Citation
Application of travel speed ranges in air quality modeling

Material Information

Title:
Application of travel speed ranges in air quality modeling
Creator:
Haire, Katherine Marie
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
90 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Air -- Pollution -- Law and legislation -- United States ( lcsh )
Air quality -- Effect of speed limits on ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 89-90).
Thesis:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Civil Engineering
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Katherine Marie Haire.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
33981224 ( OCLC )
ocm33981224
Classification:
LD1190.E53 1995m .H53 ( lcc )

Downloads

This item has the following downloads:


Full Text
APPLICATION OF TRAVEL SPEED RANGES
IN AIR QUALITY MODELING
by
Katherine Marie Haire
B.S., University of Missouri, 1985
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
1995


This thesis for the Master of Science
degree by
Katherine Marie Haire
has been approved
by
Bruce Janson
Sarosh Khan
///?£'
Date


Haire, Katherine Marie (M.S., Civil Engineering)
Application of Travel Speed Ranges in Regional Air Quality Modeling
Thesis Directed by Associate Professor Bruce Janson
ABSTRACT
The passage of the Intermodal Surface Transportation Efficiency Act (ISTEA)
of 1991 and the Clean Air Act Amendments (CAAA) of 1990 interrelated
transportation planning with air quality planning. Through the development
of these two key legislative actions, federal funding was established to allow
transportation projects to focus on measures which will assist an area in
lowering the mobile source air pollutants emitted. The Congestion
Mitigation and Air Quality Improvement (CMAQ) program funds are available
to nonattainment areas for use in transportation projects which can
demonstrate a tangible air quality benefit. Specific to the CMAQ funding
category is a group of projects identified as traffic flow improvement
programs. The implementation of these programs result in the reduction in
congestion without providing additional lane mileage. As provided in the
legislation, an analysis method must be developed to indicate the tangible"
air quality benefits received.
The development of the analysis methodology focused on the characteristics
of the transportation network and the associated levels of emitted carbon
monoxide (CO) pollutants. Travel speeds identified in the travel demand
forecasting model are divided into two subgroups, critical and "non-critical"
travel speeds. The critical travel speeds are those which emit the higher


levels of CO air pollution (travel speeds less than 30 mph and greater than
55 mph). The noncritical travel speeds are those within the 30 to 50 mph
range that emit a lower, steady level of CO air pollution. Improvements of
the traffic flow through the critical range resulted in a "tangible" air quality
benefit as indicated by the estimated emissions reduction for the traffic
network analyzed in this thesis. The traffic flow improvements applied to the
noncritical travel speeds resulted in no change in the air quality. It is
concluded that tangible air quality benefits as required by the legislation
will be possible or achieved only when traffic flow improvements occur in the
critical travel speed range.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.


Chapter
1. Introduction
CONTENTS
1
1.1 Legislation 2
1.1.1 Clean Air Act Amendments 3
1.1.2 Intermodal Surface Transportation Efficiency Act 6
1.1.2.1 Congestion Mitigation and Air Quality Improvement Program 6
1.1.2.1.1 CMAQ Transportation Programs 7
1.1.2.1.2 CMAQ Reporting Requirements 11
1.2 Purpose of Thesis 12
2. Review of Analysis Procedures 13
2.1 Travel Demand Forecasting 13
2.2 Travel Progression Analysis 14
2.3 Air Quality Analysis 15
3. Analysis 16
3.1 T raffic Operations 16
3.1.1 TRANSYT-7F 16
3.1.2 NETSIM 18
3.1.3. Traffic Operations Summary 19
3.2 Travel Demand Forecasting 20
3.2.1 TRANPLAN 20
3.2..1.1 TRANPLAN Trip Generation 20
3.2.1.2 TRANPLAN Trip Distribution 21
3.2.1.3 TRANPLAN Modal Split 23


3.2.1.4 TRANPLAN Assignment 23
3.2.1.5 TRANPLAN Speed Assignment 24
3.2.2 MINUTP 24
3.2..2.1 MINUTP Trip Generation 27
3.2..2.2 MINUTP Assignment Module 27
3.2.2.3 MINUTP Trip Distribution 28
3.2.2.4 MINUTP Speed Assignment 28
3.3 Air Quality Analysis 30
3.3.1 MOBILE5a 25
3.3.1.1 MOBILE5a Analysis Procedures 31
3.3.1.2 MOBILE5a Speed Considerations 32
4. Methodology 42
4.1 Analytical Procedures Utilized in Other Regions 42
4.1.1 DRCOG 42
4.1.2 FHWA 43
4.1.3 SANDAG/CALTRANS 43
4.1.4 Summary of Existing Analysis Procedures 44
4.2 Recommended Analysis Procedure 44
4.2.1 Sample Validation 46
4.2.1.1 Methodology Validation 53
4.2.1.1.1 Critical Data Set 1 53
4.2.1.1.2. Non-Critical Data Set 1 62
4.2.1.1.3 Data Set 2 70


5. Conclusions
63
References
67


List of Tables
1. National Ambient Air Quality Standards 5
2. Speed Table 25
3. Capacity Table 26
4. Baseline VMT 47
5. Baseline VHT 48
6. Baseline Average Speed 49
7. Baseline AM V/C 50
8. Baseline PM V/C 50
9. Baseline Off Peak V/C 51
10. Baseline CO Emissions 52
11. Critical Data Set 1 Average Speed 55
12. Critical Data Set 1 VMT 56
13. Critical Data Set 1 VHT 57
14. Critical Data Set 1 Average Speed 58
15. Critical Data Set 1 AM V/C 59
16. Critical Data Set 1 PM V/C 60
17. Critical Data Set 1 Off Peak V/C 61
18. Critical Data Set 1 CO Emissions 62
19. Non-Critical Data Set 1 Average Speed 63
20. Non-Critical Data Set 1 VMT 64
21. Non-Critical Data Set 1 VHT 65
22. Non-Critical Data Set 1 Average Speed 66
23. Non-Critical Data Set 1 AM V/C 67
24. Non-Critical Data Set 1 PM V/C 68
25. Non-Critical Data Set 1 Off Peak V/C 69
26. Non-Critical Data Set 1 CO Emissions 70
27. Critical Data Set 2 VMT 71
28. Critical Data Set 2 VHT 72
29. Critical Data Set 2 Average Speed 73
30. Critical Data Set 2 AM V/C 74
31. Critical Data Set 2 PM V/C 75
32. Critical Data Set 2 Off Peak V/C 76
33. Critical Data Set 2 CO Emissions 77
34. Non-Critical Data Set 2 - VMT 78


35. Non-Critical Data Set 2 VHT 78
36. Non-Critical Data Set 2 Average Speed 79
37. Non-Critical Data Set 2 AM V/C 80
38. Non-Critical Data Set 2 PM V/C 80
39. Non-Critical Data Set 2 Off Peak V/C 81
40. Non-Critical Data Set 2 CO Emissions 82


List of Figures
1. CMAQ Transportation Programs 10
2. Vehicle Classification Emission Factors 36
3. LDGT Emission Factors @ Varying Temperatures 37
4. LDGV Emission Factors @ Varying Temperatures 38
5. HDGV Emission Factors @ Varying Temperatures 39
6. HDDV Emission Factors @ Varying Temperatures 40
7. Program Year CO Emission Factors 41
8. Data Scenario CO Emission Rates 88


1. Introduction
Transportation projects have long been credited as the major contributor to
the deterioration of the urban air quality. Recently, the passage of federal
legislation has required the correlation between the implementation of
transportation projects and the resulting air quality to be quantified. The
enactment of the Intermodal Surface Transportation Efficiency Act (ISTEA) of
1991 and the Clean Air Act Amendments (CAAA) of 1990 embody the
cause and effect philosophy requiring transportation planners to stretch
beyond the previously established data boundaries. New procedures must
be adopted that bridge databases to achieve an interactive system which
can be used to evaluate the air quality impacts as they result from the
implementation of specific operational transportation projects.
Through the adoption of the ISTEA legislation, a special funding category
was established to assist urbanized areas to reduce regional mobile source
pollutant emissions. The Congestion Mitigation and Air Quality Improvement
(CMAQ) program allows urbanized areas to receive funds to implement
specified transportation programs providing tangible" air quality benefits.
CMAQ funding is divided into seven general categories including:
1) Transit Improvements
2) Shared-Ride Services
3) Traffic Flow Improvements
4) Demand Management Strategies
5) Pedestrian and Bicycle Programs
6) Inspection and Maintenance Programs
7) Other Air Quality Programs.
1


The majority of the CMAQ categories address ways to reduce vehicle miles
traveled (VMT) from the roadways. The calculation of tangible" benefits as
required for CMAQ funding is relatively straightforward for programs mainly
designed to reduce VMT. However, other CMAQ programs that improve the
networks operational characteristics (such as traffic flow improvements), as
opposed to reducing VMT, require that additional procedures be used to
quantify these benefits.
Although the CMAQ funding program requires the calculation of tangible"
air quality benefits, the analysis procedure for traffic flow improvements is
undefined. Traffic flow improvements include any program that improves the
air quality through the reduction of congestion without providing additional
lane mileage. The purpose of this thesis is to develop an analysis
procedure which may be utilized to quantify the air quality benefits of
projects within the Traffic Flow Improvement category of the CMAQ program.
The air quality benefits associated with the traffic flow improvement
programs will utilize the pollutant characteristics of carbon monoxide (CO)
due to its direct correlation with automotive emissions.
1.1. Legislation
Passage of the recent federal legislation has placed a greater emphasis on
the need to understand the relationship between vehicle emissions, regional
air quality, and the implementation of operational transportation
improvements. This section focuses on two key legislative acts, the Clean
Air Act Amendments (CAAA) of 1990 and the Intermodal Surface
Transportation Efficiency Act (ISTEA) of 1991.
2


1.1.1 Clean Air Act Amendments
The Clean Air-Act Amendments (CAAA) of 1990 are a culmination of
legislation which began in 1955 with the passage of the initial air quality
legislation (1). The 1955 legislation allowed the federal government to
conduct air quality research and assist state and local agencies through
assistance and technical support programs.
Following the 1955 Air Quality legislation was the enactment of the Clean Air
Act of 1963. The Clean Air Act of 1963 allowed the expansion of the federal
governments air quality role to provide funding, conduct research, enforce
interstate air pollution regulations, and develop pollutant level criteria to
maintain public health and welfare.
The Air Quality Act of 1967 was an attempt to strengthen the Nations air
pollution control effort. The 1967 Act designated an Air Quality Control
Region for every major metropolitan area, issued air quality criteria and
control technique information. While the 1967 Act mandated criteria which
must be met within a required time frame, the states were responsible for
developing the analysis and enforcement procedures.
As air quality continued to deteriorate, the public concern for the
environment intensified resulting in the passage of the 1970 Clean Air
Amendments. The 1970 Clean Air Amendments established the U.S.
Environmental Protection Agency (U.S. EPA) as an independent federal
agency. In addition, the amendments included setting the National Ambient
Air Quality Standards (NAAQS), designating Air Quality Control Regions
(AQCR), establishing the guidelines by which AQCRs must develop
Statewide Implementation Plans, setting automotive emission and fuel
3


standards, and enabling federal enforcement in air pollution emergencies
and interstate and intrastate air pollution violations.
The 1977 Clean Air Act Amendments set to soften the stringent regulation
set forth in the 1970 amendments. Time frames were extended to allow
additional time to achieve these goals. A most notable subject aspect of the
legislation is its enabling of the EPA to regulate the chemicals that destroy
the ozone layer.
The effort to maintain control over the pollutants emitted into the environment
continued into the next decade. The 1990 Clean Air Act Amendments
(CAAA) utilize the National Ambient Air Quality Standards (NAAQS) to
establish baseline thresholds for seven NAAQS criteria" pollutants which all
areas in the U.S. must achieve. The seven NAAQS criteria pollutants are:
1) Carbon Monoxide (CO)
2) Ozone (O3)
3) Nitrogen Dioxide (NO2)
4) Particulate Matter less than 10 microns (PM-10)
5) Sulfur Dioxide (SO2)
6) Hydrocarbons
7) Lead.
Table 1 shows the corresponding maximum threshold levels for these seven
criteria" pollutants.
Areas which violate the NAAQS thresholds for any one of the criteria
pollutants is designated as a "nonattainment" area. A nonattainment area
may be designated for a single pollutant (i.e. Carbon Monoxide) or multiple
pollutants (i.e. Carbon Monoxide, Ozone and Particulate Matter less than 10
microns). Areas which are designated nonattainment must reach
attainment for the violated criteria pollutant within the time frame stipulated
4


by the CAAA legislation. Failure to do so may risk the delay or loss of
federal transportation funding for some categories of highway projects.
Table 1 National Ambient Air Quality Standards [1, P-251]
Pollutant Averaainq Time Primarv Standard
Carbon Monoxide 8 hr 10 mg/m3 (9 ppm)
Nitrogen Dioxide 1 hr 40 mg/m3 (35 ppm)
Sulfur Dioxide Annual Average 100ng/m3 (0.05 ppm)
24 hr 365fi,g/m3 (0.14 ppm)
PM 10 Annual Arithmetic Mean 50 |ig/m3
24 hr 150 (xg/m3
Hydrocarbons 3 hr 160 |.ig/m3
Ozone 1 hr 235 ng/m3 (0.12 ppm)
Lead 3 Month Average 1.5 M,g/m3
5


1.1.2 Intermodal Surface Transportation Efficiency Act (ISTEA) of
1991
Passage of the Intermodal Surface Transportation Efficiency Act (ISTEA) of
1991 promised to change the focus of transportation. The purpose of the Act
is to develop a National Intermodal Transportation System that is
economically efficient, environmentally sound, provides the foundation for
the Nation to compete in the global economy and will move people and
goods in an energy efficient manner." (2) The ISTEA legislation establishes
a National Highway System, encourages innovative transportation
technologies, encourages public-private partnerships, and promotes
highway safety. One goal of the ISTEA legislation is to maximize the usage
of the existing transportation facilities prior to the construction of any new
single occupancy vehicle facilities.
1.1.2.1 Congestion Mitigation and Air Quality Improvement
Program (CMAQ)
Enacted by the ISTEA legislation, the Congestion Mitigation and Air Quality
Improvement Program (CMAQ) was developed to promote innovative
strategies to reduce the levels of pollutants emitted from mobile sources.
The purpose of the CMAQ program is to fund transportation projects or
programs that will contribute to attainment of national ambient air quality
standards (NAAQS) with a focus on Ozone and Carbon Monoxide. Under
certain conditions, transportation projects and programs targeting Particulate
Matter less than 10 microns are also eligible."(3)
The CMAQ program was developed to assist in the reduction of the pollutant
levels emitted from mobile sources. Mobile sources, such as automobiles,
are primarily responsible for emitting two pollutants, Ozone (O3) and Carbon
6


Monoxide (CO). Therefore, these two pollutants are the focus of CMAQ
program funds. However, since PM-10 can be attributed to mobile source
emissions, funds may be selectively used for PM-10 GMAQ programs. The
CMAQ program is designed to assist nonattainment areas in reaching
attainment or keep an attainment area from becoming a nonattainment area
for CO, Ozone or PM-10 pollutants.
1.1.2.1.1. CMAQ Transportation Programs
Nonattainment areas are required to develop strategies to achieve
attainment within the time period established by the CAAA. The
development of attainment strategies are initiated by the metropolitan
planning organization (MPO) in consultation with the Environmental
Protection Agency (EPA), the Federal Transit Administration (FTA), the
Federal Highway Administration (FHWA), and the State. Transportation
Control Measures (TCMs) are a key group of strategies that can be
proposed to achieve attainment. The TCMs may be used as emission
credits in the Statewide Implementation Plan (SIP). The SIP combines
regional emission control strategies, as well as statewide strategies, to
assure that the State as a whole also reaches attainment within the
designated time frame if it is in question.
CMAQ funds enable States to introduce transportation projects and
programs that will achieve the required air quality standards within the
designated time frame. Effective use of CMAQ funds requires a strong
planning effort for two reasons: (1) limited funding requires obtaining
maximum cost/benefit return from these expenditures, and (2) the substantial
commitment of the personnel to develop the level of analysis necessary to
illustrate a tangible air quality benefit.
7


Transportation programs eligible to receive CMAQ funding include the
following as described in (4):
- Transit Improvements to improve the air quality by enhancing the
existing transit service and providing service to more people.
- Shared-Ride Services to improve air quality by shifting people from
their single occupancy vehicle (SOV) to high occupancy vehicles.
- Traffic Flow Improvements that improve air quality by reducing
congestion without adding lane mileage.
- Demand Management Strategies which develop strategies,
techniques or programs to reduce the demand for SOV travel. Demand
management strategies improve air quality by reducing vehicle miles
traveled and vehicle trips through the implementation of alternate
transportation strategies.
- Pedestrian and Bicycle Programs that improve air quality by making
these zero emission" viable modes of transportation.
- Inspection and Maintenance Programs that improve air quality by
reducing the emissions of the Nations vehicle fleet.
- Other air quality programs that promote public outreach and
education, promote promising new technology to reduce emissions or
promote the conversion of public vehicle fleets to alternative fuels (under
certain conditions).
8


As defined in the legislation, four programs are excluded from receiving
CMAQ funds. These four excluded programs are:
- efforts to reduce emissions for extreme cold-start conditions;
- encouragement of the removal of pre-1980 vehicles;
- increased road capacity for SOVs; and
- maintenance costs for existing systems.
All requests for the CMAQ funds must be coordinated through the MPO. The
MPO is the key planning agency in the nonattainment area for both
transportation and air quality. The MPO is responsible for the development
of the Transportation Improvement Program (TIP) which includes the
federally funded highway and transit projects and TCMs found in the State
Implementation Plan (SIP). All projects that are funded under the CMAQ
program must be included in the TIP. The TIP must be in conformance with
the SIP. As appropriate, these programs may be approved by the EPA as
TCMs included in the SIP and receive emission reduction credits. All TCMs
included in the SIP that qualify for emission reduction credits shall receive
the highest priority for funding as designated by CMAQ regulations.
9


FIGURE 1 (4, p. 5)
CMAQ TRANSPORTATION PROGRAMS
20%
D Transit 1 Shared-Ride 1 Traffic Flow I Demand Mgt ID! Ped & Bike 01/M
10


1.1.2.1.2. CMAQ Reporting Requirements
Federal legislation requires the recipients of the CMAQ funds to submit an
annual report that details the level of expenditures on each program as well
as the air quality benefits realized as a result of the program. Though the
CMAQ funding regulations require tangible air quality benefits be calculated
as a result of the traffic flow improvements, the development of the
methodology used in the analysis is left to the local planning agencies.
The federal agencies are allowing the local agencies to provide their own
analysis procedures providing they are consistent with accepted practice.
This is acceptable due to the varying levels of technical expertise and
analytical approaches utilized throughout the United States. A single
method of analysis would be difficult to adopt at a Federal level due to the
localized control of applications. Therefore, the prescribed analysis method
is to be determined through consultation with the Federal Transit
Administration (FTA), Federal Highway Administration (FHWA), the U.S.
Environmental Protection Agency (EPA), the State and MPOs.
Annual CMAQ reports are produced by the regional planning organization
and submitted to the FTA, FHWA, EPA, and State. FTA, FHWA and EPA
review and accept the reports as received or recommend additional
analyses.
As stated in the CMAQ guidelines (3), the annual report shall contain all
projects funded under the CMAQ program, the dollar amount utilized in the
program, and the estimated tangible air quality benefits for that year.
11


1.2 Purpose of Thesis
The requirement of transportation planners to analyze performance
characteristics of the transportation network not found by traditional
procedures requires the refinement of these procedures. The refined
analysis procedure will enable the user to work between transportation and
air quality databases so as to estimate the tangible air quality benefits. The
purpose of this thesis is to present a methodology that links the
transportation analysis procedures to the air quality analysis procedures in a
manner that meets the analysis requirement for CMAQ funding of proposed
traffic flow improvements.
In order to relate air quality benefits to traffic flow improvements, it is critical
to identify the relationship which exists between traffic operations and the
resulting mobile source air pollutant emissions. The implementation of the
traffic flow improvement projects increase the travel speeds on the network
through improved progression. Therefore, the analysis procedure must
have the ability to relate the resulting travel speed variance as it relates to
the air quality pollutant emissions.
12


2. REVIEW OF ANALYSIS PROCEDURES
To gain an understanding of the influence that transportation infrastructure
(or supply) has on regional air quality, it is necessary to review both
transportation analysis procedures as well as air quality analysis
procedures. Transportation analysis procedures are typically divided into
two general categories, transportation planning and traffic operations. The
analyses conducted specific to the CMAQ traffic flow improvements category
focus on travel demand forecasting models and signal coordination
procedures. The air quality analysis focuses on the MOBILE5a model that is
used to estimate regional emissions based on traffic volumes and speeds.
2.1 Travel Demand Forecasting
Travel demand forecasting is critical to the assessment of future
transportation needs within a region. Travel demand forecasting occurs
through the assistance of a computer generated model that performs the
four-step transportation planning process. The four-step process consists of
Trip Generation, Trip Distribution, Modal Split, and Assignment. Trip
generation estimates the number of trips that will travel to and from an area
based on the land use characteristics of that area. These areas are called
traffic analysis zones into which the region is divided. Trip distribution
estimates the numbers of trips between zones given the production of trips
from and attraction of trips to each zone, and the travel cost between each
pair of zones. Modal split divides trips between user specified modes of
transportation (i.e., public transit or automobile). The assignment step loads
alternative routes of the network with the trips of the various modes.
13


Travel demand forecasting models estimates the network performance
characteristics necessary to formulate the input required for the air quality
model (MOBILE5a). Travel speed are estimated for the network on a link by
link basis. The output of MOBILE5a summarizes the average speeds
categorized by area type and functional classification.
The travel demand forecasting model assigns trips to alternative routes of
the network according to link travel times along these routes. Link travel
times are developed by applying the assigned link volumes to the capacity
and length of each link. The process of developing these travel time (and
thus speeds) is inherent to travel demand forecasting model, and can be
somewhat different between models. The following sections focus on two
travel demand forecasting models (TRANPLAN and MINUTP) and their use
of travel speeds.
2.2 Travel Progression Analysis
The efficient progression of traffic through an urbanized area maintains the
level of service for the transportation network, extends the service life of the
facility and is beneficial to the community. Travel time is increased through
the reduction of stops and delays which improves the networks level of
service. As the progression through the system improves, the flow capacity
through the facility is increased allowing more vehicles to travel on the
system. The community benefits through overall environmental
improvements including increased fuel efficiency and the reduction of
carbon monoxide (CO) emissions.
The development of a coordinated traffic signal system requires the
interconnection of the network traffic signals throughout the region in order
to provide communication. Once the signal system is interconnected, it is
14


necessary to develop signal timing parameters that optimize the progression
of vehicles. This section identifies two operational traffic optimization
programs: TRANSYT-7F and NETSIM.
2.3 Air Quality Analysis
Areas which are designated nonattainment must certify that all
transportation projects within the region do not contribute to a deterioration
in the air quality. The air quality analysis is conducted through a "Baseline
versus Build" versus No-Build test. The transportation projects must emit
less pollution emissions in the Build" scenario then in the Baseline and
No-Build scenarios. The conformity tests utilize the transportation network
output as input into the air quality software. For regional emission
calculations, the MOBILE5a software authorized by the EPA is the required
air quality software.
15


3. Analysis
Conducting air quality analysis requires the input of vehicle volumes and
speeds in order to predict the mobile source emissions. Obtaining vehicle
volume and speed data for an exiting transportation network can occur
through field observations or detailed traffic operations models. However,
obtaining this type of detailed information for future scenarios on a regional
level is not feasible. This section focuses on the analysis considerations for
utilizing the traffic flow software, the travel demand forecasting software, and
the air quality analysis software.
3.1 Traffic Operations
Traffic flow or progression software can be effectively utilized to determine
the widest band width allowable for a given grid network. This band width
establishes the opportunity for the platoon of vehicles to travel downstream
while maintaining a constant rate of speed thus eliminating or reducing the
vehicle stops or delays and reducing the vehicle queuing.
3.1.1 TRANSYT-7F
TRANSYT-7F (7) analyzes corridor operational characteristics to optimize
the progression of vehicles traveling through a series of signalized
intersections. The software conducts platoon dispersion analysis of the
distribution of traffic over time at specified locations in the network.
TRANSYT-7F applies an internal algorithm to distribute the platoon of traffic
evenly downstream.
16


The network optimization is simulated for a described transportation network
through a given cycle length and phasing pattern. The performance index
(PI) is determined by the network analysis of the traffic signal offsets and
cycle splits based on the weighted sum of stops and delays. A series of
iterations are conducted utilizing varying offsets and splits to calculate the
resulting performance index. Input into the program include the traffic signal
parameters (i.e. signal timing, cycle length and signal phasing), the platoon
characteristics, and the roadway characteristics. TRANSYT-7F requires a
common cycle length for all of the traffic signals in the network. The analysis
output includes the speed of the vehicles and the delay experienced by the
vehicles. The flow characteristics of the upstream traffic signal and the
dispersion of traffic as the flow moves downstream are utilized to determine
the performance index.
The TRANSYT-7F program outputs data which may be utilized to improve
network efficiency and assist in evaluation procedures. The data and
measures of effectiveness resulting from the TRANSYT-7F analysis by link
volume include; saturation flow, degree of saturation, total travel, travel time,
delay, stops, fuel consumption, maximum back of queue, and green times.
In addition, signal timing tables for each intersection with the offset from the
main controller is given. The flow of the platoon through the network may be
shown graphically using time-space diagrams.
TRANSYT-7F has proven very sensitive to the speed set within the model.(8)
Therefore, it is critical that the speed input be estimated as accurately as
possible. The optimization process determines the optimal traffic signal
offsets based on the given link speeds (or travel times). Incorrect speed
inputs would cause the TRANSYT-7F to loose a degree of accuracy in the
optimization process by calculating incorrect offsets.
17


TRANSYT-7F has the ability to determine environmental effects associated
with the progression analysis including fuel consumption and resulting
pollutant emissions. The following equations represent the macroscopic
relationship used in the program to calculate environmental impacts.
Rate of Production (Fuel) =_Aev
Cv
Rate of Production (CO emissions) = 3.3963 ( e^ ^4^61v/1000v)
where: v = velocity
A,B,C = user specified constants
3.1.2 NETSIM
NETSIM is a traffic model utilized to evaluate operational characteristics of
urban street networks (9). The model uses node and link information to
input the individual vehicles entering network. Each vehicle is then
stochastically assigned a set of performance characteristics (i.e. vehicle
type, average discharge headway, average acceptable gap, etc.) on the
network. The vehicle proceeds through the network controlled by the
assigned performance characteristics, microscopic car-following, queue
discharge, lane-switching algorithms and by the assigned link turn
percentages.
NETSIM may be utilized to analyze a variety of traffic controls such as
"STOP, YIELD, actuated or pretime traffic signals. NETSIM output
includes speed and delay measures of effectiveness as well as fuel
consumption and pollutant emissions for each vehicle type.
18


3.1.3 Traffic Operations Summary
The traffic flow software is recommended for determining the optimal signal
timing for a series of signalized intersections or a grid network of
intersections. The methodology developed must be able to be utilized on a
regional level. TRANSYT-7F does not analyze vehicle behavior. Thus the
increase in travel speeds on the network would not cause the vehicles to
alter their travel path from origin to destination to account for the improved
conditions. The software assumes that all vehicles will continue on the
same path in a more efficient manner. In actuality, vehicles will continually
shift their path to achieve the shortest route possible. This is an important
consideration as transportation facilities are upgraded or constructed.
Traffic flow software requires detailed operational level input. This
information is limited in the existing condition and becomes a process of
estimation in future conditions. Estimations of future operational conditions
will lessen the accuracy of the optimization model.
Traffic flow software would not prove to be the best method of analysis for
regional transportation networks. The limited size of the network evaluation,
as well as the lack of operational data, does not accurately represent future
regional scenarios. In addition, the traffic flow software does not account for
shifts of the vehicle patterns. As facilities are upgraded or constructed, traffic
will tend to gravitate toward the improved facilities, thus it is important to
reflect this trend in the modeling.
19


3.2 Travel Demand Forecasting
Future conditions and what if scenarios must be developed through the
travel demand forecasting models. Each transportation network alternative
developed in the model will vary the assignment and distribution process.
The travel demand software calculates the change in vehicle paths that are
attracted to an upgraded or new facility.
3.2.1 TRANPLAN
TRANPLAN (5) is a travel demand forecasting model which may be used to
analyze and evaluate the highway network. The TRANPLAN model consists
of a series of routines which allows the user to customize the development of
the model to reflect the analytical needs of the area. Following the four step
planning methodology, TRANPLAN establishes the trip generation, trip
distribution, modal split, and assignment based upon the highway network.
The highway network is initially created through an external program which
contains the x and y coordinates for the nodes and the characteristics of the
links (transportation facilities). Transportation characteristics found for each
highway network link include its functional classification, area type, hourly
flow capacity, link length, posted speed limit, and number of lanes.
3.2.1.1 TRANPLAN Trip Generation
Trip generation data is developed through a process which analyzes factors
associated with trip making patterns as a reflection of the regional
characteristics. Data obtained through travel surveys may be utilized to
supply specific regional characteristics.
20


Each traffic analysis zone (TAZ) in the region is characterized by
socioeconomic conditions. Based upon the socioeconomic characteristics
of each zone, an estimation of the number of trips is determined. Trip
generation determines the number of trips attracted to and produced by
each TAZ.
The socioeconomic characteristics (i.e., number of households, number of
employees, number of persons per household, household income level,
etc.), and the quantity by purpose of trips made are generated for each type
of lands. The trip generation rates made be determined through travel
studies or surveys, or through the Institute of Transportation Engineers (ICE)
manual.
3.2.1.2 TRANPLAN Trip Distribution
TRANPLAN generates a zone-to-zone trip table utilizing a Gravity Model
based upon the classification of the trip (i.e. trip purpose), the travel
impedance factors, zone-to-zone travel indices, and optional K-factors. The
Gravity Model was developed under the guidance of Newton Gravitational
Law (See Equation 1) which assumes that all trips starting from a given
zone are attracted by various traffic generators in other zones and that this
attraction is directly proportional to the relative attraction of the zone and
inversely proportional to the separation between the zones in the gravity
model .(5) Typically, the travel time between network zones serve as the
measure of separation described in the gravitational formula.
Equation 1: Classical Gravitational Formula
Force (f) = mass of body 1 (nrrO mass of body 2 fm21 gravitational
(distance separating ml and m2)2 constant
21


The Classical Gravitation Formula must be restructured in order to fulfill the
needs of travel demand forecasting. This restructuring must incorporate the
characteristics of a transportation network. The restructured gravitational
formula is shown in Equation 2.
Equation 2: Restructured Gravitational Formula:
T(jj) =^1!^
2 Ex(i,x)K(i,x)
where: T^jj is the trips produced in zone i and attracted to zone j
(analogous to gravitational force);
Pj is the trips produced in zone i;
Al is the trips attracted to zone j;
Ft(i,j)is the empirically derived travel time factor that expresses the
average area-wide effect of spatial separation on trip
interchange between zones that are t(i,j) apart;
t(i,j) is the travel time in minutes between zone i and zone j; and
K(j ^ is the specific zone to zone adjustment factor to allow for the
incorporation of the effect of spatial separation on trip
interchange between zones that are t(i,j) apart.
A fundamental requirement of the transportation model is that the calculated
trip attractions balance. This procedure is conducted utilizing the ratio of the
22


computed attractions versus the theoretical attraction calculation based
upon a district or zonal level. The required input is a skim table which
contains the interzonal impedances to be used in the determination of the
friction factor selection. The output from the gravity model is a trip table file
containing the zone-to-zone distributed trips for up to 15 trip purposes.
During the trip production and attraction process, the focus is on the travel
time occurring on each link. As the trips are attracted to a zone, the trip is
influenced by the travel time experienced on the link. The travel time on the
link directly correlates to the speed of the facility, the capacity of the facility,
and the length of the link.
3.2.1.3 TRANPLAN Modal Split
The modal split module is not discussed within this analysis as the highway
elements are the primary concern in the deriving of the speed on the link.
3.2.1.4 TRANPLAN Assignment
TRANPLAN has the ability to incorporate three types of assignment loading.
The options for the assignment loading include: AII-or-Nothing assignment;
Restraint Loading; or Incremental Loading. The following paragraphs
discuss each assignment type.
Restraint Loading places the selected interzonal highway trips onto the
minimum paths of the highway network. The network uses the parameter of
time, adjusted link by link, according to the curve data either specified as a
volume-to-capacity (V/C) time adjustment curve data or the default Bureau of
Public Roads (BPR) capacity restraint formula. The speed of the facility is a
reflection of the travel time impedance calculated on each link.
23


Incremental loading is performed for each iteration. A user-specified
percentage of selected interzonal highway trips is loaded on the minimum
paths during path building. Time is adjusted link by link according to user-
specified V/C speed adjustment curve data or the BPR capacity restraint
formula.
AII-or-Nothing assignment loads all selected interzonal highway trips on the
minimum paths based upon time, distance, cost, or other user specified
impedance. This is a free flow assignment with no consideration given to
the type of links or link capacities.
In the Restraint Loading and the Incremental Loading for each assignment
group, minimum and maximum speeds may be specified which forces the
model to set the adjusted speed within the range.
3.2.1.5 TRANPLAN Speed Assignments
The distribution factors inherent within the program utilize travel time as the
determining factor. The speed of the facility is a reflection of the travel time
impedance calculated on the link. TRANPLAN allows the input of individual
link speeds of the unloaded highway network. Once the network is loaded,
the speeds become a reflection of the impedance on the facilities.
3.2.2 MINUTP
The MINUTP (6) travel demand forecasting model is an interactive software
package based on the four step planning process: trip generation; trip
distribution; mode choice; and trip assignment. This section describes the
development of the MINUTP model and the speed assignment process.
24


The transportation network is built utilizing the X/Y coordinates combined
with the unbuilt network link card file. NETBLD creates the highway network
by reading the ASCII link data records that describes the highway network.
Link speeds can be coded directly or through the use of a look up table
based on link speed class (See Table 2). Link capacity is provided to the
model by a table stratified by link capacity codes and the number of lanes
(See Table 3).
Table 2: Speed Table[6, p. 12]
SPEED CLASS (mph)
FacilitvType
AreaTvpe Freeway 1 Expwy 2
1. CBD 48(11) 37(12)
2. CBD Fringe 48(21) 44(22)
3. Residential 67(31) 47(32)
4. Outer CBD 58(41) 37(42)
5. Rural 67(51) 47(52)
Arterial Collector Centroid
3 4 5
22(13) 20(14) 10(16)
29(23) 25(24) 15(25)
32(33) 28(34) 15(35)
24(43) 22(44) 15(45)
32(53) 28(54) 15(55)
25


Table 3: Capacity [Speed in mph (Service Volume in vehicles
per hour per lane)] [6, p. 14]
AREA TYPE LEVEL
OF
CBD FRINGE URBAN SUBURB RES SERVICE
Freeway 36 40 42 42 45
(1400) (1400) (1400) (1400) (1400) (1400) C
C (1750) (1750) (1750) (1750) (1750) (1750) E
r A Expressway 24 27 30 40 40 55
C (600) (600) (600) (600) (700) (700) C
1 1 (840) (840) (885) (1500) (1500) (1500) E
L i Principal 22 23 28 36 38 42
T Arterial (600) (600) (600) (600) (600) (600) C
Y (840) (840) (840) (885) (885) (885) E
T Minor 22 23 24 27 33 42
Y Arterial (350) (350) (350) (400) (400) (400) C
P c (560) (560) (560) (650) (650) (885) E
c Collector 15 20 20 20 25 30
(350) (10000) (350) (350) (350) (350) C
(500) (10000) (550) (550) (550) (550) E
26


3.2.2.1 MINUTP Trip Generation
Trip generation is developed through the production and attraction files
developed for the three trip purposes (home based work, home based other,
and non-home based) assigned to the traffic analysis zone (TAZ). The trip
generation file is combined with the impedance file and friction factor file to
yield the zone to zone matrix which contains the production and attraction
person trips by purpose. The zone to zone production and attraction files
are converted to origin and destination files by applying car occupancy
factors, adding through movements and additional tables. The built highway
network, the origin and destination file, and the turn penalty file are input into
the assignment model. The resulting loaded link network is utilized to
compute the vehicle miles traveled (VMT) on the network.
3.2.2.2 MINUTP Assignment Module
The MINUTP assignment module, ASSIGN, determines the travel time
between zones and then assigns the zone-to-zone trip values to the network
links along the paths. The assignment process may utilize the All or
Nothing, the All Shortest Path, or the Stochastic methodologies. Any of the
assignment methodologies may be iterative or incremental when combined
with capacity restraint. Additionally, the volume and/or equilibrium may be
adjusted through the iterative assignment.
PTHBLD reads the network description and determines the minimal travel
paths from each zone centroid node to each other zone centroid node. The
paths selected are the minimum impedance paths based upon each links
time and distance. Optionally, turn penalties may be assessed at selected
intersections.
27


3.2.2.3 MINUTP Trip Distribution
TRPDST reads production and attraction, impedance, and friction factor data
sets and applies the standard gravity model distribution equation for each
zonal pair in the study area. This is done individually for each of the
purposes. TRPDST iterates a selected number of times in an attempt to
obtain correct attraction totals within each zone. On the last iteration it writes
out trip matrices containing estimated trips for each zonal pair for each
phase.
3.2.2.4 MINUTP Speed Assignments
The introduction of speed assignments within the MINUTP model occurs in
the PTHBLD module. In this module, the speed categories may be
established in a link format (i.e. actual input of the speed data) or the link
speed look up table which references the speed by roadway functional
classification and area type. Congested link speeds are generated through
the assignment model where the zone-to-zone trips are assigned to the
network.
The travel demand software is built and calibrated based on the vehicle
volumes on the links. The model accurately determines the demands on the
facilities, however, the speeds of the facilities are seldomly verified. Once
the transportation network achieves the calibrated vehicle volumes, the
average speeds are not checked for reasonableness (10). Planners
assume that the modeled average speed on the network is accurate if the
vehicle volumes are calibrated.
The travel demand forecasting models typically use a speed-flow curve such
as the BPR curve to estimate the congested travel speed given the initial
28


tree-flow speed and the volume/capacity ratio (V/C). The standard equation
for the BPR curve is:
congested speed = free-flow speed
(1+0.15*V/C4)
Planning models may estimate demands over the specified capacity of the
individual facility. As this does not occur in actuality and therefore there are
no observed speed-flow curves for V/C ratios over 1.00. As a result, the
user defined speed-flow curve or the default BPR curve may be applied
inappropriately when the volumes meet or exceed the capacity. As the
network approaches a V/C ratio close to 1.00, the conditions become
unstable. This unstable condition creates difficulty in predicting the average
speed on the facility. An error in speed estimation is critical as V/C ratios
approach 1.00. At that point, the congestion causes the vehicles to slow to
very low speeds thus exponentially increasing the air pollutant emission
rates.
The speed-flow curves utilized in the travel demand forecasting models do
not account for the effects of queuing on travel speeds and demand. As a
result, the calculations of average speed may exceed those actually on the
roadway. Planning models estimate the average speed on the network
through the flow and capacity data for each link. The travel demand
forecasting models provide the best level of estimation for the future network
comparisons.
29


3.3 Air Quality Analysis
3.3.1 M0BILE5a
M0BILE5a is the air quality software authorized by the EPA to be utilized in
nonattainment areas to ascertain the regional pollution levels emitted from
mobile sources. MOBILE5a contains three input sections; the control
section, the data section and the scenario section.
The control section contains the portion of the input data that regulates the
input, output, and execution of the program. Control elements include the
inspection and maintenance program, additional input data, output emission
factors for visual inspection, and output formatted for further analysis in
another program.
The data section contains one-time application information pertaining to ALL
scenarios. The data section defines parameter values different from those
internal to the program which will be used in the calculations of all scenarios
within a given run. Examples of data section input include the annual
mileage accumulation rates, registration distribution by age, and further
control program parameters (i.e. description of inspection and maintenance
program).
The Scenario section details the individual scenarios for which emission
factors are to be calculated. These include the calender year of evaluation,
average speed to assume, and the high- or low-altitude region.
30


3.3.1.1 M0BILE5a Analysis Procedures
The transportation network for each scenario is developed and formatted for
input into the MOBILE5a model. The MOBILE5a model generates a series
of modules to obtain the all-day pollutant emissions rates for each scenario.
The focus within this analysis is on the pollutant carbon monoxide (CO) due
to the direct correlation with vehicle exhaust.
Initially, setup information obtained from the environmental regulatory
agency is applied to the transportation network to create the mobile source
pollutant emissions. Representative setup information includes: regional
vehicle registrations; inspection and maintenance programs; anti-tampering
programs; oxygenated fuels; Reid vapor pressure; average minimum and
maximum temperatures; regional vehicle mix by area type and functional
classification for a.m., p.m., and off peak periods; and the adjustment factors
to convert the peak period data (VMT mix and operating modes) into an all-
day average.
The transportation network output serves as input to an average speed
model that generates the am, pm, and off peak periods stratified by area type
and functional classification for low and normal operating speed ranges.
The average speed data files contain the transportation network data file
which will form one of the required input to later calculate the total carbon
monoxide emissions attributed to the network.
Once the network speeds have been determined the scenario data file must
be prepared for the forecast year. The all-day operating modes and ambient
temperature are loaded with the average speeds into the final forecast year
scenario data.
31


The control data section, including the one-time data records, is added to the
scenario data file to initiate the MOBILE5a program. The emission factors
attributed to the network are taken from the output of the MOBILE5a. Total
emission rates for the forecast year scenario are then calculated.
The MOBILE5a program allows for the modification of input providing
concurrence is received from the EPA, the FHWA, the FTA, and the State. An
agency may wish to alter the input of the average speed so as not to stratify
the VMT in the low and high speed categories. This may reflect the actual
air quality of the region as the low and high speed ranges contribute greater
amounts of CO pollutants then the mid range speeds. It is important to
prevent the dilution of the low and high speed categories into the mid range
speeds for the pollutant emission calculations.
3.3.1.2 MOBILE5a Speed Considerations
The average speed input is one of the most critical inputs into the MOBILESa
program. The development of the CO emission rates are based upon the
application of the mobile source pollutant emission parameters onto the
average speeds in the network.
The MOBILESa model focuses on the pollutant emissions of mobile sources.
The mobile sources are typically divided into four categories: Light Duty
Gasoline Vehicles (LDGV); Light Duty Gasoline Trucks (LDGT); Heavy Duty
Gasoline Vehicles (HDGV); and Heavy Duty Diesel Vehicles (HDDV). This
section reviews CO pollutant levels for various scenarios.
Figure 2 illustrates the CO emission factors as a reflection of vehicle
classifications and travel speeds. The heavy duty diesel vehicles (HDDV)
emit the lowest level of CO pollutants at all speeds, approaching zero
32


emissions between 40 and 50 mph. The light duty gasoline vehicles (LDGV)
emit the second lowest level of CO pollutants, followed closely by the light
duty gasoline trucks (LDGT). The heavy duty gasoline vehicles (HDGV) emit
the highest level of CO pollution, significantly greater than the other three
categories.
The CO pollutant emission rates vary by the vehicle type and ambient
temperature. As illustrated in Figures 3 through 6, the ambient temperature
alters the emission rates from the mobile sources. Figure 3 illustrates the
LDGV emission factors at three varying temperatures. At all three
temperature levels the CO emission factors are high at speeds below 10
mph, decreasing rapidly as speeds reach 30 mph, becoming relatively
stable between 30 and 50 mph and increasing slightly as speeds become
greater than 55 mph. The coldest temperature modeled (20F) emits the
highest levels of CO pollutants at all travel speeds. The warmest
temperature modeled (60F) exhibits the lowest levels of CO pollutants
emitted at all travel speeds. The mid-range modeled temperature (40F)
emits pollutants less than the 20F conditions and more than the 60F
conditions. The temperature variance creates significant differences in the
modeled pollutant levels at travel speeds less than 10 mph. As the three
temperature curves approach the stable conditions of 30 to 50 mph, the gap
between the pollutant levels narrows.
The LDGT emission factors, illustrated in Figure 4, indicate a similar set of
curves to those described above. The CO emission factors are slightly
higher for all three temperatures in the LDGT graph than in the LDGV graph.
Similarly, the widest margin between the CO pollutant emission rates occurs
in the low travel speed range of 0 to 10 mph. The margin between the
temperature curves narrows as speeds approach 30 mph, remaining narrow
through speeds of 50 mph and widening after travel speeds of 55 mph are
33


reached. The lowest illustrated temperature curve (20F) exhibits the highest
CO emission levels whereas the highest temperature curve (60F) exhibits
the lowest levels of CO emission factors.
The HDGV graph, illustrated in Figure 5, projects the highest levels of CO
pollutant emission factors as compared to the other vehicle types for all
temperature levels and travel speeds. The CO pollutant emission curve, as
described by the travel speed and ambient temperature, peaks at travel
speeds less than 10 mph. The CO pollutant emission curve decreases as
travel speeds approach 30 mph. Between 30 and 50 mph, the emission
factors remain relatively stable. As travel speeds exceed 50 mph, the CO
pollutant emission factors gradually begin to climb. Similar to the other
graphs, the lower temperature levels emit the higher CO pollutant emissions
level.
Figure 6 illustrates the overall lowest CO pollutant emission levels of the four
vehicle types. All three temperatures project the same CO emission factors
within each travel speed. The CO emission factors peak at travel speeds
under 10 mph, remaining stable between 30 and 50 mph and increasing
over 50 mph.
The MOBILE5a CO emission rates are subject to change in order to reflect
the changing conditions. The emission rates are affected by the
characteristics of the vehicle fleet including the turnover of vehicle fleets and
the improvement of automobile technology. The turnover of vehicle fleets as
well as the improvement of automobile technology result in lower levels of
CO emissions.
Figure 7 illustrates the variance in pollutant emission factors for three
planning horizons. The projected CO emission factors are expected to be
34


less in the planning year 2003 than they are in the year 1995 and 1993. As
expected, the improvement in technology decreases the levels of CO
pollutants emitted. As older vehicles are phased out of service and replaced
with new vehicles, the average CO emission rate by vehicle distribution
decreases. Again the CO pollutant emission factors are at their highest as
speeds are less than 10 mph. The CO emission factors continue to
decrease gradually until travel speeds reach 55 mph and then begin to rise.
The variance between the CO emission factors narrow as the average
speed increases.
35


FIGURE 2 (13)
Vehicle Classification Emission Factors
Speed (mph)
ILDGV LDGT BHDGV iHDDV
36


FIGURE 3 (13)
LDGT EMISSION FACTORS 9 VARYING TEMPERATURES
ILDGT ZOF LDGT 40FI LDGT 60F
37


FIGURE 4 (13)
LOGV EMISSION FACTORSVARYING TEMPERATURES
i-T-NNlOfOtftfinintffl
Speed (mph)
ILDGV20F OlIM 1LDGV60F
38


FIGURE 4 (13)
HDGV EMISSION FACTORS I VARYING TEMPERATURES
Speed (mph)
IHDGV 20f D HDGV 40F lHDGV 60T
39


FIGURE 6 (13)
HDDV EMISSION FACTORS i VARYING TEMPERATURES
5 10 15 20 25 30 35 40 45 50 55 60 65
Speed (mph)
I HDDV ZOF HDDV 40F HDOV 60F


FIGURE 7 (15)
PROGRAM YEAR CO EMISSION FACTORS
0 fm
p L a
0 A
m 9
V
200r
I oo-

20
30 40

60
50
Speed (mph)
41


4. METHODOLOGY
4.1 Analytical Procedures Utilized in Other Regions
A review of the existing analysis procedures utilized in the United States
was conducted to formulate a basis for comparison. Unfortunately, this is a
relatively new area of analysis and limited data is available.
4.1.1. DRCOG
The Denver Regional Council of Governments (DRCOG) utilizes the corridor
specific methodology to analyze each signal timing/optimization project (11).
The traffic signal coordination plans are developed for each of the three time
periods: am peak period (6:30 a.m. through 8:30 a.m.); p.m. peak period
(3:00 p.m. through 6:30 p.m.); and the off-peak (period not included in the
a.m. or the p.m. peak periods). Progression analysis is conducted on the
corridor to establish the optimal signal cycle length. Once the optimal signal
cycle length is implemented along the corridor, analysis is conducted to
ascertain the benefits achieved. The before and after scenarios are
compared as to the Travel Time (seconds), the Stopped Time (seconds) and
the Travel Speed (miles per hours). In addition, the environmental benefits
are calculated as to the reduction in Fuel Consumption (gallons) and the
Pollutant Emissions (Carbon Monoxide, Hydrocarbons, and Nitrous Oxides).
The analysis is conducted only on the progression on the major street.
Analysis is not conducted on the minor cross streets.
42


4.1.2. FHWA
An outline approach, developed by the FHWA (13) to determine the
emission reductions by the CMAQ programs, has been provided to the
MPOs to assist in the development of an analysis procedure. The outline
allows the CMAQ programs and projects to be analyzed individually or as a
group of projects which are in the same category (i.e. transit) or area (i.e.
transportation corridor). The steps in the analysis procedure are as follows:
1) estimate the emissions per trip in grams for each criteria pollutant
(using an emissions model approved by EPA which provides the most recent
estimate of emissions);
2) multiplied by the estimated number of daily trips reduced for each
criteria pollutant (preferably using a methodology consistent with the
methodology used or proposed to determine the air quality benefits of
transportation control measures in an approved State Implementation Plan
revision);
3) equals the estimated emission reductions in grams per day for
each criteria pollutant; and
4) divided by 1000 equals the estimated emission reduction in
kilograms per day for each criteria pollutant.
4.1.3. SANDAG/CALTRANS
Evaluating the Traffic Flow Improvement Programs for the
SANDAG/CALTRANS area are conducted through user inputs of speed
43


changes (13). The California Air Resource Board analyzes traffic
signalization improvements by speed changes determined from traffic flow
modeling. The traffic flow model requires the user to input network data
(links and nodes), saturation flows, traffic volumes, cruise speeds, bus stop
delays and traffic signal data.
4.1.4. Existing Analysis Procedure Summary
The three procedures differ from one another within analysis techniques.
The DRCOG method reviews corridor impacts of the transportation
implementation. This analysis supports a before and after process that may
only be implemented for the immediate time period. Future scenarios would
need to project the operational characteristics whereby loosing a degree of
accuracy. The FHWA method requires the implementation of the
transportation project to reduce the number of trips on the network. Traffic
flow improvements represent the same number of vehicles on the roadway
and therefore do not fit into this analysis procedure. The
SANDAG/CALTRANS method fits the requirements for the analyzing of the
traffic flow improvements. This procedure models the impacts of the speed
changes through traffic flow modeling.
4.2 Recommended Analysis Procedure
The CO emission factor rates (illustrated in Figure 6) show very little
variance between 30 and 50 mph. Therefore, it is determined that the CO
pollutant emission factors will remain relatively stable within this range.
Projects which are implemented within the 30 to 50 mph range will be
unable to show a significant tangible benefit for CO pollutant emissions.
44


Utilizing this hypothesis as a foundation, a methodology was developed to
formulate a tangible solution. The following steps outline a strategy for
determining the benefits associated with the implementation of the traffic
flow enhancement CMAQ program.
Step 1: Run the travel demand forecasting model for the existing
transportation system prior to the implementation of the CMAQ
program.
Step 2: Calculate the resulting average speed as a function of the VMT
and VHT output from the travel demand forecasting model by
area type and functional classification.
Step 3: Identify the functional classification and area type of
transportation facilities that are not in the 30 to 50 mph range.
Step 4: Determine if the cell (functional class by area type) is within the
range or outside the range. If within the range advance to Step
5. If outside the speed range, advance to Step 8.
Step 5: Speed within the 30 to 50 mph range, therefore conduct
localized corridor specific analysis.
Step 6: Build corridor in traffic progression analysis program.
Step 7: Conduct localized corridor air quality analysis.
Step 8: Speed outside of the 30 to 50 mph range, therefore
regionally significant analysis.
45


Step 9: Determine the revised travel speed through the aid of traffic
flow progression analysis or field surveys.
Step 10: Input the revised speed for the appropriate links into the
regional travel demand model.
Step 11: Calculate the revised mobile source emissions utilizing
MOBILE5a.
Step 12: Compare the "Build versus No Build versus Baseline"
scenarios.
4.2.1 Sample Validation
A sample transportation network was utilized to review the methodology
presented within this thesis. The sample network consists of 350 traffic
analysis zones and 4413 links covering approximately 1,350 square miles.
The loaded transportation network contains 216,666 total trips. Network
calibration is based on the comparison of link volumes to actual vehicle
volumes at cordon locations. Area Type and Functional Classification
categories reflect those discussed in the MINUTP capacity table. The
purpose of this example is to forecast traffic on a given network, develop
comparative networks to confirm, through application, that the analysis
procedure is valid.
The transportation network was generated through the four step planning
process. The Baseline" scenario reflects the existing modeled conditions
that future model modifications will be compared to. Table 4 and Table 5
illustrate the VMT and VHT respectively for the "Baseline scenario.
46


Table 4: Baseline VMT by Functional Classification and
Area Type
VMT units are Vehicle Miles
Area Type
Functional
Classification 1
1 0
2 0
3 39637.3
4 14303.2
5 11.1
6 0
2 3
581863.8 1627630.4
4443.6 1212576.1
103037.3 3133095.7
33384.7 839519.5
12152.1 299651.3
18106.6 92871.6
4 5
196779.8 1590869.7
174741.8 517580.5
193086.7 1540968.2
47204.2 995243.9
65186.0 151275.0
36529.5 48884.4
47


Table 5: Baseline VHT by Functional Classification and
Area Type
VHT units are Vehicle Hours
Area Type
Functional
Classification 1
1 0
2 0
3 1584.4
4 646.0
5 0.6
6 0
2 3
11633.5 31449.8
125.9 31803.8
3954.7 89433.1
1329.4 27975.5
549.1 10703.6
719.7 3709.0
4 5
3771.9 25399.9
4615.2 10736.1
6436.3 37644.2
1972.1 28437.4
2729.5 5042.5
1465.1 1951.8
Average speed for the Baseline" scenario must be calculated for each area
type and functional classification. This may be achieved through the
arithmetic calculation of VMT divided by VHT. The average speed
calculated for the baseline scenario is contained in Table 6.
48


Table 6: Baseline Average Speed by Functional Classification
and Area Type
Average Speed unit is miles per hour
Functional Area Type
Classification 1 2 3 4 5
1 0 50.02 51.75 52.17 62.63
2 0 35.29 38.13 37.86 48.21
3 25.02 26.05 35.03 30.00 40.94
4 22.14 25.11 30.01 23.94 35.00
5 18.50 22.13 28.00 23.88 30.00
6 0 25.16 25.04 24.93 25.05
The baseline scenario volume to capacity ratios (V/C) calculated for the AM,
PM and Off peak periods are shown in Tables 7,8, and 9 respectively.
49


Table 7: Baseline AM Volume to Capacity Ratios (V/C)
Functional
Classification 1 2
1 0 .09
2 0 .07
3 .04 .05
4 .02 .03
5 0 .01
6 0 .07
Area Type
3 4 5
.09 .07 .04
.06 .08 .05
.06 .05 CO o
.03 .02 .02
.02 .02 .01
.05 .07 .04
Table 8: Baseline PM Volume to Capacity Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .06 .05 .05 .03
2 0 .04 .04 .05 .03
3 .02 .03 .04 .03 .04
4 .01 .02 .02 .01 .01
5 0 .01 .01 .02 .01
6 0 .04 .03 .04 .02
50


Table 9: Baseline Off Peak Volume to Capacity Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .12 .12 .10 .06
2 0 .09 .08 .11 .07
3 .05 .06 .08 .07 .08
4 .03 .04 .04 .02 .03
5 0 .01 .03 .04 .02
6 0 .09 CO o .09 .05
The VMT, VHT and average speed for the baseline scenario are entered into
the MOBILE5a model. Table 10 lists the estimated CO emissions for the
baseline scenario.
51


Table 10: Baseline CO Emissions
AM Peak Period: 10,850,870 grams per day 11.96 tons per day
PM Peak Period: 13,542,100 grams per day 14.93 tons per day
OFF Peak Period: 97,740,460 grams per day 107.74 tons per day
TOTAL CO EMISSIONS: 122,133,430 grams per day 134.63 tons per day
52


4.2.1.1. Methodology Validation
As illustrated in Figure 2, the Carbon Monoxide emission rates remain
relatively constant between 30 mph and 50 mph. Therefore, any change in
the average speed for areas within that range should not create additional
regional CO pollutant emissions nor should the change create a beneficial
situation.
To verify the methodology, two sample data sets were utilized. The first data
set (Data Set 1) modified a small number of facilities within the network
whereas, the second data set (Data Set 2) modified a larger amount of
facilities within the region. In both data sets, the transportation network was
run prior to any modifications to determine which area types and functional
classifications fell within and outside of the critical speed range. Individual
analysis was performed on four separate cases: Critical Data Set 1; Non-
Critical Data Set 1; Critical Data Set 2; and Non-Critical Data Set 2. The
critical and non-critical title refers to whether the modified input free-flow
speeds are within the critical range or within the non-critical range.
The methodology will compare the two modified data sets with the initial
baseline data set to determine the impacts on the regional CO air quality.
Initially, a small corridor will be altered (Data Set 1) with higher input free-
flow speeds in the critical and noncritical areas. The second data set (Data
Set 2) will modify a larger section of input free-flow speeds within the critical
and noncritical areas. Comparisons of the two data sets will be made to the
"Baseline scenario for resulting CO air quality performance.
4.2.1.1.1 Critical Data Set 1
The initial phase is to ascertain which links will benefit from travel speed
53


improvements. The best use of the funding is to apply traffic flow
improvements to the links which will improve the network travel speeds thus
benefiting the regional air quality. Links identified as providing the most
benefit for improving regional air quality are the critical travel speeds found
in Table 11.
The first modification of Data Set 1 focused on the critical speed ranges
occurring beyond the 30 to 50 mph range. Table 11 shows the critical"
average speeds to be incorporated into Critical Data Set 1. As the travel
speeds on the critical links are improved, a corresponding improvement in
the regional air quality should result. The improvement to the regional air
quality will reflect the degree of improvements on the roadway.
The areas which are excluded from consideration (*****) are the areas with
travel speeds that fall within the 30-50 mph and therefore maintain a
constant CO pollutant emission range. The free-flow input speeds are
altered on a five mile corridor and the vehicle tables are held constant. The
network is rerun and allowed only to recalculate the assignment process.
The Vehicle Miles Traveled (VMT) and Vehicle Hours Traveled (VHT) are
recalculated for each facility type and area type.
54


Table 11: Critical Average Speed by Functional Classification
and Area Type
Average Speed unit is miles per hour
Functional Classification 1 2

1 0 ******
2 0 ******
3 25.02 26.05
4 22.14 25.11
5 18.50 22.13
6 0 25.16
Area Type
3 4 5
****** ****** ****** ****** ******
****** ****** ****** 23.94 ****** ******
28.00 23.88 ******
25.04 24.93 25.05
The new link input speeds may be obtained through traffic flow progression
analysis or observed field surveys. For illustrative purposes, one cell
(functional classification and area type) is chosen as a representative
sample to determine the impact an improvement to a single corridor would
have on the regional network.
Critical Data Set 1 will revise the travel speeds on a 5 mile corridor that is
located in the CBD (area type 1) and functionally classified as a principal
arterial (facility type 3). The overall input speed will be increased 10 mph
along the corridor.
The travel demand forecasting model is rerun with the revised travel speed
link data input. The vehicle trip tables remain constant and the revised VMT
55


and VHT are output. Table 12 and Table 13 illustrate the VMT and VHT
respectively for the Critical Data Set 1 analysis. The calculated Critical Data
Set 1 average speeds are found in Table 14.
Table 12: Critical Data Set 1 Scenario VMT by Functional
Classification and Area Type
VMT units are Vehicle Miles
Functional
Classification 1 2
1 0 590613.8
2 0 4476.0
3 24187.7 88382.9
4 39666.0 43062.1
5 125.1 12902.8
6 0 18633.3
Area Type
3 4 5
1632385.5 195581.4 1591000.2
1214082.9 177229.7 517317.9
3124055.3 191422.8 1541340.2
837065.7 47165.0 995222.5
297227.6 64796.2 151277.9
93000.1 37338.7 48896.9
56


Table 13: Critical Data Set 1 VHT by Functional Classification
and Area Type
VHT units are Vehicle Hours
Area Type
Functional
Classification 1
1 0
2 0
3 966.4
4 1133.1
5 6.3
6 0
2 3
11830.4 31542.0
126.6 31843.3
3391.9 89174.9
1719.3 27893.7
583.4 10617.7
740.6 3714.0
4 5
3750.0 25402.0
4687.2 10727.9
6380.8 37654.5
1970.5 28436.9
2713.8 5042.6
1497.6 1952.4
57


Table 14: Critical Data Set 1 Average Speed by Functional
Classification and Area Type
Average Speed unit is miles per hour
Functional Area Type
Classification 1 2 3 4 5
1 0 49.92 51.75 52.16 62.87
2 0 35.36 38.13 37.81 48.22
3 25.03 26.06 35.03 30.00 40.93
4 35.01 25.05 30.01 23.94 35.00
5 19.86 22.12 27.99 23.87 30.00
6 0 25.16 25.04 24.93 25.04
58


Critical Data Set 1 calculated volume to capacity ratios (V/C) are shown in
Tables 15, 16, and 17 for the AM, PM and Off Peak Periods respectively.
Table 15: Critical Data Set 1 AM Volume to Capacity Ratios
(V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .09 .09 .07 .04
2 0 .07 .06 .08 .05
3 .02 .04 .06 .05 .06
4 .06 .04 .03 .02 .02
5 0 .01 .02 .02 .01
6 0 .07 .05 .07 .04
59


Table 16: Critical Data Set 1 PM Volume to Capacity Ratios
(V/C)
Functional
Classification 1 2
1 0 .06
2 0 .04
3 .02 .02
4 .04 .02
5 0 .01
6 0 .04
Area Type
3 4 5
.06 .05 .03
.04 .05 .03
.04 .03 .04
.02 .01 .01
.01 .02 .01
.03 .04 .02
60


Table 17: Critical Data Set 1 Off Peak Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .12 .12 .10 .06
2 0 .09 .08 .11 .07
3 .03 .05 .08 .07 .08
4 .08 .05 .04 .02 .03
5 0 .02 .03 .04 .02
6 0 .09 .06 .09 .05
The Critical Data Set 1 average speeds are combined with the
transportation network as input into the MOBILE5a model. The resulting CO
emissions are shown in Table 18.
61


Table 18: Critical Data Set 1 CO Emissions
AM Peak Period: 10,844,900 grams per day 11.95 tons per day
PM Peak Period: 13,516,250 grams per day 14.90 tons per day
OFF Peak Period: 97,033,420 grams per day 106.96 tons per day
TOTAL CO EMISSIONS: 121,394,570 grams per day 133.81 tons per day
4.2.1.1.2 Non-Critical Data Set 1
The second phase of the validation is to illustrate the impact similar changes
have within the 30-50 mph range have on the regional air quality. It is
hypothesized that altering the non-critical average speeds within the 30 mph
to 50 mph range would not significantly alter the regional CO emissions.
Therefore, Non-Critical Data Set 1 selects an area which has a non-critical
average travel speed. Table 19 illustrates the average travel speeds which
fall within the stable CO pollutant emission factor range.
62


Table 19: Non-Critical Average Speed by Functional
Classification and Area Type
Average Speed unit is miles per hour
Functional Area Type
Classification 1 2 3 4 5
1 ****** 50.02 51.75 52.17 62.63
2 ****** 35.29 38.13 37.86 48.21
3 ****** ****** 35.03 30.00 40.94
4 ****** ****** ****** ****** 35.00
5 ****** ****** ****** 30.00
ft ****** ****** ****** ****** ******
Non-Critical Data Set 1 chose the area within the CBD Fringe (area type 2)
and functionally classified as an expressway (facility type 2). A 5 mile
section of roadway was selected within this area and the travel speed input
was increased 10 miles per hour. Once again the travel demand forecasting
model was run to generate the revised VMT and VHT. Tables 20 and 21
illustrate the VMT and VHT respectively. The average travel speeds for Non-
Critical Data Set 1 are shown in Table 22.
63


Table 20: Non-Critical Data Set 1 VMT by Functional
Classification and Area Type
VMT units are Vehicle Miles
Functional
Classification 1 2
1 0 581607.9
2 0 4464.5
3 39244.1 104229.6
4 14083.0 33228.4
5 11.1 12101.2
6 0 18073.9
Area Type
3 4 5
1626755.2 194973.7 1591032.6
1203965.3 174174.0 517353.4
3165235.1 191133.1 1539406.5
832181.3 47084.6 995646.2
293809.3 65039.0 151210.3
92883.1 36382.1 48880.2
64


Table 21: Non-Critical Data Set 1 VHT by Functional
Classification and Area Type
VHT units are Vehicle Hours
Area Type
Functional
Classification 1
1 0
2 0
3 1568.7
4 636.0
5 0.6
6 0
2 3
11638.6 31432.6
126.5 31578.1
4000.2 89680.2
1323.3 27934.4
546.9 10496.2
718.5 3709.3
4 5
3737.7 25402.5
4600.5 10731.5
6371.0 37604.4
1967.2 28449.0
2723.4 5040.4
1459.1 1951.7
65


Table 22: Non-Critical Data Set 1 Average Speed by
Functional Classification and Area Type
Average Speed unit is miles per hour
Functional
Classification 1 2
1 0 49.97
2 0 35.29
3 25.02 26.06
4 22.14 25.11
5 18.50 22.12
6 0 25.16
Area Type
3 4 5
51.75 52.16 62.63
38.13 37.86 48.21
35.29 30.00 40.94
30.01 23.93 35.00
27.99 23.88 30.00
25.04 24.93 25.04
The AM, PM and Off Peak volume to capacity ratios (V/C) for the Non-Critical
Data Set 1 scenario are shown in Tables 23, 24, and 25 respectively.
66


Table 23: Non-Critical Data Set 1 I AM Volume to Capacity Ratios (V/C)
Functional Classification 1 2 Area Type 3 4 5
1 0 .09 .09 .07 .04
2 0 .07 .06 .08 .05
3 .04 .05 .06 .05 .06
4 .02 .03 .03 .02 .02
5 0 .01 .02 .02 .01
6 0 .07 .05 .07 .04
67


Table 24: Non-Critical Data Set 1 - PM Volume to Capacity Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .06 .05 .05 .03
2 0 .05 .04 .05 .03
3 .02 .03 .04 .03 .04
4 .01 .02 .02 .01 .01
5 0 .01 .01 .02 .01
6 0 .04 .03 .04 .02
68


Table 25: Non-Critical Data Set 1 Off Peak Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .12 .12 .10 CO o
2 0 .09 .08 .11 .07
3 .05 .06 .08 .07 .08
4 .03 .04 .04 .02 .03
5 0 .01 .03 .04 .02
6 0 .09 .06 .09 .05
The Non-Critical Data Set 1 average travel speeds as well as the
transportation network are input into the MOBILE5a model. The resulting
CO emissions are shown in Table 26.
69


Table 26: Non-Critical Data Set 1 CO Emissions
AM Peak Period: 10,848,600 grams per day 11.96 tons per day
PM Peak Period: 13,543,020 grams per day 14.93 tons per day
OFF Peak Period: 97,793,360 grams per day 107.80 tons per day
TOTAL CO EMISSIONS: 122,184,980 grams per day 134.69 tons per day
4.2.1.1.3. Data Set 2
Data Set 2 improved the input tree-flow travel speeds on 50 miles of arterial
roadways stratified between various area types within the region. The travel
demand forecasting model utilized the same vehicle tables as in the
Baseline and Data Set 1 scenarios to provide the identical distributed
vehicle volumes necessary to maintain consistency for CO emission
comparison purposes.
The improvements to the input free-flow speeds were increased on 50 miles
of arterial roadways in the critical and non-critical ranges. The critical speed
input included upgrading the free-flow speeds for arterial roadways in areas
70


CBD and CBD fringe. The non-critical speed input upgraded the arterial
roadways in the residential, suburban and rural areas. Both the modified
critical and non-critical data set improvements increased the input free-flow
travel speed 10 miles per hour. Tables 27 through 40 illustrate the VMT,
VHT, average speed, V/C ratios and CO emission rates for the Critical and
Non-Critical Data Set 2.
Table 27: Critical Data Set 2 VMT by Functional
Classification and Area Type
VMT units are Vehicle Miles
Functional
Classification 1 2
1 0 588948.3
2 0 4463.6
3 23476.1 87058.6
4 40894.7 42965.0
5 125.1 12790.0
6 0 18501.0
Area Type
3 4 5
1617670.1 192605.7 1595242.7
1205892.8 177066.8 517596.6
3098364.5 192319.8 1540863.4
838136.0 48206.9 995027.4
335708.1 70317.5 148385.3
92416.3 37704.4 48923.3
71


Table 28: Critical Data Set 2 VHT by Functional
Classification and Area Type
VHT units are Vehicle Hours
Area Type
Functional
Classification 1
1 0
2 0
3 937.9
4 1164.3
5 6.3
6 0
2 3
11784.6 31254.5
126.3 31623.6
3341.3 88442.2
1714.4 27698.7
578.3 11387.9
735.3 3690.6
4 5
3691.1 25468.5
4684.1 10736.4
6286.5 37644.1
1917.3 28431.2
2795.9 4946.2
1511.7 1953.5
72


Table 29: Critical Data Set 2 Average Speed by
Functional Classification and Area Type
Average Speed unit is miles per hour
Functional Area Type
Classification 1 2 3 4 5
1 0 49.98 51.76 52.18 62.64
2 0 35.34 38.13 37.80 48.21
3 25.03 26.06 35.03 30.59 40.93
4 35.12 25.06 30.26 25.14 35.00
5 19.86 22.12 29.48 25.15 30.00
6 0 25.16 25.04 24.94 25.04
73


Table 30: Critical Data Set 2 AM Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .09 .09 .07 .04
2 0 .07 .06 .08 .05
3 .02 .04 .06 .05 .06
4 .06 .04 .03 .02 .02
5 0 .01 .03 .03 .01
6 0 .07 .05 .07 .04
Table 31: Critical Data Set 2 PM Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .06 .05 .05 .03
2 0 .04 .04 .05 .03
3 .01 .02 .04 .03 .04
4 .04 .02 .02 .01 .01
5 0 .01 .02 .02 .01
6 0 .04 .03 .04 .02
74


Table 32: Critical Data Set 2 Off Peak Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .12 .12 .10 .06
2 0 .09 .08 .11 .07
3 .03 .05 .08 .07 .08
4 .08 .05 .04 .02 .03
5 0 .01 .03 .04 .02
6 0 .09 .06 .09 .05
75


Table 33: Critical Data Set 2 CO Emissions
AM Peak Period: 10,811,980 grams per day
11.92 tons per day
PM Peak Period: 13,475,940 grams per day 14.85 tons per day
OFF Peak Period: 98,685,960 grams per day 108.78 tons per day
TOTAL CO EMISSIONS: 122,973,880 grams per day 134.69 tons per day
76


Table 34: Non-Critical Data Set 2 VMT by Functional
Classification and Area Type
VMT units are Vehicle Miles
Functional
Classification 1 2
1 0 584636.5
2 0 4456.9
3 39563.7 102034.3
4 14303.1 33504.2
5 11.1 12102.6
6 0 18236.8
Area Type
3 4 5
1632303.8 197630.6 1589658.6
1605118.7 174953.2 520493.2
3143446.8 195896.1 1435212.4
838041.0 46292.3 993646.8
298263.6 65362.4 151435.1
92657.5 36611.7 48426.0
77


Table 35: Non-Critical Data Set 2 VHT by Functional
Classification and Area Type
VHT units are Vehicle Hours
Area Type
Functional
Classification 1
1 0
2 0
3 1581.4
4 646.0
5 0.6
6 0
2 3
11698.8 31545.3
126.2 31553.5
3916.2 88216.1
1334.1 27463.7
546.9 10654.4
724.8 3700.4
4 5
3791.4 25375.1
4620.7 10490.8
6236.8 37671.1
1934.0 28387.2
2737.4 5047.8
1468.5 1933.7
78


Table 36: Non-Critical Data Set 2 Average Speed by
Functional Classification and Area Type
Average Speed unit is miles per hour
Functional Area Type
Classification 1 2 3 4 5
1 0 49.97 51.74 52.13 62.65
2 0 35.32 50.87 37.86 49.61
3 25.02 26.05 35.63 31.41 38.10
4 22.14 25.11 30.51 23.94 35.00
5 18.50 22.13 27.99 23.88 30.00
6 0 25.16 25.04 24.93 25.04
79


Table 37: Non-Critical Data Set 2 AM Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .09 .09 .07 .04
2 0 .07 .06 .08 .05
3 .04 .05 .06 .05 .06
4 .02 .03 .03 .02 .02
5 0 .01 .02 .03 .01
6 0 .07 .04 .07 .04
Table 38: Non-Critical Data Set 2 PM Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .06 .06 .05 .03
2 0 .05 .04 .05 .03
3 .02 .03 .04 .03 .04
4 .01 .02 .02 .01 .01
5 0 .01 .01 .02 .01
6 0 .04 .03 .04 .02
80


Table 39: Non-Critical Data Set 2 Off Peak Volume to Capacity
Ratios (V/C)
Functional Area Type
Classification 1 2 3 4 5
1 0 .12 .12 .10 .06
2 0 .09 .08 .11 .08
3 .05 .06 .08 .07 .08
4 .03 .04 .04 .02 .03
5 0 .01 .03 .04 .02
6 0 .09 .06 .09 .05
81


Table 40: Non-Critical Data Set 2 CO Emissions
AM Peak Period: 10,729,380 grams per day
11.83 tons per day
PM Peak Period: 13,401,020 grams per day 14.77 tons per day
OFF Peak Period: 64,993,440 grams per day 71.64 tons per day
TOTAL CO EMISSIONS: 89,123,840 grams per day 98.24 tons per day
82


5. Conclusions
The CMAQ funding program focuses on projects which will enable the
nonattainment region to reach attainment according to the CAAA. As many
regions throughout the United States are utilizing CMAQ funds, it is critical
that a methodology be developed to ascertain the air quality benefits
received. The implementation of any CMAQ program is required to reduce
the mobile source pollutant emissions. However, the amount of pollutants
reduced will vary with each program.
A comprehensive implementation of a signal progression system through a
congested urbanized area will effect many travelers thus reducing a greater
percentage of the pollutant emissions. The public service provided through
the traffic flow improvements are far greater than those received with any
other form of transportation. Public transit, which is the second most
common form of transportation, serves a small segment of the urban
population. In the Colorado Springs urbanized area approximately 2% of
the population utilize public transportation as their mode of travel (14).
Increased service may improve the amount of persons served, however,
trends in alternative modes of transportation indicate a continuing decrease
in usage among carpooling, vanpooling, and public transit. Other modes of
transportation, including walking and bicycling, have even less of an impact
on the regional travel. Programs such as enhanced Inspection and
Maintenance for vehicles require the legislative implementation and
enforcement of regulatory policies. These legislative acts often receive
negative public input and therefore local agencies are reluctant to
implement such requirements. Thus, the most beneficial impact on regional
air quality is the traffic flow improvement programs.
83


The purpose of this thesis was to develop a methodology which may be
applied to a regional level to ascertain the air quality benefits received as a
result of the implementation of a traffic flow improvement program. Travel
demand forecasting models and the regional air quality emission model
(MOBILE5a) should be used to develop the tangible CMAQ program air
quality benefits. The usage of travel demand forecasting is an essential part
of the CO emission conformity determination. The travel paths in the
critical and non-critical cells are altered to reflect the increased travel
speeds on the links. Utilizing the traffic progression software does not alter
the selection of travel paths. However, the traffic progression software does
efficiently analyze the network queuing effects which would be needed in
congested urbanized areas.
Every input into the MOBILE5a model significantly weighs the CO pollutant
emission factor output. The distribution of vehicle types, the average age of
the vehicle fleets, the ambient temperature and the design year all create
important parameters in the conformity determination. As illustrated in
Figure 2 the distribution of the types of vehicles on the roadway can effect
the emission rates. Since the influence of the vehicle types can significantly
alter the emission factor rates, every effort should be made to ensure the
modeled scenario data represents the actual distribution of vehicles on the
road. Obtaining vehicle distribution information may be obtained through
vehicle registration and field surveys. Field surveys are necessary in order
to determine the vehicle distribution of external trips utilizing the regional
facilities.
The modeled design year is affected by the vehicle population thus
influencing the CO emission factor rates. Older vehicles (especially pre-
1980 model years) emit higher CO pollution levels than the newer vehicles.
Figure 7 illustrates the impact the turnover of vehicle fleets have on the CO
84


emission factors. The 2003 CO emission factors are lower than the factors
for the 1993 and 1995 planning years. The lower future CO emission factors
for future planning horizons assist the region in achieving conformity even
though the VMT is continuing to increase in future years. Although programs
have been developed nationally to remove the older higher polluting vehicle
fleets, the CMAQ program will not fund projects of that nature. The thesis
focused on the analyzing procedure of the model years as opposed to the
elimination of a particular type of vehicle.
Although all MOBILE5a inputs are critical to the evaluation of the CO
emissions, it is the average speed component which is the foundation of the
CMAQ conformity determination. In the calculation of the "tangible benefits,
the MOBILESa model will utilize the same input data for all three scenarios.
The vehicle types, ambient temperature and CO emission factors do not
change as a result of the transportation model. However, the transportation
model does recalculate the VMT, VHT and average speed of each scenario
as a result of the implementation of traffic flow improvements. The
comparison of the "Build", "No-Build", and Baseline scenarios only alter
the transportation characteristics of the MOBILESa model as a result of the
travel demand forecasting model. The emission analysis input remains
unaltered. Therefore, it is critical that the transportation networks be
enhanced to further the accuracy of the model output to ensure that the
information provided to the emissions models contain the accuracy needed
to provide an accurate analysis.
The formulation of the analysis methodology utilized the travel demand
forecasting model as input into the mobile emission model. The validation of
the methodology was conducted through a baseline or existing scenario and
two test scenarios, Data Set 1 and Data Set 2.
85


Data Set 1 revised a set of critical and non-critical travel speeds within a
single 5 mile corridor in the network. The Data Set 1 CO emission results
indicated that altering the travel speeds within a single small corridor
(arterial facility in the CBD) may improve the regional network CO emissions
over the Baseline and Non-Critical Data Set 1 scenarios. The Non-Critical
Data Set 1 improvements (expressway facility in the CBD fringe) created
only a slight shift in the VMT and a small improvement of travel speeds in the
network. The Non-Critical Data Set 1 scenario did not significantly improve
the regional transportation network and therefore did not alter the CO total
pollutant emissions from the baseline scenario.
Data Set 2 implemented traffic flow improvements on a large number of
facilities within the region. Again the data set was divided into two sets of
speed ranges; critical and non-critical. The CO emissions for the non-critical
range was lower than those within the critical range. The Non-Critical Data
Set 2 had lower CO emission rates for all time periods than the Critical Data
Set 2 and Baseline scenarios. This occurrence may be attributed to the fact
that the alterations to the free-flow input travel speeds reflected only the
critical and non-critical speed range criteria and did not reflect the
importance of VHT or VMT on the facility. The effect of altering a link in the
non-critical speed range which transports a large number of vehicles may
have an increased benefit over the altering of a critical link that carries a low
number of vehicles. The sample scenario represents a relatively
uncongested transportation network which may not alter the travel paths of
vehicles to improved links as critically as an severely congested network.
Figure 8 compares the Data Set 1 and Data Set 2 CO emissions for both the
critical and non-critical scenarios with the CO emissions from the Baseline
scenario. Transportation programs implemented must be able to show
regional pollutant levels below the Baseline and No-Build scenarios.
86


Conformity determination may be approved for any level of CO emission
improvement. Critical Data Set 1 indicated a 0.82 tons per day CO emission
reduction over the baseline scenario which meets the CAAA conformity
criteria. Non-Critical Data Set 1 did not show an improvement over the
Baseline scenario and therefore does not meet the CAAA conformity criteria.
Critical Data Set 2 was not able to show conformity for the traffic flow
improvements. Non-Critical Data Set 2 indicated conformity with the CAAA
for the implementation of the traffic flow improvement program.
The results of this thesis indicate that the critical'' travel speed is important in
the development of traffic flow improvement program strategies. However,
the travel speed is not an independent determining factor within the CO
emission conformity analysis. Additional variables must be reviewed to
determine the role the changes will have on the region. VMT and VHT on
the facilities within the network are crucial to the air quality CO emission
analysis. As shown in the methodology validation, the improvement of link
traffic flow within the critical data set must also reflect the areas with high
VMT and VHT in order to produce the most benefit to the region. The
implementation of the traffic flow improvements must include areas within
the critical travel speed range, and maintain high VMT and VHT volumes.
87


(fie FIGURE 8
DATA SCENARIO CO EMISSION RATES
Data Scenario
I AM Peak D PM Peak 1 Off Peak I Total
88


REFERENCES
(1) Godish, Thad, 1991. Air Quality. 247-251, Lewis, Michigan.
(2) U.S. Department of Transportation, Federal Highway Administration,
1992. A Summary: Intermodal Surface Transportation Efficiency Act of
1991.
(3) U.S. Department of Transportation, Federal Highway Administration,
1992. Further Guidance on the Congestion Mitigation and Air Quality
Improvement Program (CMAQ Program).
(4) U.S. Department of Transportation, Federal Highway Administration,
1994. A Guide to the Congestion Mitigation and Air Quality Improvement
Program. No. FHWA-PD-94-008, HEP-41/1-94(40M)E.
(5) THE URBAN ANALYSIS GROUP, 1993. URBAN/SYS User Manual
Supplement and Installation Instructions for the DOS and OS/2 Operating
Systems.
(6) COMSIS Corporation, 1992. MINUTP Travel Demand Forecasting
Training Seminar.
(7) U.S. Department of Transportation, Federal Highway Administration,
1986. "TRANSYT-7F Traffic Network Study Tool (Version 7F)."
(8) Yagar, Sam and Case, E.R., 1981. "Using TRANSYT for Evaluation."
The Application of Traffic Simulation Models. Special Report 194.
Transportation Research Board, National Academy of Sciences,
Washington, D.C.
89


(9) Hurley, Jamie W. Jr. and Radwan, Ahmed E., 1981. Traffic Flow
Simulation: User Experience in Research." The Application of Traffic
Simulation Models. Special Report 194, Transportation Research Board,
National Academy of Sciences, Washington, D.C.
(10) Dowling, Richard and Skabardonis, Alexander, 1992. Improving
Average Travel Speeds Estimated By Planning Models." Transportation
Research Board No. 1366, National Academy Press, Washington D.C.
(11) Denver Regional Council of Governments (DRCOG) 1994. Technical
Briefs: Signal Timing/Optimization Report.
(12) Shrouds, Jim, 1992. CMAQ Annual Reporting. U.S. Department of
Transportation, Federal Highway Administration.
(13) Heiken, J.G., Austin, B.S., Eisinger, D.S., Shepard, S.B., Duvall, L.L.,
1991. Estimating Travel and Emission Effects of TCMs." SYSAPP-91 /117.
(14) Pikes Peak Area Council of Governments, 1992. Regional Travel
Survey."
(15) Al-Deek, H., Wayson, R., and Radwan, A. E., 1995. Methodology for
Evaluating ATIS Impacts on Air Quality. Journal of Transportation
Engineering, 376-384.
90


Full Text

PAGE 1

APPLICATION OF TRAVEL SPEED RANGES IN AIR QUALITY MODELING by Katherine Marie Haire B.S., University of Missouri, 1985 A thesis submitted to the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Science Civil Engineering 1995

PAGE 2

This thesis for the Master of Science degree by Katherine Marie Haire has been approved by Bruce Janson Sarosh Khan Date

PAGE 3

Haire, Katherine Marie (M.S., Civil Engineering) Application of Travel Speed Ranges in Regional Air Quality Modeling Thesis Directed by Associate Professor Bruce Janson ABSTRACT The passage of the lntermodal Surface Transportation Efficiency Act (ISTEA) of 1991 and the Clean Air Act Amendments (CAAA) of 1990 interrelated transportation planning with air quality planning. Through the development of these two key legislative actions, federal funding was established to allow transportation projects to focus on measures which will assist an area in lowering the mobile source air pollutants emitted. The Congestion Mitigation and Air Quality Improvement (CMAQ) program funds are available to nonattainment areas for use in transportation projects which can demonstrate a "tangible" air quality benefit. Specific to the CMAQ funding category is a group of projects identified as traffic flow improvement programs. The implementation of these programs result in the reduction in congestion without providing additional lane mileage. As provided in the legislation, an analysis method must be developed to indicate the "tangible" air quality benefits received. The development of the analysis methodology focused on the characteristics of the transportation network and the associated levels of emitted carbon monoxide (CO) pollutants. Travel speeds identified in the travel demand forecasting model are divided into two subgroups, "critical" and "non-critical" travel speeds. The critical travel speeds are those which emit the higher

PAGE 4

levels of CO air pollution ( travel speeds less than 30 mph and greater than 55 mph). The noncritical travel speeds are those within the 30 to 50 mph range that emit a lower, steady level of CO air pollution. Improvements of the traffic flow through the critical range resulted in a "tangible" air quality benefit as indicated by the estimated emissions reduction for the traffic network analyzed in this thesis. The traffic flow improvements applied to the noncritical travel speeds resulted in no change in the air quality. It is concluded that "tangible" air quality benefits as required by the legislation will be possible or achieved only when traffic flow improvements occur in the critical travel speed range. This abstract accurately represents the content of the candidate's thesis. I recommend its publication.

PAGE 5

CONTENTS Chapter 1. Introduction 1.1 Legislation Clean Air Act Amendments lntermodal Surface Transportation Efficiency Act 1 2 3 6 1.1.1 1.1.2 1.1.2.1 1.1.2.1.1 1.1.2.1.2 1.2 Congestion Mitigation and Air Quality Improvement Program 6 CMAQ Transportation Programs CMAQ Reporting Requirements Purpose of Thesis 2. Review of Analysis Procedures 2.1 Travel Demand Forecasting 2.2 Travel Progression Analysis 2.3 Air Quality Analysis 3. Analysis 3.1 Traffic Operations 3.1.1 TRANSYT-7F 3.1.2 NETSIM 3.1.3. Traffic Operations Summary 3.2 Travel Demand Forecasting 3.2.1 TRANPLAN 3.2 .. 1.1 TRANPLAN Trip Generation 3.2.1.2 TRANPLAN Trip Distribution 3.2.1.3 TRANPLAN Modal Split 7 11 12 13 13 14 15 16 16 16 18 19 20 20 20 21 23

PAGE 6

3.2.1.4 TRANPLAN Assignment 23 3.2.1.5 TRANPLAN Speed Assignment 24 3.2.2 MINUTP 24 3.2 .. 2.1 MINUTP Trip Generation 27 3.2 .. 2.2 MINUTP Assignment Module 27 3.2.2.3 MINUTP Trip Distribution 28 3.2.2.4 MINUTP Speed Assignment 28 3.3 Air Quality Analysis 30 3.3.1 MOBILE5a 25 3.3.1.1 MOBILE5a Analysis Procedures 31 3.3.1.2 MOBILE5a Speed Considerations 32 4. Methodology 42 4.1 Analytical Procedures Utilized in Other Regions 42 4.1.1 DR COG 42 4.1.2 FHWA 43 4.1.3 SANDAG/CALTRANS 43 4.1.4 Summary of Existing Analysis Procedures 44 4.2 Recommended Analysis Procedure 44 4.2.1 Sample Validation 46 4.2.1.1 Methodology Validation 53 4.2.1.1.1 Critical Data Set 1 53 4.2.1.1.2. Non-Critical Data Set 1 62 4.2.1.1.3 Data Set 2 70

PAGE 7

5. Conclusions References 63 67

PAGE 8

List of Tables 1. National Ambient Air Quality Standards 2. Speed Table 3. Capacity Table 4. Baseline -VMT 5. Baseline VHT 6. Baseline Average Speed 7. Baseline-AM V/C 8. Baseline -PM V/C 9. Baseline Off Peak VIC 10. Baseline CO Emissions 11. Critical Data Set 1 Average Speed 12. Critical Data Set 1 -VMT 13. Critical Data Set 1 -VHT 14. Critical Data Set 1 -Average Speed 15. Critical Data Set 1 -AM V/C 16. Critical Data Set 1 -PM V/C 17. Critical Data Set 1 Off Peak V/C 18. Critical Data Set 1 CO Emissions 19. Non-Critical Data Set 1 -Average Speed 20. Non-Critical Data Set 1 -VMT 21 Non-Critical Data Set 1 -VHT 22. Non-Critical Data Set 1 -Average Speed 23. Non-Critical Data Set 1 -AM V/C 24. Non-Critical Data Set 1 -PM V/C 25. Non-Critical Data Set 1 -Off Peak VIC 26. Non-Critical Data Set 1 -CO Emissions 27. Critical Data Set 2 -VMT 28. Critical Data Set 2 -VHT 29. Critical Data Set 2 Average Speed 30. Critical Data Set 2 -AM V/C 31. Critical Data Set 2-PM V/C 32. Critical Data Set 2Off Peak V/C 33. Critical Data Set 2 CO Emissions 34. Non-Critical Data Set 2 -VMT 5 25 26 47 48 49 50 50 51 52 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78

PAGE 9

35. Non-Critical Data Set 2 -VHT 36. Non-Critical Data Set 2-Average Speed 37. Non-Critical Data Set 2-AM V/C 38. Non-Critical Data Set 2-PM VIC 39. Non-Critical Data Set 2Off Peak V/C 40. Non-Critical Data Set 2CO Emissions 78 79 80 80 81 82

PAGE 10

List of Figures 1. CMAQ Transportation Programs 10 2. Vehicle Classification Emission Factors 36 3. LDGT Emission Factors@ Varying Temperatures 37 4. LDGV Emission Factors@ Varying Temperatures 38 5. HDGV Emission Factors@ Varying Temperatures 39 6. HDDV Emission Factors @ Varying Temperatures 40 7. Program Year CO Emission Factors 41 8. Data Scenario CO Emission Rates 88

PAGE 11

1. Introduction Transportation projects have long been credited as the major contributor to the deterioration of the urban air quality. Recently, the passage of federal legislation has required the correlation between the implementation of transportation projects and the resulting air quality to be quantified. The enactment of the lntermodal Surface Transportation Efficiency Act (ISTEA) of 1991 and the Clean Air Act Amendments (CAAA) of 1990 embody the "cause and effect" philosophy requiring transportation planners to stretch beyond the previously established data boundaries. New procedures must be adopted that bridge databases to achieve an interactive system which can be used to evaluate the air quality impacts as they result from the implementation of specific operational transportation projects. Through the adoption of the ISTEA legislation, a special funding category was established to assist urbanized areas to reduce regional mobile source pollutant emissions. The Congestion Mitigation and Air Quality Improvement (CMAQ) program allows urbanized areas to receive funds to implement specified transportation programs providing "tangible" air quality benefits. CMAQ funding is divided into seven general categories including: 1) Transit Improvements 2) Shared-Ride Services 3) Traffic Flow Improvements 4) Demand Management Strategies 5) Pedestrian and Bicycle Programs 6) Inspection and Maintenance Programs 7) Other Air Quality Programs. 1

PAGE 12

The majority of the CMAQ categories address ways to reduce vehicle miles traveled (VMT) from the roadways. The calculation of "tangible" benefits as required for CMAQ funding is relatively straightforward for programs mainly designed to reduce VMT. However, other CMAQ programs that improve the network's operational characteristics (such as traffic flow improvements), as opposed to reducing VMT, require that additional procedures be used to quantify these benefits. Although the CMAQ funding program requires the calculation of "tangible" air quality benefits, the analysis procedure for traffic flow improvements is undefined. Traffic flow improvements include any program that improves the air quality through the reduction of congestion without providing additional lane mileage. The purpose of this thesis is to develop an analysis procedure which may be utilized to quantify the air quality benefits of projects within the Traffic Flow Improvement category of the CMAQ program. The air quality benefits associated with the traffic flow improvement programs will utilize the pollutant characteristics of carbon monoxide (CO) due to its direct correlation with automotive emissions. 1.1. Legislation Passage of the recent federal legislation has placed a greater emphasis on the need to understand the relationship between vehicle emissions, regional air quality, and the implementation of operational transportation improvements. This section focuses on two key legislative acts, the Clean Air Act Amendments (CAAA) of 1990 and the lntermodal Surface Transportation Efficiency Act (ISTEA) of 1991. 2

PAGE 13

1.1 :1 Clean Air Act Amendments The Clean Air-Act Amendments (CAAA) of 1990 are a culmination of legislation which began in 1955 with the passage of the initial air quality legislation (1 ). The 1955 legislation allowed the federal government to conduct air quality research and assist state and local agencies through assistance and technical support programs. Following the 1955 Air Quality legislation was the enactment of the Clean Air Act of 1963. The Clean Air Act of 1963 allowed the expansion of the federal governments air quality role to provide funding, conduct research, enforce interstate air pollution regulations, and develop pollutant level criteria to maintain public health and welfare. The Air Quality Act of 1967 was an attempt to strengthen the Nation's air pollution control effort. The 1967 Act designated an Air Quality Control Region for every major metropolitan area, issued air quality criteria and control technique information. While the 1967 Act mandated criteria which must be met within a required time frame, the states were responsible for developing the analysis and enforcement procedures. As air quality continued to deteriorate, the public concern for the environment intensified resulting in the passage of the 1970 Clean Air Amendments. The 1970 Clean Air Amendments established the U.S. Environmental Protection Agency (U.S. EPA) as an independent federal agency. In addition, the amendments included setting the National Ambient Air Quality Standards (NAAQS), designating Air Quality Control Regions (AQCR), establishing the guidelines by which AQCR's must develop Statewide Implementation Plans, setting automotive emission and fuel 3

PAGE 14

standards, and enabling federal enforcement in air pollution emergencies and interstate and intrastate air pollution violations. The 1977 Clean Air Act Amendments set to soften the stringent regulation set forth in the 1970 amendments. Time frames were extended to allow additional time to achieve these goals. A most notable subject aspect of the legislation is its enabling of the EPA to regulate the chemicals that destroy the ozone layer. The effort to maintain control over the pollutants emitted into the environment continued into the next decade. The 1990 Clean Air Act Amendments (CAAA) utilize the National Ambient Air Quality Standards (NAAQS) to establish baseline thresholds for seven NAAQS "criteria" pollutants which all areas in the U.S. must achieve. The seven NAAQS "criteria" pollutants are: 1) Carbon Monoxide (CO) 2) Ozone (03 ) 3) Nitrogen Dioxide (N02 ) 4) Particulate Matter less than 10 microns (PM-1 0) 5) Sulfur Dioxide (S02) 6) Hydrocarbons 7) Lead. Table 1 shows the corresponding maximum threshold levels for these seven "criteria" pollutants. Areas which violate the NAAQS thresholds for any one of the criteria pollutants is designated as a "nonattainment" area. A nonattainment area may be designated for a single pollutant (i.e. Carbon Monoxide) or multiple pollutants (i.e. Carbon Monoxide, Ozone and Particulate Matter less than 10 microns). Areas which are designated "nonattainment" must reach attainment for the violated criteria pollutant within the time frame stipulated 4

PAGE 15

by the CAAA legislation. Failure to do so may risk the delay or loss of federal transportation funding for some categories of highway projects. Table 1 National Ambient Air Quality Standards [1, p.251] Pollutant Averaging Time Primary Standard Carbon Monoxide 8 hr 10 mgtm3 (9 ppm) Nitrogen Dioxide 1 hr 40 mgtm3 (35 ppm) Sulfur Dioxide Annual Average 1 00flg/m3 (0.05 ppm) 24 hr 365f.tg/m3 (0.14 ppm) PM10 Annual Arithmetic Mean 50 flg/m 3 24 hr 150 flg/m3 Hydrocarbons 3 hr 160 Ozone 1 hr 235 flQ/m3 (0.12 ppm) Lead 3 Month Average 1.5 flQ/m3 5

PAGE 16

1.1.2 lntermodal Surface Transportation Efficiency Act (ISTEA) of 1991 Passage of the lntermodal Surface Transportation Efficiency Act (ISTEA) of 1991 promised to change the focus of transportation. The purpose of the Act is "to develop a National lntermodal Transportation System that is economically efficient, environmentally sound, provides the foundation for the Nation to compete in the global economy and will move people and goods in an energy efficient manner." (2) The ISTEA legislation establishes a National Highway System, encourages innovative transportation technologies, encourages public-private partnerships, and promotes highway safety. One goal of the ISTEA legislation is to maximize the usage of the existing transportation facilities prior to the construction of any new single occupancy vehicle facilities. 1.1.2.1 Congestion Mitigation and Air Quality Improvement Program (CMAQ) Enacted by the ISTEA legislation, the Congestion Mitigation and Air Quality Improvement Program (CMAQ) was developed to promote innovative strategies to reduce the levels of pollutants emitted from mobile sources. The purpose of the CMAQ program is "to fund transportation projects or programs that will contribute to attainment of national ambient air quality standards (NAAQS) with a focus on Ozone and Carbon Monoxide. Under certain conditions, transportation projects and programs targeting Particulate Matter less than 10 microns are also eligible."(3) The CMAQ program was developed to assist in the reduction of the pollutant levels emitted from mobile sources. Mobile sources, such as automobiles, are primarily responsible for emitting two pollutants, Ozone (03 ) and Carbon 6

PAGE 17

Monoxide (CO). Therefore, these two pollutants are the focus of CMAQ program funds. However, since PM-1 0 can be attributed to mobile source emissions, funds may be selectively used for PM-1 0 CMAQ programs. The CMAQ program is designed to assist nonattainment areas in reaching attainment or keep an attainment area from becoming a nonattainment area for CO, Ozone or PM-1 0 pollutants. 1.1.2.1.1. CMAQ Transportation Programs Nonattainment areas are required to develop strategies to achieve attainment within the time period established by the CAAA. The development of attainment strategies are initiated by the metropolitan planning organization (MPO) in consultation with the Environmental Protection Agency (EPA), the Federal Transit Administration (FTA), the Federal Highway Administration (FHWA), and the State. Transportation Control Measures (TCMs) are a key group of strategies that can be proposed to achieve attainment. The TCMs may be used as emission credits in the Statewide Implementation Plan (SIP). The SIP combines regional emission control strategies, as well as statewide strategies, to assure that the State as a whole also reaches attainment within the designated time frame if it is in question. CMAQ funds enable States to introduce transportation projects and programs that will achieve the required air quality standards within the designated time frame. Effective use of CMAQ funds requires a strong planning effort for two reasons: (1) limited funding requires obtaining maximum cost/benefit return from these expenditures, and (2) the substantial commitment of the personnel to develop the level of analysis necessary to illustrate a "tangible" air quality benefit. 7

PAGE 18

Transportation programs eligible to receive CMAQ funding include the following as described in (4): Transit Improvements to improve the air quality by enhancing the existing transit service and providing service to more people. -Shared-Ride Services to improve air quality by shifting people from their single occupancy vehicle (SOV) to high occupancy vehicles. Traffic Flow Improvements that improve air quality by reducing congestion without adding lane mileage. -Demand Management Strategies which develop strategies, techniques or programs to reduce the demand for SOV travel. Demand management strategies improve air quality by reducing vehicle miles traveled and vehicle trips through the implementation of alternate transportation strategies. -Pedestrian and Bicycle Programs that improve air quality by making these "zero emission" viable modes of transportation. -Inspection and Maintenance Programs that improve air quality by reducing the emissions of the Nation's vehicle fleet. -Other air quality programs that promote public outreach and education, promote promising new technology to reduce emissions or promote the conversion of public vehicle fleets to alternative fuels (under certain conditions). 8

PAGE 19

As defined in the legislation, four programs are excluded from receiving CMAQ funds. These four excluded programs are: efforts to reduce emissions for extreme cold-start conditions; -encouragement of the removal of pre-1980 vehicles; increased road capacity for SOV's; and maintenance costs for existing systems. All requests for the CMAQ funds must be coordinated through the MPO. The MPO is the key planning agency in the nonattainment area for both transportation and air quality. The MPO is responsible for the development of the Transportation Improvement Program (TIP) which includes the federally funded highway and transit projects and TCM's found in the State Implementation Plan (SIP). All projects that are funded under the CMAQ program must be included in the TIP. The TIP must be in conformance with the SIP. As appropriate, these programs may be approved by the EPA as TCMs included in the SIP and receive emission reduction credits. All TCM's included in the SIP that qualify for emission reduction credits shall receive the highest priority for funding as designated by CMAQ regulations. 9

PAGE 20

FIGURE 1 (4, p. 5) TRANSPORTATION PROGRAMS 20% D Transit I SharedRide I T raffle Flow D Demand Mgt ml Ped & Bike 1J VM 10

PAGE 21

1.1.2.1.2. CMAQ Reporting Requirements Federal legislation requires the recipients of the CMAQ funds to submit an annual report that details the level of expenditures on each program as well as the air quality benefits realized as a result of the program. Though the CMAQ funding regulations require tangible air quality benefits be calculated as a result of the traffic flow improvements, the development of the methodology used in the analysis is left to the local planning agencies. The federal agencies are allowing the local agencies to provide their own analysis procedures providing they are consistent with accepted practice. This is acceptable due to the varying levels of technical expertise and analytical approaches utilized throughout the United States. A single method of analysis would be difficult to adopt at a Federal level due to the localized control of applications. Therefore, the prescribed analysis method is to be determined through consultation with the Federal Transit Administration (FTA), Federal Highway Administration (FHWA), the U.S. Environmental Protection Agency (EPA), the State and MPO's. Annual CMAQ reports are produced by the regional planning organization and submitted to the FTA, FHWA, EPA, and State. FTA, FHWA and EPA review and accept the reports as received or recommend additional analyses. As stated in the CMAQ guidelines (3), the annual report shall contain all projects funded under the CMAQ program, the dollar amount utilized in the program, and the estimated tangible air quality benefits for that year. 11

PAGE 22

1.2 Purpose of Thesis The requirement of transportation planners to analyze performance characteristics of the transportation network not found by traditional procedures requires the refinement of these procedures. The refined analysis procedure will enable the user to work between transportation and air quality databases so as to estimate the tangible air quality benefits. The purpose of this thesis is to present a methodology that links the transportation analysis procedures to the air quality analysis procedures in a manner that meets the analysis requirement for CMAQ funding of proposed traffic flow improvements. In order to relate air quality benefits to traffic flow improvements, it is critical to identify the relationship which exists between traffic operations and the resulting mobile source air pollutant emissions. The implementation of the traffic flow improvement projects increase the travel speeds on the network through improved progression. Therefore, the analysis procedure must have the ability to relate the resulting travel speed variance as it relates to the air quality pollutant emissions. 12

PAGE 23

2. REVIEW OF ANALYSIS PROCEDURES To gain an understanding of the influence that transportation infrastructure (or supply) has on regional air quality, it is necessary to review both transportation analysis procedures as well as air quality analysis procedures. Transportation analysis procedures are typically divided into two general categories, transportation planning and traffic operations. The analyses conducted specific to the CMAQ traffic flow improvements category focus on travel demand forecasting models and signal coordination procedures. The air quality analysis focuses on the MOBILE5a model that is used to estimate regional emissions based on traffic volumes and speeds. 2.1 Travel Demand Forecasting Travel demand forecasting is critical to the assessment of future transportation needs within a region. Travel demand forecasting occurs through the assistance of a computer generated model that performs the four-step transportation planning process. The four-step process consists of Trip Generation, Trip Distribution, Modal Split, and Assignment. Trip generation estimates the number of trips that will travel to and from an area based on the land use characteristics of that area. These areas are called traffic analysis zones into which the region is divided. Trip distribution estimates the numbers of trips between zones given the production of trips from and attraction of trips to each zone, and the travel cost between each pair of zones. Modal split divides trips between user specified modes of transportation (i.e., public transit or automobile). The assignment step loads alternative routes of the network with the trips of the various modes. 13

PAGE 24

Travel demand forecasting models estimates the network performance characteristics necessary to formulate the input required for the air quality model (MOBILE5a). Travel speed are estimated for the network on a link by link basis. The output of MOBILE5a summarizes the average speeds categorized by area type and functional classification. The travel demand forecasting model assigns trips to alternative routes of the network according to link travel times along these routes. Link travel times are developed by applying the assigned link volumes to the capacity and length of each link. The process of developing these travel time (and thus speeds) is inherent to travel demand forecasting model, and can be somewhat different between models. The following sections focus on two travel demand forecasting models (TRANPLAN and MINUTP) and their use of travel speeds. 2.2 Travel Progression Analysis The efficient progression of traffic through an urbanized area maintains the level of service for the transportation network, extends the service life of the facility and is beneficial to the community. Travel time is increased through the reduction of stops and delays which improves the network's level of service. As the progression through the system improves, the flow capacity through the facility is increased allowing more vehicles to travel on the system. The community benefits through overall environmental improvements including increased fuel efficiency and the reduction of carbon monoxide (CO) emissions. The development of a coordinated traffic signal system requires the interconnection of the network traffic signals throughout the region in order to provide communication. Once the signal system is interconnected, it is 14

PAGE 25

necessary to develop signal timing parameters that optimize the progression of vehicles. This section identifies two operational traffic optimization programs: TRANSYT-7F and NETSIM. 2.3 Air Quality Analysis Areas which are designated "nonattainment" must certify that all transportation projects within the region do not contribute to a deterioration in the air quality. The air quality analysis is conducted through a "Baseline" versus "Build" versus "No-Build" test. The transportation projects must emit less pollution emissions in the "Build" scenario then in the "Baseline" and "No-Build" scenarios. The conformity tests utilize the transportation network output as input into the air quality.software. For regional emission calculations, the MOBILE5a software authorized by the EPA is the required air quality software. 15

PAGE 26

3. Analysis Conducting air quality analysis requires the input of vehicle volumes and speeds in order to predict the mobile source emissions. Obtaining vehicle volume and speed data for an exiting transportation network can occur through field observations or detailed traffic operations models. However, obtaining this type of detailed information for future scenarios on a regional level is not feasible. This section focuses on the analysis considerations for utilizing the traffic flow software, the travel demand forecasting software, and the air quality analysis software. 3.1 Traffic Operations Traffic flow or progression software can be effectively utilized to determine the widest band width allowable for a given grid network. This band width establishes the opportunity for the platoon of vehicles to travel downstream while maintaining a constant rate of speed thus eliminating or reducing the vehicle stops or delays and reducing the vehicle queuing. 3.1.1 TRANSVT-7F TRANSYT-7F (7) analyzes corridor operational characteristics to optimize the progression of vehicles traveling through a series of signalized intersections. The software conducts platoon dispersion analysis of the distribution of traffic over time at specified locations in the network. TRANSYT-7F applies an internal algorithm to distribute the platoon of traffic evenly downstream. 16

PAGE 27

The network optimization is simulated for a described transportation network through a given cycle length and phasing pattern. The performance index (PI) is determined by the network analysis of the traffic signal offsets and cycle splits based on the weighted sum of stops and delays. A series of iterations are conducted utilizing varying offsets and splits to calculate the resulting performance index. Input into the program include the traffic signal parameters (i.e. signal timing, cycle length and signal phasing), the platoon characteristics, and the roadway characteristics. TRANSYT-7F requires a common cycle length for all of the traffic signals in the network. The analysis output includes the speed of the vehicles and the delay experienced by the vehicles. The flow characteristics of the upstream traffic signal and the dispersion of traffic as the flow moves downstream are utilized to determine the performance index. The TRANSYT-7F program outputs data which may be utilized to improve network efficiency and assist in evaluation procedures. The data and measures of effectiveness resulting from the TRANSYT-7F analysis by link volume include; saturation flow, degree of saturation, total travel, travel time, delay, stops, fuel consumption, maximum back of queue, and green times. In addition, signal timing tables for each intersection with the offset from the main controller is given. The flow of the platoon through the network may be shown graphically using time-space diagrams. TRANSYT-7F has proven very sensitive to the speed set within the modei.(S) Therefore, it is critical that the speed input be estimated as accurately as possible. The optimization process determines the optimal traffic signal offsets based on the given link speeds (or travel times). Incorrect speed inputs would cause the TRANSYT-7F to loose a degree of accuracy in the optimization process by calculating incorrect offsets. 17

PAGE 28

TRANSYT-7F has the ability to determine environmental effects associated with the progression analysis including fuel consumption and resulting pollutant emissions. The following equations represent the macroscopic relationship used in the program to calculate environmental impacts. Rate of P-roduction (Fuel)= AeBv Cv Rate of Production (CO emissions)= 3.3963 ( e0.014561vt1 ooov) where: v = velocity A, B, C = user specified constants 3.1.2 NETSIM NETSIM is a traffic model utilized to evaluate operational characteristics of urban street networks (9). The model uses node and link information to input the individual vehicles entering network. Each vehicle is then stochastically assigned a set of performance characteristics (i.e. vehicle type, average discharge headway, average acceptable gap, etc.) on the network. The vehicle proceeds through the network controlled by the assigned performance characteristics, microscopic car-following, queue discharge, lane-switching algorithms and by the assigned link turn percentages. NETSIM may be utilized to analyze a variety of traffic controls such as "STOP", "YIELD", actuated or pretime traffic signals. NETSIM output includes speed and delay measures of effectiveness as well as fuel consumption and pollutant emissions for each vehicle type. 18

PAGE 29

3.1.3 Traffic Operations Summary The traffic flow software is recommended for determining the optimal signal timing for a series of signalized intersections or a grid network of intersections. The methodology developed must be able to be utilized on a regional level. TRANSYT-7F does not analyze vehicle behavior. Thus the increase in travel speeds on the network would not cause the vehicles to alter their travel path from origin to destination to account for the improved conditions. The software assumes that all vehicles will continue on the same path in a more efficient manner. In actuality, vehicles will continually shift their path to achieve the shortest route possible. This is an important consideration as transportation facilities are upgraded or constructed. Traffic flow software requires detailed operational level input. This information is limited in the existing condition and becomes a process of estimation in future conditions. Estimations of future operational conditions will lessen the accuracy of the optimization model. Traffic flow software would not prove to be the best method of analysis for regional transportation networks. The limited size of the network evaluation, as well as the lack of operational data, does not accurately represent future regional scenarios. In addition, the traffic flow software does not account for shifts of the vehicle patterns. As facilities are upgraded or constructed, traffic will tend to gravitate toward the improved facilities, thus it is important to reflect this trend in the modeling. 19

PAGE 30

3.2 Travel Demand Forecasting Future conditions and "what if" scenarios must be developed through the travel demand forecasting models. Each transportation network alternative developed in the model will v,ary the assignment and distribution process. The travel demand software calculates the change in vehicle paths that are attracted to an upgraded or new facility. 3.2.1 TRANPLAN TRANPLAN (5) is a travel demand forecasting model which may be used to analyze and evaluate the highway network. The TRANPLAN model consists of a series of routines which allows the user to customize the development of the model to reflect the analytical needs of the area. Following the four step planning methodology, TRANPLAN establishes the trip generation, trip distribution, modal split, and assignment based upon the highway network. The highway network is initially created through an external program which contains the x and y coordinates for the nodes and the characteristics of the links (transportation facilities). Transportation characteristics found for each highway network link include its functional classification, area type, hourly flow capacity, link length, posted speed limit, and number of lanes. 3.2.1.1 TRANPLAN Trip Generation Trip generation data is developed through a process which analyzes factors associated with trip making patterns as a reflection of the regional characteristics. Data obtained through travel surveys may be utilized to supply specific regional characteristics. 20

PAGE 31

Each traffic analysis zone (TAZ) in the region is characterized by socioeconomic conditions. Based upon the socioeconomic characteristics of each zone, an estimation of the number of trips is determined. Trip generation determines the number of trips attracted to and produced by each TAZ. The socioeconomic characteristics (i.e., number of households, number of employees, number of persons per household, household income level, etc.), and the quantity by purpose of trips made are generated for each type of lands. The trip generation rates made be determined through travel studies or surveys, or through the Institute of Transportation Engineers (ICE) manual. 3.2.1.2 TRANPLAN Trip Distribution TRANPLAN generates a zone-to-zone trip table utilizing a Gravity Model based upon the classification of the trip (i.e. trip purpose), the travel impedance factors, zone-to-zone travel indices, and optional K-factors. The Gravity Model was developed under the guidance of Newton Gravitational Law (See Equation 1) which assumes that all "trips starting from a given zone are attracted by various traffic generators in other zones and that this attraction is directly proportional to the relative attraction of the zone and inversely proportional to the separation between the zones in the gravity model ."(5) Typically, the travel time between network zones serve as the measure of separation described in the gravitational formula. Equation 1: Classical Gravitational Formula Force (f) = mass of body 1 (m 1) mass of body 2 (m2) gravitational (distance separating m1 and m2)2 constant 21

PAGE 32

The Classical Gravitation Formula must be restructured in order to fulfill the needs of travel demand forecasting. This restructuring must incorporate the characteristics of a transportation network. The restructured gravitational formula is shown in Equation 2. Equation 2: Restructured Gravitational Formula: T( ') PiAiFtn i\Kti i\ I,J 11u.-lhu l: Ex(i,x)K(i,x) where: T (i,j) is the trips produced in zone i and attracted to zone j (analogous to gravitational force); Pi is the trips produced in zone i; AI is the trips attracted to zone j; Ft(i,j) is the empirically derived travel time factor that expresses the average area-wide effect of spatial separation on trip interchange between zones that are t(i,j) apart; t(i,j) is the travel time in minutes between zone i and zone j; and K(i,j) is the specific zone to zone adjustment factor to allow for the incorporation of the effect of spatial separation on trip interchange between zones that are t(i,j) apart. A fundamental requirement of the transportation model is that the calculated trip attractions balance. This procedure is conducted utilizing the ratio of the 22

PAGE 33

computed attractions versus the theoretical attraction calculation based upon a district or zonal level. The required input is a skim table which contains the interzonal impedances to be used ,in the determination of the friction factor selection. The output from the gravity model is a trip table file containing the zone-to-zone distributed trips for up to 15 trip purposes. During the trip production and attraction process, the focus is on the travel time occurring on each link. As the trips are attracted to a zone, the trip is influenced by the travel time experienced on the link. The travel time on the link directly correlates to the speed of the facility, the capacity of the facility, and the length of the link. 3.2.1.3 TRANPLAN Modal Split The modal split module is not discussed within this analysis as the highway elements are the primary concern in the deriving of the speed on the link. 3.2.1.4 TRANPLAN Assignment TRANPLAN has the ability to incorporate three types of assignment loading. The options for the assignment loading include: All-or-Nothing assignment; Restraint Loading; or Incremental Loading. The following paragraphs discuss each assignment type. Restraint Loading places the selected interzonal highway trips onto the minimum paths of the highway network. The network uses the parameter of time, adjusted link by link, according to the curve data either specified as a volume-to-capacity (V/C) time adjustment curve data or the default Bureau of Public Roads (BPR) capacity restraint formula. The speed of the facility is a reflection of the travel time impedance calculated on each link. 23

PAGE 34

Incremental loading is performed for each iteration. A user-specified percentage of selected interzonal highway trips is loaded on the minimum paths during path building. Time is adjusted link by link according to user specified V/C speed adjustment curve data or the BPR capacity restraint formula. AU-or-Nothing assignment loads all selected interzonal highway trips on the minimum paths based upon time, distance, cost, or other user specified impedance. This is a free flow assignment with no consideration given to the type of links or link capacities. In the Restraint Loading and the Incremental Loading for each assignment group, minimum and maximum speeds may be specified which forces the model to set the adjusted speed within the range. 3.2.1.5 TRANPLAN Speed Assignments The distribution factors inherent within the program utilize travel time as the determining factor. The speed of the facility is a reflection of the travel time impedance calculated on the link. TRANPLAN allows the input of individual link speeds of the unloaded highway network. Once the network is loaded, the speeds become a reflection of the impedance on the facilities. 3.2.2 MINUTP The MINUTP (6) travel demand forecasting model is an interactive software package based on the four step planning process: trip generation; trip distribution; mode choice; and trip assignment. This section describes the development of the MINUTP model and the speed assignment process. 24

PAGE 35

The transportation network is built utilizing the X/Y coordinates combined with the unbuilt network link card file. NETBLD creates the highway network by reading the ASCII link data records that describes the highway network. Link speeds can be coded directly or through the use of a look up table based on link speed class (See Table 2). Link capacity is provided to the model by a table stratified by link capacity codes and the number of lanes (See Table 3). Table 2: Speed Table [6, p.12] SPEED CLASS (mph) Facility Type Freeway Expwy Arterial Collector Centroid 1 a ;! Area Type 1. CBD 48(11) 37(12) 22(13) 20(14) 10(16) 2. CBD Fringe 48(21) 44(22) 29(23) 25(24) 15(25) 3. Residential 67(31) 47(32) 32(33) 28(34) 15(35) 4. OuterCBD 58(41) 37(42) 24(43) 22(44) 15(45) 5. Rural 67(51) 47(52) 32(53) 28(54) 15(55) 25

PAGE 36

Table 3: Capacity [Speed in mph (Service Volume in vehicles per hour per lane)] [6, p.14] AREA TYPE LEVEL OF CBD FRINGE URBAN SUBURB RES SERVICE Freeway 36 40 42 42 45 (1400) (1400) (1400) (1400) (1400) (1400) c (1750) (1750) (1750) (1750) (1750) (1750) E F A Expressway 24 27 30 40 40 55 c (600) (600) (600) (600) (700) (700) c (840) (840) (885) (1500) (1500) (1500) E L I Principal 22 23 28 36 38 42 T Arterial (600) (600) (600) (600) (600) (600) c y (840) (840) (840) (885) (885) (885) E T Minor 22 23 24 27 33 42 y Arterial (350) (350) (350) (400) (400) (400) c p (560) (560) (560) (650) (650) (885) E E Collector 15 20 20 20 25 30 (350) (10000) (350) (350) (350) (350) c (500) (10000) (550) (550) (550) (550) E 26

PAGE 37

3.2.2.1 MINUTP Trip Generation Trip generation is developed through the production and attraction files developed for the three trip purposes (home based work, home based other, and non-home based) assigned to the traffic analysis zone (TAZ). The trip generation file is combined with the impedance file and friction factor file to yield the zone to zone matrix which contains the production and attraction person trips by purpose. The zone to zone production and attraction files are converted to origin and destination files by applying car occupancy factors, adding through movements and additional tables. The built highway network, the origin and destination file, and the turn penalty file are input into the assignment model. The resulting loaded link network is utilized to compute the vehicle miles traveled (VMT) on the network. 3.2.2.2 MINUTP Assignment Module The MINUTP assignment module, ASSIGN, determines the travel time between zones and then assigns the zone-to-zone trip values to the network links along the paths. The assignment process may utilize the All or Nothing, the All Shortest Path, or the Stochastic methodologies. Any of the assignment methodologies may be iterative or incremental when combined with capacity restraint. Additionally, the volume and/or equilibrium may be adjusted through the iterative assignment. PTHBLD reads the network description and determines the minimal travel paths from each zone centroid node to each other zone centroid node. The paths selected are the minimum impedance paths based upon each link's time and distance. Optionally, turn penalties may be assessed at selected intersections. 27

PAGE 38

3.2.2.3 MINUTP Trip Distribution TRPDST reads production and attraction, impedance, and friction factor data sets and applies the standard gravity model distribution equation for each zonal pair in the study area. This is done individually for each of the purposes. TRPDST iterates a selected number of times in an attempt to obtain correct attraction totals within each zone. On the last iteration it writes out trip matrices containing estimated trips for each zonal pair for each phase. 3.2.2.4 MINUTP Speed Assignments The introduction of speed assignments within the MINUTP model occurs in the PTHBLD module. In this module, the speed categories may be established in a link format (i.e. actual input of the speed data) or the link speed look up table which references the speed by roadway functional classification and area type. Congested link speeds are generated through the assignment model where the zone-to-zone trips are assigned to the network. The travel demand software is built and calibrated based on the vehicle volumes on the links. The model accurately determines the demands on the facilities, however, the speeds of the facilities are seldomly verified. Once the transportation network achieves the calibrated vehicle volumes, the average speeds are not checked for reasonableness (1 0). Planners assume that the modeled average speed on the network is accurate if the vehicle volumes are calibrated. The travel demand forecasting models typically use a speed-flow curve such as the BPR curve to estimate the congested travel speed given the initial 28

PAGE 39

free-flow speed and the volume/capacity ratio (V/C). The standard equation for the BPR curve is: congested speed = free-flow speed (1 +0.15*V/C4) Planning models may estimate demands over the specified capacity of the individual facility. As this does not occur in actuality and therefore there are no observed speed-flow curves for V/C ratios over 1.00. As a result, the user defined speed-flow curve or the default BPR curve may be applied inappropriately when the volumes meet or exceed the capacity. As the network approaches a VIC ratio close to 1.00, the conditions become unstable. This unstable condition creates difficulty in predicting the average speed on the facility. An error in speed estimation is critical as VIC ratios approach 1.00. At that point, the congestion causes the vehicles to slow to very low speeds thus exponentially increasing the air pollutant emission rates. The speed-flow curves utilized in the travel demand forecasting models do not account for the effects of queuing on travel speeds and demand. As a result, the calculations of average speed may exceed those actually on the roadway. Planning models estimate the average speed on the network through the flow and capacity data for each link. The travel demand forecasting models provide the best level of estimation for the future network comparisons. 29

PAGE 40

3.3 Air Quality Analysis 3.3.1 MOBILE5a MOBILE5a is the air quality software authorized by the EPA to be utilized in nonattainment areas to ascertain the regional pollution levels emitted from mobile sources. MOBILE5a contains three input sections; the control section, the data section and the scenario section. The control section contains the portion of the input data that regulates the input, output, and execution of the program. Control elements include the inspection and maintenance program, additional input data, output emission factors for visual inspection, and output formatted for further analysis in another program. The data section contains one-time application information pertaining to ALL scenarios. The data section defines parameter values different from those internal to the program which will be used in the calculations of all scenarios within a given run. Examples of data section input include the annual mileage accumulation rates, registration distribution by age, and further control program parameters (i.e. description of inspection and maintenance program}. The Scenario section details the individual scenarios for which emission factors are to be calculated. These include the calender year of evaluation, average speed to assume, and the highor low-altitude region. 30

PAGE 41

3.3.1.1 MOBILESa Analysis Procedures The transportation network for each scenario is developed and formatted for input into the MOBILE5a model. The MOBILE5a model generates a series of modules to obtain the all-day pollutant emissions rates for each scenario. The focus within this analysis is on the pollutant carbon monoxide (CO) due to the direct correlation with vehicle exhaust. Initially, setup information obtained from the environmental regulatory agency is applied to the transportation network to create the mobile source pollutant emissions. Representative setup information includes: regional vehicle registrations; inspection and maintenance programs; anti-tampering programs; oxygenated fuels; Reid vapor pressure; average minimum and maximum temperatures; regional vehicle mix by area type and functional classification for a.m., p.m., and off peak periods; and the adjustment factors to convert the peak period data (VMT mix and operating modes) into an all day average. The transportation network output serves as input to an average speed model that generates the am, pm, and off peak periods stratified by area type and functional classification for low and normal operating speed ranges. The average speed data files contain the transportation network data file which will form one of the required input to later calculate the total carbon monoxide emissions attributed to the network. Once the network speeds have been determined the scenario data file must be prepared for the forecast year. The all-day operating modes and ambient temperature are loaded with the average speeds into the final forecast year scenario data. 31

PAGE 42

The control data section, including the one-time data records, is added to the scenario data file to initiate the MOBILESa program. The emission factors attributed to the network are taken from the output of the MOBILESa. Total emission rates for the forecast year scenario are then calculated. The MOBILESa program allows for the modification of input providing concurrence is received from the EPA, the FHWA, the FTA, and the State. An agency may wish to alter the input of the average speed so as not to stratify the VMT in the low and high speed categories. This may reflect the actual air quality of the region as the low and high speed ranges contribute greater amounts of CO pollutants then the mid range speeds. It is important to prevent the dilution of the low and high speed categories into the mid range speeds for the pollutant emission calculations. 3.3.1.2 MOBILES a Speed Considerations The average speed input is one of the most critical inputs into the MOBILESa program. The development of the CO emission rates are based upon the application of the mobile source pollutant emission parameters onto the average speeds in the network. The MOBILESa model focuses on the pollutant emissions of mobile sources. The mobile sources are typically divided into four categories: Light Duty Gasoline Vehicles (lDGV); Light Duty Gasoline Trucks (LDGT); Heavy Duty Gasoline Vehicles (HDGV); and Heavy Duty Diesel Vehicles (HDDV). This section reviews CO pollutant levels for various scenarios. Figure 2 illustrates the CO emission factors as a reflection of vehicle classifications and travel speeds. The heavy duty diesel vehicles (HDDV) emit the lowest level of CO pollutants at all speeds, approaching zero 32

PAGE 43

emissions between 40 and 50 mph. The light duty gasoline vehicles (LDGV) emit the second lowest level of CO pollutants, followed closely by the light duty gasoline trucks (LDGT). The heavy duty gasoline vehicles (HDGV) emit the highest level of CO pollution, significantly greater than the other three categories. The CO pollutant emission rates vary by the vehicle type and ambient temperature. As illustrated in Figures 3 through 6, the ambient temperature alters the emission rates from the mobile sources. Figure 3 illustrates the LDGV emission factors at three varying temperatures. At all three temperature levels the CO emission factors are high at speeds below 10 mph, decreasing rapidly as speeds reach 30 mph, becoming relatively stable between 30 and 50 mph and increasing slightly as speeds become greater than 55 mph. The coldest temperature modeled (20F) emits the highest levels of CO pollutants at all travel speeds. The warmest temperature modeled (60F) exhibits the lowest levels of CO pollutants emitted at all travel speeds. The mid-range modeled temperature ( 40F) emits pollutants less than the 20F conditions and more than the 60F conditions. The temperature variance creates significant differences in the modeled pollutant levels at travel speeds less than 10 mph. As the three temperature curves approach the stable conditions of 30 to 50 mph, the gap between the pollutant levels narrows. The LDGT emission factors, illustrated in Figure 4, indicate a similar set of curves to those described above. The CO emission factors are slightly higher for all three temperatures in the LDGT graph than in the LDGV graph. Similarly, the widest margin between the CO pollutant emission rates occurs in the low travel speed range of 0 to 10 mph. The margin between the temperature curves narrows as speeds approach 30 mph, remaining narrow through speeds of 50 mph and widening after travel speeds of 55 mph are 33

PAGE 44

reached. The lowest illustrated temperature curve (20F) exhibits the highest CO emission levels whereas the highest temperature curve (60F) exhibits the lowest levels of CO emission factors. The HDGV graph, illustrated in Figure 5, projects the highest levels of CO pollutant emission factors as compared to the other vehicle types for all temperature levels and travel speeds. The CO pollutant emission curve, as described by the travel speed and ambient temperature, peaks at travel speeds less than 10 mph. The CO pollutant emission curve decreases as travel speeds approach 30 mph. Between 30 and 50 mph, the emission factors remain relatively stable. As travel speeds exceed 50 mph, the CO pollutant emission factors gradually begin to climb. Similar to the other graphs, the lower temperature levels emit the higher CO pollutant emissions level. Figure 6 illustrates the overall lowest CO pollutant emission levels of the four vehicle types. All three temperatures project the same CO emission factors within each travel speed. The CO emission factors peak at travel speeds under 10 mph, remaining stable between 30 and 50 mph and increasing over 50 mph. The MOBILE5a CO emission rates are subject to change in order to reflect the changing conditions. The emission rates are affected by the characteristics of the vehicle fleet including the turnover of vehicle fleets and the improvement of automobile technology. The turnover of vehicle fleets as well as the improvement of automobile technology result in lower levels of CO emissions. Figure 7 illustrates the variance in pollutant emission factors for three planning horizons. The projected CO emission factors are expected to be 34

PAGE 45

less in the planning year 2003 than they are in the year 1995 and 1993. As expected, the improvement in technology decreases the levels of CO pollutants emitted. As older vehicles are phased out of service and replaced with new vehicles, the average CO emission rate by vehicle distribution decreases. Again the CO pollutant emission factors are at their highest as speeds are less than 1 0 mph. The CO emission factors continue to decrease gradually until travel speeds reach 55 mph and then begin to rise. The variance between the CO emission factors narrow as the average speed increases. 35

PAGE 46

FIGURE 2 (13) 350 300 !: G) = 250 I E 200 Cl' s 1so 11.1 I Iii Q II. 100 u 50 0 Vehicle Classification Emission Factors 0 ln 0 lJj C) ... ... N N tl1 tl1 V Speed (mph) I UXJV D ux;r I HDGV II Hoov 36

PAGE 47

FIGURE 3 (13) LDGT tMISSION rACTORS @ VARYING TtMPtRATURtS 1.11 0 If) .,.. .,.. Speed I LOGT D LOGT I LOGT 37

PAGE 48

FIGURE 4 (13) LDGV EMISSION @ VARYING TEMPERATURES ,.. ,.. 0 IJ) N N 0 IJ) 0 IJ) I() I() \j' \j' Speed (mpn) 0 IJ) 0 IJ) ll1 ll1 ID ID I LDGV D LDGV I LDGV 38

PAGE 49

FIGURE 4 (13) HDGV tMISSION @ VARYING TEMPERATURES J 10 1 J lO lJ JO JJ 60 6J IHDGV lOf D HDGV IHDGV sor 39

PAGE 50

FIGURE 6 (13) HDDV @ J 10 1 J 30 3J 40 4J JO JJ 60 6J Speed I HDDV D HDDV 40f I HDDV 60f 40

PAGE 51

FIGURE 7 (15) =n" I 0 100 .. u Ill U e e IW "-= 0 'Y 0 PROGRAM YEAR CO 10 (mpn) 40 1 CO D 1998 CO CO 41 JO 60

PAGE 52

I 4. METHODOLOGY 4.1 Analytical Procedures Utilized in Other Regions A review of the existing analysis procedures utilized in the United States was conducted to formulate a basis for comparison. Unfortunately, this is a relatively new area of analysis and limited data is available. 4.1.1. DRCOG The Denver Regional Council of Governments (DRCOG) utilizes the corridor specific methodology to analyze each signal timing/optimization project (11). The traffic signal coordination plans are developed for each of the three time periods: am peak period (6:30a.m. through 8:30a.m.); p.m. peak period (3:00p.m. through 6:30p.m.); and the off-peak (period not included in the a.m. or the p.m. peak periods). Progression analysis is conducted on the corridor to establish the optimal signal cycle length. Once the optimal signal cycle length is implemented along the corridor, analysis is conducted to ascertain the benefits achieved. The before and after scenarios are compared as to the Travel Time (seconds), the Stopped Time (seconds) and the Travel Speed (miles per hours). In addition, the environmental benefits are calculated as to the reduction in Fuel Consumption (gallons) and the Pollutant Emissions (Carbon Monoxide, Hydrocarbons, and Nitrous Oxides). The analysis is conducted only on the progression on the major street. Analysis is not conducted on the minor cross streets. 42

PAGE 53

4.1.2. FHWA An outline approach, developed by the FHWA (13) to determine the emission reductions by the CMAQ programs, has been provided to the MPO's to assist in the development of an analysis procedure. The outline allows the CMAQ programs and projects to be analyzed individually or as a group of projects which are in the same category (i.e. transit) or area (i.e. transportation corridor). The steps in the analysis procedure are as follows: 1) estimate the emissions per trip in grams for each criteria pollutant (using an emissions model approved by EPA which provides the most recent estimate of emissions); 2) multiplied by the estimated number of daily trips reduced for each criteria pollutant (preferably using a methodology consistent with the methodology used or proposed to determine the air quality benefits of transportation control measures in an approved State Implementation Plan revision); 3) equals the estimated emission reductions in grams per day for each criteria pollutant; and 4) divided by 1000 equals the estimated emission reduction in kilograms per day for each criteria pollutant. 4.1.3. SANDAG/CALTRANS Evaluating the Traffic Flow Improvement Programs for the SANDAG/CALTRANS area are conducted through user inputs of speed 43

PAGE 54

changes (13). The California Air Resource Board analyzes traffic signalization improvements by speed changes determined from traffic flow modeling. The traffic flow model requires the user to input network data (links and nodes), saturation flows, traffic volumes, cruise speeds, bus stop delays and traffic signal data. 4.1.4. Existing Analysis Procedure Summary The three procedures differ from one another within analysis techniques. The DR COG method reviews corridor impacts of the transportation implementation. This analysis supports a before and after process that may only be implemented for the immediate time period. Future scenarios would need to project the operational characteristics whereby loosing a degree of accuracy. The FHWA method requires the implementation of the transportation project to reduce the number of trips on the network. Traffic flow improvements represent the same number of vehicles on the roadway and therefore do not fit into this analysis procedure. The SANDAG/CALTRANS method fits the requirements for the analyzing of the traffic flow improvements. This procedure models the impacts of the speed changes through traffic flow modeling. 4.2 Recommended Analysis Procedure The CO emission factor rates (illustrated in Figure 6) show very little variance between 30 and 50 mph. Therefore, it is determined that the CO pollutant emission factors will remain relatively stable within this range. Projects which are implemented within the 30 to 50 mph range will be unable to show a significant "tangible" benefit for CO pollutant emissions. 44

PAGE 55

Utilizing this hypothesis as a foundation, a methodology was developed to formulate a "tangible" solution. The following steps outline a strategy for determining the benefits associated with the implementation of the traffic flow enhancement CMAQ program. Step 1: Step 2: Step 3: Step 4: Step 5: Step 6: Step 7: Step 8: Run the travel demand forecasting model for the existing transportation system prior to the implementation of the CMAQ program. Calculate the resulting average speed as a function of the VMT and VHT output from the travel demand forecasting model by area type and functional classification. Identify the functional classification and area type of transportation facilities that are not in the 30 to 50 mph range. Determine if the cell (functional class by area type) is within the range or outside the range. If within the range advance to Step 5. If outside the speed range, advance to Step 8. Speed within the 30 to 50 mph range, therefore conduct localized corridor specific analysis. Build corridor in traffic progression analysis program. Conduct localized corridor air quality analysis. Speed outside of the 30 to 50 mph range, therefore regionally analysis. 45

PAGE 56

Step 9: Step 10: Step 11: Step 12: Determine the revised travel speed through the aid of traffic flow progression analysis or field surveys. Input the revised speed for the appropriate links into the regional travel demand model. Calculate the revised mobile source emissions utilizing MOBILE5a. Compare the "Build" versus "No Build" versus "Baseline" scenarios. 4.2.1 Sample Validation A sample transportation network was utilized to review the methodology presented within this thesis. The sample network consists of 350 traffic analysis zones and 4413 links covering approximately 1 ,350 square miles. The loaded transportation network contains 216,666 total trips. Network calibration is based on the comparison of link volumes to actual vehicle volumes at cordon locations. Area Type and Functional Classification categories reflect those discussed in the MINUTP capacity table. The purpose of this example is to forecast traffic on a given network, develop comparative networks to confirm, through application, that the analysis procedure is valid .. The transportation network was generated through the four step planning process. The "Baseline" scenario reflects the existing modeled conditions that future model modifications will be compared to. Table 4 and Table 5 illustrate the VMT and VHT respectively for the "Baseline" scenario. 46

PAGE 57

Table 4: Baseline VMT by Functional Classification and Area Type VMT units are Vehicle Miles Functional Area Type Classification 1 g_ 1 0 581863.8 1627630.4 196779.8 1590869.7 2 0 4443.6 1212576.1 174741.8 517580.5 3 39637.3 103037.3 3133095.7 193086.7 1540968.2 4 14303.2 33384.7 839519.5 47204.2 995243.9 5 11.1 12152.1 299651.3 65186.0 151275.0 6 0 18106.6 92871.6 36529.5 48884.4 47

PAGE 58

Table 5: Baseline VHT by Functional Classification and Area Type VHT units are Vehicle Hours Functional Area Type Classification 1 g_ .4. 1 0 11633.5 31449.8 3771.9 25399.9 2 0 125.9 31803.8 4615.2 10736.1 3 1584.4 3954.7 89433.1 6436.3 37644.2 4 646.0 1329.4 27975.5 1972.1 28437.4 5 0.6 549.1 10703.6 2729.5 5042.5 6 0 719.7 3709.0 1465.1 1951.8 Average speed for the "Baseline" scenario must be calculated for each area type and functional classification. This may be achieved through the arithmetic calculation of VMT divided by VHT. The average speed calculated for the baseline scenario is contained in Table 6. 48

PAGE 59

Table 6: Baseline Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 g_ 1 0 50.02 51.75 52.17 62.63 2 0 35.29 38.13 37.86 48.21 3 25.02 26.05 35.03 30.00 40.94 4 22.14 25.11 30.01 23.94 35.00 5 18.50 22.13 28.00 23.88 30.00 6 0 25.16 25.04 24.93 25.05 The baseline scenario volume to capacity ratios (V/C) calculated for the AM, PM and Off peak periods are shown in Tables 7,8, and 9 respectively. 49

PAGE 60

Table 7: Baseline -AM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ 1 0 .09 .09 .07 .04 2 0 .07 .06 .08 .05 3 .04 .05 .06 .05 .06 4 .02 .03 .03 .02 .02 5 0 .01 .02 .02 .01 6 0 .07 .05 .07 .04 Table 8: Baseline -PM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ i 1 0 .06 .05 .05 .03 2 0 .04 .04 .05 .03 3 .02 .03 .04 .03 .04 4 .01 .02 .02 .01 .01 5 0 .01 .01 .02 .01 6 0 .04 .03 .04 .02 50

PAGE 61

Table 9: Baseline Off Peak Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g 1 0 .12 .12 .10 .06 2 0 .09 .08 .11 .07 3 .05 .06 .08 .07 .08 4 .03 .04 .04 .02 .03 5 0 .01 .03 .04 .02 6 0 .09 .06 .09 .05 The VMT, VHT and average speed for the baseline scenario are entered into the MOBILE5a model. Table 10 lists the estimated CO emissions for the baseline scenario. 51

PAGE 62

Table 10: Baseline CO Emissions AM Peak Period: PM Peak Period: OFF Peak Period: 10,850,870 grams per day 11.96 tons per day 13,542,100 grams per day 14.93 tons per day 97,740,460 grams per day 107.74 tons per day TOTAL CO EMISSIONS: 122,133,430 grams per day 134.63 tons per day 52

PAGE 63

4.2.1.1. Methodology Validation As illustrated in Figure 2, the Carbon Monoxide emission rates remain relatively constant between 30 mph and 50 mph. Therefore, any change in the average speed for areas within that range should not create additional regional CO pollutant emissions nor should the change create a beneficial situation. To verify the methodology, two sample data sets were utilized. The first data set (Data Set 1) modified a small number of facilities within the network whereas, the second data set (Data Set 2) modified a larger amount of facilities within the region. In both data sets, the transportation network was run prior to any modifications to determine which area types and functional classifications fell within and outside of the critical speed range. Individual analysis was performed on four separate cases: Critical Data Set 1; Non Critical Data Set 1; Critical Data Set 2; and Non-Critical Data Set 2. The critical and non-critical title refers to whether the modified input free-flow speeds are within the critical range or within the non-critical range. The methodology will compare the two modified data sets with the initial baseline data set to determine the impacts on the regional CO air quality. Initially, a small corridor will be altered (Data Set 1) with higher input free flow speeds in the critical and noncritical areas. The second data set (Data Set 2) will modify a larger section of input free-flow speeds within the critical and noncritical areas. Comparisons of the two data sets will be made to the "Baseline" scenario for resulting CO air quality performance. 4.2.1.1.1 Critical Data Set 1 The initial phase is to ascertain which links will benefit from travel speed 53

PAGE 64

improvements. The best use of the funding is to apply traffic flow improvements to the links which will improve the network travel speeds thus benefiting the regional air quality. Links identified as providing the most benefit for improving regional air quality are the critical travel speeds found in Table 11. The first modification of Data Set 1 focused on the "critical" speed ranges occurring beyond the 30 to 50 mph range. Table 11 shows the "critical" average speeds to be incorporated into Critical Data Set 1. As the travel speeds on the critical links are improved, a corresponding improvement in the regional air quality should result. The improvement to the regional air quality will reflect the degree of improvements on the roadway. The areas which are excluded from consideration (******)are the areas with travel speeds that fall within the 30-50 mph and therefore maintain a constant CO pollutant emission range. The free-flow input speeds are altered on a five mile corridor and the vehicle tables are held constant. The network is rerun and allowed only to recalculate the assignment process. The Vehicle Miles Traveled (VMT) and Vehicle Hours Traveled (VHT) are recalculated for each facility type and area type. 54

PAGE 65

Table 11: Critical Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 1 1 0 ...... ...... ..... ****** 2 0 ****** ****** 3 25.02 26.05 4 22.14 25.11 23.94 5 18.50 22.13 28.00 23.88 ****** 6 0 25.16 25.04 24.93 25.05 The new link input speeds may be obtained through traffic flow progression analysis or observed field surveys. For illustrative purposes, one cell (functional classification and area type} is chosen as a representative sample to determine the impact an improvement to a single corridor would have on the regional network. Critical Data Set 1 will revise the travel speeds on a 5 mile corridor that is located in the CBD (area type 1) and functionally classified as a principal arterial (facility type 3). The overall input speed will be increased 10 mph along the corridor. The travel demand forecasting model is rerun with the revised travel speed link data input. The vehicle trip tables remain constant and the revised VMT 55

PAGE 66

and VHT are output. Table 12 and Table 13 illustrate the VMT and VHT respectively for the Critical Data Set 1 analysis. The calculated Critical Data Set 1 average speeds are found in Table 14. Table 12: Critical Data Set 1 Scenario VMT by Functiona' Classification and Area Type VMT units are Vehicle Miles Functional Area Type Classification 1 g i Q. 1 0 590613.8 1632385.5 195581.4 1591000.2 2 0 4476.0 1214082.9 177229.7 517317.9 3 24187.7 88382.9 3124055.3 191422.8 1541340.2 4 39666.0 43062.1 837065.7 47165.0 995222.5 5 125.1 12902.8 297227.6 64796.2 1s12n.9 6 0 18633.3 93000.1 37338.7 48896.9 56

PAGE 67

Table 13: Critical Data Set 1 VHT by Functional Classification and Area Type VHT units are Vehicle Hours Functional Area Type Classification 1 g_ 1. 1 0 11830.4 31542.0 3750.0 25402.0 2 0 126.6 31843.3 4687.2 10727.9. 3 966.4 3391.9 89174.9 6380.8 37654.5 4 1133.1 1719.3 27893.7 1970.5 28436.9 5 6.3 583.4 10617.7 2713.8 5042.6 6 0 740.6 3714.0 1497.6 1952.4 57

PAGE 68

Table 14: Critical Data Set 1 Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 g_ 1 0 49.92 51.75 52.16 62.87 2 0 35.36 38.13 37.81 48.22 3 25.03 26.06 35.03 30.00 40.93 4 35.01 25.05 30.01 23.94 35.00 5 19.86 22.12 27.99 23.87 30.00 6 0 25.16 25.04 24.93 25.04 58

PAGE 69

Critical Data Set 1 calculated volume to capacity ratios (V/C) are shown in Tables 15, 16, and 17 for the AM, PM and Off Peak Periods respectively. Table 15: Critical Data Set 1 -AM Volume to Capacity Ratios {V/C) Functional Area Type Classification 1 g_ 1 1 0 .09 .09 .07 .04 2 0 .07 .06 .08 .05 3 .02 .04 .06 .05 .06 4 .06 .04 .03 .02 .02 5 0 .01 .02 .02 .01 6 0 .07 .05 .07 .04 59

PAGE 70

Table 16: Critical Data Set 1 -PM Volume to Capacity Ratios {V/C) Functional Area Type Classification 1 g i 1 0 .06 .06 .05 .03 2 0 .04 .04 .05 .03 3 .02 .02 .04 .03 .04 4 .04 .02 .02 .01 .01 5 0 .01 .01 .02 .01 6 0 .04 .03 .04 .02 60

PAGE 71

Table 17: Critical Data Set 1 Off Peak Volume to Capacity Ratios (VIC) Functional Area Type Classification 1 g_ i 1 0 .12 .12 .10 .06 2 0 .09 .08 .11 .07 3 .03 .05 .08 .07 .08 4 .08 .05 .04 .02 .03 5 0 .02 .03 .04 .02 6 0 .09 .06 .09 .05 The Critical Data Set 1 average speeds are combined with the transportation network as input into the MOBILE5a model. The resulting CO emissions are shown in Table 18. 61

PAGE 72

Table 18: Critical Data Set 1 -CO Emissions AM Peak Period: PM Peak Period: OFF Peak Period: 10,844,900 grams per day 11.95 tons per day 13,516,250 grams per day 14.90 tons per day 97,033,420 grams per day 106.96 tons per day TOTAL CO EMISSIONS: 121,394,570 grams per day 133.81 tons per day 4.2.1.1.2 Non-Critical Data Set 1 The second phase of the validation is to illustrate the impact changes have within the 30-50 mph range have on the regional air quality. It is hypothesized that altering the non-critical average speeds within the 30 mph to 50 mph range would not significantly alter the regional CO emissions. Therefore, Non-Critical Data Set 1 selects an area which has a non-critical average travel speed. Table 19 illustrates the average travel speeds which fall within the stable CO pollutant emission factor range. 62

PAGE 73

Table 19: Non-Critical Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 2. i 1 ....... 50.02 51.75 52.17 62.63 2 ****** 35.29 38.13 37.86 48.21 3 ...... 35.03 30.00 40.94 4 35.00 5 ..... 30.00 6 Non-Critical Data Set 1 chose the area within the CBD Fringe (area type 2) and functionally classified as an expressway (facility type 2). A 5 mile section of roadway was selected within this area and the travel speed input was increased 10 miles per hour. Once again the travel demand forecasting model was run to generate the revised VMT and VHT. Tables 20 and 21 illustrate the VMT and VHT respectively. The average travel speeds for Non Critical Data Set 1 are shown in Table 22. 63

PAGE 74

Table 20: Non-Critical Data Set 1 VMT by Functional Classification and Area Type VMT units are Vehicle Miles Functional Area Type Classification 1 g_ 1 1 0 581607.9 1626755.2 194973.7 2 0 4464.5 1203965.3 174174.0 3 39244.1 104229.6 3165235.1 191133.1 4 14083.0 33228.4 832181.3 47084.6 5 11.1 12101.2 293809.3 65039.0 6 0 18073.9 92883.1 36382.1 64 1591032.6 517353.4 1539406.5 995646.2 151210.3 48880.2

PAGE 75

Table 21: Non-Critical Data Set 1 VHT by Functional Classification and Area Type VHT units are Vehicle Hours Functional Area Type Classification 1 g i 1 0 11638.6 31432.6 3737.7 2 0 126.5 31578.1 4600.5 3 1568.7 4000.2 89680.2 6371.0 4 636.0 1323.3 27934.4 1967.2 5 0.6 546.9 10496.2 2723.4 6 0 718.5 3709.3 1459.1 65 25402.5 10731.5 37604.4 28449.0 5040.4 1951.7

PAGE 76

Table 22: Non-Critical Data Set 1 Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 g_ 1 0 49.97 51.75 52.16 62.63 2 0 35.29 38.13 37.86 48.21 3 25.02 26.06 35.29 30.00 40.94 4 22.14 25.11 30.01 23.93 35.00. 5 18.50 22.12 27.99 23.88 30.00 6 0 25.16 25.04 24.93 25.04 The AM, PM and Off Peak volume to capacity ratios (V/C) for the Non-Critical Data Set 1 scenario are shown in Tables 23, 24, and 25 respectively. 66

PAGE 77

Table 23: Non-Critical Data Set 1 -AM Volume to Capacity Ratios (V /C) Functional Area Type Classification 1 g 1 Q. 1 0 .09 .09 .07 .04 2 0 .07 .06 .08 .05 3 .04 .05 .06 .05 .06 4 .02 .03 .03 .02 .02 5 0 .01 .02 .02 .01 6 0 .07 .05 .07 .04 67

PAGE 78

Table 24: Non-Critical Data Set 1 -PM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ 1 1 0 .06 .05 .05 .03 2 0 .05 .04 .05 .03 3 .02 .03 .04 .03 .04 4 .01 .02 .02 .01 .01 5 0 .01 .01 .02 .01 6 0 .04 .03 .04 .02 68

PAGE 79

Table 25: Non-Critical Data Set 1 Off Peak Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 l 1 1 0 .12 .12 .10 2 0 .09 .08 .11 3 .05 .06 .08 .07 4 .03 .04 .04 .02 5 0 .01 .03 .04 6 0 .09 .06 .09 The Non-Critical Data Set 1 average travel speeds as well as the transportation network are input into the MOBILESa model. The resulting CO emissions are shown in Table 26. 69 2 .06 .07 .08 .03 .02 .05

PAGE 80

Table 26: Non-Critical Data Set 1 CO Emissions AM Peak Period: PM Peak Period: OFF Peak Period: 10,848,600 grams per day 11.96 tons per day 13,543,020 grams per day 14.93 tons per day 97,793,360 grams per day 107.80 tons per day TOTAL CO EMISSIONS: 122,184,980 grams per day 134.69 tons per day 4.2.1.1.3. Data Set 2 Data Set 2 improved the input free-flow travel speeds on 50 miles of arterial roadways stratified between various area types within the region. The travel demand forecasting model utilized the same vehicle tables as in the Baseline and Data Set 1 scenarios to provide the identical distributed vehicle volumes necessary to maintain consistency for CO emission comparison purposes. The improvements to the input free-flow speeds were increased on 50 miles of arterial roadways in the critical and non-critical ranges. The critical speed input included upgrading the free-flow speeds for arterial roadways in areas 70

PAGE 81

CBD and CBD fringe. The non-critical speed input upgraded the arterial roadways in the residential, suburban and rural areas. Both the modified critical and non-critical data set improvements increased the input free-flow travel speed 10 miles per hour. Tables 27 through 40 the VMT, VHT, average speed, V/C ratios and CO emission rates for the Critical and Non-Critical Data Set 2. Table 27: Critical Data Set 2 VMT by Functional Classification and Area Type VMT units are Vehicle Miles Functional Area Type Classification 1 g 1 0 588948.3 1617670.1 192605.7 2 0 4463.6 1205892.8 177066.8 3 23476.1 87058.6 3098364.5 192319.8 4 40894.7 42965.0 838136.0 48206.9 5 125.1 12790.0 335708.1 70317.5 6 0 18501.0 92416.3 37704.4 71 1595242.7 517596.6 1540863.4 995027.4 148385.3 48923.3

PAGE 82

Table 28: Critical Data Set 2 VHT by Functional Classification and Area Type VHT units are Vehicle Hours Functional Area Type Classification 1 g_ 1 1 0 11784.6 31254.5 3691.1 2 0 126.3 31623.6 4684.1 3 937.9 3341.3 88442.2 6286.5 4 1164.3 1714.4 27698.7 1917.3 5 6.3 578.3 11387.9 2795.9 6 0 735.3 3690.6 1511.7 72 25468.5 10736.4 37644.1 28431.2 4946.2 1953.5

PAGE 83

Table 29: Critical Data Set 2 Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 2. Q. 1 0 49.98 51.76 52.18 62.64 2 0 35.34 38.13 37.80 48.21 3 25.03 26.06 35.03 30.59 40.93 4 35.12 25.06 30.26 25.14 35.00 5 19.86 22.12 29.48 25.15 30.00 6 0 25.16 25.04 24.94 25.04 73

PAGE 84

Table 30: Critical Data Set 2 -AM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ a .Q. 1 0 .09 .09 .07 .04 2 0 .07 .06 .08 .05 3 .02 .04 .06 .05 .06 4 .06 .04 .03 .02 .02 5 0 .01 .03 .03 .01 6 0 .07 .05 .07 .04 Table 31: Critical Data Set 2 -PM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ a .Q. 1 0 .06 .05 .05 .03 2 0 .04 .04 .05 .03 3 .01 .02 .04 .03 .04 4 .04 .02 .02 .01 .01 5 0 .01 .02 .02 .01 6 0 .04 .03 .04 .02 74

PAGE 85

Table 32: Critical Data Set 2 Off Peak Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 2. 1 0 .12 .12 .10 .06 2 0 .09 .08 11 .07 3 .03 .05 .08 .07 .08 4 .08 .05 .04 .02 .03 5 0 .01 .03 .04 .02 6 0 .09 .06 .09 .05 75

PAGE 86

Table 33: Critical Data Set 2 CO Emissions AM Peak Period: PM Peak Period: OFF Peak Period: 10,811,980 grams per day 11.92 tons per day 13,475,940 grams per day 14.85 tons per day 98,685,960 grams per day 108.78 tons per day TOTAL CO EMISSIONS: 122,973,880 grams per day 134.69 tons per day 76

PAGE 87

Table 34: Non-Critical Data Set 2 VMT by Functional Classification and Area Type VMT units are Vehicle Miles Functional Area Type Classification 1 g 1 0 584636.5 1632303.8 197630.6 2 0 4456.9 1605118.7 174953.2 3 39563.7 102034.3 3143446.8 195896.1 4 14303.1 33504.2 838041.0 46292.3 5 11.1 12102.6 298263.6 65362.4 6 0 18236.8 92657.5 36611.7 77 1589658.6 520493.2 1435212.4 993646.8 151435.1 48426.0

PAGE 88

Table 35: Non-Critical Data Set 2 VHT by Functional Classification and Area Type VHT units are Vehicle Hours Functional Area Type Classification 1 g_ 1 0 11698.8 31545.3 3791.4 2 0 126.2 31553.5 4620.7 3 1581.4 3916.2 88216.1 6236.8 4 646.0 1334.1 27463.7 1934.0 5 0.6 546.9 10654.4 2737.4 6 0 724.8 3700.4 1468.5 78 25375.1 10490.8 37671.1 28387.2 5047.8 1933.7

PAGE 89

Table 36: Non-Critical Data Set 2 Average Speed by Functional Classification and Area Type Average Speed unit is miles per hour Functional Area Type Classification 1 g i Q. 1 0 49.97 51.74 52.13 62.65 2 0 35.32 50.87 37.86 49.61 3 25.02 26.05 35.63 31.41 38.10 4 22.14 25.11 30.51 23.94 35.00 5 18.50 22.13 27.99 23.88 30.00 6 0 25.16 25.04 24.93 25.04 79

PAGE 90

Table 37: Non-Critical Data Set 2 -AM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ 1 0 .09 .09 .07 .04 2 0 .07 .06 .08 .05 3 .04 .05 .06 .05 .06 4 .02 .03 .03 .02 .02 5 0 .01 .02 .03 .01 6 0 .07 .04 .07 .04 Table 38: Non-Critical Data Set 2 -PM Volume to Capacity Ratios (V/C) Functional Area Type Classification 1 g_ 1 0 .06 .06 .05 .03 2 0 .05 .04 .05 .03 3 .02 .03 .04 .03 .04 4 .01 .02 .02 .01 .01 5 0 .01 .01 .02 .01 6 0 .04 .03 .04 .02 80

PAGE 91

Table 39: Non-Critical Data Set 2 Off Peak Volume to Capacity Ratios (V/C} Functional Area Type Classification 1 g_ 1 0 .12 .12 .10 .06 2 0 .09 .08 .11 .08 3 .05 .06 .08 .07 .08 4 .03 .04 .04 .02 .03 5 0 .01 .03 .04 .02 6 0 .09 .06 .09 .05 81

PAGE 92

Table 40: Non-Critical Data Set 2 CO Emissions AM Peak Period: PM Peak Period: OFF Peak Period: 10,729,380 grams per day 11.83 tons per day 13,401,020 grams per day 14.77 tons per day 64,993,440 grams per day 71.64 tons per day TOTAL CO EMISSIONS: 89,123,840 grams per day 98.24 tons per day 82

PAGE 93

5. Conclusions The CMAQ funding program focuses on projects which will enable the nonattainment region to reach attainment according to the CAAA. As many regions throughout the United States are utilizing CMAQ funds, it is critical that a methodology be developed to ascertain the air quality benefits received. The implementation of any CMAQ program is required to reduce the mobile source pollutant emissions. However, the amount of pollutants reduced will vary with each program. A comprehensive implementation of a signal progression system through a congested urbanized area will effect many travelers thus reducing a greater percentage of the pollutant emissions. The public service provided through the traffic flow improvements are far greater than those received with any other form of transportation. Public transit, which is the second most common form of transportation, serves a small segment of the urban population. In the Colorado Springs urbanized area approximately 2% of the population utilize public transportation as their mode of travel (14). Increased service may improve the amount of persons served, however, trends in alternative modes of transportation indicate a continuing decrease in usage among carpooling, vanpooling, and public transit. Other modes of transportation, including walking and bicycling, have even less of an impact on the regional travel. Programs such as enhanced Inspection and Maintenance for vehicles require the legislative implementation and enforcement of regulatory policies. These legislative acts often receive negative public input and therefore local agencies are reluctant to implement such requirements. Thus, the most beneficial impact on regional air quality is the traffic flow improvement programs. 83

PAGE 94

The purpose of this thesis was to develop a methodology which may be applied to a regional level to ascertain the air quality benefits received as a result of the implementation of a traffic flow improvement program. Travel demand forecasting models and the regional air quality emission model (MOBILE5a) should be used to develop the tangible CMAQ program air quality benefits. The usage of travel demand forecasting is an essential part of the CO emission conformity determination. The travel paths in the "critical" and "non-critical" cells are altered to reflect the increased travel speeds on the links. Utilizing the traffic progression software does not alter the selection of travel paths. However, the traffic progression software does efficiently analyze the network queuing effects which would be needed in congested urbanized areas. Every input into the MOBILE5a model significantly weighs the CO pollutant emission factor output. The distribution of vehicle types, the average age of the vehicle fleets, the ambient temperature and the design year all create important parameters in the conformity determination. As illu$trated in Figure 2 the distribution of the types of vehicles on the roadway can effect the emission rates. Since the influence of the vehicle types can significantly alter the emission factor rates, every effort should be made to ensure the modeled scenario data represents the actual distribution of vehicles on the road. Obtaining vehicle distribution information may be obtained through vehicle registration and field surveys. Field surveys are necessary in order to determine the vehicle distribution of external trips utilizing the regional facilities. The modeled design year is affected by the vehicle population thus influencing the CO emission factor rates. Older vehicles (especially pre1980 model years) emit higher CO pollution levels than the newer vehicles. Figure 7 illustrates the impact the turnover of vehicle fleets have on the CO 84

PAGE 95

emission factors. The 2003 CO emission factors are lower than the factors for the 1993 and 1995 planning years. The lower future CO emission factors for future planning horizons assist the region in achieving conformity even though the VMT is continuing to increase in future years. Although programs have been developed nationally to remove the older higher polluting vehicle fleets, the CMAQ program will not fund projects of that nature. The thesis focused on the analyzing procedure of the model years as opposed to the elimination of a particular type of vehicle. Although all MOBILE5a inputs are critical to the evaluation of the CO emissions, it is the average speed component which is the foundation of the CMAQ conformity determination. In the calculation of the "tangible" benefits, the MOBILE5a model will utilize the same input data for all three scenarios. The vehicle types, ambient temperature and CO emission factors do not change as a result of the transportation model. However, the transportation model does recalculate the VMT, VHT and average speed of each scenario as a result of the implementation of traffic flow improvements. The comparison of the "Build", "No-Build", and "Baseline" scenarios only alter the transportation characteristics of the MOBILE5a model as a result of the travel demand forecasting model. The emission analysis input remains unaltered. Therefore, it is critical that the transportation networks be enhanced to further the accuracy of the model output to ensure that the information provided to the emissions models contain the accuracy needed to provide an accurate analysis. The formulation of the analysis methodology utilized the travel demand forecasting model as input into the mobile emission model. The validation of the methodology was conducted through a baseline or existing scenario and two test scenarios, Data Set 1 and Data Set 2. 85

PAGE 96

Data Set 1 revised a set of critical and non-critical travel speeds within a single 5 mile corridor in the network. The Data Set 1 CO emission results indicated that altering the travel speeds within a single small corridor (arterial facility in the CBD) may improve the regional network CO emissions over the Baseline and Non-Critical Data Set 1 scenarios. The Non-Critical Data Set 1 improvements (expressway facility in the CBD fringe) created only a slight shift in the VMT and a small improvement of travel speeds in the network. The Non-Critical Data Set 1 scenario did not significantly improve the regional transportation network and therefore did not alter the CO total pollutant emissions from the baseline scenario. Data Set 2 implemented traffic flow improvements on a large number of facilities within the region. Again the data set was divided into two sets of speed ranges; critical and non-critical. The CO emissions for the non-critical range was lower than those within the critical range. The Non-Critical Data Set 2 had lower CO emission rates for all time periods than the Critical Data Set 2 and Baseline scenarios. This occurrence may be attributed to the fact that the alterations to the free-flow input travel speeds reflected only the critical and non-critical speed range criteria and did not reflect the importance of VHT or VMT on the facility. The effect of altering a link in the non-critical speed range which transports a large number of vehicles may have an increased benefit over the altering of a critical link that carries a low number of vehicles. The sample scenario represents a relatively uncongested transportation network which may not alter the travel paths of vehicles to improved links as critically as an severely congested network. Figure 8 compares the Data Set 1 and Data Set 2 CO emissions for both the critical and non-critical scenarios with the CO emissions from the Baseline scenario. Transportation programs implemented must be able to show regional pollutant levels below the Baseline and No-Build scenarios. 86

PAGE 97

Conformity determination may be approved for any level of CO emission improvement. Critical Data Set 1 indicated a 0.82 tons per day CO emission reduction over the baseline scenario which meets the CAAA conformity criteria. Non-Critical Data Set 1 did not show an improvement over the Baseline scenario and therefore does not meet the CAAA conformity criteria. Critical Data Set 2 was not able to show conformity for the traffic flow improvements. Non-Critical Data Set 2 indicated conformity with the CAAA for the implementation of the traffic flow improvement program. The results of this thesis indicate that the "critical" travel speed is important in the development of traffic flow improvement program strategies. However, the travel speed is not an independent determining factor within the CO emission conformity analysis. Additional variables must be reviewed to determine the role the changes will have on the region. VMT and VHT on the facilities within the network are crucial to the air quality CO emission analysis. As shown in the methodology validation, the improvement of link traffic flow within the critical data set must also reflect the areas with high VMT and VHT in order to produce the most benefit to the region. The implementation of the traffic flow improvements must include areas within the critical travel speed range, and maintain high VMT and VHT volumes. 87

PAGE 98

FIGURE 8 DATA SCENARIO CO EMISSION RATES c 150 0 ,.., ... g : 100 ...... & 50 = 1-0 u ..., -+--Bsline NonCrit 1 Crit 1 NonCrit 2 Crit 2 Data Scenario I AM Peak D PM Peak I Off Peak I Total 88

PAGE 99

REFERENCES (1) Godish, Thad, 1991. Air Quality, 247-251, Lewis, Michigan. (2) U.S. Department of Transportation, Federal Highway Administration, 1992. "A Summary: lntermodal Surface Transportation Efficiency Act of 1991." (3) U.S. Department of Transportation, Federal Highway Administration, 1992. "Further Guidance on the Congestion Mitigation and Air Quality Improvement Program (CMAQ Program)." (4) U.S. Department of Transportation, Federal Highway Administration, 1994. "A Guide to the Congestion Mitigation and Air Quality Improvement Program." No. FHWA-PD-94-008, HEP-41/1-94(40M)E. (5) THE URBAN ANALYSIS GROUP, 1993. "URBAN/SYS User Manual Supplement and Installation Instructions for the DOS and OS/2 Operating Systems." (6) COMSIS Corporation, 1992. "MINUTP Travel Demand Forecasting Training Seminar." (7) U.S. Department of Transportation, Federal Highway Administration, 1986. "TRANSYT-7F Traffic Network Study Tool (Version 7F)." (8) Yagar, Sam and Case, E.R., 1981. "Using TRANSYT for Evaluation." The Application of Traffic Simulation Models. Special Report 194, Transportation Research Board, National Academy of Sciences, Washington, D.C. 89

PAGE 100

(9) Hurley, Jamie W. Jr. and Radwan, Ahmed E., 1981. ''Traffic Flow Simulation: User Experience in Research." The Application of Traffic Simulation Models. Special Report 194, Transportation Research Board, National Academy of Sciences, Washington, D.C. (10) Dowling, Richard and Skabardonis, Alexander, 1992. "Improving Average Travel Speeds Estimated By Planning Models." Transportation Research Board No. 1366, National Academy Press, Washington D.C. (11) Denver Regional Council of Governments (DRCOG) 1994. "Technical Briefs: Signal Timing/Optimization Report." (12) Shrouds, Jim, 1992. "CMAQ Annual Reporting." U.S. Department of Transportation, Federal Highway Administration. (13) Heiken, J.G., Austin, B.S., Eisinger, D.S., Shepard, S.B., Duvall, L.L., 1991. "Estimating Travel and Emission Effects of TCM's." SYSAPP-91/117. (14) Pikes Peak Area Council of Governments, 1992. "Regional Travel Survey." (15) AI-Deek, H., Wayson, R., and Radwan, A. E., 1995. "Methodology for Evaluating ATIS Impacts on Air Quality." Journal of Transportation Engineering, 376-384. 90