400180001B
ۥ
\
//••;"•*
^—
transportation
air quality
analysis
«••€• «•
•> €•
sketch planning methods
Volume II
-------
NOTICE
This information document is isssued by the Office of
Transportation and Land Use Policy, U.S. Environmental Protection
Agency, in response to Section 108(f) of the Clean Air Act. This
document provides a range of quantitative analytical techniques to
evaluate transportation measures and packages of alternative measures,
and is designed to assist State and local air pollution control
agencies and Section 174 lead planning agencies to perform
transportation-air quality planning. Although examples are provided,
this document is not intended as a substitute for the necessary
case-by-case analysis by approporiate local planning organizations.
A limited number of copies of this document are available from EPA
Regional Offices. Additional copies may be obtained, for a nominal
cost, from the National Technical Information Services, 5285 Port Poyal
Road, Springfield, Virginia 22151.
This technical report was furnished to the Environmental
Protection Agency by Cambridge Systematics, Inc., Cambridge,
Massachusetts 02142, in fulfillment of Contract No. 68-01-4977. The
opinions, findings, and conclusions expressed are those of the authors
and not necessarily those of the Environmental Protection Agency or of
cooperating agencies. Mention of company or product names is not to be
considered as .an endorsement by the Environmental Protection Agency.
-------
Technical Report Documentation Page
1. Report No. 2. -Government Accession No.
EPA 400/1 -80J|[-001b
4. Title end Subtitle*
TRANSPORTATION AIR QUALITY ANALYSIS -
SKETCH PLANNING METHODS
Volume II: Case Studies
7. Author's)
9. Performing Organization Nome and Address
Cambridge Systematics, Inc.
238 Main Street
Cambridge, Massachusetts 02142
12. Sponsoring Agency Name end Address
Office of Transportation & Land Use Policy
Environmental Protection Agency
401 M Street S.W.
Washington, D.C. 20460
3. Recipient s Catalog No.
5. Report Date
December, 1979
6. Performing Organization Cod*
78015
8. Performing Organisation Report No.
10. Work Unit No. (TRAIS)
11. Contract or Grant No.
68-01-4977
13. Type of Report and Period Covered
Final Report
14. Sponsoring Agency Code
15. Supplementory Notes
Part of a two-volume final report describing sketch planning methods for trans-
portation and air quality planning.
16. Abstract
Analytical methodologies are described (in Volume 1) and illustrated (in Volume II
for use by metropolitan planning organizations and other state and local transportation
agencies in analyzing the air quality potential of candidate urban transportation
measures. As sketch planning techniques, the methods are designed to produce first-
cut estimates of a proposed transportation measure's impact for a relatively small in-
vestment of time and effort. Quantitative methods oriented to auto restricted zones,
high occupancy vehicle priorities, transit improvements, parking programs, carpool/
vanpool incentives, and staggered work hours are provided. The methods use worksheet,
programmable calculator, and computerized approaches to apply disaggregate behavioral
models. They can be used to predict traveller demand as a function of transportation
system characteristics, transportation facility operations as a function of their usage
and their physical characteristics, and special impacts including vehicular emissions,
fuel consumption, and operating costs. Guidelines are provided both to those respon-
sible for designing the transportation-air quality analysis approaches in specific
local areas, and to those who will carry out these analyses. In addition, references
are provided to documents which provide additional detail on the methods.
17. Key Words
Air Quality Planning
Urban Transportation Planning
Sketch Planning
Transportation Systems Management
IB. Distribution Statement
19. Security Classif. (of this report)
unclassified
20. Security Clossif. (of this page)
unclassified
21* No. of Pages
22. Price
Form DOT F 1700.7 (8-72)
Reproduction of completed poge authorized
loO '•>',' Dp,1.7 boviT St. oat, Iluom Io70'
Chica-io. IL ' 60604
-------
EPA 400/1-80-001 b EPA Contract No. 68-01-4977
TRANSPORTATION AIR QUALITY ANALYSIS
SKETCH PLANNING METHODS
VOLUME II
DECEMBER 1979
Prepared for:
Environmental Protection Agency
Office of Transportation and Land Use Policy
in Cooperation with the
Urban Mass Transportation Administration
Office of Planning Methods and Support
-------
PREFACE
The sketch planning methods described in this handbook have been
assembled and illustrated under contract to the U.S. Environmental
Protection Agency and the Urban Mass Transportation Administration in
order to provide assistance to local transportation agencies as they
conduct the planning and analyses required to develop the transportation
portions of State Implementation Plans. The handbook is presented in two
volumes:
Volume I: Analysis Methods
Volume II: Case Studies
The project was performed by Cambridge Systematics, Inc. Earl R.
Ruiter, Project Manager, and John H. Suhrbier, Principal Responsible,
provided the overall direction and management of the work performed. The
development of the handbook benefited greatly from the advice provided by
Marvin L. Manheim in the sketch planning applications of transportation
analysis and programmable calculator methods areas, and by Adolph D. May
(University of California, Berkeley), and Frederick A. Wagner
(Wagier-McGee, Inc.) in the area of highway facility operations. The
development or enhancement of specific analysis methods was carried out
by Ellyn S. Eder and Melissa M. Laube. Additional major contributors to
the handbook were Elizabeth A. Deakin, Lance A. Neumann, Daniel S. Nagin,
Terry J Atherton, Scott D. Nason, William D. Byrne, and Greig Harvey.
-------
Important contributions have been made by EPA staff members Chris
Shaver, David Levinsohn, Gary Hawthorn, and Joseph Ossi. Their support
and individual inputs have been very much appreciated. The contents of
this report, however, reflect the views of Cambridge Systematics, Inc.,'
and they are fully responsible for the facts, the accuracy of the data,
and the conclusions expressed herein. The contents should not be
interpreted as necessarily representing the views, opinions, or policies
of the Environmental Protection Agency, the Urban Mass Transportation
Administration, or the United States Government.
-------
m
TABLE OF CONTENTS
Page
INTRODUCTION
A.
B.
C.
CASE
A.
B.
C.
D.
E.
F.
G.
H.
I.
CASE
A.
B.
C.
D.
E.
F.
G.
H.
I.
CASE
A.
B.
C.
Purpose of this Handbook
The Range of Sketch Planning Techniques Available
General Structure of the Case Studies
STUDY I: FREEWAY FACILITY RESERVED FOR CARPOOLS AND BUSES
Problem Presentation
Proposed Transportation Measures
Selection of Analysis Techniques
Overview of the Analysis
Defining the Scope of the Analysis
Input Data Development
Description of Model Application
impact Assessment
Interpretation of Results
STUDY II: DOWNTOWN AUTO RESTRICTED ZONE
Problem Presentation
Proposed Transportation Measures
Selection of Analysis Technique
Overview of the Analysis
Defining the Scope of the Analysis
Input Data Development
Description of Model Application
Impact Assessment
Interpretation of Results
STUDY III: BUS PRIORITY STRATEGIES FOR A RADIAL URBAN CORRIDOR
Problem Presentation
Proposed Transportation Measures
Selection of Analysis Technique
i- 1
i- 5
i- 7
I- 1
I- 1
I- 3
I- 1
I- 7
I- 9
1-36
1-38
1-51
II- 1
II- 1
II- 4
II- 5
II- 7
II- 8
11-22
II-M1
11-51
III- 1
III- 3
III- 3
-------
IV
LIST OF TABLES
Page
I- 1 Detailed Trip Generation Characteristics 1-12
I- 2 Trip Generation Parameters 1-14
III-l Tract and Block Input Data from "Block Statistics" 111-10
Boston Urban Area
III-2 Tract Income Data from Census Tracts, Boston, SMSA III-ll
III-3 Tract Auto Ownership Data from census Tracts, Boston 111-12
SMSA
III-U Station Characteristics 111-23
III-5 Line Segment Characteristics III-2M
III-6 Intersection Characteristics III-2M
III-7 Equilibrium Transit Travel Times under Alternative 111-55
Policies
III-8 Equilibrium Transit Ridership under Alternative 111-55
Policies
III-9 Impacts Under Alternative Policies 111-59
111-10 Bus Emissions of Major Pollutants under Alternative 111-63
Policies
III-ll Estimated Auto and Bus Emissions Impacts 111-69
-------
IV-1 Estimated versus Adjusted Values of Alternative IV- 1
Specific Constant Terms
IV-2 Changes in Travel Behavior and Travel-Related Impacts IV-12
IV-3 Percent Changes in Travel Behavior and Travel-Related IV-15
Impacts of Ridesharing Promotion
IV-U Changes in Travel Behavior and Travel-Related Impacts (A) IV-18
IV-5 Changes in Travel Behavior and Travel-Related Impacts (B) IV-21
V-l Representative Trips for the Busway Option V-10
V-2 Subzone Percentages Busway Network V-12
V-3 Level of Service Assumptions V-14
V-1! Impact Summary for Busway Transit Option V-26
VII-1 Peak Period Travel Times VII-11
VII-2 Traffic Signal Timing Optimization Impacts VII-16
VII-3 Impacts of Traffic Signal Master Control Improvements VII-17
VII-U Freeway Ramp Control System Impacts on Average Speed VII-20
VII-5 Freeway Ramp Control System Impacts on Proportion VII-21
of Freeway Minute-Miles Congested
VII-6 Worksheet for Steps 1-6, Work Travel VII-23
VII-7 Travel Time Elasticities VII-26
VII-8 Worksheet for Steps 1-13, Work Travel VII-30
VII-9 Worksheet for Steps 1-13, Non-Work Travel VII-31
VII-10 Summary of Impacts of Combined Traffic Engineering VII-32
Actions
-------
VI
LIST OF FIGURES
Page
I- 1 Basic Steps for HOV Lane Analysis I- 5
I- 2 Information Flow in Manual VMT Worksheet Analysis I- 6
I- 3 Auto Ownership Distribution I-10
I- U Study Corridor and Proposed HOV Lanes 1-17
I- 5 Freeway/Arterial Diversion Curves 1-26
II-l Steps in Analysis of ARZ Policy Scenarios II- 6
II-2 Auto Ownership Distribution 11-10
II-3 CBD Entry and Destination Zones 11-18
III-l Study Corridor with Residential Analysis Zones III- 2
III-2 Integrated Analysis Steps and Programs III- 6
III-3 Sample HHGEM Worksheet 111-13
III-U Generated Household Characteristics I11-16
III-5 Datasheet (2MODE-AGG) 111-20
III-6 Simplified Bus Route Representation 111-21
III-7 Base Case Equilibrium Data Flows 111-25
III-8 User Worksheet-Binary Mode Choice with Aggregation 111-27
(2MODE-AGG)
III-9 Datasheet (2MODE-AGG) 111-29
-------
VII
111-10 User Instructions - Bus Route Simulation 111-32
III-ll Worksheet C-l Base Data III-U6
III-12 Worksheet C-2 Changes in Transportation Level of Service III-M7
111-13 Worksheet C-H Program Steps III-H8
III-1U Equilibrium Data Flows for Testing Effects of III-54
Transportation Measures
111-15 User Worksheet - Energy Consumption on Bus Route 111-57
III-16 User Worksheet - Bus Pollutant Emissions 111-60
IV-1 SRGP Data Preparation Steps IV- »J
IV-2 Effectiveness of Combining Incentive and Disincentive IV-23
Measures
V-l Busway Network V- 2
V-2 Major Steps in Sketch Planning Procedure V- 5
V-3 1987 Transit Zones V- 7
V-U System Description Log Sheet V- 9
V-5 Basic Level-of-Service Coding V-15
V-6 Mapping Matrix V-17
V-7 Work Mode Choice Sheet V-18
V-8 Non-Work Mode Choice Sheet V-18
V-9 Peak Period Operating Cost Sheet V-21
V-10 System Capital Costing Sheet V-23
V-ll Environmental Impacts Sheet V-2M
-------
VI11
VI-1 Location of Eastshore Freeway Corridor VI- 2
VI-2 Detailed Corridor Map VI- 3
VI-3 Eastshore Freeway Facility Data VI- 6
VI-4 Sample Origin-Destination Data for Eastshore Freeway— VI- 8
Time Slice One
VI-5 Distribution of Passenger Occupancies on the Eastshore VI-10
Freeway
VI-6 San Pablo Avenue Facility Data VI-11
VI-7 Queueing Diagram of Freeway Congestion before Entry VI-13
Control
VI-8 Eastshore Freeway—Freeway Summary before Entry Control VI-15
VI-9 Eastshore Freeway—Arterial Summary before Entry Control VI-16
VI-10 Eastshore Freeway Ramp Control Summary: No Bypass VI-18
Priority Ramp Lanes, No Diversion
VI-11 Eastshore Freeway Differential Effects: Entry Control VI-20
without Bypass Priority Ramp Lanes, No Diversion
VI-12 Effects of Short- and Longer-term Diversion for Ramp VI-21
Control without Priority Entry
VI-13 Eastshore Freeway Ramp Control Summary: Priority VI-23
Ramp Bypass for Vehicles with Three or More Occupants,
No Diversion
VI-1U Effects of Ramp Control with Priority Entry for VI-24
Vehicles with Three or More Occupants
VI-15 Effects of Ramp Control after Diversion and Modal VI-25
Shift with Priority Entry for Vehicles with Two or
More Occupants
-------
INTRODUCTION
A. Purpose of this Handbook
This handbook presents case studies illustrating the application of
a selection of techniques for transportation-air quality planning. The
emphasis is on "sketch planning" techniques—ones which can produce a
first-cut estimate of a proposed transportation measure's impact for a
relatively small Investment of time and effort.
The analysis approaches illustrated here should be useful in
developing the transportation portions of State Implementation Plans
(SIP's), as required to meet national ambient air quality standards under
the Clean Air Act, as amended (42 U.S.C. 1857 et seq.). The Act, and the
joint Environmental Protection Agency/Department of Transportation
guidelines issued pursuant to it, call for the analysis of a number of
transportation measures which potentially could improve air quality
(Table 1). The effects these measures would have—on travel; the
transportation system; energy conservation; and a host of other social,
environmental, economic, and financial concerns; as well as on air
quality—must be evaluated within a broadly participatory, interactive
planning process. Furthermore, the analyses must be completed
expeditiously, in order to meet legislative deadlines for SIP adoptions
and submittals. These combined requirements necessitate analytic
capabilities which produce results quickly and yet provide accurate
information on a wide range of impacts.
-------
i-2
TABLE 1
Reasonably Available Transportation-Air Quality Measures
to be Analyzed for SIP Revisions
Transportation System Measures;
• improved public transit (short- and long-range)
• exclusive bus and carpool lanes
• area-wide carpool programs
• private car restrictions
• on-street parking controls
• pedestrian malls
• park-and-ride and fringe parking lots
• employer programs to encourage carpooling and vanpooling, mass
transit, bicycling and walking
• bicycle lanes and storage facilities
• staggered work hours (flexitime)
• road pricing to discourage single-occupancy auto trips
• traffic flow improvements
Vehicle and Equipment Measures:
• inspection and maintenance programs
• alternative fuels or engines and other fleet vehicle controls
• other than light duty vehicle retrofit
• extreme cold start emission reduction programs
• controls on extended vehicle idling
• vapor recovery
^-Note: This report focuses on the first category of measures, those
affecting the transportation system. The second catgory of measures are
considered in some of the emissions estimation techniques, however.
-------
i-3
Metropolitan Planning Organizations (MPO's) for the most part are
taking the lead in transportation-air quality analysis. Often, however,
the MPOs' analytical techniques center around large computer modelling
systems which originally were designed to investigate the effects of long-
term activity shifts and/or major capital investments in transportation.
Such model systems tend to be too costly, too data-demanding, and too time
consuming for use in analyzing the numerous alternatives to be considered
in transportation-air quality planning. Because the models were not
intended for analysis of transportation operations, management policies, or
small changes in facilities, they often omit variables necessary to study
such measures. Some of the variables which are included enter the models
in ways that cannot respond to the influence of proposed actions, or that
can do so only by receding or reprogramming. Because the models were meant
for regional studies, they frequently are so "aggregate" (i.e., coarse-
scaled) that small or localized change cannot be discerned. The models
usually do not have the capability to focus on a subarea or corridor except
after considerable modification, and then only with a great deal of work.
They rarely distinguish among various socioeconomic groups or other popu-
lation subsamples. Thus, the conventional model systems are not well-
suited for transportation-air quality planning—nor for most transportation
system management (TSM) or transportation-energy conservation planning
efforts, which likewise emphasize quick response analysis of management and
operations policies and small capital investment projects. Methods which
are cheaper, quicker, more flexible, and more responsive are needed.
-------
In recent years, a number of sketch planning tools have been devel-
oped. They range from specialized techniques designed to address a parti-
cular type of measure or impact to general-purpose procedures and method-
ologies, and they cover a smililarly broad range of sophistication and
complexity. In Volume I of this handbook, a selection of these sketch
planning tools is reviewed, their applicability to various analysis
problems is evaluated, and the resource requirements for each technique are
assessed. Special emphasis is given to techniques suitable for
transportation-air quality analysis, although most of the methods are more
generally applicable to transportation system management planning and
transportation energy conservation planning.
This handbook is designed with two major types of readers in mind:
• Those who are designing transportation-air quality analysis
approaches, and in doing so must select an appropriate set of
analysis techniques.
• Those who are conducting transportation-air quality analyses,
"and require reference material on particular analysis methods.
Volumes I and II both contain material useful to each of these types
of readers. Volume I includes a discussion of the terminology used in
the handbook, overviews of various kinds of analysis methods,
descriptions of specific methods, pointers to further reference material,
and analysis design issues. The case studies included in this volume
i
provide extended examples of specific analysis methods. These examples
will be useful to designers of analysis approaches in understanding the
applicability of the methods, and to those carrying out these approaches
in understanding the steps which must be carried out as specific methods
are used.
-------
i-5
B. The Range of Sketch Planning Techniques Available
Numerous sketch planning techniques have been developed in recent
years. Three categories of such techniques are presented in this
handbook:
• Travel demand analysis methods are those which predict travel-
ler behavior in response to change in the transportation
system. Techniques for trip generation, destination choice,
mode share, and route choice are all reviewed.
• Facility operations analysis methods predict the operating
characteristics of transportation facilities as a function of
changes in capacity and operating policy. Intersection
improvements, signal timing, capacity changes, and flow meter-
ing all are addressed by the techniques discussed.
• Special impact analysis methods focus on particular effects of
transportation changes. Methods for assessing changes in vehi-
cular emissions and fuel consumption are presented.
Within these categories, the techniques are further classified by the
technology used in applying them:
• Manual methods are techniques which utilize worksheets, formu-
las, nomographs and the like to carry out hand calculations, as
well as approaches for making use of data or study results from
other urban areas.
• Programmable calculator methods are adaptations of manual
methods which, by capitalizing on recent developments in inex-
pensive calculating equipment, allow for more detailed and pre-
cise analysis at no significant increase in effort.
• Computer-based methods are model systems for which time and
expense are minimized by making simplifying assumptions or
otherwise limiting the scope of the analysis.
Table 2 illustrates how the classification is used in this hand-
book. The methods were selected as representatives of the range of
approaches which have been developed. The various techniques can accom-
modate different amounts and types of data, can be used at different levels
of detail, and require various levels of staff expertise or experience,
-------
TABLE 2
Typology of Sketch Planning Techniques
Technology
Demand
Facility Operations
Special Impacts
Manual
worksheet mode choice
quick response urban
travel estimation
techniques
systematic data
analysis
traffic flow formulae
graphical techniques
areawide traffic
engineering method
transfer of experience
emissions worksheets
auto fuel consumption
and operating costs
Programmable
Calculator
HHGEN
2MODE-AGG
3MODE(VAN)-AGG
BUS
BUSPOL
ENERGY
Computer
CAPM
SRGP
transit sketch planning
procedure
TRANSYT
FREQ
MOBILE1
-------
i-7
labor, and other resources. Furthermore, techniques which are applicable
to the full range of transportation-air quality measures have been
included. The handbook thus should be useful in disparate settings and
under disparate conditions.
An index to the methods listed in Table 2 is provided in Table 3.
This table shows the case studies in this volume in which the various
methods are demonstrated, as well as the policies illustrated in each case
study. Also, for each general class of method, and for each individual
method, the appropriate section in Volume I in which it is discussed is
indicated.
To illustrate the use of Table 3, consider first the analyst who
wishes to see examples of the analysis of parking programs. Under this
column, both Case Studies I and IV appear. Case Study I illustrates a
number of manual demand and impact methods, and Case Study IV illustrates
SRGP, a computer demand method. Secondly, consider an analyst who wishes
to obtain information on calculator demand methods. The table provides
Volume I section references to the general overview of this class of
methods (2.2), and to the three methods described in detail in this
handbook (2.2.1, 2.2.2, 2.2.3). The table also shows that two case studies
(II and III) illustrate the use of these methods.
C. General Structure of the Case Studies
The remainder of this volume consists of seven case studies drawn from
previous work done by the authors of this handbook and others to illustrate
the application of sketch planning analysis methods to transportation-air
-------
TABLE 3
Cross-Reference Table for Analysis Methods and Case Studies
VOLUME II: CASE STUDIES BY POLICY CLASS
METHOD
Manual Demand
Pivot Point Mode Choice
Synthetic Mode Choice
Quick Response Urban Travel
Estimation
Systematic Data Analysis
Calculator Demand
HHGEN - Household Samples
2MODE-AGG - Synthetic Mode Choice
3MODE(VAN)-AGG - Pivot Point Mode
Choice
Computer Demand
CAPM - Community Aggregate
Planning Model
SRGP - Short-Range Generalized
Policy Analysis
Transit Sketch Planning
Manual Facility Operations
Traffic Flow Formulae
Graphical Techniques
Areawide Traffic Engineering
Transfer of Experience
VOL I
SEC-
TION
2.1
2.1.1
2.1.1
2.1.2
2.1.3
2.2
2.2.1
2.2.2
2.2.3
2.3
2.3.1
2.3.2
2.3.3
3.1
3.1.1
3.1.2
3.1.3
3.1. A
Auto
Restric-
ted
Zones
II
II
110V
Prior-
ities
I
I
I
III
III
III
Traffic
Flow
Improve-
ments
II
II
VII
Transit
System
Improve-
ments
I
I
I,H
III
III
II, III
IV
V
Park-
ing
Pro-
grams
I
I
I
IV
Pric-
ing
Poli-
cies
Carpool/
Vanpool
Incent-
ives
I
I
I
IV
I^HMI^H^^^HMHMMMMM
Stag-
gered
Work
Hours
^^^^^^•^•i
CO
-------
TABLE 3 (Con't)
Cross-Reference Table for Analysis Methods and Case Studies
••••••••••••••••••••
VOLUME II: CASE STUDIES BY POLICY CLASS
METHOD
Calculator Facility Operations
BUS - Bus Operations
Computer Facility Operations
TRANSYT - Arterial Street Systems
Analyzer
FREQ - Freeway System Analyzer
Manual Impacts
Emissions Worksheets
Auto Fuel Consumption. and Operating
Costs
Calculator Impacts
BUSPOL/ENERGY - Bus Emissions and
Fuel Consumption
Computer Impacts
MOBILE1 - Auto Emissions
VOL I
SEC-
TION
3.2
3.2.1
3.3
3.3.1
3.3.2
A.I
4.1
4.2
4.3
4.3
4.4
4.4
Auto
Restric-
ted
Zones
II
II
HOV
Prior-
ities
III
VI
1,111
III
Traffic
Flow
Improve-
ments
VI
II, VII
II
Transit
System
Improve-
ments
III
I, II, III
II
III
Park-
ing
Pro-
grams
I
Pric-
ing
Poli-
cies
Carpool/
Vanpool
Incent-
ives
I
Stag-
gered
Work
Hours
I
VO
-------
i-10
quality planning. A common structure is generally used to present each
case study, with minor variations adopted where these are required. The
sections included in the common structure are presented as subsections
here, with a discussion of their functions. Also, the sections of Volume I
which discuss more generally the issues or procedures addressed in the
specific case study context are identified.
1. Problem Presentation
This section introduces the case study, discussing briefly the
air quality problem which is addressed and its geographical,
demographic, and transportation system contexts.
2. Proposed Transportation Measures
The nature of the transportation measures which are proposed as
partial solutions of the air quality problem are discussed. The
proposed new or modified transportation facilities and operational
strategies are discussed, and their expected impacts on both travel
patterns and air quality are identified.
3. Selection of Analysis Techniques
The analysis techniques used in the case study are identified,
with references to their discussion in Volume I. Also, the reasons
for selecting each technique are presented. General discussions of
the analysis selection issue are included in Sections l.t and 5.1 of
Volume I.
-------
H. Overview of the Analysis
In cases where multiple techniques are illustrated, the major
analysis steps are described and their relationships to each other are
discussed. Section 5.1 of Volume I addresses this topic more
generally.
5. Defining the Scope of the Analysis
This section presents a rationale for the market segmentation
approach chosen for use in the analysis, as an example of the
application of the guidelines presented in Section 5.3 of Volume I,
and describes the criteria used to define each group of actual or
potential travellers identified as a market segment.
6. Input Data Development
The all-important process of specifying the values of input
variables for the chosen analysis methods is described. Data sources
are identified, and alternative sources are mentioned as an aid to
understanding how specific local data availability issues can be
addressed. Section 5.2 of Volume I discusses general issues in
representing transportation system changes, and Section 5.5 discusses
alternative data sources more generally.
7. Description of Model Application
The step-by-step process of applying the chosen analysis
techniques to evauate the proposed transportation measures is
discussed. Where applicable, illustrative worksheets are included,
completed to show the calculation process which was carried out. This
section thus provides examples of the use of the methods presented in
Chapters 2, 3, and 4 and of the worksheets included in Appendices A
through E of Volume I.
-------
i-12
8. Impact Assessment
This section summarizes the travel, air quality, and energy
conservation impacts predicted for the proposed transportation
measures.
9. Interpretation of Results
The results summarized in the former section are interpreted in
terms of their significance, their impact on further analysis and
implementation decisions, and their limitations due to the nature of
the input data, scope of the analysis and applicability of the
analysis methods. It should be noted that the results are only meant
to illustrate the use of the sketch planning methods to evaluate a
specific transportation-air quality problem in a specific geographic,
demographic, and transportation context. The results should not be
used as general evaluations of the proposed transportation measures or
as indicative of the impacts to be expected if these measures are
implemented in other contexts.
-------
CASE STUDY I
FilEEWAY FACILITY RESERVED FOR CARPOOLS AND BUSES
-------
CASE STUDY I: FREEWAY FACILITY RESERVED
FOR CARPOOLS AND BUSES
A. Problem Presentation
An urban area with population 1.5 million has serious violations
of the health standard for ozone during the summer. Within the area
is a travel corridor which includes a major radial freeway. This
freeway generates 34,000 vehicle trips to the CBD and fringe areas
of the central city during the morning peak period. One of the
largest travel generators in the region, the corridor suffers from
increasing congestion, particularly on the radial freeway. In addi-
tion to the congestion, major construction projects proposed for the
downtown area are expected to cause delays to CBD-bound commuters
from the corridor.
B. Proposed Transportation Measures
To help alleviate the increasing congestion and the air quality
problems in the area, a plan incorporating two complementary actions
has been proposed. The major feature of the proposal is the construc-
tion of a reversible two-lane facility on the median strip of the
existing freeway. These lanes will provide peak direction flow for
buses and carpools with four or more occupants. Expansion of bus
services is also proposed to encourage the use of the high occupancy
vehicle (HOV) lanes.
-------
1-2
Due to the existing congestion on the freeway, significant time
savings for the line-haul portion of the trips made by vehicles eligible
to use the HOV lanes are expected. The project involves adding new
lanes to the freeway, so congestion and travel times are not expected
to increase for drive alone commuters and those in two and three person
carpools, except those bound for the downtown area who will be delayed
by the construction there.
Other factors are also expected to change the level of transpor-
tation service for trips in the corridor. As a result of a program
to encourage carpooling, major employers in the fringe area (represent-
ing about half of the total employment in this area) are planning to
offer preferred parking spaces to carpoolers. Single occupant auto
drivers working at these employment sites should be inconvenienced to
some degree. An additional measure intended to encourage the use of
public transit in the corridor is the creation of special bus lanes in
the downtown area, routed so as to avoid the construction-related delays
to be experienced by autos.
It is argued that these measures will result in a shift by people
living in the corridor and working in the central city from autos to
buses, from single-occupant autos to carpools, and from two- and
three- occupant autos to four-person and larger carpools. These shifts
in mode usage should lead to decreases in VMT, vehicle trips, HC and
NOX emissions (which cause formation of ozone), and gasoline consumption
in the corridor.
-------
1-3
The direct impacts of the proposed measures on transportation
levels of service in the corridor are as follows:
Change
HOV Lanes
Construction
Employer Parking Policy
(Affects one-half of
employees in the fringe
area)
Bus Frequency
Improvements
LOS Impact
Reduced transit in-vehicle travel
time (IVTT) to central city (fringe
and CBD)
Reduced IVTT for 4+ person carpools
to central city
Increased IVTT for all autos to the
CBD
Increased out-of-vehicle travel time
(OVTT) for drive alone trips to fringe
areas
Decreased OVTT for all carpools to
fringe areas
Decreased OVTT for transit trips to
the central city
The problem is to predict the magnitude of the vehicle emissions,
fuel consumption, and other impacts of the proposed measures, given the
set of expected LOS impacts. It is expected that the proposed measures
will be completely in place by 1982, which has therefore been chosen as the
analysis year.
C. Selection of Analysis Techniques
Because the proposed measures are in the early stages of planning
and evaluation, a quick and somewhat gross estimate of the effects of the
plan is desired to determine if further, more detailed analysis is justified,
The limited resources available for this initial analysis dictate that
-------
1-4
manual techniques be used, drawing on existing sources of data to the max-
imum degree possible. For these reasons the manual pivot-point mode choice
(2.1.la) and the emissions worksheets (A.I) methods are chosen for the
analysis of the proposed transportation measures. In addition, portions
of the quick response urban travel estimation (2.1.2) and systematic data
analysis (2.1.3) techniques are used to specify base case conditions.
D. Overview of the Analysis
Seven major steps are involved in the application of the manual
analysis techniques to the evaluation of the HOV lanes and accompanying
measures:
Step 1 - Identify the effects of the HOV lane and other factors;
Step 2 - Segment the population into homogeneous subgroups;
Step 3 - Quantify existing mode splits for each sub-group;
Step 4 - Quantify the level-of-service changes for each sub-group;
Step 5 - Apply FEA VMT analysis worksheets to each subgroup;
Step 6 - Determine corridor-wide VMT and auto trip changes;
Step 7 - Determine corridor-wide emissions changes.
These steps are illustrated in Figure 1-1. Information flows between
steps are illustrated by arrows. Steps 5 and 6 involve the utilization
of the FEA VMT analysis worksheets, while Step 7 involves the use of
the manual emissions worksheets.
Figure 1-2 illustrates the use of the manual pivot-point mode choice
method in carrying out the analysis. Steps labelled with numbers indicate
the use of specific worksheets. Worksheets I, II-A, II, V, and V-A are
used only once. The remaining worksheets are used once for each popula-
The section numbers following specified analysis methods refer to the
location of their description in Volume I. In addition, master copie0
of all worksheets used are provided in Appendices A and D of Volume I.
es
-------
1-5
Identify Effects
of HOV Lane
and other factors
Segment Affected
Population into
Homogeneous Subgroups
Quantify Existing
Mode Splits for
Each Subgroup
Quantify LOS
Changes for
Each Subgroup
Apply VMT
Analysis Worksheets
to each Subgroup
Determine
Corridor - Wide
VMT Changes
Determine
Corridor - Wide
Emissions Changes
FIGURE 1-1
Basic Steps for HOV Lane Analysis
-------
II-A* III-A* Xyes
Record
Data
All Sub
Determine LOS Estir
ja 1 ». Changes hy _^, Change
Carpool Size Carpoo
Base / „ , \
/ Inlnnt- \ — . .......
for / „ f C }
Ern,,pfi ~^l Subgroup !_A*
\^^ / Determine
. ». Rasp
VMT
IV * HI*
Estimate Estimate
Revised , Revised Work
VMT " Trip Model ^
Shares
/ More ^^ yes
X. Subgroups!/'
V-A* | | y*
Determine Determine
Corridor Corridor
Average Mode Average VMT *• XO Emissi
Share Shifts Shifts
nate ^/^ ^\^
. ^^ More >v
»R in . . »-f ^s
L Size ^^X. Sub8roups?x'
no
II*
f \ Determine
Select \ Changes in LOS
\ Subgroup j* By Mode for
\ / All Subgroups
1 1
ons Analysis Worksheets
H
*Worksheet number
FIGURE 1-2
Information Flow in Manual VMT Worksheet Analysis
-------
1-7
tion subgroup to be analyzed. The following section describes the identi-
fication of population subgroups.
E. Defining The Scope of the Analysis
Steps 1 and 2 define the scope and level of detail of the analysis.
The more effects identified and more subgroups considered, the more com-
plex (and accurate) the analysis becomes. The major impacts of the proposed
measures were outlined in Section B. These effects will be the only ones
considered in this initial evaluation of the proposal.
The purpose of segmenting the affected population is to obtain a
number of groups which face approximately the same changes in level-of-
service, and which have the same set travel opportunities available to
them. For example, under the proposed measures, commuters bound for the
CBD will expedience somewhat different impacts with respect to level-
of-service than those bound for the fringe areas. For this reason, a
classification of the corridor's commuters based on the location of their
workplace is suggested.
Individuals not owning automobiles are likely to respond very differ-
ently to changes in the level-of-service of transportation modes than those
who have one or more vehicles available for their use. Once again, a
segmentation of the market is suggested, with auto-owning households
being considered independently from those without autos available for
their use. Those not owning autos are assumed not to have the drive alone
mode available, but may use the transit or carpool modes. The popula-
tion of the corridor is relatively homogeneous with respect to other char-
acteristics, such as income, which are likely to affect travel behavior.
-------
1-8
For this reason, no further segmentation of the population need be done
for this analysis.
in situations where there is a great deal of variation within a
population with respect to income, work trip length, current mode to work
or trip-making behavior, a more extensive segmentation of the population
would be useful. Also, if different types of change in transportation
level-of-service will be experienced by certain groups (such as in the
case of improved transit service targeted to low income areas within a
city) the segmentation should be detailed enough to account for any dif-
ferences which might lead to variations in travel behavior response. An
additional consideration in the definition of market segments is the level
of detail and accuracy desired for the analysis in relation to the budget
available for preparing the input data and doing the calculations.
The four subgroups implied by the two levels of segmentation
described above are:
I. CBD-bound with all modes available
II. CBD-bound without the drive-alone mode
III. Fringe area-bound with all modes available
IV. Fringe area-bound without the drive alone mode.
Vehicle and person-trip counts on the existing highway—taken yearly
by the local traffic department—indicate that 60 percent of the trips
originating on the corridor are bound for the CBD, while 40 percent are
destined for the fringe area.
Income projections, developed as part of a land use study for the
metropolitan area, indicate that the mean income (in 1970 dollars) for
-------
1-9
residents of the study corridor will be $11,000 in 1982. Because no pro-
jection of auto ownership levels is available, this income level must be
used to estimate the distribution of households by auto ownership category
for the corridor. Curves relating auto ownership level and income,
grouped by urban area size category are available in the report, Trip
Generation Analysis, prepared by FHWA. The curve for Philadelphia,
Pennsylvania was chosen for use in estimating auto ownership in the
study corridor (See Figure 1-3). This selection was based on the fact
that the urban area under study, like Philadelphia, is an older eastern
city with relatively good transit service. Reading the auto ownership
levels for 1970 income level $11,000 yields: zero autos, 11.5 percent;
one auto, 59.0 percent; two autos, 28.0 percent; three or more autos,
1.5 percent. The 11.5 percent zero-auto ownership level and the distribu-
tion of trips by destination imply the following distribution of house-
holds in the corridor by subgroup:
I. - 53.1%
II. - 6.9%
III. - 35.4%
IV. - 4.6%
These values are entered on column 2 of Worksheet I, Base Data.
This use of available data to assist in specifying the base case
illustrates a systematic data analysis method (2.1.3).
F. Input Data Development
The input data required for the analysis are defined on Worksheets I,
II-A, and II.
U.S. Department of Transportation, Federal Highway Administration;
Trip Generation Analysis, August, 1975.
-------
1-10
PHILADELPHIA. PENNSYLVANIA
4,020,420 5,346
1500 3000 SON 7000 1000 11.000 13.500
17.500
22.500
37.000
INCOME
FIGURE 1-3
Auto Ownership Distribution
-------
1-11
Socioeconomic, mode choice, and tripmaking statistics are required
by Worksheet I for each of the four market segments to be analyzed.
Typical sources for the types of data required include:
• Nomographic/formula techniques or other estimation procedures
• U.S. Census
• Recent or updated home interview surveys
• Traffic counts
• Transit ridership data
• Projections done in past years
• National average figures and relationships
Household Data
In this case, income projections based on past Census data were
used to determine the average income for the corridor and the average
number of workers per household (column 3 and A) in 1982. No informa-
tion on non-work tripmaking was available from the land use study which
provided the projections, so a portion of the quick response urban travel
estimation technique (2.1.2) was employed to estimate the non-work trip-
making behavior of the corridor's households. Table 2 of NCHRP Report 187
(repeated here as Table 1-1) provides average tripmaking rates by urban
area size, income range, and auto ownership level, as well as the typical
distribution of number of autos owned for each income range and the per-
centage of trips by major purpose. For urban areas between 750,000 and
1,000,000 population, and the $10 - 12.5 thousand income range, the follow-
ing information is presented:
Sosslau, et al., Comsis Corporation; Quick-Response Urban Travel Esti-
mation Techniques and Transferable Parameters: Users Guide; NCHRP
Report 187; 1978.
-------
1-12
TABLE 1-1
Detailed Trip Generation Characteristics
URBANIZED AREA POPULATION: 250,000-750,000
LncomeRange
1970 $
(OOO's)
0-3
3-lt
>t-5
5-6
6-7
7-8
8-9
9-10
10-12.5
12.5-15
15-20
20-25
25+
Wt. Avg.
Avg Autos
Per HHd
O.U7
0.77
0.88
1.01
1.10
1.21*
1.33
l.ltO
1.58
1.72
1.88
2.01*
2.08
l.Ul
Average
Daily Person
Trips per HHe
3.3
5.8
6.9
8.1*
9-5
10.9
11.7
12. It
13.5
lit. 6
15-5
16.0
16.2
ll£
% HH by Autos Ownedb
0
58
38
29
20
lit
8
6
It
2
2
2
1
1
12
1
37
50
57
62
65
61t
60
57
1*6
36
26
20
17
ItU
2
5
10
12
16
19
25
30
35
1*6
53
58
59
61
37
3+
0
2
2
2
2
3
It
It
6
9
lit
20
21
7
Average Daily Person Trips
Per HH by Mo. of Autos /HHC
0
l.lt
2.0
2.5
2.9
3.5
lt.0
U.6
5.2
5.5
5.5
5.2
5-0
5.0
lt.3
5.6
7. It
8.0
9.0
9.7
10.6
11.0
11.2
11.3
ll.lt
11.5
12.0
12.0
10.0
9.3
10.8
11.5
12.1
12.7
13.5
1U.2
lit. 8
15.6
l6.lt
16.7
16.7
16.7
llt.lt
9.3
U.I
11.9
12.7
13.5
lU.lt
15.3
16.2
17.6
18.8
19-1
18.6
18.6
15.8
% Average
Tri™ I
10
13
21
22
22
20
20
20
20
20
20
20
20
20
; Daily Person
iv Purp<7«**
67
61*
57
56
55
55
55
55
55
53
52
50
50
55
23
23
22
22
23
25
25
25
25
27
28
30
30
25
1970 $
(OOO's)
0-3
3-1*
It- 5
5-6
6-7
7-8
8-9
9-10
10-12.5
12.5-15
15-20
20-25
25+
Wt. Avg
Avg Autos
Per HIT
O.lt7
0.68
0.78
0.8U
0.95
1.06
1.16
1-25
l.ltl
1.60
1.77
1-95
2.02
1.31
URBANIZED AREA POPULATIOK: 750,000-2,000,000
Daily Person
Trips per HHe
1-9
3.7
It. 5
5.1
5.8
6.5
7.2
7.7
8.5
9. It
9-9
10.6
11.0
7.6
% HH b
0
58
ItO
32
28
22
16
12
9
5
2
2
2
2
15
y.Aut
T \~
37
52
58
60
62
63
63
61
56
It5
35
21*
20
U8
os Ov
1 2~"
5
8
10
12
15
20
23
27
3U
1.6
51
56
58
32
nedb
~3+
0
0
0
0
1
1
2
3
5
7
12
18
20
6
Averag
Per HH
0.8
2.0
2.5
2.7
2.9
3.0
3.2
3.3
3.U
3.6
3.6
3.7
lt.0
3.1
by No
3.2
U.5
5.0
5.6
6.1
6.5
6.9
7.1
7.5
7.5
7.6
7.0
7.0
6.5
r Person Trips
of Autos /HHC
5.7
7.0
7.5
8.0
8.6
9.2
9.5
9.9
10.1
10.7
10.7
U.I
11.3
9-5
7.3
9.2
10.0
U.o
U.6
12.2
12.6
13.0
13.6
Hi. 2
lit. 6
1U.8
lit. 8
12.6
% Average Daily Person
Trips by Purpose*
lit
22
28
26
28
27
27
27
26
25
21*
2lt
23
25
66
59
53
53
53
53
53
5U
53
53
53
53
5U
5
-------
1-13
Average Daily
Person-Trips per
Household
Auto Ownership Level
0
3.4
1
7.5
2
10.1
3+
13.6
The distribution of households by auto ownership was previously
determined to be:
Percentage
of Total
Households
Auto Ownership Level
0
11.5
1
59.0
2
28.0
3+
1.5
Applying this distribution to the average person-trip rates shown
above yields:
.115 X 3.4 + .590 X 7.5 + .280 X 10.1 + .015 X 13.6 = 7.85 person trips
day
Both home-based non-work (HBNW) and non-home-based (NHB) trips were
included in the non-work trip category for this analysis. Table 1-1 further
indicates that 53 percent of the total person trips are home-based non-
work and 21 percent ar,e non-home based. Table 1-2 (Table 3 from NCHRP
187) indicates that for large urban areas, the number of auto driver
trips is approximately 53 percent of homebased non-work trips and 60
percent of non-home based trips. The average daily number of non-work
auto trips per household in the corridor can be approximated as:
7.85 Total Trips .53 HBNW Trips .53 Auto Trips 7.85 Total Trips
Day X Total Trips X HBNW Trips Day
,21 NHB Trips
Total Trips
.60 Auto Trips
NHB Trips
3.19 Auto Trips
Day
-------
TABLE 1-2
TRIP-GENERATION PARAMETERS
1-14
I PART A - TRIP PRODUCTION ESTIMATES
Urbanized
Area
Population
50,000- 100,000
100,000- 250,000
250,000- 750,000
750,000-2,000,000
Average
Daily
Person
Trips
Per HT
1>»
ll»
12
8
% Average Daily
Person Trips
by Mode11
Public
Transit
2
6
8
13
Auto
Passenger
bO
30
31
30
Auto
Driver
58
6k
61
57
% Average Daily
Person Trips
by Purpose*
HBW
16
20
20
25
HBNV
61
57
55
5"t
NHB
23
23
25
21
Auto Person Trips
as a % of Total
Person Trips c
HEW
96
88
8U
7U
HBNV
99
97
96
93
NHB
98
9U
92
86
Auto Driver Trips
as a % of Total
Person Trips'*
HBW
70
64
62
56
HBNV
54
54
54
53
NEB
68
66
64
60
[PART B - USEFUL CHARACTERISTICS FOR TRIP ESTIMATION
Urbanized
Area
Population
50,000- 100,000
100,000- 250,000
250,000- 750,000
750,000-2,000,000
External Travel Characteristics
% of Total External
Trips Passing
Through Area
21
15
10
It
f of Total External
Trips to
the CBDe
22
22
18
12
Total Areavide
Truck Trips
as a % of
Areawide Auto
Driver Trips '
27
17
16
16
PART C - TRIP ATTRACTION ESTIMATING RELATIONSHIPS g
(All Population Groupings for either Vehicle or Person Trips)
TO ESTIMATE TRIP ATTRACTIONS FOR AN ANALYSIS AREA
HBW Trip Attractions » Fx Q.7 (Analysis Area Total
/Analysis Area\
HBNW Trip Attractions « Fg 10.01 Retail j +
V Employment /
/Analysis Area\
NHB Trip Attractions « Fj 2.0 I Retail j +
\ Employment /
Where: ?j, Fg and Fj are areavlde control factors.
TO DEVELOP AREAWIDE CONTROL FACTORS, USE:
T m Areawide Productions for HBW Trips
1 1.7 (Areavlde Total Employment)
- w Areavlde Productions for HBNW Trips
2 f~ f Areavide \ / Areavide \
10.0 I Retail j + 0.5 [Non-Retail] +
[_ \Employment/ \Employment/
. „ Areavlde Productions for NHB Trips
3 r / Areavide \ / Areavlde \
2.0 I Retail ] * 2.5 I Non-Retail) +
£_ \£mployBeny yinploymeny
USE:
Employment }J
/Analysis Area\ Analysis Ar
0.5l Non-Retail 1 + 1.0 1 Dwelling
\^ Employment / ^ Units
/Analysis Area\ Analysis AJ
2.5 1 Non-Retail 1 + 0.5 1 Dwelling
\Employment / \ Units
/AreavideV~|
1.0 (Dwelling)
\ Units /J
1
1
/AreawldeVI
0.5 Dwelling)
\ Units /J
a. Froa Table 2.
b. Source: Kef. (19).
c. Source: Orlgin-D«stin>tion Surveys.
d. Calculated using c and Table 12, Chapter Five.
e. Source: Ref. (20).
f. Source: Ref. (19).
g. Source: Office of Planning Methodology and Technical Support, OMTA.
Source: NCHRP Report #187, Quick Response Urban Travel Estimation Techniques
and Transferable Parameters.
-------
1-15
All of these trips must be made by households owning autos. Since
11.5 percent of the corridor's households own no autos, the average daily
frequency of non-work auto trips for auto-owning households must be:
3.19 T .885 - 3.61 tr.ips * 2 trips - 1.80 round trips
day round trip day
This figure is entered on column 4 of Worksheet I for the auto own-
ing subgroups,while "0" is entered for the two subgroups without autos.
Base Work Trip Modal Shares
Screenline counts on the freeway and major arterials leading from
the corridor into the central city and alighting counts on transit routes
serving the city from the corridor were conducted to determine:
• The total volume of auto person trips from the corridor
to the central city;
• The volume of transit riders in the corridor; and
• The distribution of auto occupancy for central city work trips
The information from these counts may be used to determine the base
mode shares required for the manual worksheet analysis.
Counts were conducted on a typical workday and yielded the following
results:
Total Auto Person Trips - 48,910
One Person - 23,230
Two Person - 16,042
Three Person - 5,016
Four+ Person - 4,622
Average Size of Four+ Carpool - 4.60
-------
1-16
Total Auto Person Trips to Fringe Area - 19,564 (40% of Total)
Total Auto Person Trips to CBD - 29,346 (60% of Total)
Total Transit Passengers to Fringe - 5,707 (40% of Total)
Total Transit Passengers to CBD - 8,560 (60% of Total)
License plate numbers were recorded for a sample of the autos pass-
ing the screenline and vehicle registrations were checked to determine
the residence location of the driver. It was found that 45 percent of
the vehicles passing the screenline had origins outside the corridor, or
in locations within the corridor which would be unaffected by the HOV
lanes or improved bus service (areas A on the map of the corridor shown in
Figure 1-4). Similar checks were made for bus passengers using boarding
counts on a sample of the routes, yielding an estimate of 10 percent of
downtown bound passengers on corridor routes who would be unaffected by
the proposed measures.
The person trip counts and the percentages of unaffected trips yielded
the following estimates of travel volumes and mode shares for central city
bound trips in the corridor which could potentially be affected by the
HOV lane proposal:
Total
Share
m
Auto Trips (Occupancy)
I
12,776
32.2
2
8,823
22.3
3
2,759
6.9
4+
2,542
6.4
Transit
Trips
12,840
32.2
Total
Trips
39,740
100.0
Worksheet I provides space for recording base mode shares of drive-
alone, carpoolj transit, and other modes only. Because two- and three-
person carpools will be affected differently from four-person-and-larger
carpools, the base shares for both of these size categories must be
-------
1-17
0 miles
FIGURE 1-4
Study Corridor and Proposed HOV Lanes
-------
1-18
recorded for this analysis. It was assumed that the mode shares for CBD and
fringe area workers were approximately the same. The average mode
shares for all households in the corridor are:
Drive Alone (DA): 32.2%
Two-, Three-person Carpools(SR2 ) : 29.2%
Four-plus person Carpools(SR^+) : 6.4%
Transit 32.2%
Adjustments must be made to these average mode shares when record-
ing the base work trip mode shares for each market segment since the sub-
group not owning autos can not use the drive-alone mode. These adjust-
ments were accomplished as follows:
T SR
a a a a
Where:
DA = Drive Alone share for zero auto-owning households
o
T = Transit Share for the Zero Auto-Owning Subgroup
o
SR = Shared-Ride Share (all carpools)
3
T = Average Transit Share
SR = Shared-Ride Share for Zero Auto-Owning Subgroup
BA, = a SR,
1
"l
Where:
DA.. - Drive-Alone Share for Auto-Owning Subgroup
DA = Average Drive-Alone Share
a
-------
1-19
SR., - Shared-Ride Share for Auto-Owning Subgroup
T. = Transit Share for Auto-Owning Subgroup
Fo « Fraction of Households Owning Zero Autos
F = Fraction of Households Owning One or More Autos
Applying these formulas to the average mode shares for the corridor,
and assuming that the breakdown of carpools by size (82.0 percent of
carpools are two or three persons) is not dependent on auto ownership
yielded the following mode shares for each market segment:
Subgroups 2 and 4: DA_ = 0
'356 = .525
.322 + .356
T = -322 = 475
0 .322 + .356 '*n
SR23= .525(.820) = .431
SR4+= .525(.18) = .094
Subgroups 1 and 3: DA. = -^fff = .364
i.
SP - .356 - .525(.115) _ .,,,
SR1 ~ 7885 - ~ >334
_ .322 - .475C.115)
1 ~ .885 ~ •***
SR23= .334 (.82) = .274
SRA+= .334 (.18) = .060
The mode shares for each subgroup are entered on colums 5 through 8 on
Worksheet I. Note that the share for other modes is zero in this case.
Also note that if the average mode shares had been different for the
fringe and CBD areas , the above calculations would have been carried out
separately for each subgroup.
-------
1-20
Average Carpool Size
The average carpool size, required in column 9, may be calculated
from the distribution of auto occupancies available in the screenline
count data. Once again, distinction must be made between two- and three-
person, and larger carpoola. The average size for two- and three-person
carpools is;
2( -223 -069 ,
223 + .069 \223 + .069 ' ~ '
The average occupancy of larger carpools was found to be 4.6 according
to the screenline counts .
Average Trip Length
The final pieces of base data required for the analysis are work and
non-work trip lengths. In this analysis, average work trip length was
determined using approximate measurements on a map of the area and a know-
ledge of the distribution of traffic entering the freeway along its
length. Figure 1-4 shows the planned entrance ramps for the freeway HOV
lanes. The percentage of the total freeway traffic currently entering
at each point is known to be:
Ramp 1 (8.5 miles from the CBD) - 40%
Ramp 2 (11.5 miles) - 20%
Ramp 3 (15.0 miles) - 20%
Ramp 4 and beyond (16.5 miles) - 20%
The fringe area and the CBD are about 5 miles apart, so the freeway
portion of average trip lengths in the corridor is 7 miles for
fringe area work trips and 12 miles for CBD work trips. The
-------
1-21
population distribution in the corridor and the location of major arterials
paralleling the freeway indicate that the average freeway and arterial
access distance is about 2 miles, and central city circulation accounts
for about 0.5 miles on the average. The total trip lengths of 14.5 and
9.5 miles for CBD and fringe trips are recorded in column 10 of Worksheet I.
Non-work trip length data was not available for the area, so a 3-mile
trip length was assumed for auto-owning households based on national aver-
age trip length figures. Non-work trip lengths for households without autos
are not needed since it is assumed that these households make no auto trips.
Average Daily VMT
The information required in the final three columns of Worksheet I
will be calculated using Worksheet I-A. Worksheet I-A is executed for
each of the four population subgroups using data from Worksheet I. The
entries for subgroup 1 are shown as an example. As in the case of the
base data, separate consideration of two-and three-person and four-plus
person carpools is necessary. Base autos used per worker for subgroup 1
is calculated:
.274 (2,3 carpool share) T 2.2 /persons \ + .060 (4+ carpool share) -f
V.2,3 carpool/
4.6 /'Persons \ + .364 (Drive-alone share) = .502 autos
\4+ Carpool/ worker
Similar calculations are made for the other three subgroups.
Changes in Transportation Level of Service (LOS)
Worksheet II-A is used with the slight modifications shown to record
LOS changes for carpools of differing size. The base carpool size shares
-------
POLICY: HOV Lan<
Population
Subgroup*
1
2
3
4
*1 CBD Destir
2 CBD Destii
3 Ftliigfc! DBtJ
4 Fringe Des
-i
II
0)
N »
• *
*
2.2 ./
/ \.6
2 2 /
/
/4.6
2 ? >X
'/
X^4.6
2X
/ 4.6
Average
Trip Length
I1
14.5
14.5
9.5
9.5
o *
i
I -
3
-
3
-
Average Daily VMT
I
\
I
I
m
s
**Two shared-ride mode, shares must be recorded for each subgroup:
***Two average carpool sizes, 2 and 3 person, and 4+ person are required also:
X. = 2 and 3 person carpools
Y = 4+ person carpools
-------
I-A. BASE VMT
1-23
Base
Shared Ride
Modal Share
.274
.060
Base
Autos Used
per Workers
RK TRIP VMT
Base Average
CarpodSize
2-2 / ft H
4.6 ^
No. of Workers
per Household
1.2 )
Base
Drive Alone
Modal Share
.364
Work Trip
Length
C 14.5
POLICY: HOV Lane
SUBGROUP: 1
Base
Autos Used
C Worker
,502
Base
Household
Work VMT
X 2X) = 17.47
2. BASE HOUSEHOLD NON-WORK TRIP VMT
No. of Non-Work
Auto Trips
Non-Work
Trip Length
3. BASE HOUSEHOLD TOTAL VMT
Base
Household Non-
Work VMT
-------
1-24
and average size of four-plus person carpools required by Worksheet
II-A have already been determined. The LOS changes must be estimated
separately for each population subgroup and carpool size and recorded
on Worksheet II-A.
1. GBP Workers (Groups 1 and 2)
Two- and Three-person Carpools - All autos will experience delays
due to construction in the downtown area. The average diversion of autos
due to the construction is expected to be about .5 mile. Current travel
speeds downtown are 17 mph, so the diversion translates into 1.76 minutes
in each direction or 3.5 minutes of additional travel time for the round
trip. This is the only impact expected for two- and three-person carpools
since they cannot use the HOV lane.
Four-plus person carpools - These autos will experience the 3.5-
minute delay downtown, but will also save time because they can use the HOV
lane. As noted earlier, the average length of the freeway section of
downtown-bound trips is 12 miles, or 24 miles round trip. Because of con-
gestion during the peak hour, the average speed on the freeway is now 22 mph.
The average speed for the HOV lanes is expected to be about 50 mph. The
average time saved on the freeway for four-plus person carpools will thus be:
24 miles 60 min. 24 miles 60 min 0, , . .Q
-r-r : x —: - T^ T— x —: « 35.7 minutes
22 mph hr. 50 mph hr
To obtain these time savings, some of the carpoolers will have to go
out of their way to get to the freeway (rather than using one of the major
arterials in the corridor which would provide a more direct route to the
central city). Standard diversion curves may be used to determine the degree
-------
1-25
to which the HOV lanes will cause carpoolers to divert from their current
1 2
routes on the. major arterials. Curves developed by Moskowitz, based
on time savings via a freeway versus arterials, and the extra distance
required to use the freeway, are particularly useful for this analysis
(See Figure 1-5). In order to apply the curves, a knowledge of the distri-
bution of the population by distance from the freeway is necessary. Land
use maps of the corridor indicate that in 1982, the following distribution
will exist:
Distance from Percent of Corridor
Freeway Population
0-1 miles 30.0%
1-2 miles 26.3%
2-5 miles 30.2%
5+ miles 13.5%
Currently, 26 percent of the corridor traffic to the central city uses
the freeway. Given the current freeway speed of 22 mph and arterial speeds
averaging 20 mph, it is assumed that all of the travellers currently using
the freeway must live within one mile of the freeway, since there are no
time savings associated with diversions to the freeway from greater distances.
However, with the institution of reserved lanes for buses and carpools, di-
version can be expected to occur to a much greater degree.
For a compilation of typical diversion curves, see Martin, B.V., F.W. Memmott,
and A.J. Bone, Principals and Techniques of Predicting Future Demand for
Urban Area Transportation, January, 1963.
2
K. Moskowitz, "California Method of Assigning Diverted Traffic to Proposed
Freeways," Highway Research Bulletin 130. 1956.
-------
1-26
+ 2
1
-2 0+2 +lj +6 +8 +10 +12
Distance Saved Using Expressway-Miles
FIGURE 1-5
Freeway/Arterial Diversion Curves
Source: Moskowitz. HRB Bulletin 130
-------
1-27
The net one-way time savings for each of the above distance categories
may be calculated:
36.7 minutes
N
20 mph
x 60
where:
the average distance from the freeway
These time savings may then be used in combination with the extra distance
associated with freeway travel to enter the route diversion curves and
determine the proportion of the population in each distance category which
will use the freeway. The net travel time savings, extra distance travelled
and percent of the four-plus person carpools within each distance category
diverting to the freeway are shown below:
Distance
Category
0-1
1-2
2-5
5+
Net Travel
Time Savings
16.9 min.
13.9 min.
2.9 min.
0 min.
Extra Distance
Travelled
0.5 mile
1.5 miles
3.5 miles
6.5 miles
Percent of 4+
Person Carpools Using
Freeway HOV Lanes
100%
80%
53%
0%
The average round-trip travel time savings for all carpools in the corridor
may be calculated from the percent diversion for each distance category,
the percent of the population in each category, and the round-trip net time
savings, recalling that 26 percent of the population (all within one mile
of the HOV lanes) would not divert from their current route at all:
-------
1-28
.26(36.7) + 1.0(.30 - .26) (16.9)(2) +
.80(.263)(13.9)(2) + .53(.302)(7.9)(2) = 19.2 minutes
These carpoolers will be delayed, along with all other auto traffic in the
downtown area, by 3.5 minutes. Therefore, the net time savings will be 15.7
minutes overall for four-plus person carpools bound to the CBD. This is
entered under AlVTT for four-plus person carpools on Worksheet II-A for
sub-groups 1 and 2.
2. Fringe Area Workers (groups 3 and 4)
Two- and Three-person Carpools - Approximately half of the
employees in the fringe area will be affected by the employer carpool
parking policy. Among these employees, single-occupant auto drivers will
have longer walks, on the average, from their parking spaces to their
offices, while carpoolers will have shorter walks. The current average
walk distance for all employees is approximately 700 feet* Under the pre-
ferential policy, the average walk distance for all carpoolers will be
350 feet. Walking speed averages 264 feet per minute (3 mph), so the one-
way walk time savings for carpoolers averages 1.33 minutes. This savings
is doubled for the round trip, but only half the fringe area employees
would be affected, so the average impact for all employees is a 1.33 minute
round trip savings. This is recorded on Worksheet II-A in column 5—AOVTT.
No other changes in LOS are expected for two- and three-person carpoolers.
Four-plus Person Carpools - These travellers will enjoy the 1.33-
minute OVTT savings from preferential parking, plus travel time savings on
the HOV lanes of the freeway. The freeway travel time savings, based on a
-------
1-29
II-A. CHANGES IN TRANSPORTATION LEVEL OF SERVICE BY CARPOOL SIZE
5
•g
5
•£
8
>
s
"5
j
S
*•*
S
p
Q.
8
£
_C
tf>
0)
i
O
"5
o
O
_
w
o>
"•
*
8
Q.
(0
0
o
1
CO
CM
Promotion and
Matching
Incentives
0,1
Out-of -Pocket
Travel Cost
AOPTC3+
Out-of-Vehide
Travel Time
AOVTT3+
In-Vehide
Travel Time
AIVTT3+
Promotion and
Matching
Incentives
0,1
Out-of -Pocket
Travel Cost
AOPTC2
<
Out-of-Vehide
Travel Time
AOVTT2
In-Vehide
min.
Travel Time
AIVTT2
Average Size of
min.
4+ Person Carpools
II
0 |
N
(A
Population
Subgroup
i
&
c
en O
** (0
CN) d)
4
o
o
-
i— i
i
o
o
o
Ul
CO
vO
-a-
00
i-H
•
CM
00
oo
1— I
o
o
^
1 — 1
1
o
o
o
"?
CO
vO
•*
oo
<—<
CM
00
oo
CM
o
o
CO
. — 1
1
CO
1
o
o
CO
CO
1
o
vO
**
oo
,-H
CM
oo
oo
CO
o
o
CO
CO
i-H
I
VO
CO
1
o
o
CO
1
o
vO
"*
oo
*"t
CM
oo
oo
•
"*
0)
c
-------
1-30
7-mile average distance is estimated to be 22.9 minutes for fringe area
four-plus person carpoolers. Using the same distance categories, diver-
sion curves, and calculations as in the CBD-bound case, the following
results are obtained:
Distance
Category
0-1
1-2
2-5
5+
Net Travel
Time Savings
10 min.
7 min.
1 min.
3.5 min.
Extra Distance
Travelled
0.5 mile
1.5 miles
3.5 miles
5.0 miles
Percent of 4+
Person Carpools Using
Freeway HOV Lanes
90%
70%
18%
0%
Average Round trip savings:
.26(22.9) + .90(.30 - .26)(20) +
.70(.263)(14) + .18(.302)(2) = 9.36 minutes
This savings figure is entered under AIVTT for four-plus person carpools
for subgroups 3 and A.
As illustrated in Figure 1-2, changes in carpool size as a result of
the differential impact of the transportation measures on two-, three-, and
four-plus person carpools must now be determined. The new proportions of
carpools by size will then be used to calculate the average level-of-service
impact on the carpool mode as a whole. Worksheet III-A is used to estimate
the changes in the distribution of carpool sizes. The analysis of this
particular carpool/bus lane proposal requires a non-standard use of
Worksheet III-A, since only carpools with four or more members are eligible
to use the reserved-lane facility. The only change required is to include
-------
1-31
two- and three-person carpools in the calculations of change in utility for
carpools with two members, and only carpools of four or more members in the
three-plus person carpool calculations. The average two- and three-person
carpool size of 2.2 is substituted for the two-person carpool size, and the
average four-plus carpool size of 4.6 is used for the average three-plus
carpool size in the calculation of overall average carpool size.
Worksheet III-A is used once for each of the population subgroups.
The calculations for subgroups 1 and 2 are provided as an example (the
calculations for groups 1 and 2, and groups 3 and 4 are identical). All
of the data required for Worksheet III-A has been recorded on previous
worksheets.
The revised changes in shared-ride LOS calculated in Section 3 of
Worksheet III-A represent the average changes in service quality for the
ridesharing mode, given the changes in carpool size which are likely to
occur because of the difference in impact of the planned transportation
measures on four-plus person carpools relative to smaller carpools.
These average shared-ride LOS changes are now entered on Worksheet II;
along with the LOS changes for the other modes described in the following
sections.
Drive Alone
The only impact on drive-alone trips to the CBD will be the 3.5-minute
delay for downtown circulation due to construction. This will affect
subgroup 1 only, since subgroup 2 does not have the drive-alone option
available.
-------
1-32
!!l-A. ESTIMATION OF CHANGES IN CARPOOL
SIZE
POLICY1 HOV Lane
SUBGROUP:
1/2
1. CHANGE IN UTILITY FOR EACH CARPOOL SIZE AryTT2
2,3 Person Carpool
Trip Length AOVTT2
Income AOPTC2 Carpool Size
+ -29.0 * 13,440 X 0
Imi 1.1 . i.ijiiiiii.iin .. !•• I tmiiuL.il 11 .1 11II I Him I.IF! BmMMim^H
TOTAL CHANGE j"-053
4 + Person Carpool
Average 4+
Carpool Size
TOTAL CHANGE I+.236
2. REVISED CARPOOL-SIZE SHARES
Total Share Revised Share
2,3 Person I.
3. REVISED CHANGES IN SHARED
RIDE LEVEL-OF-SERVICE (LOS)
Share = | 1-006 I
Revised 2,3 2,3 Person Revised14+ 4+.Person Shared Ride
rwviOWJ -*• > -* *- 9 -~f • TSTx ..— -•——-- -T -, m ~-~~.
Person Share ALPS Person Share ALPS
AOPTC I .773
Revised 2 23 Person Revised 4+ Average 4* Revised Average
,
Person Share Carpool Size Person Share Carpool Size
Carpool Size
-------
1-33
Half of the drive-alone commuters bound to the fringe area will be
affected by the preferential carpool parking policy planned by some employers.
The average walking distance for those who do not carpool is expected to
increase to 1,500 feet from the current average of 700 feet. At 264 feet
per minute (3 mph), the 800-foot increase in walking distance will delay
drive-alone commuters 3.0 minutes. Half the employees will experience
the delay twice daily, so the round-trip average change in OVTT is +3.0
minutes. Again, this will only impact auto-owning households (subgroup 3).
These LOS changes are entered on Worksheet II under Drive-Alone.
Transit
Two important changes in transit level of service are associated with
the proposed plan. Travel times will be reduced on those routes which use
the HOV lanes, and wait times will be lowered due to higher frequency
service.
It is assumed that the distribution of bus passengers entering the
freeway high-occupancy vehicle lanes will be similar to that for carpools
and autos, and that the average one-way distance for buses going to the
CBD will be 12 miles on the HOV lanes, and the average to the fringe area
7 miles. It is also assumed that buses will travel 50 mph on the lanes,
compared to 22 mph under current conditions. According to the plans of
the bus operator, approximately 60 percent of the buses serving the corridor
will use the HOV lanes. Therefore, the average time savings for central
city-bound bus passengers in the corridor will be:
-------
1-34
run Tt^..r^-c12 miles 12 miles. ,_ minutes
CBD-Bound: (.-TT: r— — TT : ) x 60 , • ••—"• n •
22 mph 50 mph hours
x 0.6 (fraction using HOV lanes) x 2(°ne"W^y tji?s)
round trip '
=22.0 minutes
Similarly,
-, . „ , , 7 miles 7 miles x ,_
Fringe area-Bound: ( - — — ) x 60 minutes per hour
2.2. mph 50 mph
x 0.6 (fraction using HOV lanes) x 2(°ne^y t^'iPs)
6 round trip
=12.8 minutes
These average savings in bus IVTT will be enjoyed by all residents in the
corridor, regardless of auto ownership.
The bus fleet serving the corridor will be doubled to take advantage
of the HOV lanes and encourage their use by commuters in express buses.
On the average, it is estimated that headways will be cut in half, although
headways on some routes will be improved more than on others. The average
headway for radial routes in the corridor is now 20 minutes, and should be
reduced to 10 minutes when the new services are in place. Assuming waiting
times are approximately half the headway, a 5—minute average reduction in
waiting times (OVTT) for all central city-bound bus passengers will
occur. No changes in bus fares are planned.
The average drive-alone,shared-ride, and transit changes in level of
service should now be entered on Worksheet II. This completes the prepara-
tion of input data for the analysis, and the analyst is ready to begin
Step 5 using Worksheet III.
-------
II. CHANGES IN TRANSPORTATION LEVEL C^JERVICE (ROUND TRIP)
'55
H
0
oc
1
s
i
^
Out-of-Pocket
Travel Cost
AOPTCt *
Out-of-Vehide
Travel Time
A OVTT, min.
In-Vehicle
Travel Time
AIVTTt min.
Carpool Prom-
otion & Match-
ing Incentives
0,1
Out-of-Pocket
Travel Cost
AOPTCs, 4
Out-of-Vehide
Travel Time
AOPTCsr min.
In-Vehicle
Travel Time
A IVTTW min.
Out-of-Pocket
Travel Cost
AOPTCda $
Out-of-VeNcle
Travel Time
AOVTTda min.
In-Vehicle
Travel Time
AlVTT^ min.
Population
Subgroup
o
o
I
o
CM
CM
I
O
o
o
vC
oo
*
I
o
o
m
CO
+
T-t
O
o
m
i
o
CM
CM
I
O
O
o
<£>
00
I
1
1
1
CM
O
O
m
i
oo
CM
i-H
1
O
O
CO
CO
•— I
1
o\
oo
1
o
o
CO
+
o
CO
o
o
t
m
i
oo
CM
.-1
1
O
o
CO
CO
t— 1
1
00
.— I
1
1
1
0)
c
O
o.
-------
1-36
G. Description of Model Application
Steps 5 and 6 of the analysis are accomplished using Worksheets III,
IVs V, and V-A. Worksheet III is executed four times, once for each of
the four population subgroups being analyzed to determine new mode shares
as a consequence of the changes in LOS. The worksheet for subgroup 1
provides an example of the flow of data from previous worksheets and the
calculations required for each subgroup.
The change in level of service required for each mode is taken from
Worksheet II for each market segment. In this case, drive-alone AIVTT is
3.5, shared-ride AIVTT is -.86, transit AIVTT is -22.0, and transit AOVTT
is -5.0. All other level-of-service changes are zero. The required trip
length and income statistics for subgroup 1 are obtained from Worksheet I,
along with the base mode shares required in Part 2 of the Worksheet. The
shared-ride share used in Worksheet III is the combined two-, three-, and
four-plus person share (33.4 percent for sub-group 1.). The A utility
inputs in Part 2 of Worksheet III are the "total changes" calculated for
each mode in Part 1 of the same worksheet. The mode shares calculated
in Part 2 reflect the predicted travel response of the particular subgroup
to the proposed plans. Thus it is possible through the use of the manual
worksheets to isolate the impacts of proposed transportation measures on
particular population groups.
The new mode shares calculated using Worksheet III are then used,
along with data from Worksheet 1-A (base work and non-work VMT, trip
lengths, and workers per household) and Worksheet III-A (revised average
-------
1-37
III. ESTIMATION OF REVISED WORK-TRIP
MODAL SHARES
t CHANGE IN UTILITY FOR EACH MODE
Drive Atone AUTILITY ==
1
4 -.16 •*-
POLICY:
SUBGROUP
-.015 ;
'rip Length
14.5 ;
Income
4 -29.0 -r-
HOV Lane
1
AivUste
>< +3.5
AOVTTda
K o
AOPTCda
13,440 |X| 0
__
TOTAL CHANGE
Shared Ride AUTILITY = ~'015 X
-.86
Trip Length AOVTTsr
4- -.16 -f- 14.5 X
0
Income AOPTCsr
4- -29.0 4 13,440 X
0 -
Incentives
4- -29 X
—
Average Car
pool Size
f- 2.5
_
-0.053
0
0
-0.053
+0.013
0
o
0
TOTAL CHANGE
IVTT,
Transit
AUTILITY
2. REVISED MODAL SHARE
Base Modal Share AUtility
Revised Share
Drive Alone = I -364 |X EXP
Shared Ride ~
Transit
Vanpool
Other
Tbtal =
-------
1-38
carpool size) in executing Worksheet IV for each population subgroup.
The revised work and non-work VMT calculated for each subgroup on Work-
sheet IV is then summarized and translated into average corridor-wide
changes in work and non-work VMT on Worksheet V. Worksheet V-A is used
to summarize the predicted mode share impacts and calculate corridor-
wide average mode shares under the new conditions. An example of the
use of Worksheet IV is provided for subgroup 1, and the attached Work-
sheets V and V-A show the predicted impacts of the proposed HOV lane
plans for all population subgroups in the corridor.
In order to isolate the impacts of the proposed plans on four-plus
person carpools, the revised two- and three-person, and four-plus person
carpool shares calculated in Worksheet III-A for each market segment must
be applied to the revised overall carpool share calculated on Worksheet III .
This is done on the sample Worksheet V-A. Note that the vanpool sections
of Worksheets III, IV, V, and V-A are ignored for this analysis. Also, the
non-work VMT section of Worksheet IV can be ignored for subgroups 2 and 4
since these subgroups do not make non-work auto trips.
H. Impact Assessment
To determine the impacts of the mode shifts and percentage changes in
VMT summarized on Worksheets V and V-A on HC and NOx emissions in the cor-
rider (Step 7 of the analysis), total VMT changes must be calculated. The
distribution of auto person trips by auto occupancy level (pg. 11) yields:
= 18j743 auto trlps
We know that 40 percent of the work trips average 9.5 miles one way and
60 percent average 14.5 miles one way. The total base work VMT is then:
-------
1-39
IV. ESTIMATION OF CHANGES IN VMT
1. REVISED HOUSEHOLD WORK TRIP VMT
Revised Revised Revised
Shared Ride Average Drive Alone
Modal Share Carpool Size Modal Share
Revised
Vanpool
Modal Share
1 - *
.300
Revised
Autos Used
per Worker
.426
+ | 2.5
+
No. of Workers
per Household
X | 1.2
X
Average
Vanpool No. of Workers
Occupancy per Household
x| -
X
.306
Work Trip
Length
14.5
Work Trip
Length
—
>
X
X
POLICY HOV Lane
SUBGROUP 1
Revised
Autos Used
Per Worker
.426 |
2.0
Vanpool
Circurty
Factor
- |X| 2.0
Auto VMT
= 14.82
Vanpool
VMT
=
Total Revised
Household
Work VMT
14.82
2. CHANGE IN HOUSEHOLD WORK TRIP VMT
Revised Household Base Work Change in
Work-Trip VMT Trip VMT Work Trip VMT
| 14.82
—
17.47
-
-2.65
j.
Base Work
Trip VMT
17.47 |x
100
I Change in
rk Trip VMT
= -15.17
3. REVISED HOUSEHOLD NON-WORK TRIP VMT
Base Revised
Autos Used Autos Used No. of Workers
/ per Worker per Worker ^ per Household
Change in
Autos Remaining
per Household
._..„_ in ..
jtos Remaining
per Household
(Per.
P
0912
.426
1.0
x
)
I-
1.
3ase
Work
10.
2
Non-
VMT
80
+.0912
Revised Non-
Work VMT
10.88
4. CHANGE IN TOTAL HOUSEHOLD VMT FROM BASE
Revised Work Revised Non- Revised
Trip VMT Work VMT Total VMT
14.82
4
10.88
Change in
Total VMT
X 100
Change in
Total VMT
= -9.09
-------
POLICY'• HOV T.nnp
Work
Trip :
Total:
Population
Subgroup
1
2
3
4
AVERAGE HOUSEHOLD VMT
Work Trip VMT
Base
17.47
7.52
11.45
4.77
A
-2.65
-1.49
-1.28
-0.42
Total VMT
Base
28.27
7.52
22.25
4.77
A
-2.57
-1.49
-1.22
-0.42
X
o-rt(/>
o o
3 C
•o
.531
.069
.354
.046
Average
AVMT
Average
VMT
AVERAGE -
Subgroup
%AVMT
NORMALIZED HOUSEHOLD VMT
Work Trip VMT
Base
9.28
0.52
4.05
0.22
14.07
A
-1.41
-0.10
-0.45
-0.02
-1.98
Total VMT
Base
15.01
0.52
7.88
0.22
23.63
A
-1.36
-0.10
-0.43
-0.02
-1.91
Percent Change
in VMT
Work
-15.17
-19.81
-11.18
-8.81
Total
-9.09
-19.81
-5.5
-8.81
i
o
i\
o
o
m
CO
h-f
I
•P-
O
-------
POLICY
HOV Lane
II
1
2
3
4
Drive
B*
.364
0
.364
0
* B=Base
Alone
R*
.306
0
.316
0
AVERAGE MODAL SHARE
Shared Ride
B**
.274
.060
.431
.094
.274
.060
.431
.094
t* X X=
Y Y=
R
.232
.068
.335
.098
.256
.065
.374
.095
Transit
B
.302
.475
.302
.475
R
.394
.567
.363
.531
Other
B
0
0
0
0
R
0
0
0
0
Vanpooling
B
0
0
0
0
R
0
0
0
0
=2,3 Person Carpools AVERAGE =
=4+ Person Carpools
s
S.""
^3
§!a
8LO
3-3
•p
.531
.069
.354
.046
z-
ubqroi
ID
NORMALIZED MODAL SHARE
Drive Alone
B
.193
0
.129
0
.322
R
.162
0
.112
0
.274
Shared Ride
B
.145
.032
.030
.097
.021
.020
.004
.292
.063
R
.123
.036
.023
.091
.023
.017
.004
.254
.070
Transit
B
.160
.033
.107
.022
.322
R
.209
.039
.129
.024
.401
Other
B
—
—
—
—
—
R
—
—
—
—
—
Vanpooling
B
—
—
—
—
—
R
—
—
—
—
—
o
o
m
-------
1-42
(0.4 x 9.5 + 0.6 x 14.5) x 18,743 x 2 = 468,575 VMT/day
This VMT must be apportioned among the four population subgroups. The
total work VMT accounted for by each market segment is given by:
VMT = F x HHVMT T HHVMT X VMT
a a a t t
where :
VMT - Work VMT for subgroup A
cl
F = Fraction of population in subgroup A
Si
HHVMT = Household work VMT for subgroup A (from Worksheet I-A)
£1
HHVMT = Average Household Work VMT for all subgroups
VMT = Total work VMT for all subgroups
The average household work VMT for the corridor (HHVMT ) is calculated
by adding the product of HHVMT and F for all subgroups:
EL
17.47(.531) + 7,52(.069) + 11.45(.354) + 4.77(.046) = 14.07 miles/day
The subgroup daily work VMT may now be calculated:
TT 1 7 tl.~l
VMT1 = 14 07 X 468»575 = 308,939 miles/day
Similar calculations for each subgroup yield work trip base VMT which may
be used in conjunction with the percent change in VMT figures summarized
on Worksheet IV for each subgroup to obtain total VMT changes:
-------
1-43
Subgroup
1
2
3
4
Base
Work VMT
308,979 mi.
17,283 mi.
135,005 mi.
7,308 mi.
Proportional
Changes in
Work VMT
-.152
-.198
-.112
-.088
Revised
Work VMT
262,014
13,861
119,884
6,665
Similar estimates must be made for non-work VMT as well, but they
need only be made for subgroups 1 and 3. Worksheet I-A indicates that
the base household non-work VMT for both subgroups 1 and 3 is 10.8 miles.
The base total non-work VMT for these groups may be found by simply taking
the ratios of base household non-work VMT to work VMT and multiplying by
the subgroup work VMT totals above. Worksheet IV then provides the propor-
tional change in non-work VMT for the two subgroups required to develop
the following estimates of non-work VMT changes:
Subgroup
1
3
Base
Non-Work VMT
191,012
127,341
Proportional
Changes
+ .007
+ .005
Total
Change
+ 1,337
+ 637
The emissions worksheets method (4.1) may now be used to forecast HC and
NOx emissions changes in the corridor. Reference to Emissions Tables
D.I through D.6 in Volume I is required.
Worksheet VI-A is used twice to calculate base and revised corridor
VMT and the number of trips made by trip purpose using the base and re-
vised VMT by population subgroup as calculated above. The average travel
-------
1-44
speeds for each population subgroup are assumed to be the same for the
base and revised cases. Alternatively, actual average vehicle speeds
could be calculated based on the mode split between 4+ person carpools
and autos, and knowledge of trip speed and length for each mode. Based
on this assumption, start-up and travel emissions may be calculated
for the corridor as a whole, rather than for each market segment, using
Worksheets VI-C and VI-D. Worksheet VI-B is used to determine the per-
centage of cold starts for work and non-work trips for the study year.
Because parking duration data is not available for the study corridor,
Emissions Table D.I provides the necessary 1982 cold start percentages.
Worksheets VI-C and VI-D are each executed twice, once for the
base case, and once for the revised situation. The use of the worksheets
is straightforward, using data from the Emissions tables and work-
sheets as noted. A summer ambient temperature of 75 degrees is assumed
based on historical data.
The cold start HC factor for work trips is obtained by simple linear
interpolation in Emissions Table D.2. TWO dimensional interpolation is re-
quired to determine the cold start HC factor for non-work trips. The procedure
for accomplishing two (and three) dimensional interpolation is described
in Vol. I, SectionD.4. The calculations for the study corridor appear
below, with steps corresponding to those outlined in Section D.4.
-------
1-45
1. Identify bracketing values.
Xj_ - 70,
Z1 = 50,
X2 = 80
Z2 = 60
2. Read emissions factors for bracketing combinations from Table D.2.
Combination
70 , 50
70 , 60
80 , 50
80 , 60
Emission Factor
6.2
6.9
5.7
6.2
3.
Compute fv and f .
A L*
75 - 70
80
57
70
50
60 - 50
.5 (Ambient temperature of 75 degrees)
•7 (57 percent cold starts)
4. Compute emissions factor.
ef = .5 x .3 x 6.2 +
.5 x .7 x 6.9 +
.5 x .3 x 5.7 -f
.5 x .7 x 6.2
= 6.37
The value 6.37 is entered for the non-work cold start HC factor on
Worksheet VI-C, column 8.
-------
VT-A. INPUT TRAVEL DATA SUMMARY FOR EMISSIONS
KSTTMATION
Population
Subgroup
1
2
3
4
V =
Work VMT
(I or IV)1
308,979
17,283
135,005
7,308
468,575
TOTAL WOR.X VMT
jTJBase Alternative
I [Revised Alternative
Forecast Year:
1982
Average
Work Trip
Distance
(miles) (I)
(,1-L or L)
14.5
14.5
9.5
9.5
\"** —
Number of
One-Way
Work Trips
21,309
1,192
14,211
769
37,481
V =
Pol- icy: HOV Lane
Non-Work
VMT
(I or IV)
191,012
0
127,341
0
318,353
Average
Non-Work
Trip
Distance
(miles) (I)
3
-
3
0
\~~* _
Non-Work
Trips
63,671
42,447
106,118
H
1
£
TOTAL WORK TRIPS
I
TOTAL NON-WORK VMT
TOTAL NON-WORK TRIPS
I
Source Worksheets are indicated in parentheses where applicable.
VMT and trips on worksheets I and IV in Appendix A must be multiplied
by the number of households per population subgroup.
I
143,599
TOTAL TRIPS
-------
VI-A. INPUT TRAVEL DATA SUMMARY FOR EMISSIONS
I I Base Alternative
ESTIMATION
ion
P
^ ^_
TOT
Work
tri
r»iimh
Work VMT
(I or IV) 1
262,014
13,861
119,884
6,665
402,424
AL WORK VMT
•
Average
Work Trip
Distance
(miles) (I)
14.5
14.5
9.5
9.5
£
Number of
One-Way
Work Trips
18,070
956
12,619
702
32,347
i — i
^c-»_
TOTAL WORK TRIPS TOTAL
1
Revised Alternative
Forecast Year:
Policy:
Non-Work
VMT
(I or IV)
192,349
0
127,978
0
320,327
HOV Lane
NON-WORK VMT
sheets are indicated In parentheses where applicable.
ps on worksheets I and IV in Appendix A must be multiplied
Average
Non-Work
Trip
Distance
(miles) (I)
3
_
3
_
=
E^m ,
TOTAL
1
139,122
Non-Work
Trips
64,116
_
42,659
_
106,775
NON-WORK TRIPS
1
TOTAL TRIPS
-------
VI-B. COLD START FRACTIONS
I x| Base Alternative
I [ Revised Alternative
Policy: HOV Lane
Subgroup:
All
Forecast Year; 1982
1. If Daily Parking Duration Data are Available:
% of Auto Trips by % of Auto Trips with % of Auto Trips by % of Auto Trips with
Catalyst-Equipped Parking Duration Non-Catalyst Equip- Parking Duration % of Auto Trips
Vehicles (Table D.I) 1 hour ped Vehicles 4 hours with Cold Starts
i
-t>
oo
2. If Daily Parking Duration Data are not Available
% of Work Trip Cold Starts (Table D.2)
90
% Non-Work Trip Cold Starts (Table D.2)
57
-------
VI-C. AUTO START-UP AND EVAPORATIVE EMISSIONS
I x I Base Alternative
II Revised Alternative
Policy: HOV Lane
Forecast Year:_
1982
Temperature:
75
Work Trips
Non-Work Trips
'
(1)
Population
Subgroup
All
TOTALS
(2)
X Cold
Starts.
(VI-B)1
90
(3)
Trips
(IV-A)
37,480
Polluc-
ant2 1
HC(c)
C0(c)
N0x(c
HC(c)
C0(c)
N0x(c
HC(h)
*C(c)
C0(c)
<0x(c
fl
D.2
D.3
D.4
D.fi
D.2
D.3
D.4
n.fi
D.2
D.3
D.4
TOTALS
(4)
Start-Up
Factors
8.3
-
3.5
6.0
HC
/. CO
Subgroups NOx
(5)
Emissions •
Col. 3 X Col. 4
(grams)
311,092
-
131,184
224,886
535,978
-
131,184
/
(6)
% Cold
Starts
(VI-B)
57
(7)
Trips
(VI-A)
106,118
1
(8)
Start-Up
Factors
6.37
-
3.0
6.0
HC
X co
Subgroups NOx
\
(9)
Emissions -
Col. 3 X Col. <
(grams)
675.972
-
318,354
636,708
Ir312.680
318.354
Source Worksheets are indicated In parentheses
where applicable
(c) indicates cold start factor
(h) indicates hot soak factor
both work and non-work start-up factors
obtained from the indicated tables
Work Trip Start-
up Emissions
(grams)
I
.(>
VD
Non-Work Trip
Start-Up
Emissions
(grams)
Total Start-Up
Emissions
(grams)
-------
VI-C. AUTO START-UP AND EVAPORATIVE EMISSIONS
Base Alternative
Revised Alternative
Policy:
HOV Lane
Forecast Year:_
1982
Temperature:
75
Work Trips
I
Non-Work Trips
(1)
Population
Subgroup
All
/
(2)
Z Cold
Starts.
(VI-B)1
90
(3)
Trips
(IV-A)
32,347
u
•HP*
"o a
a. S
HC(c)
C0(c)
!JOJc(c
HC(h)
«C(c)
C0(c)
N0x(c
»C(h)
HC(c)
C0(c)
Y0x(c
IC(h)
<*>
o ja
& H
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
TOTALS TOTALS
(4)
Start-Up
Factors
8.3
—
3.5
6.0
HC
X^1
Subgroups NOx
\
(5)
Emissions •
Col. 3 X Col. U
(grams)
268,480
-
113,215
194,082
462,652
113,215
/ N
(6)
Z Cold
Starts
(Vl-B)
57
1
(7)
Trips
(VI-A)
106,775
(8)
Start-Up
Factors
6.37
-
3.0
6.0
HC
"S ^ CO
f •>
Subgroups NOx
(9)
Emissions •
Col. 3 X Col. 4
(grama)
680,156
-
320,325
640.650
1,320,806
320,326
Source Workaheeta are Indicated In parentheses
where applicable
(c) indicates cold start factor
(h) indicates hot soak factor
both work and non-work start-up factors
obtained from the Indicated tables
Work Trip Start-
Up Emissions
(grams)
Non-Work Trip
Start-Up
Emissions
(grams)
I
Ui
O
Total Start-Up
Emissions
(grams)
-------
1-51
The NOx factors required by Worksheet VI-C are obtained in a similar
manner from Emissions Table D.4. The work factor is obtained directly and the
non-work factor requires simple interpolation. The final input to Worksheet
VI-C, Hot Soak HC factor, is read directly from Emissions Table D.5 for 1982.
Worksheet VI-D is used to determine travel, or VMT-related, emissions
for the base and revised alternatives. For simplicity, average speeds of 20
mph for work trips and 30 mph for non-work trips are assumed. Emissions fac-
tors for 1982 and average speeds of 25 and 30 mph may be read directly from
Emissions Table D.6 and entered on Worksheet VI-D. Worksheet VI-A supplies
the necessary VMT total work and non-work figures for the corridor.
Worksheet VI-E may now be used to summarize the net HC and NOx emis-
sion reduction associated with the HOV lane. The data sources for this
worksheet are noted for each item. All of the emissions changes are given
in grams per day.
I. Interpretation of Results
Large reductions in ozone forming emissions and total VMT are predic-
ted for the proposed combination of carpool and bus enhancements in the
corridor. The 67,500 mile reduction in daily work VMT (Worksheet III-l)
represents a decrease of more than 14 percent from the current level of
468,375 miles per day. The predicted HC and NOx emissions decreases
shown on Worksheet VI-E are each on the order of one-fifth of a ton per
day. Although some of the input data items for the analysis procedure
represented somewhat rough estimates of existing conditions in the cor-
ridor, the magnitude of the predicted impacts on VMT and vehicle emissions
-------
VI-D. AUTO TRAVEL EMISSIONS
Base Alternative
I Revised Alternative
Policy:
HOV Lane
Forecast Year:
1982
Work Trips
1
Non-Work Trips
1
\ /
(1)
Population
Subgroup
All
(2)
Average
Speed
20
(3)
VMT .
(VI-A)1
468,575
TOTALS
(«)
Auto Travel
Factors
(Table D.7)
HC 1.9
CO
NOx 1-9
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
v» HC
V CO
NOx
(5)
Emissions •
Col. 3 X Col. 4
(grams)
890,290
890,290
890,290
890,290
(6)
Average
Speed
30
J
1
(7)
VMT
(VI-A)
118 351
-
\
Sube
(8)
Auto Travel
Factors
(Table D.7)
"C 1.3
CO
NOX 2 .3
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
2^ co
NOx
roups
(9)
Emissions -
Col. 7 X Col. 8
(grams)
413,859
732,212
413,860
732,210
^Source Worksheets are indicated in
parentheses where applicable
Work Trip Travel
Emissions
(grams)
I
Ul
ho
1.304,150
1,622,500
Non-Work Trip
Travel Emissions
(grams)
Total VMT Travel
Emissions
(grams)
-------
VI-D. AUTO TRAVEL EMISSIONS
Base Alternative
Revised Alternative
Policy:
HOV Lane
Forecast Year:
1982
Work Trips
I
Non-Work Trips
(1)
Population
Subgroup
All
/
(2)
Average
Speed
20
\ / \
(3)
VMT
(VI-A)1
402,424
TOTALS
(A)
Auto Travel
Factors
(Table D.7)
HC 1.9
CO
NOx 1-9
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
-------
VI-E. SUMMARY OF CHANGES IN EMISSIONS
Revised Alternative
Policy:
HOV Lane
Forecast Year: 1982
(1)
Population
Subgroup
All
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOX
Base Emissions
(2)
Trip-Related
(VI-C)1
1,848,658
—
449,538
TOTALS
Source Worksheets are Indicated in
parentheses where applicable
(3)
Travel
(VI-D)
1,304,150
-
1,622,500
HC
ECO
NOx
Sub-
groups
(4)
Total
(Col . 2 + Col . 3)
3,152,808
-
2,072,038
3,152,808
2,072,038
Total Base
Emissions
(grams)
Revised Emission
(5)
Trip-Related
(VI-C)
1,783,368
-
433,540
(6)
Travel
(vi-n)
1,181,031
-
1,501,358
z
Sub-
giro u
a
(7)
Total
[Col-5+Col. 6)
2.964.39S
-
1,934.89?
HC
CO
NOx
(8)
Change in
Total
Emissions
[Col. 4~CoL 7)
-188.409
-
-137,140
-188,409
-137,140
Total Change
ps in Emissions
(grans)
(9)
Percent
Change in
Emissions
(Col. 8/CoLA)
x 100
-6.0
-
-6.6
-6.0
-6.6
Percent
Change,
Total
Emissions
I
Ln
-------
1-55
indicate that significant beneficial impacts can be predicted for the
proposal with a high degree of confidence.
In general, the types of mode shifts which were anticipated prior
to the analysis are predicted to occur. The corridor-wide average mode
share estimates from Worksheet V-A indicate that the major impact of the
proposed measures is an increase in bus patronage—on the order of 25 per-
cent. Both drive-alone commuters and carpoolers are predicted to switch to
buses, with the overall share for each auto mode decreasing. However,
the number of four-person-or-larger carpools is predicted to increase by
about 11 percent in response to the greatly improved travel speeds these
carpools will enjoy . The combination of lower overall carpool share and
higher four carpool share implies an increase in the average size
of carpools in the corridor. From the impacts predicted using the manual
analysis techniques, it can be seen that the proposal will achieve its
major goals of decreasing fuel consumption and vehicle emissions through
increases in the use by commuters of large carpools and buses.
The air quality and fuel conservation potential of the planned
measures as generally proposed is quite certain; further, more detailed
analysis is not required to establish the desireability of pursuing imple-
mentation. However, in the project's later planning stages, it
may be desirable to do a more detailed analysis to select the most attrac-
tive detailed implementation plan. This analysis would necessarily be at
a much more specific and complex level, requiring the use of computer-based
techniques such as SRGP.
Note that all percentage figures apply only to trips between the study
corridor and the CBD and fringe areas. On a regionwide basis, percentage
changes will be much smaller.
-------
CASE STUDY II
DOWNTOWN AUTO RESTRICTED ZONE
-------
CASE STUDY II: DOWNTOWN AUTO RESTRICTED ZONE
A. Problem Presentation
The major retail and office employment center of a large eastern
city experiences recurring heavy traffic congestion throughout much of
the day. During the peak travel period, traffic tie-ups in the area lead
to congestion problems throughout the central business district of the
city. This congestion contributes to the poor air quality of the metro-
politan area as a whole (which is in violation of the ozone health
standard), and is the major cause of high levels of carbon monoxide in the
CBD. There is also a general consensus within the downtown business
community that the current traffic situation within the area is contribu-
ting to declining retail sales and a general deterioration of the area's
shopping environment.
B. Proposed Transportation Measures
In response to the congestion problems in the area and the desires
of retailers to improve shopping conditions, a plan incorporating the
removal of traffic from sections of two major streets in the shopping
district has been developed. The plan also provides for the rerouting
of auto traffic around the auto restricted zone (ARZ) and new routings
for buses on the auto restricted streets and reserved lanes on streets
adjacent to the ARZ. The plan has reached a high level of maturity
including a detailed assessment of the likely changes in the traffic flows
on streets in the CBD and the specification of bus routings and stops in
-------
II-2
the area. However, no analysis of how the proposed measures will affect
the mode choices of travellers to the ARZ or vehicle emissions in the
metropolitan area has been conducted. Additionally, there is some con-
troversy over whether a free shuttle bus service proposed by some of the
retailers should be provided between the ARZ and the parking lots and
major transit terminals which surround it. A quick analysis is necessary
to determine the relative air quality impacts of the planned ARZ with and
without the proposed shuttle bus service.
Work and non-work trips to the ARZ must be considered in the analysis
since the proposed changes will have a direct impact on the relative
levels of service offered by the different modes for all trips. The auto
restrictions proposed under the ARZ plan will tend to increase in-vehicle
travel time for all auto users destined to the immediate area proposed for
auto restriction. Increased circuity due to the necessity for autos to
travel around, rather than through, the auto restricted area will be pri-
marily responsible for the increase in travel time for auto trips. Out of
vehicle travel time for both auto and transit will be reduced because of
the ARZ. The removal of traffic from the ARZ area will lead to easier and
faster walking for all pedestrians, while the routing of bus lines direct-
ly into the area will shorten the walk from the bus stop to the final
destination for transit riders. Since on-street parking is currently not
available on the proposed auto restricted streets, and access will continue
to be provided to all off-street parking facilities, walking distances
for auto travellers will not change. If shuttle bus service is provided
to the parking garages and transit terminals, further out-of-vehicle
-------
II-3
travel time reductions will be experienced by both auto and transit users.
In summary, the following level-of-service changes are associated
with the ARZ proposal:
Transportation Measure
1. Removal of Autos from
Shopping District Streets
2. Free Shuttle Service
Level-of-Service Impact
- Increased in vehicle travel
time (IVTT) for auto users
to the ARZ
- Reduced out-of-vehicle
travel time (OVTT) for auto
users
- Reduced OVTT for transit
users
Reduced OVTT for auto
users
Reduced OVTT for transit
users
All of the proposed actions will be taken by 1982. Because the short-
term effects of the auto restrictions are of primary interest, 1982
is selected as the analysis year.
The proposed measures are expected to affect both work and non-work
travel to the ARZ. The extra access time for auto trips is expected to
lead to a shift to transit, which already carries the majority of the
area's workers. Improved (reduced) walk times from transit stops to em-
ployers in the area will further improve the service offered by transit
relative to auto.
The overall effect of the proposed measures on the number of people
travelling to the ARZ for non-work purposes is uncertain. Some argue that
-------
II-4
the improved walking conditions associated with the removal of traffic
from the area's streets will attract more people. Others contend that
making the area more difficult to reach by auto will make it less
attractive than other shopping areas in the city, causing an overall
reduction in the numbers of non-work trips to the area.
C. Selection of Analysis Technique
Because the removal of autos from the streets in the core of the
city's retail center has been accepted as a desireable action and since
the ARZ plan has reached a fairly high level of detail with respect to
auto traffic and bus routing, a detailed demand and air quality analysis
is not required. A simple comparison of the impacts of the ARZ with and
without the proposed shuttle service, yielding order of magnitude esti-
mates of the emissions changes likely to occur under each scenario, will
be sufficient to make a decision on the provision of the shuttle service.
Because the amount of time allocated to the analysis is limited, and a
programmable calculator is readily available, the program 3MODE(VAN)-AGG
(2.2.3) is selected for computing the mode share and VMT impacts of the
two ARZ scenarios. The manual emissions worksheets (4.1) and auto fuel
consumption and operating costs method (4.2) are used to develop air quality
impact estimates based on the predicted VMT changes. In addition, system-
atic data analysis methods (2.1.3) are used to specify base case conditions.
A non-standard application of the calculator program is necessary
for the analysis of non-work trips, A model of joint destination and
mode choice for non-work trips is used in place of the standard work trip
model incorporated into the calculator program. This is accomplished by
The section numbers following specified analysis methods refer to the
location of their description in Volume I.
-------
II-5
replacing the default model coefficients in the program with user-
supplied coefficients based on the non-work model. The worksheets for
preparing input data for the program and for summarizing the calculator
outputs are also used in a non-standard way, which will be described in
subsequent sections.
D. Overview of the Analysis
The analysis of the effects of the ARZ on downtown and region-wide
air quality requires five basic steps:
Step 1 — Identify components of each policy scenario
(shuttle and non-shuttle)
Step 2 - Segment the affected population into homogeneous
sub-groups or market segments and develop base
data for each.
Step 3 - Estimate and quantify the transportation level-
of-service impacts of each scenario for each
trip type.
Step A ~ Estimate new mode shares and VMT using the
3MODE(VAN)-AGG program.
5 - Estmiate changes in vehicle emissions using the
manual auto emissions worksheets.
The relationship between these steps is shown in the flow chart of
Figure II-l. The first three steps involve the preparation of data for
input to 3MODE(VAN)-AGG in step 4. Changes in the level-of-service
experienced by each of the market segments identified in step 2 must be
determined separately for each of the two ARZ scenarios, while the rest
of the input data is identical for the shuttle and non-shuttle alternatives.
-------
STEP 1
Enumerate
Components of
Policy Scenario
STEP 2
Segment
Affected
Population
Select Market
Segment for
Analysis
STEP 3
Estimate Changes
in Level of
Service for
Market Segment
STEP 5
r
Estimate ARZ
Area Vehicle
Emissions
Changes
Estimate Area-
wide Changes in
Vehicle
Emissions
|
MANUAL EMISSIONS WORKSHEET ANALYSIS
More
Market
Segments?
Integrate Results
with those for
Previous Market
Segments
Calculate
Revised Mode
Shares and VMT
for Market
Segments
3MODE(VAN)-AGG RUNS
FIGURE H-1
Steps in Analysis of ARZ Policy Scenarios
(All steps repeated for each scenario)
-------
II-7
The calculator program provides estimates of the changes in VMT by
travel mode for work and non-work trips under each scenario. These
estimates are then used in the step 5 analysis of vehicle emission
impacts.
E. Defining the Scope of the Analysis
The relevant components of each policy scenario were identified in
section B above; therefore, step 1 of the analysis has been completed.
Step 2 involves determining how detailed the analysis should be (the
more market segments identified, the more detailed and complex the
analysis becomes). Budget limitations for the analysis, the availability
of data pertaining to different population groups and the desired
accuracy of the predictions developed during the analysis are the major
considerations in choosing the scope and level of detail of the analysis.
For the analysis of the ARZ proposal, two levels of distinction
between the affected travelers were used. Work and non-work trips to the
ARZ are analyzed separately, and for each trip type, households with and
without autos available for their trip are considered independently. This
categorization of travellers leads to the following four market segments:
• Workers with the drive alone mode available
• Workers without the drive alone mode available
• Non-work travellers with an auto available
• Non-work travellers with no auto available
The non-work model to be used in the analysis has only two modes: auto
and transit. Non-work travellers without autos must use transit. However,
workers in the ARZ can use the shared-ride mode regardless of the avail-
ability of the auto to them.
-------
II-8
The changes in level-of-service associated with the three ARZ scenar-
ios will potentially affect the mode choices of both work and non-work
travellers. A distinction between work and non-work trips is crucial to
the development of accurate impact estimates because of the different
choices available to each. The attractiveness of the ARZ is reflected
in part by the level of transportation service available, and will determine
the total number of non-work trips destined to the area. On the other
hand, the volume of work trips is fixed in the short-run, being determined
by the total employment in the area. Thus, in analyzing non-work trips
both destination and mode choice must be determined, while only mode choice
is of interest for work trips.
Separating the population into groups with and without the drive
alone or auto mode available leads to greater accuracy in the prediction
of revised mode shares. This distinction is particularly important for
the analysis of this ARZ proposal because of the high proportion of those
living in the city who do not own autos (18 percent).
F. Input Data Development
Worksheets C-l and C-2 fpr the 3MQDE(VAN)-AGG procedure define
the data required for input to the program for each market segment. Much
of the information required for this analysis is available from detailed
planning studies conducted for the ARZ, and transportation system data
bases developed as part of the metropolitan area's on-going transportation
planning process. Such data is typically available for a large city with
a well developed metropolitan transit system.
-------
II-9
Base Data
Hourly total person trip tables for the 1982 analysis year provide
most of the base data required for worksheet C-l . Average trip lengths
were calculated using the trip tables at the district level and a map of
the urban area (to determine district-to-district distances). Annual
household income is not required in the analysis since none of the
measures involve changing the cost of any of the modes. However, the
calculator program requires that the default value '"1" be entered for
income. No information was available on average carpool size for trips
to the ARZ area, so the default value of 2.5 (the national average for
carpool size) is assumed. The total population of the two work trip
market segments was determined by summing all of the AM peak period
and 25 percent of the PM peak period ARZ-bound trips, and then appor-
tioning these trips among auto owning and non-auto owning households.
Figure II-2 shows a typical distribution of auto ownership by 1970
income level for a large city with a well developed transportation system.
For the metropolitan area under study, the 1972 U.S. Census County and
City Data Book reported the following distribution of households by
income level:
If a trip table for the region was not readily available, the quick re-
sponse urban travel estimation techniques (2.1.2) could be used to deter-
mine the total number of work and non-work trips to the area.
2
All of the AM peak period trips and a smaller proportion of the PM peak
period trips to the ARZ are assumed to be work trips.
-------
11-10
PHILADELPHIA. PENNSYLVANIA
4,020.420 5.346
IN
10
SO
8 »
o
.0
«
30
20
10
0
1500 3000 $000 7000 9000 11,000 13,500 17,500 22,600
INCOME
37,000
FIGURE II-2
Auto Ownership Distribution
-------
11-11
Household Annual
Income Range
Under $3,000
$3,000 - 4,900
$5,000 - 9,999
$10,000 - 14,999
$15,000+
Percentage of
Total Households
6.1
6.8
18.7
29.8
30.1
The overall proportion of households owning no autos may be calculated
by reading the average percentage of zero auto owning households for each
income level from Figure II-2, and then multiplying the percentage by
the proportion of all households in the income bracket:
$3-5.000 $5-7,000
(.535) (.068) + (.370) (.087)
Income Under .$3 ,.000
.65 (0 Auto ffiA .061 HH<3K
VHH<3K ; x Total HH
$7-10.000 $10-15.000 $15.000+
(.210) (.185) + (.085) (.298) + (.030) (.301) = .181
The 18 percent zero auto owning proportion was applied to the total work
trip volume of 128,748 round trips per day to obtain the market segment
populations for sub-groups 1 and 2 on the Worksheet C-l.
Potentially, any of the urban area's 3,654,200 residents could travel
to the ARZ for a non-work purpose. Therefore, the entire population of
the region classified by auto ownership, forms market segments 3 and 4.
As with workers, 18% of the non-work population is assumed to have no auto
available.
Vanpools do not operate to the ARZ area now, and none are expected
to start up as a result of the auto restrictions. However, the default
values phown on worksheet C-1 for average vanpool size and vanpool
-------
11-12
circuity factor must be entered into the calculator program. They will
not be used in any of the internal calculations or subsequent analysis
steps, however.
Base Mode Shares
Work Trips
For market segments 1 and 2 combined (all workers in the ARZ area),
the total number of trips made to the downtown area by transit may be
determined from alighting counts taken on each of the bus routes terminat-
ing in the ARZ and each of the six subway stops serving the area, during
the a.m. and p.m. peak periods. Once again all of the a.m. peak period
and one-quarter of the p.m. peak period trips downtown are assumed to
be work trips. Counts of transit passengers to downtown by hour were
conducted recently as part of the planning process for the ARZ proposal.
According to the counts, the total number of daily transit trips classi-
fied as work trips is 94,872—out of 128,748 total work trips. The
aggregate transit share for all workers is then 73.7 percent, leaving
an auto share of 26.3 percent. Data from annual screenline traffic
counts indicate that 66 percent of the persons entering the CBD by auto
are auto drivers and 34 percent are auto passengers. The aggregate auto
driver and auto passenger mode shares are then:
If no auto occupancy data were available for the study area, the quick
response urban travel estimation techniques (2.1.2) could have been used
to determine carpooling activity.
-------
11-13
Auto Driver = .263 (auto share) x .660 fAuto drivers^- .174
share \ Auto users 1
Auto Passenger = .263 x .340 = .089
Share
For use in the calculator program analysis, these mode shares must be
converted to drive alone and shared ride shares. The following formulas
may be used for this conversion:
^ . . i „,_ 4 ^ ^ • ot_ Auto Passenger Share
Drive Alone Share = Auto Driver Share ° —•
(Average Carpool Size -1)
_,,,„.,„, . _. _. . Auto Passenger Share
Shared Ride Share = Auto Passenger Share + ,. •—•= °. „. rr-
& (Auto Carpool Size -1)
Assuming the national average carpool size of 2.5, and using the above
auto driver and passenger shares yields:
Drive Alone Share = .174 - (25-!) " -114
OflQ
Shared Ride Share = .089 + 05-1) = -149
Because those not owning autos must have a zero share for the drive
alone mode, the mode shares of the two market segments making up the
work trip population must be different. To convert from total population
mode shares to market segment mode shares, the following formulas (based
on the assumption that the ratio of transit riders to carpoolers is the
same for the non-auto owning sub-group as the entire population) are used:
-------
11-14
SR T
DA = 0 SR = — r^— T -
o o SR + T o SR + T
DA,. SR,. - SR F T - T F
where DA = Drive alone share for households without autos
SRQ = Shared ride share for households without autos
To = Transit share for households without autos
DAj^ = Drive alone for households with autos
SRi = Shared ride share for households with autos
Ti = Transit share for household with autos
DAt = Drive alone share for all households
SRt = Shared ride share for all households
Tt = Transit shares for all households
FO, FI = Fraction of household without and with autos
The calculations are crried out sequentially, with the mode shares for
households without autos being determined first. For the downtown
workers the resulting mode shares are:
DAo - ° SRo = .149^.737 - -168
T = '737 .114 __0
o .149 + .737 = '832 DA1 ~ ~^2 = -139
SR = -149 ~ -165(.18) ..145 _ .737 - .832(.18) = .716
1 .82 i i82
These shares are entered on Worksheet C-l for subgroups 1 and 2.
-------
11-15
Non-Work Trips
Although worksheet C-l calls for base work trip modal shares, it
may be used to record non-work trip making data for input to the non-work
model to be used in this analysis. Because the model predicts changes in
mode share for auto and transit only, only these modes need to be considered.
The model predicts changes both in the mode choice of travellers to the CBD,
and the absolute volume of non-work trips to the area as a result of the
auto restrictions. The base mode shares reported on Worksheet C-l for non-
work trips must, therefore, reflect the existing mode shares and the pro-
portions of the region-wide non-work trips which are bound for the study
area. This is accomplished by using the "other" mode share column to
record the proportion of all of the urban area non-work trips which are not
bound for the study area (regardless of mode). The drive alone and transit
columns of Worksheet C-l then reflect both the total volume and mode share
of non-work trips to the area.
The 1980 trip tables by hour (alternatively the quick response urban
travel techniques (2.1.2) could be used) indicate that 5 percent of the non-
work trips in the region were destined to the downtown area. Therefore, the
"other" mode share for subgroups 3 and 4 is .95. Transit alighting counts
in the area and the person trip tables ahow that the overall non-work
transit share is 59 percent. Employing an analysis similar to that used
to determine the market segment mode shares for work trips, the following
shares were calculated for non-work trips!
For simplicity, the non-auto owning households were assumed to have a
zero non-work auto mode share. However, a non-zero share for these
households could be assigned to reflect their use of carpools for non-
work trips. This share could be determined from census data, or could
be calculated from a known shared ride and drive alone mode split for non-
work trips using adjustments similar to those used above to determine the
work trip mode splits for households without autos.
-------
11-16
Households without Autos:
Auto Share - 0%
Transit Share - 100%
Households with Autos:
Auto Share 50%
Transit Share - 50%
Multiplying these shares by the ARZ area's 5% share of total non-work trips
yields the base mode shares recorded on Worksheet C-l. (The auto share is
recorded in the Drive Alone column.)
Changes in Level-of-Service
Step 3 of the analysis procedure involves quantifying the level-
of-service impacts of the proposed auto restrictions in terms of changes
in in-vehicle and out-of-vehicle travel time by mode for the shuttle and
non-shuttle scenarios. Worksheet C-2 is used to record the level-of-
service changes in a format for input to 3MODE(VAN)-AGG.
Auto Trips - In order to determine the average impact of the auto
restrictions on in-vehicle travel time for autos, peak and off peak trips
made to the area by auto were classified (using the 1980 trip table,
and counts at entry points to the CBD) according to entry location
and destination point (parking location) within the CBD. The analysis
zones used to define the entry and destination points of CBD trips are
shown in Figure II-3. Any of the numbered zones were considered
potential entry locations and all zones were potential destinations for
The drive alone and carpool modes will experience identical level-
of-service changes. Therefore, in making estimates of changes .in LOS,
they are treated together. In recording the estimated LOS changes
on Worksheet C-2, identical entries will be made for the drive alone
and carpool modes.
-------
11-17
ARZ Base Data
WORKSHEET C-1
BASE DATA
POLICY: Downtown ARZ
Population
Subgroup
l.Work Trips
All Modes
2. Work. Trips
No Drive Alone
3.Non-Work
All Modes
4 . Non-Work
No Auto
Average round
Trip Length (mi)
16
16
16
16
Annual
Household
Income ($)
*
[1 ]
[1]
[1]
[1]
Average
Carpool Size
[2.5]
[2.5]
[2.5]
[2.5]
Population
105,573
23,175
3,096,
780
S57,420
Average
Vanpool Size
[10]
[10]
[10]
[10]
if
•< o
[1.5]
[1.5]
[1.5]
[1.5]
Base Work Trip Modal Shares
sr
.139
0
.025
0
O
o>
I
.145
.168
0
0
1
.716
.832
.025
.050
1
o_
0
0
0
0
9
1
0
0
.95
.95
'[ ] = Default Value
-------
11-18
FIGURE II-3
CBD Entry and Destination Zones
-------
11-19
auto trips. Changes in the routing of auto trips for each zone pair were
determined by examining the new traffic flow patterns (one-way streets,
streets closed to autos, etc.) included in the ARZ plan. Entry and
destination point pairs with significant auto delays due to more cir-
cuitous routing are identified below, along with the estimated dela> to
autos and the proportion of the total peak period downtown-bound auto
person-trips represented by trips between those zones.
Zone Pair
I -ARZ
II-IV
III-I
III-V
I I I- ARZ
IV-II
IV-ARZ
v-in
I-III
Delay in Minutes
(round trip)
2.0
0.5
3.0
3.0
2.0
0.5
2.0
3.0
3.0
Proportion of
Hour Downtown
.043
.069
.006
.007
.020
.069
.021
.006
.006
Total Peak
Auto Trips
A similar table is required for non-work auto trips. Multiplying the
round trip delay for each zone pair by the percentage of total auto
trips made between the zones and summing over all zone pairs yields the
average in-vehicle travel time increase for CBD-bound auto trips (AlVTT)
«s recorded on Worksheet C-2 for the shuttle and non-shuttle alternatives:
These IVTT increases for auto trips are not affected by the provision of
shuttle bus service since the time on the shuttle is considered to be
out-of-vehicle travel time.
-------
11-20
Work = + .314 min
Non-work = +1.258 min
Improved walking conditions will reduce out-of-vehicle travel time
(OVTT) for all auto trips bound for the ARZ. The CBD entry point and
destination trip table shows that 8.4 percent of all CBD auto work trips
and 33.7 percent of non-work auto trips are bound for the immediate ARZ
area. Average round-trip walk time savings are estimated at 1.0 minute
for work trips, and 1.5 minutes for non-work trips, leading to net OVTT
savings of .084 minutes for work trips and .506 minutes for non-work
trips with no shuttle service.
The shuttle service is estimated to lead to 3.5 minutes in out-of-
vehicle travel time savings for those who use it. In previous analyses,
it was estimated that 8.5 percent of the downtown-bound auto work trips
would use the shuttle service. However, only 40 percent of the ARZ-bound
auto travellers will be served by the shuttle (the remaining 50 percent
will still save 1.0 minutes). The overall average OVTT savings for the
shuttle alternative may be calculated:
085 Buttle trips t „ ,. shuttle , ARZ non-shuttle
total trips ' savings ' All ARZ
n«R All ARZ . ARZ walk .
.085 ,———— 1.0 . = .348 minutes
total trips savings
The corresponding calculation for non-work auto trips is:
-------
11-21
shuttle trips 3.5 minutes + .3 ff? °g-^uttle
total trips All ARZ
337 ARZ triPS — x 1.5 minutes = .750 minutes
total trips
(Due to parking location differences for work and non-work trips, a
higher proportion of non-work auto trips will be served by the shuttle.)
The above IVTT and OVTT changes are the only auto level-of-service
impacts anticipated for the ARZ proposal.
Transit Trips
In-vehicle travel time for transit is not anticipated to change
appreciably due to the ARZ measures. However, significant OVTT savings
are anticipated for both work and non-work trips bound to the ARZ areas for
the non-shuttle scenario. Hourly alighting counts at the transit stops
within the CBD indicate that for both work and non-work trips, 26.8 percent
of the transit passengers bound for downtown locations are travelling
to the immediate ARZ area. The combined round trip OVTT savings
associated with easier walking in the ARZ and shorter walks due to new
bus stop locations within the ARZ are estimated to be approximately two
minutes, leading to an average CBD transit OVTT reduction of; .268(2)=.532
minutes. This figure is entered on worksheet C-2 for all market segments
for the non-shuttle alternative.
It is expected that approximately 6.5 percent of the CBD bound
transit passengers will use the proposed shuttle, enjoying a 3.5 minute
OVTT savings. None of these shuttle users will be destined to the ARZ,
Unlike auto trips to the area, transit trip-making to the downtown
area is similar for work and non-work trips.
-------
11-22
however, because transit stops will be located directly on the auto
restricted streets. The total transit OVTT reduction for the shuttle
alternative is therefore estimated to be; .532 + .065(3.5) = .762 minutes.
All market segments on Worksheet C-2 will experience this average
transit OVTT savings under the shuttle scenario.
Only travel time changes are expected to occur as a result of the
proposed auto restrictions. No travel cost impacts are associated with
either the shuttle or non-shuttle alternative. Because the drive alone
and carpool impacts are identical, and vanpooling is not expected to occur
as a result of the ARZ, no further level of service change estimates or
entries on Worksheet C-2 are required.
G. Description of Model Application
Four runs of the calculator program 3MODE(VAN)-AGG are required to
estimate new mode splits and VMT for the two ARZ scenarios, one each
for work and non-work trips for each scenario. The program runs are
straightforward, using none of the optional subroutines. The non-work runs
involve the use of user-supplied mode choice coefficients, which were
derived from a joint choice model of shopping trip frequency, destination,
and mode choice. The work and non-work runs for the non-shuttle alternative
are described in detail below. The runs for the shuttle scenario analysis
are identical, except for differences in the level of service changes as
noted on Worksheet C-2 for the shuttle option.
-------
WORKSHEET C-2
Level-of-Servlce Changes: Non—Shuttle Alternative
POLICY'-Downtown ARZ Non-Shuttle Alteri
CHANGES IN TRANSPORTATION LEVEL OF SERVICE
(all data represent round trips)
atlve
Population
Subgroup ¥
1
2
3
4
Drive Alone
OS?
2F
3if
^H|
*? ^
5
0
0
-
0
t> * p
o§ *
5* a
"if
T1!
-.536
-.536
-.536
-.536
O rl O
I1*
'Si?
«•"•• if
si *
0
0
0
0
Carpool
I> ^ 5"
m
£|j
.314
.314
0
0
t> ^ 0
m
" i *
I1!
-.084
-.084
0
0
> ^ o
o$ £
5^a
o o •
1& j
2 s
0
0
0
0
y o o
!fl
a «• a
M?
a1 If
0
0
0
0
Vanpool
t> =" ?
?ii
£f*
l
.314
.314
0
0
t> ^ 0
0 | &
3 ao
H°g^
1§I
-.084
-.084
0
0
> =f o
05 5.
3*°
00'
^8 |
3"s
0
0
0
0
3' S $
* 5'l
fit
J^3
•o» &3
Ci ' '
0
0
0
0
I
NJ
U>
1 - Work Trips, All Modes
2 - Work Trips, No Drive Alone
3 - Non-Work, All Modes
4 - Non-Work, No Auto
-------
11-24
Work Trips
Two passes through the calculator program are used to estimate the
combined impact of the ARZ on downtown-bound workers with and without autos
available for their trip. Worksheet C-4 is used to prepare the input
data from Worksheets C-l and C-2 for direct input to the Calculator.
One pass through the program is described in detail as an example of the
required order for calculation steps.
Because user supplied coefficients are not used for the work trip
analysis, the first input value required is defined by Step 4 on Worksheet
C-4. Market segment data from Worksheet C-l is entered for market
segment 1. The optional carpool subroutine is not used in this run, so
Step 6 defines the next set of required input data—base mode shares for
market segment 1. The level-of-service changes estimated for the non-
shuttle alternative and market segment 1 (Worksheet C -2) are then entered
under Step 8 on Worksheet C-4. Note that the changes for drive alone
and carpool are identical, and that "0" is recorded for all LOS variables
not affected by the ARZ proposal. These are the only program steps
requiring input data. Data may now be entered into the programmable
calculator, and market segment 1 analyzed.
After the program is read into the calculator, key "A" is pressed
to initialize the program. Beginning at Step 4, data items are entered
sequentially in the exact order they appear on Worksheet C-4. Key "R/S"
is pressed between each item to enter it into the program. Data entry
continues through step 6e after which "R/S" is pushed. The calculator
will display and print the average number of autos used per worker in
the market segment at this point. "R/S" is pressed a second time, and
the base total VMT is displayed and printed. Step 8 is begun by pressing
-------
11-25
WORKSHEET C -4 PROGRAM STEPS
639.39
(3MODE(VAN)-AGG-2(B)/7900110/ESE)
ARZ, Non-Shuttle:
Work Trips (DA AVAIL)
USER INSTRUCTIONS
STEP
JL_
2
3
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
31
3m
3n
3p
4
4a
4b
Ac
4d
PROCEDURE
Read-jiard(s) -partitioning is 639.39
.Initialize Storage Registers and store —
de.fa.uLt .Coefficients
If-.de fault co^f fi.cienjts_are_usej3 v .skip
to STEP 4
OPTIONAL - Enter user-supplied coefficients
coefficient of IVTT-DA mode
coefficient of OVTT/DIST-DA mode
coefficient of OPTC/Y-DA mode
coefficient 4-DA mode
coefficient 5-DA mode
coefficient of IVTT-SR mode
coefficient of jDVTT/pISTj-SR mode
coefficient of OPTC/Y-SR mode
coefficient of incentives-SR mode
coefficient 5-SR mode ......
coefficient of IVTT-T mode
coefficient of OVTT/DIST-T mode
coefficient of OPTC/Y-T mode
coefficient 4-T mode
coefficient 5-T mode
IF A MISTAKE IS MADE, BEGIN AGAIN AT STEP 3a
rw.«ii»M»»w»»MMMi
Enter market segment data (from worksheet C-.
round trip distance in miles 1 0
household average annual income in $ 4 0
enter "1" if income is not needed for this
segment. _ _ ..
Average carpool occupancy i 0
Default - 2.5
Market segment population . . ...
ENTER
Co Tn CTRP
r"™~
e jDA*
62DA =
6 3°A — --
6^DA =
6 5DA -
e LSR -
02SR"5
9 3SR -
e 4sR -
65SR=
8xT - :
62T -",
e3T -
94T -.
65T ^
LD.
DIST= 16.". '
Y = 1
OCCCp= 2.5
POP .105573
A
.i—
STO
CTO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
**
R/S
R/S
R/S
R/S
PRESS
«»
00
m
UJ.
.02
03
04 '
05
06
07
08
09
10
11
.12
13
14
kau
~
.
LU
— •
DISPLAY
.29
...
- .-_. .....
-
-
— —
..
.. . —
-
-------
11-26
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
STEP
4e
4f
5
5a
5b
USER INSTRUCTIONS
PROCEDURE
Average vanpool occupancy 4 0
default — 10 "" "
Vanpool circuity factor j4 0
default —1.5 -
If a mistake is made, press [Cjand begin "
vtth STEP 4a
If carpool s ub routine" is Trot used,-
go to STEP 6
-.m.m.^^~~~— ~~— -™^~~-^~
OPTIONAL - Carpool Subroutine (data from vox
Enter average occupancies of two classes of
carpools: ••• •
1. Average occupancy carpool class 1
2. Average occupancy carpool class 2 -
Enter level of service and mode share data
1. AIVTT - carpool class 1
2. AOVTT - carpool class 1.
3. AOPTC - carpool class 1
4. AINCENT - carpool class 1 (0/1)
5. A- - carpool class 1**
6. -carpool class 1 "share -as -a "fractiorr
of all carpools *
with STEP 5a
7. AIVTT- carpool class".2
8. AOVTT - carpool class 2
9. AOPTC- - carpool class 2
10. A INCENT -~ carpool class 2 (0/1)
11. Ac - carpool class 2**
12. carpool class 2 share as a fraction""
of all carpools*
If a mistake is made, press C1 and begin
with STEP 5 a
* These two values should sum to 1.0
** If default coefficients are used, these
A's must equal zero.
ENTER
OCCyp - 10
CIRC - 1.5
GO TO STEP
-~~~— ~
ksheet C-3)
OCCCP1" '
occcp2- —
A -
A =
A «
A =
A'= - -
srpi =
A -
A -
A -
A -
A -
SCP2=
PRESS
R/S
R/S
,6^
K.tt.1
2nd
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
C'
2nd
. . _
E'_
DISPLAY
. . . :
38
•- •
-------
11-27
WORKSHEET C.-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
5c
6
6a
6b
6c
6d
6e
fif
fig
7
PROCEDURE
carpool subroutine results :
1. New average occupancy for all carpools
2. New tiarpool change in level of - ._
service (for all carpools)
Enter A LOS for carpool class 1
Enter A LOS for carpool class 2
To repeat for another change in level of
service, press |D I and repeat STEP 5c2-
Step 5 results can also "be recorded on Work
Store modal shares for this market segment
(from Worksheet C-l)
Base drive-alone share .as _a fraction**
(i.e., .645, not 64.5)
Base carpool.share. as a fraction **
Base transit share as a fraction **
Base vanpool share as a fraction**
Base "other" share as a fraction **
If a mistake is made, press GO TO 1/X|
and begin with STEP 6a •
Base VMT
If vanpool sub routine -is not used, go to —
STEP 8
OPTIONAL Vanpool Subroutine -
If VP = 0 in 6d and the revised. vanpool. .
share 4 0, enter a "base" vanpool share on
which to pivot. Defaults —
.14 for firms with <_ 2000 employees
.06 for firms with >^ 2000 employees
r-p
If a mistake is made, press IB I and repeat
STEP 7
* printed but not displayed
** these five fractions must sum to 1
ENTER
ACP1= 2nd TJ*
ACP2"~"
sheet" C -5 .
SDA= _iH?_
Scp= -145
ST - -716
SVP- °
sn ; . ' o
GO TO STEP 8
.
1SVP *
F
R/S
S~
R/S
R/S
^.
R/S
R/S
R/S
R/S
R/S
s/s
R/S
R/S
>RESS
^
-'
\
J
DISPLAY
*OC&-0 =
A/W.-1--197 _
VMTT - 332766
-------
11-28
"D". Level-of-service changes are now entered in the exact order shown on
the worksheet, including zero values, with "R/S" being pushed between
each LOS entry.
Pressing "B" begins step 10 in which the predicted impacts of the ARZ
are calculated and printed by the program. Each time printing ceases, R/S
is pushed, through the printing of base autos per worker (step lOa). Care
must be taken not to press R/S after this final value is printed. At this
point, all calculations for market segment 1 are complete, and market seg-
ment 2 may be analyzed. Key "C" is pressed and the input data for market
segment 2 are entered exactly as for market segment 1, beginning at step 4.
After all steps for market segment 2 are completed, "E" is pressed to obtain
average impacts for market segments 1 and 2 combined. Each of the results
listed under step 11 of Worksheet C-4 will be printed in sequence. The re-
sults for each market segment and for work trips to the CBD as a whole as
recorded on the calculator tape are shown below in the sequence in which
they are printed.
Calculator Output for Non-Shuttle Scenario
Work Trips
DA Available1
0.197
332766.096
.1377108186
14538.54425
232616.708
.1436551705
15166.10731
97063.08681
.7186340109
75868.34843
0.
Q.
0.
0.
329679.7948
.1951728868
DA Not Available1
0.0672
24917.76
0.
0.
0.
.1661921319
3851.502658
24649.61701
.9339078681
19323.49734
0.
0.
0.
0.
24649.61701
.0664758528
Summary
14538.54425
232616.708
19017.60997
121712.7038
95191.84578
0.
0.
357683.856
354329.4119
-3354.444142
-.9378237473
Results2
29
30
31
32
33
34
35
36
37
38
39
^These values correspond to step 10 of Worksheet C-4.
These values correspond to step 11 of Worksheet C-4.
-------
11-29
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
8
8a
8b
8c
8d
8e
8f
8g
8h
8i
81
8k
81
8m
8n
80
9
9a
9b
9c
9d
9e
PROCEDURE
Enter .changes in level of service from Works
If there are NO changes in. level of service,
Press [B]and go to STEP 10; otherwise,
Drive Alone changes:
AIVTTDA
AOVTTM
AOPTCDA ....... .'....
A 4 *^
A-50A **
If there are no additional changes in LOS,
press (B~] and go to STEP 10
Carpool changes from Step 5 or Worksheet C
if -optional STEP 5 was used:
AlVTTrp - - -
AOVTTcp
AOPTCrp
AINCENTcp *
A^CP **
If there are no additional changes in LOS,
press [§] and go to STEP 10
Transit changes:
AIVTTT
A OVTTT - ' - ~
A OPTCT
A4T **
A5T **
If a mistake is made, press [GO TOj IT/xl
and begin at STEP 6a
If the vanpool share in 6d = 0 and STEP -7 — •
was not used, GO TO STEP 10
OPTIONAL: Enter vanpool .change .in LOsTrrpjq
A IVTT^
AOVTTvp
A OPTCyp
A INCENTvp ***
A 5 ** .
* This must equal 0 or 1 unless the carpool
must be between 0 and 1 inclusive.
** If default coefficients are used, these /
*** 9d must equal zero if STEP 7 was used.
ENTER
heet -C-2)
' A = .314
-A = -.084
A
A . 0
A = _D
-5
-A -—• -31*.
A • -.084
A - Q_
A - P_
A - 0
A = 0
"A _ ' -.536
A - °
A = 0
A = 0
— - -
. _.
GO TO STEP
Worksoeet C-
A -
A - ^
A -
A = '
A =
subroutine ;
's must equa^
PRESS
JD
B
R/S
R/S
R/S
R/S
R/S
B
R/S
R/S
R/S
R/S
R/S
B
R/S
R/S
R/S
R/S
R/S
10
-2^
R/S
R/S
R/S
R/S
R/S
s us
0.
_.
-.
id,
. .
..
...
—
-
n wh
DISPLAY
__.._
— .
- -
. . . . ._ .
_ „ ,
.
—
- •
.ch case it
-------
11-30
WORKSHEET
C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
10
10 a
lOb
lOc
10 d
lOe
10 f
10g
lOh
101
10j
10k
101
10m
lOn
ll
PROCEDURE
Market se_gment results
New DA share *
New DA volume
New DA VMT
New CP share *
New CP volume
New CP. VMT - ' -
New T share *' ' ~"
New T "volume • - -• . — -
New VP share *
New VP "volume .- - . . - —
New" VP VMT ' "="
New Other share * '
New VMT for this market segment
New Autos per wprker * ..._._
These results can also be recorded on Works
To analyze another market segment and have
the results aggregated with previously
analyzed market segments, press [CTjand GO.
TO STEP 4
To print** aggregated results of_ all pre-
vious market segments, press [E]
These results can also be recorded on Works
To analyze a new policy (no aggregation wit
previous market segments), press JA]and GO
TO STEP 2. (Since this sets memories to zero
data from previously tested policies must b
copied before A is pressed, and user-
supplied coefficients, if any, must be re-
entered in STEP 3.
* Printed but not displayed
**See Comments (Section 4) for retrieving
data without a printer.
ENTER
eet C-6
GO TO STEP 4
. - - -
leet C-6
PRESS
B
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
C
E
A
---
DISPLAY
s~'pA-nr8™~"
VOL'DA=14532
VMT'^A=232519
SVp' =.144
VOL'cp=15158
VMT'CP=97010
S'T =".'719
VOL'T =75884
S'VP g 0"
VOLIvp=__0_
VMT'vp- °
V L
VMT£fVr=3'29~4~99'
A/W --..195—
.
£VOL'DA=
2VMT'DA=
ZVOL'Cp=
^VMT'Cp=
ZVDL'T~="" ~
ZVOL'vp'
IVMT'vp= '
ZVMTxOT=
ZVMTTOT- '
AIVMTT"OT= "
%A^VMTTOT=
-------
11-31
Worksheet C-4 can be used as a guide for transforming the calculator
output to Worksheet C-6.
Non-Work Trips
The procedure for executing the calculator program for non-work
trips is somewhat different from that for work trips, because of the use
of user-supplied coefficients for the mode choice model.
The program is initialized by pressing "A" and step 3 is executed
immediately to enter the user-supplied coefficients. The coefficients to
be used in this analysis were derived from a joint choice model of shopping
trip frequency, destination, and mode choice developed using 1968 Washington,
B.C. data. They reflect the different choices available to non-work
travellers, and the different responses they make to changes in the trans-
portation system compared to work travellers. Using these coefficients,
both the choice of mode of ARZ bound travellers and the total number of ARZ
trips will be estimated. The coefficients are:
• IVTT - 0.0582
• OVTT/DIST - 0.459
2
These must be entered for the drive alone mode (Steps 3a and 3b) and
the transit mode (Steps 3k and 31). Only these modes need be considered,
since the model only predicts auto and transit mode choices.
1 Adler, T. J., and M. E. Ben-Akiva, "Joint Choice Model of Frequency,
Destination and Mode Choice for Shopping Trips." Transportation
Research Record 569, 1975
2
Note that "drive alone" means all auto trips for the non-work analysis
-------
WORKSHEET
11-32
C-4 PROGRAM STEPS
639.39
(3MQDE(VAN)-AGG-2(B)/7900110/ESE)
ARZ, Non-Shuttie:
NON-WORK TRIPS
(DA AVAILABLE)
USER INSTRUCTIONS
STEP
JL._
..2.
3
Ja
3b
3c
3d
3e
3f
3g
3h
31
3j
3k
31
3m
3n
3o
4
4a
4b
4c
4d
PROCEDURE
.*ead_card fa) ^-Partitioning is 639.39
.Initialize Storage Register? and stpre —
default Coefficients
If, .default .«Q£fficien_ts__are_usfidK._akip
to STEP 4
ENTER
USE OEXIONAI
PRESS
---
A
STB
OPTIONAL - Enter user-supplied coefficients
coefficient of IVTT-DA mode
coefficient of OVTT/DIST-DA mode
coefficient of OPTC/Y-DA mode
coefficient 4-DA mode
coefficient 5-DA mode
coefficient of IVTT-SR mode
coefficient of _pVTT/pIST_-SR mpde_
coefficient of OPTC/Y-SR mode
coefficient of incentives-SR mode
coefficient 5-SR mode .. . .
coefficient of IVTT-T mode
coefficient of OVTT/DIST-T mode
coefficient of OPTC/Y-T mode
coefficient 4-T mode
coefficient 5-1 mode
IP A MISTAKE IS MADE, BEGIN AGAIN AT STEP Ja
9 l^A V,SS2
6 2DA g-.549
e 3DA — -
9 4DA "
6 5DA =
B^R'-.SSZ
ft osu =-549
63SR =
6 4SR -_
e^sR-
8iT -
62T -
63! "
64T -.-__
95T -
,T,n»™™»i.^»»»»-»»»»»»» •••••••
Enter market segment data (from worksheet C'
round trip distance in miles j 0
household average annual income in $ ^ 0
enter "1" if income is not needed for this
segment. __
Average carpool occupancy ^ 0
Default = 2.5
Market segment population
,a.._.
DIST= 16
. J = 1_ .
OCCcp= 2.5
POP =3,096,780
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
.SJEQ
S1Q_
J5TQ
R/S
R/S
R/S
R/S
3\
00
01
02
03
04 '
05
06
07
08
09
10
11
_12_
JL3..
1.4
_
Ikjk
-
—
.. ..
DISPLAY
.29
-
-
.
_
...
-------
11-33
WORKSHEET C-4
(continue9)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
4e
4f
5
5a
5b
PROCEDURE
Average vanpool occupancy ^ 0
default — 10 " "
Vanpool circuity factor ^ 0
detault — 1.5
If a mistake is made, press [Cjand begin *
with STEP 4a
If carpool s ub routine- is irot used,-
go to STEP 6
OPTIONAL - Carpool Subroutine (data from wot
Enter average occupancies of two classes of
car-pools : - - - - - -
1. Average occupancy carpool class 1
2. Average occupancy carpool class 2 '
Enter level of service and mode share data
1. AlVTT - carpool class 1
2. AOVTT - carpool class 1
3. AOPTC - carpool class 1
4. AINCENT - carpool class 1 (0/1)
5. A^ - carpool class 1**
6. carpool: class 1 "share -as -a -fraction- -
of all carpools *
If a mis-take is made, --p-ress ^j and begin --
with STEP 5a
7. AIVTT- carpool class-.2
8. AOVTT - carpool class 2
9. AOPTC- - carpool class 2
0. A INCENT -"carpool class 2 (0/1)
1. Ac ~ carpool class 2**
2. carpool class 2 share as a fraction"
of all carpools*
If a mistake is made, press |C]J and begin
with STEP 5 a
* These two values should sum to 1.0
* If default coefficients are used, these
A's must equal zero.
ENTER
OCCyp = l.Q
CIRC = 1.5
GO TO -STEP 6
ksheet 1C- 3)
occcpl=
occcro=
A -
A =
A =
A -
A'. • -
srpi"
A =
A"= ' "
A -
A -
A =
OTTO ,
LJt Z
R/S
R/S
«
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
''RESS
-
:
'
'
2nd
i
—
E'
—
—
- -
DISPLAY
- -
- - • - -
38
_ — ._- ,
-
.. ......
. — ...__...
—
--
-------
11-34
After entering the user supplied model coefficients, the calculator
program is executed in a manner similar to that for work trips, with
changes in the base data, base mode shares, and level of service changes
as noted on Worksheet C-2. Market segment 3 is analyzed first, with
the market segment 4 analysis being" initialized at the completion of
market segment 3's calculations by pressing "C" as in the work trip
analysis. At this point Step 4 is executed for market segment 4 (the
coefficients of the mode choice model need not be entered a second time).
The worksheets corresponding to the non-work analysis of the non-shuttle
alternative appear below, along with the resulting calculator output.
CALCULATOR OUTPUT FOR NON-WORK TRIPS
DA Available1
0.025
1238712.
.0248991324
77107.13537
1233714.166
0.
0.
0.
.0252650738
78240.37535
0.
0.
0.
.9498357937
1233714.166
.0248991324
DA Not Available1
0.
0.
0.
0.
0.
0.
0.
0.
.0507288846
28277.29484
0.
0.
0.
.9492711154
Q.
0.
2
Summary Results
77107.13537
1233714.166
0.
0.
106517.6702
0.
0.
1238712.
1233714.166
-4997.834136
-.4034702284
29
30
31
32
33
34
35
36
37
38
39
2These values correspond to step 10 of Worksheet C-4.
These values correspond to step 11 of Worksheet C-4.
-------
11-35
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
.5c
6
6a
6b
6c
6d
6e
6f
6g
7
PROCEDURE
carpool subroutine results:
1. New average occupancy for all carpools
2. New tiarpool change in level of -_..,
service (for all carpools)
Enter A LOS for carpool class 1
Enter A LOS for carpool class 2
To repeat for another change in level of
service, press D^J arid repeat STEP 5c2.
Step 5 results can "also Ve' recorded on Wor
To (Continue with^ £TEP fi " ' ^
Store modal shares for this market segment
(from Worksheet -C-l) •
Base drive-alone .share, as .a .fraction **
(i.e., .645, not 64.5)
Base carpeol-share, as a fraction **
Base transit share as a fraction **
Base vanpool share as a fraction **
Base "other" share as a fraction **
Base autos per. worker - .-
If a mistake is made, press JGO TO| 1/X
and begin with STEP 6a •
Base VMT
If vanpool subroutine -is not used, go to —
STEP 8
OPTIONAL Vanpool Subroutine -
If VP = 0 in 6d and the revised, vanpool. .
share ^ 0, enter a "base" vanpool share on
which to pivot. Defaults —
.14 for firms with <_ 2000 employees
.06 for firms with J> 2000 employees
If a mistake is made, press B | and repeat
STEP 7
* printed but not displayed
** these five fractions must sum to 1
ENTER
A ' -' ^
CPl=2jjd_H!
Vp2=
sheet~-C-5.
Scp= n
ST = .025
svp=
S0 = .95
GO TO_STEP 8
•SVP =
PRESS
R/S
,--~-
R/S
R/S
rftfl
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
~^N
*IL
k.m.K
: :
1AI
DISPLAY
*occcp =
\ =
Cr — — — ^— ;
^^ •!• II ^^ « <•
-
A/W-=--025 _
VMT^,,,- 123871
::'.:
^^^^^^^
-------
11-36
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
8
8a
8b
8c
8d
8e
8f
8g
8h
81
8j
8k
81
8m
8n
80
9
9a
9b
9c
9d
9e
PROCEDURE
Enter _ch_anges in level of service from Works
If there are NO changes in level of service,
Press [|]and go to STEP 10; otherwise, .
Drive Alone changes:
AIVTTM
AOVTTM
AOPTCDA
A 4-.. **
A-5BA **
If there are no additional changes in LOS,
press (if) and go to STEP 10
Carpool changes from Step 5 or Worksheet C
if -optional STEP 5 was used:
AivTtCP •- •- -
AOVTTcp . . .
AOPTCpp
. V** ~ — " ~
AINCENTCP *
A5cp **
If there are no additional changes in LOS,
press [B] and go to STEP 10
Transit changes:
A IVTT
A OVTTj
A OPTCT
A4T **
A5T **
If a mistake is made, press |GO T0| |l/x|
and begxn at STEP oa~ •
If the vanpool share in 6d = 0 and STEP -7 —
was not used, GO TO STEP 10
OPTIONAL: Enter vanpool ..change .in LOS~7X?EPJi
A IVTT...,
VP - • - - —
AOVTT^
A OPTCyp
A INCENTyp ***
A 5 **
* This must equal 0 or 1 unless the carpool
must be between 0 and 1 inclusive.
** If default coefficients are used, these ^
*** 9d must equal zero if STEP 7 was used.
ENTER
heet C-2)
A = 1.2-Sfi
-A = -'.506
A - 0
A = o
ss ^^™iL™«^_
-5
• A
A - °
A - 0
A = _Q
A •= 0
A = 0
'A = --536
A = 0
A = 0
A = 0
- _.
GO TO ..STEP 10
Worksheet c"
A =
A =
A =
A = •
A =
subroutine :
's must equa!
1
JD
B
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
B
R/S
R/S
R/S
R/S
R/S
-b
R/S
R/S
R/S
R/S
R/S
s us
0.
=RESS
ed, a
—
-
n wh
DISPLAY
. _
-
_ ^_
. .
— ^. .
.ch case it
-------
11-37
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
10
lOa
lOb
10 c
10 d
lOe
10 f
10g
lOh
10 i
10j
10k
10]
10m
lOn
ll
PROCEDURE
Market segment results
New DA share *
New DA volume
New DA VMT
New CP share *
New CP volume'
New CP. VMT
New T share * ' -
New T ••• volume • - _...._
New VP share *
New VP volume . - .. . -
New VP VMT " "
New- Other share * ~
New VMT for this market segment
New' Autos per wprker * - --
These results can also be recorded on Works
To analyze another market segment and have
the results aggregated with previously
analyzed market segments, press [T) and GO
TO STEP 4
To print** aggregated results of_ all pre-
vious market segments, press (jfj
._....-.
These results can also be. recorded on Works
To analyze a new policy (no aggregation wit
previous market segments), press JTJand GO
TO STEP 2. (Since this sets memories to zero
data from previously tested policies must b
copied before A is pressed, and user-
supplied coefficients, if any, must be re-
entered in STEP 3.
* Printed but not displayed
**See Comments (Section 4) for retrieving
data without a printer.
ENTER
leet C-6
GO TO STEP 4
. . . -T— » — -
. . .
ieet C-6
PRESS
B
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
C
E
A
-
DISPLAY
S'DAjj=j024 __
VOL'nA=73294
VMT'rA=lJ.72,69!
S'rp' = °~"
VOL'Cp-V
VMT'CP= ' ~ "
S'T ='.'025
VOL'T =78339
S'yp - 0 •-"
VOL'vp=" 0 -
VMT'W= 0
S'0 =--.951
VMTfnT=5I7r^98
A/W1 =;02T"'
.
SVOL'DA=;
ZVMT'nA."~ ""
ivoi/CP=
IVMT'CP=
ZVDL'T~= " ~
EVOL'vp^'
ZVMT'vp'
EVMTxoT=
7"VMT ' " ' -~~
ivniTOT-
AIVMTT-0T= "
%AXVMTTOT=
.29
-------
11-38
The printed results for each market segment and the entire population
are entered on Worksheet C-6, once again using Worksheet C-4 as a
reference for the printed output values, and noting that the mode share
for the "other" mode refers to all non-work trips made to non-CBD
locations regardless of travel mode. The drive alone mode share reported
by the program is the share of all non-work person-trips in the urban
area made to the CBD by the auto. The transit share shown for market
segments 3 and 4 has a similar meaning.
Copies of Worksheet C-6 for the non-shuttle and shuttle scenarios
appear below. The change in VMT for work and non-work VMT are calculated
by subtracting the "New VMT" total from the "Base VMT" total for each
trip type.
Caution must be used in interpreting the VMT change estimates
calculated for non-work trips. Unlike the change in work VMT, the estimate
does not represent a direct measure of the areawide change in auto VMT.
Work trips are fixed with respect to destination and the VMT changes
shown on Worksheet C-6 are due solely to mode changes. Both the mode
and destination of non-work trips may change however. The VMT change
shown on Worksheet C-6 for non-work trips is the change in the VMT
accounted for by auto trips to the CBD. This change in CBD-bound VMT
could be caused by shifts of ARZ-bound travellers to transit from autos
or shifts in the destination of non-work auto trips from the CBD to other
locations in the urban area. Consequently, the areawide non-work VMT
impact of the ARZ proposals can not be determined without making assumptions
concerning the relative proportion of auto trips formerly made to the CBD
-------
WORKSHEET C-6
Results by Market Segment
Policy Downtown ARZ Without Shuttle
RESULTS BY MARKET SEGMENT
CO
vo
Population
Subgroup
1. Work Trips
with Drive Alone
2. Work Trips
without DA
3. Non-Work Trip
with DA
4. Non-Work Trip
without DA
TOTALS Work
Non-Work
Base
Autos
Per
Worker
.197
.067
l
.025
0
X^
Base
VMT
332,766
24,918
1,238,712
0
i 357, 684
1,238,712
New Drive Alone
Share
.138
0
.024
0
X
Vol
14,530
0
75,319
0
14,531
73,294
VMT
232,490
0
L, 172,691
0
232,490
1,172,699
New Carpool
Share
.144
.166
0
0
X
Vol
15,15£
3,84<
0
0
19,007
VMT
97,010
24,634
0
0
121,644
New Transit
Share
.719
.834
.026
.053
X
Vol
75,885
19,326
51,263
29,776
J5.210
.06A95
New Vanpool
Share
0
0
0
0
X
Vol
0
0
0
0
0
0
VMT
0
0
0
0
0
0
New
Other
Share
0
0
.949
.947
X
New
VMT
329,500
24,634
1,172,699
0
354, 133
1,172,69
New
Autos
Per
Worker
.195
.066
.024
0
X
Work
Non-Work
Change
in VMT
- 3,551
-66,013
Percent
Change
in VMT
-1.0
-5.3
-------
WORKSHEET
C-6
Results by Market Segment
Policy; Downtown ARZ with Shuttle
RESULTS BY MARKET SEGMENT
Population
Subgroup
1. Work Trips
with DA
2. Work Trips
without DA
3. Non-Work Trip
with DA
4. Non-Work Trip
without DA
Work
TOTALS
Non-Work
Base
Autos
Per
Worker
.197
.067
.025
0
X
Base
VMT
332,766
24,918
1,238,712
0
357,684
1,238,712
New Drive Alone
Share
.138
0
.025
0
X
Vol
14,539
0
76,935
0
14,538
77,107
VMT
232,616
0
1,233,714
0
232,617
1,233,714
New Carpool
Share
.144
.166
0
0
X
Vol
15,16(
3,852
0
0
19,017
0
VMT
97,063
24,650
0
0
121,713
0
New Transit
Share
.719
.834
.027
.053
X
Vol
75,868
19,322
82,865
29,776
95,192
106^518
New Vanpool
Share
0
0
0
0
X
Vol
0
0
0
0
0
0
VMT
0
0
0
0
0
0
New
Other
Share
0
0
.948
.947
X
New
VMT
329,500
24,634
1,233,71
0
354,329
1,233,712
New
Autos
Per
Worker
.195
.066
4.025
0
X
Work
Non-Work
Change
in VMT
-3355
-4998
Percent
Change
in VMT
-0.9
-0.4
-------
11-41
which are diverted to other destinations vs. auto trips to the CBD
which are diverted to transit, but still bound for the CBD.
However, it is possible to predict the changes in auto VMT within
the CBD, and the resulting reduction in CO emissions, related to the
proposed auto restrictions directly from the calculator outputs, if the
proportion of the total VMT which occurs in the CBD is known. Assuming
that 1/16 of the average trip mileage occurs in the CBD, non-work VMT in
the downtown area is reduced by 4,125 vehicle miles per day for the
non-work shuttle alternative, and 312 vehicle miles per day for the
shuttle option. These CBD VMT reductions may be used in the manual
emissions worksheets to determine the CO emissions impact of the competing
proposals.
H. Impact Assessment
Emissions, fuel consumption, and operating costs impacts can be
determined for portions of the travel affected by the ARZ using the methods
discussed in Chapter 4 of Volume I. The emissions impacts of the alterna-
tive ARZ plans are analyzed using separate sets of the manual emissions
worksheets. To analyze both the shuttle and non-shuttle alternative, the
worksheets must be executed three times; once each for the base case and
each of the alternative revised cases. Only the base case and non-shuttle
worksheets are discussed in detail for illustrative purposes. The calcula-
tions for the shuttle alternative are identical with the exception of dif-
ferences in the VMT values used in the procedures. The forecast year of
1982 used in the VMT analysis is assumed.
Areawide emissions impacts may be determined for work trips (Market
Segments 1 and 2) because the VMT impacts predicted by 3MODE(VAN)-AGG
refer to areawide total changes in work VMT. However, for non-work trips
-------
11-42
to the ARZ, only the change in vehicle emissions within the downtown area
may be calculated. The program predicts changes in the volume of auto tra-
vel to the downtown area, but because these changes could occur as a result
of either shifts to transit by auto users or simply changes in the destin-
ation of auto shopping trips, the net areawide change in VMT can not be
ascertained. The estimated change in downtown auto tripmaking and VMT
may be used to determine the impact of the proposed ARZ alternatives on
CO concentrations in the downtown area, which suffers from frequent
violations of the CO health standards.
Worksheet VI-A is used to record the base and revised work VMT and
trip lengths for market segments 1 and 2, and non-work VMT and trip
lengths for market segments 3 and A. The non-work VMT recorded on
Worksheet VI-A is the portion of downtown-bound VMT within the downtown
area (1/16 of the total non-work VMT recorded on Worksheet C-6) .
Similarly, the trip distance recorded for non-work trips is 0.5 mile,
corresponding to the one-way length of each non-work trip which occurs
in the downtown area. Table D.2 provides the required percentage of
cold starts for work and non-work trips for Worksheet VI-B.
Worksheets VI-C and VI-D are used to calculate the base and revised
emissions associated with downtown trip making. Emissions Tables D.3,
D.4 and D.5 provide the required cold-start emissions factors for use in
Worksheet VI-C with an ambient temperature of 50° used to enter the tables.
Only the CO factor need be obtained for non-work trips since only the
concentration of CO in the downtown area is of interest with respect to
non-work trips. The execution of Worksheet VI-D is straightforward,
-------
VI-A. INPUT TRAVEL DATA SUMMARY FOR EMISSIONS
ESTIMATION
Population
Subgroup
1
2
3
4
E =
Work VMT
(I or IV) 1
332,766
24,918
-
-
357,684
•
•
Average
Work Trip
Distance
(miles) (I)
(III or I)
8
8
Number of
One-Way
Work Trips
41,596
3,115
44,711
TOTAL WOR;;. VMT
TOTAL WORK TRIPS
I
IxlBase Alternative
| [Revised Alternative
Forecast Year: 1982
Policy:
Non-Work
VMT
(I or IV)
-
-
77,420
0
77,420
ARZ
*
Average
Non-Work
Trip
Distance
(miles) (I)
-
-
0.5
0.5
Non-Work
Trips
-
-
154,840
0
154,840
I
-C-
TOTAL NON-WORK VMT
TOTAL NON-WORK TRIPS
I
Source Worksheets are indicated in parentheses where applicable.
VMT and trips on worksheets I and IV in Appendix A must be multiplied
by the number of households per population subgroup.
TOTAL TRIPS
-------
VI-A. INPUT TRAVEL DATA SUMMARY FOR EMISSIONS
ESTIMATION
Population
Subgroup
1
2
3
4
£ =
Work VMT
(I or IV)1
329,500
24,634
-
-
354,134
«
•
Average
Work Trip
Distance
(miles) (I)
(IV or I)
8
8
-
-
^"^
Number of
One -Way
Work Trips
41,188
3,079
-
-
44,267
x
___
Base Alternative
Revised Alternative - Non-Shuttle
Forecast Year: 1982
Policy:
Non-Work
VMT
(I or IV)
-
-
73,293
0
73,293
ARZ
Average
Non-Work
Trip
Distance
(miles) (I)
-
-
0.5
0.5
£
Non-Work
Trips
-
-
146,587
0
146,587
1-1
TOTAL WORK VMT
TOTAL WORK TRIPS
TOTAL NON-WORK VMT
TOTAL NON-WORK TRIPS
Source Worksheets are indicated in parentheses where applicable.
VMT and trips on worksheets I and IV in Appendix A must be multiplied
by the number of households per population subgroup.
TOTAL TRIPC
-------
VI-B. COLD START FRACTIONS
_XJ Base Alternative
Revised Alternative
Policy:
ARZ
Subgroup:
All
Forecast Year:
1982
1. If Daily Parking Duration Data are Available;
% of Auto Trips by % of Auto Trips with % of Auto Trips by % of Auto Trips with
Catalyst-Equipped Parking Duration Non-Catalyst Equip- Parking Duration % of Auto Trips
Vehicles (Table D.I) 1 hour ped Vehicles 4 hours with Cold Starts
2. If Daily Parking Duration Data are not Available (Table III-2)
of Work Trip Cold Starts (Table D.2)
90
% Non-Work Trip Cold Starts (Table D.2)
57
-------
VI-B. COLD START FRACTIONS
I J Base Alternative
1. If Daily Parking Duration Data are Available;
XJ Revised Alternative
Policy:
ARZ
Subgroup:
All
Forecast Year:
1982
% of Auto Trips by % of Auto Trips with % of Auto Trips by % of Auto Trips with
Catalyst-Equipped Parking Duration Non-Catalyst Equip- Parking Duration % of Auto Trips
Vehicles (Table D.I) 1 hour ped Vehicles 4 hours with Cold Starts
X
M
I
2. If Daily Parking Duration Data are not Available
% of Work Trip Cold Starts (Table D.2)
90
% Non-Work Trip Cold Starts (Table D.2)
57
-------
VI-C. AUTO START-UP AND EVAPORATIVE EMISSIONS
| x| Base Alternative
II Revised Alternative
Policy: ARZ
Forecast Year:
1982
Temperature:
50 F
Work Trips
Non-Work Trips
(1)
Population
Subgroup
1,2
(2)
Z Cold
Starts.
(VI-B)1
90
(3)
Trips
(IV-A)
44,711
1
u
3
r-ltM
i-l U
o a
o. 3
hc(c)
C0(c)
N0x(c
HC(h)
*C(c)
C0(c)
N0x(c
HC(h)
HC(c)
C0(c)
10x(c
*C(h)
fc «
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
TOTALS TOTALS
Start-Up
Factors
12.3
199
3.4
6.0
HC
X co
Subgroups NOx
(5)
Emissions -
Col. 3 X Col. 4
(grams)
549,945
8,896,489
153,017
268,266
818,211
8,896,489
152,017
(6)
* Cold
Starts
(VI-B)
57
(7)
Trips
(VI-A)
154,840
1
(8)
Start-Up
Factors
CO 130
HC
"S ^ co
f •>
Subgroups NOx
(9)
Emissions -
Col. 3 X Col. It,
(grams)
20,129,200
20,129,200
M
Source Worksheets are Indicated In parentheses
where applicable
(c) Indicates cold start factor
(h) Indicates hot soak factor
both work and non-work start-up factors
obtained 'from the Indicated tables
Work Trip Start-
Up Emissions
(grams)
Non-Work Trip
Start-Up
Emissions
(grams)
Total Start-Up
Emissions
(grams)
-------
VI-C. AUTO START-UP AND EVAPORATIVE EMISSIONS
D
Base Alternative
Revised Alternative - Non-Shuttle
Policy:
ARZ
Forecast Year:
1982
Temperature:
50 F
Work Trips
I
Non-Work Trlpa
I
(1)
Population
Subgroup
1,2
TOTALS
(2)
Z Cold
Starts.
(VI-B)1
90
(3)
Trips
(IV-A)
44,188
Pollut-
ant2
HC(c)
C0(c)
N0x(c
HC(h)
HC(c)
C0(c)
N0x(c
HC(h)
HC(c)
C0(c)
V0x(c
!IC(h)
PI
u eg
tu H
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
TOTALS
Start-Up
Factors
12.3
199
3.4
6.0
HC
V CO
Subgroups NOx
(5)
Emissions "
Col. 3 X Col. 4
(grams)
543,512
8,793,412
150,239
265,128
808,640
8,793,412
150,239
(6)
Z Cold
Starts
(VI-B)
57
(7)
Trips
(VI-A)
146,587
1
(8)
Start-Up
Factors
CO 130
HC
} "" CO
/ J
Subgroups NOx
(9)
Emissions -
Col. 3 X Col. 4
(grams)
19,056,310
19,056,310
I
.t-
CO
Source Worksheets are indicated in parentheses
where applicable
(c) indicates cold start factor
(h) Indicates hot soak factor
both work and non-work start-up factors
obtained from the indicated tables
Work Trip Start-
Up Emissions
(grains)
Non-Work Trip
Start-Up
Emissions
(grams)
Total Start-Up
Emissions
(grams)
-------
VI-D. AUTO TRAVEL EMISSIONS
X
Base Alternative
Revised Alternative
Policy: ARZ
Forecast Year:
1982
Work Trips
1
Non-Work Trips
\ r
(1)
Population
Subgroup
,2
3,4
(2)
Average
Speed
25
_
(3)
VMT ,
(VI-A)1
357,684
_
TOTALS
(4)
Auto Travel
Factors
(Table D.7)
HC 1.6
CO 24.3
NOx 2. 1
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
v» HC
2_, co
NOx
Suberouns
(5)
Emissions •
Col. 3 X Col. 4
(grams)
572,294
8,691,721
751,136
572,294
8,691,721
751,136
(6)
Average
Speed
15
-J
n
(7)
VMT
(VI-A)
77,420
-
Sube
(8)
Auto Travel
Factors
(Table D.7)
HC
CO
NOx
HC
co 37.0
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
V HC
JL co
NOx
roups
(9)
Emissions -
Col. 7 X Col. 8
(grams)
2,864,540
L2, 864, 540
I
-f>
vo
Source Worksheets are Indicated In
parentheses where applicable
Work Trip Travel
Emissions
(grains)
Non-Work Trip
Travel Emissions
(grams)
Total VMT Travel
Emissions
(grams)
-------
VI-D. AUTO TRAVEL EMISSIONS
J Base Alternative
Revised Alternative
Policy:
HOV Lane - Non-Shuttle
Forecast Year: 1982
Work Trips
I
\ /
Non-Work Trips
I
(1)
Population
Subgroup
1,2
3,4
(2)
Average
Speed
25
-
(3)
VMT
(VI-A)1
354,134
-
TOTALS
(4)
Auto Travel
Factors
(Table D.7)
HC 1 • 6
co 24.3
NOx 2 • 1
HC
CO
NOx
HC
CO
NOx
1IC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
Z/ co
NOx
Subgroups
(5)
Emissions "
Col. 3 X Col. 4
(grams)
566,614
8,605,456
743,681
566,614
8,605,456
743,681
(6)
Average
Speed
30
15
(7)
VMT
(VI-A)
-
73,293
,
(8)
Auto Travel
Factors
(Table D.7)
HC
CO
NOx
HC
co 37.0
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
1IC
^~] CO
„ . NOx
Subgroups
(9)
Emissions "
Col. 7 X Col. 8
(grams)
2,711,841
2,711,841
I
Ul
o
Source Worksheets are Indicated In
parentheses where applicable
Work Trip Travel
Emissions
(grams)
Non-Work Trip
Travel Emissions
(grams)
Total VMT Travel
Emissions
(grams)
-------
11-51
again considering only CO emissions in the downtown area for non-work
trips. An average travel speed of 25 mph is assumed for work trips. For
the downtown portion of non-work trips, an average speed of 15 mph is
used, based on observations of traffic flow in the CBD. The predicted
base and revised emissions for the ARZ proposal are summarized in
Worksheet VI-E, which is also used to calculate the net and percentage
change in emissions for the revised alternative.
For work trips, the emissions figures for Worksheet VI-E are taken
directly from the previous worksheets. However, for non-work trips,
start-up emissions calculated in Worksheet VI-C represent the total
round trip emissions while only one-half of the total one-way non-work
trips begin in the downtown area. Therefore, to obtain the start up
CO emissions in the downtown area alone, the total emissions for the
base and revised cases must be divided by two.
Auto fuel consumption and operating cost estimates of CBD-bound
vehicle trips for work can be estimated using the manual method presented
in Section 4.2 and Appendix E of Volume I. The Base, Shuttle, and Non-
Shuttle input values are shown in Table II-l, along with the resulting
estimated fuel consumed and operating costs. The inputs reflect illustra-
tive projections of average vehicle weights, automotive technology, gaso-
line costs, and auto maintenance costs for the analysis year, 1982.
I. Interpretation of Results
Care must be taken in interpreting the predicted impacts of the ARZ
measures. The changes in work trip emissions shown on Worksheet VI-E
reflect the net change in areawide emissions associated with the1auto re-
-------
VI-E. SUMMARY OF CHANCES IN EMISSIONS
Revised Alternative
Policy:
ARZ, No Shuttle
Forecast Year: l982
(1)
Population
Subgroup
1,2
3,4
(ARZ only)
HC
CO
NOx
1IC
CO
NOx
HC
CO
1IC
CO
NO,
Base Emissions
(2)
Trip-Related
(VI-C)1
818,211
8,897,489
152,017
10,064,600
TOTALS
Source Worksheets are Indicated In
(3)
Travel
(VI-D)
572,294
8,691,721
751,136
2.864,540
HC
V""* CO
L-J NOx
Sub-
groups
(4)
Total
(Col. 2 + Col. 3)
1.390.505
17,589,210
903,153
12.929.140
1,390,505
17,589,210
903,153
Total Base
Emissions
(grams)
Revised Emissions
(5)
Trip-Related
(VI-C)
808 .640
8.793.41:
150, 23<
9.528.18C
(6)
Travel
(vi-n)
566r614
8.605.456
743,681
2,711,841
I
Sub-
groi
(7)
Total
CoL5+Col. 6)
1,17s, 2.54
17,398.86?
893,920
J? ,750,071
HC
;CO
NOx
(8)
Change In
Total
Emissions
Col. 4-CoL 7)
- 15,251
-190,342
- 9,233
-fiROJIQ
-15,251
-190,342
- 9,233
Total Change
ips In Emissions
(grans)
(9)
Percent
Change In
Emissions
(Col. 8/OoW)
x 100
-1.1
-1.1
-1.0
_<; T
-1.1
-1.1
-1.0
Percent
Change,
Total
Emissions
V
r-o
parentheses where applicable
-------
VI-E. SUMMARY OF CHANGES IN EMISSIONS
Revised Alternative
Policy:
ARZ, With Shuttle
Forecast Year: 1982
(1)
Population
Subgroup
1,2
3 4
(ARZ only)
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOX
Base Emissions
(2)
Trip-Related '
(VI-C)1
818,211
8,897,489
152,017
10,064,600
TOTALS
Source Worksheets are indicated in
parentheses where applicable
(3)
Travel
(VI-D)
572,294
8,691.721
751,136
2,864.540
HC
Zco
NOx
Sub-
groups
(4)
Total
(Col. 2 + Col. 3)
1,390.505
17,589,210
903,153
12.929.140
1,390,505
17,589,210
903,153
Total Base
Emissions
(grams)
Revised Emissions
(5)
Trip-Related
(VI-C)
808r640
8.793.41;
150, 23<
10,023,910
(6)
Travel
(VI-P)
566,614
8.605.456
743,681
2,852,959
Z
Sub-
gron
(7)
Total
[CoLS+CoL 6)
],37S,254
17,398,86?
893,920
12,876,86?
He'
;co
NOx
(8)
Change in
Total
Emissions
[Col. 4-CoL 7)
- 15.251
-190,342
- 9,233
- 52,271,
-15-, 251
-190,342
- 9,233
Total Change
ps in Emissions
(grans)
(9)
Percent
Change in
Emissions
(Col. 8/OolA)
x 100
-1.1
-1 .1
-1.0
-0.4
-1.1
-1.1
-1.0
Percent
Change,
Total
Emissions
CO
-------
11-54
TABLE II-l
Auto Fuel Consumption and Operating Cost Impacts for Work Trips
Variable
Inputs
WT
RFC
TEMP
GCOST
RMTCOST
TIME
DIST
TPHH*HHS
Estimated
Values
FUEL
(gal/day)
OCOST
($/day)
Alternative
Base
3.37
1.0
7
1.50
0.0595
19.2
8
44,711
24,428
$57,924
Shuttle
3.37
1.0
7
1.50
0.0595
19.2
8
44,260
24,182
$57,340
Non-Shuttle
3.37
1.0
7
1.50
0.0595
19.2
8
44,267
24,185
$57,349
Source
auto industry projections
auto industry projections
local weather data
Department of Energy
projections
local economic projections
DIST x 60/SPEED; SPEED
from Worksheet VI-D
Worksheet VI-A
Worksheet VI-A
I Procedure presented in
> Volume I, Appendix E
-------
11-55
strictions. The percentage changes, however, refer only to downtown-bound
travel. To develop an estimate of the areawide percentage change in
vehicle emissions, the total work trip vehicle emissions in the region
would have to be calculated and compared to the predicted emissions changes
associated with the ARZ proposals.
Also, the predicted reduction in downtown area CO estimates for non-
work trips should not be interpreted as a net reduction of CO emissions
throughout the urban area. This reduction is due in part to an overall
change in non-work auto use and in part to a shift in the destination
of non-work auto trips. Thus, a portion of the decrease in CO emissions
in downtown is accompanied by a similar increase in CO emissions in other
parts of the urban area. The information available from the analysis is
not sufficient to determine to what extent CO emissions in other areas
would increase if auto restriction is undertaken downtown.
The fuel consumption and operating cost impacts are also limited to
a portion of total urban travel: they reflect only changes in CBD-bound
work travel. Because no changes are predicted in average work trip length
or in average work trip speed due to the implementation of the ARZ, the
computed changes in fuel consumption and operating costs vary directly
with the changes in the number of auto work trips.
From the results summarized in Worksheets C-6 and VI-E, and in
Table II-l, it is apparent that the work trip VMT and emissions impacts of
both the shuttle and non-shuttle alternatives are marginal and that the
availability of the proposed shuttle service will have essentially no
impact on the travel behavior of downtown workers. More significant
impacts on vehicle emissions and a greater shuttle service impact on
-------
11-56
travel behavior are predicted for non-work trips, however. A 5.3 percent
decrease in auto VMT in the downtown area and an overall reduction in total
trip-making to the downtown area of 1.6 percent is predicted for the non-
shuttle alternative with a corresponding 5.3 percent reduction in CO
emissions. Adding the shuttle to the ARZ proposal makes the downtown area
much more attractive as a non-work destination, relative to the non-shuttle
alternative and leads to an overall increase in non-work trip making to
the area. Slightly fewer downtown-bound non-work trips are predicted for
the ARZ with shuttle than for the base case, and the impact of the ARZ
with shuttle on CO concentrations in the downtown area would be negligible.
In evaluating the relative merits of the shuttle and non-shuttle
alternatives, the importance of the viability of downtown as a major
shopping district would have to be weighed against the desire to reduce
CO concentrations in the area. The non-shuttle alternative is predicted
to lead to a decrease in total non-work travel to the area, while the
with-shuttle option will lead to little or no change in CO emissions
downtown. Another consideration in using and interpreting the results
of this analysis is that changes in the attractiveness of the ARZ
caused by factors other than changes in transportation level of service
have not been included in the analysis. If the area is made more
attractive as a shopping destination through the addition of new shopping
space or beautification measures, somewhat different impacts with respect
to total trip making to the area than those predicted by this simple
analysis would be expected. However, the relative impact of the shuttle
and non-shuttle alternatives on non-work travel to the ARZ is accurately
reflected by the results of this analysis.
Calculated using the base and revised "other" mode shares on Worksheets
C-l and C-6 to determine the number of trips bound for other parts
of the urban area.
-------
CASE STUDY III
BUS PRIORITY STRATEGIES FOR A RADIAL URBAN CORRIDOR
-------
CASE STUDY III; BUS PRIORITY STRATEGIES
FOR A RADIAL URBAN CORRIDOR1
A. Problem Presentation
A one mile wide radial travel corridor in a major urban area, served
by a single express bus route, has been identified as a candidate for
bus priority treatments. High peak period auto volumes in the corridor
are contributing to the metropolitan area's air quality problems, which
include violations of the ozone and CO health standards. Operations on
the bus route are adversely affected by conflicts with autos in the
traffic stream and the contribution of buses idling in traffic to the
area's air quality problems is an increasing concern, particularly for
the residents of the corridor.
A four-lane major arterial in the corridor serves work trips between
the corridor's residential areas and the major employment centers in the
central city. The section of the arterial under study is 5.3 miles in
length, stretching from the residential sections to a major transit transfer
facility where the bus route serving the corridor terminates. (See Figure
III-l) . Downtown-bound transit passengers transfer to a rapid transit
line at the transfer facility, while auto users continue into the central
city by a number of routes.
This Case Study was originally developed as part of the Responsive Analysis
Methods Project conducted by the Center for Transportation Studies,
Massachusetts Institute of Technology. It has been adapted from material
in: Responsive Transportation Analysis; Pocket Calculator Methods,
Volume II, Examples of Transportation Analyses, by M.L. Manheim, P. Furth,
and I. Salomon, MIT, August, 1978.
-------
Analysis Zones
Bus
Route
Terminals
Transit
Line
FIGURE III-l
Study Corrdior with Residential Analysis Zones
H
I
1x5
-------
III-3
B. Proposed Transportation Measures;
In order to encourage increased transit usage by improving the service
on the corridor's bus route, two bus priority measures have been proposed
for implementation by 1982:
1) Preferential Signal Control for buses at intersections;
2) Preferential Signal Control and Preferential Lanes for buses.
These measures are also expected to reduce bus pollutant emissions and
lower the cost of providing bus service in the corridor. Because the
volume of traffic is below the capacity of the arterial, no major impact
on auto level of service is expected for either measure. Reduced in-vehicle
travel time for bus trips is the only transportation level-of-service
change anticipated.
C. Selection of Analysis Technique
A detailed evaluation of the proposed measures' impacts on mobile
source emissions and transit operating costs is required to support a
decision on implementation. Both auto and bus emissions are of interest,
as well as bus fuel consumption (a measure of the operating cost impact
of the measures).
However, the budget and data available for the analysis are limited,
and the metropolitan area's computer-based model system is not well suited
to a detailed analysis of specific transit measures such as the two proposed
for the study corridor. Programmable calculator methods provide the
accuracy required for the analysis, but do not require extensive data or
costly computerized calculations to support accurate predictions.
-------
III-4
Calculator program 3MODE(VAN)-AGG (2.2.3)1 is selected for the analysis
of the mode choice impact of the proposed measures because of its efficiency
in using available data and its accuracy in representing behavioral decisions.
Since bus fuel consumption and emissions are of critical interest, programs
BUS (3.2.1), BUSPOL and ENERGY (A.3) are also selected. Finally, auto
emissions are predicted using the auto emissions worksheets (4.1).
The base data available for the corridor is limited to 1970 census
tract and block data. No mode split information is available and the
analysis budget is not large enough to allow traffic or transit counts to
be made. In order to develop base case data, calculator programs HHGEN
(2..2.1) and a two-mode calculator mode choice program (2MODE-AGG) are
selected to develop a sample of residential households for forecasting use
and to estimate existing mode splits based on the current levels of service
offered by the auto and transit modes in the corridor. Using this set of
programmable calculator methods, an estimate of the fuel consumption and
emissions impact of the proposed priority measures can be made using only
census and level-of-service data.
D. Overview of the Analysis
The analysis of the effects of changes in operating policies on bus
ridership and environmental factors consists of three major tasks:
• Collection, estimation, and preparation of data
• Equilibration of bus ridership and bus performance
• Calculation of the environmental impacts of the equilibrium bus
ridership
The section numbers following specified analysis methods refer to the
location of their description in Volume I. In addition, master copies
of worksheets are provided in Appendices C and D of Volume I.
-------
III-5
The three tasks consist of the nine steps shown in the analysis flow
diagram of Figure III-2. The six calculator programs used in the analysis
are also shown.
Task 1. Collection, Estimation, and Preparation of Data
This task includes Steps 1 to A. The data items needed to exercise
these steps include census data for the analysis corridor, approximate
existing travel times and costs for bus and auto trips, information on
the work trip destination for residents of the corridor, and detailed
data on the bus route operating in the corridor.
Census tract and block data from 1970; specifically total population,
population in group quarters, type of housing, auto ownership, income and
housing tenure; is required for Step 1. The output from program HHGEN
in Step 1 is combined with work trip destination data developed in Step 2
and level of service estimates for auto and transit from Steps 3 and 4 for
input to Task 2. Step A also includes an in-depth analysis of the corridor
bus route including traffic signal settings, line segment lengths (distances
between stops and intersections), and the location of bus stops. This
information is used by program BUS in Task 2.
Task 2. Equilibration
Because the bus route simulation program is ultimately used to test
the effects of different operating policies on route performance, it is
important that, in base case conditions, the predictions of ridership and
performance are in equilibrium. Ridership and performance on a bus route
are functions of each other:
Detailed descriptions of the programs may be found in M.L. Manheim, P.
Furth, and I. Salomon, Responsive Transportation Analysis Pocket Calculator
Methods. Volume 3; Program Library, Center for Transportation Studies,
Massachusetts Institute of Technology, October, 1978.
-------
III-6
Sample of HH
Characteristics
Collection,
Estimation, and
Preparation of
Data
I
I
Assign HH Randomly to
Destinations
Estimate Auto LOS
Estimate Transit LOS
Equilibration of
Ridership and
Performance
Estimate Bus Ridership
Simulate
Bus Performance
NO
/ Ridership and Performance
~( Consistent?
YES
Change Transit LOS
to Test New Policy?
I
Environ-
mental
Impacts
I
Incremental Revision
of Bus Ridership
NO i
Analysis o
Consumptio
f
n
Energy
9
\
Analysis of
Emissions
NO
FIGURE II1-2
Integrated Analysis Steps and Programs
Calculator
Program
HHGEN
2MODE-AGG
BUS
3MODE(VAN)-AGG
ENERGY
BUSPOL
-------
III-7
Ridership = D (Performance) (1)
Performance = S (Ridership) (2)
Equation (1) is the demand function, in which increases in transit
ridership demands are predicted as transit preformance improves. Equation (2)
is the supply function, in which decreases in transit performance are
predicted, due to congestion and longer boarding times, as transit ridership
increases. Equilibrium occurs when the values of ridership and per-
formance satisfy both Equations (1) and (2). Such values are termed
consistent, or equilibrium values. In this case study, equilibrium values
are found by iteratively applying in turn the separate programmable
calculator routines which represent Equations (1) and (2). This is done
in Steps 5-7, using either program 2MODE-AGG or program 3MODE(VAN)-AGG
for Equation (1), and program BUS for Equation (2).
Each time one of these programs is used, its input data are taken
from the most recent prediction of the other program. Iteration stops when
the input and output values of the ridership and performance variables no
longer change. When transit service is changed by the proposed transpor-
tation policies, equilibration is again necessary so that bus ridership
and performance reflect the influences that each has on the other.
Task 3. Environmental Impacts
When mode shares have been determined with ridership/performance
equilibration, bus energy consumption and emissions can be forecast based on
bus vehicle-miles travelled, travel speed in the corridor, and assumptions
about average fleet characteristics. Two programs, ENERGY and BUSPOL, are
-------
III-8
used to determine bus environmental impacts in Step 8 and 9, respectively.
In Step 9, auto emissions are also estimated using the auto emissions
worksheets.
E. Defining the Scope of the Analysis
Because the bus route serving the corridor in the peak period
carries almost exclusively downtown-bound passengers, only downtown work
trips are considered in the analysis. The route operates as a local
service only in the residential area shown in Figure III-l. Outside this
area, inboud stops are made only to discharge passengers. Therefore, the
analysis will focus on workers in the residential section of the corridor
who are bound for the central city during the morning peak hour. The
residential area is partitioned into three analysis zones, with the
divisions between zones roughly corresponding to census tract boundaries.
Only residents within one-half mile of the arterial on which the bus route
operates are considered because the number of bus passengers walking over
one-half mile to the route is assumed to be negligible.
A sample of 18 households, six from each analysis zone, are
generated for forecasting use. If more time was available for the analysis,
a somewhat larger sample could be employed. The amount of calculation
required to complete the analysis is highly related to the number of
households generated since the mode choices for each household must be
determined for the base case and for each proposed transportation measure.
-------
III-9
F. Input Data Development
Task 1 involves the development of base data for use in the analysis.
Program HHGEN is used along with some manual methods to develop a sample
of households whose behavior will be estimated. Also, the auto and transit
level of service for each analysis zone is determined for use in Task 2.
Inputs to HHGEN are provided by published census data at the tract
and block level. The output is a number (determined by the analyst) of
«
households with characteristics selected at random from distributions
based.on the census data. These households are representative of the
population of the study area. Six households represent the population of
each zone. Within each analysis zone, three census blocks were selected
at random and two households with at least one worker were generated for
each block, based on data specific to the block and the tract in which it
was located. The census data used as input to HHGEN for one of the blocks
is shown in Tables III-l, III-2, and III-3, the following HHGEN worksheets
(Figure III-3) illustrate the input format for the program. For this block,
three households were generated, two of which had at least one worker.
Figure III-4 illustrates the calculator output of HHGEN for each of three
generated households.
Step 2 involves assigning work places for each of the workers in
the households generated by HHGEN. Eleven destination areas listed in
the Census table "Social Characteristics of the Population"1 were candi-
dates for the work trip destination of workers in the household sample.
1This table reports work locations for tracts. Each block being analyzed
within a tract is assumed to have the same workplace distribution.
-------
TABLE III-l
Tract and Block Input Data from "Block Statistics", Boston Urban Area
Characteristics of Housing Units and Population, by Blocks: 1970-Con.
Middlesex County, Mass.
(Data e»clvoe vocont seasonal ond vacant migratory housing units For minimum bow lor derived figures (percent, overage. yraball. see le»l|
Dlocics
Within
Census
Tracts
503
504
5O5 . ...
SO* . ...
507
50S
601 . ...
602
603
604
605
607
608
6C9
610 . ..
611 . .
612
3561/
*"™^1o3
104
105
IU6
10;
<§$::::.
^^***>
Total
popu-
la-
tion
t _J
165
I3A
12?
165
302
311
7'J9
144
148
106
VO
275
719
266
141
45
50
W
6tltJ
119
55
49
43
131
Percent ol total population
^•^^^
\ f InA Un- «2
1 group 1 der years
Ne-Vqiwr / 18 ond
* gro XJcts/yeors over
6 26 21
32 19
12 43
1 18 76
6 41
- 29 15
- 21 19
28 15
- 79 14
19 73
21 26
20 22
19 17
- 31 II
32 5
72 29
36 16
34 18
CT) 28 17
^/ 24 4
45 1
- 22 20
35 18
- 19 28
21 M
- 21 29
1 Q 24 27
28 18
Year-round
lock
ng
some
or an
plumb
tacli
Total IKS
52
44 1
65 1
68 1
163
III 2
78 3
46
44
38
32
87 1
69
85 1
42
18
14
33 1
2411 22
18J
76
21
17
14
51
20
52 1
55
lousing units
Umtsin-
Struc
s^°^ tures
/One- \ ol
I unit 1 10 or
Mlruc I moie
yio.r-j/ unili
3
3
3 33
4 22
5 106
9 1
2
5
2
2 12
_
19
II
12
II
8
^*4 56
13
13
14
31
(OwnerJ
lack-
some Aver- ^»-^^
pr oH age fAver.^
plumb- num 1 age |
mg ler 1 vrlue 1 Per-
©lac* a< 1 (dol I cent
I** rooms ^ois)/Negru
27-61
19 - 7 1
15-62
24 1 57
24 - 65
40 1 62 22100
32 1 69 ...
22 - 62
71 59 18000
70 - 65
11 - 74
36 1 57
33 57
45 6 1 21000
21 - 69 19500
13 - 58 17400
10 - 60 I99UO
15 1 f>5 21300
(945) 7 59 /J3MXj)
']j - 54 - _
_
19 - 49 IA9DO
13 - 52 I1DJO
14 54 25500
48 - 51 15000
17 - 54>—- S
Occupied housing units
-
/Renter ^\
\.r~-^
lock- f — v
ing Aver^^
some Aver- / age \
or all age I con- 1
plumb- nuin- 1 tract 1
ing t«r 1 rent 1 Per-
f*\ loc* ol V (dot- / cent
ITolnlJ ties roams ^ors)^legro
25 - 5.3 117
24 1 58 137
50 1 3.6 146
42 - 4.1 157 2
136 - 3.4 Ii9
70 1 S.O 144
43 2 5.4 137
24 - $.6 178
21 - 51 115
17 - 52 119
21 - 4.7 156
48 - 5.2 135
35 53 121
38 49 138
20 - 4.5 136
4
4 ...
17 - 46 109
t£S> 14 41 <5?> -
^*7Wj - 32 ijo |
25 46 84
2
4
- - _ —
3
(TV i 33 (J62)
1.01 armor*
— , persons
per room
With
al
ptomb-
ing
focili
Total tics
1 1
_ _
2 1
2 2
$ S
1 1
_ .
4 4
1 1
_ _
» -.
3 3
2 2
1 1
I 1
.
2 2
144 142
8 8
12 12
1 1
2 2
- -
2 7
1 1
2 2
3 3
/W!M>
/ room-
/ ers.
f"^^ Wilnl twid
/ One- Veino!?! en.
person 1 hcodl or
house- / »f \ !•>-.']
ilmljl/ LOT !, \_ •.,/
S 6
• 5 1
27 II 2
18 5 -
65 17 1
IS 17
10 8 2
5 4
2 2
5 5 1
562
14 10
2 II -
9 II 4
572
432
32-
7 3 1
C?T> 253
-------
TABLE III-2
Tract Income Data from Census Tracts, Boston SMSA
Income Characteristics of the Population: 1970-Continued
lOoto bond an lampie. u« mi. For minimum bolt lor *>rtved I'tyurn (ptrctnl. nwdfen. tic.) and mrqnlno ef lymMt. ttt Hull
Census Tracts
IMCOMt IN m» OF FAMIUtt AND
UNilctAUO INDIVIDUALS
$100010$! V/9...
$3000ta$lV9?
$4.QCO to V 9<9
$5< — - — -^
*.'.tj.l I.LU..'» . .1
Balance ol SWSA In MNMCMK County -Con.
Tract
3387
9JJ
70
73
15
70
30
5J
75
5'
40
94
137
178
730
6
5
$11 307
$11 992
1 797
$10 <06
$11 148
344
$765'
Troct
3383
IM
4
5
8
4
6
13
9
34
41
37
114
351
183
45
$19 330
$72 794
1 078
$16 677
$19 675
774
$5 667
$7 787
Tract
3384
1 J»J
13
9
17
5
12
32
9
36
71
93
117
481
31?
118
$70 790
$25 451
1 437
$19 117
$7) 950
177
$3 500
$4 549
Tract
3185
i j»7
6
17
17
37
44
38
67
36
137
720
634
787
38
$17 571
$18 9VO
1 944
$14 909
367
$3 864
$5 525
Tract Intel
3561 3567
1 219 1 7J7
19 72
9 63
21 75
78 77
J6 54
40 96
72 49
69 IOJ
67 137
178 91
184 313
713 337
2«5 415
74 35
$11 091 $11 3.13
$11 774 $11 872(
1 467 2 397
$9 V96 $? 996
$10 476 $10 285
748 MO
$3 4tO $5 ?»
$4 458 $5 126
0?v
1 tM
15
75
77
91
39
69
60
143
150
179
361
370
316
55
7 459
$9 657
J9 859
'if.fjL
GT*M)
>5 VW
S Tract
/ 3544
1 *45
36
77
47
45
40
57
61
78
114
17}
249
331
740
767
40
$14 316
JI5 943
7 545
$13 271
$14 877
3CO
$5971
$4 753
Troct Trocl
3565 3564
15 78
57 74
30 II
39 64
73 67
34 85
94 67
89 142
147 139
139 164
39B 356
410 700
571 699
50 709
•10
$11 (14 $13 005
$17 781 $14 014
2 635 36/0
$10 601 $11 419
$10 062 $12 110
499 873
$3 625 $5 604
$4 789 $6 010
Trnct
3567
) m
17
1$
42
50
81
59
06
110
110
170
764
376
687
186
19
$13 753
$14 530
3 703
$10669
$11 976
981
$5 512
$6 191
Troct Trocl
3571 3577
1 J3» 1 fir
7 12
II II
19 14
23 6
43 70
27 77
19 19
35 71
61 3)
61 44
117 101
143 113
400 331
213 159
67 57
>16 750 $16 017
$19 V4I $19 950
1 45? 1 276
$14 797 $13 564
$17 439 $17 655
709 209
$5 9(2 » 115
$7 859 $6 486
Trocl Tract
3573 3574
t OS* 7M
5
4 5
78 13
79 7
41 19
40 31
24 41
67 36
65 35
48 17
172 94
741 143
707 711
41 79
7 25
$11 C60 $13 657
$13 117 $16 804
1 349 906
$10 595 $11 716
$11 556 $14 357
793 258
$5 392 $6 143
$5 930 $6 193
-------
TABLE III-3
Tract Auto Ownership Data from Census Tracts, Boston SMSA
Structural, Equipment, and Financial Characteristics of Housing Units: 1970-Continued
C""'?i Tract*
IDolo b«fd on lampto. SM lent, fat minimum txnt for derived figures (percent, median. «lc ) and memting of tymtwts. in te«l|
Tracl
330?
Tract Tract Troct Tract
.1.183 3384 3.1(15 3541
Bolootct of SMSA «
Trncl /Tract i
3542 1 354.1
n Middlesex County — Con.
V Troct
1 3544
Tracl Tract Tract Tract
3545 3544 3547 3571
Tracl
3572
Trocl
3573
AUTOMOBIICS AVAIlADlt
j .
Won ..
704
(00
301
28
W
J74
145
44}
145
II?
170
344
in
15}
10
441
734
777
155
I5»
39fl
(70
355
44
734
440
1 40?
3AB
52
5J?
371
I \ 417
351
14
V 37»
400
\l 3I«
\ 87B
1 14V
/ 158
«A^»
409
1 411
»A4
III
Tfl?
(4?
1 70?
eo)
13!
37?
W
-------
111-13
Generation of Sample Households (HHGEN-I(A))
USER WORKSHEET 1
Tract 3563
Step
1
2
3
4
5
6
Procedure
Set partition
Read banks la, 2a,
and 3
Enter tract data
(from Table 2 of
Block Statistics)
Computes interme-
diate values (above
values not
retained)
Enter tract data
from Census Tracts,
Table P-4 and
Table H-2
Compute adjustment
factors (auto own-
ership values from
step 5 not re-
tained).
To review other
adjustment factors:'
Enter
2
/ c / 5
population .= 6*43
% in grp. qtrs = 1
i 1-unit = 444
# owner = 945
avg value =3 23500
1 renter = 141S
avg rent =$ 727"
t 1-person HH = 517
§ HH with RBL = 4J_
mean fam. Inc =H 2 86
median U.I. Inc.* 5466
# HH with 1 car =JAL!
f HH wi th 2 cars =_jcLjL
# HH with 3+ cars = ^
# HH with 0 cars=_57Q_
Press
2nd Op 17
A
IEZS3
EZD
Ik/si
'IZI3
_J/SJ
R/SI orFD'
IR/SJ
JR/SJ
fRTsl
iRcTiirgi
ROlfTTl
Display/Print
799.19
population
% in grp qtrs.
9 1-unit
$ owner
avg value ($)
# renter
avg rent ($)
# 1-person HH
# HH with RBL
Mean Fam. Inc. ($)
Median U.I. Inc ($)
# HH with one car
# HH with 2 cars
2 HH with 3+ cars
# HH with 0 cars
2+ cars adj. factor
2+ cars adj. factor
1+ cars adj. factor
Income adj. factor
FIGURE II1-3
Sample HHGEN Worksheet
-------
111-14
Generation of Sample Households (HHGEN-I(A))
USER WORKSHEET 2
Tract 3553
Block 110
Step
7
8
9
10
n
12
13
Procedure
Enter block data (from
Block Statistics,
Table 2).
(To generate house-
holds based only on
tract level data skip
this step and qo to
step 9.)
Compute intermediate
values (above values
not retained).
Read banks Ib & 2b
To make program write
output automatically
on magnetic cards
(see step 12):
Generate n households
(may be repeated as
desired).
To write output, onto
magnetic cards
(see step 10):
Review characteris-
tics of household
generated.
Note: 2 cars = 2 or more
3 workers = 3 or more
HTYPE = J1 1'faml"|y house
Co multifamily
Enter
population = 735
% in grp.qtrs.= 0
# 1-unit =_44_
1 owner =_£2_
avg. value =259pQ
3? renter = W
avg rent =_T££_
i? 1 -person HH = 71
# HH with RBL = Q_
Number of house-
holds desired,
n = _3_
Place blank card
into card reader
during execution.
After card is run
through, replace
for each household
generated
Press
1
R/SI
R7S|
R/Sl
,RZS
RTSl
[R/Sl
R/S
I2nd!tst flq
S3
ED
IRCLJDZl
ED SI
(RHJEIF
iRCUjgg
Display/Print
population
% in grp.qtrs.
# 1-unit
JT owner
avg. value
# renter
avg. rent
# 1 -person HH
# HH with RBL
Household charac-
teristics
(printed only)
# cars
household size
HTYPE
income ($1000)
# workers
FIGURE III-3 (Cont.)
-------
111-15
Generation of Sample Households (HHGEN-1(A)/780317/PGF)
USER WORKSHEET 3
Step
14
Procedure
To generate households
from a different block
in the same tract,
read banks la and 2a
and go to step 7. To
generate households
from a different tract
go to step 1.
Enter
Press
Print/Display
FIGURE III-3 (Cont.)
-------
111-16
Three Households Generated
1. IF 1. Lives in one-family unit
19547. INC Income « $19,547
i. IJKR One worker
5. HHSZ Household size = 5
1. CRRS Auto ownership = 1
0. MF 2. liives in multi -family unit
15902. INC Income « $15,902
2. UKR Two workers
3. HH:-:Z Household size = 3
2. CRRS Auto ownership = 2
0. MF 3. Lives in multi -family unit
2907. INC income = $2,907
0. UKR Does not work
1. HHS2 Household size = 1 (unrelated individual)
0. CRRS Auto ownership • 0
FIGURE III-4
Generated Household Characteristics
-------
111-17
Each was assigned a zone number. The block for which the household
generation process has been illustrated has the following distribution
of work locations:
Destination
1
2
3
4
5
6
7
8
9
10
11
Number
245
552
58
569
182
961
42
15
18
0
116
Percent
.087
.195
.021
.201
.064
.340
.015-
.005
.030
.000
.041
Cumulative
Percent
.087
.282
.303
.504
.568
.909
.924
.929
.959
.959
1.000
Workers are assigned to work locations randomly. A random number between
0 and 1 generated by the programmable calculator, or obtained from a random
number table, is assigned to each worker. The work location assigned to
each worker is the first zone with cummulative frequency higher than the
random number generated for the worker. The random numbers and corres-
ponding work locations for the three workers in the block under analysis
are;
-------
111-18
Worker
1
2
3
Random Number
.2201
.5223
.1383
Work Location
Zone
3
5
2
The work trip length for each worker was determined by measuring the
centroid to centroid distance between the analysis block and the work
location zone on a map of the urban area.
The estimation of auto and transit level of service (Steps 3 and 4)
is accomplished manually. The level-of-service data items required for
auto trips between each work trip origin/destination pair are in-vehicle
travel time (IVTT), out-of-vehicle travel time (OVTT), and out-of-
pocket travel costs (OPTC). Auto OVTT at the origin was taken as one
minute for those living in multi-family structures, zero for single
family houses. OVTT at the destinations ranged from one to five
minutes, depending on parking availability and the density of develop-
ment, as indicated by the height of the buildings at each destination.
Auto IVTT was estimated based on"empirical knowledge of the area, aided
by field checks, and the trip length measures on a map of the area.
Auto OPTC was taken as auto running cost (4 cents per mile) plus the
average daily parking cost at the destinations, obtained by a brief
telephone survey of parking lot owners.
-------
111-19
Transit out-of-vehicle time at the work trip origin consists of
walk time plus half the headway. Walk time was based on the specific
household location within the zone (defined by the census block) and its
distance from the bus route. Walk times at the destination were based
on average distances between transit routes in the destination
zone. Transfer times for transit were taken as one-half the headway.
Transit in-vehicle time has two components: the time on the express
route to the transfer point and time beyond the transfer point. The
time beyond the transfer point to a given destination is assumed to be the
same for workers in all zones and was kept constant when considering
policies affecting the study corridor. The time to the transfer point
depends on the block in which the worker lives and is affected by the
proposed bus priority measures. The existing in-vehicle travel time for
transit was estimated using bus schedules and average subway speeds.
Transit fares were recorded, consistent with present values.
A completed Datasheet for one of the analysis zones is shown in
Figure III-5. The market segment size for each worker is determined
by dividing the total number of workers living in the census tract by
the number of workers in the tract generated for forecasting use.
(The table showing work locations for the tract also provides the
total number of workers.) A similar datasheet is required for each
of the remaining analysis zones.
Also in Step 4, the characteristics of the bus route operating in
the corridor are ascertained. Figure III-6 shows the simplified set of
stations, intersections and line segments used to represent the route.
-------
FIGURE 'IIX-5
DATASHEET .(2MODE-AGG-1(A)
Worker
f
—
1
2
3
4
5
6
TOTAL
Distance
to Work
mm
5.6
5.3
3
8
5.5
3
Yearly
Income
($)
|STO||08|
16810
11181
13205
10731
8173
8173
Autos/
Licensed
Driver
|STO||09(
1
.5
1
.5
1
0
INPUT DATA
Market
Segment
Size
mm
751
751
751
751
751
751
AUTO
IVTT
(min)
0
20
19
13
22
20
13
OVTT
(min)
fRDTj
5
2
4
4
4
4
OPTC
(*)
IRUN}
23
21
162
276
22
162
TRANSIT
IVTT
(min)
0
22.1
25.1
14.1
29.1
17.6
14.1
OVTT
(min)
17
17
8.5
18
16
11.5
|2nd| [
OPTC
U)
75
75
25
75
50
25
a
OUTPUT
Auto
Vol ume
i
.ransit
Volume
M
H
H
S3
1
-------
111-21
Bus direction
Station #1 Intersection Station //2 Line Station #3
#1 Segment (Terminal
Line Segment Line #3 Station)
#1 Segment #2
FIGURE III-6
Simplified Bus Route Representation
-------
111-22
A more detailed representation would yield a more accurate picture of
the operation of the route, but because bus operations along the route
are not a critical factor in the analysis, this simplified model is
adequate. For any analysis the stops and intersections used in
BUS would be an aggregation of the actual stop and intersection
pattern in order to keep the number of computations at a reasonable
level.
Tables Hl-4, III-5, and III-6 show the station, intersection, and
line segment characteristics required as input by BUS. The arrival rate
at each station in passengers per minute is determined from the passenger
demand calculated in Task II by 2 MODE-AGG and is assumed to be evenly
split between the two stations and constant throughout the two-hour time
period. The line segment lengths were determined from a map of the corri-
dor, and assumptions concerning where the aggregated stops should be placed
to best represent boarding patterns on the route. A constant auto volume
of 1500 vph is assumed based on observations of the corridor. The number
of lanes refers to the one-way lane capacity of the arterial. Traffic
signal data could be obtained from the traffic engineering departments of
the cities along the corridor, or even a brief observation of each signal
or a sample of signals.
G. Description of Model Application
Baje Modal Shares
Programs 2MODE-AGG and BUS are used to determine base transit shares
and bus operating characteristics for each of the market segments (repre-
sented by one of the workers in the generated sample). Figure III-7 shows
-------
II1-23
TABLE III-4
Station Characteristics
Station #
Arrival rate, VgTA
Fraction alighting, 3
1
*
0.0
2
*
0.2
3 (Terminal)
0.0
(1.0)
* To be determined by 2MODE-AGG
-------
111-24
TABLE II1-5
Line Segment Characteristics
Line Segment #
Length, d (ft.)
Limiting speed, V, TM
(ft/ml n) LIM
Auto volume, q.
(veh/min) A
Number of lanes, n
1
5000
3080
(= 35 mph)
25
2
2
1000
3080
25
2
3
5000
3080
25
2
TABLE III-6
Intersection Characteristics
Intersection #
Signal cycle length,
c (min)
Fraction green time,
X
Auto volume, q.
(veh/min) A
Number of lanes, n
1.17
0.6
25
-------
111-25
Transit Fare, Auto Level-of-
Service, Worker- Socio-
economic Characteristics,
Trip Length
Initial Estimate of
Transit Travel Times
\ n -
Predict Ridership
2 MODE-AGG
Auto Users
Bus Ridership
Predict Performance
BUS
Output Transit
Travel Times
NO
Output Transit Travel Time Consistent
with Most Recent Transit Travel Time
Input to 2 MODE-AGG ?
YES
Equilibrium Condition
Reached
FIGURE III-7
Base Case Equilibrium Data Flows
-------
111-26
the flow of data between the two programs, which are run iteratively until
bus ridership and bus performance are in equilibrium. An initial bus
ridership estimate is developed using 2MODE-AGG and this ridership is
used to BUS to determine the level of service which could be offered at
that ridership level. If the level of service predicted by BUS is signifi-
cantly different from that used as input to 2MODE-AGG to predict bus rider-
ship, a new run of 2MODE-AGG is made using the output of BUS for the
level-of-service variables.
A 2MODE-AGG user worksheet for the first worker is shown in Figure
III-8, illustrating the required order of input values and the outputs
produced by the program for each worker. Program steps 5 through 7 are
repeated 18 times, to complete the initial estimate of base mode shares and
transit volumes for use in program BUS. Figure III-9 shows completed pro-
gram datasheets including transit and auto volumes for each worker. The
total estimated work trip transit ridership is 2767, which translates into
a passenger arrival rate of 11.5 passengers per minute at each of the two
stations in the corridor.
This passenger demand rate, along with the other intersection, line
segment, and station information collected for the bus route is entered
into program BUS to determine if the level-of-service assumed in the
estimation of bus pasenger demand is consistent with the service which
can be provided at the resulting passenger demand rate.
Following the sample user worksheets for 2MODE-AGG is an illustrative
run of BUS (Figure 111-10). This sample run assumes a headway of ten
minutes and a passenger arrival rate corresponding to a lower demand
-------
111-27
USER WORKSHEET - BINARY MODE CHOICE WITH AGGREGATION (2 HODE-AGG-UA)
Default values in [ ]. Trip variables take on one-way values.
Step
1.
2.
3a.
b.
c.
4.
5.
Procedure
Read card, sides A and B
Set printing option if using
printer-
Load default values of coef-
ficients
Enter any alternate values
desired:
(1 ) Coefficient of IVTT [-.03]
(2) Coefficient of OVTT [-0.34]
(3! efficientofOPTC[-50]
(4) Coefficient of income
[.0000895]
(5) Coefficient of AALD [2.84]
(6) Auto constant [-2]
List coefficients
Initialize modal volume
accumulation registers
Enter socio-economic charac-
teristics
(1) Distance to work
(2) Yearly income ($)
(3) Autos available/licensed
driver
(4) Market segment size
Enter
a, s
i
a «
a -
a »
:onst -
d = 5.6
Inc = 16810
AALD = 1.0
Pop = 751
Press
2nd Stflgl
E
IsTolfofl
ISTO] f02~|
fSTOlfOSl
BED El
r$ro] tou
fSTOl |06]
[RUN]*
fRum*
fRUfl*
(RUN]*
[RUN]*
mm
IsTolfoTl
ISTOl |08)
I STOl |09)
Isum
Print/Display
al
a2
a3
a4
a5
const
FIGURE III-8
-------
111-28
Step
6.
7.
8.
9.
10.
Procedure
Enter auto level-of-service
data
(1) In-vehicle travel time fain)
(2) Out-of-vehrle travel time fain!
(3) Out-of-pockettravelcostU)
See note **.
Enter transit level-of-service
and compute modal shares and
volumes.
(1) In-vehicle travel time (min)
(2) Out-of-veh. travel time(min)
(3) Out-of-pocket travel cjs>t.(£;
Read modal shares (P ) and
volumes (V ).
m
(Accumulates auto volume in
register 98, transit volume in
register 99.)
^ ^ **
See note
Repeat Steps 5-7 for every
worker in sample (unchanged
values in Step 5 need not be
entered).
Print/display accumulated
modal volumes
To do parametric variations
over a sample go to Step 4.
Enter
IVTTA= 20
OVTT.= 5
OPTC, = 23
A •- - -
IVTTT* 22.1
OVTTT= 17
OPTCr= 75
Repeat for al 1
Press
a
1103
^^
IRUN
EOi
[RUI?
JRUNl*
FRUN!*
_____ *
|RUN
households
llndl [Cj]
fRUNj*
Print/Display
PA =
VA =
PT s
VT =
auto volume
transit volume
FIGURE III-8 (Cont.)
-------
DATASHEET (2MODE-AGG-HA)
Worker
it
—
1
2
3
4
5
6
TOTAL
Distance
to Work
mm
5.6
5.3
3
8
5.5
3
Yearly
Income
($)
fSTO"|{08|
16810
11181
13205
10731
8173
8173
Autos/
.icensed
Driver
|STO||09|
1
.5
1
.5
1
0
INPUT DATA
Ma v> I/at
narKct
Segment
Size
fSTOJlTol
751
751
751
751
751
751
AUTO
IVTT
(min)
0
20
19
13
22
20
13
OVTT
(min)
flDlT]
5
2
4
4
4
4
OPTC
(«)
|RUN|
23
21
162
276
22
162
TRANSIT
IVTT
(min)
E
22.1
25.1
14.1
29.1
17.6
14.1
OVTT
(min)
mm
17
17
8.5
18
16
11.5
|?nril 1
OPTC
(#)
75
75
25
75
50
25
SI
OUTPUT I
Auto •
Vol ume
724
645
666
423
690
0
Transit
Volume
27
106
85
328
61
751
FIGURE III-9
VO
-------
DATASHEET (-2MODE-AG6-1 (A)
Worker
f
—
13
14
15
16
17
18
TOTAL
Distance
to Work
mm
12
7.5
5
7.6
7.5
5
Yearly
Income
($)
[STOl{08|
23590
8901
15075
4651
10676
12350
Autos/
Licensed
Driver
|STO 09|
.5
.5
1
1
.33
.5
INPUT DATA
Marmot
Segment
Size
HOD0
332
332
332
332
332
332
AUTO
IVTT
(min)
0
38
26
19
27
26
19
OVTT
(min)
KUN|
2
3
3
5
3
3
OPTC
(0
|RUN|
48
30
20
31
30
20
TRANSIT
IVTT
(min)
E
43.9
29.9
26.4
34.4
29.9
26.4
OVTT
(min)
ism
12.5
10
10.5
21
16
11.5
OPTC
(O
75
50
25
75
50
25
||ndj [F]
OUTPUT i
Auto
Volume
295
226
316
310
222
263
I
Transit
Volume
37
106
17
22
110
69
FIGURE III-9 (Cont.)
I
w
o
-------
DATASHEET.(2MODE-AGG-1(A)
Worker
if
—
7
8
9
10
11
12
TOTAL
Distance
to Work
jSTO]p7|
6.5
9.6
9.6
.6
9
4
Yearly
Income
($)
[STO][08|
'-•• • . . .1 C,i._i 1
14911
3782
8821
18173
15338
14457
Autos/
Licensed
Driver
|STO |09|
l
0
.5
.66
.5
.5
INPUT DATA
Market
Segment
Size
[STOllTti
398
398
398
398
398
398
AUTO
IVTT
(min)
0
23
28
30
30
25
16
OVTT
(min)
|RUN|
3
3
6
5
5
4
OPTC
(*)
IRUN}
26
90
280
280
235
166
TRANSIT
IVTT
(min)
HD
23.8
34.3
28.3
28.3
25.3
20.3
OVTT
(min)
15
17
10
10
16.5
13.5
• •-i ..... fm
l^dT [
OPTC
U)
-\rl
[RDTI
50
50
50
50
50
50
93
OUTPUT
Auto
Vol ume
378
0
107
291
260
304
"r tins it
Vol ume
20
398
291
107
138
94
FIGURE III-9 (continued)
i
CO
-------
111-32
USER INSTRUCTIONS - Bus Route Simulation [BUS-2(A),-2(B)]
A. Initialization
Step
A-l.
A-2.
A-3.
A-4.
A-5.
Procedure
(TI 52) Load card 3
(TI 59) Set partition
Load banks 1 and 2
Initialize (stores default
value automatically)
Replace default values if
desired *.
(1) Random number seed
[501; 0-199017]
(2) Vehicle capacity per lane
at saturation flow (vpm/
lane) [30]
(3) Boarding/alighting rate
(min/pax) [0.0333]
(4) Bus capacity, seated and
standing (pax) [75]
(5) Vehicle-acceleration
(ft/min2) [15,840 =
3 mph/sec]
(6) Clearance time between
busses (min) [0.25]
(7) Congestion curve factor
[0.10,0.05-0.20]
Enter equivalent bussts per
minute, where
qB = busses per minute, and
k = autos per bus equivalency
factor [2]
Set printing option if using
printer
Enter
6
[U&t dt^autt v
to St&p A-4\
Y =
QCL =
b =
vc =
a =
VIR"
OLK
J =
QD= 0.1
k * 2
Ppess
tMlMIQ
mm
1&LLZA tlVLQUC
[sron mi
rsroi mi
fSTOl QT |
[sToi ini
fSTOl (D^j
tsrg og
(2nd |St fl
til
Print/Display
/ioot; &kip
0.2
FIGURE 111-10
-------
111-33
USER INSTRUCTIONS - Bus Route Simulation [BUS-2(A) ,-2(B)]
A. Initialization - p. 2
Step
A-6.
A-7.
Procedure
Set up the Detailed Simulation
Record for the route being
studied. The first page shoulc
be entitled "Header;" it re-
cords the passage of the bus
through the origin station and
the first line segment. The
header is followed by a se-
quence of "Station" and "Inter-
section" Records in the same
order as stations and inter-
sections occur on the route.
These records also include a
line segment, since all sta-
tions and intersections are
separated by line segments.
Finally there should be a Ter-
minal Station Record for the
last station. Altogether there
should be one record for each
station and intersection
modeled on the route. This
series of records will record
the movements of 4 busses
through the route; to simulate
more than. 4 bus runs, a dupli-
cate of the set of records is
required for each 4 additional
runs.
For the first run, values for
*LAST' *^e time the previous
bus left the intersection, and
V. £py» the number of passengers
left unbearded at a station by
the previous bus, must be ini-
tialized by the user. (Usually
Vi rrr should be -ero and t, flc:T
L L. 1 1 L>r*^ 1
estimated knowing the average
headway.)
Enter
The. VetaU.e.d S
organized and
jjo-tdowoig. Th
0^ the U6eA In
the. &te,p-by-&t
£ottou)e.d £o>i
loute. &e.gme.nt.
'^LEFT ^ ^^
*LAST Jii> i.nU
bllA OM
Press
UJUtud out
_ Jiejno4.ning
tfiu.ctLon&
jp piocerfote
one 0^ eacJ
•LaLLzed at
iaLize.d to
•iveA minuA
Print/Display
-------
111-34
USER INSTRUCTIONS - Bus Route Simulation [BUS-2(A) ,2(B)]
A. Initialization - p. 3
Step
Procedure
Enter
Press
Print/Display
A-8,,
Simulate running the first bus
through each segment of the
route, recording times and
passenger volumes on the De-
tailed Simulation Record com-
piled in step 7, following
the User Worksheets B, C, D
and E for line segments, in-
tersections, stations, and
terminal stations, respective!.
To run second and subsequent
busses, see User Worksheet F.
*
As many of these values as desired may be changed, and they may be
changed at any time during the simulation. In particular, the vehicle
capacity per lane (qCL) and the congestion curve factor (J) may be different
for different line segments over the route. Once changed, the new values
remain in memory unless changed again.
-------
111-35
USER INSTRUCTION'S - Bus Route Simulation [BUS-2(A) ,-2(B)]
B. Line Segment
(Data Q4.v
-------
BUS-2(A),(B)
111-36
USER INSTRUCTIONS- Bus Route Simulation [BUS-2(A) ,-2(B)]
C. Intersection
(Data. Qlvtn heAe.
IntesiAe.cti.on #1,
Bo*
Step
C-l.
C-2.
C-3.
C-4.
Procedure
(TI 52 only):
Read card 1 (not necessary if
card 1 is already in memory)
Enter data:
(1) Time the previous bus left
this intersection
(2) Signal cycle length (min)
(3) Fraction of cycle length
which is effective green
time
it
(4) Auto volume (vpm)
*
(5) Number of lanes
Computed expected delay (min)
and cumulative clock time (mini
Record t-j- and tcy^ on the simu-
lation record for this run.
For the next run, record tpUM
as t. .<., for this intersection.
Enter
t,fle.T =-7.50
LHo 1
c = 1.17
X = 0.60
QA - 25*
r\
n - 2*
(See Example.
Press
ISTO |oi|
STOI 105J
STO [06
STOJ |07|
ISTO losl
o
I RUN
S-cmu&L&con
Print/Display
tr » 0.174
^UH = 2'674
Record, p. 21
**Notes at end of Worksheet E.
-------
111-37
USER INSTRUCTIONS - Bus Route Simulation [BUS-2(A) ,-2(B)]
D. Station
(Vcuta g-tuen fieAe ion Station #1, Bui
Step
D-l.
D-2-.
D-3.
D-4.
Procedure
(TI 52 only:) Read card 2.
Enter data:
(1) Time the previous bus left
i*hic c^^^irvn
UN I -> d La L I UH
(2) Number of pax left un-
boarded as the last bus
left this station
(3) Mean passenger arrival
rate (pax/mi n)
(4) Fraction of pax on the
vehicle alighting at this
station
Compute:
(1) Number of pax left un-
boarded
(2) Number of pax boarding
(3) Number of pax alighting
(4) Total delay (min)
(5) Cumulative clock time
\iu i \\ 1
Recall total volume on bus
Record the above values
on the simula-
tion record for this run. For
the next run record V^py as
VLEFT *nd record Wl as tlAST
for this station.
Enter
IAST= Jrt
VLFFT = °
7/1
•Z
VCT/\
O 1 1\
l\
3 = °
[See Example. Si
Press
|STO| |oi
STO [02J
STO |05
ISTO [06|
@
IRUN **
RUNI
IRUN **
[RUN]**
IRCLJ |T6
nutation Re<
Print/Display
V'LEFT ' -±-
a
VOM =
Un
VOFF = °
ur r
tr = .550
1 -.—..- .. . . ..
*CUH" '55°
V = 9
.ond, p.1 . 1
Notes at end of worksheet E.
-------
111-38
BUS-2(A),(B)
USER INSTRUCTIONS - Bus Route Simulation [BUS-2(A) ,-2(B)]
E. Terminal Station
("Data, g-iven HeA.e fan TvunLnaJi Station, Bo6 #J)
Step
E-l.
E-2.
E-3.
Procedure
(TI 52 only): Read card 3.
Compute:
(1) Time spent unloading pax
(min)
(2) Cumulative clock time
(min)
(3) Running time over the
whole route for this bus
(min)
(4) Total # pax carried in this
run
Record the above values on
the simulation record.
Enter
(See Example S
Press'
1 2nd"1 [F|
|RUNJ**
RUN j **
i 1**
|RUN|
{.mutation Re
Print/Display
tr = 0.366
1. • • " —
tCUM= 5.996
v,un
tn = 5.996
VCUM = l3
cold, p. 4-.)
Items (4) and (5) need not be entered if q. and n have the same values
as when they were last entered., whether in a line segment or an inter-
section, and are still stored in registers 07 and 08.
**,
R/S| on TI 59. Not to be pressed if printing option is set.
-------
111-39
Simulation Record - p.l
Header
STATION
(card 2)
ENTER
VSTA=1-2
PRESS
fSTO'|05|
e = jj_i!3io3
LINE SEGMENT- #J_
(card 1)
ENTER PRESS
d" 5QOO
n = 2
ISTO! 105!
IST01 106|
ISTO (OB
INPUT/
OUTPUT
ENTER
*LAST
VLEFT
RECORD
V
VLEFT
VON
VOFF
tCUM
V
ENTER
tLAST
RECORD
V
(I/ lenKu
Bus Simulation Records
Bus i i
t = 0.0
= -10.0
0
Q
9
0
B 0.550
= 0.550
9
v
= -9.45
[eAtimcuted cu>
0.550 - 10.01
- 7.949
« 2.499
- (9)
Bus | 2
t_ - 10.0
U " "" ""
= .550
0
0
S
0
= .516
= JO. 576
t
T
*'**
= 2.499
= 7.949
= 72.466
- m
i>u.mj tine. 4(
Bus i
t0 =
s
0
ss
_
s
=
=
J
1
***
s
=
s
gmenti and i
Bus I
tfl =
s
s
s
_
.
—
—
s
V
***
s
s
_
•*"•"«« '
Press
^0
ISTO (oil
STO| f02|
• 1**
JRUNI
SI**
IRUNI**
i 1**
[RUN]
[RcTj [Tel
ISTO] (oTj
USD**
**
to be pressed if printing option is set. |R/Slon TI 59.
is usually the same for every bus.so it is sufficient to enter tT (as cal-
culated for the first bus) and press {£]. See sheet Fof the User Instructions.
-------
111-40
Simulation Record - p.2
Intersection
INTERSECTION
1 1
(card 1)
ENTER PRESS
c« l-l?
A.= °'6
qA=(25)*
n= (2)*
(q. and n
[STOl |05
ate. ifie.
LINE SEGMENT *_£_
(card 1)
ENTER PRESS
d= 1000
Y..IM-3080
Lin —
Q« « (25)*
A J
n = (2)*
STO) 105|
13 El
STOi 1071
mm
INPUT/
OUTPUT
ENTER
^AST
RECORD
IT
*CUM
V
&ame ai <
ENTER
*LAST
RECORD
*T
^UM
V
Bus Simulation Records
Bus f J
= -7.50
0.174
2.674
(9)
in Line. Segmen
V
s -7.33
.557
s 3.230
= (9)
Bus f _JL_
= 2.674***
m 0.174
= 72.540
M
t *1 and
V
= 3.230***
.557
= J3./99
= («)
Bus 1
***
a
s
_
=
-tkiA need
V
***
8
a
s
••
s
Bus I
***
=
.
not be eni
^ r
***
s
e
=
Press
(srof [oT|
RUN **
IRCLJ [Te]
ted ago^n. )
{STO] [01]
f— — i**
JRUNl
El OH
Need not be entered if values are the same as when last entered
**fe/SJon TI 59. Mot ;n be pressed if printing option is set
***See note on Header sheet.
-------
111-41
Simulation Record - p.3
Station.
STATION f 2_
(card 2)
ENTER PRESS
VSTA*0.6
B «= 0.2
UTOlEJl
ST01|06)
LINE SEGMENT #J_
(card 1)
ENTER PRESS
d* 5000
vi-TMs-2Q80.
QA ei25)*
n
« s (2)*
iSTOifOS)
ISTOlloe]
ISTOllo?!
fSTO)(Q8]
INPUT/
OUTPUT
ENTER
*LAST
VLEFT
RECORD
V
VLEFT
VON
VOFF
*t
*CUM
V
ENTER
\AST
RECORD
*T
1CUM
V
Bus Simulation Records
Bus * 1
-5.77
0
» 0
4
2
« .450
« 3.6&0
n
1
J
» -5.32
B 1.949
s 5. 650
* (11}
Bus # 2
= .3..5*
0
* 0
= 5
= 7
= .450
= 13.647
12
1
1
= 5.^'"
(See noie o
4/ieet FJ
s 7.949
s J5.596
= (12]
Bus 1
B
C
S
=
S
St
m
S
1
I
***
sr
IteeA I»t4^u
^
s
s
—
Bus I
s
s
s
s
s
=
8
B
1
I
***
S
cJJiom
s
s
s
Press
fSTO] {Oil
fSTOJ (02]
IRUN **
. 1 **
IRUN]
._ j if ^
IRUN
,- « **
{RUN
mm
|sTol gT)
i— — - 1**
IRUN]
(RCL| DU
*
***
Need not be entered if values are the same as when last entered
«.
Jon TI 59. Not to be pressed if printing option is set
See note on hp.adpr s
-------
111-42
Simulation Record - p.4
terminal Station
TERMINAL
STATION
(card 3)
INPUT/
OUTPUT
RECORD
*T
tCUM
*R
VCUM
Bus Simulation Records
Bus | 1
.366
= 5.995
" 5.996
= M"
Bus 1 2
= Q.400
= 75.996
5.99^
T3
Bus #
_
=
Bus I
=
w
=
Press
r— i **
[RUN]
t — i **
[RUN!
,_.., **
[RUN
felon TI 59. Not to be pressed if or(ntil,9 opticm 1
1s ,.t
-------
111-43
level than that predicted by 2MODE-AGG. It is included for illustrative
purposes as the actual BUS runs for the corridor analysis were not
available. The following input values and procedures would be changed
for the analysis of a route with three-minute headways and a passenger
demand rate of 11.5 passengers per minute:
• Step A-4 - qfi = 0.33
• Step B-2 (1) 1^^ = -2-45
. Step C-2 (2) tLAST = -0.50
. Step D-2 (1) tLAST = -3.0
• Step D-2 (3) V = 11.5
• Computations in D-3 and E-2 would reflect the higher demand rate
rate and bus frequencies.
• Additional line segments and one additional station would be
included in the run.
• The simulation record would also reflect the above-mentioned
changes .
To achieve equilibrium for the base case conditions on the bus route,
two iterations (two runs of 2MODE-AGG and BUS) were required. The final
results predicted average bus loads of 57 passengers into the transfer
point, with just over 28 minutes required to run the entire length of the
route.
Step 7 of the analysis involves the prediction of incremental
changes in bus ridership due to the proposed bus priority measures.
To estimate the level-of-service impact of the two measures, program
BUS is run with the following modifications to the base case:
-------
III-AA
• increased effective green time and cycle length at the inter-
section to reflect the preferential signal control for buses
(applies to both proposed priority strategies).
• On each line segment and at the intersection, the number of
lanes is reduced to one and auto volume to zero to represent
the reserved bus lane.
From the two runs of BUS under bus priority conditions (one run for each
policy being analyzed), estimates of the improvement in bus IVTT were
obtained.
Program 3MODE(VAN)-AGG is then used to determine the incremental
increase in bus ridership which would result from the predicted improve-
ment in bus level of service. Because no information on carpooling
activity is available for the study corridor, a two-mode (auto and transit)
model is incorporated into the 3MODE(VAN)-AGG program by changing the
mode choice model coefficients. This is accomplished by entering user-
specified coefficients in Step 3 of the program. The coefficients of the
two-mode model are:
• IVTT: -.03
• OVTT: -.11
• OPTC: -.005
The coefficients required by 3MODE(VAN)-AGG are IVTT, OVTT/DIST, and OPTC/Y,
whre "DIST" is round trip length and "Y" is household income. In order
to enter the coefficients as shown above, the OVTT and OPTC coefficient
must be adjusted and entered separately for each market segment being
analyzed as follows:
OVTT/DIST = .11 x trip length
OPTC/Y = -.005 x household income.
-------
111-45
The proper coefficients for market segment 1 would be:
OVTT/DIST - -.11 x 11.2 « -1.23
OPTC/Y = -.005 x 16,810 - -84.5
None of the other optional 3MODE(VAN)-AGG program subroutines are used
in the analysis of the bus priority strategies. The only other non-
standard procedure used is representing the overall auto mode share by
the "drive alone" share in the program rather than by a drive-alone and
a carpool share.
A set of user worksheets with entries for one of the market segments
under analysis, reflecting the signal preemption only priority option
follows. The data used to develop the information on Worksheet C-l
(Figure III-ll) is available from the 2MODE-AGG datasheet for the equili-
brium base case. The auto and transit volumes predicted by 2MODE-AGG are
used to calculate the base work trip model shares for "drive alone" and
transit. The run of BUS reflecting the alternative bus priority measures
provides the estimated change in transit in-vehicle travel time shown on
Worksheet C-2 (Figure 111-12). This is assumed to be the only significant
change in level of aservice associated with the bus priority measures; all
other level-of-service changes are zero. Worksheet C-A (Figure 111-13) is
used to organize the data for input to the 3MODE(VAN)-AGG and provides
step-by-step directions for program execution.
3MODE(VAN)-AGG must be run 18 times (once for each representative
worker) with the results being accumulated in the program automatically.
After the bus ridership level for each bus priority alternative has been
determined, the new passenger demand rate per hour is used as input for
-------
III-A6
FIGURE III-ll
WORKSHEET C-l|
BASE DATA
POLICY' Signal
Population
Subgroup
1
Average round
Trip Length (mi)
11.2
Annual
Household
Income ($)
16 , 810
P
!i
1
(2.5)
Population
751
Average
Vanpool Size
10
ffQ^
*3|
?£•!
•<£
1.5
Base Work Trip Modal Shares
%*
§!'
.964
O
*
I
-
s1
.036
f
•o
o
2.
-
O
1
-
-------
WORKSHEET C-2
FIGURE 111-1:2
POLICY: Signal Pre-Emption
CHANGES IN TRANSPORTATION LEVEL OF SERVICE
(all data represent round trips)
s§
1
Drive Alone
-fi
•IP
0
Out-of-Vehicte
Travel Time
AOVTT,), (mia)
0
Oul-of-PocKet
Travel Cost
£OPTCd.(«)
0
Carpool
-31
31!
* !*
r
0
Iff
H * 2.
Jffjf
I3!
0
> s^ O
o$ £
3^4
«°9 j
-~f
«£• &
0
5 If
J5^
p. SL
1?
^ §3
•^" *
0
Transit
^ii
Slf
-HH!
1*
-8.8
^a1?
0$ ?
!':i
0
> 5*9
Q* £
|ia
Poi
a|
s *
0
Vanpool
t> =« 3"
-------
lli-48
WORKSHEET C-4 PROGRAM STEPS
639.39
(3MODE(VAN)-AGG-2(B)/7900110/ESE)
USER INSTRUCTIONS
STEP
_1_
2
r
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
31
3m
3n
v»
jy .
4
4a
4b
4c
Ad
PROCEDURE
Read card(s") -Partitioning is 639.39
Initialize Storage Registers and store —
"default Coefficients.
If default coefficients jaze used, skip
to STEP 4
OPTIONAL - Enter user-supplied coefficients
coefficient of IVTT-DA mode
coefficient of OVTT/DIST-DA mode
coefficient of OPTC/Y-DA mode
coefficient 4-DA mode
coefficient 5-DA mode
coefficient of IVTT-SR mode
coefficient _of JjyTT/pISTj-SR jnode^
coefficient of OPTC/Y-SR mode
coefficient of incentives-SR mode
coefficient 5-SR mode ..._..
coefficient of IVTT-T mode
coefficient of OVTT/DIST-T mode
coefficient of OPTC/Y-T mode
coefficient 4-T mode
IF A MISTAKE IS MADE, BEGIN AGAIN AT STEP__3a
Enter market segment data (from worksheet C>
round trip distance in miles ^ 0
household average annual income in $ i* 0
enter "1" if income is not needed for this_
segment. . . __ _. . .. —
Average carpool occupancy ^ 0 _
Default - 2.5.
Market segment population .
ENTER
6 1DA*-.Q3
62DA-ii2J
6 3DA "-8A5
6 4DA - 0_
6 qDA - 0
8 LSR - Q
9 2SR "-=-03.
6 3SR -^1^3
8 «SR — 8_^5
65SR- 0
61T "-.03
62T -^^
63!. — €^5
94T ••_!?
65T - 0
-1) .....
DIST- 11*2
Y - 15.810
°c.cCpV^I
POP - 751
F
A
^ VB^ V
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
STO
SIQ
SIQ_
STO
— —
R/S
R/S
R/S
R/S
>RESS
wJ
00
01
02
03
04
05
06
07
08
09
10
11
-JL2.
OlJ
»1
- • -
V^k V^.
„
DISPUY
.29
...
•^.•^r^^~w^:^
...
FIGURE 111-13
-------
IIL-49
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
4e
4f
5
5a
5b
---
PROCEDURE
Average vanpool occupancy 4 0
default ~ 10
Vanpool circuity factor ^ 0
default — 1.5
If a mistake is made, press [Cjand begin "
with STEP 4a
If carpool subroutine- is Trot used,-
go to STEP 6 .'.....
OPTIONAL - Carpool Subroutine (data from wol
Enter average occupancies of two classes of
carpools : •- - - - -
1. Average occupancy carpool class 1
2. Average occupancy carpool elass 2
Enter level of service and mode share data
1. AlVTT - carpool class 1
2. AOVTT - carpool class 1.
3. AOPTC - carpool class 1
4- AINCENT - carpool class 1 (0/1)
5. ACJ - carpool class 1**
6. --carpool class 1 "share -as- -a -fraction-
of all carpools *
If a mistake is made,- -p-ress C* ] and begin —
with STEP 5 a
7. AIV.TT- carpool clsss".2
8. AOVTT - carpool class 2
9. AOPTC- - carpool class 2
10. A INCENT -~ carpool class 2 (0/1)
11. Ac - carpool class 2** - —
12. carpool class 2 share as a f ractioh ~
of all carpools*
If a mistake is made, press (cj and begin
with STEP 5 a
* These two values should sum to 1.0
** If default coefficients are used, these
A's must equal zero.
ENTER
OCCyp = IQ
CIRC = 1-5
.
~~~~~~
ksheet C-3)
OCCCP1=
occcr^=
A =
A - '
A =
A =
A = -
SpPl"
- - ...
A =
A = ' "
A =
A =
A -
S =
CP2
R/S
R/S
.... .
^
imj
2nd
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
=>RES£
KW
L*W
Cf
•
2nd
T,
.*•
E'
—
-
DISPLAY
--- • - -
38
.. ..
- - .
- •
_.~
. — ........
—
_..
._. .
-------
111-50
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
5c
6
6a
6b
6c
6d
6e
6f
6g
7
PROCEDURE
carpool .subroutine results.:_
1. New average occupancy for all carpools
2. New Carpool change in level of -,._
service (for all carpools)
Enter A LOS ' for carpool class 1
Enter A LOS for carpool class 2
To repeat for another change in level of
service, press JDJ and repeat STEP 5c2.
Step 5 results can also be 'recorded on Wo'rV
To continue with^STEP^fi ^^^.^^^.^^^^
Store modal shares for this market segment
(from Worksheet C-l)
Base drive-alone share, as .a .fraction **.
(i.e., .645, not 64.5)
Base carpool-share. as a fraction **
Base transit share as a fraction **
Base vanpool share as a fraction**
Base "other" share as a fraction **
If a mistake is made, press |GO TO 1/X|
and begin with STEP 6a •
Base VMT
If vanpool subroutine -is not used, go to- •-
STEP 8
OPTIONAL Vanpool Subroutine -
If VP = 0 in 6d and the revised. vanpool. .
share 4 0, enter a "base" vanpool share on
which to pivot. Defaults —
.14 for firms with £ 2000 employees
.06 for firms with >^ 2000 employees -
If a mistake is made, press |B'| and repeat
STEP 7
* printed but not displayed
:* these five fractions must sum to 1
ENTER
.
ACPl"
sheet C -5.
SDA- ~964
S = °
ST = .036
SQ ; .
„._
F
R/S
R/S
R/S
2nd
R/S
R/S
R/S
R/S
R/S
*/fi
R/S
R/S
>RESS
-
__-. .
D'
. ..
DISPLAY
*OC
-------
111-51
WORKSHEET C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
.8
8a
8b
8c
8d
8e
8f
8g
8h
8i
8j
8k
81
8m
8n
80
9
9a
9b
9c
9d
9e
PROCEDURE
Enter _changes in level of service from Works
If there are NO changes in level of service,
Press [jj] and go to STEP 10; otherwise, ..
Drive Alone changes:
AIVTTDA ' " "
Ltf*.
A OVTT
AoPTC A -
A 4DA **
A'50A **
If there are no additional changes in LOS,
press [H and go to STEP 10
Carpool changes from Step 5 or Worksheet C
if -optional STEP 5 was used:
AIVTTCP
AOVTTrp
vi
AOPTCcp
iINCENTcp *
A5cp **
If there are no additional changes in LOS,
press [¥) and go to STEP 10
Transit changes:
^ IVTT
•— J- » j. J.fn
AOVTT^ -
A OPTCT
A4T **
A5T ** "
If a mistake is made, press [GO TO] |l/x|
and begin at STEP 6a
If the vanpool share in 6d «= 0 and STEP -7 —
was not used, GO TO STEP 10
OPTIONAL: Enter vanpool .change .in LOsTfVpji
AIVTT^
AOVTTyp
A OPTCyp
A INCENTvp ***
A5**.
* This must equal 0 or 1 unless the carpoo]
must be between 0 and 1 inclusive.
** If default coefficients are used, these L
** 9d must equal zero if STEP 7 was used.
ENTER
heet C-2)
A = °
A = 0
A =__0
A = _Q
-5
- A -.r".o _
A -. 0
A = °
A = 0
A » Q.
A = -8.8
^ =
A = ___ —
"A -~ o
" 'A ". "6
"woVkVoeTt^c"
A = ^
A -
"A - '
'A = -
A «
subroutine :
's must equa!
1
J>
B
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S
B
R/S
R/S
R/S
R/S
R/S
" "
-£>*
R/S
R/S
R/S
R/S
R/S
s us
0.
=RESS
„,.
sd, i
„ ,
_
...
-
n wh
DISPLAY
._ .,. ^ „.
• - - -- ™
. -
.
'..
.__.__•- •"*-- - --~
—
. . _
— ..
...
- -
.ch case it
-------
111-52
WORKSHEET
C-4
(continued)
(3MODE(VAN)-AGG-2(B)/790110/ESE)
USER INSTRUCTIONS
STEP
10
10 a
lOb
10 c
10 d
lOe
10 f
10g
lOh
101
10j
10k
10]
10m
lOn
ii
PROCEDURE
Market segment results
New DA share *
New DA volume
New DA VMT
New CP share * SEE~SUMMARY~ "
New CP volume' ' ". ^^ '
New CP. VMT
New T share *
New T 'volume • - -• . —
New VP share * -
New -VP' volume ......
New VP VMT "~ ~ " '
New Other share * " "
New VMT for this market segment
New- Autos per wprker * - --
These results can also be recorded on Works
To analyze another market segment and have
the results aggregated with previously
analyzed market segments, press [Qand GO.
TO STEP 4
To print** aggregated results of all pre-
vious market segments, press [JEJ
"" " "
...,.._
. . -- .. " —
These results can also be recorded on Works
To analyze a new policy (no aggregation wit
previous market segments), press [A]and GO
TO STEP 2. (Since this sets memories to zero
data from previously tested policies must b
copied before A is pressed, and user-
supplied coefficients, if any, must be re-
entered in STEP 3.
* Printed but not displayed
**See Comments (Section 4) for retrieving
data without a printer.
ENTER
-
.
eet C-6
....
leet ;C-6
P
B
R/S
R/S
R/S
R/S
R/S
R/S
R/S
R/S"
C
E
A
RESS
•
_
' "~
-
DISPLAY
S'pi •*-
VOL'.p.A-
VMT 'PA-
S'CP' =
VOL'CP=
VMT'CP=
oi _• '
b rp
VOL'T.= -
S'
* VP ~
voL'yp-j;
VMT'VP=
„....
VMT-j.nT=. '
A/WT =
£VOL'M=
ZVMT'm=
ZVOL'Cp=
EVMT'cp=
EVDL'T =
EVOL'vp"
EVMT'vp=
ZVMT-roT3
IVMT^OT= •
AlVMTj'ox=
%AL'VMTTOT«
-------
111-53
another run of BUS. Each program is run iteratively, using the output
of the previous program as part of the program input until equilibrium is
reached between bus ridership and bus level of service. Figure 111-14
illustrates this iterative procedure.
Tables III-7 and III-8 show the average bus travel time and ridership
predictions developed for the base case and the two levels of bus priority
treatment. The results were compiled from information recorded on
Worksheet C-6 for each of the 18 representative worktrips or market
segments and the simulation records from Program BUS. Market segments 1
through 6 represented origin Zone I; 7 through 12, Zone II; and 13 through
18, Zone III. These results are used as input to the Step 8 analysis of
bus energy consumption and Step 9 analysis of emissions. 3MODE(VAN)-AGG
also calculates total work trip auto VMT for the base and alternative
cases. However, in converting the program to a two-mode model, all
auto person trips were classified as drive-alone for predictive purposes.
This leads to an overestimate of auto VMT because each auto person trip is
translated into a vehicle trip for VMT calculation purposes, while in
actuality, the auto occupancy for worktrips in the corridor is somewhat
greater than 1.0. Assuming the national average of 1.2, the auto VMT calcu-
lated by 3MODE(VAN)-AGG must be divided by 1.2 for the base and each alter-
native case.
Program ENERGY predicts the per-hour fuel consumption rate of a bus
route. It also calculates the number of buses required to cover the
route and the actual layover time which would occur with that number of
Alternatively, auto occupancy estimation techniques contained in NCHRP
Report 187 could be used to determine the approximate corridor auto
occupancy.
-------
111-54
New TSM Bus
Operating Policies
Base Case Ridership
Predict Performance
BUS
Transit Travel Time
Transit Travel Time Consistent
with the Base Case or most
recently used Transit Travel
Time input to Forecast Bus
Ridership?
Predict Ridership
2MODE-AGG
Base Case or
"Latest"Mode Shares
Revised Bus
Ridership
Revised
Auto Users
Equilibrium
Condition Reached
FIGURE III- 14
Equilibrium Data Flows for Testing Effects
of Transportation Measures
-------
111-55
TABLE III-7
Equilibrium Transit Travel Times under Alternative Policies
One-way travel time from zone centroid to
in-bound terminus, minutes
(percent change from base case)
Origin Zone Base Case Signal Control S1Jigt\al Cont^01 +
6 6 Preferential Lane
I
II
III
14.1
(-)
20.3
(-)
26.4
(-)
9.7
(-31.2%)
14.2
(-30.0%)
19.1
(-27.7%)
8.8
(-37.6%)
12.8
(-36.9%)
17.2
(-34.8%)
TABLE III-8
Equilibrim Transit Ridership under Alternative Policies
One-way ridership (percent change from base case)
_,,_ no oj i o ^ i Signal Control +
Origin Zone Base Case Signal Control „ ,- . , T
0 e Preferential Lane
I 1358 1420 1431
(-) (+4.6%) (+5.4%)
II 1048 1109 1125
(-) (+5.8%) (+7.3%)
III 361 422 438
(-) (+16.9%) (+21.3%)
TOTAL 2767 2951 2994
(-) (+6.6%) (+8.2%)
-------
111-56
buses. Required inputs for the program include the route length (in
miles), the headway, the average grade (important in hilly areas), the
round-trip travel time, and the minimum layover time (usually set by
agreement between the bus operator and the drivers). Program BUS provides
the round-trip travel time estimate and the minimum layover is assumed to
be ten minutes, consistent with the current policies of the bus operator.
All other values required by ENERGY (with the exception of grade, which is
assumed to be zero) are available from base data. The procedure is illus-
trated by the worksheets for the calculation of base case bus route energy
consumption and fleet requirements shown in Figure 111-15. Table III-9
shows the calculator output for each of the three runs. From these data,
estimates of the per-hour operating cost of the route for the three scen-
arios (base, signal pre-emption, signal pre-emption with reserved lane)
can be developed for evaluation purposes.
In Step 9, the impacts of the proposed bus priority strategies on the
emissions of both buses and autos in the corridor are estimated. Program
BUSPOL calculates the per-hour total emissions in grams of the bus fleet
required to serve the route. The required inputs to the program are system
(route) bus VMT per hour (10.6 miles round-trip x 20 runs per hour = 212
miles), and the average effective speed (available from the output of ENERGY)
The program reports total hourly and per-mile emissions of CO, HC, NOx,
particulates, and SOx in grams. The program is run three times with varying
average effective speeds corresponding to the base case and the two alter-
native bus priority strategies. A sample run for the base case is illus-
trated by the worksheets in Figure 111-16 and the calculator output is
shown in Table 111-10.
-------
111-57
USER WORKSHEET - Energy Consumption on Bus Route
{ENERGY-1(A)]
Step
1.
2.
3.
4.
5.
6.
7.
Procedure
Read card 1
Enter grade (0-5%)
Read Card 2
Set print option if using
printer
Enter route characteristics
(1) Route length, round
trip (mi)
(2) Route travel time, round
trip (min)(excluding
layover
(3) Headway (min)
(4) Minimum schedule slack
time (min)
Compute fuel consumption
Read:
(1) effective operating
speed (mph)
(2) fuel consumption per mile
(3) VMT/hr
(4) fuel consumption per hour
Compute schedule characteristics
(1) # vehicles required
(2) Round trip travel time,
including layover (min)
Enter
Gr. = Q
d = ™.6
en c
TT = 52'S
K = 3
ST = 10
Press
EJ
|2nd|Stflg
m
|STO|TO"
isTOilin
JSTOJJ12J
|STO|[l3]
H
IKORf
, *
| RUN
(RUN)*
PI
*
JRUM!
Print/Display
I
i
= 12.045
= 0.227
» 212. 0
= 4S.182
m 21
= 63.0
FIGURE 111-15
-------
111-58
Step
8.
Procedure
(3) Layover time, round trip
(min)
To change any of the variables
except grade, enter the new
value(s) using the necessary
part(s) of Step 5; repeat
Steps 6 and 7.
To change the grade, repeat
the procedure from Step 1 .
Enter
Press
*
RUN|
Print/Display
= 10.2
FT
Not to be entered if print option is set.
-------
111-59
TABLE III-9
Impacts Under Alternative Policies
(Example Calculator Output)
Impact
Effective operating
speed (MPH
Fuel Consumption per
mile (gal/mi)
VMT/hr
Fuel consumption/hr
(Gal/hr)
# vehicles required
Round trip route time
(min)
Round trip layover
Base Case
12.045
0.227
212.000
48.182
21.000
63.000
10.200
Policy 1
16.649
0.184
212.000
39.098
17.000
51.000
12.800
Policy 2
18.488
0.175
212.000
37.070
15 . 000
45.000
10.600
-------
111-60
FIGURE III- 16
USER WORKSHEET - Bus Pollutant Emissions (BUSPOL-1)
Step
1
2'
3
4
Procedure
Read card (card 1 for TI 52)
Set printing option if using
printer
Load memory registers
List memory contents (optional):
0) Emission factor
(parti culates)
(2) Emission factor (SOY)
A
(3) NOX factors:
(a) ROAD
(b) IDLE
(c\ IIRRAN
(d) emission factor
(4) HC factors:
(a) ROAD
(b) IDLE
(c) URBAN
(d) emission factor
(5) CO factors
(a) ROAD
(b) IDLE
(c) URBAN
(d) emission factor
Enter
(Note.: Tk
&te.p may be.
omitted;
9"td ' ^
Jabt&> 1 and
2)
Ppess
J2ndl St fig
CD
|2nd |B'
• 1 1 i
2nd Q3
RUN *
[RUNl*
[RU¥j*
IRUN 1*
(RUNl*
(RUN!*
[RUN!'*
*
IRUN
[RUN]*
IRUN 1*
[RUN 1*
|RUN|*
[RU¥|*
Print/Display
=
=
-
=
=
_
-
=
-
=.
—
B
-------
111-61
FIGURE 111-16 (Cont.)
Step
5
6
7
8
9
10
11
12
Procedure
Read card 2
Enter system Vehicle Miles
Traveled
Enter effective speed (mph)
Compute CO emission (grams)
- per vehicle-mile
- total
Compute HC emission (grams)
- per vehicle-mile
- total
Compute NOX emission (grams)
- per vehicle-mile
- total
Compute emission of (grams)
- particulates per vehicle-mile
- total particulates
- S0¥ per vehicle-mile
A
- total SOX
Perform Steps 8-11 automati-
cally (TI 59 with printer only)
•
Enter
VMT = 2? 2
s = 12.04
(Wote: Step
12 peAj$oim4
S*ep4 8-11
CUUJtQWOJU(£.oJL-
ty]
Ppess
DO
m
[RUN]*
ron*
| RUN
2nd] J E '
I RUN
RUN
OS*
|2nd 1A'|
Print/Display
VMT
s
(printed only)
= .22. OB
= 46B1.75
-
4.46
= 945.34
23.25
= 4925.65
= 1.3
= 275.6
= 2.8
= 593.6
same as for
Steps 8-11
-------
111-62
FIGURE III- 16(Cont.)
Step
13.
' Procedure
To vary VMT or speed, go to
Steps 6 and 7 and proceed.
Enter
Press
" "" ' i
Print/Display
., . *
Notes: Not to be pressed if printing option is set (Step 2).
R/S on TI 59.
Steps 8-11 may be performed in any order.
-------
111-63
TABLE 111-10
Bus Emissions of Major Pollutants under Alternative Policies
(Calculator Printout)
System VMT
Effective Speed
CO emission (gms)
- per vehicle-mile
- total
HC emissions (gms)
- per vehicle-mile
- total
NOx emissions (gms)
- per vehicle-mile
- total
Particulates emissions
- per vehicle-mile
- total
SOx emissions
- per vehicle mile
- total
Base Case
212.
12.04
22.08
4681.75
4.46
945.34
23.25
4928.65
1.30
275.60
3.80
593.60
Policy 1
(same)
16.65
21.43
4542.82
4.08
863.94
21.79
4618.71
1.30
275.60
2.80
593.60
Policy 2
(same)
18.49
20.65
4376.81
3.92
831.60
21.79
4619.10
1.30
275.60
2.80
593.60
-------
111-64
The auto emissions worksheets (4.1) were used to estimate the impact
of the predicted increase in bus patronage on auto air pollutant emissions
in the corridor. The emissions model incorporated in the worksheets
requires the number of auto work trip and total auto VMT for the base case
and each alternative under study. Average travel speeds, the ambient
temperature, and the analysis year determine the auto emissions rates for
each scenario. Worksheet VI-A may be used to calculate the change in auto
trip-making based on VMT estimates from 3MODE(VAN)-AGG and average trip
length data. Alternatively, the total auto person trip volume reported by
2MODE-AGG (base case) and 3MODE(VAN)-AGG (alternative bus priority strategies)
may be used directly to determine auto vehicle trips. The base case auto
volume of 6120 (found by summing the market segment auto voluems in
Figure III-5) predicted by 2MODE-AGG must be adjusted to reflect carpool-
ing in the corridor. Using an auto occupancy value of 1.2 yields 5,100
vehicle round-trips per day, or 10,200 one-way trips. VMT is reported by
3MODE(VAN)-AGG for the base case and each alternative. Total base work
VMT in the corridor is 72,030. The revised worktrip auto vehicle volume
and total work trip VMT are estimated to be:
• Signal control - 9,890 one-way work trips; 70,100 VMT per day.
• Signal control and preferential lane - 9,750 one-way work trips;
69,620 VMT per day.
Because example applications of these worksheets also appear in Case
Studies I and II, the completed worksheets have not been included in
this Case Study.
-------
111-65
An analysis year of 1982 is assumed for determining the work trip
vehicle emissions impact. Non-work travel behavior is not expected to
change significantly as a result of the bus priority measures. An
ambient temperature of 50° is used, based on historical weather data. All
subgroups will be analyzed together, using the VMT and trip values developed
above.
Worksheet VI-B is used to determine the percentage of autos
starting cold. Since parking duration data is not available for the
corridor, Emissions Table D.I provides the percentage of cold starts for
1982 work trips.
Worksheets VI-C and VI-D are executed three times, once for each
scenario, including the base case, under analysis. Worksheet VI-E is
completed for each bus priority alternative and is used to compare the
performance of each measure relative to base conditions. The execution
of Worksheets VI-C and VI-D will be described for the base case, and both
executions of VI-E will be illustrated.
All subgroups are analyzed in one pass in both VI-C and VI-D. The
percentage of cold starts for work trips is obtained from VI-B, and the
number of trips from the information above. Start-up emissions factors
for VI-C are obtained from Emissions Tables D.2, D.3, D.4, and D.6,
using the analysis year of 1982, the ambient temperature of 50°, and the
90° cold-start percentage to enter the tables. No interpolation is
necessary to obtain emission factors for this set of circumstances.
Emissions Table D.7 provides auto travel emission factors for Worksheet
VI-D, using average speed and analysis year to determine the appropriate
-------
111-66
emissions rate. Based on several test speed runs on the arterial
serving the study corridor, an average work trip speed of 25 mph is
used. Once again, the emissions factors may be read directly, without
interpolation. The execution of the worksheets for each alternative bus
priority strategy is identical to the base case, with the only changes
being the number of trips in VI-C and auto VMT in VI-D.
Worksheet VI-E is used to calculate the percent change in vehicle
emissions attributable to the proposed bus priority measures, which may
then be used as input to a rollback or similar model to determine the
change in ambient concentrations of pollutants in the atmosphere.
Column 9 of Worksheet VI-E for the two alternatives indicates that signal
control alone will lead to a 3 percent reduction in each pollutant, and
that the addition of a preferential lane leads to an additional 1 percent
decrease in pollutant emissions.
H. Interpretation of Results
Table III-ll summaries the estimated impacts of the proposed bus
priority measures on auto and transit emissions in the study corridor.
With the exception of NOx, the change in pollutant emissions for buses
is insignificant in comparison to the predicted reduction in auto
emissions. In most instances, from an areawide standpoint, transit
emissions may be ignored when determining the air quality impact of tran-
portation measures, due to the much higher proportion of total emissions
Because existing auto volumes in the corridor are not high in relation to
the capacity of the area's streets, the average speed of auto trips is
not expected to change with the provision of an exclusive bus lane. In
other applications, some increase in congestion might be expected which
would have to be reflected in the emissions worksheets as a lower average
speed.
-------
VI-E. SUMMARY OF CHANGES IN EMISSIONS
Revised Alternative
Policy: Bus Priority Signal Control
Forecast Year:
1982
(1)
Population
Subgroup
All
HC
CO
NOx
HC
CO
HOx
HC
CO
NOx
HC
CO
NOX
Base Emissions
(2)
Trip-Related
(VI-C)1
186 r 660
2,029,800
35,700
TOTALS
Source Worksheets are indicated in
parentheses where applicable
(3)
Travel
(VI-D)
115.248
1,750,329
151,263
1IC
Zco
NOx
Sub-
groups
(4)
Total
(Col. 2 + Col. 3)
301.908
3,780,129
186.963
301,908
3,780,129
186,963
Total Base
Emissions
(grams)
Revised Emission
(5)
Trip-Related
(VI-C)
180,987
L, 968, 110
34.615
(6)
Travel
(VI-D)
112,16f
1,703,43(
147. 21C
z
Sub-
grot
a
(7)
Total
(CoLS+Coi. 6)
29% 147
3,671,540
181.825
HC
;co
NOx
(8)
Change in
Total
Emissions
Col. 4-CoL 7)
- 8r761
-108,589
- 5.138
- 8,761
-108,589
- 5,138
Total Change
ips in Emissions
(grans)
(9)
'ercent
Change In
Emissions
(Col. 8/Oo LA)
x 100
- 2.9
- 2.9
- 2.7
-2.9
-2.9
-2.7
Percent
Change,
Total
Emissions
H
M
I
-------
VI-E. SUMMARY OF CHANGES IN EMISSIONS
Revised Alternative
Policy: Rng Prlnr-H-y Signal Control & Pref. Lane
Forecast Year:
1982
(1)
Population
Subgroup
All
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
"Ox
Base Emissions
CD
Trip-Related
(VI-C)1
186,660
2,029,800
35,700
TOTALS
Source Worksheets are indicated In
(3)
Travel
(VI-D)
115,248
1,750,329
151,263
lie
Zco
NOx
Sub-
groups
(*>
Total
(Col. 2 + Col. 3)
301,908
3,780,129
186,963
*\n\ Qnft
3,780,129
186,963 _j
Total Base
Emissions
(grams)
Rev sed Emission
(5)
Trip-Related
(VI-C)
178,425
1,940,250
34,165
(6)
Travel
(VI-D)
111,360
1,691,280
146,160
Z
Sub-
groti
a _
(7)
Total
[CoLS+CoL 6)
289,785
3,631,53C
180,285
HC
;CO
NOx
(8)
Change in
Total
Emissions
Col. 4-CoL 7)
- 12,123
-148,599
- 6,678
- 12,123
-148,599
- 6,678
Total Change
ips In Emissions
(grano)
(9)
Percent
Change in
Emissions
(Col. 8/OolA)
x 100
-4.0
-3.9
-3.6
-4.0
-3.9
-3.6
Percent
Change,
Total
Emissions
I
ON
00
parentheses where applicable
-------
111-69
TABLE III-ll
Estimated Auto and Bus Emissions Impacts
(grams per day)
Pollutant
HC
CO
NOx
Change in Daily Emissions
Signal Control
Auto
-8,761
-108,589
-5,138
Bus
-405
-695
-1,550
Signal Control and
Preferential Lane
Auto
-12,123
-148,599
-6,678
Bus
-569
-1,525
-1,550
-------
111-70
related to auto travel. However, as the air quality impact estimates
for this example illustrate, if nitrogen oxide levels are of major con-
cern, or if changes in ozone concentration are to be estimated based on
the change in total HC and NOx emissions, the impact of the proposed
measures on bus NOx emissions should be considered.
Although modest decreases in total vehicle emissions are predicted
by the analysis, the estimated impact is large enough that beneficial
air quality impacts may be attributed to either of the bus priority
strategies with a high level of confidence. Further analysis of the two
proposed alternatives might focus on.a more in-depth assessment of the
transit operations and traffic flow implications of the measures. A
more detailed analysis of the bus route using either BUS, with a greater
number of stations and intersections; or a computerized analysis tool
such as TRANSYT-6C would be justified in order to assess fleet require-
ments and traffic flow changes. The ultimate choice between the two
alternative bus priority strategies will depend on the more in-depth
assessment of the performance of each which could be provided by a more
detailed operational analysis.
-------
CASE STUDY IV
AREAWIDE ASSESSMENT OF TRANSPORTATION MEASURES
-------
CASE STUDY IV: AREAWIDE ASSESSMENT OF TRANSPORTATION MEASURES
A. Problem Statement
An urban area currently not attaining minimum health standards for
CO and ozone concentrations is beginning the process of analyzing trans-
portation/air quality measures for inclusion in its 1982 SIP revision.
Through its involvement in planning studies for a number of major transit
development alternatives, the area's MPO has developed an extensive and
up-to-date UTPS-based model system, including highway and transit networks
and zonal population, socioeconomic and employment projections at five-
year intervals to the year 2000. However, the MPO has found the travel
demand models incorporated in their model system expensive to use and
insensitive to the types of short-range transportation measures commonly
included in air quality improvement strategies. The MPO wishes to
develop a capability to systematically analyze the air quality, fuel
consumption and modal choice impact of a range of measures it is con-
sidering for inclusion in its SIP revision.
B. Proposed Transportation Measures
Because of the area's non-attainment status, the MPO is required to
consider the entire range of reasonably available control measures. Illus-
trative of the types of action the MPO hopes to evaluate are the following
four measures:
-------
IV-2
• Increase service frequencies on all CBD-bound transit routes
during the peak hour;
• Initiate carpool matching and promotional programs at all major
employers (i.e., firms with over 50 employees);
• Restrict the number of employer-provided parking spaces used by
single-occupant autos;
• Combine the restriction of single-occupant auto parking spaces
with the promotion of carpooling.
Through such measures, the MPO hopes to reduce the use of the single-
occupant auto, while increasing the utilization of transit and carpooling.
Energy conservation as well as environmental protection benefits are
expected to accrue if a successful program of transportation control
measures can be devised.
C. Selection of Analysis Method
Because the urban area has an up-to-date UTPS-based model system,
SRGP (2.3.2) can be implemented relatively inexpensively and can provide a
flexible and inexpensive tool for analyzing a large number of alternative
transportation measures at the urban area level. SRGP is more desirable
in this instance than the available manual techniques because of its ability
to quickly analyze a large number of strategies. At the same time, the
computational power of the computer-based method will allow a high degree
of accuracy and the ability to isolate the effects of alternative policies
on special groups within the urban area. Therefore, SRGP has been selected
for application to the urban area's SIP planning process.
The section numbers following specified analysis methods refer to the
location of their description in Volume I.
-------
IV-3
D. Input Data Development
A work program for setting up SRGP for use in a given urban area was
described in Volume I. The basic steps in the work plan involve:
• developing base year data files;
• adjusting model coefficients (model calibration);
• updating base year data to analysis year conditions.
SRGP requires three types of base data: a sample of households, transpor-
tation levels of service for peak and off-peak trips, and land use
characteristics.
The primary source of household information available for the case
study urban area is an areawide home interview survey (HIS) conducted in
1971. Level-of-service data were available in the form of standard UTPS
networks representing peak and off-peak conditions in 1975. Zonal popu-
lation, employment, and land use data were available from a study conducted
in 1970 which developed forecasts of these data in five-year increments to
the year 2000. The travel behavior models within SRGP are calibrated
on base year data from 1971 assuming that no major land use, population,
and employment pattern changes had occurred since the 1970 study.
The sequence of data processing steps required to develop the
household, level-of-service, and zonal input files is shown in Figure IV-1.
The first step was to reduce the household file from 15,000 (the sample
size of the HIS) to 2,506 households in order to minimize the data
processing costs of subsequent steps, while maintaining an adequate sample
size for analysis purposes. The size and composition of the household
sample used in SRGP depends on the desired accuracy of the impact
predictions generated, the degree to which the impacts on specific groups
-------
Hou.«ho!ti
OK-P.ik LOS
Zon«l
<
FIGURE IV-1
SRGP Data Preparation Steps
-------
IV-5
are of interest, and the budget constraints placed on data processing. The
trip file was also shortened to include only work trips associated with
the reduced household file.
Both the household and trip files were sorted by ascending home zone
number and, within a given home zone, by ascending household identifica-
tion number. The trip file was further sorted by ascending person number
within each household.
Peak period transit level-of-service data in the form of skimmed
trees was obtained from basic transit network data using standard
UTPS software; skimmed trees for auto were already available. At this
point, a FORTRAN program was developed to merge these data into a single
file in the appropriate format for input to SRGP.
Developing the off-peak level-of-service file was straightforward.
Skims of 1975 off-peak networks for both auto and transit were already
available. For transit, UMATRIX was used to combine IVTT and OVTT to
form the required total travel time variable. The resulting transit
skims were then combined with the auto skims and transit fare matrix
into a single file using a FORTRAN program.
The zonal data file development was also straightforward. A simple
FORTRAN program was used to merge the 1970 zonal population, employment,
and land-use data with the required zonal transportation level-of-service
variables.
Once the base-year data were organzied into the required input files, the
SRGP model system was calibrated to the analysis area by adjusting the alter-
native specific constant terms of each component model iteratively until
the regional travel behavior predicted by the model system matched that
-------
IV-6
TABLE IV-1
Estimated versus Adjusted Values of Alternative
Specific Constant Terms - Case Study IV
Model/Constant Term
Estimated
Value
Adjusted
Value
Worker Household Auto Ownership
1 Auto Constant
2 Auto Constant
Non-Worker Household Auto Ownership
1 Auto Constant
2 Auto Constant
Work Trip Mode Choice
Drive Alone Constant
Shared Ride Constant
Shopping Trip Destination/Mode Choice
Auto Constant
Soc/Rec Trip Destination/Mode Chocie
Auto Constant
Shopping Trip Generation
Shop Expansion Factor
Soc/Rec Trip Generation
Soc/Rec Expansion Factor
4.989
5.689
-.8695
-8.357
-4.697
-3.658
-1.080
1.844
1.0
1.0
6.295
7.615
-.9092
-8.6119
-1.867
-3.657
1.0906
4.714
.9496
1.6106
-------
IV-7
observed for the base year. For this purpose, observed travel data were
taken from trip summaries and other tabulations available from the home
interview survey. Both the original and adjusted values for the constant
terms are shown in Table IV-1 for all of the SRGP models.
E. Description of Model Application and Impact Assessment
Data Updating
To analyze the impacts of implementing selected transportation poli-
cies in future analysis years, it was necessary to use input data that
reflected conditions in these years. While the available level-of-service
data already satisfied this requirement, the household data file and
certain elements of the land use data file required updating to reflect
future conditions.
The household data were updated using a procedure in which household
expansion factors are modified to represent expected changes in the dis-
tribution of selected population characterisitcs within the urban area.
Selection of the actual household characteristics used in updating was
constrained by the availability of projections for these characteristics
in future years. For the example urban area, projections of employment
and population were available in five-year increments for each of 34
geographic divisions within the metropolitan area. Areawide household
income projections were also available in five-year increments for the
region. Employment figures, though, were only available from 1970 data
The details of this procedure are given in Cambridge Systematics,
Urban Transportation Energy Conservation, Vol. II; Analytical Procedures
for Estimating Changes in Travel Demand and Fuel Consumption, prepared
for U.S. Department of Energy, September, 1978.
-------
IV-8
and projections for the year 2000. The values for these variables
in other years were obtained by linear interpolation between the 1970
and 2000 levels.
As an illustration of the household updating process, consider an
upper income household living in a suburban area. In the year for which
the household file is representative (the year of the HIS in this case),
households with these characteristics represented 20 percent of the urban
area's population. In the analysis year, however, such households
are projected to represent 40 percent of the urban area population. To
update the household file to the analysis year, the household expansion
factor for all upper income households in suburban areas would be increased
by a factor given by:
Analysis Year Fraction
Base Year Fraction
.40
.20
- 2.0
The expansion factor for other households would similarly be reduced to
reflect their lower representation in the total population.
Zonal data must also be updated to reflect population, employment,
and land use changes. The five-year projections of zonal characteristics
were used to update key zonal variables for the example urban area,
using multiplicative factors similar to those used for the household file.
Household data files and zonal data files were developed for each of the
analysis years of interest using these updating techniques. Level-of-
service changes in the analysis years reflecting the policy measures to
-------
IV-9
be evaluated are communicated to SRGP by means of various input parameter
options. The following sections illustrate how some of these options
are executed for analysis purposes.
F. Representing Transportation Measures in the Model System
Once the necessary household, zonal, and level-of-service data
(reflecting changes in the quality of service caused by general population
growth and other trends) has been assembled for the analysis year and
used to update the SRGP input files, policy options may be evaluated for
the analysis year. Example applications indicating the statements
required to analyze the following policy measures are illustrated
below:
• Increase frequency of transit service to the CBD during peak
periods;
• Initiate carpool matching and promotional programs at all
employers with 50 or more employees;
• Restrict the number of employer-provided parking spaces used by
single-occupant vehicles;
• Combine the parking restrictions above with employer-based
carpool and vanpool promotion and matching.
1. Increase Frequency of Transit Service to CBD - With this measure,
headways on all transit routes to the CBD are to be reduced by 50 percent
during the peak period. Because this measure can be expressed directly
in terms of changes in levels of service, its representation is fairly
straightforward. The first step in translating this measure into model
system parameters is to identify those individuals who would be affected
by such a measure. Assuming that measures applied only during peak
periods will influence work travel only, this measure would affect all
-------
IV-10
workers with employment locations in the CBD. This measure then must be
transformed into changes in the appropriate model variables. In this case,
doubling peak period transit frequency would be represented by a 50 percent
decrease in the transit headway variable for the work trip mode choice
model.
For this measure, only four statements are required to represent the
level-of-service changes in SRGP. First, it is necessary to establish
a correspondence between the traffic zone system used in developing the
basic input data files for SRGP, and the geographic areas to which the
proposed measure applies. This is accomplished by identifying two
districts (the CBD and the rest of the urban area) and then defining the
districts with the appropriate traffic zone numbers using the following
three s ta t ement s:
&PARAM DIST=2 &END
&EQUIV DIST=1, Z= &END
&EQUIV DIST=2, Z= SEND
With these two areas defined, the only other statement necessary is
that which actually indicates the variable to modify, how to modify it,
and who is affected by this change. For this measure, the statement
would be:
&WORK 0=1, 2, D=l, FAC(7)=.5 &END
The statement specifies that this "update" is applied to work trips (&WORK)
from origin districts 1 and 2 (0=1, 2), which encompases the entire urban
area, to destination district 1 (D=l), which is the CBD. The update
itself (FAC(7)=0.5) specifies that the seventh field in the work trip
data segment (e.g., - transit headway) is to be multipled by 0.5.
-------
IV-11
When SRGP is run with these options set, the model system processes
each household as it would in the base case with revisions reflecting
the data update to the analysis year, except that whenever a work trip
destined to the CBD is encountered, transit headway is halved, and new mode
choice probabilities are calculated using this revised headway.
Once processing of the household data file is completed, various
statistics on travel-related impacts accumulated throughout the run
are printed, both as expanded totals for the entire urban area and
as household averages. In addition, if a file of base case statistics
for the analysis year had been created in an earlier run, this information
could now be read and compared with travel-related impacts predicted for
this measure. The results of this comparison then would be printed in
the form of percentage changes from base case statistics.
Table VI-2 presents selected impacts predicted for increasing transit
frequencies in terms of percentage changes from the base case. Also shown
are expanded areawide totals and average work trip values for the base
case. Note that while this measure affects CBD work trips only, the impacts
are presented for the entire urban area. The percentage changes given in
Table IV-2, then, represent the changes associated with CBD work travel
relative to all work travel. Since in this example work trips to the CBD
account for 10 percent of all work trips, the corresponding percentage
changes for CBD workers alone are about ten times as great as those
given in Table IV-2. On an areawide basis, the impact of this measure on
auto travel is relatively small—VMT decreases by only .4 percent. The
corresponding decreases in emissions range from .3 percent for NOx to
.5 percent for CO.
-------
TABLE IV-2
Changes in Travel Behavior and Travel-Related Impacts
IMPACT (% Change from Base)
TRANSPORTATION
CONTROL
MEASURES
Areawide Base
(Totals)
Areawide Base
(Average Round Trip)
Headways Reduced by
50% (CBD Only)
WORK TRIP MODE SHARES
Drive
Alone
306,000
0.78
-.5
Shared
Ride
76,200
0.19
-.8
Transit
10,800
0.03
3.6
Vanpool
0
-
0
WORK
TRIP
VMT
mi /day
5,030,000
12.99
-.4
WORK TRIP
FUEL CON-
SUMPTION
gal /day
413,000
1.06
-.4
WORK TRIP AUTO
EMISSIONS
kg /day
HC
50,900
0.12
-.4
CO
727,000
1.17
-.5
NO
X
19,600
0.03
-.3
-------
IV-13
2. Employer-based Carpool Matching and Promotion - With this measure,
employers in the urban area are assumed to initiate promotional campaigns
encouraging carpooling and to provide carpool matching assistance for
their employees. Since carpooling incentives such as these are feasible
only for organizations with a relatively large number of employees, the
most logical criterion for determining the availability of such incentives
to an individual worker is employer size. However, this information was
not available for individual workers in the urban area under study. In-
stead, employer size distribution data on a zonal or district level was
used to determine for each traffic zone the fraction of workers in a
given zone employed by companies of a specified size or larger.
To forecast the impacts of carpool incentives, a random number
(between 0 and 1) is generated for each work trip to determine whether or
not that worker will benefit from the particular carpool incentives
being modelled. If the number generated is less than or equal to the
fraction of workers employed by large companies in the worker's work zone,
the carpool incentives are available to that worker; if the number is
greater, the worker does not benefit from the carpool incentives.
Because employer-based incentive programs such as these cannot be
easily reduced to time and cost terms, SRGP incorporates a specific
variable to represent the existence of an employer-based carpool promotion
program. For workers exposed to the promotion campaign, the variable
In this case, incentive programs at employers with at least 50 employees
were compared with similar programs only at firms employing 250 or more.
-------
IV-14
takes on the value "1," for others, "0." This variable is then incor-
porated into the utility function for the shared ride mode for work trips.
Since this measure is essentially areawide in scope, no district
definitions or zonal equivalences are required; its representation is
communicated to SRGP by means of a single control card:
&POOLS ZCPOOL=20, PMATCH=.287 &END
The keyword ZCPOOL=20 indicates that this run will include employer-based
incentives, and that the required employer size data are located in field
20 on the zonal data file. PMATCH=.287 specifies that these incentives
will include matching and promotion, which are represented by adding .287,
the estimated coefficient value for the carpool incentive variable dis-
cussed earlier, to the shared-ride utility function for those workers
employed by large employers. Note that in addition to matching and pro-
motional incentives, it is also possible to represent any employer-based
travel time or cost-related measures such as preferential carpool parking,
differential parking charges, etc.
The results of this measure, expressed in percentage changes from
the base case, are shown in Table IV-3. As shown, the implementation
of carpool matching assistance and promotional programs by organizations
with 50 or more employees results in a decrease in areawide work trip
VMT of 1.4 percent and decreases in auto emissions of 1.3 percent for
hydrocarbons, 1.1 percent for carbon monoxide, and 1.5 percent for nitrogen
oxides. When such carpool incentives are offered only by firms with 250 or
more employees, the resulting decrease in VMT is .8 percent, with decreases
-------
TABLE IV-3
Percent Changes in Travel Behavior and Travel-Related
Impacts of Ridesharing Promotion
IMPACT (Percent Change from Base)
TRANSPORTATION
CONTROL MEASURE
Carpool Program
(50 Employees)
Carpool Program
(250 Employees)
WORK TRIP MODE SHARES
Drive
Alone
-2.9
-1.5
Shared
Ride
+13.0
+ 6.8
Transit
-10.1
- 6.6
Vanpool
-
-
WORK
TRIP
VMT
(mi/ day)
-1.4
-0.8
WORK TRIP
FUEL CON-
SUMPTION
(gal/day)
-1.2
-0.6
WORK TRIP AUTO
EMISSIONS
(kg/day)
HC
-1.3
-0.7
CO
-1.1
-0.6
NOx
-1.5
-0.8
t—»
Ul
-------
IV-16
in auto emissions ranging from .6 to .8 percent. The difference between
these results (expressed on an areawide basis) is due to the difference in
the number of employees affected. When 50 employees is used as the lower
bound for employer size, approximately 60 percent of the work force is
reached by the measure; when a size of 250 employees is used, only 30
percent of the work force is reached.
3. Restrictions on Employer-Provided Parking - The measure considered
here is also employer based, but is defined as a disincentive to single-
occupant auto use rather than an incentive for carpooling. This measure
is a parking supply restriction which limits the number of parking
spaces that an employer can provide to single-occupant vehicles by setting
a maximum value based on the number of employees. This measure is assumed
to apply only to organizations, either public or private, employing at
least 250 people since the practical problems associated with implementing
and enforcing such a measure for a large number of small firms, many of
which may use facilities jointly with other firms (if indeed they provide
parking at all), would be extremely difficult. In addition, firms
located in the CBD are assumed to be unaffected since much of the parking
there is in commercial facilities rather than employer-provided. It also
is assumed that the necessary restrictions could be applied to prohibit
the use of alternative on-street parking nearby. A contingency measure
such as this normally would not be included in a SIP. Its presentation
here, then, is for illustrative purposes.
-------
IV-17
In examining parking supply restrictions, the most appropriate analysis
approach depends on the nature and severity of the restriction. In those
cases where supply restrictions affect only excess capacity (e.g., when
the number of parking spaces available after the restriction is still
greater than or equal to the number of spaces utilized before the restric-
tion) , the immediate impact would be increased search and walk time—for
those individuals forced to use more inconveniently located spaces—and
possibly increased parking costs if those spaces affected by the restric-
tion are free or relatively inexpensive. This type of restriction, then,
would be represented straightforwardly in terms of changes in travel time
and cost. The magnitudes of these changes would depend on the character-
istics of those spaces affected by the restriction.
On the other hand, if restrictions are to be imposed which reduce
the total supply of parking in an area below the level that is currently
used, an iterative forecasting procedure is required to calculate supply/
demand equilibrium. In such a procedure SRGP would proceed through the
household sample once, and expand the results to estimate the number of
auto trips to the area affected by the supply restriction. From this, there-
quired number of parking spaces could be calculated and compared with the
number of spaces that would be available with the parking restriction
in effect to determine the extent to which auto travel must be reduced.
Then, a "shadow price" or penalty price (in terms of excess travel time)
for utilizing a constrained resource would be estimated and applied in
th*> appropriate utility functions of the demand models, and SRGP would
be rerun to produce revised travel patterns. This sequence would be
repeated, altering the shadow price until the demand for parking falls
within the limits of the total available parking supply.
-------
TABLE IV-4
Changes in Travel Behavior and Travel-Related Impacts (A)
IMPACT (% Change from Base)
TRANSPORTATION
CONTROL
MEASURES
Areawide Base
(Average/Round Trip)
Restrictions on Employer-
Provided Parking:
-Ratio of drive-alone
spaces /employees ;
0,3 at large employers
-Ratio of drive-alone
spaces/employees;
0V4 at large employers
WORK TRIP MODE SHARES
Drive
Alone
0-78
-7.1
-11.0
Shared
Ride
0.19
26.6
41.1
Transit
0.03
9.2
13.8
Vanpool
-
—
—
WORK
TRIP
VMT
(tai/day)
12.99
-4.1
-6.3
WORK TRIP
FUEL CON-
SUMPTION
(gal/day)
1.06
-3.6
-5.6
WORK TRIP AUTO
EMISSIONS
( kg/day )
HC
0.12
-3.8
-5.8
CO
1.17
-3.7
-5.7
NOx
0.03
-4.5
-6.9
M
I—'
CO
-------
IV-19
The measure considered here was defined as a maximum value for the
ratio of the number of parking spaces provided by an employer for use by
single-occupant vehicles, divided by the number of employees. Two values
for this ratio were examined: 0.4 and 0.5. Since the existing value was
0.76, both levels result in reductions in parking supply below the level
currently utilized, and therefore the second analysis approach described
above was used in each case. The shadow price necessary to achieve the
0.5 value was an increase in walk time of 35 minutes. To achieve the 0.4
value, an increase of 50 minutes was required. The values of these
shadow prices reflect the severity of restriction represented by the
imposition of parking space quotas.
The results of these measures are presented in Table IV-4 as percen-
tage changes relative to areawide work travel characterisitcs. Because
these measures are primarily disincentives to driving alone, both the
shared ride and transit mode shares increase. However, because transit is
not available to all workers, whereas the shared ride alternative is, the
percentage increase in shared-ride on an areawide basis (ranging from 26.6
percent to 41.1 percent) is greater than that for transit (9.2 percent to
13.8 percent). These changes in travel behavior translate into reductions
in areawide work trip VMT ranging from 4.1 percent to 6.3 percent, and
reductions in work trip auto emissions ranging from 3.8 percent to 6.9 per-
cent. For evaluation purposes, these reductions must be compared with the
severity of restriction represented by the shadow price values required to
reduce drive-alone travel to levels matching the availability of parking
spaces.
The size of the shadow price gives insight into the reasons behind the
unpopularity of severe parking supply, restrictions.
-------
IV-20
4. Combined Measures; Employer-Sponsored Carpool/Vanpool Programs
with Restrictions on Employer-Provided Spaces - The purpose of examining
this combination of individual measures is to explore the synergistic
effects of combining actions that are essentially disincentives for
driving alone with measures designed as incentives for using modes other
than the single-occupant auto. The employer-based parking restrictions
and carpool programs are identical to the individual measures described
in the previous two examples. Employer-based vanpool programs offer an
additional ride-sharing alternative to the single-occupant auto. As in
the case of employer-based carpooling incentives, vanpooling programs
are most frequently organized by larger employers; therefore, the same
employer size criterion for determining the availability of vanpooling
is used. Further, experience has shown that commuters choosing vanpooling
as their mode to work typically have relatively long trip lengths.
Therefore, another constraint imposed on the availability of vanpooling
as an alternative is trip length (a ten-mile, one-way trip length is used
as a lower bound). Employer-sponsored vanpool programs are represented
in the demand model system by adding a new mode to the set of available
modes for the work trip for those commuters who are employed by large
organizations and have long work trips.
The results of these measures are presented in Table IV-5 for the
incentives and disincentives implemented individually as well as in
combination. As shown, this particular combination of measures is about
More detailed documentation of the vanpool option is presented in
Cambridge Systematics , Urban Transportation Energy Conservation. Vol. Hi
Analytical Procedures for Estimating Changes in Travel Demand and Fuel
Consumption, prepared for US Department of Energy, Sept. 1978.
-------
TABLE IV-5
Changes in Travel Behavior and Travel-Related Impacts (B)
IMPACT (% Change from Base)
TRANSPORTATION
CONTROL
MEASURES
Areawide Base
(Average/Round Trip)
Carpool/Vanpool Program
( 250 employees)
Restrictions on Employer-
Provided Parking:
-Ratio of drive-alone
spaces /employees ;
0.5 at large employers
-Ratio of drive-alone
spaces/employees ;
0.4 at large employers
Carpool/Vanpool Program
plus Restrict Employer-
Provided Spaces to High
Occupancy Vehicles;
Ratio of drive-alone
spaces /Employees :
0.5
0.4
WORK TRIP MODE SHARES
Drive
Alone
9.78
-2.8
-7.1
-11.0
-10.8
-14.6
Shared
Ride
0.19
4.0
26.6
41.1
31.4
44.9
Transit
0.03
-3.6
9.2
13.8
3.8
7.3
Vanpool
-
(.015)*
—
—
(.02)*
(.02)*
WORK
TRIP
VMT
mi/day
12.99
-3.7
-4.1
-6.3
-8.9
-11.3
WORK TRIP
FUEL CON-
SUMPTION
gal /day
1.06
-2.6
-3.6
-5.6
-7.0
-9.0
WORK TRIP AUTO
EMISSIONS
kg/day
HC
0.12
-3.0
-3.8
-5.8
-7.6
-9.7
CO
1.17
-2.9
-3.7
-5.7
-7.4
-9.5
NOx
0.03
-3.5
-4.5
-6.9
-9.0
-11.5
<
* Represents new share rather than percentage change.
-------
IV-2 2
14 percent more effective in terms of VMT reduction than the sum of the
two component measures taken individually. (The corresponding increase
in effectiveness in terms of work trip auto emissions reductions ranges
from 10 percent to 13 percent.) A convenient way to illustrate the
synergistic effects of combined measures is to graph the single and
combined measure impacts as shown in Figure IV-2. The lower curve
represents the effectiveness of the supply restrictions alone; the middle
curve represents the summed effect of the supply restriction and employer-
sponsored carpool and vanpool programs taken individually. The upper
curve represents the effectiveness of these measures implemented in
combination. The shaded area, then, represents the increased effective-
ness attributable to the synergistic effect of the combined implementation
of these measures.
G. Interpretation of Results
Organizing the SRGP output into tables and graphs such as those
presented in the previous section can be a significant aid to analyzing
and interpreting the predicted impacts of proposed measures. Of particular
value is the comparison of the single and combined effects of a number
of measures. In many cases, the synergistic effect of the combined imple-
mentation of carpooling incentives and auto use disincentives may lead to
a cost-effective transportation control strategy, where single actions
cannot be justified based on their benefits and costs. Because none of
the reasonably available control measures alone is likely to lead to
emissions reductions large enough to bring an urban area with serious
-------
IV-23
O
M
H
co
W
w
OT
w
ce:
o
z
CO
01
3
w
CO
01
•H
4J
c
0>
O
c
•H
cn
•H
n
•a
c
CO
0)
a)
o
C
M
C
•§
O
CO
w
a)
e
a
O)
14-1
w
00
rH
oo
PERCENT REDUCTION IN AREAWIDE WORK TRIP VMT
-------
IV-24
air quality problems into conformance with the health standards for air
pollutant concentrations, designing effective packages of measures will
be an important aspect of air quality/transportation planning. The
technique of comparing single and multiple measures can aid in the
development of a cost-effective package of measures which can lead to
significant improvements in air quality.
-------
CASE STUDY V
ANALYZING A PROPOSED BUSWAY NETWORK
-------
CASE STUDY V: ANALYZING A PROPOSED BUSWAY NETWORK
A. Problem Statement
A small but densely settled city with relatively poor existing transit
service is considering a number of major public transit improvement pro-
jects. Among the proposed new systems is a network of busways providing
line-haul express service to the CBD, supported by improved conventional
fixed-route bus service and demand responsive feeder bus service in
residential areas. In order to choose between the proposed busway option
and the other possible alternatives, the local planning agency and deci-
sion makers need to know what the transit ridership, VMT, and air pollution
impacts of the proposed system will be. The urban area has implemented the
standard UTPS model system for use in analyzing its transportation system,
but has found the standard aggregate passenger demand models difficult and
expensive to use. Because a large number of alternative solutions to the
area's public transportation needs (including the proposed busway) are under
initial consideration, a fast and inexpensive method of screening the range
of alternatives is desired, so that detailed analyses can be conducted for
only the most promising alternatives.
B. Proposed Transportation Measures
The proposed network of busways and supporting bus services are shown
in Figure V-l. Three busways, serving the northwest, southeast and south-
west travel corridors in the city would be built under this plan. Several
express bus routes would use each of the busways and three express routes
would use the existing street network. A number of local fixed-route
-------
• Express Bus Stops/Stations
FIGURE V-l
Busway Network
TRANSIT STRATEGY OPTION
BUSWAY NETUOM
TRANSIT POLICY OPTION
Moderate Service Level
LINE HAUL _ Express Buses on/off Busway
~ Conventional Fixed Route
COLLECTION (Residential)
.PEAK HOURS - "Rationalized" Telebus
OFF-PEAK - Telebus
DISTRIBUTION - Overlapping Fixed Routes
(CBD, Activity Centers)
FIXED EXPRESS/
FEEDER ROUTE BUSWAY
on
demand
550 ft. 550 ft.
800 ft. 4000 ft.
STATION
SPACING
AVERAGE CBD
AVERAGE SUB-
URBAN
SERVICE HOURS
PEAK
OFF-PEAK
SERVICE FRE-
QUENCY
PEAK
OFF-PEAK
NIGHT
FARES
ADULT
* INCLUDES FREE TRANSFER TO OTHER SERVICES
5 hrs S hrs 5 hrs
13 hrs 13.5 hrs -/? hrs
12 min 12 min 30/12 min
20 min 20 min -/20 min
30 min 30 min
30$
<
-------
V-3
services from the existing transit system would remain in service, supple-
mented by Dial-a-Ride services (termed "Telebus") in outlying suburban
residential areas. Coordinated transfers are planned between nearly all
services.
In order to maximize the usage of the routes using the new busways,
a package of auto disincentives is planned for implementation when busway
construction is completed. Increased parking costs for CBD auto commuters,
and a limit on the construction of new parking spaces in the CBD are the
major features of the auto disincentive plan.
The busway plan, if implemented, could be completed by the mid 1980's.
An analysis year of 1987 was chosen to coincide with the deadline for the
attainment of the health standards for air pollution since the expansion
of bus service in the city is anticipated to have a significant impact on
air quality, particularly in the city's downtown area.
C. Selection of Analysis Technique
The important factors affecting the choice of the analysis technique
to use in evaluating the busway alternative (and the other proposed transit
development projects) are:
• The budget available for the analysis of the proposals is
limited. At this early stage, when a range of alternatives
is being explored, a detailed analysis of all of the pro-
posals using the full UTPS model system is not possible.
• A large number of alternatives must receive an initial
screening. These alternatives represent a significant
change in existing conditions and will lead to the provi-
sion of completely new services in many areas of the city.
Because the level-of-service changes associated with the
proposed new transit systems are not incremental, the manual
and calculator pivot-point demand analysis methods are not
appropriate.
-------
Y-4
• Many of the proposals involve new construction, so that the
changes being evaluated are facility- rather than service-
oriented. This fact> combined with the limited budget avail-
able for implementing new computer analysis capabilities and
the lack of a recent home interview survey, argues against
the use of SRGP, which is more applicable to policy measures
(such as carpooling programs and parking management strategies).
The transit sketch planning procedure (Volume I, Section 2.3.3)
selected for the analysis is particularly well-suited to the analysis and
initial screening of a large number of transit system development alter-
natives. Its simplified structure and the incorporation of manual work-
sheet techniques makes it relatively inexpensive to set up; and the
utilization of computer steps using UTPS programs reduces the computational
burden associated with analyzing a large number of alternatives.
I). Overview of the Analysis
The sequence of steps followed in applying the transit sketch plan-
ning procedure is shown in Figure V-2. First, a zone system is established
by aggregating existing traffic zones to form up to 100 analysis zones.
Formulation of the transit alternative is accomplished by means of system
description log sheets, which summarize specific characteristics of the
alternative, and a simplified transit network coded through the use of
UTPS program HR. Next, a sample of representative trips are chosen for
detailed demand analysis, the results of which will be mapped onto all
other origin/destination pairs. Market segments based on transit access
mode are defined for these zone pairs, and service levels are developed
for each market segment. Then, travel demand models are applied manually
to estimate mode shares for representative zone pairs. These representa-
tive mode shares are then mapped onto all origin/destination pairs
-------
V-5
FIGURE V-2
Major Steps in Sketch Planning Procedure
1
Develop Zone System
2
Formulate Transit
Alternative(s)
System Description
(Log Sheet, Map)
Code Transit Network
(HR)
Choose Representative
Zone Pairs, Mapping
Matrix
6,7
Develop Service Levels
for Representative
Zone Pairs
Estimate Mode Shares
for Representative
Zone Pairs
Demand Model(s)
Socio-economic Data
Market Segments
9
Map Representative
Mode Shares onto All
Trips (UMODEL)
10
Assign Transit Trips
to Network
(UROAD)
Trip Table
Mapping Matrix
Peaking Factors
Market Segments based
on Service Level
11
Cost/Revenue
Estimation
12
Environmental Impact
Estimation
-------
V-6
using a user-coded version of the UTPS program UMODEL. UMODEL produces
transit trip tables which are assigned to the transit network using UROAD.
A series of manual worksheets is completed to estimate the transit system's
cost and performance and environmental impacts.
E. Input Data Development
Steps 1 through 7 of the analysis procedure involve the development
of input data and an analysis framework.
Step 1; Establishment of a Zone System—The first step in applying
the transit sketch planning procedure is the definition of a zone system.
Three criteria were used to define the zonal structure:
• the boundaries of each zone are located to avoid having multiple
transit routes with differing levels of service in the same zone;
• the number of intrazonal trips was kept to approximately 5 percent
using an approximate relationship between interzonal trip making
rates and zone size included in the sketch planning model;
• zone boundaries were constructed to conform, as much as possible
with existing census tract boundaries.
The 32-zone system shown in Figure V-3 was developed for use in the analysis
of each of the proposed transit alternatives. Between 30 and 40 zones will
typically be adequate for use in analyzing a small to medium-sized city.
A somewhat larger zone system would be required to analyze an entire
large urban area.
Step 2: Formulation of Transit Alternative—The busway alternative
considered in this illustration was described in Section B. The alternatives
to be analyzed must be developed in at least the level of detail shown in
Figure IV-1. Headways, station spacings, service hours, and fares
developed for all transit services, and routes are defined at a level
-------
V-7
FIGURE V-3
1987 Transit Zones
-------
V-8
specific enough to indicate the level of transit service available in each
of the analysis zones defined in Step 1. A map such as the one shown in
Figure V-l is adequate for the purpose of defining the proposed transit
system.
Step 3: Network Coding—A network was coded to correspond to the
busway system shown in Figure V-l. The network consisted of 56 centroid
connectors and 75 links representing the transit services. No specific
route structure or headways are coded in the network explicitly—only
link travel times are -used. Also, no feeder system information is coded
in the network. Each of these considerations is dealt with manually in
defining transit level of service between zones. The coded transit net-
work is used only for assigning transit trips once the transit mode share
has been determined using manual or programmable calculator procedures.
Finally, no highway network was coded. Highway zone-to-zone travel times
used in the demand model were obtained from actual surveys conducted in
1976, and were assumed to increase by 10 percent in the 1987 analysis year.
Step 4: Detailed System Description—The system description log sheet
sheet shown in Figure V-4 is used to specify the characteristics of the
proposed transit services. All of the variables shown on the system
description log sheet must be relatively constant over the system being
analyzed.
Step 5; Choice of Representative Trips—Seventeen zone pairs were
selected to represent trip making throughout the urban area (up to 18 zone
pairs may be used in the sketch planning procedure). The representative
zone (0-D) pairs were selected so as to cover the range of trip types in
the urban area. Table V-l shows the 0-D pairs selected and the type of
-------
V-9
Option:
Time Period;
Busway
Peak
Linehaul 1
Linehaul 2
Feeder
Bus
Bus
Minibus
Linehaul 1
Linehaul 2
Feeder
Express
Local
Dial-a-ride
Technology:
Operation:
Station Spacing: Linehaul 1
Linehaul 2
Vehicle Size:
Feeder Service: Area of Zones 0.8
0.8
0.1
mi.
mi.
Linehaul 1
Linehaul 2
Feeder
50
50
24
seats
seats
seats
mi.'
Integrated with linehaul_
If separate, headway
or separate
minutes
Linehaul Service 1: Headway, CBD routes
12
Headway, non-CBD routes
Average number of
intermediate stops
all
Linehaul Service 2: Headway, CBD routes
12
Headway, non-CBD routes
Average number of
intermediate stops
all
minutes
minutes
(CBD routes)
minutes
minutes
(CBD routes)
Fare:
Linehaul 1 :
Linehaul 2:
Zone:
Feeder:
Transfer:
45
30
0
45
0
Speed:
Linehaul 1:
Linehaul 2:
CBD Linehaul:
Feeder:
20
12
10
cents
cents
cents (trips over
cents
miles)
mph (with
mph (with
or without
or without
stops)
stops)
mph (with stops)
mph (with stops)
Other:
See supplementary sheet describing auto disincentives.
FIGURE,V-4
System Description Log Sheet
-------
v-io
TABLE V-l
Representative Trips for the Busway Option
1KDEX 0-D PAIR TRIP CHARACTERISTICS
1 23-13 Refill a r Transit + Telebus; Mcdlua Distance
to CBD
2 8-13 Regular Transit + Telebus; Short Distance
to COD
3 15-13 Express Bus to CBD
4 25-13 Regular Transit; Medium Distance to COO
5 16-13 Regular Transit; Short Distance to CBD
6 6-1 Crosstown Telebus - Telebus Direct
7 3-1 Crosstown Short Distance; Regular Transit -
Telebus
8 31-6 Crosstown Medium Distance; Regular Transit
Telebus + Transfer
9 18-3 Crosstown Medium Distance; Regular Transit
+ Telebus
10 19-9 Crosstown Medium Distance; Regular Transit
11 15-7 Crosstown Short Distance; Regular Transit
12 28-7 Crosstown Long Distance; Regular Transit
13 22-22 Intra-zone Telebus
14 17-17 In,tra-zone Regular Transit
IS 29-13 Busway to CEO
16 22-8 Crosstown Long Distance Busway
17 25-15 Crosstown Medium Distance BUJWJ/
-------
V-ll
trip represented by each. Note that the characteristics of the transit
alternative under study, as well as the location of the zones in the 0-D
pair define the type of trip a given pair is chosen to represent. The
number of representative 0-D pairs necessary to reflect the range of trip
types is a function of the characteristics of the urban area under study,
and the extent and variety of transit services being analyzed.
Once representative zone pairs had been identified, a mapping matrix
was developed relating each possible zone-pair to one of the selected
representative trips. The classification of zone pairs relies to a great
extent on analyst judgment, based on a knowledge of the urban area and the
proposed transit system. Travel demand model coefficients aid in the
classification process by identifying the most sensitive variables to con-
sider in matching zone pairs and representative trips.
Step 6; Market Segmentation Based on Service Level—Each zone was
then broken into market segments defined by the type of access to transit
service. Table V-2 shows the way in which the market segmentation is pre-
pared as input for the model. Each of the 32 zones is divided into three
"subzones": near, middle, and far. The near subzone is the area of the
zone located within 0.25 miles of the linehaul transit route; potential
transit users are assumed to walk to the route from this area. The middle
subzone is the area beyond 0.25 miles of the linehaul route, but within 0.25
miles of a feeder service; individuals in this subzone are assumed to use
the feeder service. The far subzone is the area beyond a 0.25 mile walk to
either a linehaul or feeder route; park-ride or kiss-ride are assumed to
Trips between zones are assumed to have the same transportation level of
service, regardless of direction.
-------
V-12
TABLE V-2
Subzone Percentages Busway Network
1
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Near
Sub-Zone
20
5
20
10
100
75
100
70
90
100
100
85
100
100
50
10
75
75
75
75
5
0
10
10
100
0
0
0
0
0
0
0
Middle
Sub- Zone
80
95
80
0
0
25
0
30
0
0
0
0
0
0
0
90
0
0
0
25
95
100
90
90
0
100
100
100
100
100
100
100
Far
Sub-Zone
0
0
0
90
0
0
0
0
10
0
0
15
0
0
50
0
25
25
25
0
0
0
0
0
0
0
0
0
0
0
0
0
Total
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
-------
V-13
be the only transit access options available. (Note that the procedure
assumes that if feeder service is available, it will be used; park-ride
or kiss-ride are not available options for users in the middle subzone.)
In the transit demand analysis, market segments are defined on an
origin/destination basis for each zone pair. Thus a near-subzone to near-
subzone trips corresponds to a walk-to-transit/walk-from-transit market
segment. Transit trips are presumed not to serve trips to far-subzones
in the destination zone, thus six possible market segments exist for each
zone pair, with the size being determined by the product of the sub-zone
percentages for the origin and destination zone. Thus, using figures from
Table V-2, the Walk/Walk market segment from zone 1 to zone 9 is .20 x .90,
or 18 percent of the total zone 1 to zone 9 market. The UTPS program UMODEL
described later computes the market segment proportions for all 0-D pairs
in the urban area.
Step 7: Development of Service Levels for Representative Zone Pairs—
For each of the 17 representative zone pairs, it is necessary to code the
cost (OPTC), in-vehicle travel time (IVTT), and out-of-vehicle time (OVTT)
information for the three modal alternatives used in the demand model: auto
drive-alone, auto shared-ride, and transit. This information is developed
by the analyst using the transit system map or network, the system descrip-
tion log sheet, a general description of the road system, and general
assumptions such as those outlined in Table V-3. Figure V-5 shows the first
sheet (of two) used to record the service levels for 0-D pair 23-13. Note
that service levels are found separately for each market segment in the zone
pair; the second sheet contains the remaining four market segments not in-
cluded in the first sheet.
-------
Y-14
TABLE V-3
Level of Service Assumptions (1987)
Auto Out-of-Pocket Travel Costs
Cl 987)
Auto In-Vehicle Travel Time
Auto Out-of-Vehicle Travel Time
Auto Daily CBD Parking Cost
Carpool Auto Occupancy Rate
Carpool In-Vehicle Travel Time Penalty
Carpool Out-of-Vehicle Travel Time
Penalty
Transit In-Vehicle Travel Time
Transit Fares
Walk Times to Transit
Wait Times at Trip Origin
Feeder-Line Haul Transfer
12c per mile.
Taken from speed and travel time
data provided by City Traffic
and Engineering Departments
Two minutes at non-CBD origins or
destinations.
Three minutes at CBD origins or
destinations.
$1.20
2.5 persons
Five minutes
One minute
Taken from Transit System
running boards.
30c non-Telebus trips.
45C trips with Telebus.
Three minutes for route.
0.50 minutes for Telebus.
h headway for fixed route.
Three minutes for Telebus.*
Four minutes for transfers which
occur in CBD.*
% headway for non-CBD transfers.
Three minutes.*
Assumptions for Auto Disincentives
Auto Out-of-Vehicle Travel Time
Auto Daily CBD Parking Cost
Six minutes for CBD origins or
destinations.
$2.40
Coordinated transfers.
-------
V-lf
Year lit?
Systeei »..«.«mf_ Operating Policy narrate
Description .
Origin !S Destination l3 Auto Distance 3-<
(All Values' are One-Way Values)
DRIVE ALOHE
OPTC:_t: _ min/Bile (jLJLWph) x ?.
-------
V-16
F. Description of Model Application
Steps 8 through 10 involve the estimation of mode shares for the
transit, carpool, and drive-alone modes. First, the representative work
and non-work mode shares for each market segment in each 0-D pair are cal-
culated using either manual worksheets or a programmable calculator version
of the worksheets. The representative zone pair results are then mapped
onto total trips in the city using the mapping matrix of Figure V-6 and a
user coded version of the UTPS Program UMODEL. The predicted zone-to-zone
transit trips are then assigned to the transit network coded in Step 3
using the UTPS program UROAD.
Step 8; Estimate Mode Shares for Representative Zone Pairs—Figures
V-7 and V-8 show the worksheets used to calculate the transit mode splits
for each of the representative trips for the peak and off-peak period for
0-D pair 23-13. Mode splits are calculated for each zone pair and each market
segment separately using the information coded in Figure V-4. For this
study, the modelling did not consider park-and-ride or kiss-and ride trips
because of the city's small size; trips originating in the far subzone of
each zone were assumed to not have transit service available. Both of the
demand models used in this application of the transit sketch planning pro-
cedure were disaggregate travel demand models transferred from Washington,
B.C., and updated for application to this urban area.
For a description of the technique for updating disaggregate travel demand
models, see Atherton, Terry J. and Moshe Ben-Akiva, "Transferability and
Updating of Disaggregate Travel Demand Models," Transportation Research
Record 610, Transportation Research Board, Washington, D.C., 1977.
-------
V-17
Kap of Representative Trips
From/To Zone
t 1 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 19 30 31 32
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
13
7 13
7 6
II II
I) 16
6 7
II II
9 16
10 17
U 17
1 IS
II 16
1 15
II IS
10 17
J 17
10 17
9 12
9 12
17 17
17 17
17 16
17 16
17 16
9 16
10 10
9 17
17 16
17 16
17 16
17 16
17 16
13
16 14
16 11
7 M
11 II
16 16
17 17
17 17
IS 4
16 10
15 *
15 10
17 17
17 17
17 17
12 12
12 12
17 17
17 17
16 16
16 16
16 16
16 16
10 9
17 17
16 16
16 16
16 16
16 16
16 16
14
11 13
II U 14
16 8 10
17 11 8
17 II 11
15 1 4
16 11 U
15 I *
IS 10 10
II 11 11
17 10 II
17 10 U
12 10 1)
12 9 11
17 9 12
17 9 12
16 17 12
16 17 17
16 17 12
17 9 10
10 10 10
17 17 10
16 17 12
16 17 12
16 17 12
16 17 17
16 17 17
13
10 Id
10 U
16 15
U II
16 4
15 10
17 10
17 10
17 11
12 U
12 10
17 12
12 17
16 16
17 16
16 16
17 9
12 10
10 10
15 12
15 12
12 12
16 16
16 16
14
5 14
11 10
5 5
H 5
11 3
It 1
It 3
II 5
10 5
12 17
17 17
12 15
12 15
12 15
9 <>
10 4
10 2
12 15
15 15
12 IS
12 15
12 15
14 .
10 14
10 5
1 3
11 1
10 3
II 5
11 5
12 17
12 17
12 IS
12 15
12 15
10 4
9 *
10 2
12 15
12 15
12 15
12 15
12 15
14
10 14
1 10
11 U
1) 11
11 10
17 17
17 17
15 15
15 15
15 IS
11 10
11 10
10 10
15 17
15 17
15 17
15 17
15 17
13
II 14
11 11
10 10
17 12
17 12
17 12
17 17
17 17
10 II
10 10
7 II
17 12
17 12
17 12
17 12
15 12
14
10 14
10 10 13
10 10 6
10 8 8
12 10 8
12 10 8
3 9 17
10 10 12
9 9 17
12 12 9
12 12 9
12 12 12
12 12 17
12 12 17
13
6 13
8 6 13
8 7 7
17 17 17
12 12 12
17 17 17
9 12 12
9 12 12
12 9 9
17 6 16
17 16 16
13
II 14
12 10 13
17 9 10 13
17 10 12 8 13
8 11 12 6 8 13
8 II 12 17 8 8 13
16 17 12 17 16 16 16 13
16 17 12 17 16 16 16 6 13
FIGURE V-6
Mapping Matrix
-------
FIGURE V-7
Work Mode Choice Sheet
FIGURE V-8
Non-Work Mode Choice Sheet
System Burjtat
Year 1987
Description
Origin t3
Destination
Distance 3.4 miles
SOCIO-ECOHOHIC ATTRIBUTES
U'p « -3.066 + .000044 DINC 14975
U'C • -1.777 * 3.35 AALO .S3 4-.896 BW .54
-.998 DCITY Z . * .000044 DINC 14975
U'S « -1.448 + 1.68 AALD .83 -.514 DCITY _Z_
«• .000044 DINC 1497S * .036 NUORK l.SS
+ .00084 DTECA 0
TOTAL UTILITIES
B! • -27/INC 1 • ..OOIS 6j • -.OZi< B3
U Transit
HOPE U-m (OPTC x BI * IVTT x 6? * OVTT X t3) • Urn e Mode Share
(U'c)
* \--zo I
(U's)
• -.046/OIST - -
Drive ^
Alone
+2 {
Car-
Pool
... to *2
J --JJ
) •-Z.Ot.3C
Transit
HR-NR 0 *2 (_
.NR-MIO _0_ +2 (_
MID-NR _0_ +2 (_
HID-MID_0_ *2 (_
FAR-HR +2 (_
U p
FAR-MID +2 (_
. -Z.gj .31 .28 (MN)
- (FM)
System ""«**>
Operating Policy Hodefatt Year l987
Description
Origin 29
Destination 13
Distance 4.8 miles
SOCIO-ECOHOHIC ATTRIBUTES
U'C « -.462 + .605 AAC 1.38
TOTAL UTILITIES
-.JiTj|(U-c)
Bj • -.0203/1NC 2 • -.002S 63 • -.0220/D1ST • -.0210 83 • -i.lil
HOPE (OPTC xjli * OVTT x fl, + ln(lVTT+QVTT) x g,)- Um e"m shire'*
Auto -..J65 -2(.
Transit:
NR-NR 0 -2(_
HID-NR 0 -2(_
MID-MID_0_ -2
FAR-NR _0_ -2(_
FAR-MID 0 -2(
)•
I • .JS_ (UN)
. ^ (Ml)
• -Of-
-------
V-19
No market segmentation based on socioeconomic characteristics was used
in this particular application. This decision was made in part due to re-
source constraints, and also because of the relative homogeneity of house-
holds in many zones in the city. In the most general version of the model,
however, Figures V-7 and V-8 would reflect market segmentation by auto
ownership level or other socioeconomic characteristics as well as location
within the zone. The variables used in these worksheets are defined in
Figure V-5.
Step 9: Map Representative Mode Shares Onto All Trips (UMODEL)—A
user-coded version of UMODEL is used for this step; the inputs to the
program are:
• a set of directional peaking factors, derived from local highway
and transit system data if possible; otherwise default values are
included in the model;
• proportion of households in each subzone or service level market
segment, by zone;
• mapping matrix that translates representative 0-D pairs onto all
0-D pairs;
• mode shares calculated for each representative trip, by market
segment.
Note that UMODEL must be run separately for peak and off-peak analyses
since different service levels, service level market segments, mode shares,
and peaking factors normally will exist for each of these periods. UMODEL
produces four basic outputs:
• a set of zone-to-zone person trip tables by mode
• a trip end summary of person trips by mode
• a trip end summary of transit trips by access mode (subzone)
• a trip end summary of transit revenue by access mode; an additional
input—fares by service market segment for each representative zone
pair—is required if this output is desired.
-------
Y-20
Step 10; Network Assignment (UROAD)—The transit trip tables produced
by the UMODEL program are assigned to the networks prepared earlier using
the UROAD assignment program. All transit trips are assigned on an all-or-
nothing basis (UROAD selects the single shortest path for each 0-D pair
and assigns all trips to that path.) The link loads are plotted on network
maps and balanced manually, if required.
G. Impact Assessment
Steps 11 and 12 of the sketch planning procedure involve the use of
manual calculation worksheets to estimate the cost and performance charac-
teristics of the proposed busway system, and the change in fuel consumption,
emissions, and. traffic accidents for both autos and transit.
Step 11; Cost and Performance Estimation—A series of manual work-
sheets are used in this step to determine the bus system's performance and
capital and operating costs. The peak period annual operating costs are
calculated using the worksheet shown in Figure V-9. Route mile and headway
data are used to calculate the hourly vehicle miles required to cover the
system in the peak hour on lines 1 through 3. The annual vehicle mileage-
related cost for the fixed route system is calculated in line 4, assuming
a 43-cent-per-mile operating cost. Similar calculations are used to deter-
mine the annual mileage-related costs for the demand responsive feeder
service in Lines 5 and 6. Line 7 is used to determine the number of
vehicles required to provide the peak period service, and Line 8 uses this
information to calculate the annual vehicle-hour related expenses, based
on an hourly cost of $15.63. Finally, other operating expenses are cal-
culated in Line 9, and the total annual operating expense for the proposed
-------
V-21
System ttaaaH Operating Policy Hodtfat' Servic* Ltoel
PEAK PERIOD OPERATING COST
1. VH]»C -RMlJ"
2.
KM
3. VH^ • RM1J£ J£J.x60/l|Pj Z2
5- VHe/d .1^1^==
X C
,h P.«3 (_lW£pk.hr./yr)
X 60
\ "fix -=— Hdr -^~
6. CVM,., - VH .. S80 -C-M 0.43 X (M00pk.hr,
C/0 C/u—^—~ C/o ^^— -
hr/yr)
., 60 x VM]?C «0_ 60 x VH"P 35 60 x
7 . flV^ * - 'h - + - 'h - +
(50-L)SS. x
Year
Zones* x J
60 x VH"P 35 60 x VM^ 123
- 'h - + - 1h - - + Extras 80
(60-L)JSx SexfT£ (60-L) «_x 5^20.
8. CH
9. COP
« CFO
d J£J x
x G _^
_^_ x NS
x tlV Z33 /u .8
{7so|VH.
r«
,lh
)oc
,lh
exp
.39S« CTH
Ih
VH
JCVH.
Mh
HDwJd
3.25S.V|CH
Ui»_>dcopoth
copg^HToTir
Number of demand-responsive zones based on peak demand-responsive demand.
FIGURE V-9
Peak Period Operating Cost Sheet
-------
V-22
system is determined by adding lines 4, 6, 8 and 9. Similar calculations
are carried out on the worksheet in Figure V-10 to determine the total
capital cost of the system based on the number of vehicles of each type
required, the per vehicle cost, and the cost of the guideway. Both total
and equivalent annual capital costs are calculated.
Step 12; Estimate Environmental Impacts—Simple manual worksheets
for estimating changes in emissions, fuel consumption and accidents are
available within the sketch planning procedure. Changes in annual auto VMT
are used as base data in the calculations. The base case VMT for the
analysis year must be determined by running the model system with a transit
system reflecting existing conditions. The VMT predicted for each transit
alternative under study is then compared to the base VMT to determine VMT
changes. The worksheet for calculating emissions is shown in Figure V-ll.
The annual change in total emissions is calculated solely from the predicted
change in VMT for autos and buses. Auto emissions rates for CO, HC, and
NOx are determined from Table D.7 in Volume I, based on an assumed travel
speed, and bus emission rates acquired from EPA Report AP-42. A more
sophisticated and accurate appraisal of the air quality impact of the alter-
native transit systems could be made using the auto emissions worksheets
(see Volume I, Section 4.1).
H. Interpretation of Results
The manual sketch planning procedure as applied to the busway alterna-
tive provides a comprehensive set of estimates of the impact of transit
development alternatives. Data on capital and operating cost, revenue,
Environmental Protection Agency, "Compilation of Pollutant Emissions
Factors," EPA Report AP-42, Second Edition, 1976.
-------
Y-23
System liuaoan Operating Policy HoAtrot* Technology Bui
CAPITAL COST (Repeat this block for each technology)
Tine Period IS87
M.
Ih
CV
Ih
CV
Ih
erf
»ih '
NV
CV
art
•rt
CV'
art
-------
SYSTEM
Busway
OPERATING POLICY
Moderate
COMMENTS 1987 analysis year
ENVIRONMENTAL EFFECTS _
(All VMT Figures in Annual Miles x 10 )
UROAD VMT, THIS ALTERNATIVE
UROAD VMT, TARGET YEAR BASE
CHANGE IN HIGHWAY VMT
STANDARD BUS VMT, THIS ALTERNATIVE
STANDARD BUS VMT, TARGET YEAR BASE
CHANGE IN STANDARD BUS VMT
MINI BUS VMT, THIS ALTERNATIVE
MINI BUS VMT, TARGET YEAR BASE
CHANGE IN MINI BUS VMT
AIR POLLUTION
ACO = AVMTA x COA -283,500 + AVMT^ x
AHC = AVMT x HC - 22,680 + AVMT^ x HC_ +6,924
ANOX= AVMT. x NOX. - 35.154 + AVMTOT1 x NOX,_+6,500
A A bo oa
143,333
150,500
-7,167
(XI)
1,226
994
232"
(X3)
870
214
656
(X5)
+5,134 + AVMT,,
MJ
+6,924 + AVMT,,.
n.
,,+6,500 + AVMT,_
199,827
204,000
-4,173
(X2)
1,194
958
239
(X4)
822
929
-107
(X6)
x CO +65,880
L> MIS
x HC + 8,113
L) M13
_xNOX^_ +4,337
X1+X2
X3+X4
X5+X6
-212,486
- 6,643
- 24,317
-11,340 AVMT.
471 AVMT,
SB
549 AVMT,
MB
ACO(kg)
AHC(kg)
ANOX(kg)
FIGURE V-ll
Environmental Impacts Sheet
-------
V-25
ridership, VMT, fuel consumption and air quality are developed in an
integrated procedure. Table V-5 illustrates the type of system impact
summary which can be developed for each of a range of alternative trans-
portation systems. The information available from the procedure is
sufficient to screen out a small subset of attractive options for further
development and detailed analysis.
-------
OPERATING CHARACTERISTICS
ANNUAL SERVICE HOURS
ANNUAL VEHICLE MILES
Large Bus Miles
Mini Bus Miles
Total Bus Miles
Auto Miles (000's)
ANNUAL RIDERSHIP
ANNUAL COSTS (1976 $)
Operating Costs
Capital Costs
Total Costs
Peak Hours Off Peak Totals
207,000 182,600 389,600
1,226,000
(1,962,000)
870,000
(1,392,000)
2,096,000
(3,354,000)
143,333
(229,333)
1,194,000 2,421,000
(1,910,000)(3,874,000)
822,000 1,692,000
(1,315,000)(2,707,000)
2,016,000 4,113,000
(3,225,000)(6,581,000)
199,827 343,159
(319,722) (549,056)
7,051,000 6,318,000 13,369,000
5,013,000 3,665,000 8,679,000
5,833,000 4,073,000 9,906,000
2,007,000 - 2,007,000
10,686,000
11,913,000
ANNUAL REVENUES (1976 $)2,343,000 1,650,000 3,993,000
ANNUAL DEFICIT
8,693,000
7, 920,000
Note: Figures in Italics shou "high" trend in driver wages.
Figures in Parenthesis show metric equivalents in hn.
TABLE V-4
Imoact Summary for Busway Transit Option
SERVICE CHARACTERISTICS
ONE-WAY ROUTE MILES
Peak Hours
90(144)
Conven. Fixed Route
Fixed Route Feeder
Telebus Feeder(approx.)l16(186)
Express Bus 17.5(28)
Express Bus (Busway) 25.6(50)
Total Route Miles 249(399)
FEEDER ZONES
Fixed Route Lines
Buses/Line
Telebus Zones 29
Buses/Zone 2
VEHICLES REQUIREMENTS
Articulated Buses 25
Standard Buses 75
Mini Buses 73
UNIT CHARACTERISTICS
PER CAPITA
Off-Peak
79(126)
71(114)
25.6(50)
175.6(281)
15
1
Annual Bus-Miles
Annual Ridership
Annual Cost
Annual Deficit
Annual Auto-Miles
PER PASSENGER
Cost
Deficit
21.7 (34.7)
70.5 70.5
$ 56.4 $ 62.8
$ 35.3 $ 41.8
1810 (2896)
.799
.501
$.891
$.592
<
NJ
-------
CASE STUDY VI
FREEWAY RAMP METERING IN A HEAVILY CONGESTED CORRIDOR
-------
CASE STUDY VI: FREEWAY RAMP METERING
IN A HEAVILY CONGESTED CORRIDORl
A. Problem Presentation
The Eastshore Freeway, a major urban highway on the eastern side of
San Francisco Bay, serves commuter traffic destined for San Francisco,
Oakland, and Berkeley from the Northeast quadrant of the Bay Area (see
Figure VI-1). Rapid suburban development oriented around the corridor
has lead to increasing peak hour congestion in recent years, with a
spillover effect on parallel arterial routes.
The Eastshore Freeway Corridor begins at the junction of three major
freeways: 1-80 from San Francisco via the Bay Bridge, State Route (SR) 17
from Oakland and the South Bay Area, and 1-580 from Oakland and the
eastern suburbs. The corridor is approximately 10 miles long.
The freeway is heavily travelled throughout the study section but
congestion is particularly bad at the ends. In general, demand diminishes
along the study section but at a slower rate than capacity, resulting in
bottlenecks and congestion under the present situation. On San Pablo Avenue,
a parallel arterial street, traffic is moderate except for several
critical intersections with heavy cross-flow, such as University Avenue,
Cutting Boulevard and Barrett Avenue (see Figure VI-2).
San Pablo Avenue is a four-lane (occasionally six-lane) signalized
major arterial located parallel to the freeway and extending for the en-
tire length of the study section. Figure IV-2 shows its location in
relation to the freeway and the placement of signalized intersections.
This case study was originally developed by the Institute of Transportation
Studies, University of California, Berkeley. It has been adapted from material
in: Jovanis, Paul P., Wai Ki Yip, and Adolf D. May, FREQ6PE - A Freeway Priority
Entry Control Simulation Model, Institute of Transportation Studies, University
of California, Berkeley, November 1978.
-------
VI-2
FIGURE IV-1
Location of Eastshore Freeway Corridor
-------
VI-3
CENTRAL AVE.
LEGEND
Oj = Origin t
Dj E Destinotion j
: Signolized intersection
PIERCE AVE.
MARIN AVE.
]1_G1LMAN ST.
UNIVERSITY AVE.
FIGURE VI-2
Detailed Corridor Map
ASHBY AVE.
STANFORD
-------
VI-4
B. Proposed Transportation Measures
Funding limitations and the substantial capacity constraint of the
Bay Bridge (five lanes fed by three major freeways) have prompted the con-
sideration of non-construction alternatives for mitigating congestion in
the corridor. Ramp metering has been prominent among these proposals.
Two approaches to ramp metering are under consideration; one in which
vehicles entering the freeway are controlled without regard to type, and
one affording priority ramp entry for multiple occupant vehicles. These
strategies are studied for the PM peak only, in the northbound direction,
although, in principal, southbound AM congestion could have been evaluated
in a similar analysis.
C. Selection of Analysis Technique
Because of the complexity involved in the detailed planning and eval-
uation of freeway ramp metering on a specific highway, manual sketch plan-
ning techniques are not appropriate for this analysis. The computer model
FREQ6PE was specifically developed for analyzing freeway measures such as
ramp metering. Given its computational power, the data requirements and
run expenses for the program are modest. FREQ6PE is, therefore, chosen to
analyze the ramp metering options for the Eastshore Freeway.
D. Overview of the Analysis
The major effort involved in the ramp metering analysis is the prepara-
tion of input data for the FREQ program. The program itself produces
vehicle emissions and fuel consumption impact estimates, as well as a
Jovanis, Paul P., Wai KiYip, and Adolf D. May, FREQ6PE - A Freeway Priority
Entry Control Simulation Model, Institute of Transportation Studies, Univer-
sity of California, Berkeley, November 1978. Also see Volume I, Section 3.3.2
-------
VI-5
description of traffic flow on the Eastshore Freeway and San Pablo Avenue.
FREQ requires a detailed description of freeway and entrance ramp
geometry, origin-destination flows, and freeway operating policies,
including:
• "time slices" defined to capture the distinct phases of freeway
flow
• r,o more than 40 subsections of the freeway and their capacity,
length, and other characteristics
• similar subsections of the parallel arterial
• ramp capacities
• origin-destination volumes for autos and buses
• base flows on arterial subsections.
The following section describes the development of input data for the
analysis.
E. Input Data Development
The input data required for the FREQ6PE model are of three major
types: freeway facility data, freeway demand data, and parallel
arterial route data. Each of these three data types is described below.
The ten-mile study section was divided into twenty-seven sub-
sections with boundaries signifying changes in traffic demands and/or
capacities. The applicable data for each subsection are shown in
Figure VI-3. For example, subsection 20 is three-lanes wide, has a
capacity of 5880 vehicles per hour, and a length of 1100 feet. It has
an origin (on-ramp) at the beginning of the subsection, designated as
a special ramp because of the extended length of the merge area. The
-------
SUB
SEC
1
2
3
4
5
6
7
8
9
X
10
±\J
11
12
13
14
15
16
JLV
17
A f
Ifl
AVI
IP
JL7
70
*. V
21
22
23
24
25
26
27
NO.
LNS
5
5
5
4
4
4
4
4
4
5
4
3
3
3
3
3
3
3
J
3
*/
3
CARLSON ON TO POTRERO OFF
POTRERO OFF TO CUTTING ON
CUTTING ON TO GRADE CHGE
GRADE CHGE PT TO MACDONALD
MACDONALD OFF TO SAN PABLO
SAN PABLO OFF TO SAN PABLO
SAN PABLO ON TO SOLANO ON
SOLANO OFF TO S. PABLO ON
DAM ROAD OFF TO DAM RD ON
DAM RD ON TO RD 20 OFF
* INDICATES USER-SUPPLIED SPEED-FLOW CURVE NUMBER
RAMP LIMITS - 1500.
OFF-RAMP 6 LIMIT = 2200.
FIGURE VI-3
Eastshore Freeway Facility Data
-------
VI-7
subsection has a 2 percent up-grade and a smooth surface, and is on a
tangent. Three percent of the vehicles are buses and trucks, and 30
percent of these are diesel powered.
The freeway design speed is used to indicate the appropriate
speed-volume/capacity relationship for each subsection. These V/C
curves are internal to the program. A sixty-five mph curve was used
for subsections 6 through 27 whereas a fifty-five mph curve was used
for subsections 1 through A because of heavy merging traffic due to
the three upstream freeways. An intermediate user supplied curve was
used for subsection 5. A capacity value of 1500 vehicles per hour
was selected for all ramps except for off-ramp 6, which was assigned a
value of 2200 vehicles per hour. Because of the heavy weaving in many
areas of the study section, an internal model option was used to
decrease the capacity of certain sections.
The afternoon peak in this corridor extends roughly from 3:30 PM
to 6:00 PM. Fifteen minute time "slices" were selected for analysis,
yielding a total of 10 separate time periods for the peak. Two sets of
origin and destination tables were prepared for each time slice: one
for passenger vehicles and the other for buses. An example 0-D table,
for passenger vehicles during the first time slice, is shown in Figure
VI-A. All buses enter the freeway at the mainline origin and leave
These 0-D tables were developed from observed flows at each of the on-
and off-ramps and at the mainline start and end points. A computer
program was used to synthesize 0-D voluems consistent with these marginal
totals. For some FREQ applications, policy conclusions could be affected
by the accuracy of this 0-D approximation, in which case it may be desirable
to measure the actual flows more directly.
-------
DESTINATION NUMBER
n
M
25
1
2
3
4
5
6
7
8
9
10
11
1
228
0
0
0
0
0
0
0
0
0
0
2
253
14
0
0
0
0
0
0
0
0
0
3
270
15
35
0
0
0
0
0
0
0
0
4
67
4
9
13
0
0
0
0
0
0
0
5
87
5
11
17
15
0
0
0
0
0
0
6
268
15
35
51
46
0
0
0
0
0
0
7
19
1
2
4
3
0
0
0
0
0
0
8
86
5
11
16
15
5
0
0
0
0
0
9
36
2
5
7
6
2
7
0
0
0
0
10
57
3
7
11
10
3
10
9
0
0
0
11
31
2
4
6
5
2
6
5
14
0
0
12
50
3
7
9
8
3
9
8
23
0
0
13
28
2
• 4
5
5
2
5
4
13
10
0
14
88
5
12
17
15
5
16
13
40
32
0
15
284
17
37
55
49
17
52
43
130
104
64
I
00
FIGURE VI-4
Sample Origin-Destination Data for Eastshore Freeway - Time Slice One
-------
VI-9
the freeway at one of five off-ramps: Ashby, University, Central,
Carlson and the mainline destination. During the period 3:30 to 6:00
PM, a total of 103 buses use the freeway, with an average occupancy
of thirty-three persons.
The model also requires an occupancy distribution for passenger
vehicles at each ramp, as shown in Figure VI-5. The proportions of single
occupant autos and two-, and three-person carpools, as well as the propor-
tion of buses, are specified for each ramp. In addition, the average occu-
pancies for passenger vehicles carrying three or more persons and for
buses are specified for each on-ramp.
The arterial, like the freeway, is divided into twenty-seven
subsections. Each subsection corresponds to a freeway subsection and
requires three types of descriptive information:
• capacity in vehicles per hour;
• free-flow speed in miles per hour; and,
• a variable indicating whether or not there are signals on the
subsection, and if there are signals, whether the progression
is poor or good.
Figure VI-6 provides these data for San Pablo Avenue. For example,
subsection 1 has a capacity of 3630 vehicles per hour, no signals, and
a free-flow speed of 30 miles per hour. Subsection 2 has a capacity
of 1971 vehicles per hour, contains signals with good progression, and
has a free-flow speed of twenty-five miles per hour.
•^Arterial data for this case were derived from previous TRANSYT modelling
work. Typically, one would develop more approximate speed and capacity
data for such an arterial based on number of lanes, lane widths, turning
movements, and, if possible, more detailed field measurements. The accu-
racy requirements for these data are less stringent than for the freeway
itself.
-------
ON-
RAMP
1
2
3
A
5
6
7
8
9
10
11
OCCUPANCY
1
0.744
0.751
0.751
0.751
0.751
0.751
0.751
0.751
0.751
0.751
0.751
2
0.168
0.170
0.170
0.170
0.17Q
0.170
0.170
0.170
0.170
0.170
0.170
3 OR
MORE
0.078
0.079
0.079
0.079
0.079
0.079
0.079
0.079
0.079
0.079
0.079
BUSES
o.pio
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
CARPOOL
(3+)
OCCUPANCY
3".2
3.2
3.2
3.2
3.2
3.2
3.2
3.2
3.2
3.2
3.2
BUS
OCCUPANCY
33
0
0
0
0
0
0
0
0
0
0
FIGURE VI-5
Distribution of Passenger Occupancies on the Eastshore Freeway
-------
VI-11
Subsection
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Free-Flow Speed
30
25
25
25
25
27
20
28
23
26
26
26
20
20
26
23
29
30
30
24
24
27
25
27
30
30
30
Capacity
3630
1971
1971
1686
4918
2224
2681
2529
2155
2515
4300
2211
2325
2475
3331
2023
2000
1954
2722
5938
2524
1358
3055
2322
3367
2014
4400
Arterial
Type
NS
G
G
G
NS
G
G
G
G
G
NS
G
G
G
G
G
G
G
G
NS
G
G
G
G
G
G
NS
NOTE 1: NS = No Signals; G = Signalization with Good Progression; and P
Signalization with Poor Progression.
FIGURE VI-6
San Pablo Avenue Facility Data
-------
VI-12
The flow data for the arterial consist of estimates for each
subsection for each time slice. The flows are entered on a separate
set of cards. In this application, there are ten such sets of cards.
Field observations indicated an average occupancy of 1.2 persons per
vehicle for San Pablo Avenue, which was used in this model.
F. Description of Model Application and Impact Assessment
Detailed program documentation describes the input format
required by FREQ6PE, which will not be described here. The model
produces extensive, detailed output for each time slice and subsection,
including: flows, volume-capacity ratios, speeds, queue lengths,
fuel consumption, and emissions. Attention should be directed to the
congestion queue diagram and to the freeway and arterial summary tables.
Base Conditions
The queueing diagram of freeway congestion for the pre-control
situation is shown in Figure VI-7. The vertical scale is time and the
horizontal scale is distance along the freeway. Time moves up the
diagram and traffic moves from left to right. The diagram identifies
three separate bottlenecks in subsections 5, 20 and 25, and their
accompanying upstream congestion. Congestion first begins at the start
of time slice 3 (4:00 PM) and all congestion is over at the end of
time slice 9 (5:45 PM). The bottlenecks and accompanying queues are
separate from one another and there is a slight offset in the time
periods of congestion.
Jovanis, Yip, and May, op.cit.
-------
en
f
M
O
2!
W
10
9
8
7
6
5
4
3
2
1
— «?
citf?
3&
«*$
4jg
rf?5<
vx^
(XV
$?
$$
^
/
cxx
^
-«-ir
o
xVS
VS^
$v
^
vv
w
^
^
%$
^
^^^^
vV
$
$
$
$•
x/-
x>
y*
\
\
I
m
^
6:00
5* /. c
t*»5
5* in
: ju
5. •) c
" ij
5< nn
!UU
4. /. c
• ** J
A. 30
A. I C
• JL J
4* nn
3.45
3:30
23 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
h-•
u>
SUBSECTION NUMBER
FIGURE VI-7
Queueing Diagram of Freeway Congestion before Entry Control
-------
VI-14
The bottleneck at subsection 25 is a minor one, resulting in small
congestion delays. The congestion due to bottlenecks at subsections 5
and 20 are much more severe. The queueing diagram illustrates a rather
complex congestion pattern on the freeway. Just as congestion ends at
the upstream end, the queue due to the bottleneck at subsection 20 is
at its peak. Again, as congestion ends upstream of subsection 20, the
queue in subsections 23 and 24 is at its peak. We may expect this
congestion pattern to be reflected in the ramp metering plan and
associated traveller responses.
The freeway and arterial summary tables are shown in Figures VI-8
and VI-9. Summary data are presented for each time slice and for the
entire analysis period. This information serves as the base condition,
and all entry control results are compared with it. For example, in
Figure VI-8, the total freeway travel time is 3981 passenger hours,
while the arterial time is 1251 passenger hours. The total travel
distance is 184,297 passenger miles for the freeway and 28,323 passenger
miles for the arterial. The arterial thus carries 15 percent of the
corridor passenger miles, yet because of a lower level of service
accounts for 30 percent of corridor passenger hours.
Entry Control Analysis
The two entry control situations analyzed included ramp metering
without priority ramp lanes and ramp metering with priority ramp lanes.
Three analysis phases were undertaken for entry control without bypass
priority ramp lanes: analysis with no diversion, analysis with
-------
T.S.
1
2
3
4
5
6
7
8
9
10
Total
Where:
T.S.
FRWY
RAMP
TOT.
VH
PH
SPD
FUEL
11. C.
CO
FRWY T. T. RAMP DEL.
VH. PH VH PH
221 297 0 0
240 328 0 0
303 412 0 0
305 416 0 0
350 482 0 0
375 524 0 0
333 493 0 0
280 428 0 0
200 311 0 0
189 289 0 0
2798 3981 0 0
TOT. FRWY
VH
221
240
303
305
350
375
333
280
200
189
2798
- Time Slice Number.
T.T. » Freeway Travel Time, excluding
DEL. » Ramp Delay
FRWY T.T. - Total Freeway Travel Time,
•• Vehicle Hours
T.T
PH
297
328
412
416
482
524
493
428
311
289
3981
FRWY
VM
12122
13075
14282
14314
14265
14151
13437
12426
10815
10566
129454
any ramp delay.
including ramp
T.D.
PM
16284
17830
19441
19507
19631
19790
19891
18993
16799
16130
184297
delay.
SPD
MPH
54.8
54.4
47.2
46.9
40.7
37-8
40.3
44.4
54.0
55.8
47.6
FUEL
Gal.
742
797
849
852
867
868
842
749
670
653
7890
HO
KR
33
35
40
40
43
44
41
36
29
28
370
CO
KR
317
343
390
392
425
435
405
348
286
275
3617
NOX
Kg
60
64
62.
61
55
52
54
52
54
53
566
Begin
Time
3:30
3:45
4:00
4:15
4:30
4:45
5:00
5:15
5:30
5:45
NO • Nitrous Oxide emissions in kilograms.
A
Begin Time - the time at the beginning of
the slice.
- passenger hours
- Speed in miles per hour
• Fuel consumption in gallons
- Hydrocarbon emissions in kilograms
- Carbon Monoxide emissions in kilograms
FIGURE VI-8
Eastshore Freeway - Freeway Summary before Entry Control
-------
T.S.
1
2
3
4
5
6
7
8
9
10
Totals
ART. T.T.
VH
86
88
95
100
117
119
121
122
107
88
1043
PH
103
105
114
120
141
143
146
147
128
105
1251
ART. T,D.
VM
2017
2077
2200
2312
2618
2652
2651
2674
2351
2051
23602
PM
2420
2493
2639
2775
3142
3183
3181
3208
2821
2462
28323
SPD
MPH
24
24
23
23
22
22
22
22
22
23
23
FUEL
GAL.
188
193
204
215
243
246
246
248
218
191
2192
HC
Kg
8
8
9
9
11
11
11
11
10
8
94
CO
Kg
71
73
79
84
100
102
106
106
93
74
-*•». «
889
N°x
Kg
2
3
3
3
3
3
3
3
3
2
27
Begin
Time
3:30
3:45
4:00
4:15
4:30
4:45
5:00
5:15
5:30
5:45
Where:
T.S. = Time Slice Number.
ART. T.T.» Arterial Travel Time; VH - Vehicle Hours, PH = Passenger Hours.
ART. T.D.=* Arterial Travel Distance; VM *• Vehicle Miles, PM - Passenger Miles.
SPD = Speed in miles per hour.
FUEL « Fuel Consumption in gallons.
H.C. • • Hydrocarbon emissions in kilograms.
CO , •* Carbon Monoxide emissions in kilograms
NOX ( - Nitrous Oxide emissions in kilograms
DeRin Time - The time at the beginning of the time slice.
FIGURE VI-9
Eastshore Freeway - Arterial Summary Before Entry Control
-------
VI-1 ?
short-term diversion, and analysis with short- and longer-term
diversion. The "No Diversion" impacts are those that can be expected
to occur on the first day of the operation with a particular entry
control plan. The "Short-Term Analysis" takes place from two to four
weeks after the start of entry control. The "Longer-Term Analysis"
takes place approximately two to four months after the start of entry
control.
For the no-diversion case, the ramp metering plan and the resulting
ramp queue model, are shown in Figure VI-10. The metering plan calls
for seven ramps to be controlled for varying periods of time between
4:00 PM and 5:15 PM. It clearly reflects the congestion pattern
illustrated by Figure VI-7. The bottleneck in subsection 5 is relieved
by metering on-ramps 2 and 3 during time slices 3, 4 and 5. Equity is
reflected in this portion of the metering plan through control of on-ramp
2 despite the location of the bottleneck just downstream of on-ramp 3.
Tighter control at ramp 3 alone would have relieved the congestion with
lower total delay but at the cost of very high delays for travellers
entering the freeway at on-ramp 3.
The congestion caused by the bottlenecks at subsections 20 and 25
is relieved by metering ramps 4, 6, 7, 8 and 9 for varying times between
equity considerations of freeway entry control have been a source
of much discussion in recent years. Equity here refers to the more
equal sharing of ramp metering delays. Through this more equal sharing,
more people will pay a small amount in time penalty rather than have a
small number of users pay a very large amount (in increased ramp delays),
The benefits of entry control still accrue to those who get on the
freeway and are now able to travel congestion-free. The costs are
merely distributed to more freeway users.
-------
•1
4:00 - 4:15 PM
4
4:15 - 4:30 PM
e
4:30 - 4:45 PM
4:45 - 5:00 PM
5:00 - 5:15 PM
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles' (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
ON-RAMP NUMBERS
2
180
26
0
295
30
0
180
47
0
3
269
107
0
504
137
0
4
784
29
0
900
17
0
6
180
38
0
7
367
21
0
464
35
0
248
68
0
8
218
15
0
290
21
0
208
79
0
500
55
0
9
900
65
0
900
156
0
900
24
0
00
Legend; On Ramp 2 Is Powell Street
3 Is Ashby Avenue
4 Is University Avenue
On-Ratnp 6 Is Pierce Street
7 is Central Avenue
8 is Carlson Avenue
On-Ramp 9 is Cutting Boulevard
FIGURE VI-10
Eastshore Freeway Ramp Control Summary; No Bypass Priority Ramp Lanes, No Diversion
-------
VI-19
4:00 PM and 5:15 PM. Once again, control at on-ramp 9 would have been
sufficient to relieve congestion, but massive queues would have
accumulated. To avoid these queues, on-ramp 9 is only controlled
during time slices 5, 6 and 7, and ramps 4, 6, 7 and 8 are controlled
during other time periods. The control at on-ramp 4 during time slices
5 and 6 is an attempt to spread the queues resulting from control to
many on-ramps. On-ramp 5 is not metered at all because of heavy input
demand (approximately 1200 vphs). Any control at on-ramp 5 would have
resulted in very large queues and possible under-utilization of the
freeway. In order to prevent on-ramp 5 from being under control, a
"No Control" program option was exercised for this ramp from time slice
3 through time slice 8.
A comparison of corridor impacts for the no-diversion and base
cases is shown in Figure VI-11. There are varying effects depending
on the impacts under consideration. Small travel time savings (2%) are
counterbalanced by increases in fuel consumption and vehicle emissions.
Similar results for the short and longer term are shown in Figure VI-12.
These indicate that even after a longer-term shift in corridor travel
patterns, the net effect of undifferentiated ramp metering is to reduce
the average delay experienced by travellers, while increasing fuel
consumption and pollutant emissions. This is because moderate average
speeds under the congested base conditions are not necessarily worse
for auto efficiency than the higher speeds made possible by ramp meter-
ing. Also, ramp queues with idling vehicles would exacerbate the nega-
tive emissions and fuel consumption effects.
-------
VI-20
Travel Time
Fuel Consumption
Hydro- Carbons
Carbon-Monoxide
Nitrous Oxide
yeh-hr
pass-hr
gal
kg
fcg
kg
Differences
Freeway
-84
-123
71
5
83
50
Alt. Rte.
0
0
0
0
0
0
Total
-84
•123
71
5
83
50
% Change
-2
-2
1
1
2
8
Note: Differences are values after control minus values before control.
FIGURE VI-11
Eastshore Freeway Differential Effects: Entry Control
Without Bypass Priority Ramp Lanes, No Diversion
-------
VI-21
Travel Time
Fuel Consumption
HC
CO
NOX
veh-hr
pass-hr
gal.
kg
kg
kg
Percent
from
Short-term
Diversion
-2
-3
1
0
1
9
Change
the base case
Longer-term
Diversion
-1
-2
1
1
2
9
FIGURE VI-12
Effects of Short- and Longer-term Diversion for Ramp
Control Without Priority Entry
-------
VI-22
The second metering scheme permits priority entry at each ramp for
vehicles with three or more occupants. This implies a shoulder or
second lane at each on-ramp for use by priority vehicles to bypass
queued non-priority vehicles and gain unrestricted access to the free-
way. The revised metering scheme with no diversion is shown in Figure
VI-13. It is virtually identical to Figure VI-10. The impacts for
each state of diversion are shown on Figure VI-14. A fourth type of
diversion—mode shift—is included here. In the previous example, no
mode shift was likely because of the equal effects on all modes of the
metering scheme. However, priority entry yields further benefits for
transit and shared-ride users, at the expense of low-occupancy autos
and trucks. Thus, some kind of modal effect is certain. As Figure
VI-14 shows, the predicted modal effect is very small. In fact, the
overall difference in fuel consumption and emissions between the
non-priority and priority cases is very small. The travel time savings,
however, are much higher because it is high occupancy vehicles which
benefit from this strategy.
Several other results from this analysis are of particular note.
The priority entry scheme also was analyzed for 2+ occupant vehicles,
with further improvement in most of the impact measures (see Figure
VI-15). The same analysis was redone without consideration of equity
among ramps, yielding a marked improvement in all categories except
NOX emissions. Thus the practical need to distribute the negative
effects of ramp metering geographically appears to be significant con-
straint on environmental and energy benefits.
-------
TIME SLICE
A: 00 - A: 15 PM
A: 15 - A: 30 PM
A: 30 - A: 45 PM
A:A5 - 5:00 PM
5:00 - 5:15 PM
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles' (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
Metering Rate (vph)
Queued Vehicles (veh)
Diverted Vehicles (veh)
ON-RAMP NUMBERS
2
180
20
0
2A8
30
0
180'
A2
0
3
19A
112
0
AOA
141
0
4
653
44
0
826
34
0
356
123
0
6
180
31
0
180
47
0
180
63
0
7
357
15
0
180
90
0
447
65
0
420
77
0
8
180
28
0
270
62
0
270
87
0
180
121
0
9
835
9-
0
900
42
0
900
108
0
<
M
I
UJ
Legend: On Ramp 2 is Powell Street
3 is Ashby Avenue
4 is University Avenue
On-Ramp 6 is Pierce Street
7 is Central Avenue
8 is Carlson Avenue
On-Ramp 9 is Cutting Boulevard
FIGURE VI-13
Eastshore Freeway Ramp Control Summary: Priority Ramp Bypass for Vehicles with Three
Or More Occupants, No Diversion
-------
VI-24
Travel Time
Fuel Consumption
HC
CO
NOX
veh-hr
pass-hr
gal
kg
kg
kg
Percen!rom the base ci£tnge
no short-term longer-term
diversion diversion diversion
-1
-4
1
2
3
8
-3
-6
1
0
1
8
-3
-5
1
1
2
9
model
shift
-3
-5
1
0
1
9
FIGURE VI-14
Effects of Ramp Control with Priority Entry for
Vehicles with Three or More Occupants
-------
VI-25
Travel Time
Fuel Consumption
HC
CO
NOX
Veh - hr
pass - hr
gal
kg
kg
kg
Percent
from th
Equity
-A
-6
1
0
0
9
Change
e base case
No Equity
-6
-7
0
-2
-1
9
FIGURE VI-15
Effects of Ramp Control After Diversion and Model Shift With
Priority Entry for Vehicles with Two or More Occupants
-------
VI-26
G. Interpretation of Results
This entry control analysis and the FREQ6PE model in general must
be viewed in light of theoretical and empirical limitations, including:
• Vehicle emission estimates are based on the EPA modal emissions
model and reflect a vehicle fleet without current emission
control devices. Although this should not radically affect the
percentage changes, future applications should be preceded by
an update of the emissions models. This also is true of the
fuel consumption model.
• Parallel route traffic is computed assuming all traffic diverts
to a simulated single arterial. In practice, the diversion is
likely to occur on many routes, so the resulting arterial
congestion may be less than for a single route. The probable
implication is that the estimates of arterial impedence are high,
and actual costs under field conditions may be less.
• No buses are present at freeway on-ramps in the Eastshore Free-
way Corridor; therefore, only mode shift to carpools is simulated.
In other operating environments, where buses are present, larger
mode shifts may occur. Again, the implication is that the analysis
results may be slightly conservative.
• All entry control results assume no change in temporal or total
demand. The effect of temporal demand shifts due to entry
control are very difficult to estimate. As the total demand
level increases, however, it is possible that greater benefits
will accrue with priority entry. This hypothesis is supported
by sensitivity analyses.*
• All results are for entry control alone. The effects of combined
TSM tactics were not evaluated. Examples of combined tactics
are entry control with variable work hours, and/or entry control
with parking measures.
• The analysis did not consider potential accident and/or safety
improvements which may be significant.
• Enforcement and bypass lane construction costs were not considered
because of their highly site-specific nature.
Jovanis, et al.. op.cit.
-------
CASE STUDY VII
AREAWIDE ANALYSIS OF TRAFFIC ENGINEERING MEASURES
-------
CASE STUDY VII: AREAWIDE ANALYSIS OF
TRAFFIC ENGINEERING MEASURES1
A. Problem Presentation
The central sections of radial freeways in a medium-sized
Midwestern city experience recurring congestion during peak travel
periods. Most of the arterials in the central city have heavy traffic
flows throughout the day, with congestion becoming severe during the
peak periods in many locations. Although the area has a relatively
extensive network of interconnected traffic signals, during recent
years not enough attention has been given to maintaining timing plans
which reflect the current pattern of traffic flow. Consequently,
delays to traffic occur throughout the day because of inefficient
signal operation.
Air quality is a major concern in the area, as it experiences
some of the worst air pollution conditions in the nation. A majority
of the HC and CO pollutants in the region's air are attributable to
automobile travel. The local Metropolitan Planning Organization (MPO)
is in the process of evaluating transportation measures for inclusion
in the 1982 SIP revision submittal. A variety of measures are being
evaluated, including actions to improve traffic flow. There is concern
however, that these actions may encourage additional travel and, thus,
more severe air pollution (as well as increased fuel consumption).
Adapted from Wagner, Frederick A., Urban Transportation Energy Conser-
vation Vol. IV; Analysis of Traffic Engineering Actions, prepared under
subcontract to Cambridge Systematics, Inc. for the U.S. Department of
Energy, September 1978.
-------
VII-2
B. Proposed Transportation Measures
Three major types of traffic engineering action are being evaluated:
1. Freeway surveillance and control utilizing ramp metering
to manage the demand entering congested sections of free-
ways- This action applies only during peak periods and only
for congested zones of the freeway system.
2. Optimization of traffic signal timing to make the most
efficient use possible of the existing traffic signal sys-
tem. This action is a continuing effort requiring periodic
measurements of traffic patterns in the signalized network
and computation of corresponding optimum timing plans for
different periods of the day either through manual or com-
puterized techniques. This action is applicable to the
entire system of surface arterial highways, both within and
outside of congested zones.
3. Implementation of improved master control systems for the
signalized network. This usually entails providing real-
time computer control systems to allow a more flexible
range of timing plans and/or traffic-responsive control al-
gorithms which select or adjust signal timing based on
actual fluctuations of traffic patterns. Such systems are
most applicable within congested zones and on major access
routes to high-activity centers, rather than areawide.
Ultimately, such systems might be implemented throughout
the arterial system, but for short to medium term planning
-------
VII-3
purposes, the more limited application is more realistic
to achieve.
Each of these measures will improve the flow of traffic on the
highway sections to which they are applied, thereby decreasing average
travel times and per-mile fuel consumption and emissions rates for autos.
Each may tend to encourage additional auto travel in the region as well,
prompting concern about their efficacy in addressing environmental pro-
tection and energy conservation objectives. An analysis technique sensi-
tive to both the supply and demand impacts of the proposed measures is re-
qured in order to evaluate the degree to which they support efforts to
improve air quality in the region.
C. Selection of Analysis Technique
The manual areawide traffic engineering analysis method (3.1.3)
combined with the emissions worksheets (4.1) are chosen for application
to the evaluation of the proposed measures. Several aspects of the evalua-
tion point to the use of these techniques:
• Both supply and demand impacts must be considered. The
manual sketch planning technique includes estimation of
the increase in auto travel associated with improved
traffic flow as well as travel time and fuel consumption
impacts.
• The concepts to be evaluated are not developed in detail.
Specific locations or highway facilities have not been
identified as candidates for any of the proposed actions.
A general assessment of the areawide impact of the im-
plementation of these improvements where appropriate is
desired.
• An analysis of three distinct measures is required within
a limited budget. Based on the general assessment of the
effectiveness of the measures, more detailed studies may
be conducted later. An inexpensive approach, able to use
The section numbers following specified analysis methods refer to the
location of their description in Volume I.
-------
VII-4
available data is required, suggesting the use of a
manual technique.
The anual missions orksheets can interface directly
with the manual traffic engineering analysis technique
and yield accurate estimates of changes in HC, CO and
NOX emissions.
D. Overview of Analysis
The analysis is carried out in a well-defined sequence of steps
as shown below. As with the other manual techniques described in
this volume, a primary component of the analysis is market segmen-
tation. In this case, different highway facility types, or functional
classes, provide the basis for segmenting the areawide travel market,
as noted in Step 1. The impact of the proposed traffic engineering
actions is estimated separately for each highway functional class.
The results are then aggregated to develop estimates of the areawide
impact of each action. Work and non-work travel are also analyzed
separately, with the overall impact being determined by combining the
results of the separate analyses.
The steps involved in the areawide analysis of traffic engineering
measures are:
1. Segment the total highway network into relatively homo-
geneous functional classes characterized by varying values
of travel time per mile, and travel time elasticity.
2. Estimate the proportion of areawide VMT using each func-
tional highway class.
3. Estimate the average travel time per mile for each func-
tional highway class, and areawide average travel time.
4. For each proposed traffic engineering action, estimate the
fraction of each highway class affected by the action.
-------
VII-5
5. For each proposed traffic engineering action, estimate
the proportional shift in travel time caused by the
action.
6. For all actions combined, estimate the proportional shift
in travel time on each functional class of highway, the
resulting new travel time on each highway class, and the
new areawide travel time.
7. Compute the areawide proportional shift in travel time
caused by the combined traffic engineering actions.
8. Estimate the elasticity of travel time for each functional
highway class, and the areawide elasticity of travel time
to changes in VMT.
9. Estimate the areawide elasticity of VMT to changes
in travel time.
10. Estimate the proportional change in areawide VMT, at
the new equilibrium point resulting from traffic engineering
actions.
11. Estimate the proportional change in areawide travel time
at the new equilibrium point resulting from the traffic
engineering actions.
12. Estimate the elasticity of fuel consumption rate to changes
in travel time.
13. Estimate the proportional change in areawide fuel consumption
caused by the traffic engineering actions.
14. Compute the combined work and non-work impacts on VMT,
travel time, and fuel consumption. (Steps 1-13 are
repeated for work and non-work travel.)
15. Estimate the proportional change in auto emissions using
the manual emissions worksheets.
E. Input Data Development
Steps 1, 2 and 3 involve setting up the analysis and acquiring
base input data on the characteristics of auto travel in the urban
area. These steps are described in detail below.
Step 1: Segment the Total Highway Network Into Functional Classes
-------
VII-6
As described above, the concept of travel segmentation is
used in the manual traffic engineering analysis procedure.
Travel in the urban area is subdivided into the following five
functional classes of highway facilities:
1. Smooth flow sections of freeways
2. Congested or queued zones of freeways
3. Line-haul sections of arterial streets carrying mainly
through traffic where conditions result in higher
speed operation than on congested zone arterials
4. Congested zone sections of arterial streets where
physical, traffic control, land use intensity, and
traffic demand levels result in slow speed traffic
operation (mainly located in high-employment centers)
5. Local streets, off the major arterial system, which
mainly provide access to residential land uses
Segmentation of travel into these classifications aids in the
analysis of potential impacts of traffic engineering actions in
several ways:
• Determination of average travel time for each segment
permits a more systematic estimate of average travel
time for the total network either under existing con-
ditions or under modified conditions impacted by the
application of various traffic engineering actions.
• Segmentation of VMT into the proportions occurring
on each functional class provides a mechanism for
estimating what portion of VMT is affected by any
given engineering action.
• Specification of travel time elasticities to changes
in VMT for each functional classification permits
estimation of this elasticity for the whole network.
The degree to which travel in the area could be segmented
was controlled by the availability of data for the highway func-
tional classes which are identified by the segmentation scheme.
-------
VII-7
Steps 2 and 3 describe the data required for each functional
class of highway.
Step 2; Estimate the Proportion of Areawide Work and Non-Work
VMT Using Each Functional Highway Class, Pv-
With respect to VMT on freeway segments, it is assumed that
all freeway traffic operates in the smooth flow regime for non-
work travel, but that some portions of the freeway operation are
in the congested regime during work travel periods. Separation
of total freeway work trip travel, P , ~, into its two components
requires knowledge of the overall average freeway travel time per
mile, t1 _, plus assumed values of travel time for the smooth flow
and congested flow components, t, and t_. Based on data from
other cities, t1 =1.3 and t~ = 3.5 minutes per mile. Then:
P = P
v2 vl+2
3.5 -
P = P P
vl vl+2 v2
For the urban area under study, the U.S. DOT's 1974 National
Transportation Report, Urban Data Supplement, gives the fol-
lowing information:
Proportion of
Total VMT
Freeway 0.22
Arterial 0.71
Local 0.07
-------
VII-8
For work travel, the freeway portion of this data is further
subdivided:
1 „ = 1.45 minutes per mile1
t - 1.3
Pv2 - Pvl+2 ( —
3.5 - 1.3
= 0.22 ( i^ — ) = 0.015
2.2
Pvl = 0.22 - 0.015 = 0.205
The estimate of the portion of arterial VMT in congested zones
presented a more difficult problem since an areawide inventory
of travel time was needed to do this precisely. However, it
was hypothesized that a substantial portion of the congested
arterial sections are in CBD areas and the remainder are found
in a few other high-activity centers in the urban area.
Complete traffic count data was available for the CBD as
part of the ongoing traffic engineering program of the region's
central city. This data was used to estimate VMT in the CBD.
It was found that CBD VMT equaled approximately 2 percent of
the areawide total. Assuming that the CBD accounts for about
half the total areawide congested zone traffic, the estimate
of the portion of VMT in congested zones was raised to 4 percent
1 This value was obtained from UTPS highway network model validation
runs, conducted as a standard aspect of the urban area's ongoing
transportation planning process.
-------
VII-9
of arterial VMT or 4-0.71) = 3 percent of areawide total VMT.
The following breakdown of work and non-work VMT into the
five highway functional classes emerged from this analysis:
Work Non-Work
Functional Class, i PVJ PVJ
1. Smooth flow freeways 0.205 0.222
2. Congested zone
freeways 0.015 0
3. Line-haul arterials 0.68 0.68
4. Congested zone
arterials 0.03 0.03
5. Local streets 0.07 0.07
Step 3: Estimate the Average Travel Time per Mile for Each
Functional Highway Class, t., and Areawide, t
The t. values for the individual functional highway classes
were estimated as follows:
• For cases where all freeways are operating with
smooth flow, e.g. during non-work travel periods,
the following relation was assumed:*•
t]_ -1.1 + 0.2 £
where:
V_ _ the average volume to capacity ratio
C for all freeways
These assumptions are based on empirical evidence from a number of
cities. See Wagner, Frederick A., Urban Transportation Energy
Conservation, Volume IV;Analysis of Traffic Engineering Actions,
prepared under subcontract to Cambridge Systematics Inc. for the
U.S. Department of Energy, September 1978.
-------
VII-10
For congested (queued up) zones of freeways it
was assumed that
1
t2 = 3.5
As shown in Step 2, above, composite freeway
travel time during peak periods, t^+2 > ^s used
to estimate the proportion of VMT operating in
the smooth flow regime, PVI, and in the congested
flow regime, PV2- The composite freeway travel
time needed in this computation can be obtained
from available freeway performance data, from
new samples of travel time on representative sec-
tions of freeway, or by using freeway flow simu-
lation models to estimate existing conditions.
In this case, freeway flow simulation was used to
estimate t „.
For general arterials and congested zone arterials,
estimate 13 and t^ from available travel time data
or from new samples of travel time data collected
on a representative sample of routes. (Alterna-
tively, use street traffic flow simulation models
to estimate t3 and t^ for representative routes
or zones of the network).
For local streets, estimate t^ from available data
or from small samples in representative neighborhoods.
For this urban area, substantial peak period travel time data
were available. Table VTI-1 summarizes the analysis and inter-
pretation of the available data to obtain estimates of t for dif-
ferent functional highway classes:
The basic results are:
Functional Class, i
Work Travel
t,,
1 l'3} 1 45
2 3.5; •"-• D
3 2.83
4 4.11
5 3.0
1 See footnote on previous page.
-------
TABLE VII-1
Peak Period Travel Times
Functional Type Location
Congested Zone CBD
General Arterial South Corridor
East Corridor
West Corridor
Average for General Arterials:*
Freeways East Corridor
West Corridor
Central EB
Central WB
North Corridor
South Corridor
Average for Freeways:*
Average
Speed, Mph
14.6
21.0
19.2
22.8
21.2
30.5
43.5
48.4
48.4
36.6
47.6
41.5
Average Travel
Time, Min/Mile
4.11
2.86
3.13
2.63
2.83
1.97
1.38
1.24
1.24
1.64
1.26
1.45
Length
of Routes
—
5.0
33.2
48.4
86.6 Total
7.8
8.0
9.0
9.0
6.0
6.0
45.8 Total
M
*Average values weighted by the lengths of test routes.
-------
VII-12
Given the values of t^ and P . for individual functional
highway classes, areawide average travel time, t, is estimated
by:
t = Z P . t,
i vi i
Jor this urban area,
t = 2.58 minutes per mile, or 23.3 mph
The travel time characteristics for non-work travel are signifi-
cantly different than for work travel since the preponderance of
non-work trips are made outside of peak periods. No substantive
field data were available for off-peak travel, hence, it was
necessary to base non-work travel time estimates on the following
set of assumptions:
• The distribution of VMT by highway class for non-work
travel is the same as for work travel
• Negligible freeway congestion is encountered by non-
work travel. Therefore, average freeway travel time
was assumed to be 1.2 minutes per mile (50 mph) -
all in the smooth flow regime.
• For travel on local streets, it was assumed that work
and non-work travel times per mile are equal since
local streets are so sparsely loaded.
• For arterial streets, it was assumed that an average
elasticity of travel time of 0.17 could be used to
estimate the difference between non-work period and
work period travel times.* Assuming that hourly
traffic volumes during non-work trip periods are 40
percent lower than during work trip periods, then
average travel times should be approximately
40(0.17) = 7 percent lower for non-work trips.
See Wagner, op. cit. for a discussion of the basis for this assumption.
-------
VII-13
These assumptions yield the following estimates of t. for non-
work travel:
Non-Work Travel
Functional Class, i i
1 1.2
2
3 2.63
4 3.82
5 3.00
These data plus the P . values for non-work travel from Step 2
can then be used to calculate areawide average travel time, t,
for non-work travel. For this urban area,
t = 2.38 minutes per mile, or 25.2 mph
F. Description of Analytical Procedure
Steps 4 through 14 involve the application of the manual traffic
engineering analysis method. These steps are described below.
Step 4: For Each Proposed Traffic Engineering Action j, Estimate
the Fraction of Each Functional Highway Class Affected by the
Action F
This step defines where in the highway network an action
is implemented, or in other words, what portion of VMT within
an individual highway class is affected by a particular action.
This permits an accurate representation of actions that are de-
signed for specific subareas or corridors, as well as those which
are applied throughout the area on all highways of a given func-
tional class. The three proposed traffic engineering actions
-------
VII-14
were assumed to apply to the highway network as follows:
1. Freeway surveillance and control was assumed to be
implemented on all congested zones of the freeway
system. Therefore, the incidence of this action is
represented by setting F£I = 1.00.
2. Optimization of traffic signal timing is assumed to
be undertaken for the entire system of surface ar-
terials both within and outside of congested zones
(i.e., for functional highway classes 3 and 4). The
incidence of this action is represented by setting
F32 = 1.00 and F42 = 1.00.
3. Implementation of improved master control systems for
the signalized network was assumed only within con-
gested zones and in other high activity centers ini-
tially. The incidence of this action is represented
in approximate form by setting: F^ ~ 0.5 and F43 = 1.00.
In other words, the action is conservatively assumed to
apply to all congested zone arterial traffic and to half
the remaining arterial VMT.
Step 5: For Each Proposed Traffic Engineering Action j, Estimate
the Proportional Shift in Average Travel Time on Affected Locations,
T±j, Caused by the Action, PSt..
This impact is defined as the shift in average travel time
that would occur under given fixed levels of VMT on those portions
of the highway network where an action is applied. Estimates of
PSt.. may be made based on observed impacts resulting from the same
type of action implemented in other places (e.g., either other
locations in the same urban area or in some other similar urban
areas). Tables VII-2, VII-3, and VII-4, for example, summarize the
impacts of traffic signal system projects and freeway traffic control
projects in several cities, and provide estimates of the order of
magnitude of PSti. for these actions.
-------
VII-15
Table VTI-2 summarizes impacts of traffic signal timing
efforts in several cities encompassing a wide range of types of
surface arterial highways. The impact averages approximately 12
percent - a figure used for this analysis. Alternatively, recent
retiming projects could have been used as examples of the potential
of this action in the urban area if accurate before and after data
on travel speeds were available.
Table VII-3 summarizes the impacts of various projects in
which the master control system for traffic singals was upgraded.
The size of the impact is strongly dependent on the base condition
of the traffic control system. Data in this figure are organized
into four sets which vary with respect to the base conditions.
The average impact for each set is summarized below:
PS
Base Condition
1. Fully or Partially Non-Interconnected
Signals with Old Timing Plans -0,32
2. Interconnected Signals with Single
Dial Master Control and Old Timing
Plans -0.17
3. Non-Interconnected Signals, with
Traffic Activated Control of Indivi-
dual Signals -0.12
4. Interconnected Signals with Three
Dial Master Control and Actively
Managed Timing -0.06
The base conditions in the metropolitan area under study are
fairly well advanced and are most accuratedly represented by 4,
above. Therefore, the impacts of improved master control of traf-
fic signals is estimated at pSt.. = 0.06, a 6 percent improvement
-------
VII-16
TABLE VII-2
Traffic Signal Timing Optimization Impacts
Location
Toronto Central Area
Toronto Suburban Area
San Jose - CSD
Los Angeles - Inner City
(Broacvay - Figueroa)
Los Angeles - Inner City
(rice Boulevard)
i Los Aageles - Inner Citv
(Wilshire Boulevard)
Kacon, Georgia CBD
l
Inglewood, California
Citywide
i
Montgoaery, Alaba=a CBD
j
!
Charlotte, NC CSD Fringe
Washington, DC CBD
K\aber of
latersectlons
68
51
46
26
6
45
54
60
50
10
40
Study Method
Field Measurement
Field Measurement
TRAKS Siaulation
TRANS Simulation
TRAKS Sinulation
TRAKS Siaulation
SICO? Siculation
SIGOP Sirulation
TRAKSYT
Simulation
TRANS Simulation
UTCS-1 CKETSIM)
Tine of Day
7-9
10-12
1-3
4-6
7-9
10-12
1-3
4-6
4-6
3-i, 5:30-6
4-5:30
2:30-3:30
4:30-5:30
AX Peak
7:45-8:45
4:45-5:45
7-10
3-6
AM Peak
Off Peak
PM Peak
5-6
Off Peak
Averaee Speed, Men
Before
15.8
17.1
15.5
13.7
21.3
28.2
27.5
21.7
15.4
17.4
15.4
21.1
20.2
13.1
12.7
11.7
22.9
22.0
16.31
19.09
17.94
7.68
11.97
After
16.5
17.5
16.3
13.7
21.0
29.1
28.1
20.9
15.7
'20.6
18.9
24.9
21.5
14.4
14.4
13.7
30.9
30.0
20.24
20.26
19.87
8.66
13.22
Percent
Change
+ 4.4
+ 2.3
+ 5.2
0
- 1.4
+ 3.2
+ 2.2
- 3.7
+ 1.9
+21.1
+22.7
+18.0
+ 6.4
+ 9.9
+13.4
+17.1
+35.0
+36.0
+24.1
+ 6.1
+10.8
+25. S
+10.4
Average, All Locations: +11.8
Average, Toronto & San Jose + l.fc
Average, All Others: +18.4
Note: Toronto and San Jose were aggressively ear.aged, coaputerized signal systems in the BZFOS.E case
Source: 'Wagner, op. cit.
-------
TABLE VII-3
Impacts of Traffic Signal Master Control Improvements
Nature of Master Control System Improvement
Location
Time of Day
Percent Improvement
In Speed Or Travel
Time
Fully or Partially Non-Interconnected Signals
with Old Timing Plans
Versus
Interconnected Signals, Advanced Master
Control, and Optimized Signal Timing
Interconnected Signals, Single Dial Master
Control, Old Timing Plans
Versus
Interconnected Signals, Computerized Master
Control, Optimized Timing
Columbus, Georgia CBD
New York City:
Northern Boulevard
Hillside Avenue
Union Turnpike
Roosevelt Avenue
Thunder Bay, Ontario
CBP
Group Average
Charleston, SC
CBD
Raleigh, NC
CBD
Raleigh, NC
Arterials
*San Jose, CoSts Rica
CBD
Group Average
Pre-AM Peak
AM Peak
Off Peak
Pre-PM Peak
PMPeak
Peak
Peak
Peak
Peak
AM Peak
Off Peak
PMPeak
AM Peak
AM Off Peak
PM Off Peak
PMPeak
AM Peak
Off Peak
PMPeak
AM Peak
Off Peak
Pm Peak
All times of
Day
49
34
40
22
32
34
39
26
32
19
20
8
32
12
16
20
21
6
12
12
30
10
39
13
17
-------
TABLE VII-3 (continued)
Percent Improvement
In Speed Or Travel
Nature of Master Control System Improvement
Non-Interconnected Signals with Traffic
Actuated Controllers
Versus
Interconnected Signals, Computerized Master
Control, and Optimized Signal Timing
Interconnected Signals, Three or More Dial
Control (Minimum of Three Timing Plans),
Actively Managed Timing
Versus
Interconnected Signals, Computerized
Master Control, and Optimized Timing Plans
*Eslimated by traffic simulation methods.
Source: Wagner, op. cit.
Location
West London, England
Glasgow, Scotland
Group Average
San Jose, CA - CBD
Wichita Falls, TX
Washington, DC - CBD
*San Diego, CA - CBD
Stockholm, Sweden
*San Jose, Costa Rica
Group Average
Time of Day
Peak
AM Peak
Mid-day
PM Peak
Peak
Peak
AM Peak
Off Peak
PM Peak
7 Hrs., Peak
and Off Peak
Peak
All Times of
Day
Time
9
n
9
18
12
5
10
5
6
8
2
7
8
6
00
-------
VII-19
in mean travel time.
Table VII-4 summarizes the impacts on average
speed of major freeway traffic surveillance and control projects
in six cities. The average improvement in travel speeds for free-
flow and congested zones combined is approximately 20 percent.
Table VII-5 translates these results into estimates of the re-
duction in the proportion of the freeway which is congested. The
data indicate that surveillance and control can eliminate approxi-
mately 60 percent of the congested freeway zones. In other words,
of all freeway sections currently experiencing congestion (i.e.,
average travel time of 3.5 minutes per mile), approximately 60
percent could be upgraded to smooth-flow status (i.e., average
travel time of 1.3 minutes per mile). Thus, the proportional
shift in travel time for all currently congested freeway zones is:
PSt.. = 0.60 ( — — ) = -0.38 .
1J 3.5
Alternatively, instead of relying on estimates of impacts from
other locations, traffic simulation models could be used to develop
location specific impacts. Models are available for both signalized
arterial street situations and for freeway situations. The former
type of model (e.g., the UTCS-1 model and the TRANSYT model) can be
used to estimate impacts of traffic signal timing optimization and
master control system improvement. The latter type (e.g., the FREQ
model) can be ussd to estimate the impacts of freeway ramp control
systems.
-------
VII-20
TABLE VII-4
Freeway Ramp Control System Impacts on Average Speed
Location
Minneapolis I-35W
Northbound
(Inbound)
Southbound
(Outbound)
Chicago, Eisenhouwer
Expressway
Eastbound
(Inbound)
Los Angeles, Santa
Monica Freeway
Eastbound
(Inbound)
Houston, Gulf Freeway
Northbound
(Icbound)
Los Angeles Harbor
Freeway
Southbound
(Outbound)
Detroit, Lodge Freeway
Northbound
(Outbound)
Length,
Miles
16.6
16.6
12.7
12.7
9. A
r 9.4
13.5
6
4
4
6
Time of Day
7:15-8:15
6:30-9:00
4:30-5:30
3:3—6:30
2 hour
AM Peak
4 Hour
AM Peak
6:30-9:30
7:00-8:00
3:45-6:15
3:45-6:15
2:30-6:30
Averages All Data
Averages for Data
Including Ranp Delays
Before
Ramp
Control
33.8
43.9
33.7
38.5
30.3
37.7
36.2
20.4
25.9
25.9
27.3
32.8
34.2
After
Ramp
Control
45.5
50.1
40.1
45.7
33.0
39.7
50.6
32.6
40.3
40.3
36.4
41.4
Percent
Difference
vs.
Before
34
14
19
19
9
5
40
60
55
55
33
26
After,
Including
Ramp
Delays
43.0
48.5
38.6
44.4
41.4
37.4
37.4
32.6
40.8
Percent
Difference
vs.
Before
27
10
15
15
14
44
44
19
19
Source: Wagner, op. cit,
-------
VII-21
TABLE VII-5
Freeway Ramp Control System Impacts on Proportion
of Freeway Minute-Miles Congested
Location
Minneapolis, I-35W
Northbound
(Inbound)
Southbound
(Outbound)
Chicago, Eisenhower
Expressway
Eastbound
(Inbound)
Los Angeles, Santa
Montica Freeway
Eastbound
(Inbound)
Houston, Gulf
Freeway
Northbound
Los Angeles, Harbor
Freeway
Southbound
(Outbound)
Detroit, Lodge Freeway
Northbound
(Outbound)
Length ,
Miles
16.6
16.6
12.7
12.7
9.4
9.4
13.5
6
4
6
6
Time of Day
7:15 - 8:15
6:30 - 9:00
4:30 - 5:30
3:30 - 6:30
2 Hr. AM Peak
4 Hr. AM Peak
6:30 - 9:30
7:00 - 8:00
3:45 - 6:15
2:30 - 6:30
2:30 - 6:30
1
Averages
Proportion Congested
Before
Ramp
Control
.25
.07
.25
.16
.34
.17
.20
.76
.49
.43
.43
.31
After
Ramp
Control
.05
0
.13
.05
.27
.135
0
.28
.13
.20
.20
.12
Percent
Reduction
80
100
48
69
21
21
100
63
73
53
53
61
Source: Wagner, op. cit.
-------
VII-22
Step 6: For All Actions Combined, Estimate the Proportional Shift
in Travel Time on Each Functional Class of Highway, PSt,, the
Resulting New Travel Times of Each Highway Class, tn., and the
New Areawide Average Travel Time, tn
These estimates are calculated as follows: first, for
each functional highway class, the impact of the combined actions
is,
PSt± - [(1+ F±1 PStil)(l + F±2 PSti2)...(l
and
For the areawide network, the new average travel time is
t = Z P t ,
n I v ni
The intermediate results for steps (1) through (6) are summarized
in Table VII-6 which represents work travel characteristics in
the urban area.
Step 7; Compute the Areawide Proportional Shift in Average Travel
Time Caused by the Combined Traffic Engineering Actions, PSt
This is the proportional shift in travel time that would occur
if VMT were held constant; in other words, the proportional shift in
the supply curve. A simple proportion change equation is applied to
the original travel time, t, and the new travel time, tn, as
follows:
-------
VII-23
TABLE VII-6
Worksheet for Steps 1-6, Work Travel
Step(l) Step (2) Step(3) Step(4) Step(5) Step(6)
Functional
Class,! P t. Action F.. PS. PS t
ni
1. Smooth .205 1.3 1-3
flow
freeway
2. Congested 0.015 3.5 1 1.00 -0.38 -0.38 2.17
zone
freeway
3. General 0.68 2.83 2 1.00 -0.12
arterial 3 0.50 -0.06 -0.146 2.42
4. Congested 0.03 4.11 2 1.00 -0.12 -0.173 3.40
zone 3 1.00 -0.06
5. Local 0.07 3.00 3.00
street
t-2.58
-------
VII-24
t - t
In this case,
2.26 - 2.58
PS
2.58
Step 8: Estimate the Elasticity of Travel Time to Changes in
VMT for Each Functional Highway Class, etj, and for the Area-
wide Highway Network, et
The value e represents the degree to which travel speeds
or travel time rates degrade with additional traffic volume.
Guidelines for typical et. values observed in practice are
summarized below:1
• For non-work travel (off-peak periods)
Freeways - Assume all sections of freeway
operate in the smooth flow regime.
where V/C = the average volume to capacity ratio
during off peak periods. A reasonable V/C = 0.5,
so et = 0.08.
General arterials and arterials in congested zones
et3 = 0.17
et4 = 0.17
Local streets
Since traffic volumes are very light.
1 For a detailed discussion of the development of et. estimates, see
Wagner, op.cit., Chapter II. 1
-------
VII-25
For work travel (peak periods)
- Freeways - Assume at least some portion of the
freeway system is congested. Then eti+2, the
travel time elasticity for all freeway sections
combined, ranges from approximately 2 to A, de-
pending on the severity of existing congestion,
as shown below.
Average Freeway Travel Time for All Sections,
Range of
h+2
1.37 - 1.52
1.52 - 1.78
1.78 - 2.17
Midpoint of
'1+2
1.445
1.65
1.95
Elasticity of
Travel Time,
etl+2
2.12
3.39
4.43
In this urban area, since
General arterials
et3 = 0.17
- Congested zone arterials
et, =1.37 (observed range of et,
from 0.56 to 2.82)
Local streets
The estimates of travel time elasticities for individual highway
classes are summarized in Table VII-7.
Alternatively, freeway and street traffic simulation models
can be used to estimate et. for specific highway situations.
-------
VII-26
TABLE VII-7
Travel Time Elasticities
Fvmctional Class, i
1. Smooth flow freeways
2. Congested zone free-
ways
3. General arterials
4. Congested zone
arterials
5. Local Streets
Areawide
Work Non-Work
P'i
0.22 2.12 0.08
0.68 0.17 0.17
0.03 1.00 0.17
0.07 0 0
et = 0.612 et = 0.14
-------
VII-27
Simulation models are powerful in this regard since VMT can be
varied at will by specific amounts, upward or downward, holding
all other variables constant, and et. computed from the simula-
tion output.
After estimating et. for individual highway segments, the
areawide et can be computed by
et = ZPV. et±
Notice that the two freeway segments are combined in this step,
i.e.
e=P e + P e + P + P p
et vl+2 tl+2 + v3 t3 v4 v5 et5
The computed e values are also shown in Table VII-7.
Step 9; Estimate the Areawide Elasticity of VMT to Changes in
Travel Time, ev
The value ev represents the sensitivity of the volume of
auto travel to changes in travel time rate or speed. Several
ongoing efforts are aimed at estimating reliable values for e .
In this study, previous applications of the disaggregate travel
demand forecasting models described in Volume I were used to
estimate short-range values of ev, and yielded the following
results:l
Work travel, ev = -0.01
Non-work travel, ev = -0.15
1 See Wagner, op. cit., Chapter II.
-------
VII-28
The short-range VMT elasticities come about as a result of
work trip mode shifts, whereas non-work trip VMT elasticities
result primarily from modified trip frequency and trip length.
Longer-range VMT elasticities are known to be higher than
the short-range values since they are affected additionally by
changes in land use and choice of residential location. Since
a whole host of non-transportation forces, which were outside the
scope of this study, significantly influence these longer range
responses, sensitivity tests using various higher values of ev
were used to investigate the possible range of longer term im-
pacts. These sensitivity tests are not described here, however.
Step 10: Estimate the Proportional Change in Areawide VMT at
New Equilibrium Point Resulting from the Traffic Engineering
Actions, PCV
This computation accounts for the interaction between supply
and demand for highway facilities. Improving the level of ser-
vice or travel speed on the highway network (supply) will lead to
additional auto travel which offsets, to some degree, the improve-
ment in highway level of service. The demand for and supply of
highway travel (VMT and travel time rate) will eventually reach a
new equilibrium point. The equilibrium change in VMT, relative to
the base condition is given by :
e
PC = PS ( -
v t
1 - e e
v t
-------
VII-29
Results for this urban area are given on worksheets in Tables
VII-8 and VII-9, and are summarized in Table VII-10.
Step 11: Estimate the Proportional Change in Areawide Average
Travel Time at the New Equilibrium Point Resulting from the
Traffic Engineering Actions, PCf
This also accounts for supply-demand interaction and is
1
computed by :
PC = PS. ( )
L t 4
1 - e e
v t
Results for this computation also are shown on worksheets in
Tables VII-8, VII-9, and VII-10.
Step 12: Estimate the Elasticity of Fuel Consumption Rate Per
Mile to Changes in Travel Time, e-p
Research shows that ep is a function of the base condition
areawide travel time, t. For the 1976 VMT weighted mix of vehicles
in the U.S. fleet: 1
0.0201t
0.085 - 0.201t
In this case,
• Non-work travel, t = 2.38, e_, = 0.360
r
• Work travel, t - 2.58, ep = 0.379
See Wagner, op. cit. for derivations of these relationships
-------
VII-30
TABLE VII-8
Worksheet for Steps 1-13. Work Travel
Step(l)
Functional
Class, i
1.
2.
3.
4.
5.
Step (7)
Step (8)
Step (9)
Step (10)
Step (11)
Step (12)
Step (13)
Step(2)
P
Vi
0.205
0.015
0.68
0.03
0..07
t = 2.58
PSt =
et =
ev =
PC
v
pct =
eF "
- PCF =
Step(3) Step(4) Step(5) Step(6) Step(8)
t. Action F.. PSt^ PStj
1.3
3.5 1 1.00 -0.38 -0.38
2.83 2 1.00 -0.12 -0.146
3 0.50 -0.06
4.22 2 1.00 -0.121 -0.173
3 1.00 -0.06)
3.00
'*
*n-t = 2.26-2.58 = -0.124
t 2.58
2?V.e = 0.612
• * c •
i i
-0.01
A
x~^w . \» } ~* v« A. £^L \« M« //.io\^
t l-e e. 1 + .01 (.612;
v t
,^/l\ «^-.^/ 1 \
r"t(l-e e} 0.1-4(j + >01 {<612))
v t
0.020H 0.0201 (2.58)
0.085 + 0.020H = 0.085 + 0.0201 (2.58)
s %
1.3
2.12
2.17
2.42 0.17
3.40 1.00
3.00 0
= 2.26
-0.0012
-0.123
= 0.379
(1 + PC ) (1 + e.,PC J - 1 = (1.0012) (0.9534 - 1) = -0.04
-------
VII-31
TABLE VII-9
Worksheet for Steps 1-13, Non-Work Travel
Step(l)
Step(2) Step{3) Step(4) Step(5) Step(6) Step(8)
Functional
Class, i P t. Action F.. PS PS t e
1.
2.
3.
4.
5.
Step (7)
Step (8)
Step (9)
Step (10)
Step (11)
Step (13)
0.22 1.20 1.20j 0.08
0 - 1 X - J
0.68 2.63 2 1.00 -0.12J -0.146 2.25 0.17
3 0.50 -0.06)
0.03 3.82 2 1.00 -0.12) _0.173 3.16 0.17
3 1.00 -0.06)
0.07 3.00 3.00 0
t = 2.38 t = 2.10
Pr *» ~ * 2.10 - 2.38 Q ..g
t t 2.38 •
et = ZPV.et = 0.14
ey = -1.015
PC - r- ( % V one/ -o.is K 001-
PCv P-t ^l-evej * °'118 (^H.15(.14)j °'017
/ 1 \ f I \
TT* — T>1 1 ..-,,.. 1 _fi 11R !»».. .— ... - 1 — -0 11*1
PC - P^ I J - 0.118 (Kil5(.14)) 0.115
\ v t/ \ >
0.020H 0.0201(2.38) ,n
T" 0.085 + 0.020H 0.085 + 0.0201(2.38) U'JDU
PC- = (1 +PC )(1 + e_PCj - 1 = (1.017) (.9586) = -0.025
t V f t
-------
VII-32
TABLE VII-10
Summary of Impacts of Combined Traffic Engineering Actions
Proportion Change in Areawide
VMT Travel Time Fuel Consumption
PC.. PC. PC,
Work Travel +0.0012 -0.123 -0.045
Non-Work Travel +0.017 -0.115 -0.025
Work + Non-Work +0.0112 -0.1180 -0.0324
NOTE: Traffic engineering actions tested are comprehensive combinations
of (1) freeway surveillance and control, (2) traffic signal timing
optimization, and (3) improvement of master traffic signal control
system.
-------
VII-33
Step 13; Estimate the Proportional Change in Areawide Fuel
Consumption Resulting from the Traffic Engineering Actions, PCp
This is computed by:
PCF = (1 + PCy)(l + ep PCt) - 1
Sample calculations for Steps 1 through 13 for both work travel
and non-work travel are presented in Tables VII-8 and VII-9.
The findings are summarized in Table VII-10.
Step 14; Compute the Combined Work Trip and Non-Work Trip Impacts
on Areawide VMT, Average Travel Time, and Fuel Consumption
If Fw is the fraction of total travel that is work travel,
and the subscripts "w" and "nw" are used to depict work and non-
work impacts, then,
PC
v
= F PC.. + (1 - F ) PCV
w vw „' vr
PC = F PC,- + (1 - F ) PCt
t w cw w cnw
PCF = Fw PCFw + (1 * Fw)
In this case, FW = .37.
The results for work and non-work travel combined are shown
in Table VII-10.
G. Air Quality Impact Assessment
Two factors associated with the comprehensive traffic engineering
improvement program will influence auto emissions in the urban area:
-------
VII-34
1. The increased average travel speeds will lower the average
HC and CO emission rates and rasie the NOx emission rate.
2. The increase in auto VMT will tend to offset the effect of
lower HC and CO emission rates and will further increase
the total NOx emissions in the region.
The emissions worksheets (4.1) may be used to examine the relative impact
of each of these factors and to estimate the net change in total auto
emissions associated with the proposed traffic engineering actions.
Because areawide traffic engineering procedure used in this analysis
calculates only the proportional change in VMT, rather than the absolute
values of the base case and revised total VMT, an independent estimate of
base case VMT and the total number of trips (or average trip length) must
be obtained for work and non-work travel in the region. Previous travel
demand modelling work, based on a 1971 home interview survey, provided
estimates of the necessary data for 1980 and 1985. Linear interpolation
was used to develop estimates for the 1982 analysis year assumed in this
evaluation. Lengths for each type of trip were not necessary because
both VMT and total vehicle trip estiamtes were available from the previous
modelling work. The base case VMT and trip volume estimates are entered
in worksheet VI-A.
The revised case VMT and trip volume may be calculated separately
for work and non-work trips by applying the proportional changes in VMT
shown in Table VII-10 to the base case VMT and trip volume values.
For example, the revised case work VMT is given by:
It is assumed that change in the number of trips in the urban area
is proportional to the change in VMT.
-------
VI-A. INPUT TRAVEL DATA SUMMARY FOR EMISSIONS
ESTIMATION
Population
Subgroup
All
V~* —
Work VMT
(I or IV) 1
5,036,000
5,036,000
Average
Work Trip
Distance
(miles) (I)
-
D
Number of
One-Way
Work Trips
336,900
336,900
TOTAL WORX VMT
TOTAL WORK TRIPS
I
I xI Base Alternative
| [Revised Alternative
Forecast Year:
Policy: Areawide Comprehensive Traffic Engineering
E=
Non-Work
VMT
(I or IV)
8,674,000
8,674,000
Average
Non-Work
Trip
Distance
(miles) (I)
-
£
Non-Work
Trips
1,373,000
1,373,000
U)
U1
TOTAL NON-WORK VMT
TOTAL NON-WORK TRIPS
I
Source Worksheets are indicated in parentheses where applicable.
VMT and trips on worksheets I and IV in Appendix A must be multiplied
by the number of households per population subgroup.
1
1,709,900
TOTAL TRIPS
-------
VI-A. INPUT TRAVEL DATA SUMMARY FOR EMISSIONS
ESTIMATION
Population
Subgroup
All
^r-» _
Work VMT
(I or IV) 1
5,030,000
5,030,000
Average
Work Trip
Distance
(miles) (I)
-
Number of
One-Way
Work Trips
336,500
336,500
TOTAL WORK VMT
TOTAL WORK TRIPS
I
JBase Alternative
jRevised Alternative
Forecast Year:
Policy: Areawide Comprehensive Traffic Engineering
Non-Work
VMT
(I or IV)
8,529,000
8,529,000
*
Average
Non-Work
Trip
Distance
(miles) (I)
-
£
Non-Work
Trips
1,350,000
1,350,000
I
w
TOTAL NON-WORK VMT
TOTAL NON-WORK TRIPS
I
Source Worksheets are indicated in parentheses where applicable.
VMT and trips on worksheets I and IV in Appendix A must be multiplied
by the number of households per population subgroup.
1
1,686,500
TOTAL TRIPS
-------
VII-37
Revised VMT = Base VMT x (1 + PC (work))
= 5,030,000 (1 + .0012)
= 5,036,000
The revised case VMT and trip volume estimates are entered on a separate
copy of worksheet VI-A.
Worksheet VI-B is used to record the base and revised case per-
centage of cold starts for work and non-work trips. No parking dura-
tion data was available for work or non-work travel in the region, so
estimated figures for cold start trip percentages were obtained from
Table D.I in Volume I.
Worksheet VI-C is then executed twice (once each for the base and
revised cases) to determine areawide start up vehicle emissions. Work-
sheets VI-A and VI-B and emissions Tables D.2 through D.5 provide the
necessary data for the calculations. A 40-degree ambient temperature
is assumed since the region experiences its highest air pollution concen-
trations during the winter months.
Travel (VMT) related emissions for the base and revised cases are
calculated using worksheet VI-D. Both the VMT change and the change in
average travel speed associated with the traffic engineering actions are
considered in the calculations. The base case work and non-work average
travel time rates, were calculated in step 6 of the manual traffic en-
gineering analysis procedure. These travel time rates may be converted
to travel speeds with the following formula:
-------
VI-B. COLD START FRACTIONS
1. If Daily Parking Duration Data are Available:
_XJ Base Alternative
Xj Revised Alternative
Policy: Areawide Traffic Engineering
Subgroup :
Forecast Year: 1982
% of Auto Trips by % of Auto Trips with % of Auto Trips by % of Auto Trips with
Catalyst-Equipped Parking Duration Non-Catalyst Equip- Parking Duration % of Auto Trips
Vehicles (Table D.I) 1 hour ped Vehicles 4 hours with Cold Starts
i
u>
oo
2. If Daily Parking Duration Data are not Available
% of Work Trip Cold Starts (Table D.2)
90
Z Non-Work Trip Cold Starts (Table D.2)
57
-------
VI-C. AUTO START-UP AND liVAPORATIVE EMISSIONS
Base Alternative
II Revised Alternative
Policy: Areawide Traffic Engineering
Forecast Year:
1982
Temperature:.
Work Trips
Non-Work Trips
(1)
Population
Subgroup
All
(2)
Z Cold
Starts.
(VI-B)1
90
(3)
Trips
(IV-A)
336,500
u
a
HCM
*-4 «J
£ S
WC(c)
C0(c)
10x(c
HC(h)
1C(c)
C0(c)
N0x(c
HC(h)
IC(c)
C0(c)
10x(c
IC(h)
• Is
8-8
h. H
D.2
D.3
0.4
0.6
D.2
0.3
0.4
0.6
D.2
D.3
D.4
D.6
TOTALS TOTALS
(4)
Start-Up
Factors
14.6
245.0
3.5
6.0
HC
/ j CO
Subgroups NOx
(5)
Emissions "
Col. 3 X Col. 4
(grams)
4,913,000
82,554,000
1,178.000
2,019,000
6,943,000
82,443,000
1,178,000
(6)
Z Cold
Starts
(VI-B)
57
(7)
Trips
(VI-A)
1,350,000
1
(8)
Start-Up
Factors
10.31
159.0
3.1
6.0
HC
^ ^ CO
f J
Subgroups NOx
(9)
Emissions •
Col. 3 X Col. «
(grams)
13,919,000
214.650.000
4.185.000
8,100,000
22.019rOOO
214,650,000
4,185,000
Source Worksheets are Indicated In parentheses
where applicable
(c) Indicates cold start factor
(h) Indicates hot soak factor
both work and non-work start-up factors
obtained from the Indicated tables
Work Trip Start-
Up Emissions
(grams)
Non-Work Trip
Start-Up
Emissions
(grams)
I
10
Total Start-Up
Emissions
(grams)
-------
VI-C. AUTO START-UP AND EVAPORATIVE EMISSIONS
Base Alternative
Revised Alternative
Policy: Areawide Traffic Engineering
Forecast Year:
1982
Temperature:
Work Trips
Non-Work Trips
Source Worksheets are Indicated In parentheses
where applicable
(c) Indicates cold start factor
(h) Indicates hot soak factor
both work and non-work start-up factors
obtained from the Indicated tables
Work Trip Start-
Up Emissions
(grams)
/ N
(1)
Population
Subgroup
All
TOTALS
(2)
Z Cold
Starts.
(VI-B)1
90
(3)
Trips
(1V-A)
336,900
Pollut-
ant2
lC(c)
C0(c)
N0x(c
lC(h)
HC(c)
C0(c)
N0x(c
HC(h)
HC(c)
C0(c)
10x(c
lC(h)
O .0
M a
u.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
D.2
D.3
D.4
D.6
TOTALS
(A)
Start-Up
Factors
14.6
245.0
3.5
6.0
HC
> y CO
Subgroups NOx
(5)
Emissions -
Col. 3 X Col. 4
(grams)
4,919,000
82,541,000
1 .180rOOO
2.021.000
6,940,000
82,541,000
1,180,000
/
(6)
7. Cold
Starts
(VI-B)
57
(7)
Trips
(VI-A)
1,373,000
1
(8)
Start-Up
Factors
10.31
159.0
1 1
6.0
HC
X co
Subgroups NOx
\
(9)
Emissions •*
Col. 3 X Col. 4
(grams)
14,156,000
218,307,000
L ?S6 nnn
8.238.000
22.394.000
218.307.000
4,256,000
Non-Work Trip
Start-Up
Emissions
(grams)
I
-C-
o
Total Start-Up
Emissions
(grams)
-------
VI-D. AUTO TRAVEL EMISSIONS
[x | Base Alternative
[ [ Revised Alternative
Policy: Areawide Traffic Engineering
Work Trips
1
Forecast Year: 1982
Non-Work Trips
1
\ /
(1)
Population
Subgroup
All
(2)
Average
Speed
23.3
(3)
VMT
(VI-A)1
5,030,000
TOTALS
(«)
Auto Travel
Factors
(Table D.7)
HC 1.7
co 26.0
NOx 2.0
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
v> HC
Z-r C0
NOx
Suhprouns
(5)
Emissions -
Col. 3 X Col. 4
(grams)
8,551,000
130.780.000
10.060.000
8.551.000
1 3D 7«n nnn
10,060,000
(6)
Average
Speed
25.2
-
n
(7)
VMT
(VI-A)
8,529,000
-
r
Sube
(8)
Auto Travel
Factors
(Table D.7)
»c 1.6
C0 24 . 1
NOx 2.1
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
j> J CO
NOx
roups
(9)
Emissions -
Col. 7 X Col. 8
(grams)
13.646.000
20-^49,000
17,911.000
13,646,000
205,549,000
17,911,000
M
M
I
22.197,000
336.329.000
27,971,000
Source Worksheets are indicated in
parentheses where applicable
Work Trip Travel
Emissions
(grams)
Non-Work Trip
Travel Emissions
(grams)
Total VMT Travel
Emissions
(grams)
-------
VI-D. AUTO TRAVEL EMISSIONS
Base Alternative
Revised Alternative
Policy: Areawide Traffic Engineering
Work Trips
1
Forecast Year: 1982
Non-Work Trips
1
\ /
(1)
Population
Subgroup
All
(2)
Average
Speed
26.5
(3)
VMT
(Vl-A)1
5,036,000
TOTALS
(4)
Auto Travel
Factors
(Table D.7)
HC 1.48
co 22.86
NOX 2.14
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
v- HC
X, co
NOx
(5)
Emissions -
Col. 3 X Col. 4
(grains)
7,4S3,oon
115,123,000
10.777.000
7.453.000
115,123,000
10.777.000
(6)
Average
Speed
28.5
(7)
VMT
(VI-A)
*,674,00(
+
Sube
(8)
Auto Travel
Factors
(Table D.7)
HC 1.36
co 21.10
NOX 2.24
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
HC
CO
NOx
T-, HC
) - CO
NOx
roups
(9)
Emissions -
Col. 7 X Col. 8
(grams)
11,797.000
183,021,000
19.430.000
11,797.000
183,021,000
19,430,000
I
-P-
K3
19.250.000
298.144.000
30,207,000
Source Worksheets are Indicated in
parentheses where applicable
Work Trip Travel
Emissions
(grains)
Non-Work Trip
Travel Emissions
(grams)
Total VMT Travel
Emissions
(grams)
-------
VII-43
60
s = T
where:
s = travel speed in miles per hour
t = travel time rate in. minutes per mile
The revised alternative travel speed is calculated by applying
the proportional changes in travel time for work and non-work travel
shown in Table VII-10 to the base case travel time rates from step 6
to determine the revised average travel time rate. These rates are
then converted to average travel speeds using the above formula.
The 1982 analysis year and the average travel speeds calculated
are used to enter Table D.6 for the base and revised cases and for
work and non-work travel. Linear interpolation is required to
determine the emissions factors for each pollutant. The VMT
figures from Worksheet VI-A complete the input data for Worksheet
VI-D.
Finally, Worksheet VI-E is used to summarize the total vehicle
emissions for the base and revised cases, and to calculate the per-
centage change in emissions for each pollutant.
H. Interpretation of Results
As Worksheet VI-E shows, the improved travel speeds associated
with the proposed traffic engineering measures outweigh the associated
increase in VMT in determining HC and CO emissions impacts. Meaning-
ful reductions in the production of these pollutants are predicted,
while NOX emissions are predicted to increase.
While the emissions impacts calculated in this example provide a
reliable sense of the order of magnitude of the influence of traffic
engineering actions, several relatively simple changes in the analysis
-------
VI-E. SUMMARY OF CHANGES IN EMISSIONS
Revised Alternative
Policy: Areawide Traffic Engineering
Forecast Year:
1982
(1)
Population
Subgroup
IIC
CO
NOx
IIC
CO
NOx
HC
CO
NOx
HC
CO
NOX
Base Emissions
(2)
Trip-Related
(VI-C)1
28,962,000
297, 093,000
5,363,000
TOTALS
Source Worksheets are indicated in
parentheses where applicable
(3)
Travel
(VI-D)
22,197,000
336,329,000
27,971,000
HC
Zco
NOx
Sub-
groups
(4)
Total
(Col. 2 + Col.3)
51,159,000
633,422^000
33,334,000
Total Base
Emissions
(grams)
Revised Emisslor)
(5)
Trip-Related
(VI-C)
29,334,000
300J848DOO
5,436 POO
(6)
Travel
(VI-I))
19,250,000
29 a 144. 000
30^07,000
I
a
(7)
Total
CoL5+Col. 6)
48,584, 000
5?8 392 P 000
35 64 3 POO
HC
CO
NOx
(8)
Change in
Total
Emissions
Col. 4-Col. 7)
- 2,575,000
"~ «U. _/l T[) Jlfjfl
+2,309,OOC
- 2,575.000
-34,4300)0
f 2 ,309 POO
Sub- Total Change
groups in Emissions
(grans)
(9)
Percent
Change in
Emissions
(Col. 8/CoLA)
x 100
-5.0%
-S. 4%
+6.9%
-5.0%
-5.4%
+6.9%
Percent
Change,
Total
Emissions
<
t-H
M
I
-------
VII-45
could improve the estimates. For example, separately calculating the
change in auto emissions accounted for by travel on each of the highway
functional classes, rather than the aggregate average emissions rates
for all facilities in the urban area would provide a more accurate air
quality impact estimate. This extra accuracy would be useful in com-
paring the proposed traffic engineering actions with other transporta-
tion measures which may be competing for limited funding.
The positive impact of the traffic engineering measures on both
auto emissions and fuel consumption in the region (see Worksheet VI-E and
Table VII-10) indicate that they are quite desirable. The next step in
the planning process would be the identification of specific highway
facilities which would benefit from these measures and the development of
cost estimates for their implementation. Once specific opportunities have
been identified, computer analysis techniques such as TRANSYT and FREQ
would be applied for detailed planning purposes.
•U.S. GOVERNMENT PRINTING OTICE : 1980 0-311-726/3706
------- |