
i- -o c: T-
fO C O +-»
 J-J to -O
to to i
S- S- S_ C
••- -i- O ••-
tO CO QJ S_
QJ QJ en o
E E ro
13 3 S_
to to QJ
tO tO > QJ

-------
                               CHAPTER 3
                   EXPANDED  REGIONAL BUS SERVICE
3.1  INTRODUCTION
     Throughout the nation,  metropolitan  areas  have  been  improving bus service
levels on a region-wide basis.   Due to  lower  average fares  (Figure 3.1) in-
creased bus-miles of service (Figure 3.2),  and  new bus equipment  (Figure 3.3),
and spurred by last winter's fuel  shortage  conditions, bus  ridership has swung
in an upward direction after thirty years of  steady  decline (Figure 3.4).  The
availability of operating subsidy  assistance  under provisions of  the 1974
Mass Transportation Act should  serve to further stimulate bus service expan-
sion and continue the momentum  of  current trends  in  coming  years.
     Study analyses presented in this chapter address the energy  conservation
and air quality impacts of regional  bus service improvements.  Specifically,
three types of service improvements have  been analyzed:
     •  fare reductions;
     •  increased bus-miles  of  service  due  to new routes, route
        extensions, and headway reductions; and
     t  demand actuated service (dial-a-ride).
Data inputs have been derived primarily from  five case studies although
numerous other references, and  data summaries maintained  by the American
Public Transit Association have been employed as  well.  The case  studies
involve:
                                      66

-------
 10

 Ul
 0
Ul
I
                                        3.1
         TREND OF AVERAGE FARE
     35.
     30
as.
so .
     IS
                                  i     1    I    I
                                        3.2
         TREND OF VEHICLE  MILES
.   1.5OO ..
X
10
  1,250 .
  1.0OO
                                                3.3
         TREND  OF  NEW  MOTOR BUSES
  5,000
ffl 4,000 .

!
0 3,000 .
a
3 2,000 .
  1,000
        YEAP
                i     i    r    i    r    i     i
               'B5   'BB  '67  '68  '69  '7O   '71
                            67
                                          -7B   "73  '74

-------
     to
     a
m  ui
11  o

     ui
     CO


     s

     U!


     UI

     UI
     a

     u.
     0

     a

     ui
     a
                                                                              . 
-------
                            Fare         Increased         Demand
                         Reduction       Bus-miles        Actuated
Atlanta                      X               X
San Diego (I)                X               X
          (ID                               X
Washington                                   X
Orange County,
  California                                                  X
     Table 3.1 summarizes selected operating characteristics and 'before1 and
'after1 conditions for each of the case studies.
3.1.1  METHODOLOGY
     The model developed for studying the impact of expanded regional  bus
service using input data from each of the five case studies incorporates key
factors as noted in Table 3.2  A detailed flow chart is included in Appendix A.
     Data limitations have required two key assumptions to be invoked.   Data
regarding the prior travel mode of new bus service users was only available
for the Atlanta case study and it was utilized for the other four cases as
well.  In addition, the number of diverted auto trips was treated parametri-
cally for other case studies.  A second assumption was required to estimate
diverted auto trip lengths.  Average trip length relationship based on  regional
population provided a basis for estimating the modal value for diverted auto
trip lengths, with lower and upper bounds assumed to be 80 percent and  110
percent of the modal trip length.
     Auto and bus fuel consumption rates and air pollutant emissions factors
are incorporated as random variables in the model.  Modal, lower bound, and
                                      69

-------
       CO  CO 4->
       ui  i-  C
       T3  «5  CO
       CO U~  U
                                       LT>
                                                     LTJ
                                                     CM
                    Lf>
                    CM
                    O
                    en
                     i
                    Ln
                    CM
                              Q.  00


                           >>J=  C
                          i— tO   -o
       >,•—  c
       r— -r-  rt3
       •r-  £  00
       
              to
       CU CO •(->
       00 S~  C

       ca u.  u
           CL oo

       >>^="c
       •—  01  ITS
       •i—  i.  01
       "T3  CO  3
       o -a  o
          •r- .C
          a: •(->
           to  01
           co -o

       r— -r-  03
       •r-  £  l/>
       03     3
       O  to  O
           3 -C
          03 -M
       4->  O  O1

       E 4J  O

       X <— i—
       O  3 •—

       a. o  s
       O.O- *-*
                                       LO
                                              in
                                              co
             LT>
             ro
             00     •—
             Lf)     CM
LO
CO
CO
OO
CM
CO
                                              en
                                              Lf)
                                                     CM
                                                           ro
                                              o

                                              ro
                                 00
                                        CM
                                        pv.
                                        (Tl
                                         C
                                         rt)
                                              co    ^-^
                                              r-~    CM
                                              cr>    i—
                                              r—    CTi
                                                            *-^    CT*
                                                             I     i—
                                               C
                                               O
                                                      o
                                                      CT
                                                      O)
                           ai
                           C    -r-
                           •r-    Q
                                               to
                                               ia
                                                     00
                                         o
                                         a>
                                         CO
                                         c
                                         rO
                                        OO
                           C

                           O
                           o

                           a>
                           en
                           c
                           fO

                           o
                                70

-------
                          K
CM

CO
CQ
        CO
        LU
        oo
        oo
        ;S  if
        =3  >
        _
        O  CO
        a.
            CO
        en  =3
        i—  CQ
Q  «C


I—


LU

O
^

—
§
















*JJ7
.*£
a
-J CO
CJ _J
58
a.

0 5


LU C£
1> Lu
h- 0

t/i vi
^2 CX.
O LU


LU a

a.

sna
«

-
-

















O
LU
h-
ce
LU
S

CO
e
«
u.
0

o:
LU
—
Z3

O
Lu


LLJ
>

O










4»



Q.
Qi
h-

^
3

a
UJ

ae
LU
a

Lu
o


O
-

-
•

-












z
o
CO
CO
?•
LU
h-
"ZL
13
0


Q
LU

C£
LU
>

Q
-

•
-

•













_^
0
u-
a.
7"
oo
0

_,
LU
Lu

J—

^^


CO

CO



-

•


















on
O

oo
s:
LU

1 —

-rE


OO

CO

dial oinv

*













3"
0

£3
s
LU
CC
CO
•^
o
CO

5:
LU
O

CO

O>


^_
tu
2:

-













s
o
M
c3
g
LU
2C
CO
^
o
CO

s:
LU
y

Lf>

cn


K~
LU


-




















00
O
3i
1
00
—I
LU
ut

in

^


h-
LU


-











00
>=,
"^
a
LU
O
Q£
a.
2:
~*

^—
3
CO

£
CO

— J
LU
	 -
Lu

I—
LU


indino
                                                71

-------
 upper  bound  values were estimated according to the methodology described in
 Chapter  2.   All energy savings have been expressed in terms of equivalent
 gallons  of refined gasoline.
     From data in Table 3.1, four case studies provide estimates of rider-
 ship elasticity to increased bus-miles of service, i.e.,  Aridership as
                                                          A bus-miles
 follows:
          Atlanta                      0.6-1.3*
          Washington                   0.3
          San Diego (1973)             1.7-1.9*
          San Diego (1975)             0.4
The values noted with an asterisk have been discounted for the expected effect
                              1 o
of concurrent fare reductions.      It seems that fare reductions may have had
a greater impact than would be  expected by application of historical relation-
ships.   With regard to system load factors, an elasticity value of less than
one means that the service improvements have resulted in a lower system load
factor than for before the improvements were made.

3.2  ATLANTA
     Following initiation of public transit operations in the Atlanta metro-
politan area in early 1972, passenger fares were lowered from a previous base
fare of 40 cents to 15 cents.   This reduction generated  substantial  new rider-
ship, requiring the acquisition on an emergency basis of used buses  from other
transit operations to be reconditioned and put into  service to relieve passenger
overloads.  Through November 1972, a total  of 117 service changes were made
 Holland, D. K.   A Review of Reports Related to the Effect of Fare and Service
 Changes in Metropolitan Public Transportation Systems.   Prepared for Federal
 Highway Administration, June 1974.
2
 Highway Users Federation.  Transit  Fare and Ridership:   A Review. Dec. 1974.
                                      72

-------
which increased the annualized bus miles of operation from approximately 19
million at the date of acquisition to approximately 22 million annual bus
miles of service.  The service changes  were made  primarily  in  the  area
of improved headways and expanded service periods.  There were 85 such changes,
and in addition, 13 lines were extended, 14 lines were revised and five new
lines were installed .
3.2.1  NEW RIDERSHIP
     In response to the noted changes in bus service introduced in the Atlanta
region, weekday bus ridership increased approximately 28 percent over the
expected level.  In this case, the expected level  of ridership accounts for
a continuation of the declining ridership trend experienced  in preceding years.
Discounting for this trend results in an average weekday ridership increase of
approximately 13 percent.
     There is extensive empirical data regarding patronage response to fare
changes from numerous fare increases and, more recently, from fare reductions
such as in Atlanta and San Diego.  Applying this data, the Atlanta fare reduc-
tion of sixty percent (from 40 to 15 cents) would  generate a 10-20 percent
patronage gain.  Consequently, the remaining 8-18  percent ridership increase
may be attributed to the greater bus-miles of service, or expressed as an
elasticity with respect to increased bus-miles:
     E.m «  AN/N      =  8-18   =  0.6-1.3
      OT  ABM/BM          14
     In other words, a one percent increase in the number of bus-miles operated
would generate an expected 0.6-1.3 percent increase in bus usage.   The overall
fare reduction program resulted in increased load  factors for existing service,
 Metropolitan Atlanta Rapid Transit Authority.   Analysis of Transit Passenger
 Data.  October, 1973.
                                     73

-------
 causing additional equipment to be deployed to accommodate heavier passenger
 loadings.  However, ridership response to increased bus-miles operated
 (assuming bus trip lengths are constant or do not increase substantially)
 will serve to reduce the overall system load factor if the elasticity with
 respect to increased bus-miles is less than unity.
3.2.2  Diversion of Automobile Trips
     Table 3.3 summarizes the previous travel mode of new bus riders as deter-
mined via an interview survey program following introduction of service improve-
ments.  Only 42 percent of the new bus trips had been previously made by auto-
mobile drivers, representing a total of nearly 22,000 auto trips removed from
streets and highways each day.  Significantly, over one-half of these former
auto trips occurred during the morning and evening peak periods (see Table 3.4),
and their diversion would be expected to relieve traffic congestion to generate
possible secondary impacts on energy consumption and air pollutant emissions
for remaining auto drivers.
     The proportion of new transit riders diverted from auto passenger and
walking modes should be noted.  As shown in Table 3.3, approximately one-quarter
of the new riders were attracted from these modes, both of which are already
efficient from an energy consumption and air pollution standpoint.  In addition,
nearly 22 percent of the new transit trips were not made previously at all.
3.2.3  NET ENERGY CONSUMPTION SAVINGS
     Figure 3.5 shows the output probability distribution of fuel savings due
to the introduction of bus service improvements in the Atlanta region.  The
estimated mean fuel savings is approximately 9,300 gallons per day or less
than one-half percent of regional daily fuel consumption for passenger transpor-
tation purposes.   Savings varied from a 99-percentile lower bound of approxi-
mately 4,300 gallons to an upper bound of 13,100 gallons, reflecting uncer-
tainty associated with auto trip diversion factors listed in Table 3.2.
                                       74

-------
                          TABLE 3.3
                 PREVIOUS TRAVEL MODE FOR
                  ATLANTA NEW BUS RIDERS
Previous Mode
Auto Driver
Auto Passenger
Walk
Other
No Trip
              TOTAL
Number
21,642
11,324
 2,328
 5,343
11,151
51,788
Percent
  42
  22
   5
  10
  21
 100
                            75

-------
                           TABLE  3.4
                 WEEKDAY DISTRIBUTION  OF
               PREVIOUS AUTOMOBILE DRIVERS

Time Period                   Number                Percent
6-9 a.m.                      4,990                  23.0
9 a.m.  - 3 p.m.               5,582                  25.8
3-6 p.m.                      7,506                  34.7
Remainder of day              3,564                  16.5
               TOTAL         21,642                 100
                             76

-------
10
ri
o
jj
Q
Ul
a

a
a

(0

8
UJ
u.
                                                                                          - D
                                                                   0)
                                                                   Q

                                                                   4
                                                                   a
                                                                   a
                                                                                                o
                                                                                            CO   Q
                                                                                                z
                                                                                                I
                                                                                                Ul
                                                                                                0

                                                                                                jj
                                                                                                Q
                                                                                                III
                                                                                                a
                                                                                            .
                                                                                                g
0
a
                                      ID
                                      d
                                                                         (U
                                                                         d
                                                   =10
                                              77

-------
3.2.4  NET AIR POLLUTANT EMISSIONS REDUCTION
     The expected distribution of CO emissions reduction is plotted in
Figure 3.6 with an average reduction of approximately 13 tons per day.
Hydrocarbon emissions would be reduced by 670 kilograms  (kg)  using 1975
average Environmental Protection Agency data.

3.3  WASHINGTON
     Following a pattern similar to that already described  for Atlanta, the
Washington Metropolitan Area Transit Authority assumed regional  bus service
operations in mid-1972, and initiated a program of  service  expansion,  public
information, and equipment replacement.  During 1974, bus miles  of service
in the Washington region were expanded by approximately  16  percent, reflect-
ing a combination of increased service frequency on some lines,  routing modifi-
cations, and the introduction of new lines primarily in  outlying suburban
areas.
     Patronage response to these improvements has not been  high—an increase
of approximately five percent over the preceding year has been recorded .  This
corresponds to a direct elasticity with respect to  increased  bus-miles of
about 0.3, less than the 0.6 - 1.3 range estimated  for Atlanta.   As was noted
in discussing Atlanta results, low ridership response is resulting in  overall
system load factor to be lowered.
     No survey data exists regarding the prior mode or other characteristics
of Washington's new transit riders, and consequently, information from the
 Washington Metropolitan Area Transit Authority.
                                      78

-------
                                                                                                       0
                                                                                                       01
(0   2


1
      Q
      Ui
      a
      0   G
      3i
      en

      ui
                                                                     ID
                                                                     «•   „
                                                                         (0

                                                                         0
                                                                         \-
                                                                         w

                                                                         (0
                                                                     «   2
                                                                     r   0
                                                                         in
                                                                         to


                                                                         ui


                                                                     tu   8

                                                                     P
                                                                                                           Q
                                                                                                           UI
                                                                                                           a
                                                                                                       10
                  q
                  r
q
d
ID

0
tu
d
                                                            =10
                                                    79

-------
Atlanta case study has been employed where required and additional parametric
studies carried out.
3.3.1  NET ENERGY CONSUMPTION SAVINGS
     The distribution of expected fuel savings for Washington is illustrated
in Figure 3.7 using Atlanta data to establish the mean of the function des-
cribing the number of diverted automobiles.  Note that there is an estimated
0.56 probability of increased total fuel  consumption in this case with an
increase of 250 gallons expected per day on the average.
     In Figure 3.8, the sensitivity of estimated daily fuel  consumption changes
to the number of auto trips diverted is illustrated.  A breakeven point of
'zero' net fuel savings is reached when approximately 45-50 percent of new
bus trips are diverted from automobile driving.  With 75  percent auto diversion,
average fuel savings of approximately 4,100 gallons per day would result
assuming 1975 auto fleet and trip length characteristics.   Furthermore,
expected savings would be eliminated if either (a) the average trip length
of diverted auto trips were approximately 4.6 miles or shorter, or (b) the
diverted auto trips were made by automobiles averaging 24-25 miles per gallon
in city driving conditions.  Breakeven probability distributions for both
auto trip length and average gasoline consumption are plotted in Figure 3.9.
     If a 40 percent improvement in automobile gasoline consumption efficiency
is assumed, expected net fuel savings have a higher probability of being
negative.  As shown in Figure 3.10, fuel  consumption would increase unless
more than approximately 70 percent of new bus riders have been diverted from
auto driving.
                                       80

-------
is Z
U

D
111
DC



0
1°

(0
  0
2 t
0 7

°l
              CD
              d
                   ID
                   d
Oj
d
                                            (0
                                            cu
                                               0
                                               z

                                               I
                                               J
                                               111
                                               2 <
                                               0 0
                                               Q
                                               w -i
                                               "3
                                               aw
                                            CU
                                            i
                                           - to
                       81

-------
0
ri
     \
                                                               \
                                                                 \
                            cxva U3d sNomva aNVsnaHJ.1
                                   SQNIAVS nan=i a
                                32

-------
ui
u

Ul
EC
DC
3
U
U
0

u.
0
J
5
0
a
a
.8  .
.6  .
                                     3.9


           BREAK-EVEN  VALUES

           WASHINGTON, D.C.
                         4567     89



                         AUTO TRIP LENGTH C MILES 1
in

i
in

i
ui
u

Ul
a

5
U
U
0
u.
0
>•

J

0
<  -a
ID
0
a
a
         .01  .as  .03  .04   .oe   .06   .07  .as  .09


                         AUTO PUEL CONSUMPTION

                           C GALLONS PER MILE]
                                                   .10
                           83

-------
\
     \





0
r
n













in
UI
u
5
(£
2
0
3)
a
ui

a
0
S
IL
G
EFFECT






2
g
u
Q
ui

2
g
COIMSUMPT
111
LL
2














GTOIM, D.C.
IMIHSV/W

\ \
\
\ \ '
\
\
» "'
\ \
\
\
a v
£
n
|
0
§
0
i
M-
0
JJ
(Q
•Q
I
a
w

\ g
a 5
ffl \ £
\ iSm m
(L 2
a f
•\ ^
ui 2
\ a a
\ ^0
\ \ go
\ \ s
\ * III
o \ a
to \ \
\ \
\ \ \
\ \ \
^ \ \
8 V \ %
* \ \ %
•<»\ \ ^
\ \ \
\\ \
\ ® ®
x \
\ \ \
-8 x \
\ \ \

     ni
C AVO
                           cu
                     SIMOmVEJ ONVSnOHXJ
                   SONIAVS
                 84

-------
3.3.2  NET AIR POLLUTANT EMISSIONS REDUCTION
     Assuming 1975 auto fleet characteristics and data from Atlanta survey
results, the improved Washington bus service program has generated CO emis-
sions reductions as shown in Figure 3.11 with an average reduction of nearly
six tons per day.  The average reduction amount varies between an estimated
three tons and 11 tons per day as the percentage of new transit riders assumed
to be diverted from auto driving increases from 25 percent to 75 percent (see
Figure 3.12).  1975 HC emissions would be reduced by approximately 260 kg,
using Atlanta survey results regarding diverted auto characteristics.
     In terms of requirements for meeting National Ambient Air Quality stan-
dards in the Washington region, the bus service improvements have made an
effective contribution.  Table 3.5 summarizes these requirements as specified
in the Transportation Control Strategy which calls for a total CO emissions
reduction of approximately 633 tons in the 6 a.m. to 2 p.m. period, including
21 tons attributed to implementation of transit and car pooling programs .
Assuming that one-half of the estimated CO emissions reduction due to expanded
bus service may be credited to the maximum eight hour period, bus service
improvements have contributed about fifteen percent of the required strategy
amount.   If an upper bound level  corresponding to 75 percent auto diversion
is assumed, the contribution amount is approximately doubled to nearly one-
quarter of the strategy requirement.
  District of Columbia Government, et al  Additions and Revisions to the
  Implementation for the Control of Carbon Monoxide, Nitrogen Oxides,
  Hydrocarbons and Photochemical Oxidants for the District of Columbia
  Portion of the National Capital Interstate Air Quality Control Region,
  April, 1973.
                                      85

-------
                                                                                                          cu
                                                                                                          r
r   0)
r   2
n   s
      (0
      (0

      111
      o   u
      °d
      Q   o
                     Q
                     r
                          r^
                          (0

                          0
                          M

                          (0
                     O   2
                          0
                          0)
                          (0


                          111

                                                                                                               D
                                                                                                               Ul
                                                                                                               oc
                a
                r
(0
d
(0
o
CU
o
                                                             =ia
                                                      86

-------
\
\

01 ui
*• 5
n n
2
0
i
UJ
^
Q
0
i
IL
0
u
UJ
• 1
IL
IL
UJ



W
Q
3)
(0
UJ
0
u
DUCTION OF
UJ
cc

2

\ \ \
\ \
x \ X
\ \ \
\ \ \
\ \ \
\ \ \
\ ® \
\ \ \
\ \ ^\
^ \ v.
XA \ o\
^ \ W
U \ \ \
• NP\ \ ^
I \\\
o \\\
h S \ \
0 \\ \
I \\ \


^t
5
u D
a


ID
(D N[
5
Q
AUTO
(0
„ 0
1 1
SO 4C
PERCENTAGE PREVI




tfniooiovcuQ
r r r
CAVQ U3d SNOJ.)
                        87

-------
                             TABLE 3. 5
           SUMMARY OF 1975 TRANSPORTATION  CONTROL
       STRATEGY  FOR WASHINGTON METROPOLITAN REGION
Total  CO Emission Reduction Required
Vehicle Turnover
   Additional  Reduction Required
Transportation including improved
  mass transit, increased terminal
  costs, reciprocal enforcement of
  parking tickets, and car pool
  locator service
Vehicle Inspection and Maintenance
Retrofit of all pre-1975 Autos
Retrofit of all pre-1975 Light Duty Trucks
Selective Control of Goods Movement
Aircraft Taxiing Emission Reductions
                                TOTAL
6 a.m.-2 p.m.
    Tons
    633
    343
    290
     21
Percent
of Total
Reduction
   100
    54
47
177
34
56
30
7
28
5
9
5
    708
   112
                                  88

-------
     A second aspect of net air pollutant emissions reduction involves the
geographic orientation of bus riding in the Washington area.   Bus use patterns
are highly oriented towards the central area of the region, and less intensive
as the distance from the central area increases.  This pattern is generally
typical of many urban areas as well.  Consequently, it may be considered that
transit improvements will have relatively greater impact on air quality con-
ditions in central portions of the region.  For comparison, Figure 3.13 shows
that Washington's air quality problems may be more pronounced in outlying
suburban areas as was the case for the day shown during the summer of 1974.
Even though transit may be effectively contributing towards attainment of
air quality standards, it is not well suited to reduce auto usage outside of
central area travel corridors where a large proportion of urban area travel
and related air pollution problems may take place.

3.4  SAN DIEGO
     In August 1972, the San Diego Transit Corporation lowered fares to a
flat 25 cents (regardless of trip length) from the previous fare structure
of a 40 cent base fare plus zone charges up to a total of ninety cents for
a trip.  The fare reduction resulted in a very sharp increase in patronage,
and continued growth has been sustained since that time.  Some of the more
recent growth may be attributed to other factors including an extensive
public information program by San Diego Transit.
                                      89

-------
3
                                                        •
        AIR  QUALITY INDEX  READINGS
                 WASHINGTON, D.C.
                        JULY
                                          Silver Spring 115
                  Lewinsville
                (Balls Hill Rd. &
             Lewinsville Rd ) 105
AIR  QUALITY
    INDEX
 0-25 Good
 26-50 Fair
 51-75 Poor
 76-100 Unhealthy
 101-250 Hazardous
 251-750 Dangerous
                  • Seven Corners 115
                  Arlington 50
                           /
              f Alexandria 125
                                                      Cheverly 135
                                                            /  /
                                                      Suitland  125
    Engleside
(Route 1,South of Alexandria) 110
  Note:  The air quality  index is  a
         composite measure of several
         air pollutant concentrations
                              90

-------
     A second major service change in the region has involved implementation
of the fiscal year 1975 action plan.  This plan was envisioned to add some
63 percent more bus miles of service.  As of February, 1975, implementation
of the action plan has been frozen, with slightly more than half of the
incremental route miles in service.  The freeze was the result of a combina-
tion of factors including sharply increased operating costs, the lack of
sufficient buses to implement the complete program, increased maintenance
requirements on the buses because of heavier loads and greater usage, and the
disappointing patronage on many of the action plan routes already implemented.
It is uncertain when the remaining routes will  be implemented.
     For this study, introduction of the fuel action plan has been assumed
to provide a second San Diego case study.  Ridership estimates derived for
this case have been used for study analyses reported in a subsequent section .
3.4.1  NEW RIDERSHIP
     The 1972 fare reduction program resulted in average passenger fares
dropping from 28 cents to 23 cents, a change of approximately 39 percent.
Applying empirical relationships, ridership would be expected to increase by
6-13 percent in response to lower fares.  However, coupled with a 29 percent
increase in bus-miles operated, a dramatic increase of 16 percent was recorded
over a two-year period.  From this data, an elasticity with respect to increased
bus-miles of between 1.7 and 1.9 may be inferred—approximately double the
average value computed for Atlanta and six times that estimated for Washington,
D. C. areawide bus service improvements.
 De Leuw, Gather & Company.  "Preliminary Evaluation of Existing Transit
 Operations," Working Paper No. 1.  Prepared for the Comprehensive Planning
 Organization of the San Diego Region, April, 1974.
                                      91

-------
     The 1975 action plan involved expansion of bus-miles of service by 63
percent, with ridership projected to increase by 26 percent.  In this case,
direct ridership elasticity with respect to bus-miles is approximately 0.4,
resulting in a lower system load factor following implementation of service
improvements.
     No survey information was available for either case study.   Data inputs
regarding the mode shift characteristics of new riders for the Atlanta case
study have been utilized for modeling analyses.
3.4.2  NET ENERGY CONSUMPTION SAVINGS
     Figure 3.14 shows the net fuel  savings probability distribution function
associated with the 1972 fare reduction and service expansion program.  The
average savings of nearly 5,000 gallons per day represents less  than one
percent of daily 1975 passenger transportation energy utilization in the
San Diego region.  Figure 3.15 summarizes the results of varying diverted
auto trip lengths on net fuel savings.   The bounded area encompasses 99-
percentile output values generated by sampling at random from input variable
distributions for diverted auto trip characteristics.  Fuel  savings results
except when average diverted auto trip  lengths are less than approximately
two miles.
     Assuming that average auto fuel consumption efficiency is improved by
forty percent, expected fuel savings remain positive with an average of about
2,300 gallons per day as shown in Figure 3.16.
     As would be expected, estimated fuel savings due to the 1975 action
plan improvements which involve an increase of twice as many new bus-miles
                                     92

-------
sr z
                                      01
    z
  3 S
  to r

  ll
                                      o
                                      r
                          0
  Z 0
  n Q
                                      10
  Ul
  d
                                      tu
OB
                  10
                               CU
                    93

-------
        \
ss
n
  a


  ii
  « F

  oil
  u g
  j u
  Ul EC
  ? Ul
  2 0
  0 Q
  h U

  S!
  Q
  Ul
  d (0
          \
                ID
                      INII
                 94

-------
                                         ID





(0
ri


















0
F
a
5
ONSUI
U
UJ
UL
2

Q
F
B
Q
UJ
DC









2
0
F
UJ
a
Ul
a
*

0
(D
Ul
D
2
a
0)







m


i
g
u
"3
0
P

(Q
D
4*
2
a
E
™

Q OB U)
aaiMatdtin

























































i
« 01
0 0
q
r

q
- 01


0
- 0
0
" tri
0
" ui


0


0
01

0


F*
(0
z
Q
i
Q
4
3
Q
w
1
IL
2.
0
U
U
a
V
niva






95

-------
of service and lower patronage response than for the fare reduction program
would be smaller.  Average energy consumption would be slightly reduced with
0.44 probability of increased consumption (see Figure 3,17),  Expected fuel
savings over a range of diverted auto trip lengths are summarized in Figure
3.18, indicating actual energy use increases when average trip lengths are
less than eight miles.  No data is available regarding the trip length
characteristics of new bus riders in the San Diego region, although average
trip lengths are probably less than eight miles.
     If average auto gasoline efficiency is improved by 40 percent, increased
fuel consumption results with high probability for expected diverted auto
trip characteristics (see Figure 3.19).

3.4.3  NET AIR POLLUTANT EMISSIONS REDUCTION
     The San Diego fare reduction program would lower CO emissions by approxi-
mately 7.5 tons per day and HC emissions by 400 kg per day for 1975 auto fleet
characteristics.   The CO emissions reduction probability distribution function
is plotted in Figure 3.20   In Figure 3.21   the range of CO emissions reduc-
tions is shown for varying diverted auto trip lengths including 99-percentile
minimum and maximum values.
     For the 1975 action plan improvement package, the net CO emissions reduc-
tion distribution function is shown in Figure 3.22 with a mean reduction of
5.4 tons per day.  The expected reduction level varies according to the
average length of diverted auto trips as plotted in Figure 3.23.   For 1975
national auto fleet characteristics, an average net HC reduction of approxi-
mately 240 kg has been estimated.
                                      96

-------
  Q

nfc
                                      0
                                      r
(0

0 j
u a


n
                                        - 0
                                        - 10
C U
  2 Q
  0


  §D

  gz
  UJ
Q

|



10

Q
U

J
Ul

2|


?3

2 <
0 u



§S
D ,1
UJ (fl


>S
^ I-
H 10
          CD
                   (0
                                01
                                      01
                   97

-------
n z
r o
   <
u  J
u  a



si
U.  h
                \
                                                (0
                                                111
                                                J
                 \
                  \
                    \
                      \
                        \
                         \
                          \
                                         \
   ID
                      01
                                      01
                                      i
(0
              CAVQ M3d 8IM01-IVO
                            IMI iMaixonaau
                        98

-------
ss
n E
h
8
H 2
2°
     0
     4J


     l


     I

     "5
  Qg
  I™ «
  U Q

  Q
  ui
  EC
   h
     ID
               =to
             99

-------
H   y!
                                                                                          01
                                                                                          r
                                                                                          0
                                                                                          r
                                                                                          •   8
                                                                                              0
                                                                                              p
                                                                                              M
                                                                                              i
                                                                                        - tu
           o
           r
to
0
to
0
(U
d
                                                  =10
                                          100

-------
\
         \
            \
r  W
cu  2

ri  oj
     (0
     in


     8
         Z
         g


         y
         D

         0
         s
     E


     *
     UJ
     DC
         l
                \
                                                                                     - o>
                                                                                       - 10
                                                                                              in
                                                                                              ui
                                                                 z

                                                                 Q

                                                                 UI
                                                                 J

                                                                 a

                                                                 S


                                                                 o


                                                                 3

                                                                 Q
                                                                 UI

                                                                 a


                                                                 \
                                                                 Q
                                                         - 01
         (0
   CU
   r
               0
a
                                                     CAVQ M3d 8NOJ.1

                                                     oo IMI
                                            101

-------
q

r
q

0
ID

Q
tu

d
                                 do  Aj.i-iiavaaucj
                                                                 (U
                                                                 p
                                                                    in


                                                                    \
                                                                    w

                                                                 .  1

                                                                    0
                                                                    (0
                                                                    (0


                                                                    111


                                                                 •  8
                                                                    g



                                                                 «§
                                                                    Q
                                                                    111
                                                                    DC
                            102

-------
           \
n  «
*l2'
   "z
   0 0
   o
   z o
   0 0
   h y!
   0 Q
   UJ <
   a (o
            X
             \
               \
                \
                                        0
                                       - (0
                                           111
                                           J
                                           i
                                           111

                                           a
                                           S
                                           0


                                           \

                                           Q
                                           111
                                           III


                                           a
                                       - ai
   a
   r
          O
                       CAVQ


                SIMOI8SIIAI3 OCJ Nl
                              SISIOJ.)
                     103

-------
3.5  ORANGE COUNTY
     Located in suburban Los Angeles, Orange County Is the fastest growing
metropolitan area in the State of California with a population of approxi-
mately 1.8 million persons.  The county is characterized by low density
suburban development, served by an extensive street and freeway network,
typical of its Southern California location as well as many other large
urban areas throughout the country.
     To compete with the automobile in this environment of diffused travel
patterns, the Orange County Transit District has undertaken implementation
of a public transportation system comprised of a number of community or local
area bus systems which serve intra-community travel needs, while simultaneously
serving as collection and distribution subsystems for an extensive county-
wide network of bus routes designed to facilitate inter-community travel.
In February 1973, community demand responsive (dial-a-ride) service was
initiated in LaHabra as a pilot project.   At the present time, seven addi-
tional communities are receiving local demand actuated service, and county-
wide fixed route, fixed schedule service  is being operated.  Neither system
is considered complete, and systems are growing in ridership as service is
being offered in increasing amounts where practically no service was offered
before.
     The results of the Orange County program are of national interest and
potential application.  As was mentioned  in interpreting the Washington find-
ings regarding the central area corridor orientation of conventional transit
services, urban areas throughout the country are addressing the question of
penetrating the automobile market in suburban areas.  The approach under
                                      104

-------
development in Orange County offers a potential  means for providing a trans-
portation alternative.
3.5.1  NEW RIDERSHIP
     The Orange County Transit District operates 524,000 miles of bus service
monthly, providing at least one hour service in  selected communities.  At
the present time, average weekday ridership is about 24,000 riders, equivalent
to only 0.5 percent of the 4.5 million daily trips made by county residents.
3.5.2  NET ENERGY CONSUMPTION SAVINGS
     Figures 3.24 and 3.25 snow the estimated probability distributions for
net fuel consumption increases resulting from Orange County bus service intro-
duction, assuming that 25 percent and 50 percent of new riders are diverted
from auto driving.  In both cases, five miles was assumed as the modal  value
for the diverted auto trip length input function.  In Figure 3.26  average net
fuel savings are shown for combinations of number of diverted auto trips and
diverted auto trip length.
3.5.3  NET AIR POLLUTANT EMISSIONS REDUCTION
     The introduction of county bus service has  resulted in lower 1975  air
pollutant emissions levels dependent on the degree of auto diversion and
the length of diverted trips.  Average net CO emissions reductions are
summarized in Figure 3.27 for a range of diverted trip characteristics.  For
example, an estimated reduction of between 1.0 and 3.4 tons per day results
if 25 percent of new bus trips were formerly auto driver trips ranging  from
2.5 to 7.5 miles in length on the average.  For  the same diverted trip  charac-
teristics, between 20 and 130 kg of HC emissions would also be eliminated.
                                      105

-------
                                   0

3
(U
ri














2
g
a -
i °
mi JS
(0 g
2u! o
8™i .-
•J is
rf *•
jU.a
Ul .«•
^•B M • *
D)-44
u.h q
ZmLm
2^1 1 1 1
J 1 1 1 1 1
— 3 m r o u> « ni
g8«
0 a)
hH! P
U§Q
2la
D3lO
uigcu
rroS

0
- 10 ^
ui 3
0)
0
u
u «
-« 21
l\
0 Q
c
u Q
h « §2
0 ,^
S3
>!
i-
o Q





- 0
1
106

-------
10
01
ri












z
g
a
D
COIMS
j
ui
D
IL
Z

z
D
REDUCTI






ALIF.
U
M

Z
0
u

ORANGE

1
r



o
'55
0)
TJ
a

P
B
a

c
0)
u
0)
a
a
10

i i i i
n iq $ 01

                            - ID
i
6
1
a
2
                               a

                               i
                               0
                               u
                               III «
                               2§
                               zS
                               2<
                               go
                               G!

                               li
                               5JE
                               r
                             n
                             i
                             «
                             i
33N3tlfcinCIQa
        AJLIIIBVBOUd
107

-------
                                                                 Ul
q
cd
Q
ui
                       CAVCI
 q          c
 oi

siMQ-nvQ do
        IMI
q
ni
                                 108

-------
                                                      0
   Cfl

M £
m 0
n fn
   UI
      M

   o >•"
   " H

   ii

   §8
   c ui
   n ID
   D 2
UJ  DC
ff  0
                                                      q
                                                      CD
                                                      a
                                                      to
                                                         I

                                                         0

                                                         111
                                                         a
                                                         i-
                                                         Q
                                                         III

                                                         EC
                                                         111


                                                     - P  D
                                                      q
                                                      ai
                                           01
                                CAVO tl3cl SNOJL)
                         as MI SNaixanaau
                         109

-------
                               CHAPTER 4

                 NEW CORRIDOR EXPRESS BUS SERVICE

4.1  INTRODUCTION
     Four categories of corridor  express bus  service may be identified:
     •  Busways on Freeway Rights-of-Way--the modification of a freeway
        by the construction of new  bus-only lanes or the screening off
        of existing lanes.
     •  Reserved Freeway Lanes--either  in the peak flow direction or
        contra-flow where reverse direction traffic volumes are suffi-
        ciently light to permit 'wrong-way' bus operations on reserved
        lanes.
     •  Special Bus Ramps and  Ramp  Metering—the provision of special
        bus ramps to and from  freeways  or preferential bus entry to
        freeways that are controlled  by ramp  metering techniques.
     •  Arterial Bus Priority  Treatments—includes the provision of
        exclusive bus lanes located along curbs or in street medians,
        and bus priority operation  through pre-emption of traffic
        signals.
Any of the above categories can be  altered by the incorporation of carpools
onto the exclusive lane whether the lane be a concurrent or contra-flow
lane, an exclusive on or off ramp,  or a toll  plaza bypass.
     Major examples of the implementation of  busway operations may be
found throughout the United States.  The two  projects used for case study
                                    110

-------
analysis are indicated with an asterisk, and will  be discussed in greater
detail following this introductory section and brief methodology descrip-
tion.
Busways
     Shirley Highway, Washington, D.  C.*
     San Bernardino Freeway, Los Angeles*
     1-95, Miami
Reserved Freeway Lanes
     1-495 in New Jersey through the  Lincoln Tunnel  to N.  Y.  City
     U. S. 101  to the Golden Gate Bridge into San  Francisco
     Long Island Expressway, New York City
     San Francisco Bay Bridge toll plaza bypass
Special Bus Ramps
     Blue Streak project, Seattle
     Commuter Club bus service, Reston, Virginia
     Harbor Freeway, Los Angeles
4.1.1  METHODOLOGY
     Analysis of the fuel consumption savings and  air pollutant emissions
reductions associated with busway development projects was carried out
using a computer simulation model (BUSWAY), which  employed treatment of
key input and output variables as having probability distribution functions,
and not just single values.  The distribution functions were  developed
to reflect the range of uncertainty associated with  input  and output
variables.
                                    Ill

-------
     Table 4.1 summarizes the factors which affect the net energy conser-
vation and air quality impacts of busway development and operations.   The
table indicates which of these factors have been programmed in BUSWAY,
the extent of BUSWAY analyses conducted, factors for which some data  is
available but not included in project studies, and factors for which  no
data has been found.
     BUSWAY assumes a corridor configuration as shown in Figure 4.1.   It
further incorporates assumptions that (1) passenger demand is uniformly
distributed within the specified service area which means that approxi-
mately the same number of buses enter the busway at each entry point, and
(2) buses circulate through a portion of the service area for passenger
collection and distribution before entering the busway.   While it is
believed that the reported results are significant and that input data has
been carefully prepared, it is cautioned that findings are dependent  on
some data items for which little or no  empirical data exists and on  assump-
tions employed in the simulation model regarding corridor and bus service
characteristics.
     Energy savings are expressed in equivalent gallons  of refined gasoline
with bus diesel fuel consumption converted for case studies analysis.

4.2  SHIRLEY HIGHWAY BUSWAY
     Initial busway operation started September, 1969 with continuing
improvements through the present time.  The busway is a double, reversible
lane (both lanes inbound in the a.m., outbound in the p.m.), 11 miles long,
                                    112

-------
             TABLE 4.1
FACTORS AFFECTING ENERGY CONSUMPTION
 AND AIR POLLUTANT EMISSION CHANGES
   FOR NEW EXPRESS BUSWAY SERVICE

a.
(/I ts>
CO LU
o
s
z
fuj h-

O UJ CO
(J t/) UJ


z
O i/">

a

1/1
CO CO
CC UJ
a: ce
p



LU
O
i
l/l
LU
tJ


h-
CJ Z 0
Z CC <
O UJ Q_


£
o

NEW RIDERSHIP
PEAK PERIOD
OFF-PEAK LOAD FACTOR
PREVIOUS TRAVEL MODE INCLUDING DIVERSION
FROM CARPOOLS
LENGTH OF BUSWAY
CORRIDOR TRIBUTARY AREA
ENTRY/EXIT SPACING
BUS UNIT FUEL CONSUMPTION
BUS UNIT CO EMISSIONS
NUMBER OF AUTOS DIVERTED
DIVERTED AUTO TRIP LENGTH
TIME OF DAY OF DIVERTED AUTOS
REGIONAL LOCATION OF DIVERTED AUTO TRIPS
DIVERTED AUTO UNIT GASOLINE CONSUMPTION
DIVERTED AUTO UNIT EMISSIONS
NUMBER OF TRIPS DIVERTED FROM
FORMER BUS SERVICE
FORMER BUS SERVICE LOAD FACTOR
BUS UNIT FUEL CONSUMPTION
BUS UNIT EMISSIONS
NUMBER OF PARK-AND-RIDE PASSENGERS
NUMBER OF KISS-AND-RIDE PASSENGERS
NUMBER OF FEEDER 3US PASSENGERS
UNIT FUEL CONSUMPTION FOR AUTO ACCESS TRIPS
UNIT EMISSIONS FOR AUTO ACCESS TRIPS
FEEDER BUS UNIT FUEL CONSUMPTION
FEEDER BUS UNIT EMISSIONS
RESIDENCE LOCATION CHANGE
REDUCED AUTO OWNERSHIP
REDUCED TRAFFIC CONGESTION
OTHER USE OF AUTO
NET 1975 HC EMISSIONS REDUCTION
NET 1975 FUEL SAVINGS
NET FUEL SAVINGS WITH IMPROVED AUTOS
NET 1975 CO EMISSIONS REDUCTION
1
1
1
1
1
t
1


1

1
»


1
t


1
1
1




«







1


1
1
1
1


1

1
i


1



t
1
I




t


1











t
1




1
1

•
1
•



1
•
1
1








1



»
1
1


1





,



1
1
1













•





1
1
•
•


1
|
,































«
|







I
1








































1
1
1
g
                     113

-------
                            4/1

GENERALIZED BUSWAY MODEL
             ll
                      •Service Area
            0

            0.
            X
            111
               114

-------
built into the median strip of the freeway (see Figure 4.2).  Most bus
lines routed via the busway circulate through neighborhood areas prior
to entering the busway, although several park-and-ride lots are located
along the length of the busway.
     Ridership counts have been made periodically to measure growth, and
a.m. peak period ridership has risen from an initial 1900 passengers to
11,500 passengers as of October, 1974, representing approximately 40 percent
of all peak period person trips in the freeway corridor.  Since its opening,
construction in conjunction with freeway widening has caused significant
traffic congestion through two-lane bottleneck sections of the adjoining
freeway.  Under these conditions, bus users have enjoyed additional travel
time savings which may be removed when construction is completed.
     Input data for study analysis has been derived from the results of
an on-board bus survey conducted in Fall, 1973  plus additional information
from earlier reported surveys carried out by the National Bureau of Stan-
                                 o
dards Technical Analysis Division .
4.2.1  DIVERSION FROM AUTO DRIVING
     Approximately 40 percent of the bus trips made on the Shirley busway
each day were formerly automobile trips, made either in the Shirley Highway
corridor or in some other part of the region prior to a change in residence
 Personal communication with the Technical Analysis Division, National
 Bureau of Standards.
 U. S. Department of Transportation, Urban Mass Transportation Administra-
 ti on.  The Shirley Highway Express Bu.s-on-Freeway Demonstration Project- -
 Second Year Results.  November, 1973.
                                    115

-------
01
     in  01
                             116

-------
location (see Table 4.2).  Of the remainder, an estimated 35 percent pre-
viously used other regional bus service and 15 percent were auto passen-
gers, either as part of a car pool or as a regular passenger with a spouse
or other individual.
4.2.2  BUSWAY ACCESS MODE
     One of the key factors in evaluating the energy savings and CO emissions
reduction of busway and rail transit corridor systems is the mode of access
used to reach the corridor system.  For the Shirley busway, survey results
indicate that nearly one-third of the busway users drive or are driven to
use the busway service.  This continued auto use, involving a 'cold start',
is significant in lowering the potential impact on energy savings and air
pollutant emissions.  The access mode breakdown for the Shirley busway is
summarized in Table 4.3.
4.2.3  BUSUAY INFLUENCE ON RESIDENCE LOCATION CHOICE
     Nearly 75 percent of the busway users have located within the Shirley
Highway corridor within the preceding five years.  Of this group, approxi-
mately 37 percent expressed that the availability of busway service had
a 'definite' influence on their residence location decision.  An additional
19 percent indicated that its availability had a 'slight'  influence, while
the largest portion of new residents using the busway responded that it
had 'no' effect on their new residence location.
     The development of the Shirley busway and other major fixed transit
facilities may exert a strong influence toward energy consumption and
pollutant emissions levels via their influence on land use patterns,
                                    117

-------
                              TABLE 4.2
                    PREVIOUS TRAVEL  MODE OF
            SHIRLEY BUSWAY RIDERS,  A,M, PEAK PERIOD
                           Old Residents^            New Residents^
Mode                       Number   Percent           Number   Percent
Auto Driver                1860       16
Carpool  Driver               300        3               3220        28
Alternate Carpool Driver     480        4
Auto Passenger               420        4
Bus                        1000        9               3070        27
Other                       610       _5                520         5
               TOTAL        4700       41               6800        59
Note:  (a)  Totals may not add due to rounding
       (b)  New residents are those who have  located in the Shirley Corridor
            following introduction of busway  service.
                                   118

-------
                            TABLE 4.3
                      MODE  OF ACCESS  FOR
           SHIRLEY BUSWAY  RIDERS,  A.M.  PEAK PERIOD

Mode                          Number                    Percent
Park and Ride                   2710                       24
Walk                           7760                       67
Dropped Off                     1010                        9
Other                           120                       J_
            TOTAL              11,500                     100
                             119

-------
hypothesized to tend to encourage higher development densities.  The cited
survey findings are not sufficient for judging whether or not the Shirley
busway has generated net positive or negative impacts in this regard.
4.2.4  BUSWAY PASSENGER LOAD FACTORS
     Buses using the Shirley busway during morning and afternoon peak periods
carried an average of approximately 49 passengers according to June, 1974
survey results.  Buses have a seated capacity of 47 to 52 passengers
depending on the seat configuration.  Note that in deadheading or making
a scheduled trip in the reverse direction, passenger loads will be zero or
small, thereby reducing the average load factor by approximately one-half.
For bus service operated on parallel arterials within the corridor, passen-
ger loads in the peak direction averaged between 26 and 42 per bus counted
in June, 1974.  During off-peak periods, buses operating in the corridor
average less than 10 passengers per bus at the maximum load point.
     These load factors were utilized as input to BUSWAY runs for both the
Shirley and San Bernardino busway case studies.
4.2.5  NET ENERGY CONSUMPTION SAVINGS
     It is estimated that the Shirley busway has generated a net fuel
consumption savings of approximately 3,100 gallons per day, with 99-
percent probability that the amount of savings has not been less than
1300 gallons or in excess of 4,800 gallons as illustrated in Figure 4.3.
     For comparison, recall that the areawide bus service improvements
for the Washington area generated an energy savings of only 250 gallons
per day with significant probability of actual increase conditions depend-
ing on auto diversion characteristics.
                                     120

-------

u.
b
I
   g
   a

   in
h  Q
r
H  UJ
D  Z
!"  J
       a
                                                                       _o
                                                                       - 00
                                                                       -CD
CO
o
                              h-
                                                     jv	^

                                               •CM   .—

                                                    Q
                                                                            Q
                                                                            LU
                                                                            CO
                                                    = 5

                                                      LL
                                                    z. o
                                                    O
                                                    — co
                                                    h- Q
                                                    O Z
                                                    ^ <
                                                    Q CO
                                                    LU Z3
                                                    ce o
                                                      I
                                                    >- H-
                                                                             <
                                                                             Q
                  S
           cq
           d
                                  m
                                  d
                 C\j
                 o"
                                            do Ainievaoad
                                   121

-------
     Assuming that diverted autos consume 40 percent less gasoline on the
average, the distribution of expected fuel savings shifts resulting in
mean savings of 1,200 gallons per day, or a range of between approximately
50 gallons and 2,400 gallons of fuel saved daily (see Figure 4.3).
     Further investigation of energy use components 'before' and 'after1
busway opening illustrates the effects of auto use for system access and
increased bus operations on potential energy savings.   Elimination of
approximately 9,200 daily auto trips accounts for a potential gasoline
savings of nearly 7,000 gallons.  However, increased bus fuel consumption
and gasoline used for auto trips to and from the busway serve to reduce
net savings to approximately 3,100 gallons as shown in Figure 4.4.  If
it were assumed that all busway users were diverted from autos with an
average occupancy of 1.4 persons per auto, a fuel savings of about 13,000
gallons per day, instead of the actual estimated amount of 3,100 gallons,
would be determined.
4.2.6  NET AIR POLLUTANT EMISSIONS REDUCTION
     Assuming 1975 automobile fleet composition, use of the Shirley busway
has reduced CO emissions by an average amount of 3.5 tons per day (see
Figure 4.5).  This represents a contribution of about ten percent towards
meeting the maximum eight hour CO reduction assumed for Vehicle Miles
Traveled (VMT) reduction measures as part of the transportation control
strategy for the Washington region to meet ambient air quality standards
by 1975.  Hydrocarbon emissions have also been reduced by approximately
50 kg per day on the average.
                                   122

-------
                             1*

                             i
a
ui

u.
Q


4

Ul

gz
LL   0
UJ   C
ID   a


&   2
3   D
U)
     (0
_   0
Q   (j
                a
                ui

                u.
                                                                             (0


                                                                             00
Uj   0

ff   a
=   Ul


(0   Ul
                Ul
                UL
                Ul

                0
                                                                             0
              o
              o
              o
              o>
                                            o
                                            o
                             (0
o
o
o
CM
                                    123

-------
10
  Z
  0

bE

g°
u m

*n
I (0

• g
D 0
0) (0


>!
^
a
i o
(0 U
                                  o

                                ,0 ^


                                  CO


                                  o


                                  CO
                                - CO
                                  o
                                  o
                                 (O
                                  cc
                                • CM •-.

                 3DN3yan3oo do AiniavaoMd
                 124

-------
4.3  SAN BERNARDINO BUSWAY
     The San Bernardino busway has been constructed along a railroad right-
of-way, either within the median or adjacent to the San Bernardino freeway
east of the Los Angeles central business district (see Figure 4.6).  It is
approximately eleven miles in length and has two lanes with bi-directional
operations.  In contrast  with Shirley busway operations in which buses
enter the busway at several intermediate locations following a turn off the
busway for passenger collection, San Bernardino busway operations are
focused at a single terminal located at the eastern terminus of the busway.
The El Monte terminal provides parking for over 1,500 automobiles as well
as facilities for transferring from feeder buses and 'kiss-and-ride1 access.
In addition, buses using the busway are routed through adjacent communities
prior to arriving at the El Monte terminal to pick up transferring passen-
gers and continuing into downtown Los Angeles.  Two intermediate stations,
one located adjacent to California State University at Los Angeles (3.5
miles from downtown) and the second at the County Hospital (about two miles
from downtown),  were recently opened as additional destination points for
busway service.
     Busway ridership has increased from approximately 500 passengers at
opening in January, 1973 to about 4,800 passengers in February,  1975  during
the three hour a.m. peak period.  The system has sustained two years of
continuous patronage growth, with this growth enhanced by improvements
including a fare reduction to 25 cents and the introduction of contra-flow
                                    125

-------

-------
bus lanes in downtown.  Current ridership represents about a thirty per-
cent share of total peak period person trips made in the corridor.
     Looking at peak period, peak direction person trips, the busway volume
is now approximately 60 percent of the volume carried by one of the parallel
freeway lanes.
     Input data for the case study analysis has been provided from on-board
surveys carried out in November, 1974 by Bigelow-Crain Associates for the
Southern California Association of Governments (SCA6) and via contact with
the Southern California Rapid Transit District.  Additional information
describing first-year operating results was also available from earlier
surveys conducted by Grain & Associates for SCAG .   A review of key input
data items and case study analyses follows.
4.3.1  DIVERSION FROM AUTO DRIVING
     The November, 1974 survey results regarding the former mode of San
Bernardino busway users are summarized in Table 4.4.  Adjusting for those
respondents who indicated the busway as their only mode of travel, approxi-
mately 79 percent of all riders have been diverted from auto driving,
representing 5,200 autos removed from the freeway and parallel corridor
arterials during the a.m. peak period.  This is nearly twice the shift
noted from the Shirley busway, with the difference primarily attributable
to greater existing bus ridership prior to busway opening in the latter
i nstance.
 Grain & Associates.  "First Year Report—San Bernardino Freeway Express
 Busway Evaluation".  Draft Report prepared for the Southern California
 Association of Governments, January, 1974.
                                     127

-------
                              TABLE 4.4
                      PREVIOUS TRAVEL ME OF
                   SAN  BERNARDINO BUSWAY  USERS

Previous Mode                     Number               Percent
Always used busway                   650                   10
Auto Driver (alone)                 3830                   59
Auto Driver (with passengers)        450                   7
Auto Passenger                       720                   11
Alternate Carpool Driver             720                   11
Non-busway Bus                       130                   2
                        TOTAL      6500(a)                100

Note:  (a)  Estimated  total weekday ridership, February  1975
Source:  November,  1974 on-board survey conducted  by  Bigelow-Crain Associates
         for the  Southern California Association of Governments.
                                   128

-------
     Nearly one-third of the busway users formerly traveled in an automobile
carrying two or more persons.  However, the  average auto occupancy for
all automobile trips diverted to the busway is approximately 1.24 persons
per auto, which is slightly lower than the estimated corridor average of
1.3 persons per auto.  This difference does not appear to indicate that new
transit users are more likely to have been carpool participants than not.
4.3.2  BUSWAY ACCESS MODE
     In sharp contrast to Shirley busway access characteristics, approxi-
mately 72 percent of San Bernardino busway riders reach the system by
automobile (see Table 4.5).  This serves to reduce potential fuel savings
and air quality improvement since an automobile trip involving a cold
start is still made in the corridor.
     Of the auto access total, over three-quarters (55 percent of total
busway riders) use park-and-ride lots.  For these users, the system does
not provide full economic benefit due to reduction in household auto owner-
ship requirements which may be possible for other new users.
4.3.3  USE OF AUTO LEFT AT HOME
     To begin, recall that approximately 5,200 auto trips were diverted
to the busway each day, but that nearly 3,100 autos are parked in park-
and-ride lots each day.  Then, assume that the difference (2,100 autos)
represents a reasonable estimate of the number of autos left at home by
former auto commuters.   The November, 1974 survey inquired regarding
the use of these autos and found that 61 percent were idle during the day.
                                   129

-------
                              TABLE 4.5
                       MODE OF  ACCESS  FOR
                   SAN BERNARDINO BUSWAY USERS

Access Mode                       Number                Percent
Park and Ride                      3550                    55
Walk                              1500                    23
Dropped Off                       1100                    17
Feeder Bus                          350                   	5_
                   TOTAL          6500(a)                 100

Note:  (a)   Estimated total weekday ridership, February, 1975.
Source:   November, 1974 on-board survey  conducted by Bigelow-Crain Associates
         for  the Southern California Association of Governments.
                                   130

-------
his implies a limit for potential automobile ownership reduction of nearly
1,300 autos, a significant potential economic benefit directly attributable
to development of the San Bernardino busway.  This does not account for
new busway users who may have replaced a commuting automobile with one
in poorer repair for shorter access trips to the busway terminal.
     The remaining 800 autos left at home during the day were used for a
variety of trip purposes as summarized in Table 4.6.  This added usage
will serve to cancel some of the fuel savings and pollutant emissions
reduction achieved by diversion to the busway to the extent that these auto
trips were not previously made.
4.3.4  BUSWAY INFLUENCE ON RESIDENCE LOCATION CHOICE
     The pre-busway survey carried out in April, 1972  inquired regarding
the influence of bus service on residence location choice.  Interestingly,
the breakdown of responses are similar to those reported for the Shirley
busway.   Approximately 39 percent of the bus riders indicated that bus
service availability had influenced their choice.  About 19 percent responded
that it had slight influence, while the greatest number (46 percent)
answered that their residence location decision was based on non-transit
factors.
     No corresponding data was collected for persons changing residence
location following busway opening and development.
 Grain & Associates.
                                    131

-------
                             TABLE 4.6
              USE OF COMMUTING CAR LEFT  AT HOME
                                                 Percent
Not used                                           61
Another person takes It to work                      13
Shopping,  errands                                   17
School, university                                   5
Other                                               4
                        TOTAL                      100
                          132

-------
4.3.5  NET ENERGY CONSUMPTION SAVINGS
     Application of BUSWAY using input data as already described for 1975
automobile fleet characteristics generated an expected fuel savings dis-
tribution as shown in Figure 4.7 with an average value of about 4,300
gallons per day, and less than one percent probability of savings less than
2,200 gallons or more than 6,300 gallons of fuel.  Assuming a 40 percent
improvement in auto gas consumption rates, the fuel savings distribution
shifted with a lower mean savings of 2,100 gallons per day (see Figure
4.7).
     Recall that 800 autos left at home during the day are used for trips
previously made.  If this trip-making required an average of one gallon
per day per auto, the estimated 1975 fuel savings would be reduced by
approximately 20 percent to 2,500 gallons per day.  The estimated savings
with more efficient autos would be about 40 percent less or 1,300 gallons
per day.   This amount of fuel consumption represents additional travel
of between 12 and 22 miles per day, which appears to be a reasonable value
for induced trips.
     Figure 4.8  summarizes estimated auto and bus energy use components
'before1  and 'after' busway opening.  As was found for the Shirley busway,
the potential savings due to elimination of corridor auto trips is lowered
from approximately 8,200 gallons per day  to 3,500 gallons per day when
increased bus fuel consumption and gasoline used for system access and
induced auto trips are considered.
                                   133

-------















i
(0
m
0
Q
n
CT
2
t
UJ
ID
Z

(D

















(0
SAVING

J
Ul
IL
&
<
Q
0
F
a
Z
tfi
ol
\%
Ul
D Z
Ul j
£0
o ft
a x
a Q
1


q at oo i«.  2
                                                 cc O
                                                   I
                                                 >- 1—
                                            _o
                                                 ^ O
                                                 O
                                                 ~ CO
                                                 a: O
                                                   I
                                                 >- f—
3
3
in
d
n
o'
                   do Ainiavaoad
        134

-------
c Ava/siMcmvo) NoixdiAinsNoa ian=i
              135

-------
      Another aspect of potential fuel savings (and CO emissions reduction)
may overwhelm computed savings from diversion of auto trips to more effi-
cient transit vehicles.
      Traffic volumes on the San Bernardino freeway have been at or near
maximum capacity for the last ten years.   Volumes have ranged from about
140,000 to 170,000 vehicles per day in both directions.   This volume is
so high that speeds during the peak hours have been unstable with maximum
speeds being between 30 and 35 mph.  The  implication of this unstable flow
condition is that, if the busway were able to draw a sufficient number of
drivers from their cars, freeway speeds would increase sufficiently to
cause a significant increase in travel speed to the remaining auto users.
      To indicate the volume-speed relationships more clearly, Figure 4.9
shows approximate peak hour volume-speed  relationships for four lanes of
an eight-lane freeway.  The relationship  shows that higher speeds occur
when traffic volumes are low, and the speeds drop off sharply at higher
volume levels (6000 to 7000 vehicles per  hour).  At volumes higher than
7000 vehicles per hour, the speeds become increasingly unstable and, at
capacity, vary between zero and a few miles per hour (mph).  The San Ber-
nardino freeway peak hour volume is about 7500 vehicles, in the unstable
flow range between 30 and 35 mph.
      Assuming that one-half of the estimated 3800 auto trips diverted to
the busway occured in the peak hour, the  peak hour freeway volume would
be only 5600 vehicles resulting in an average speed increase to approxi-
mately 50 mph.  Referring to Figure 4.10  this  speed change would improve
average auto gasoline consumption by roughly 10 percent or about 0.7
                                   136

-------
         70
         60
      CO
      0
      LLJ
      HI
      Q.
      CO
      o

      I
      LU
      > 50

      LU
      ID
      <
      £T
      LU
     ^  40
     X
     O
     cr
     Q_
     Q-
        30
        20
                                              SAN BERNADINO

                                             FREEWAY AT THE'

                                                 END OF 1973
                         I
I
                             MAXIMUM
                             CAPACITY OF
                             FOUR LANES
                                                                      UNSTABLE

                                                                        FLOW
         3000          4000           5000          6000          7000          8000

             PEAK-HOUR TRAFFIC VOLUMES  FOR FOUR LANES OF AN EIGHT LANE FREEWAY
APPROXIMATE  VOLUME-SPEED  RELATIONSHIP
                                       137

-------
0  2
«
    o
    u

    in

    u.

    o
           ID
           h
           a
         *.<»>
a
to
o
                                                   _ o
                                                     .
                                                    0

                                                    Ol
0

ID

Q
          O1IIAI tJ3d SNO11VQ)

                       aisinosva oxnv
                           138

-------
gallons over the eleven mile corridor length.  For all remaining autos
during both the a.m. and p.m. peak hours only, a gasoline savings of 7,800
gallons would accrue.  This is nearly double the average amount computed
above without consideration of volume-speed relationships.
      However, the hypothesized increases in freeway speeds may affect the
route selection of other auto drivers.  Some drivers on more circuitous
routes paralleling the freeway are likely to switch to driving on the free-
way.  In this case, net speed increases may still occur on the freeway and
perhaps result on the circuitous routes as well, generating net gasoline
savings within the corridor.  In addition, the peak period may shorten
so that peak hour volumes remain approximately the same in response to im-
proved flow conditions as drivers reschedule their individual trip times
and additional fuel consumption savings do not occur.  No data was available
to confirm actual freeway conditions in the San Bernardino corridor.  The
potential magnitude of gasoline savings indicates that further studies should
be undertaken in this area.
4.3.6  Net Air Pollutant Emission Reduction
      Assuming unit emission factors representative of the 1975 national
automobile fleet and different driving cycle conditions within the corridor,
it is estimated that the San Bernardino busway development has reduced
corridor CO emissions by an average amount of 3.5 tons per day (see Figure
4.11) and HC emissions by approximately 240 kg.
       To place  these estimates  in  perspective,  the  transportation control
 strategy for the Los Angeles  region  promulgated  by  the  Environmental
                                    139

-------
   i
"3
9 (0
9 2
< o
   05
Z fn
1 n
111 "'
m
< o
(0 U
                                            _o
                                            r -co
                        CO
                        2:
                        O
                        I-


                        co
                        z
                        o

                        09
                        CO

                        £
                        UJ

                        O
                        U
                                               Q
                                               IJJ
                                               o:
                                               Q
              s
                 CO
                 o"
CD
0
CM
d
                            do Ainiavaoya
                     140

-------
Protection Agency calls for a reduction of 550 tons of CO emissions per day
through VMT reduction measures in order to meet 1977 ambient air quality
standards.  The contribution of the San Bernardino busway towards achieve-
ment of these standards would be less than one percent.  The stringent
measures required to meet standards in the Los Angeles area make plan imple-
mentation unlikely.   Studies assessing the reasonableness of the Los Angeles
1977 implementation plan considered the maximum transit share (without auto
restrictive measures) to be about 80 tons for the region.  In this context,
the busway is contributing less than four percent towards attainment.
      Figure 4.12 illustrates that the estimated air pollutant emissions
reductions actually include worsened conditions in suburban areas adjacent
to the El Monte terminal.  For CO emissions, a net reduction of 4.4 tons
results along the busway and in downtown, but an increase of approximately
0.9 tons occurs in the outer corridor suburbs.  For HC emissions, a similar
pattern results with an increase of approximately 40 kg in the El Monte
terminal vicinity.  However, the polluting effects of HC occur with formation
of regional photochemical oxidants and localized distribution effects are
not significant.
                                    141

-------
01
  ID
  Ul
^ o
ii


M
m o

°z"
5°

II
Z h
a D
ui m
ID g
2 H
i fs
5 Q
          142

-------
                              CHAPTER  5
               NEW CORRIDOR RAIL TRANSIT SERVICE

5.1  INTRODUCTION
     This report chapter considers the potential  energy consumption  savings
and air quality improvements which may be derived from new  corridor  rail
transit service.  At the present time, major rail  transit system development
projects are underway in several large U. S. cities  including Atlanta,
Baltimore, and Washington, D. C.  Two case studies were selected for review
in this study phase—Bay Area Rapid Transit (BART) in  the San Francisco Bay
area and the Lindenwold High-Speed commuter rail  line  in the Philadelphia
metropolitan area.  Both are modern rail  transit  service examples, represent-
ing the current state-of-the-art in high technology  and performance  standards.
     Before describing data and analyses carried  out regarding  a comparison
of the energy and air pollutant emission characteristics of rail transit,
some background information on two key aspects  relating to  comparisons with
other transit modes should be introduced.  First,  any  comparison of  relative
energy usage must differentiate between petroleum and  non-petroleum  energy
consumption.  Electric energy may be developed  from  either  petroleum sources
or hydroelectric, nuclear, coal, geothermal, and  other non-petroleum sources.
Hence, electrically-powered transit systems offer a  greater degree of ver-
satility in terms of energy source than either  gasoline or  diesel powered
buses.  Second, there are substantial inefficiencies associated with the
generation and transmission of electrical power.   To provide a  true  basis
                                     143

-------
for comparison, it is necessary to consider requirements for 'source1
energy (i.e., the energy equivalents in BTU's of the primary fuels).   While
estimates of these losses vary, source energy consumption may be expected
to be 3.0 - 3.5 times system energy requirements.
5.1.1  METHODOLOGY
     For analysis of the Lindenwold High-Speed line, a model (CORAIL)
similar to that used for busway studies was employed.  It invoked most of
assumptions regarding corridor and service characteristics already outlined
for BUSWAY in Chapter 4, although specification of feeder bus access  to the
line haul system was permitted.  Key factors affecting potential energy savings
and air pollutant emissions reduction were incorporated as summarized  in
Table 5.1.
     It was found that only preliminary data regarding BART was available
without undertaking more extensive data collection than possible for  this
study.  Furthermore, CORAIL was developed for corridor applications and
could not be used for BART system analysis.  Some  data tabulations for key
factors have been included for comparison with corresponding data for other
transit improvement projects.
5.2  BAY AREA RAPID TRANSIT (BART)
     BART has achieved considerable publicity, both positive and negative,
since its opening in September, 1972.  The system  encompasses 75 miles of
modern, high-speed rail transit serving San Francisco, Oakland, and other
east bay communities in the San Francisco Bay area.
                                      144

-------
             TABLE  5.1
FACTORS AFFECTING ENERGY CONSUMPTION
 AND AIR POLLUTANT EMISSION CHANGES
    FOR NEW RAIL TRANSIT SERVICE
Q.
t— •—
>- Z
1/5 (/I
Z Q:
< UJ
ec. o
>— *-
cc
CORRIDOR
SERVICE
DESCRIPTION
AUTO DIVERSION
Qi t-O
UJ ff
ST on tu
o: o a
O ca >—
Uu CC
UJ
a
UJ
2:
02 i/>
UJ J—
h- O
11
OUTPUT
NEW RIDERSHIP
PEAK PERIOD LOAD FACTOR
OFF-PEAK LOAD FACTOR
PREVIOUS TRAVEL MODE INCLUDING
DIVERSION FROM CARPCOLS
LENGTH OF LINE
CORRIDOR TRIBUTARY AREA
STATION SPACING
RAIL TRANSIT UNIT ENERGY CONSUMPTION
RAIL TRANSIT UNIT EMISSIONS
NUMBER OF AUTOS DIVERTED
DIVERTED AUTO TRIP LENGTH
TIME OF DAY OF DIVERTED AUTOS
REGIONAL LOCATION OF OIVERTEO AUTO TRIPS
DIVERTED AUTO UNIT GASOLINE CONSUMPTION
DIVERTED AUTO UNIT EMISSIONS
NUMBER OF TRIPS DIVERTED FROM
FORMER BUS SERVICE
FORMER BUS SERVICE LOAD FACTOR
BUS UNIT FUEL CONSUMPTION
SUS UNIT EMISSIONS
L NUMBER OF PARK-AND-RIDE PASSENGERS
NUMBER OF KISS-AMD-RIDE PASSENGERS
NUMBER OF FEEDER BUS PASSENGERS
UNIT FUEL CONSUMPTION FOR
AUTO ACCESS TRIPS
UNIT EMISSIONS FOR AUTO ACCESS TRIPS
FEEDER BUS UNIT FUEL CONSUMPTION
FEEDER BUS UNIT EMISSIONS
ACCESS TRIP LENGTH
RESIDENCE LOCATION CHANGE
REDUCED AUTO OWNERSHIP
REDUCED TRAFFIC CONGESTION
OTHER USE OF AUTO
NET 1975 CO EMISSIONS REDUCTION
NET 1975 HC EMISSIONS REDUCTION
NET 1975 FUEL SAVINGS
NET FUEL SAVINGS WITH IMPROVED AU'OS



0



1

1

1



,
1


1
1
1







1





1


1
t
1
1
1

1
1

1


,



1
1
1




1


1













1




1
1

1
1
1



1
1
1
(









1



1
1
1


1
1




1



1
1
t




1








1





1

1



1
1
,
t
1
1
1
t
1
,
1
1
t
1





















1
t







t
t










































1
1
1
1
                    145

-------
     BART is currently the subject of an intensive impacts analysis research
program including a detailed assessment of regional energy consumption and
air quality implications.  In this section, selected data items describing
BART ridership characteristics including estimated auto diversion and system
access mode are presented.  It should be pointed out that BART is still moving
towards full system operation, and the following data tabulations should be
treated as preliminary in this regard.
5.2.1  CORRIDOR MODE CHOICE AND DIVERSION TO BART
     BART initiated transbay service into San Francisco in September, 1974
with an immediate effect on transbay corridor trip patterns.  For the peak
period, a comparison of transit passenger counts and San Francisco Bay Bridge
traffic volumes  (taken prior to BART service introduction and following it
in October 1974) indicates that:
     t  total  peak period corridor trips have remained approximately
        constant;
     •  corridor transit mode choice has increased from 46.5 percent to
        48.8 percent using BART and bus service; and
     t  the increased total transit usage may be explained by reduced
        average auto occupancy, although both changes are small  and
        may be due to normal  data fluctuations.
Table 5.2 lists the number of trips transported  by each mode during the two-
hour a.m.  peak period.  For all day, it appears  that BART may be generating
new trips or absorbing new trips which would have mode in any case:
 Institute of Transportation and Traffic Engineering, University of
 California at Berkeley.
                                     146

-------
                       +-> +J
                        C -r-
                        Ol (/>
                        U C
                        S- (O
                        CO i-
                       Q. t-
       00

       CO
O-
co

UJ
Q
O
                           (O
                          CQ
                                 I—     CM
                                 CO     r—
                                 CO     i—
CO
o
                                 CM     CO
                                 CM     i—
             ro
CM
LT>

UJ
—1
CQ
^_



OQ
h—
CO

UJ
CD
CO

1 	 1
UJ
0_

N,/*

CM
CT1
CM
A
cn

tn
^
^
CM
c




LO
1

                                                                  S-
                                                                  
                                                                  C
                                                                  LU

                                                                  O
                                                                  <0 CO
                                                                  S- r-
                                                                  r- CO
                                                                     ^.
                                                                  T3 i.
                                                                  C CO
                                                                  1T3 CQ

                                              Ol
                                              o>
                                              C
                                              03
                                                                  O R3
                                                                  •»-> C
                                                                  S- S-
                                                                  o o
                                 C i—
                                 (O  fO
                                 S- O
                                 <4-  O
                                 o

                                 CO -(->
                                 •»-> •!-
                                 3  «rt
                                 4->  S.
                                 •r-  (U
                                 •M  >
                                 «/> 'r-
                                 C  C
       S-
  "    CO
•—     ^3
•r-     O
 S-     4->
 O.    U
«C     O
                                              C
                                              CO
                                              O
                                              S_
                                              CO
                                              Q.
(U
o
                              147

-------
     •  the total number of daily corridor trips has increased by
        about eight percent with automobile trips remaining approximately
        constant;
     •  corridor transit mode choice has increased from 22.1 percent to
        27.1 percent using BART and bus service; and
     •  new BART ridership may be accounted for as being about three-
        fifths diverted from bus service, and two-fifths trips not made
        six months previously.
     The split of daily corridor trips by travel mode is summarized in
Table 5.3.
5.2.2  PREVIOUS TRAVEL MODE
     In addition to the inferences derived from the aggregate transbay corridor
statistics which suggested no BART diversion from peak period automobile usage,
the results of an onboard rider survey conducted in May, 1973 on the Richmond-
Fremont BART line (prior to opening of east-west service including transbay
operations) provides data regarding the prior mode of travel of BART users.
These results, summarized in Table 5.4, show that approximately 39 percent
of BART trips represent former automobile trips, and another 13 percent of
the surveyed riders indicated that the trip had not been previously taken.
5.2.3  MODE OF BART ACCESS
     As has already been noted, the degree of potential  energy consumption
savings and air pollutant emissions reduction due to the introduction of new
corridor transit service may be effectively lowered by high auto use for
system access.  For BART's east bay line, nearly one-half of BART's
                                     148

-------
                              in
                              <=.
                              

UJ
_J
co
      D_
      CO
            oo
      LU   -ZL
      ca   o
      o   —'
>~   ^
CQ   —'
00

UJ

o

00

<_>
sr

ce:
u_


<
CO
 0)
 en

 CO
 (/I

 (O
a.
                              t/i

                              co
                        g
                        CQ
                        3
                                    r-.
                                    IO
            o
            OQ
                                    to
                                    o
                              cr>    in
                              co    CM
                              o    o    •—
                                          +
                              ro
                              CM
                  CO
                  CM
                              00
                              O
                        <=C
                                    ro
                                                                   O)
                                                                   c
                                                             
                                                                  +-> •!-
                                                                   3  Ul
                                                             00 •!-
                                                             c  c
                              •r-
                               J-
                               a.
                              <
                                           1-

                                           
-------
                          TABLE  5.4
                  PREVIOUS TRAVEL MODE
                OF BART PASSENGERS, 1973
Previous  Mode                          Percent
Auto Driver                              ,«
Auto Passenger                           ,,-
Bus                                     2g
Other                                    5
Trip not  made
                         TOTAL          TOO
Source:   Office of Research, Bay Area Rapid Transit District.
                          150

-------
passengers traveled to stations as an auto driver (35 percent) or were dropped
off (13 percent).  Only 14 percent used feeder bus service for system access,
while the remainder arrived as auto passengers (11 percent), by walking
(24 percent) or other means as summarized in Table 5.5.  Further analysis of
survey results showed that the mode of system access varied substantially
from station to station along the line.  At outlying suburban stations with
adequate parking, the level of auto access was higher as would be expected.
For example, over three-quarters of the passengers arriving at the Fremont
station located at the southern terminus of the line drove automobiles or
were dropped off.
5.3  LINDENWOLD RAPID TRANSIT
     Since February 1969, the Philadelphia-Lindenwold line has provided high-
speed access to the commercial centers of Camden and Philadelphia for commuters
from the densely populated suburbs in South New Jersey.   The 14.5 mile line
extends from Lindenwold, New Jersey to Camden and over the Franklin Bridge
to Philadelphia, and includes several advanced train technology elements
including automatic train operations and automated fare collection.
     The Federal Energy Administration, Office of Transportation Research
is currently sponsoring a study intended to examine the energy consumption
characteristics of the Lindenwold line including its potential impact assuming
the line was not in operation.  This study is scheduled for completion in
May, 1975.
     Selected data regarding Lindenwold line ridership characteristics which
have a significant influence on the magnitude of potential energy consumption
                                      151

-------
Mode
Auto Driver
Auto Passenger
Dropped Off
Bus
Walk
Bicycle
Other
          TABLE 5,5
MODE  OF BART ACCESS, 1973

                       Percent
                        35
                        11
                        13
                        14
                        24
                          2
                        _J	
       TOTAL            100
Source:  Office  of Research, Bay Area Rapid  Transit District.
                            152

-------
savings and air pollutant emissions reduction follows.  This data is based
on a series of on-board rider surveys performed by the Delaware River Port
Authority  (DRPA) in 1970, one year following service introduction.
5.3.1  CORRIDOR MODE CHOICE AND DIVERSION TO THE LINDENWOLD LINE
     DRPA passenger survey indicated that the new Lindenwold service diverted
approximately 8,500 daily automobile trips from bridge river crossings in
1970, representing about one-quarter of the daily system ridership of 30,000
passengers (see Table 5.6).  Note that nearly twice as many of the new rider-
ships were diverted from other transit modes.  Examination of bridge traffic
trend data does not confirm the decrease in auto trips expected from survey
findings.
5.3.2  MODE OF ACCESS TO THE LINDENWOLD LINE
     The 1970 on-board surveys found that ninety percent of all Lindenwold
riders reached the system via auto.   Of the auto access total, nearly three-
quarters involved parking at station locations while the remainder were
dropped off.   Survey results are tabulated in Table 5.7.
5.3.3  ACCESS AND LINE HAUL TRIP LENGTHS
     Working  with the same survey results and passenger count data,  Boyce,
et al., estimated that the average airline distance for station access trips
was 2-5 miles, depending on station  location along the length of the line .
In comparison, the average transit line haul trip length was estimated to
be approximately nine miles.
 Boyce, D. E., et al.   Impact of Access Distance and Parking Availability
 on Suburban Rapid Transit Station Choice.  Prepared for the Office of the
 Secretary, U. S. Department of Transportation, PB 220 694, November, 1972.
                                    153

-------
                           TABLE 5.6
                    PREVIOUS  TRAVEL  MODE
              OF LINDENWOLD  PASSENGERS, 1970
Previous Mode                   Number                 Percent
Auto Driver                     8,500                   28
Auto Passenger                  3,500                   12
Bus                           10,800                   36
Commuter Train                  3,300                   11
Trip not made                   3,900                   13
                  TOTAL        30,000                   100
                            154

-------
                            TABLE 5.7
                ESTIMATED MODE OF ACCESS  FOR
                 LINDENWOLD LINE  USERS, 1970

Mode                       Number               Percent
Park and Ride               10,100                 67
Dropped Off                  3,400                 23
Feeder Bus                    750                  5
Walk                         750                  5
                 TOTAL      15,000                100
                         155

-------
5.3.4  NET ENERGY CONSUMPTION SAVINGS1
     From simulation analysis using described data inputs, it was estimated
that the Lindenwold line is producing an average peak period energy savings
equivalent to approximately 800 gallons with four percent probability of
increased energy consumption when possible variations in diverted auto trip
and gasoline consumption characteristics are considered (see Figure 5.1).
     If a forty percent improvement in automobile gasoline consumption is
assumed, the estimated probability of increased overall energy use is approxi-
mately 30 percent with average peak period savings of only 200 gallons (see
Figure 5.2).
     Figure 5.3 summarizes energy consumption 'before' and 'after' Lindenwold
line development, as was presented in the preceding chapter for the Shirley
Highway and San Bernardino busways.  In this case, auto gasoline consumption
for system access is approximately one-half of the estimated amount of gasoline
saved by autos diverted from commuting.  If all Lindenwold line riders were
assumed to be former auto commuters, a net peak period energy savings of
approximately 5,700 gallons would be obtained.  Since only 40 percent of new
riders previously commuted by automobile, the resultant energy savings are
substantially lower.
5.3.5  NET AIR POLLUTANT EMISSIONS REDUCTION
     Assuming 1975 national automobile fleet characteristics, model results
indicate peak period CO emissions reduced by approximately 1.3 tons and HC
 Energy consumption savings are expressed in equivalent gallons of gasoline
 at the source (one gallon equals approximately 13 KWH of electrical energy).
                                       156

-------
r 2
10 Q
UMPT
CO
0
u
J
UJ
u.

2
Q
U
Q
UJ
(£





Ul
J
Q
UJ
UJ
a
CO
0
i
Q
0
Ul
D
2
j



a















u tf












U-LOIMS 1













01






§
Q
* m
a
w
- n o

a
n
i a
IM IIM FUEU COIMBl
0
\
111
L 0 •
- r









MG PEAK PERIOD
mm
a
3


                                   - 01
     =ia
157

-------
(U
ui









2
MPTIO
COIMSU
J
Ul
U.
2
IMOIJLG
Q
Ul
t£


LIIME
D
Ul
Ul
a
(0

(3
I
Q
a
Ul
Q
j

"
0
"5
a
3
(I)
0
u
0)
*•
0
(0
•a
Q)
I
a

i i
^ CD











ID «
(0
0
J
(D
Q









tu
3
OIM CTHOUSA
P
a
D
10
-w 8
j
Ul
- r
Q






n
Q
PER
IIMS PEAK
°


       =10
                                        (U
                                        I
158

-------
n
id
     o


     0   n
     u.   u
     ui   >
     ID   0
ui
LU
          S
          0
a   ffi
to   a

i   *
3   4

Jg

3*
0   uj

§   LL
2   <
UI
     j   <
a
u

S
       a
       ui
                              i
                              3
                             .
                              Q

                              0
                              0
                              O
                                                                               o
                                                                          to

                                                                          <
                                                                          a:
                                                                   an
                                                                               £
                 u.
                 ui
                 o
                                                                               a

                                                                               5
                    o
                    o
                    o
                                           8
                                           8
                                       159

-------
emissions reduced by an estimated TOO kg due to Lindenwold rail  transit
service.  Figure 5.4 shows the probability distribution function determined
for CO emissions reduction.

5.4  RAIL TRANSIT AIR POLLUTANT EMISSIONS
     Electric vehicles have no direct emissions.  Their contribution to
air pollution stems from additional  loads placed upon the power plants serv-
ing the system.  Furthermore, power  plants do not contribute to the CO or HC
problem in significant amounts.  However, power plants may be a major source
of particulates (TSP), and sulfur and nitrogen compound pollutants.
     Using typical conversions and emission factors published by the Environ-
mental Protection Agency, Table 5.8  summarizes the potential contribution of
total transit system operations energy for BART and the Lindenwold line on
the level of estimated 1975 regional TSP and S02 pollutant emissions.  In
each case, the impact of transit energy utilization is negligible.
                                     160

-------
Pf
in I
0
5)
(0
i
Ul
0
u
2

EDUCTION
a
Ul
2
j
Q
111
Ul
a
(0

(3
I

D
i
UJ
Q
2
j








































- in

„
0)
• tf 2
(0
1
3
MISSION
PERIOD!
Ul
0 4
1 1
i a
REDUCTION IIM C
CMORNINQ PE

00
(0
IU.
                      AJ.mavaotdd
             161

-------
    03 4->
    -  0)
   -^  O
    1/5  S.
    C  O)
    O 0.
CO

o
                                                 o
                                                 o
CO
zr
o


•—'
CD   
cr>   z:

-   1
Q
LU   CD
I—   t—
  O
O  «/!•!-
CO t-  CT|
    £  (U
   UJ QC
    (O  (/I
    3  C
    C  (O
    C  S-
   <=C H-
    S_
    ITS •»->
    (U C
   >- 0)
   ^-. U
    V) S_
    C <1>
    O Q.
    t/1 (O
    C-— •
    o
Q_ -i- C
l/> t/> O
    •t—
    E
       a:
    ra (/>
    3 C
    c to
    c s_
CO
cr>
CM
  ^
cr>
CO
             o
co
                   oo
                                                 evi
                                                 vo
                                                 CO
                   a>
       LO
       ro
                                     o

                                     CD
             in
             CO
       o

       o
       o
       CJ)
       oo
                                                                           1.
                                                                           O
                                                                           a.
                                                                           
                                ra
                                z

                                CVI
                                             o
                                             c
                                             
                                o

                                D-
                                                                            to
                                                                           +-1
                                                                            c
                                                                            O)


                                                                            o
                                                                            S-
 > O
 c ai-i-
 c s- P—
 «£ QJ i—
    c T-
    uj  s
                                                                              to
                                                                             •r—
                                                                              C
                                                                              S-
                                                                              o
   4-
   o

   O) «*
   -i-> r~-
   ra CTi
  • •)-> r—
to co

§1.+:
•p- O  tO
•(-> 4-  3
ro     cn
S_ C  3
a> o <
Q.-I-
O 4->   "
   3 =
                                      •!->  S-
                                       CO 4J
                                                   (/I •!-

                                                     TJ
                                             O)
                                                  oo  c
                                                      o
                                                                                                 (O
                                                                                              C 4J
                                                                                             •r- S-
                                                                                                 o
                                                                                              c a.
                                                                                              o «/>
                                                                                             •i- C
                                                                                             -!-> 10
                                                                                              3 t.
                                                                                             J3 t—
                                                                                             •r-
                                                                                              S- C
                                                                                             •(-> rs
                                                                                              to -Q
                                                                                             •p- S-
                                                                                             -o rs

                                                                                              to t-
                                                                                              ai o
                                                                                              o
                                                   i.
                                                   3
                                                         CO
                                                                           c
                                                                           o
                                                                           •r-     3  Ol -P-
                                                                               to
                                                                               (U
                                                                            <+- o

                                                                            S- 3
                                                                            o o
                                                   o
                                                                                           04
                                                                                        c o
                                                                                        ra
                                                                                        s-  »
                                                                                     CO (— > U P—
                                                                                     cn-p-  .a o
                                                                            o]o

                                                                             - S-
                                                   p— O
                                                   res OJ
                                                   CU P-
                                                   z cu
 CO
 co
 CsJ
                                                                              «a-    cr>
                                                                              r-» w> p—
                                                                              en cu
                                                                              i— -Ml/5
                                                                                  ra cu
                                                                                • E E
                                                                               CU •!- 3
                                                                             1  C •(-> CO
                                                                             i  3  to
                                                                              o uj <:
       a.

       ^"H
       O ra

       cu cu
       co O

       ^ z
       >>CQ
       o>
       S- >!
       CU -Q
       c
O -P-
CO •!->
    to
>>'P-
cn j_
i-  
c  o
CU  ra

o  re
•p- -C
s- o
•)->
o   •>
cu  c
p-  O
O) -i-
    -u
O)  ra
0)-t->
rO  J-
S-  O
O)  Q-
>  to
                                              «  03
                                             >>£
                                             P— •!-
                                             10 •»->
                                             a>  to
                                                                                        p- S- CTi
                                                                                         ra r- r-

                                                                                         O H-  -

                                                                                        '^°^

                                                                                         S^S
                                                                                            CU  •»
                                                                                         to E  to
                                                                                         CU -P  E
                                                                                         E t.  
                                                                                         (o cu  >>
                                                                                        
-------
                               APPENDIX A
                 METHODOLOGY DESCRIPTION AND INPUT DATA

     Three computer models were developed in the course of this  study to
estimate the impact on total energy consumption and selected air pollutant
emissions of transit alternatives:
     •  Bus service improvements such as increased bus miles of
        operation, fare reductions, and demand actuated service;
     •  Corridor service in which buses circulate through
        neighborhoods collecting passengers, run express on a
        reserved facility, and then distribute passengers in a
        city's central business district; and
     •  Corridor transit service in which the line haul trip is
        via modern rapid rail to and from the CBD with passenger
        access by bus, auto, and walking.
     A description of the model methodology follows including review of the
data used for each of the eight case studies of new/improved transit services.

INPUT DATA
     All three models were set up in such a way that they would  accept input
data in either of two forms:
     • Fixed value in cases where the available information was  such
       that no uncertainty existed as to the actual number to be used
       in the mode.  This was the case, for example, for before  and
                                     163

-------
        after annual bus miles of operation for the Atlanta areawide
       bus service improvements case study.  This data was readily
       available from routine system reports.
     • Random value in cases where the actual value for a case study
       variable was not known with certainty.  This was used, for
       example, for automobile gas consumption and air pollutant
       emission rates to reflect fleet characteristics.
     Where no direct survey or other data was available, random values were
generated from a distribution function defined by probable upper and lower
bounds, and by the expected modal  value.  The upper and lower bound and
modal values were obtained using available data from studies for other areas
having similar characteristics and from other secondary sources.  For example,
former auto trip lengths for new transit users, when not available, were
obtained by using empirical relationships between trip length and metropolitan
population from research studies .  The probability distribution would be
established by using this point estimate as the modal  value with a given
percentage then subtracted and added to obtain lower and upper bound values
of the distribution.
     The lower bound.modal, and upper bound values served to define an
approximate beta probability distribution function, from which the values
                                           2
used in each model  interaction were sampled .   Figure A.I illustrates a
typical probability distribution function defined in this manner.  The
values of X-,, XM and X2 are the lower bound,modal, and upper bound values
 Alan M. Voorhees & Associates.   "Factors and Trends in Trip Lengths,"
 National Cooperative Highway Research Program Report No.  48.   1968.
2
 Pei, R. Y. and I. F. Kan.   "A decision Aid to Transportation  Planners."
 Paper presented at the 41st meeting of ORSA, New Orleans.
                                      164

-------
     P(x)
f(x)
                                     AM

                         FIGURE A.I
                   TYPICAL BETA DISTRIBUTION
                                165

-------
respectively, of variable x.  For any value of x, there is an associated
probability of occurrence P(x).  All values in the interval are possible
but those near the modal value have largest probability, whereas those
further from the mode have smaller probabilities of occurrence.  Note that
the distribution is not necessarily symmetrical and may be skewed as required,
     Once the fixed input variables had been determined, they entered all
calculations as a single value.  The random variables however, would take  a
different value for each iteration, but always according the established
probability distribution function.  Iterations results were stored to
develop probability functions for output variables.  All models were run
for two hundred iterations for each case study.
OUTPUT DATA
     As a result of the uncertainties associated with the random variables
used as input, each iteration would give a different result for the savings
in fuel consumption, and the change in total CO and HC emissions.  At the
end of the iterative process, the stored results would be used to deter-
mine output variable distributions.
AREAWIDE BUS SERVICE MODEL
     The areawide model estimates the impact of improvements in regional
bus service such as expanded annual busmiles, fare changes, and new demand
actuated service.  It operates in several steps as follows:
     (1)  Define the lower bound modal and upper bound values of input
          random variable, namely:
                                     166

-------
     • percentage of previous auto drivers
     • auto trip  length
     • unit auto  fuel  consumption
     • unit auto  pollutants  emission
     • unit bus fuel  consumption
     • unit bus pollutant emission
(2)   Read the  breakdown of former  auto  trips  into  suburban, corridor,
     and urban, as percentage of total  trip  length.
(3)   Read the  transit ridership and the busmiles of  travel for  the
     "before"  and "after" conditions.
(4)   Read the  average trip length  of diverted autos.   If  unknown,
     use city  population to  obtain this value as:
              ATL = 0.46 P(0'19)                           , and
     then establish the range of trip lengths as 80  percent and 120
     percent of ATL for the  lower  and upper  bound  values,  respectively.
(5)   Obtain the change in transit  ridership  and in bus miles of travel.
(6)   Estimate  effect on ridership  due to changes in  fare  (if any).
(7)   Start iterative process for a total  of  200 iterations:
     • sample  all beta distributions to obtain one value  for each
       random  variable
     • estimate auto VMT reduction
     • estimate cold start effect
     • estimate fuel  savings and reductions  in pollutants
     • store results,  and start a  new iteration.
                               167

-------
     (8)  At the end of the iterations,  determine the mean and  probability
          distribution of iteration results.
     (9)  Print the results.
     The areawide model was applied to five case studies—Atlanta,  Washington,
D. C., San Diego (two cases), and Orange County.  Tables A.I  through A.5
show the values used in each  case study  to define the beta distribution  of
each input variable.  Figure  A.2 shows an example of the output results  for
the reduction of CO emissions in the Atlanta  case study.
     For the San Diego case studies, parametric analysis of diverted auto
trip lengths was carried out.  Fixed values for this variable of 2.5, 5.0 and
7.5 miles per former auto trip were assumed.   Tables A.6 and  A.7 list data
values used in these parametric studies.
CORRIDOR BUSWAY MODEL
     The Busway model assumes a corridor bus  system operating with  residential
passenger collection, express line haul  by bus on separate right-of-way,
and passenger transfer and distribution  in the central  business district
of the city.  It has the following general steps:
     (1)  Define the low, modal  and high values for each of the following
          random variables:
          • unit auto fuel consumption
          • unit bus fuel consumption (local  service)
          • unit bus fuel consumption (on busway)
          • unit auto pollutant rates for suburban, highway and urban
            driving cycles
                                     168

-------
      a
      o
      a.


      O
      2
co
<













S-
r~

T3
CU
X
•r—
U_












CU
-O
 u
to «3
3 a;
CQ ^*,
•=• o.












o
CO
CO
<~o
CO
CM
O
CT>
O
un
CO
CO
rO
-o
V
CU
CO
^


a.
CO
cu
-a
•r-
C£.
•r—
CO
C
It)
s_
1—


tn



CM




o
CM














*"""^
65
a>

10



i-O
in













•*~ ^
t/)
cu
•r-
"^^-
^T
CD
C
CU
a.
t—

o
.
CO
o

o
un



o
*











CO
o

-^

^

^3
CO

CU

_J
o:
c
o
•r-
a.
5
co
O •!-
(_) g
CU 'fO
rj en
1 1. — ^
































                                            169

-------
CM
CQ
      O
      Q.
      (J
      UJ
      on
     O
     UJ
          CI3













e
CO
+J
^
0)
5.
O
OJ
CO







O)
3
TJ
0
~j2
cS

.c
x
OJ
-a
o
s:


3
o
— I

_^
CD
X
Lu

-o
O)
X
u_








cu
IB
m












LO
a>
«^-
..^
LO
LO
00
VO



CO
ai

i
on
o
X
CO
co ra
3 aj
ry^ ^_
1— S_
s: ai
> CL












O
0
00
LO
LO

r-
CO
co
CO
ra
-a
a>
QJ
3

Q.
•r-
.C
co
O)
-o
CO
(13
H-


LO


CM
^J-



O
CM













£?
cu
I t
(T3
a:
version
Q


Crt
00
•
P-*.



CO
UD













^~*»
CO
OJ
is
_£.
cn
c
-j
o.
1 —

o
co
o
o
•
LO
lv-


o
1^,












CO
o

3

fO
CC.
C
o
a.
3
CO
0 ^
0 E
U- < 	
CM
CO
O
co
CM
•
O



*
<~>









CO
 	 ^






























                                                     170

-------
ro

-------
      LU
      O
      o


      a:
      o   a.
      o
«3-    Q   t-i
      «   I—
<    30

LU    LU
_1    CU   O
CQ    ^   O
*?£         11 i
r—    Z   >—i
      >-.   Q

      Q   Z
      CO
      LU













i-

0
-a
ro
CC

.C
cn
3:
0)
-o
o



o
1




-^
cu
X
LL.


-o
cu
X
•r-
U_








01
jQ
 C3.












O
O
VO

CO
o
o
o
co
A
CM
00

oo
ro
Ol
ai
s


a.
00
QJ
•t—
00
c
ro
S_
1—


LO

CM



O
CM



















• — "•
6-5
O)
ro
o;
version
•r-
Q


CO
00

O
00



3


















'•^
oo
QJ
i
-C
01
c
cu
_J
Q.
H-


O
CO
o
o
LO



o

















o

^
,,£
a;
c*.
c
o
•r-
00 ^^
o a\
C_J 	 ,

to
CM
CO
CJ
CM



C>
r—

















00
O)
00
3
CQ
OJ
ro
C
o
00
00 - — -
E 5
LU \
0 CTi
CJ - —
OO
0
0

 E
CU 'ra
3 Ol
LL. 	



































                                                         172

-------
      Q
      O



      D;
      o
      o
CO
          UJ
      Q
      UJ












OJ
*j
"^
OJ
i.
0
t-
O)
CQ





O)
3
fO
E
O
T3
C
ro
CtL


-C
d
•r-
^
s:  Q.












O
O
o
^r
CM
O




LO
ro
-a
^£

O)

"-
D.
!S
1/1
Ol
•o
•r—
I/I
c
rd
S_
1 —











in o in
CM in r»-














s«
**~^
OJ
I <
C£.
Diversion

in o ur
r>. in cv
CM in oc
in o tr
CM in r*



O OC
CM «a- ic


















i/i
OJ
r~
•r-
-C
CD
C
OJ
_J
Q.
S-
1 —


o
PO
o
o
in
^


0
^?















OT
O
4-J
3

CU
ro
Ct
C
Q
•r^
If
0 _

in
CM
ro
2
CM


°l
d
















0)
l/l
CQ

5
Qi
C
o
•r—
LO - — •
•r— T—
S £
UJ -^.
s_
O CT>
(_i -—
CO
o
d
en
o
o


m
O
d











0
4-*
3
^C

 £
LL. . 	 .
CM
ro
0
oo
CM
d


^
*—m
d











i/i
i/i
3
03


-------
                                                                                                                Q
                                                                                                                (U
                                                                                                                    I
                                                                                                                    0

                                                                                                                    to
                                                                                                                4  2
                                                                                                                r  0
                                                                                                                    3)
                                                                                                                    2

                                                                                                                    ui

                                                                                                                ru  g
                                                                                                                     0

                                                                                                                     \
                                                                                                                     Q
                                                                                                                     111
                                                                                                                     EC
                                                                                                                .
                                                                                                                ID
u
              q
              r
q
Q
10

Q
OJ

0
                                                     174

-------
             >-
             <
        —I   «C
        LU
        S   «->
        O   i—i
        "Si   C£.
       CL.
       O
     <
     Q.
co
LU   a.
a:   H-<
«C   Qi
            O
       O  >—i
       LU  |—
       OO  O
       =>  Z3
            O
       OO  LU
       LU  C£.
             az.
             u.
             o
             LU
             *—4
             a
             zr
             oo














01
•*~^

£
O
C
fO

r-
CT>
"
CU
o
2:




3
0
	 1


T3
CU
X
•r*
U-
-a
cu
X
•r*
U-








cu
rO
•r—
na














^
A
O
0
O

A
OO



„ 	 ^
00
cu

i
OO
o
X
oo TO
3 CU
CO >,
> a.













o
o
OO
t\
CM
OO
0
LO

A
p^
LO
OO
fO
TJ
cu
cu
3

Q.
0)
•i—
Q:
c
03
S.
1—


LO


CM





O
CM















3-S
(1)
OL
Diversion













LO O LO
CM LO IV.











*'— ^
VI
cu
i
^
en
c
cu
	 i
Q.
t—

O
%
OO
o
o
LO




o
^:













o

•=£
O)
fD
C
0
00
to ,-^
UJ -^

LO
CM
OO
•
O
CM




CT>
CO













00
CU
00
3
CO
cu
fa
C£.
c
0
•(—
I/) - — •
'i 'i
LJJ ^-,
i.
O O^
CO
o
o

en
o
o



LO
LO
o
o










00
0
4-*
3
OJ
4_)
Oi
c:
o
•r—
CL
00
5 '£
3 CTI
1 1 V 	 ,
CM
OO
O

OO
CM
C>




^
0










00
l/l
3
OO
 e
o> 'ro
3 CT>
1 1 ^^ •































                                                            175

-------
co
      o  °
      i  5
      <->
      uj  a.

      2r  "•'
      













P-.
A
r>.

«3-
«3-
^-

O



^_^
i/i

QJ
^
'i

CO
o
X
l/>
01 S_
tO (O
3 OJ
03 >,
h- U
2T 0)
> Q.












O
O

>,
T3
-a
J4
O)
OJ
3

Q.
-C
(/I
S-
O)
T3
a:
+j
i/i
c
ro
s_
I—

LO
r-~

CM
^3-




0
CM

















S3
"^"^

QJ
-(->
m
C£
Diversion












LO O Lf
CM LO 1^













^"*-
Ul
QJ
r—
sE
JZ
+J
CD
C
(U
	 i
a.
t.
i —

o
co
0
o
Lf)
r»x




O
r^
«*














to
O
j->
3

IT3
ct:
c
o
I/)
I/I -— -
E E
UJ \
S-
o ai
C_) —
§
C5

CTl
pv.
0
cp


U2
LO
O
CD










 e
~-^
^ «
-3 cn
U_ N_^
CM
CO
O

CO
CM
O



«^-
i—
O










l/l
0)
to
3
CO

dl
4->
fa
a:

c
o
4->
0
E
3
VI
0 ^
<~> E
ai (a
3 cn
1 1 ^^ ^
































                                                       176

-------
     • unit auto pollutant emission rates for trips done to park-
       and-ride and kiss-and-ride sites
     • bus pollutant emission rates
     • bus load factor (residential collection)
     • bus load factors (line haul)
     • bus load factors in the busway during off-peak periods
     • average trip length for park-and-ride and kiss-and-ride
       access  trips.
(2)  Read input data:
     • distribution of trips (percent of total  trip in suburbs,
       highway and downtown)
     • percent of new riders using park-and-ride and kiss-and-ride
       modes
     • before  and after ridership, number of routes and maximum
       and minimum bus service frequencies.
(3)  Estimate  diverted passengers, and cold  start effects.
(4)  Start iterative process:
     • generate a random value for each random  variable from its
       beta distribution
     • estimate bus-miles of travel before and after implementation
       of the  busway
     • estimate auto VMT reduced by trips diverted to bus
     • calculate new auto VMT due to park-and-ride and kiss-and-
       ride.
                               177

-------
          • estimate total fuel  savings and reduction in emission
            of pollutants for this iteration, store results
          • repeat the process for 200 iterations.
     (5)  Determine the mean values and probability distributions of
          output.
     (6)  Print results.
     The Busway model was applied to two case studies--the Shirley Busway
in the Washington, D. C. metropolitan area, and the San Bernardino Busway
in the Los Angeles area.  Tables A.8 and A.9 list data used for each case
study.
CORRIDOR RAIL TRANSIT MODEL
     The CORAIL computer model assumes corridor transit service in which
the line haul part of the trip is made via modern rapid rail transit into
the central business district of the city.  Passenger access to the system
may be by feeder bus, auto, or walking according to user-supplied data.
The model is very similar in assumptions to the Busway model.  CORAIL
incorporates, in addition to the variables used by Busway, the bus miles
of travel generated by the feeder bus service.  The gas consumption and
pollution emission of this service is also taken into consideration in
estimating system outputs.  CORAIL was applied to the Lindenwold Rapid
Rail case study, using the variables and values shown in Table A.10.
                                    178

-------
    UJ
    Q
    O
    LU
    a.
00
oo
r>
CQ
    tO  5-
UJ  =3  UJ
—I  CD  _J
ca      a£
    Q
    LU
    to
    to
    UJ
        to











CU
^L
*












O)
i_
O
'^—
cu
CD





Ol
E
O
C
01
-C
CTV
•r—

OJ
O

3
O


a;
X
u_






cu
3
rO
|
C
(O
Oi

.c
Ol
—

CU
-a
o
"~


3
O
— I

"O
a>
X
•i—
U-


























O
o
o
ro
CM












0
LO
o
•%
00




^
(T3
ai
a.
**•— '
CO
^,
a>

d
at
i/j
CO
fu
a.


LO
LO

*

O







0
LO


LO
co



o
co










CO
o
u
^t

T3
tl
O
— 1

00
3
CO











O

























ai
Oi

<%~
o
• r—
U1
i-
QJ
>
•f"
Q











^
CNI

































a:
oS
a.











o\


































C£.
°C
^
O

"^F

o
CM
O
r-^


























CU
U
2
00
Q
^_
OJ

(O

t—

0£
oQ
Q.
§

^^

O
o
LO
o
o






CO
o
<~>


O
•
O



LO
LO
o
o






c
0
•r-
-t-J
Q.
3
oo
C
O
^—^
'""• .^
CU g
3 N_^
U~ r-^
fO
O Q^
•*-* v^.-*
3
"^
CM

^D

CO
o
r—
O






CO
o

CO
CM
v
O



«
o








c
o
'r—
-t-J
Q.
E
3
CO
c
o
0

,_
0)

u_

00
3
CO














0 0
CO LO
O LO

0 O
• .
LO i—
r» «3-


0 O
^ ^
^- CM






CO
-O CO
5_ ai
3 S-
_Q C"l
3 X
00 LU

o
•
CO
o
o
LO
o
^






o
CO
CM

o
,
f^
oo


0
^
*









Q CC.
CO o3
o a.




c
o o -^
( 	 ) -r- *r™
oo E
O I/I — .
•M -r~ L_
3 E en
eT LU • —
LO
•
CM
CO
0
LO
en
o






LO
CM
CO


,
o
CM


CTv
O











to
C
O
•r-
CO
00

E
UJ

O
O

00
3
CO














CM
03
r—

O
1

-------
cr>
      UJ
      a
      o
a:
tu  >-
I—  <:

a.  co
z  r>
o  co

    o
      co
      ZD
      CQ
          o
          a:
          01
      Z  UJ
      *—i  cQ


      a  zr
      LU  

      co










^
QJ
^•*
^^—
«£










QJ
S-
o
u-
03

















O
cr>





•^
(O
cu
Q.
'
t/l
S-
ai

c
QJ
t/>
l/l
Q-

m
LO

5:


0








0
LO

LO
CO




o
CO








l/>
J_
o
u
u_


o
CD


LO
LO
o
o




c
o
• 1—
4_>
Q.
e
3
C
O
o
^— V
ai '£
3 ^^^
U. r—
fO
° en
-*-> 	 .
"^
CM
o

00
r—
C5

*
o







CM
co
O
co
CM
O


«*
r—
CD





C
O
'r—
a.
E
01
O
(_)
_
I







o
CO
CM
O
XJ



O
=3-








a cr:
CO 08
C_> Q_



C
o o •—
C_J -r— •»—
to E
O 01 "~-.
4J -r- S-
3 E en
<£. LU 	 '
LO
CM
CO
O
LO

a>
d







in
CM
CO

CO
CM




-------
<.

LU
_1
03
O   Q
CJ   i-<
     0.
O   —I
o   o
O   i
LU   1—1
00   _l







cu
•*j
*












OJ
^
o
tfr.
01
CO



01
3
(O
—^
Random \

a:
o>
o

o
_J
-o

d









V)
c
0
VI
i
UJ
O
tuJ

VI
ca
^
LO
O
2

CM
r—




O
•
CO
"

CO
•
en


LO










.c
c
a>
a.

L.
i—

LO
co

LO

O
LO
























S-
c
n:
u_
"S
a
— i
t
•r—
(O
a;
LO
d
i— *
d

co
CM
d























o
•M
Q.
3
c
o
o
'aJ
3
U_
_
>rw
to
o
d
o
d

0
d
























o
10
CO
UJ
o
o

_
gp-
C£.
                                                       181

-------
                              APPENDIX B

                         SELECTED REFERENCES
Alan M. Voorhees & Associates, Inc.   Energy Efficiencies of Urban Passenger
Transportation, for the Highway Users Federation ,  May,  1974.

Alan M. Voorhees & Associates, Inc.  "Factors and Trends  in Trip Lengths,"
National Cooperative Highway Research Program Report No. 48.  1968.

Alan M. Voorhees & Associates, Inc.   Guidelines to  Reduce Energy Consumption
Through Transportation Actions,  for the Urban Mass Transportation Adminis-
tration, May, 1974.

Argonne National Laboratory.  Handbook of Air Pollutant  Emissions from
Transportation Systems.  Prepared for Illinois Institute for Environmental
Qua 1i ty, December, 1973.

Automotive Environmental  Systems, Inc.  A Study of  Emissions from Light
Duty Vehicles in Six Cities.  Prepared for the Environmental Protection
Agency, March, 1973.

Boyce, D. C.  "Notes, on the Methodology of Urban Transportation Impact
Analysis" in Impact of the BART System on the San Francisco Metropolitan
Region, Highway Research Board Special Report III.   1972.

Boyce, D. C. and T. Noyelle.  "The Energy Consumption of the Philadelphia-
Lindenwold High-Speed Line and Other Modes of Transportation in the  South
Jersey Corridor."  Preliminary Draft Report prepared for the Federal
Energy Administration, January, 1975.

Boyce, D. C., et al.  Impact of Access Distance and Parking Availability
on Suburban Rapid Transit Station Choice.Prepared for  the Office of the
Secretary, U. S. Department of Transportation, PB 220 694, November,  1972.

Brand, D. and M. L. Manheim, eds.  Urban Travel Demand Forecasting.  Trans-
portation Research Board Special Report, 143.  1973.

Calspan Corporation.  Automobile Exhaust Emission Surveillance, A Summary.
Prepared for the Environmental Protection Agency, May, 1973.

Grain & Associates.  "First Year Report--San Bernardino  Freeway Express
Busway Evaluation".  Draft Report prepared for the  Southern California
Association of Governments, January, 1974.

Charles River Associates, Inc.  An Evaluation of Free Transit Service.  1968.

Charles River Associates, Inc.  Measurement of the  Effects of Transportation
Changes.  Prepared for the Urban Mass Transportation Administration, August,
1972.
                                    182

-------
DAVE Systems, Inc. and The Orange County Transit District,  Dial-A-Ride
Expansion Plan for Orange County, June, 1974.

De Leuw, Gather & Company.  "Preliminary Evaluation of Existing  Transit
Operations,"  Working Paper No.  1.   Prepared for the Comprehensive  Planning
Organization of the San Diego Region, April, 1974.

District of Columbia Government, et al.  Additions  and Revisions to the
Implementation Plan for the Control of Carbon Monoxide. Nitrogen Oxides,
Hydrocarbons and Photochemical  Qxidants for the District of Columbia
Portion of the National Capital  Interstate Air Quality Control Region.
April, 1973.

Dunphy, R. T.  "Transit Accessibility as a Determinant of Automobile
Ownership."  Paper presented at the 52nd Annual Meeting of  the Highway
Research Board, January, 1973.

Pels, M. F.  Comparative Energy Costs of Urban Transportation Systems.
Transportation Program Report,  Princeton University.September, 1974.

Healy, T. J.  "Energy Requirements of the Bay Area  Rapid Transit (BART)
System."  Prepared for the State of California Department of Transportation,
November, 1973.

Healy, T. J.  Energy Use of Public Transit Systems.  Prepared for the
California Department of Transportation, August, 1974.

Herendeen, R. A., An Energy Input-Output Matrix for the United States,  1963
University of Illinois, Urbana.   1973.

Hertz, D. B.  "The Risk Analysis in Capital Investment" in  Harvard  Business
Review, Volume 42, January-February, 1964.

Highway Users Federation.  Transit Fare and Ridership:  A Review.   December,
1974.

Hirst, E.  Energy Consumption for Transportation in the United States.
Oak Ridge National Laboratory,  March, 1972.

Hirst, E.  Energy Intensiveness  of Passenger and Freight Transport  Modes
1950-1970.  Oak Ridge National  Laboratory, April, 1973.

Hirst, E.  Direct and Indirect Energy Requirements  for Automobiles.  Oak
Ridge National Laboratory, February, 1974.

Hirst, E. and Herendeen, R., "Total Energy Demand for Automobiles".  Paper
presented at the International  Automotive Engineering Congress,  Detroit,
January, 1973.
                                    183

-------
Holland, D. K.  A Review of Reports Relating to the Effect of Fare and
Service Changes Th Metropolitan Public Transportation Systems.Prepared
for the Federal Highway Administration, June, 1974.

Howard R. Ross Associates, Energy Consumption by Transit Mode.  Menlo Park,
California, March 10, 1974.

Horowitz, Joel L., "Transportation controls are really needed in the air
cleanup fight".  Environmental Science and Technology. Volume 8, No. 9.
September, 1974.

Horowitz, J. L. and Pernela, L. M.  "An Analysis of Urban Area Automobile
Emissions According to Trip Type," in Transportation Research Record No. 492,
1974.

Horowitz, J. L. and L. M. Pernela.  "Comparison of Automobile Emissions
According to Trip Type in Two Metropolitan Areas."  Presented at 54th
Annual Meeting of the Transportation Research Board, January, 1975.

Lansing, N. F. and Ross, H. R.  Energy Consumption by Transit Mode.   Pre-
pared for the Southern California Association of Governments, March, 1974.
The MITRE Corporation.

Ludwick, J. S., Swetnam, 6. F., The MITRE Corporation.  A Preliminary Review
of Propulsion Requirements for an Urban Transit Bus. June, 1974.

Madison, W. G., Swetnam, G. F., The MITRE Corporation.  Transit Vehicle
Performance Model—Analysis of California Steam Bus Driving Cycle Data,
Working Paper.  November, 1973.

Metropolitan Atlanta Rapid Transit Authority.  Analysis of Transit Passenger
Data.  October, 1973.

MITRE Corporation.  Transportation Energy and Environmental Issues.   February,
1972.

National Bureau of Standards.  The Shirley Highway Express-Bus-on-Freeway
Demonstration Project - User's Reactions to Innovative Bus Features.
June, 1973.

Pei, R. Y. and I. F. Kan.  "A Decision Aid to Transportation Planners."
Paper presented at 41st Annual Meeting of the Operations Research Society
of America, New Orleans.

Rice, Richard A. "System Energy as a Factor in Considering Future Transpor-
tation", Paper presented at the ASME Annual Meeting, New York, New York,
November, 1970.

Rubin, D. et al., Transportation Energy Conservation Options, preliminary
report, October, 1973.
                                      184

-------
Smylie, J. S.  "Energy Consumption of Alternative Transport Modes  —  Bus,
Light Rail, Conventional Rail, Group Rapid Transit,  and Personal Rapid
Transit."  Prepared for the Third Intersociety Conference on Transporta-
tion, Atlanta, July, 1975.

TRW/De Leuw, Gather & Company.  Travel Impacts of Fuel  Shortage and Price
Increase Conditions.  Prepared for the U.S.  Environmental  Protection Agency,
December, 1974.

U. S. Department of Transportation, A Report  on Actions and Recommendations
for Energy Conservation Through Public Mass Transportation Improvements,
for the United States Congress, Washington, D. C., October, 1974.

U. S. Department of Transportation.  Characteristics of Urban Transportation
Systems.  May, 1974.

U. S. Department of Transportation, Urban Mass Transportation Administration.
The Shirley Highway Express Bus-on-Freeway Demonstration Project—Second
Year Results.  November, 1973.

U. S. Environmental Protection Agency.  Compilation  of  Air Pollutant
Emission Factors, Second Edition, AP-42.  September, 1973.

Transportation Center, The, Northwestern University, Development of Experi-
mental Design Methodology for Evaluating Mass Transit Demonstrations  on
Application to the Seattle Express Bus Service Demonstration Project,
September, 1973.

Wildhorn, S.  et al., The Rand Corporation.  How to  Save Gasoline:  Public
Policy Alternatives for the Automobile.  Santa Monica,  California, October,
1974.
                                    185

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1.
4.
7.
9.
12
15
REPORT NO. 2.
EPA-600/5-76-003
TITLE AND SUBTITLE
CASE STUDIES OF TRANSIT ENERGY AND AIR POLLUTION
IMPACTS
AUTHOR(S)
JAMES P. CURRY
PERFORMING ORGANIZATION NAME AND ADDRESS
De Leuw, Gather & Company
1201 Connecticut Avenue, N.W.
Washington, D.C. 20036
. SPONSORING AGENCY NAME AND ADDRESS
Office of Energy, Minerals and Industry
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
3. RECIPIENT'S ACCESSION«NO.
5. REPORT DATE
May 1976 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1HA093
11. CONTRACT/GRANT NO. 1
68-01-2475
13. TYPE OF REPORT AND PERIOD COVERED
FINAL
14. SPONSORING AGENCY CODE
EPA-ORD
. SUPPLEMENTARY NOTES
Prepared in cooperation with TRW Environmental Services and Bigelow - Grain & Assoc.
16. ABSTRACT
       The paper summarizes analysis of  the  energy consumption  and  air pollution impac
  of eight case studies of new  or  improved transit services.  The case studies include
  (a)  areawide bus service improvement  programs  involving  route extensions,  increased
  frequencies, new lines, demand responsive  service,  and  fare reductions;  (b)  new
  corridor exclusive busway service on the Shirley Highway  and  San  Bernardino  Freeway;
  and  (c) new rail transit service in the Philadelphia-Lindenwold corridor.   Prouabil-
  istic models were developed for  each of these three service improvement  scenerios to
  account for key travel demand and transportation system factors affecting  energy
  consumption and air  pollution impact levels.  Results  showed  that low patronage
  response to areawide bus improvements  as well as diversion  from prior bus  service,
  carpools, etc. and extensive  auto access (park-and-ride,  kiss-and-ride)  to  corridor
  systems reduce expected energy and air pollution gains  and may, under certain con-
  ditions found in four case studies, result in possible  energy use increases.
  Additionally, it was found that  auto use for corridor  system  access may  worsen air
  quality conditions in suburban areas in the vicinity of corridor  transit terminal
  locations.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Exhaust Emissions
Mass Transportation
Transportation Models
13. DISTRIBUTION STATEMENT
Release Unlimited
b.lDENTIFIERS/OPEN ENDED TERMS
Energy Conservation
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
12B
21B
21. NO. OF PAGES
198
22. PRICE
EPA Form 2220-1 (9-73)
186
6USGPO: 1976 — 657-695/5409 Region 5

-------