EPA-450/3-77-023
August 1977
DETERMINATION
OF PERCENTAGES
OF VEHICLES OPERATING
IN THE COLD START MODE
l),S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-77-023
DETERMINATION OF PERCENTAGES
OF VEHICLES OPERATING
IN THE COLD START MODE
by
Theodore P. Midurski and Alan H. Castaline
GCA Corporation
GCA/Technology Division
Burlington Road
Bedford, Massachusetts 01730
Contract No. 68-02-1376
Task Order 29
EPA Project Officer: James H. Wilson, Jr.
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
August 1977
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations in limited quantities from the
Library Services Office (MD35) , Research Triangle Park, North Carolina
27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
GCA Corporation, GCA/Technology Division, Burlington Road, Bedford,
Massachusetts 01730, in fulfillment of Contract No. 68-02-1376, Task Order
29. The contents of this report are reproduced herein as received from
GCA Corporation. The opinions, findings, and conclusions expressed
are those of the author and not necessarily those of the Environmental
Protection Agency. Mention of company or product names is not to be
considered as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-77-023
11
-------
ABSTRACT
Estimates of the percentages of vehicles operating in less than the sta-
bilized condition (i.e., cold mode) have been made for 60 locations in the
Pittsburgh and Providence regions. These estimates were derived using de-
tailed regional travel data for both regions.
Accurate knowledge of the percentages of vehicles operating in the cold
mode at a particular study location is a key factor in estimating carbon monox-
ide emissions. That carbon monoxide emission rates are severely affected by
operation in the prestabilized condition has been well documented through
numerous laboratory studies. While these laboratory studies have served to
characterize carbon monoxide emissions jErsm_jiol§cific_loeafrion. It is also indicated that the actual percentages of
vehiclesepexa£ing_in the cold mode may be somewhat different from
centages_as^umed^ in tfie~ Federal Test""Proce^ureT""~
iii
-------
CONTENTS
Abstract iii
Figures ............ vi
Tables vii
Acknowledgments ix
Executive Summary : x
1. Introduction . . 1
Background and Purpose ...... 1
2. Technical Aspects of the Cold Mode Phenomenon 3
Cold Mode Characterization 3
Definition of Cold Mode Operation 10
References 14
3. Technical Aspects of the Data Used in the Cold Mode Analysis . 15
Primary Data Base 16
Supplementary Data Base 19
References. . . 2.9
4. Study Methodology 30
General 30
Selection of Study Links 30
Analytical Procedure 30
Limitations of the Procedure. ....... 32
5. Results of the Analyses 33
Pittsburgh 33
Providence. . ~. . . . . - , .- . . . . . . 39
6. Conclusions of the Study 50
Generalization Regarding Cold Mode Distribution 50
Conclusions 53
Recommendations for Further Study . . 55
Appendices
A. Federal Test Procedure .56
References ........ 61
B. Overview of the Comprehensive Planning Process 62
Introduction 62
Background 62
Travel Forecasting 64
References 75
C Example of the Method Used to Identify the Percentage of
Vehicles Operating in the Cold Mode. 76
Glossary 97
v
-------
FIGURES
Number
1 Representation of choke-on time as a function of ambient
temperature 4
2 Cumulative HC and CO emissions during FTP driving cycle - Car A . . 6
3 Cumulative HC and CO emissions during FTP driving cycle - Car B . . 6
4 Cumulative HC and CO emissions during FTP driving cycle - Car C . . 7
5 Cumulative CO emissions during FTP driving cycle - Car G 7
6 Cumulative CO emissions during FTP driving cycle - Car I ..... 8
7 Cumulative CO emissions during FTP driving cycle - Car J 8
8 Average vehicle CO percent deviation versus start-up temperature. . 9
9 Temperature dependency of carbon monoxide emissions - varied
categories of vehicles 11
10 Cumulative distribution of the probable error for travel time runs
made in Providence, Rhode Island ......... 24
11 Study-link locations map, Pittsburgh, Pa 35
12 Study-link locations map, Providence, R.I. ' 43
vi
-------
TABLES
Number Page
1 Comparison of Cold Mode Cycle Length in Minutes as a Function of
Soak Duration for Ambient Temperatures of 75°F and 20°F .... 12
2 Travel-Time Analysis Results - Providence, Rhode Island 20
3 Minimum Sample Size Requirements for Travel Time Studies With A
Confidence Level of 95.0 Percent, Presented as the Number of
Observations Required 21
4 Assessment of the Reliability of the Travel Time Data Compiled
For Providence 23
5 Relative Volumes on Radial Arterials by Time Period and Direction,
Shown as a Percentage of ADT 25
6 Computed 24-Hour Composite Average Speeds for Providence .... 26
7 Comparison of the Derived and Assigned Network Speeds for
Providence .... 27
8 Analysis Link Descriptions, Pittsburgh 34
9 Percentage of Vehicles Operating in the Cold Mode, by Time
Period, For 24 Locations in Pittsburgh 36
10 Analysis Links by General Category, Pittsburgh . 37
11 Percentage of Vehicles in Cold Mode Operation by Time Period and
by Location Group For Pittsburgh 38
12 Volumes of Cold and Stabilized Vehicles by Time Period and by
Location Group For Pittsburgh 38
13 Percentage of Vehicles in Cold Mode Operation With 1 Minute
Additional Access Time, Pittsburgh, Pa 40
14 Percentage of Vehicles in Cold Mode Operation With 2.5 Minute
Additional Access Time, Pittsburgh, Pa 40
vii
-------
TABLES (continued)
Number Page
15 Distribution of Trips by Time-From-Trip-Origin - Pittsburgh . . . 41
16 Analysis Link Descriptions, Providence . . . 42
17 Percentage of Vehicles Operating in the Cold Mode, by Time Period,
For 36 Locations in Providence 44
18 Analysis Links by General Category, Providence 45
19 Percentage of Vehicles in Cold Mode Operation, by Time Period and
by Location Group For Providence . . . . 47
20 Volumes of Cold and Stabilized Vehicles by Time Period and by
21
22
23
Distribution of Trips by Time-From-Trip-Origin, Providence, R.I.
Percentage of Vehicles in Cold Mode Operation With 1 Minute
Additional Access Time Providence, R.I.
Percentage of Vehicles in Cold Mode Operation With 2.5 Minutes
Additional Access Time - Providence. R.I. ...
^ *
48
49
49
24 Percentage of Vehicles in Cold Mode Operation by Distance From
Central Business District For Two Corridors in Providence, R.I. 51
25 Percentage of Vehicles in Cold Mode Operation by Distance From
Central Business-District For-Three Corridors-in.. . . .- .
Pittsburgh, Pa. . 52
26 Suggested Ranges of Values of the Percentage of Vehicles Operating
in the Cold Mode For Various Conditions of Time and Location. . 54
viii
-------
ACKNOWLEDGMENTS
Numerous individuals made significant contributions to this overall study
effort, and GCA/Technology Division wishes to acknowledge their participation.
Ongoing project supervision was received from Mr. James H. Wilson of the Air
Management Technology Branch, Monitoring Data Analysis Division, U.S. Environ-
mental Protection Agency, who served as Project Officer.
Important technical guidance regarding cold mode operation was provided
by Ms. Marcia Williams, Office of Mobile Source Air Pollution Control, U.S.
Environmental Protection Agency. Finally, we wish to acknowledge the helpful
assistance provided by Messrs. John Brownell and Roland Frappier of the Rhode
Island Statewide Planning Program, and Messrs. Theodore Treadway and Wade Fox
of the Southwestern Pennsylvania Regional Planning Commission.
ix
-------
EXECUTIVE SUMMARY
STUDY OBJECTIVES
Numerous laboratory studies have shown that the emission rate of pollutants
such as carbon monoxide and hydrocarbons is significantly higher during the
first several minutes of vehicle operation compared with the emission rate
SfjteirjEhe veKiVle has warmed upt. In fact, the emission rate during the cold
mode cycle can be several times higher than the jate during stabilized opera-
tion. Owing to its strong influence on emission rates, cold operation must
be accounted for in any analysis of local or areawide vehicular emissions.
Laboratory studies have parameterized cold mode operation so that effects
can be accounted for in air quality analyses if the extent of the actual cold
mode operation is known. To date, little effort h^s been expended on defining
the fraction of automobile travel that typically occurs~~InThe coIH'^mojdeT In
response to this lack of information regarding the likely distributions of
cold operating vehicles that occur on urban street systems, GCA/Technology
Division, under contract to the U.S. Environmental Protection Agency, has under-
taken an analysis that focuses on determining the diurnal distributions of
cold operating vehicles that occur on urban street systems. This report
describes the methods used in and the results of the analysis.
TECHNICAL ASPECTS OF COLD MODE OPERATION
'""""
The excess emissions associated with cold operation occur as a result of
temporary "imbalances^ i"n'"various cOToBusExon system parameters. For instance,
when" an automobile's choke mechanism is~"actuatea7~^xtra fuel is drawn into the
combustion area. This extra fuel creates an imbalance in the air-to-fuel ratio,
which results in incomplete fuel combustion, and hence, high emission rates of
carbon monoxide and hydrocarbons. Other factors such as cylinder wall quench-
Iingandthe temperature_rise-time of catalytic conyerterj^also contribute to
1 high emission rates during initial operation. "~~~*
Numero_us_ studies, .have defined the emission patterns^during^CP.ld^operatipn.
These studies indicate that the rate of boTfPcafbon monoxide and hydrocarFon
emission are much greater during the first 200 to 300 seconds of operation
compared to the emission rate after temperatures, havenstabilized. Testing
also has'lshown that the magnitude of the emission rate associated with cold
operation is a function of ambient temperature; colder temperatures result in
much.higher emission rates during cold operation. Also, these tests indicate
that ambient, temperaturehas^ .relatively small e.f fect_on emission rates beyond
the_ first several minutes of operation.
-------
Recent analysis by the Environmental Protection Agency, Office of Mobile
Source Air Pollution Control, have resulted in defining the duration of__£he
gold mode cycle as a function of soak dur at ion ariH^amb ient temperature. These
relationships show, tor instance, ^IfhlHi^aTjiii^mblJlm^ 75°F ,
is about 2.4 minutes, while the duration for a vehicle soaked for 5 hours is
abdut~5 . 5 minutes; at 2X)°F , the corresponding cold mode durations would be about
JV3_jaiiaufes 'and 6.0 minutes. The relationships developed from EPA's research
were used extensively in the analysis being reported here.
STUDY APPROACH
General
The general procedure used in the study was to identify certain charac-
teristics of trips occurring on each of a number of specific highway sections,
and relating these characteristics to cold mode operation as defined by the
previously mentioned EPA analyses. Specifically, the trip characteristics
that were considered included (1) the elapsed time from the trip origin to the
study site, and (2) the duration of the cold soak prior to beginning the trip.
The primary source used to derive this information was comprehensive planning
data obtained from the regional planning agency in each study area. These
data were supplemented with field data collected by GCA/Technology Division,
where required.
Study Scope
The scope of the study permitted analyses to be conducted in two cities.
Selection of the cities was based on population and the availability and ade-
quacy of the necessary data elements. The Cities of Pittsburgh, Pennsylvania
and Providence, Rhode Island, were selected for study.
9
The intent of the study was to- define the diurnal* patterns of cold mode
distribution on various types of roadways and in different urban settings.
To accomplish this, a total of 60 highway sections were analyzed. Included
were expressways, major arterials, minor arterials, and collector streets
located in the central business district, outer fringe areas, residential
areas, etc.
Study Procedure
The analysis of each study link was based on identifying the diurnal pat-
terns of trips occurring on each link, as a function of the travel time from
each trip origin, and cold soak duration.
Regional planning data were used to identify the origin of each trip
occurring on each study link. The travel time from each origin to each study
link was then determined using both regional planning data and the results of
field studies conducted by GCA. Data files developed by the planning agencies
were then utilized to determine the diurnal trip generation patterns and park-
ing duration characteristics for various areas within the vicinity of the study
links.
xi
-------
The data described above were used as the basis fordisaggregating^tbe
total trips occurring on each link, into .over 200 indiyldjuajUcatBgories . which
rultimately"~described each trip in terms of (1) time that it. ocjgujrrgd, (2) travel
time "from its origin to the study link, and (3) the cold soak duration prior to
beginning the trip. These distributions and the data developed by EPA regarding
defining cold mode as a variable function of soak duration, were used directly
to determine the percentages of cold operating vehicles on each study link.
STUDY RESULTS
A total of 60 separate highway links (24 in Pittsburgh and 36 in Providence)
were analyzed and the cold mode distributions for each were identified. The
percentages of cold operating vehicles identified were generally much higher
than originally expected. For instance, several links in both cities were
shown to have more than 70 percent cold operating vehicles during variousjtime
periods . Individual" links were combined _in gagb_city ^^ ^o"
~ " ^
_ _ ^
estimate of~the percentage" of veKlcles^o^erating in the cold mode by location
(e.g., central business district) and time period (e.g., during the morning
peak hours). ' ' *" ---- -=-
These analyses showed wide variations in the percentage of cold operating
vehicles from location to location and by time period. For example, in
Pittsburgh, the p_ercenitages, during tthe_mornmg,,peak hours ranged from 0.4 (for
an expressway linkX_tP_ -26-.-8--Cf.or links, located near non-'CBD ^eneratror s) ; ~f 6 r
the entire day, the percentages on links within the CBD-ranged from 9.8 during
the morning peak hours, to 38.3 during the afternoon peak hours.
Sensitivity analyses were performed to determine the effect of increasing
the travel time between trip origins and study links by relatively small
amounts. These analyses indicated that the computed percentages of cold operat-
ing vehicles were indeed sensitive to small changes in travel time. For exam-
ple, the computed percentage, of cold operating vehicles in the central business
district of Providence during the morning peak hours -was -reduced . from 20.7 to
11.7 by increasing travel time by 1 minute; by adding 2.5 minutes, this per-
centage was reduced to 1.0.
CONCLUSIONS
A primary conclusion of the study is that a wide range in the percentages
of vehicles operating in the cold mode can be expected within an urban area
and that these percentages can be expected to vary significantly by tim§,o-f
-4§v. Also, it can be concluded that definite pattefns*^ay*&n&tr-tret5e^sarily
exist in the distribution of cold operating vehicles from city to city; in
fact, strong patterns may not exist from location to location within a particu-
lar city.
The data derived in the study were used as the basis for developing general
recommendations for estimating the percentage of vehicles operating in the cold
mode for various time periods and traffic environments. The recommended pro-
cess consists of selecting a value from a range of values provided for ^several
sets of conditions^ (urban^etting^and/jtime period) . The appropriate ranges are
shown in Table 26, which appears on page 54 of this document.
xii
-------
CONTENT OF THIS REPORT
The report provides detailed information on a number of aspects of (1) the
cold start mode in general; (2) the techniques used in the analysis; (3) the
data base used; and (4) the results of the entire study. Section 1 provides a
general introduction including discussion regarding the background and purpose
of^the study^
The technical aspects of cold mode operation are discussed in Section 2.
Included are discussions concerning the primary relationships between cold
operation and emission rates, and the factors that are instrumental in deter-
mining the impact of operating in the cold mqde^__Discussiqn is also provided
regarding the definition of cold operation.
Section 3 provides a discussion of the general techniques used in the
study. Included are descriptions of the mechanics of the procedure as well
as the data and data sources used. A more technical discussion of the study
methodology is provided in Section 4.
The results of the study are presented in Section 5. This section includes
discussions of the results of analyzing the 60 study links both separately
and in groups. Separate discussions are provided for Pittsburgh and Providence.
The results of the sensitivity analysis are also discussed in this section.
Section 6 discusses the conclusions of the entire analysis and offers several
recommendations for further study.
Three technical appendices are provided. Appendices A, B, and C describe
(1) the Federal Test Procedure; (2) the comprehensive planning process; and
(3) an example computation of the percent of vehicles operating in the cold
mode on a study link, respectively.
A glossary defining a number of technical terms used throughout the re-
port is also provided. , ,. _ . ..........
xiii
-------
SECTION 1
INTRODUCTION
BACKGROUND AND PURPOSE
Many techniques have been developed for analyzing the impact of vehicular
traffic on ambient concentrations of pollutants such as hydrocarbons and
carbon monoxide. A common element among all of these techniques is the ex-
pression of the rate that the pollutant is emitted into the ambient atmosphere.
Numerous studies have resulted in the identification of factors that affect
pollutant emission rates and have provided the basis for estimating the impact
that these factors have, both separately and collectively, on the actual emis-
sion rate.
A factor that is of particular importance, especially with regard to the.
analysis of carbon monoxide, is the relationship between: (1) a vehicle's
emission rate and ambient temperature, (2) the amount of time that the ve-
hicle has been operating, and (3) the engine temperature prior to starting.
During the first few minutes that a vehicle is operated, the rate of carbon
monoxide emission is significantly higher than when the vehicle has warmed
up. In fact, the rate that carbon monoxide is emitted during the first few
minutes of cold operation can be several times greater than the rate after
the engine has been operating for several minutes. The quantity of carbon
monoxide produced within an area, then, may be strongly affected by the
"amount of travel that occurs by vehicles operating in a cold condition. Owing
to its potentially strong influence on emissions, cold operation must be
accounted for in any analysis that requires dimensioning the quantity of car-
bon monoxide generated by vehicular traffic.
Numerous laboratory studies have been conducted to define the impact of
cold operation on emission rates. Also, studies have been conducted that
characterize the cold operating phase as a function of several operational
and environmental parameters. It can be concluded that cold vehicle operation
has been adequately parameterized so that its effects can be accounted for in
air quality analyses if the extent of the actual cold mode operation is known.
To date, very little effort has been expended on either defining the fraction
of automobile travel that occurs in the cold phase, or identifying a process
that could be used to make this determination. In response to this apparent
lack of information regarding the distributions of cold operating vehicles
that actually occur on urban street networks, GCA/Technology Division, under
contract to the U.S. Environmental Protection Agency, has undertaken an anal-
ysis that focuses on determining the types of cold mode distributions that
occur in several characteristically different areas and during different time
-------
periods. The analysis focused entirely on the portion of the vehicle popula-
tion consisting of light-duty, noncommerical vehicles. Throughout the study,
the primary concern was the effect of cold operations on carbon monoxide emis-
sions; however, the results of the analysis can be applied directly to hydro-
carbon emissions as well. The purpose of this report is to present a discus-
sion of both the procedures used for the study and the results.
-------
SECTION 2
TECHNICAL ASPECTS OF THE COLD MODE PHENOMENON
COLD MODE CHARACTERIZATION
Cold mode operation is defined in general terms as the first several
minutes that a vehicle is operated after having not been operated for several
hours. The primary manifestation of cold mode operation is an extraordinarily
high emission rate of both carbon monoxide,and hydrocarbons.
The excess emissions associated with cold operation occur as a result of
a temporary imbalance in the combustion system parameters. In order for
ignition to occur in a gasoline engine, the fuel introduced into the cylinder
must be vaporized and there must be present an appropriate balance or ratio
between the quantities of air and vaporized fuel. Gasoline does not vaporize
as well at lower temperatures as it does in the relatively high temperature
ranges that are typical of stabilized engine operation. Therefore, when a
"cold" engine is being jstarted, the rate of gasoline^vaporizjation that occurs
in thecombustion^ area"is much less,.than when the^eng£rie_is^operating at normal
temperature. As a result, an imbalance occurs in the air-to-fuel ratio; this
"imbalance can be so severe that ignition will not occur. To compensate for
this temporary imbalance in the air-to-fuel ratio, the fuel delivery system
is equipped with a choke mechanism, which, when activated, restricts the flow
of incoming air to the point where a vacuum occurs in the intake manifold.
The vacuum causes additional" fuel to be'drawn into the manifold resulting in
extra fuel being delivered to the combustion areas. The increase in the total
amount of fuel delivered to the cylinder compensates for the reduction in the
rate of vaporization, so that the net result is an air-to-vaporized fuel ratio
that is suitable for ignition. Although the ratio of air-to-vaporized fuel
becomes balanced when the choke is functioning, the ratio of air to total
fuel becomes imbalanced owing to an insufficient quantity of combustion air
being present. This imbalance results in incomplete fuel combustion; two
major products of incomplete combustion are carbon monoxide and various un-
burned hydrocarbon compounds.
The choke mechanism on most vehicles is actuated automatically by a Jheat
sensorincorporated into_-a^temjpera^ture-sensitive engine component such as the
manifjold. The rise-time^rom~ambl!en^^t6snta^iriz^3^Te^pe1taCuf"e"*for
the heat sensitive components generally lags^the rise-time in the_jent,ering
fuel temperature and combustion chamber temperature by various amounts of
time, thereby assuring adequate choke-on time. Studies have shown the
This ratio is commonly referred to as the air-to-fuel ratio.
-------
choke-on time to be a function of ambient temperature. Figure 1 provides an
indication of choke-on time as a function of ambient temperature.
E
uJ
S
9-
8-
7-
6-
B i-- i- T r -nr-
*-30 -ZO -IO 0 10 20 50
TEST TEMPERATURE, T
Figure 1. Representation of choke-on time as a function of ambient
temperature.
Production of excess emissions during the cold stages of vehicle opera-
tion occurs also as a result of a phenomenon known as wall quenching. Wall
quenching is a combustion phenomenon that occurs when a flame moves towards
a relatively cool cylinder wall. The cool cylinder wall slows and effectively
stops combustion in the vicinity, thus a layer of unburned and partially
burned fuel remains at the wall surface at the end of the combustion stroke.
The thickness of the layer is a function of several parameters including the
cylinder temperature and pressure, the presence of cylinder deposits, and the
wall temperature itself. Obviously, the cylinder and wall temperatures are
much lower during initial operation than .after warm-up. These lower tempera-
tures result in a thicker layer of unburned or partially burned fuel during
each stroke, which contributes to the total excess emissions associated with
cold-mode operation.
Emissions from newer vehicles equipped with catalytic converters are
affected by cold operation because of an additional factor. The converter,
which functions as the primary emission control device, does not begin to
operate efficiently until it reaches a certain temperature. The time required
to reach the most efficient operating range is generally around 200 to 300
seconds, regardless of the ambient temperature.2 Therefore, during the several
minutes that it takes for the converter to reach optimum operating temperature,
carbon monoxide and hydrocarbons emission rates are somewhat higher than dur-
ing warmed-up operation.
Temperature obviously has a considerable effect on cold mode emission
characteristics. The relative impact that ambient temperature has on emission
rates has been analyzed in the laboratory during several studies. Of par-
ticular importance is the fact that these studies conclude that the effects
of varying ambient temperatures are apparent only during the first several
-------
minutes of operation. In one set of tests,,?, the effect of ambient temperature
on carbon monoxide and hydrocarbon emission rates for three configurations of
standard 1970 production cars, and" three 1970 production cars equipped with
advanced (with respect to the 1970 model year) emission -control devices, was
analyzed. The testing involved measuring the emissipns^prpduced-by^each
vehicle as it was operated from a cold condition* through the Urban Dynamometer
Driving Schedule, which is included in the Federal Test Procedure (see
Appendix A). Tests were performed at several,
conditions for each vehicle. Cumulative emissions were then identified as a
function of elapsed operating time and temperature; test results are illustrated
in Figures 2 through 7, below. ~~
These figures show quite vividly that the emission rates of both hydro-
carbons and carbon monoxide for the test cars' are generally not sensitive
to changes in ambient temperature beyond, say,~Fhe first 200~to 300 sicon'ds-of
operation, but within the first 200 to 300 seconds, temperature has a marked
^effect on emission rates.
In another series of investigations3 on the effects of cold ambient
temperatures on light-duty vehicle emissions, nine_1973 vehicles were consid-
ered. These tests involved a standard 23-minute FTP driving schedule during
which exhaust samples were collepted for three cycles of operation. The^ first
cycle included the first 505 seconds of the schedule representing the "cold"
operating phase. The second cycle involved the stabilized phase, which rep-
resents the remainder of the 23-minute cycle. The third cycle represents a
"hot start" condition, which involves repeating the first 5.5 seconds of the
driving schedule after a 10-minute engine-off period. For this series of
tests, an ambient temperature of 60°F was established^as a. baseline condition.
Comparisons were made of carbon monoxide emissions, for each^ mode at ambient
temperatures ranging from 0 F" to 80°F.
The results of the tests are shown in Figure 8. This figure shows the
relative effect of temperature diffejrences (with respect to an ambient tem-
oeratureof 60°F) for each of" the three test phases and the composite.**;
Figure 8 sHows, again, that the relative effect of various ambient temperatures
is insignificant beyond the first few minutes of operation; this is shown by
comparing temperature effect on Bag 1 emissions (the first 505 seconds of
operation) to the Bag 2 emissions.
Cold condition implies that the vehicles had not been operated for at least
12 hours prior to testing.
HC emissions for Test Cars G, I, and J (Figures 5 to 7) were conducted but are
not reported here.
**
The composite is computed using the sum of 43 percent of Cycle 1 emissions,
100 percent of Cycle 2 emissions, and 57 percent of Cycle 3 emissions (see
Appendix A).
-------
HC
CUM
OIAMS
20
W
CAIA
MO CIO 'A- lo*y
IfTO Ptoduction
Cemmrciol tYtm. Fuel
4000b.lMrtloWt.
HC
TO'FCO
200 400
400
TIME
HC
800 1000 1200 1400
800
700
600
5" CO
*v, CUM'
400 .GIAMS
300
200
MO
0
Figure 2. Cumulative HC and CO emissions
during FTP driving cycle - Car A.
CAJtl
HC
CUM
GIAMS
WO Prarfuctton Car
_ ComMrelol Fram. Full
4500 b.
10
200 400
400 800
TMf
HC
1000 1200 1400
800
700
600
300 CO
CUM.
400
300
200
100
0
Figure 3. Cumulative HC and CO emissions
during FTP driving cycle - Car B.
-------
MC
CUM
CRAMS
30
30
20
10
CAKC
455 CIO Engirt* "B" ftody
1970 Production Cor
Commefciol Frcffl. Fuel
5000 Ib. IrwrlioWt.
I
HC
200
TO'FCO
400
600 800
TIME
SEC
1000 1200 1400
BOO
700
600
500
400 CO
CUM
300 CRAMS
200
100
0
Figure 4. Cumulative HC and CO emissions
during FTP driving cycle - Car C.
250
200
CO
CUM
OftAMS
ISO
100
so
CAIG
ISO CIO fnctm T lo«y
Commercial Unlradid Fu.1
Catalytic Converter
Al», ECU
Quick Hwt MemifoU
Quick ChoU
4500 Ib. Inwtla Wt.
200
400
«oo no
TIMI
SIC
1000
1200
1400
Figure 5. Cumulative CO emissions during
FTP driving cycle - Car G.
-------
230
200
CAII
259 CID Engine 'A' toey
Premium Unleaded Fuel
Catalytic Converter
AIR
£Gf
Production Choke
3500 fc. Inertia Wt.
1400
Figure 6.
CO
CUM.
GMMS
500
400
300
200
100
Cumulative CO emissions during
FTP driving cycle - Car I.
O'F
aoo
^b-
SEC
CAIJ
19 CD nglne
Commrreio)
Catalytic Converter
AM
Production Choke
4500 Ib. Inertia Wt.
-nte nte
Figure 7. Cumulative CO emissions during
FTP driving cycle - Car J.
-------
+150
+ 120
O +90
to
O +60
cc
u.
+30
1 °
>
0-30
in
u
-60
-90
-120
-150
-BAG I
V
BAG 2
REFERENCE CASE
I
20 40 60
INITIAL START-UP TEMPERATURE, (°F)
80
Figure 8. Average vehicle CO percent deviation
versus start-up temperature.
-------
The results of a third study** also indicated the significance of the
first few minutes of cold operation. In this study, 26 production vehicles
of various model years and configurations were tested using the FTP driving
cycle. The testing was conducted at ambient temperatures of 20°F, 50°F,
75°F and 110°F. The average emission rates for each phase of the FTP driving
cycle and a composite emission rate were defined for various groupings of
vehicles; these are shown in Figure 9.
Figure 9 provides a good indication of the magnitude of the effect that
cold mode operation has on emission rates. The effect is best illustrated by
directly comparing the Bag 1 and Bag 2 emission rates. The curves for catalyst
vehicles operating at 20°F indicate that during the first 505 seconds of
operation, the emission rate is.about 130 grams per mile while the Bag 2 rate
is about 3 or 4 grams per mile. The" Importance, then, of accounting for cold
mode operatrtonTrn any analysis of carbon monoxide is obvious. As an illus-
tration, if emissions were computed for a specific location first assuming
(1) an ambient temperature of 20°F, (2) a vehicle mix whose emission charac-
teristics were similar to those for 1973 - 1974 model year vehicles shown in
Figure 9, and (3) all vehicles operating in the warmed-up (stabilized) mode,
the result might yield the quantity X, which would represent the product of
an average emission rate (in this instance, about 60 grams per mile based on
Figure 9) and a travel factor defining the quantity of travel (in vehicle-
miles) . If a second analysis were performed using the same assumptions except
that all vehicles were operating in the cold mode (obviously an extreme assump-
tion) , the results would be the quantity 3.8 X; this reflects the difference
in emission rates for cold and stabilized operating modes for the representa-
tive vehicle population. If it were assumed that the vehicle population was
comprised of all catalyst vehicles, the difference would be much greater
(again, based on Figure 9).
DEFINITION OF COLD MODE OPERATION
It can be concluded from the previous discussion that cold-mode operation
is a time-dependent function. The standard definition of cold mode operation
is that it represents the first 505 seconds of vehicle operation following a
4 hour (minimum) engine-off period. 5 This definition implies that a discrete
function exists between cold mode operation and the cold soak period. Recall,
however, that in the previous discussion it was indicated that both ambient
and engine temperatures appear to be critical determinants of cold mode opera-
tion. The figures presented in that section indicate that ambient tempera-
ture does not have a large effect on the time required to stabilize (ambient
temperature does have a very significant effect on the rate of emissions,
however). Apparently, then, the rise-time to stabilized operating temperature
is not affected to a large extent by a differential in ambient temperature of,
say, 50°F (20°F to 70°F). However, the temperature within an engine ranges
from ambient up to about 2000°F in the combustion area (cylinder walls) and
over 200°F for fuel delivery components. Obviously, when an engine is shut
down, these components begin to cool down to the ambient temperature. If the
Cold soak is defined as the time interval that a vehicle s engine is not
operated.
10
-------
J20
200
160
120
«0
I
3
4O
!967
. l97l-'?2
320
260
240
200
160
120
80
40
Cold trontiflnt
160
HO
120
IOO
80
00
<»o
20
S'06'Wta
1967
160
140
120
100
60
60
40
20
20
40 «0
0
IOO I2O 0
TEST AMBIENT,
HOT trontitnt
Cololytl
»97l-'72
20
40 60 60 IOO 120
Figure 9. Temperature dependency of carbon monoxide emissions
categories of vehicles.
- varied
11
-------
engine is restarted before the components reach ambient temperature, the amount
of time required to again reach the hot stabilized condition would be reduced.
Limited testing6 by the U.S. Environmental Protection Agency, Office of Mobile
Source Air Pollution Control, has indicated that the time required to reach
the hot stabilized mode is indeed a continuous function of the engine starting
temperature. In these analyses, the starting temperature is considered im-
plicity to be a function of the soak time. In this connection, equations were
developed describing the required time to reach the stabilized mode as a func-
tion of cold soak period; one of these equations is
t = 3.11 S°'36 (1)*
where t = time in minutes to reach the stabilized mode
S = soak time in hours.
The time t defined in equation 1 represents the time to stabilize when the am-
bient temperature is about 75°F. Testing was also performed at temperatures
of about 20°F and an equation was derived representing the time required to
stabilize (t') again as a function of the cold soak duration; this equation is
t' = 2.61 S°*36 + 1.32 (2)
where t" = time in minutes to reach the stabilized mode at an ambient
temperature of 20°F
S = soak time in hours.
The effect of ambient temperature on the cold mode cycle length can be
seen by comparing the results of equations 1 and 2 applied to a set of arbi-
trary cold soak durations; this is presented in Table 1, below.
TABLE 1, COMPARISON OF COLD MODE CYCLE LENGTH IN MINUTES AS A FUNCTION OF
SOAK DURATION FOR AMBIENT TEMPERATURES OF 75°F and 20°F
Cold soak duration, hours
Ambient
temperature, °F
0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5
75
20
2.
3.
4
3
3.6
4.4
4.3
5.0
4.9
5.4
5.
5.
3
8
5.8
6.1
6.1
6.4
6.4
6.8
6.7
7.0
Table 1 clearly shows that the .effect of ambient temperature on the cold mode
cycle length is not nearly as severe as is the cold soak duration.
This equation has been normalized to reflect assumptions in AP-42; for analy-
ses not associated with the emission factors presented in AP-42, a slightly
different equation applies.
12
-------
Basic guidance in accounting for cold mode operation in air quality
analyses is provided in AP-42.5 Contained in AP-42 are equations for computing
cold mode adjustment coefficients to be applied to composite emission factors
for light-duty vehicles.
13
-------
REFERENCES
1. Robertson, J. E. The Impact of Vehicle Emissions on Air Quality at Low
Ambient Temperatures. SAE Paper No. 741054. October 1974.
2. Miles, Donald L. and Max Homfeld. The Effect of Ambient Temperature on
Exhaust Emissions of Cars With Experimental Emission Controls. SAE
Paper No. 451052. October 1974.
3. Polak, J. C. Cold Ambient Temperature Effects on Emissions from Light-
Duty Motor Vehicles. SAE Paper No. 741051. October 1974.
4. Ashby, H. A., et al. Vehicle Emissions - Summer to Winter. SAE Paper
No. 741053. October 1974.
5. Compilation of Air Pollutant Emission Factors, Second Edition. U.S.
Environmental Protection Agency, Research Triangle Park, N.C. 27711.
Supplement No. 5 to Publication No. AP-42. February 1976.
6. Williams, Marcia. Definition of Vehicle Cold Start Operation. Unpublished
paper. U.S. Environmental Protection Agency, Office of Mobile Sources.
May 1977.
14
-------
SECTION 3
TECHNICAL ASPECTS OF THE DATA USED IN THE
COLD MODE ANALYSIS
BACKGROUND OF THE STUDY METHODS
As indicated in Section 1, the primary objectives of this study were to
determine the distributions of "cold" vehicles operating in traffic streams at
various urban locations and, based on these determinations, develop a general
statement regarding the distributions that would be expected at similar loca-
tions in any urban area.
The general procedure used in the study was to identify certain charac-
teristics of all trips occurring on each of a number of specific street sec-
tions in two major U.S. cities, and relate these characteristics to cold mode
operation. From the previous discussion regarding the nature of the cold mode
phenomenon, it is apparent that two trip characteristics that would be of par-
ticular importance in the analysis of cold mode operation are (1) the elapsed
operating time from starting the vehicle, and (2) the duration of the cold
soak cycle prior to starting the vehicle. A primary task in this study, then,
was to identify these two characteristics for each trip occurring on each of
the street sections analyzed.
The primary data source that was used in the study was comprehensive trans-
portation and land-use planning data (details of this data will be provided
later); only limited field studies were required to supplement this data.
The scope of the study permitted analyses to be conducted in two cities.
Selection of the cities was based on two criteria. First, the comprehensive
planning data base had to be both accessible and adequate. In this connection,
numerous metropolitan planning organizations were contacted and it was deter-
mined that several apparently had adequate data and were willing to accom-
modate any requests for it. The second criterion concerned the desire to anal-
yze cities of different population categories. Specifically, the intent was
that one city (actually, the urbanized area) be within the population range of
500,000 to 1 million, while the other city's population would be greater than
1 million. From a rather extensive list of possible sites, the Cities of
Pittsburgh, Pennsylvania, and Providence, Rhode Island were selected. It is
noted that there were no compelling reasons for selecting these two particular
cities over others identified as potential study sites; Providence and
Pittsburgh appeared to offer only minor advantages over several of the other
cities.
15
-------
PRIMARY DATA BASE
A most crucial aspect of any study is the data that are used. The reli-
ability of the results of a study is obviously closely related to the quality
of the data used. It is appropriate, then, that a study report include a com-
plete description of the data including its sources, methods of derivation,
reliability and limitations, and application in the study.
The following presents descriptions of both the comprehensive planning
programs for Pittsburgh and Providence (an overview of the entire comprehensive
transportation planning process is provided in Appendix B) and of the individual
data elements from the comprehensive planning programs that were used in this
study. Also included is a description of supplementary data developed for the
study.
Pittsburgh Regional Planning Process
During March 1966, the Southwestern Pennsylvania Regional Planning Com-
mission (SPRPC), in cooperation with the U.S. Bureau of Public Roads, the U.S.
Department of Housing and Urban Development, and the Pennsylvania State Highway
Department, began a comprehensive land use and transportation planning program
for the region comprised of the Counties of Allegheny, Armstrong, Beaver,
Butler, Washington, and Westmoreland. The stated objectives of the program
were to develop a comprehensive land use and transportation plan for south-
western Pannsylvania and, at the same time, to qualify the region for continued
federal funding by satisfying the requirements of Section 134 of the Federal-
Aid Highway Act of 1962.
The initial effort focused on developing a detailed data base describing
various features and characteristics of the 4,500 square mile region. The
first task undertaken was a housing unit survey, which established the geogra-
phic location of all housing units in the region. This base provided the
.universe from which a sample was drawn for conducting home interviews..
The home interview survey was designed as a 4 percent sample of all dwel-
ling units in the urbanized portion of the planning region. Interviews were
conducted during 1966 at each household in the sample to obtain information
regarding population, employment, household socioeconomic characteristics,
and travel patterns of members of each household.
Additional inventories and surveys were conducted to obtain information
on travel parameters. These included truck and taxi surveys, roadside
interviews (external travel, surveys), on-board transit surveys, and inven-
tories of the regional transit system and the highway system.
The data obtained from the initial surveys and inventories were utilized
in the trip generation, trip distribution, modal split, and traffic and tran-
sit assignment procedures. The results of these and other analyses were used
to forecast travel patterns for future years to the year 2000 based on various
"sketch planning" scenarios; these travel patterns have provided the basis
for developing specific transit and highway improvement plans for the region.
16
-------
The Providence Planning Process
Comprehensive land use and transportation planning was initiated in the
Providence region in 1961 under the direction of the Rhode Island Statewide
Planning Program (RISPP). The intent was specifically to develop a compre-
hensive transportation and land use plan for the Providence region.
Inventories and surveys were conducted during the first stage of the
planning process and again several years later in 1964. Home interview sur-
veys were conducted in the Providence area using sample rates of 5 to 10
percent. Other surveys were also conducted to establish transit use patterns
and trip characteristics of vehicles entering the region. Inventories of
land use, housing, and the regional highway and transit systems were conducted
as well. These data were utilized to develop trip generation, trip distribu-
tion, and traffic assignment analyses required in the travel forecasting
process.
During 1970, a decision was made to update the existing origin-destination
data base. This decision was based on several factors including (1) a major
portion of Interstate Highway System in the region had changed travel patterns
and characteristics significantly since the last survey was taken, (2) the
existing data base had two base years (1961 and 1964), therefore the reliabi-
lity of models developed from the data was somewhat diminished, and (3) the
evolving planning philosophy focused on small area analysis rather than corridor
analysis, which generally required a higher level of confidence in the model's
ability to produce reliable results.
The 0-D survey update involved a small sample size (overall, about 0.42
percent of the households). Interviews were conducted both by telephone and
in person to obtain data of the type previously discussed. This update also
included external 0-D surveys and on-board transit surveys. These data, sup-
plemented with the earlier data, were used to develop new travel pattern es-
timates for the region.
Planning Data Used in the Cold Mode Analysis
The same basic types of data were used for both the Pittsburgh and the
Providence analyses. The first data element that will be discussed is trip
distribution data. Trip distribution data were provided in the form of trip
tables indicating the number of auto-driver trips generated from each analysis
zone, that used a designated network link. These tables were generated using
the Urban Planning System 360 battery program LINKUSE.1 This program identi-
fies all origin-destination pairs and the number of trip interchanges asso-
ciated with each pair that have assigned travel paths that traverse a speci-
fied link, combination of links, or node. Trip distributions of this type
were obtained for 24 links and nodes in Pittsburgh and 36 links and nodes in
Providence. The primary purpose of these data was to identify the origin
zones of all trips using each study link.
The second data element that will be discussed is travel time data.
Travel time data in the form of skim trees were obtained from each planning
agency. Skim trees are simply tabulations of the one-way travel time between
17
-------
a zone pair over the assigned network route. A criterion used in selecting
specific study links was that one end of the link had to be located at a
centreid connector. This permitted the analysis of only the skim tree data that
applied to the study links. In other words, it was specified that the skim
tree data should indicate the travel time from specific zones centroids (those
centroids located at the study link) to all other zone centroids rather than
from each zone centroid in the network to all other zones. Although this pro-
vided a significant cost saving, it also introduced some error into the analyses
by the fact that the travel times are from the zone located at the study link
to all other zones rather than from all zones to the zone adjacent to the study
link. It was assumed that the travel time from zone A to zone B is the same
as the time from B to A. In most instances this assumption is valid; however,
in instances where this assumption does not hold, the differences are quite
small. The travel times indicated zone-to-zone travel; the interest, however,
was zone-to-link travel time. This was obtained by subtracting the travel time
increment associated with the zone centroid connector (adjacent to the study
link) from the total zone-to-zone travel time. The travel time data described
here were compiled using the Urban Planning System 360 battery program FMTSKIM1
Another data element used was the home interview trip record file for both
regions. These are computer tapes containing data derived from the home inter-
views. The following list provides an example of the types of information re-
garding each person trip that are contained in the file for the Pittsburgh
region:
Origin trip purpose (home, work, shopping, school, etc.)
Origin civil division (municipality)
Destination trip purpose (home, work, shopping, school, etc.)
Destination civil division (municipality)
Mode of travel (auto driver, auto passenger, transit
passengers, etc.)
Time of trip (clock time at the midpoint of the trip)
Blocks walked, origin and destination
Type parking (curb, off street, etc.)
Time parked (minutes)
Origin SPRPC zone (analysis zone number)
Destination SPRPC zone (analysis zone number)
Home SPRPC zone (analysis zone number)
Home civil division (municipality)
18
-------
Gravity model trip purposes (home based, work, shopping, etc.)
Trip factor (expansion factor)
Highway travel time (travel time in minutes)
The data contained in these files were used to determine various trip
characteristics for different analysis districts (aggregations of analysis
zones). For example, analyses of parking characteristics were developed from
these files. An additional file - the historical record file - for each
region was also obtained. These proved useful in a number of instances where
link characteristics had to be identified.
The only other element of data obtained from the planning regions was
general background material - base maps, network maps, traffic volume data,
etc.
SUPPLEMENTARY DATA BASE
Travel Speed Analysis
One of the most critical elements in this study involved determining the
time from trip-origin distributions for each study link. As indicated, the
skim tree data obtained from the planning agencies provided this information.
Owing to the critical nature of this data element, field tests were conducted
by GCA to verify the reasonableness of the skim tree data.
Providence
Field testing in the Providence region was conducted during the period
15 to 30 December 1976. Six arterial routes radiating from the Providence
central business district (CBD) and four downtown or outlying local streets
were selected for study. The so-called "floating car" technique was used to
determine the"travel time between desighated: checkpoints along each route.
A total of three runs was made in each direction (inbound and outbound) on
all arterial routes during each of three-time periods including the morning
peak (7 a.m. to 9 a.m.), off-peak (10 a.m. to 2 p.m.) and afternoon peak
(3:30 p.m. to 5:30 p.m.); runs on the four local streets were limited to the
off-peak periods.
The results of the travel time runs are shown in Table 2.
As would be expected, the travel time runs for each time period yielded
a range of speeds for each route (see Table 2). Several methods are available
for assessing the representativeness of this type of data. The technique that
was applied here is based on the assumption that sample means of travel time
have a normal distribution for any study route. Using this assumption, tech-
niques typically associated with statistical quality control were applied to
derive estimates of the relative error for a desired confidence level, based
primarily on the observed average range for the travel time runs and the num-
ber of observations used to estimate the average range. Tables have been
19
-------
TABLE 2. TRAVEL-TIME ANALYSIS RESULTS - PROVIDENCE, RHODE ISLAND
Measured
a.m. Peak
Street Limits
Aliens Ave. Ocean St. - Eddy St.
Eddy St. Aliens Ave. - Broad St.
U.S. Route 6 Winter St. - Atwood Ave.
NJ
O Elmwood Ave. Broad St. - Warwick T.L.
Broad St. Elmwood Ave - Park St.
Smith St. Canal St. - Aliens Ave
Washington St. CBD/Fringe
Weybossett St. CBD/Fringe
Dorrance St. CBD
Adelaide St. Fringe
Direction
Northbound
Southbound
Northbound
Southbound
Eastbound
Westbound
Northbound
Southbound
Northbound
Southbound
Eastbound
Westbound
Southbound
Northbound"
Eastbound
Eastbound
Westbound
Run 1
25.2
23.5
24.2
20.0
21.0
26.5
19.2
23.5
20.3
26.1
22.8
22.0
-
-
-
-
-
Run 2
22.9
26.4
22.8
20.0
25.4
23.0
24.8
24.9
19.9
20.4
17.0
25.2
-
-
-
-
-
period
Run 3
24.5
30.9
22.3
22.0
24.3
25.5
24.7
21.8
18.9
21.2
19.1
20.9
-
-
-
-
-
Off peak
Ave.
24.2
26.9
23.1
20.7
23.6
25.0
22.9
23.4
19.7
22.6
19.6
22.7
-
-
-
-
-
Run 1
21.1
22.1
21.7
16.6
18.2
19.4
18.7
21.7
19.8
19.0
15.9
16.8
14.6
11.6
5.6
20.8
23.5
Run 2
22.3
23.4
19.5
10.3
19.6
18.3
22.4
24.2
19.9
22.4
20.2
19.3
9.8
9.6
9.0
21.4
26.3
travel
period
Run 3
24.4
23.3
16.7
14.0
21.3
17.5
25.6
21.7
17.4
19.2
16.5
17.2
12.6
10.8
11.3
20.9
24.5
speed
(mph)
p.m. peak
Ave.
22.6
22.9
19.3
13.6
19.7
18.4
22.2
22.5
19.0
20.2
17.5
17.8.
12.3
10.9
8.6
21.0
24.8
Run 1
20.2
27.2
24.7
9.8
21.0
20.9
23.5
14.0
19.8
20.8
18.4
18.9
-
-
-
-
-
Run 2
21.3
27.1.
23.8
17.1
20.1
18.1
20.2
20.5
20.3
21.8
18.1
17.7
-
-
-
-
period
Run 3 Ave.
24.3 21.9
- . 27.2
24.2
18.9 15.3
20.9 20.6
20.8 19.9
21.9
17.3
20.1
21.3
18.3 18.2
16.4 17.7
-
-
-
-
~ -
-------
developed2 indicating the relationships among the permitted error, sample size,
and average range of observed travel speed. The particular table used in this
study is presented, in part, as Table 3, below:
TABLE 3. MINIMUM SAMPLE SIZE REQUIREMENTS FOR TRAVEL TIME STUDIES WITH
A CONFIDENCE LEVEL OF 95.0 PERCENT, PRESENTED AS THE NUMBER
OF OBSERVATIONS REQUIRED*
Average range
in travel speed,
mph
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
+1.0 mph
2
3
5
6
8
10
12
15
18
21
23
27
32
34
37
Specified
+2.0 mph
2
2
3
3
4
5
6
6
7
8
9
10
11
12
14
permitted
+3.0 mph
2
2
2
2
3
3
3
4
5
5
6
6
7
8
8
error
+4.0 mph
2
2
2
2
2
3
3
3
3
4
4
5
5
6
6
+5.0 mph
2
2
2
2
2
2
3
3
3
3
3
4
4
4
5
Source: Reference 2.
In Table 3, the average range in travel speed can be estimated from the
equation:
R= E
1=1
where R = average range in travel speed (mph)
5. = the measured travel speed for th
N = the number of observations made.
S. = the measured travel speed for the i observation
21
-------
Application of this technique to the travel time data collected in Providence
yielded the results shown in Table 4.
Typically, travel speed studies are designed to produce results that have
a permitted error of about 10 percent with a confidence level of 95 percent;
attempts to improve upon this level of reliability usually cannot be justified
because of the level of field effort that would be required. Table 4 indi-
cates that most, though not all, of the derived average speeds for the Provi-
dence study are within the acceptable range of + 10 percent. Figure 10 shows
the cumulative distribution of the probable error for the entire study.
As a result of the assessment described above, it was concluded that the
travel time analyses provided an acceptable probable error (<_ 10 percent) and
would, therefore, provide a valid indication of the average travel speed for
35 of the 41 cases studied. These 35 acceptable cases, then, were used as
the basis for assessing the reasonableness of the skim tree data.
Before the measured speed data presented in Table 2 were used to eval-
uate the network data, adjustments were made to reflect a composite average
speed - that is, one that represents the average speed for both directions
of travel for a 24-hour period. This was accomplished by weighting the speeds
determined for each direction and time period by a factor reflecting the re-
lative traffic volumes for the corresponding time period and direction of
travel. To do this, several basic assumptions concerning the distribution
of traffic were made. First, it was assumed that, on a 24-hour basis, the
total volume in each direction for radial arterials is equal; hence, 50 per-
cent of the total average daily traffic (ADT) occurs in each direction (in-
bound and outbound with respect to the downtown). Secondly, it was assumed
that 20 percent of the total ADT on radial arterials occurs during each of
the peak periods (defined as the periods from 7 a.m. to 9:30 a.m. and 3:30
p.m. to 6:00 p.m.); for a 24-hour period, then, the total volumes that occur
on an arterial during the morning peak period, the evening peak period, and
the off-peak period, are (0.2.x ADT),.. (0.2.x. AD.T) . and 0.6 x ADT), respectively.
Finally, it was assumed that a distinct directional split is associated with
each of the three-time periods; during the morning peak period, 60 percent of
the traffic is inbound and 40 percent outbound; during the evening peak period,
60 percent is outbound and 40 percent inbound; and during the off-peak periods
the split is 50 percent in each direction. The assumed relative volumes
occurring on radial arterials during the three-time periods, by direction,
are indicated in Table 5.
The relative volumes shown in Table 5 were used to weight the speeds com-
puted for each case. The resultant composite speeds for each route are pre-
sented in Table 6.
Finally, a direct comparison was made of the composite speeds shown in
Table 6 and the assigned network speeds obtained from RISPP; this comparison
is shown in Table 7.
22
-------
TABLE 4. ASSESSMENT OF THE RELIABILITY OF THE TRAVEL TIME DATA COMPILED FOR PROVIDENCE
NJ
a.m. Peak period
Street
e Aliens Ave.
o Eddy St.
o U.S. Route 6
o Elmwood Ave.
« Broad St .
e Smith St.
Washington St.
Weybossett St.
Dorranco St.
o Adelaide St.
No. of
Direction observations
Northbound
Southbound
Northbound
Southbound
Eastbound
Westbound
Northbound
Southbound
Northbound
Southbound
Eastbound
Westbound
Southbound
Northbound
Eastbound
Eastbound
Westbound
3
3
3
3
3
3
3
3
3
3
3
3
0
0
0
0
0
Measured
average
speed , mph
24.2 .
26.9
23-1
20.7
23.6
25.0
22.9
23.4
19.7
22.6
19.6
22.7
-
-
-
-
-
Off-peak period
R Probable No. of
mph error, 7. observations
1.9
3.7
0.9
0.1
1.7
3.0
2.8
2.3
0.7
3.3
3.9
3.7
'-
-
-
-
-
±4
±7
±4
±4
±4
±5
±9
±7
±5
±8
±10
±9
-
-
-
-
-
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
P
Measured _
average R Probable No. of
speed, mph mph error, % observations
22.6
22.9
19.3
13.6
19.7
18.4
22.2
22.5
19.0
20.2
17.5
17.8
12.3
10.9
8.6
21.0
24.8
1.7
0.7
2.5
5.0
1.5
0.9
3.5
2.5
1.3
3.3
4.0
2.3
3.8
1.6
2.9
0.6
2.3
±4
±4
±8
±22
±5
±5
±9
±7
±5
±10
±12
±10
±18
±9
±23
±5
±7
3
2
2
3
3
3
2
2
2
2
3
3
-
'
-
-
-
m. Peak period
Measured _
average R Probable
speed, mph mph error, %
21.9
27.2
24.2
15.3
20.6
19.9
21.9
17.3
20.1
21.3
18.2
17.7
' -
-
-
-
-
2.1
0.1
0.9
4.5
0.5
2.8
3.2
6.5
0.5
1.0
0.1
1.3
-
-
-
-
-
±5
±4
±4
±18
±5
±10
±9
±17
±5
±5
±5
±6
-
-
-
-
-
-------
111
)
<
u
r 32(78%)-
U
K
CD
2E
Ul
PROBABLE ERROR (%)
Figure 10. Cumulative distribution of the probable error for travel
time runs made in Providence, Rhode Island.
24
-------
TABLE 5. RELATIVE VOLUMES ON RADIAL
ARTERIALS BY TIME PERIOD
AND DIRECTION, SHOWN AS A
PERCENTAGE OF ADT
Relative volume - % ADT
Period Inbound Outbound
a.m. Peak
p.m. Peak
Off peak
12.0
8.0
30.0
8.0
12.0
30.0
Table 7 shows fairly good agreement (good agreement being defined some-
what subjectively as being within 20 percent) between the measured and
assigned speeds in all but two instances - the Webbossett Street and Dorrance
Street routes. However, note too that these two routes involved distances of
only 0.2 and 0.1 miles, respectively, plus the speed on each route was mea-
sured only during the off-peak periods. For these reasons it is entirely
probable that the measured speeds for these two routes are not within a
reasonable range of accuracy; hence, comparisons are not appropriate.
Pittsburgh
During the week of 3 January 1977 field studies were conducted by GCA in
the Pittsburgh region. These studies focused primarily on measuring travel
time on the highway network, generally within the city. The techniques used
for both collecting and analyzing the data were essentially identical to those
used for Providence. Comparisons made as a result of the study indicated
about the same degree of agreement between measured and assigned travel time
as, did the Providence analysis. Again,, the measured and assigned .travel times
were closer for zone pairs separated by greater distances (say, more than 0.5
miles). Since no large differences in measured and observed values were de-
tected, it was concluded that the assigned zone-to-zone travel times were
acceptable for the study.
Parking Study
A second critical element in the cold mode analysis concerns the cold
soak duration. This type of information was not available directly from any
identified source in either of the two regions. Parking data were available
in various formats from the trip record files, however, so that it was pos-
sible to derive the required cold soak information. The process used was
essentially the same for both regions.
The first step in the process was to analyze the trip record files to
determine which data elements would be relevant. The file for Pittsburgh
included the following information for each auto-driver trip:
25
-------
TABLE 6. COMPUTED 24-HOUR COMPOSITE AVERAGE SPEEDS FOR PROVIDENCE
to
Average speed
Route
Aliens Ave.
Eddy St.
U.S. Route 6
Elmwood Ave.
Broad St.
Smith St.
Washington St.
Weybossett St.
Dorrance St.
Adelaide St .
A.M. peak
Direction period (mph)
Northbound
Southbound
Northbound
Southbound
Eastbound
Westbound
Northbound
Southbound
Northbound
Southbound
Eastbound
Westbound
Southbound
Northbound
Eastbound
Eastbound
Westbound
24.2
26.9
23.1
20.7
23.6
25.0
22.9
23.4
19.7
22.6
19.6
22.7.
-
-
. '-
_
-
Off-Peak
period (mph)
22.6
22.9
19.3
13.6*
19.7
18.4
22.2
22.5
19.0
20.2
17.5
17.8
12.3*
10.9
8.6
21.0 .
24.8
P.M. peak
period (raph)
21.9
27.2
24.2
15.3*
20.6
19.9
21.9
17.3*
. 20.1
21.3
18.2
17.7
-
-
-
-
24-Hour average
directional (mph)
22.9
24.6
21.0
- + .
20.8
19.9
22.3
21.4*
19.3
20.8
18.1
18.6
- **
10.9
8.6
21.0
24.8
24-Hour
composite (mph)
23.7
" +
20.5
21.94
20.1
18.3
. **
10.9
8.6
22.9
Probable error greater than 10 percent.
*Value not computed because probable error in data was considered too severe.
*V«lue considered representative even though it was derived fro* data with relative error > percent.
-------
TABLE 7. COMPARISON OF THE DERIVED AND ASSIGNED NETWORK SPEEDS FOR PROVIDENCE
o
o
o
o
e
0
0
e
Route
Aliens Avenue
U.S. Route 6
Elmwood Avenue
Broad Street
Smith Street
Weybossett Street
Dorrance Street
Adelaide Street
Total Derived composite
length, speed,
miles mph
2.6
3.8
3.6
3.1
4.4
0.2
0.1
0.5
23.7
20.5
21.9
20.1
18.3
10.9
8.6
22.9
Assigned network Actual / 7» Difference \
speed, Difference, icogpoaite - assigned x 10Q \
mph mph \ composite '
22.6
19.5
23.3
18.4
21.2
8.0
15.0
18.4
1.1
1.0
-1.4
1.7
-2.9
2.9
-6.4
4.5
4.6%
4.9%
-6.4%
-8.5%
-15.8%
26.6%
-74.4%
19.7%
-------
origin zone
destination zone
time of trip
type of parking
time parked
The data above provided the basis for determining the hourly distribu-
tion of trips departing from an area by the time they had been parked. This
was accomplished by computing the departure time for each trip (approximated
by adding the time of trip* and the time parked) and aggregating these by
the time parked; e.g., the total number of trips departing a zone between,
say, midnight and 1 a.m. was disaggregated into categories reflecting the
amount of time that they had been parked (in 1-hour increments from less than
1 hour, to between 9 and 10 hours, and an additional increment of over 10
hours). These data were developed for 12 analysis districts. Tables were
developed for each 1-hour increment of parking duration for each analysis dis-
trict, indicating the number of vehicles departing by hour of the day. The
processing was accomplished using a computer.
The result of this analysis was the definition of the fraction of total
daily trips generated in each analysis district that occur during each hour
of the day, by 1-hour increments of parking duration.
The data for Providence were processed to achieve the same result. A
different technique was used, however, owing to a different trip record
format.
*
This was defined in the file as the mid-time (clock time) of the trip.
28
-------
REFERENCES
1. FHWA Computer Programs for Urban Transportation Planning. U.S. Depart-
ment of Transportation. Federal Highway Administration. July 1974.
2. Oppenlander, Joseph C. Sample Size Determination for Travel Time and
Delay Studies. Traffic Engineering. September 1976.
29
-------
SECTION 4
STUDY METHODOLOGY
GENERAL
At the beginning of the study it was hypothesized that the distributions
of cold operating vehicles at a particular location on a street would be func-
tionally related to the type of location, the orientation of the location with
respect to the central business district, and the time of day. In this con-
nection, then, analyses were conducted of diurnal traffic patterns at numerous
characteristically different locations, in two separate cities. Again, the
analyses were based almost entirely on data derived in connection with compre-
hensive planning studies conducted during the 1960*s and early 1970's. The
specific data elements used were described previously; the purpose of this
section is to describe how these data were applied in the analyses.
SELECTION OF STUDY LINKS
The process of selecting specific links for analysis was based on the
desire to consider a variety of facility-types located in different urban
settings. Of primary interest were arterial streets although expressways and
local streets also received attention. The intent was to analyze the cold
mode characteristics of vehicles operating on (1) arterials located in the
central business district (CBD), the fringe areas, mixed commercial-residential
areas,'and residential areas; (2) expressways at one-location in each-city
only; and (3) local streets in one or two locations if adequate data were
available. The scope of the project permitted the analysis of a total of
60 links 24 in Pittsburgh and 36 in Providence.
ANALYTICAL PROCEDURE
The first step in the process involved analyzing the trip distribution
data for each study link, which resulted in defining the number of trips
generated from each analysis zone that used the study link. As indicated
previously, trip distribution data, which were provided by the regional plan-
ning agencies, included tables showing the number of trips by origin zone that
were assigned over a specific link or node. An example of this tabulation is
shown in Table C-l of Appendix C.
The next step was to identify the travel time from each origin zone to
the study link; this was accomplished using the skim tree data (see Table C-2
in Appendix C). This identified the number of trips using the link by time-
from-trip-origin; the units of time-from-trip-origin used were 2-minute
30
-------
intervals beginning with 0 to 2 minutes, through. 8 to 10 minutes. The result,
then, was a separate tabulation for each study link showing the travel time
between each analysis zone and the study link, and the number of trips from
each zone that used the study link. An example of this distribution is shown
in Table C-3 of Appendix C.
The next step involved determining the diurnal distribution of trip pro-
ductions in each analysis district. This was accomplished by analyzing the
time-of-trip distribution for each district based on the trip record file for
each city. Trip productions for each analysis district were summed by the time
that they occurred, using four time periods. The time periods included (1)
5 a.m. to 10 a.m., (2) 10 a.m. to 3 p.m., (3) 3 p.m. to 8 p.m., and (4) 8 p.m.
to 5 a.m. For Pittsburgh, this tabulation is shown in Table C-5 of Appendix C.
Trips on each study link were then summed by the analysis district from
which they originated. This distribution was applied to the data describing
the diurnal distribution of trip productions, which resulted in defining the
distribution of trips originating in each analysis district as a function of
the time that the trip occurred (see Table C-7 of Appendix C).
The distributions of travel time from trip origin to each study link were
determined from the distributions of trip origins by analysis zone for each
analysis district. Table C-8 provides an example of these distributions.
These distributions and the distribution of trips originating in each analysis
district as a function of the time that the trip occurred were used to define
the distribution of trips by time from trip origin, as a function of time for
each analysis district; an example of this distribution is shown in Table C-9
of Appendix C.
The trip record file was again analyzed to derive distributions of parking
duration (in 1-hour increments beginning with 0 to 1 hour, through 9 to 10
hours) as a function of the time period when the trips were generated, for each
analysis-district. .This distribution for Pittsburgh is shown in Table C-10.
This indicates, for instance, the fraction of total daily trips originating in
analysis district II during the period 10 a.m. to 3 p.m., that had been parked
between 6 and 7 hours prior to beginning the trip. These distributions were
applied to the distribution of link trips originating at analysis districts
by time from trip origin, as a function of time, to produce a tabulation of the
distribution of link trips by time from trip origin and by time of trip, as a
function of parking duration (see Table C-ll).
This final distribution was then used in conjunction with the data provided
by the Office of Mobile Source Air Pollution Control to identify the fraction
of cold operating vehicles occurring on each study link during each time-
period. To.minimize the inherent bias in the trip distribution used in the
analysis, the results (fraction of total time-period volume operating in the
cold mode) for each individual link were combined to form several groups with
a common element of functional category or orientation within the region. This
approach provided an indication of the cold mode distribution in areas (such
as the CBD) or in corridors at various distances from the CBD. A detailed
example for this entire procedure is provided in Appendix C.
31
-------
These final distributions were analyzed on a comparative basis to identify
any apparent patterns that would provide the basis for developing general state-
ments regarding the fraction of vehicles operating in the cold mode, at var-
ious locations within an urban area and at different times during the day.
LIMITATIONS OF THE PROCEDURE
The primary limitation in the analysis concerned the data used. In this
connection, it was indicated that several basic techniques used in developing
the data (for instance the trip distribution analysis and the traffic assign-
ment process) were based largely on theoretical relationships and a synthetic
highway network. Obviously, the relative error introduced into any analysis
that uses these'data increases as the required level of detail increases.
The analyses of individual link-trip characteristics that were performed for
this study required a great amount of detail, consequently the results are
subject to error. Of particular concern is the fact that the analysis of cold
mode operation is very sensitive to the time-from-trip-origin factor, while
at the same time, the relative error in the computed time-from-trip-origin
distribution is greatest in the time range of about 0 to 3 minutes. (This is
crucial because the duration of the cold mode phase is only several minutes.)
The reason for this is that trips can only enter the network from any zone
centroid at one point only. Consequently, trips originating close to a par-
ticular point on the network (a study link for instance) tend to be over-
estimated. An analysis was made of the sensitivity of the study results to the
distribution of trips originating very close to the study link. This was done
by adding two "penalty times" (1 minute and 2-1/2 minutes) to the computed time-
from-trip-origin; in other words, trips previously in the 0 to 2 minute range
would now be in a range of 1 to 3 minutes. As will be shown in the following
section, this added time had a very significant effect on the results.
At this point, it is not possible to dimension the accuracy of the study
results. It would be possible to measure the true distribution of the time-
from-trip-origin, however, if direct, detailed analyses were made of trip
characteristics on specific links.
32
-------
SECTION 5
RESULTS OF THE ANALYSES
PITTSBURGH
A total of 24 highway links in Pittsburgh were analyzed using data pro-
vided by the Southwestern Pennsylvania Regional Planning Commission (SPRPC)
and information obtained directly by GCA through field studies. The specific
links that were studied are listed below in Table 8 and their approximate
locations within the city are shown in Figure 11.
The focus of the analyses was to identify certain characteristics of trips
occurring on each study link and to relate these characteristics, in aggregate,
to cold mode operating parameters. Specifically, the trip characteristics that
were essential to identify were (1) the elapsed time from the trip origin to
the study link, and (2) the duration of the cold soak period prior to beginning
the trip. These data were related to cold mode operation based on the results
of studies conducted by the U.S. Environmental Protection Agency, Office of
Mobile Source Air Pollution Control.
The first result of applying the analytical process described previously
was a tabulation for each analysis link showing the percentage of total volume
occurring during each time period that was computed to be in the cold operating
mode; these percentages are shown in Table 9. Owing to certain limitations in
the basic modeling techniques used to develop the datai the results of the
analysis are more meaningful if groups of links are considered rather than
individual links. In this connection, each of the 24 links was categorized
according to its functional classification, its location with respect to the
CBD, and its setting (rural, residential, commercial, etc.); the links were
then grouped into six categories based on these attributes. The links assigned
to each category are listed in Table 10. The total number of vehicles computed
to be operating in the cold mode within each category was determined and ex-
pressed as a percentage of total trips; these are shown in Table 11. The actual
volumes of cold and stabilized vehicles are shown in Table 12.
The percentages of cold operating vehicles shown in Table 11 are in many
instances much higher than what one might intuitively expect. It is noted
that the travel times used in the analyses were representative of the network
travel time only; there was no time increment involved that would reflect the
nonnetwork operating time, which would include an initial idling period during
warm-up and the access time from a parking facility to the network. If it
were assumed that a 1-minute increment would be representative of this non-
network operating time, then the percentages shown in Table 11 would be reduced
33
-------
TABLE 8. ANALYSIS LINK DESCRIPTIONS, PITTSBURGH
No.*
Link name
Limits
Type
1 Fifth Ave.
2 Grant Ave.
3 Seventh Ave.
4 Sixth Ave.
5 Sixth Ave.
6 Sixth Ave. Bridge
7 Smithfield St.
8 Bedford Ave.
9 .Centre Ave.
10 Fifth Ave.
11 Forbes Ave.
12 Baum and Centre Aves.
13 California Ave.
14 Centre Ave.
15 Federal St.
16 Fifth Ave.
17 Forbes Ave.
18 Perryville Ave.
19 Penn-Lincoln Pkwy EB
20 Penn-Lincoln Pkwy WB
21 Fifth Ave.
22 Forbes Ave
23 Carson St.
24 Lincoln Ave.
At Liberty
At Boulevard of the Allies
At Liberty
At Grant 'Ave.
Between Wood and Smithfield Streets
Between Fort Duquesne Blvd. and River Ave.
Between Third Ave. and Blvd. of the Allies
Between Washington PI. and Crawford St.
Between Washington PI. and Crawford St.
Between Gist and Jumonvllle Sts.
Between Gist and Brady Sts.
Between Morewood and S. Millvalle Aves.
Between Yunkle and Columbus Aves.
Between Herron Ave. and Iowa St.
Between W. North Ave. and Jacksonia St.
Between S. Aiken Ave. and Ivy St.
Between Wightman St. and Murray Ave.
Between Marshall Ave. and Hawkins St.
Between Bates St. and Brady St. Bridge
Between Bates St. and Brady St. Bridge
Between S. Beliefield Ave. and S. Dlthrldge St.
Between S. Belief ield Ave. and- S. Craig St.- -
Between S. Eighteenth and S. Twentieth
Between Frankston Ave. and Meadow St.
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Collector
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Expressway
Expressway
Non-CBD Generator, Arterial
Non-CBD Generator, Arterial
Residential Collector
Mixed Collector
. *Keyed to Figure 11.
34
-------
LINK LOCATION NO.
(SEE TABLE 8)
/ LINK LOCATION
INSET
Figure 11. Study-link locations map, Pittsburgh, Pa,
35
-------
TABLE 9. PERCENTAGE OF VEHICLES OPERATING IN THE COLD MODE, BY
TIME PERIOD, FOR 24 LOCATIONS IN PITTSBURGH
Percent of vehicles operating in cold mode by time period
No.*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Link
Fifth Ave.
Grant Ave.
Seventh Ave.
Sixth Ave-
Sixth Ave-
Sixth Ave. Bridge
Smithfield St.
Bedford Ave.
Centre Ave.
Fifth Ave-
Forbes Ave.
Baum and Centre
California Ave.
Centre Ave.
Federal St.
Fifth Ave
Forbes Ave-
Perryville Ave.
Penn-Lincoln Pkwy Eu
Penn-Lincoln Pkwy WB
Fifth Ave.
Forbes Ave.
Carson St.
Lincoln Ave.
I
13.0
8.7
13.6
6.5
4.9
9.2
6.7
14.0
14.3
19.8
24.8
16.0
12.5
10.2
19.1
20.9
19.5
31.1
4.0
0.4
25.6
27.3
19.9
29.0
II
33.3
19.8
33.8
18.2
11.9
27.4
3.7
39.8
44.2
65.7
85.9
21.7
36.4
.45.3
45.3
40.4
24.6
. 24.6
4.4
0.5
41.5
33.2
31.2
37.5
III
62.2
33.0
56.3
10.2
22.3
31.2
13.4
52.0
57.3
61.8
76.8
16.3
40.1
37.8
47.4
42.8
24.7
. 26.2
8.6
0.3
52.5
46.4
31.7
33.8
IV
36.4
34.2
38.0
12.5
16.3
28.3
9.5
53.3
50.9
14.3
84.8
46.8
53.7
47.3
73.1
75.8
47.0
70.4
12.2
1.1
69.6
55.3
84.8
97.4
Total
day
41.5
25.6
39.7
11.8
15.5
25.7
9.0
42.0
44.5
44.2
69.9
23.0
35.7
34.9
45.4
43.5
27.7
.35.2.
7.4
0.5
48.7
39.7
36.9
44.7
*Keyed to Figure 11.
36
-------
TABLE 10. ANALYSIS LINKS BY GENERAL CATEGORY, PITTSBURGH
Classification
Central Business District (CBD)
Fringe* Arterials
Outer Arterials .
Expressway
Non-CBD Generators
Collectors
No.*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Links included
Fifth Avenue/Liberty Street
Grant Street/Blvd of the Allies
Seventh Street/Liberty Street
Sixth Street/Grant Street
Sixth Avenue
Sixth Street Bridge
Smithfield Street
Bedford Avenue
Centre Avenue
Fifth Avenue
Forbes Avenue
Baum/ Centre Avenue
California Avenue
Centre Avenue
Federal Street
Fifth Avenue
Forbes Avenue
Perryville Avenue
Penn-Lincoln Pkwy EB
Penn-Lincoln Pkwy WB
Fifth Avenue . -
Forbes Avenue
Carson Street
Lincoln Avenues
Description
Includes the "Golden
Triangle" bounded by
the Allegheny River
to the north, the
Monongahela River to
the south, and 1-579
to the east.
Includes links within
about 1 mile of the CBD
Includes links greater
than 1 mile from the CBD
Links are about 2.5
miles east of the CBD
Major arterials in 'the
university and hospital
areas of Oakland
Serve local traffic
ICeyed to Figure 11.
37
-------
TABLE 11. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION BY TIME
PERIOD AND BY LOCATION GROUP FOR PITTSBURGH
Percentage of the volume operating in
the cold mode during each time group
Central business district
Fringe arterials
Outer arterials
Expressway inbound
Expressway outbound
Non-CBD generator
Collector streets
I*
9.8
17.8
17.3
0.4
4.0
26.8
24.9
II*
25.3
57.3
34.0
0.5
4.3
53.5
33.9
III*
38.3
61.3
32.6
0.3
8.6
43.0
35.9
IV*
31.5
45.2
58.1
1.1
12.2
58.6
84.5
Total day
28.6
48.2
34.1
0.5
7.4
41.5
40.3
Time groups.
TABLE 12. VOLUMES OF COLD AND STABILIZED VEHICLES BY TIME PERIOD AND BY
LOCATION GROUP FOR PITTSBURGH
Location
Central business district
Fringe arterials
Outer arterials
Expressway inbound
Expressway outbound
Non-CBD generation
Collector streets
Volume operating in cold and
I*
2,143
(19,724)
429
(1,981)
3,931
(18,791)
50
(12,450)
476
(11,424)
974
(2,660)
1,835
(5,534)
II*
7,000
(20,668)
1,461
(1,089)
6,869
(13,334)
50
(9,950)
476
(10,594)
1,654
(1,438)
2,501
(4,877)
III*
17,586
(28,330)
2,754
(1,739)
12,593
(26,036)
50
(16,660)
1,744
(18,535)
3,025
(4,010)
4,547
(8,118)
(stabilized) mode*
IV*
6,703
(14,576)
1,144
(1,387)
10,746
(7,750)
100
(8,990)
1,215
(8,744)
2,555
(1,805)
4,542
(833)
Total day
33,362
(83,228)
5,788
(6,212)
34,139
(65,561)
250
(48,650)
3,911
(48,789)
8,218
(11,572)
13,426
(29,874)
Time groups.
Total assigned volumes.
38
-------
substantially; if 2.5 minutes were assumed (this might be reasonable for CBD's
or areas where large, off-street parking facilities are used) the percentages
would be reduced to an even greater degree. The impact of adding these time
increments is shown in Tables 13 and 14.
Another significant observation concerns the distribution of trip origins
by travel time from the study link. In most instances, nearly half of all
trips originating within 10 minutes of the study link originate, in zones that
are within 2.5 minutes of the study link; this can be seen in Table 15. The
point here is that the zones closest to most nonexpressway links generate a
very large fraction of the total trips using (actually, assigned to) the link.
However, the planning network and the traffic assignment process are such that
possible trip paths are very limited in number close to the zone where the
trips are generated, therefore, the trips generated from any particular zone
are forced over a relatively small number of different links in the immediate
vicinity of the zone, but then diffuse rapidly throughout the network as the
travel time from the origin increases. The result is that a significant amount
of distortion can be expected in the trip assignment process, especially during
the first stages of the trip. Unfortunately, these first stages of trip as-
signment are the most crucial in our analysis.
A further note is that the trips represented in the analysis are auto-
driver trips only. Excluded are all commercial and public transportation ve-
hicles. The effect is that the cold mode fractions shown in Table 11 are not
fractions of the total link traffic, rather it reflects about 90 percent of
the total. This means that the percentages can be reduced by up to about 6
percent if it is considered that vehicles other than private autos are in the
stabilized mode.
Again, a general conclusion that can be made is that the cold mode frac-
tions shown in Table 11 are overestimated somewhat because of limitations in
the traffic assignment process that tend to overload links with trips generated
from-nearby zones. At- this point, however, one can .only..speculate about the
degree to which this overestimating occurred.
PROVIDENCE
For Providence, a total of 36 highway links were analyzed. The data used
for the analyses were obtained from the Rhode Island Statewide Planning Pro-
gram (RISPP), and supplemented with field studies conducted by GCA. The links
analyzed are listed in Table 16 and their locations are shown in Figure 12.
The procedure used in the analysis of the Providence data was identical
to that used for Pittsburgh. The first result of applying the procedure was
the tabulation of the percentages of cold operating vehicles during each of
the four time groups for each study link; these are shown in Table 17.
The results are, again, more meaningful if portrayed in group form. Five
groupings were defined and the individual links were aggregated into these.
Table 18 identifies the categories and indicates which links are included in
39
-------
TABLE 13. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION WITH
1 MINUTE ADDITIONAL ACCESS TIME, PITTSBURGH, PA.
Percentage of time-period volume
operating in cold mode
Location
II III IV Total day
Central business district 7.1 15.2 40.8 22.3 25.2
Fringe arterials 14.3 23.0 40.5 43.8 32.1
Outer arterials 15.6 18.7 25.9 44.1 25.6
Non-CBD generator 20.7 23.7 32.9 39.1 29.9
Collector streets 20.3 17.2 25.1 71.4 29.8
_
Time groups.
TABLE 14. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION WITH
2.5 MINUTE ADDITIONAL ACCESS TIME, PITTSBURG, PA.
Location
Percentage of time-period volume
operating in cold mode
I*
Central business district
Fringe arterials
Outer arterials
Non-CBD generator
Collector streets
5
10
14
19
10
.4
.4
.2
.2
.8
II
8.
16.
11.
14.
10.
*
9
0
6
9
8
III*.
29
38
19
26
18
.2
.4
.6
.4
.8
IV* Total
17.
38.
32.
30.
57.
4
2
8
9
1
17.
25.
19.
23.
21.
day
8
5
2
8
5
*
Time groups.
40
-------
TABLE 15. DISTRIBUTION OF TRIPS BY TIME FROM TRIP ORIGIN - PITTSBURGH
Link
No.*
1
2
6
8
9
13
15
18
22
23
Link
name
Fifth Ave.
Grant Ave.
Sixth St. bridge
Bedford St.
Centre St.
California Ave.
Federal St.
Perryville Ave.
Forbes Ave.
Carson St.
Percentage
0.5 - 2.5
1615
15.0
12.8
30.5
21.9
2.8
28.4
8.5
34 '.4
28.3
of total daily link trips by time from trip
2.5 - 4.5
27.5
1.0
14.6
33.5
25.5
0
15.8
21.4
12.1
2.0
4.5 - 6.5
6.2
30.2
3.7
0
9.1
3.2
7.3
25.3
18.2
4.3
6.5 - 8.5
7.3
1.0
3.0
6.3
14.6
24.7
15.7
5.4
11.9
8.2
8.5 - 10.5
2.7
3.9
6.0
7.5
0.9
16.9
1.0
10.0
4.3
8.8
origin
>10.5
39.6
48.9
59.9
22.2
28.0
52.2
31.8
29.4
19.1
48.4
*Keyed to Figure 11.
-------
TABLE 16. ANALYSIS LINK DESCRIPTIONS, PROVIDENCE
No.'
Link name
Limits
Type
1 Atwells Ave.
2 Broadway
3 Dorrance St.
4 Dorrance St.
5 Dyer St.
6 Dyer St.
7 Washington St.
8 W. Exchange St.
9 Weybossett St.
10 Aliens Ave.
11 Aliens Ave.
12 Angell St.
13 Atwells Ave.
14 Atwells Ave.
IS Broadway
16 Charles St.
17 Douglas St.
18 Eddy St.
19 Elmwood St.
20 Hope St.
21 N. Main St.
22 Smith St.
23° Waterman St.
24 Aliens Ave.
25 Aliens Ave.
26 Atwells Ave.
27 Broad St.
28 Eddy St.
29 Eddy St.
30 Elmwood St.
31 1-95 SB
32 1-95 NB
33 Adelaide St.
34 Elmgrove Ave.
35 Blacks tone Blvd.
36 Kinaley Ave.
At Sabin and Empire Sts.
At Sabln and Empire Sts.
At 1-195 Ramp and Dyer St.
Between Weybosset and Westminster Sts.
At Fulton St.
At Custom House St.
Between Revere and Aborn Sts.
At Sabin St. and Exchange Terr.
Between Richmond St and Park PI.
Between Public St. and 1-95 Ramp
Between 1-95 Ramp and Oxford St.
Between Prospect and Thayer Sts.
Between Pasquale Ave. and Knight St.
Between Valley and Culter Sts.
Between Knight and Almy Sts.
Between Admiral and Ashburton Sts.
Between Camden and Chad Brown Sts.
Between Public St. and 1-95 Ramp
Between Broad and Friendship Sts.
Between Rochambeau and Vassar Aves.
Between Olney and Benefit Sts.
Between Orms and Calverley Sts.
Between Prospect and Thayer Sts.
Between Earnest St. and 1-95 Ramp
Between Montgomery and Washington Aves.
Between Manton and Glenbridge Aves.
Between Adelaide and Oxford Sts.
Between Earnest St. and 1-95 Ramp
Between Broad St. and Washington Ave.
Between Reservoir Ave. and Adelaide St.
Between Broad St. and Broadway
Between Broad St. and Washington St.
Between Elmwood Ave. and Broad St.
Between Lloyd and President Aves.
Between Lloyd and President Aves.
Between Eagle and Acorn Sts.
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
CBD Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Fringe Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Outer Ring Arterial
Expressway
Expressway
Local Street
Local Street
Residential Collector
Industrial Collector
*Keyed to Figure 12.
42
-------
U)
1-0
KEY:
|LINK LOCATION NO.
(SEE TABLE 14)
LINK LOCATION
Figure 12. Study-link locations map; Providence, R.I.
-------
TABLE 17. PERCENTAGE OF VEHICLES OPERATING IN THE COLD MODE, BY
TIME PERIOD, FOR 36 LOCATIONS IN PROVIDENCE
Percent of vehicles operating in cold mode by time period
No.*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Link
Atwells Ave.
Broadway
Dor ranee St.
Dorrance St.
Dyer St.
Dyer St.
Washington St.
W. Exchange St.
Weybossett St.
Aliens Ave.
Aliens Ave.
Angell St.
Atwells Ave.
Atwells Ave.
Broadway
Charles St.
Douglas St.
Eddy St.
Elmwood St.
Hope St.
N: Main: St.
Smith St.
Waterman St.
Aliens Ave.
Aliens Ave.
Atwells Ave.
Broad St.
Eddy St.
Eddy St.
Elmwood Ave .
1-95 SB
1-95 NB
Adelaide St.
Elmgrove Ave.
Blacks tone Blvd.
Kinsley Ave.
I
15.8
4.5
53.0
17.4
14.1
63.0
20.2
23.6
18.2
18.0
9.4
6.8
16.6
19.0
7.0
9.7
6.6
26.5
24.0
24.9
12.5
4.7
16.9
15.9
23.0
29.5
58.8
11.4
29.4
27.2
0.9
3.2
25.3
6.2
19.8
23.6
II
51.3
28.0
70.3
23.0
60.8
76.4
64.8
69.3
39.1
41.7
34.3
29.4
41.6
52.2
26.6
31.9
33.8
62.5
50.6
70.6
-40.2
42.2
53.0
59.6
28.1
49.3
71.8
35.4
63.6
26.9 S .-.
ia.o
19.9
43.0
40.4
61.8
48.7
III
55.3
23.7
78.3
63.0
65.9
85.8
67.9
97.3
54.0
45.6
47.7
49.4
48.4
53.7
37.2
49.8
43.7
60.7
53.6
75.1
62.3
54.4
60.1
48.0
58.6
62.9
76.7
51.0
61.9
65.1
16.7
25.5
61.5
44.3
63.7
52.5
IV
40.0
11.4
50.0
39.3
36.0
84.2
46.3
65.3
28.9
21.2
21.2
33.2
27.1
21.8
17.2
21.1
18.9
27.7
7.6
72.4
38.7
35.1
40.7
21.7
28.8
31.9
65.8
27.8
32.3
34.1
7.6
10.9
38.1
22.2
40.5
40.0
Total
day
44.7
18.1
68.8
39.6
48.5
80.5
55.3
71.1
39.9
34.4
31.7
33.9
36.8
41.1
25.1
32.5
29.4
47.8
39.1
64.4
42.9
38.7
47.6
39.5
38.9
47.8
71.2
34.1
50.2
42.8
11.8
16.7
47.0
31.5
51.6
47.3
Keyed to Figure 12.
44
-------
TABLE 18. ANALYSIS LINKS BY GENERAL CATEGORY, PROVIDENCE
Classification No.*
Central Business District (CBD) 1
2
3
4
5
6
7
8
9
Fringe Arterials 10
11
12
13
14
15
16
17
18
19
20
21
22
23
Outer Arterlals 24
25
26
27
28
29
30
Expressway 31
32
Local/Collectors 33
34
35
36
Links included
Atwells Ave./Sabin and Empire Sts'.
Broadway/Sabin and Empire Sts.
Dorrance St./I-195 Ramp
Dorrance St.
Dyer St. /Fulton St.
Dyer St. /Custom House St.
Washington St.
W. Exchange St./Sabin St.
Weybossett St.
Aliens Ave.
Aliens Ave.
Angell St.
Atwells Ave.
Atwells Ave.
Broadway
Charles St.
Douglas St.
Eddy St.
Elmwood St.
Hope St.
N. Main St.
Smith St.
Waterman St.
Aliens Ave.
Aliens Ave.
Atwells Ave.
Broad St.
Eddy St.
Eddy St.
Elmwood St.
1-95 Southbound
1-95 Northbound
Adelaide St.
Elmgrove Ave.
Blackstone Blvd.
Kinsley Ave.
Description
Includes the area bounded
by 1-95 to the west, 1-195
to the south and the
Providence River to the
nor the and east
Includes links within a
ring 1 mile of the CBD"
Includes links 1 mile
beyond the CBD
Links on portions of 1-95
forming CBD border
Serves local traffic
*Kcyed to Figure 12.
45
-------
each. The individual link distributions within each category were summed and
a composite cold mode distribution was developed for each group. These distri-
butions are shown in Table 19, while Table 20 shows the actual volumes computed.
Again, the distribution of cold operating vehicles appears to be somewhat
different from what one might expect. The same comments concerning the pos-
sibility of the cold mode fraction being overestimated apply to Providence
as well as Pittsburgh; in this connection the time-from-trip-origin distribu-
tions for a sample of 10 locations are shown in Table 21.
In order to illustrate the sensitivity of the analysis to the time-from-
trip-origin distribution, additional time increments of 1 minute and 2.5 min-
utes were added to the final time-from-trip-origin distributions computed
previously, and new percentages of cold operating vehicles were derived. The
results of adding these time increments are shown in Tables 22 and 23; the
impact of relatively small time differentials is significant.
46
-------
TABLE 19. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION, BY
TIME PERIOD AND BY LOCATION GROUP FOR PROVIDENCE
Location
Central business
district
Fringe arterials
Outer arterials
Expressway inbound
Expressway outbound
Local/collectors
Percentage of the volume during
group operating in the cold
I*
20V 7
12.1
22.0
3.2
0.9
15.1
II*
58.0
37.9
44.3
19.9
18.0
46.2
III*
67.1
48.3
56.6
25.5
16.7
52.8
IV*
41.. 9
23.5
27.6
10.9
7.6
30.7
each time
mode
Total
53.0
34.5
41.6
16.7
11.8
40.8
*,
Time groups.
TABLE 20. . VOLUMES OF COLD AND STABILIZED VEHICLES BY TIME PERIOD AND BY
LOCATION GROUP FOR PROVIDENCE
Location
Central business
district
Fringe arterials
Outer arterials
Expressway inbound
Expressway outbound
Local/collectors
Volume operating in cold and
I*
3,598
(13,766)
2,159
(7,661)
2,828
(20,595)
339
(10,374)
58
(6,347)
631
(3,553)
II*
16,432
(11,920)
5,459
(6,861)
12,159
(19,888)
2,036
(80,192)
1,101
(5,015)
2,820
(3,290)
III*
29,174
(14,319)
10,551~
(8,098)
22,475
(24,076)
4,951
(13,863)
1,855
(9,275)
4,719
(4,220)
(stabilized) mode+
IV*
6,895
(9,555)
2,343
(6,143)
4,931
(16,077)
969
(7,950)
406
(4,927)
1,039
(2,346)
Total
56,099
(50,217)
20,514
(28,760)
42,393
(80,637)
8,095
(40,378)
3,420
(25,564)
9,209
(13,373)
*Time groups.
"^"Total assigned volumes.
47
-------
TABLE 21. DISTRIBUTION OF TRIPS BY TIME-FROM-TRIP-ORIGIN - PROVIDENCE, R.I.
oo
Link
No.*
2
7
9
12
16
22
25
28
30
36
Link name
Broadway
Washington St.
Weybossett St.
Angell St.
Charles St.
Smith St.
Aliens Ave.
Eddy St.
Elmwood St.
Kinsley Ave.
Percentage
0.5-2.5
10.5'
45.4
25.2
9.8
16.9
14.9
31.5
43.0
28.9
25.7
of total
2.5-4.5
6.0
15.9
9.2
26.7
18.4
32.9
9.6
5.9
21.2
29.4
daily link
4.5-6.5
13.0
2.8
30.0
24.2
14.2
15.8
5.7
6.8
11.3
23.5
trips by
6.5-8.5
6.1
5.9
12.0
18.5
10.4
9.7
12.2
14.1
5.2
9.8
time from trip
8.5-10.5
8.0
4.9
5.9
12.5
5.8
6.6
9.9
8.0
4.5
5.1
origin
>10.5
56.4
25.1
17.7
8.3
34.3
20.1
31.1
22.2
28.9
6.5
*"Keyed to Figure 12.
-------
TABLE 22. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION WITH 1
MINUTE ADDITIONAL ACCESS TIME - PROVIDENCE, R.I.
Percentage of
I*
Central business 11.7
district
Fringe arterials 6.4
Outer arterials 12.8
Local/collectors 7.9
time period volume operating in cold mode
II* Iir* IV* Total
41.0 51.6 32.5 39.0
23.2 34.6 15.6 23.8
25.6 43.0 19.7 28.8
32.7 38.9 19.6 28.2
Time groups .
TABLE 23. PERCENTAGE OF VEHICLES
ADDITIONAL ACCESS TIME
IN COLD MODE OPERATION WITH 2.5 MINUTES
- PROVIDENCE, R.I.
Percentage of
Location
I*
Central business 1.0
district
Fringe arterials 1.0
Outer arterials 2.4
Local/collectors 1.8
time period volume operating in cold mode
II* III* IV* Total
18.2 35.8 20.8 23.1
10.8 19.5 8.1 11.8
11.2 27.7 10.7 15.6
13.4 22.2 9.4 14.6
Time groups.
49
-------
SECTION 6
CONCLUSIONS OF THE STUDY
GENERALIZATION REGARDING COLD MODE DISTRIBUTION
Obviously, one would hope that the distributions identified by this type
of analysis would exhibit definite patterns that would provide a basis for de-
veloping general statements regarding cold mode operations. Unfortunately,
patterns of this nature are not evident at this point.
Comparisons of several combinations of cold mode distributions were made
using an analogical approach. First, direct comparisons were made of the dis-
tributions for similar locations in both cities (see Tables 11 and 19), by both
time group and for the fraction of the total daily trips. Also, the relative
percentages of cold operating vehicles were considered for each time grouping
and location category in both cities but, again, the cold mode fractions do
not appear to display consistent patterns. Perhaps a valid conclusion is that
there may not necessarily be similarities in the cold mode distributions from
city to city, or at least between Pittsburgh and Providence. Factors such as
the density and distribution of parking facilities within a city would have
an extremely strong influence on the eventual distribution of cold operating
vehicles on the street network, as would the relationship between volume de-
mand and capacity on the street network. Parameters such as these would not
be expected to correlate closely among different cities; for that matter, they
would not necessarily correlate positively among different areas of a particu-
lar city.
Further comparisons were made among locations within each city. The spe-
cific intent was to determine whether or not patterns could be detected on a
basis other than the type (CBD, Fringe, etc.) of location. For both cities
the distribution at various points along major radial corridors is shown in
Tables 24 and 25.
These tables again do not indicate the types of distributions that one
might expect. For instance, it might be expected that the fraction of cold
operating vehicles would be greater at locations further from the CBD in the
morning (especially along arterials serving residential areas) and, conversely,
they would be greater closer to the CBD in the afternoon. These types of
patterns appear to a limited degree within the three corridors in Pittsburgh,
but not in Providence (see Tables 24 and 25).
50
-------
TABLE 24. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION
BY DISTANCE FROM CENTRAL BUSINESS DISTRICT
FOR TWO CORRIDORS IN PROVIDENCE, R.I.
*
No.
1
13
14
26
11
18
24
28
25
29
Link name
Atwells Ave
Atwells Ave
Atwells Ave
Atwells Ave
Aliens Ave
Eddy St
Aliens Ave
Eddy St
Aliens Ave
Eddy St.
n. . Percent of vehicles operating in
- ___. cold mode by time period
from CBD, J K
miles
0.0
0.7
1.3
2.2
0.9
1.6
2.4
I
15.
16.
19.
29.
15.
15.
24.
8
6
0
5
8
2
6
II
51.
41.
52.
49.
44.
55.
36.
Ill
3
6
2
3
6
6
5
55.
48.
53.
62.
52.
48.
59.
3
4
7
9
6
5
3
IV
40.0
27.1
21.8
31.9
23.6
21.9
29.6
Total day
44.
36.
41.
47.
37.
38.
41.
7
8
1
8
7
6
. . -
6
Keyed to Figure 12.
51
-------
TABLE 25. PERCENTAGE OF VEHICLES IN COLD MODE OPERATION
BY DISTANCE FROM CENTRAL BUSINESS DISTRICT
FOR THREE CORRIDORS IN PITTSBURGH, PA
*
No.
6
15
18
8
9
14
12
10
11
21
22
16
17
n . Percent of vehicles operating in
Link name from^BD. C°ld m°de by time period
Sixth Ave Bridge
Federal St
Perryville Ave
Bedford Ave
Centre Ave
Centre Ave
Baum & Centre Ave
Fifth Ave
Forbes Ave
Fifth Ave
Forbes Ave
Fifth Ave
Forbes Ave
mlles I II III
0.3 9.2 27.4 31.2
1.1 19.1 45.3 47.4
2.3 31.1 24.6 26.2
0.4 14.2 43.2 56.2
2.1 10.2 45.3 37.8
3.1 16.0 21.7 16.3
1.1 21.6 72.8 67.1
2.8 27.0 35.0 43.1
3.9 20.3 33.6 42.5
IV
28.3
73.1
70.4
51.5
47.3
46.8
39.0
58.9
69.7
Total day
25.7
45.4
35.2
43.9
34.9
23.0
53.2
41.7
36.6
Keyed to Figure 11.
52
-------
CONCLUSIONS
The analyses performed indicate that strong patterns in cold mode distribu-
tion do not necessarily exist from city to city; in fact, it appears that
trends do not exist on a citywide basis, either. Tables 11 and 19 do show that
relatively large fractions of arterial traffic may be in the cold mode during
the late morning, afternoon, and early evening hours. In addition, the frac-
tion of cold operating vehicles on expressways is relatively small compared
with other types of facilities.
At this point it is appropriate to provide recommendations regarding
values reflecting the percentage of cold operating vehicles for different time
periods and at several types of urban locations. Obviously, these recommenda-
tions are subjective to a large degree, and are based on a relatively limited
study. The percentages of cold operating vehicles presented here then, are
intended to provide guidance in selecting values to use in a particular in-
stance, rather than providing absolute values.
Table 26 provides suggested values to be used in accounting for cold mode
operation. Included are ranges of cold mode percentages for 10 types of road-
way facilities by time period for three separate "cases." The three "cases"
reflect the observed effect of adding small time increments (i.e., 1 minute
and 2.5 minutes) to the time-from-trip-origin distribution. Case I is based
on use of the time-from-trip-origin distribution computed originally while
Case II and Case III are based on adding a 1-minute and a 2.5-minute increment,
respectively, to the original time-from-trip-origin distribution.
In using the information provided in Table 26, a certain amount of sub-
jective judgment is required. The user must consider carefully the specific
location or area under analysis and judge which of the three cases is appro-
priate. A suggested criterion for making this decision is the nature of the
area being considered; for the very densely developed portions of an urban
area that are characterized by congested traffic-and-generally low.travel -
speeds, Case III, which reflects slow traffic movement, would be appropriate.
For less densely developed portions of an urban area where traffic movement
is not severely impeded, either Case I or II would be appropriate. The spe-
cific suggestion, then, is that the user select either Case I, II, or III
based on the relative level of traffic congestion and impedance within the
particular urban area.
Further, it is suggested that the propensity be toward selecting the
higher values within the ranges specified in Table 26 when the maximum 1-hour
percentage of cold operating vehicles is required. Also, the nature of the
specific location being analyzed should be assessed when selecting a particu-
lar value from Table 26. For instance, if the location is an intersection
located in the vicinity of several major parking facilities, a higher value
would be appropriate. Conversely, if the location is a collector street, for
instance, downstream from a busy off-ramp from an expressway, lower levels
would be appropriate.
53
-------
TABLE 26. SUGGESTED RANGES OF VALUES OF THE PERCENTAGE OF VEHICLES
OPERATING IN THE COLD MODE FOR VARIOUS CONDITIONS OF
TIME AND LOCATION
Case*
_I_
II
III
General location
CBD
Fringe areas
Outer arterials
Local/collector streets
Expressways :
Within core area and
fringes: inbound"1"
Within core area and
fringes: outbound
Outer portion of urban
area : inbound"*"
Outer portion of urban
area : outbound"*"
Special generators out-
side the CBD
CBD
Fringe aereas
Outer arterials
Local/collector streets
Expressways:
Within core area and
fringes: inbound"1"
Within core area and
fringes: outbound"*"
Outer portion of urban
area: inbound"*"
Outer portion of urban.
area: outbound^*"
Special generators out-
side the CBD
CBD
Fringe areas
Outer arterlals
Local/collector streets
Expressways :
Within core area and
fringes: inbound"1"
Within core area and
fringes: outbound"*"
Outer portion of urban
area : inbound"*"
Outer portion of urban
area: outbound"*"
Special generators out-
side the CBD
Morning peak
hours ,
10 -
10 -
15 -
10 -
3 -
1 -
3 -
3 -
25 -
5 -
5 -
10 -
5 -
2 -
1 -
2 -
2 -
15 -
1 -
1 -
5 -
5 -
1 -
1 -
1 -
1 -
15 -
20
20
25
20
5
3
5
5
40
15
15
20
20
4
3
4
4
25.
6
15
15
15
3
3
3
3
20
Midday off-peak
hours ,
20 -
25 -
30 -
35 -
15 -
15 -
2 -
2 -
30 -
15 -
20 -
15 -
15 -
10 -
10 -
2 -
2 -
20 -
5 -
10 -
10 -
10 -
10 -
10 -
1 -
1 -
10 -
50
60
50
50
20
20
4
4.
50
45
25
25
35
20
20
4
4
25
20
20
35
15
15
15
3
3 .
20
Kvun 1 ti&
,111.1 <-.l|-ly
Evening peak morning off-peak Total
hours, hours, day
40
40
30
35
20
15
2
15
45
30
30
20
25
15
10
2
10
30
25
15
15
15
10
10
1
10
20
- 70
- 65
- 60
- 55
- 30
- 20
- 4
- 20
- 60
- 50
- 45
- 45
- 40
- 25
- 20
- 4
- 20
- 35
- 40
- 40
- 30
- 25
- 20
- 15
- 3
- 15
- 30
25
20
25
20
10
5
4
10
50
20
15
15
20
10
5
3
10
35
15
10
10
10
10
5
2
10
25
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
~
50
45
60
70
15
10
6
15
65
35
45
45
70
15
10
5
15 '.'
40
25
40
15
60
15
10
4
15
35
25
25
30
30
15
10
3
10
40
20
20
20
25
10
10
3
10
25
. 15
10
15
10
10
10
2
10.
20
- 55
- 50
- 40
- 45
- 20
- 15
- 5
- 15
- 60
- 40
- 35
- 30
- 30
- 20
- 15
- 5
- 15
- 35
- 25
- 30
- 20
- 25
- 15
- 15
- 4
- 15
- 30
Case I - no.access time added; Case II - 1-minute additional access time; Case III - 2.5-minute addition-
al access time.
With respect to the CBD.
54
-------
RECOMMENDATIONS FOR FURTHER STUDY
In view of the several deficiencies and inadequacies of existing data, it
is recommended that studies be undertaken to obtain and analyze data specific
to a number of urban locations to provide an indication of the reasonableness
of the results obtained here.
55
-------
APPENDIX A
FEDERAL TEST PROCEDURE
A methodology has been developed that provides the basis for analyzing the
emission characteristics of motor vehicles in a comprehensive and consistent
manner. This methodology, referred to as the Federal Test Procedure (FTP),
establishes the standard for conducting tests of automotive emission character-
istics, which are used as the basis for certifying that a particular type of
vehicle meets the appropriate federal emission standards. The FTP prescribes
in detail the entire procedure for the conduct of dynamic testing of either
production or prototype vehicles powered by combustion engines
The FTP involves determination of mass emissions of carbon monoxide, hy-
drocarbons, oxides of nitrogen, and carbon dioxide produced by a test vehicle
during operation through a standard driving schedule. Testing is performed on
stationary vehicles using a dynamometer to simulate actual highway driving.
The driving schedule, referred to as the EPA Urban Dynamometer Driving Sched-
ule, consists of three phases - cold transient, hot stabilized, and hot tran-
sient. Each phase is comprised of a specific series of acceleration, decelera-
tion, cruise, and idle events; the duration and rates of acceleration, decelera-
tion, and cruise, and the duration of idle events are specified by the Dyna-
mometer Driving Schedule. Figure A-l provides both a trace of the instantan-
eous speed as a function of elapsed time, and a description of the events as-
sociated with each cycle.
The test procedure begins with an initial vehicle soak at ambient temper-
atures in the range of 68°F to 86°F, for a period of between 12 and 36 hours.
Upon completion of the initial soak period, the vehicle is moved (not under its
own power, however) onto a calibrated dynamometer. The vehicle is then in-
strumented and all emission collection apparatus are installed and connected
to the sampling system. Exhaust emission testing begins when the vehicle is
started, and continues through the three phases of the Dynamometer Driving
Schedule. The sampling system is designed to collect exhaust samples and ana-
lyze pollutant concentrations for each phase of the test on a separate basis.
The sample obtained for the first phase (cold transient) is referred to as
Bag 1 emissions. This sample represents the first 505 seconds (3.598 miles)
that the vehicle is operated and includes five cycles of acceleration and
deceleration from a stopped position, plus five idle events. It is of interest
to note^ that the J>05 seconds associated with cold transient emissions_reflects
a convenient point (when the vehicle is idling)at'which the sampling device
can be switched from Bag 1 to Bag 2, rather than reflecting some absolute
function between cold mode emissions and operating time.
56
-------
»0
to + 909
te+!37l
GO
t.+24T«
-r*""T.
f \
V / \ A n
A/^ J A /In/1
r \r I ml
^v
f\ /\ n Al VVN ^ /s^ rt XI & « ^\ «
'if l/l/\/u .. .11 ull UinJ \n
PHASE IDo >
ITT ^
fc + I9TI
PHASE IKk
j^^\
A \
j-i / \ A A
j^\ /* \/lA/
r w l/tm
0 2OO 400 6OO BOO 1000 I2OO 1400 16OO I BOO ZOOO 2200 2400 26OC
ELAPSED TIME.itconds
SUMMARY OF EVENTS
Time, sec
Event
Start engine, begin idle*.
Begin first driving cycle; maximum speed during first cycle is
32.A mph.
End first driving cycle, begin idle period.
End idle period, begin second driving cycle; maximum speed during
second cycle is 56.7 mph.
End second driving cycle, begin Idle.
End idle, begin third driving cycle; maximum speed during third
cycle is 36.5 mph.
End third driving cycle, begin idle.
End idle, begin fourth driving cycle's maximum speed during fourth
cycle is 30.1 mph.
End fourth driving, cycle, .begin, idle.. ....
End idle, begin fifth driving cycle, maximum speed during fifth
cycle is 35.5 mph.
End fifth driving cycle, begin idle.
End idle, begin sixth driving cycle, maximum speed during sixth
cycle is 26.8 mph.
End sixth driving cycle, begin idle.
End idle, begin seventh driving cycle, maximum speed during
seventh cycle is 27.0 mph.
End seventh driving cycle, begin idle.
End idle, begin eighth driving cycle, maximum speed during eigth
cycle is 26.5 mph.
End eighth driving cycle, begin idle.
L0+125
U0fl63
0+333
0+346
L0+397
L0+403
"0+429
0+505
'o+sio
'0+552
'0+568
C0+620
'0+645
C0+680
Figure A-l. Summary of the EPA Urban Dynamometer Driving Schedule, used in the
Federal Test Procedure.
57
-------
Time, sec
Event
0+694
L0+727
0+767
uO+957
L 0+959
0+1023
0+1052
L0+1100
0+1153
-0+1168
0+1187
0+1196
'0+1244
bO+1313
'0+1337
0+1367
0+1371
0+1971
uO+2476
End idle, begin ninth driving cycle, maximum speed during ninth
cycle is 23.5 mph.
End ninth driving cycle, speed is 0 mph, begin tenth driving cycle,
maximum speed during ninth cycle is 28.6 mph.
End tenth driving cycle, speed is 0 mph, begin eleventh driving
cycle, maximum speed during eleventh cycle is 34.3 nph.
End eleventh driving cycle, begin idle.
End idle, begin twelfth driving cycle, maximum speed during
twelfth cycle is 28.5 mph.
End twelfth driving cycle, begin idle.
End idle, begin thirteenth driving cycle, maximum speed during
thirteenth cycle is 28.0 mph.
End thirteenth driving cycle, speed is 0 mph, begin fourteenth
driving cycle, maximum speed during fourteenth cycle is 27.0 mph.
End fourteenth driving cycle, begin idle.
End idle, begin fifteenth driving cycle; maximum speed during
fifteenth cycle is 23.5 mph.
End fifteenth driving cycle, begin idle.
End idle, begin sixteenth driving cycle, maximum speed during
sixteenth cycle is 21.8 mph.
End sixteenth driving cycle, begin idle.
End idle, begin seventeenth driving cycle, maximum speed during
s-venteenth cycle is 29.1 mph.
End seventeenth driving cycle, begin idle.
End idle, begin eighteenth driving cycle, maximum speed during
eighteenth cycle is 22.0 mph.
End eighteenth driving cycle, begin idle.
End idle, shut engine off.
Restart engine, repeat idle cycles and driving cycles for the
period t0 through 10+505.
End driving cycle, speed is 0 mph, end test.
Idle in this context implies that the vehicle is stopped and the clutch
disengaged (if standard transmission) or the transmission is in the "drive"
position (if automatic).
Figure A-l (continued).
Summary of the EPA Urban Dynamometer Driving Schedule,
used in the Federal Test Procedure.
58
-------
The second phase emissions are referred to as Bag 2 emissions representing
hot stabilized operation. This sample reflects emissions generated during the
subsequent 866 seconds (3.902 miles) of the Driving Schedule. At the end of
the second phase (at fco + 1371 - see Figure A-l) the engine is shut down and
remains off for 10 minutes. After this 10 minute hot soak period, the engine
is restarted and the vehicle is operated through the remainder of the Driving
Schedule, which duplicates the first five cycles that comprise the first phase.
The emissions collected during this phase are referred to as Bag 3 emissions.
The weighted mass emissions from vehicles tested are computed from the
equation:
wm
0.43
^ + Y
ct s
}/ /D + D } + 0.57 ( Y, ^ + Y } I / D, + D \ (A-l)
/ \ ct s / \ ht s / \ ht s)
where Y^n = weighted mass emissions of each pollutant (i.e., HG, CO, NOX,
or 002) in grams per mile;
Yct = mass emissions as calculated from the cold transient phase of
the test, in grams per test phase;
Y^t = mass emissions as calculated from the hot transient phase of
the test, in grams per test phase;
Ys = mass emissions as calculated from the hot stabilized phase of
the test, in grams per test phase;
Dct = measured driving distance during the cold transient phase, in
miles ;
D^t = measured driving distance during the hot transient phase, in
miles;
Ds = measured driving distance during the hot stabilized phase, in
miles .
It can be seen from equation (A-l) that the weighted mass emissions (Y^^)
reflect the weighted average emission rate during two identical 7.5 mile trips
(Phase "L plus Phase II, and Phase Illb plus Phase II as shown in Figure A-l).
The implication is that the hot stabilized emission characteristics are
identical for both trips. Since the driving cycles in Phases I and Illb are
identical, equation (A-l) can be simplified to:
Y =
wm
=[o.43/Y..+ V }+ 0.57 /Y, . + Y \ 1
I \ ct s I yhts/J
/D.
(A-2)
and
wm
/D
(A-3)
59
-------
where D = D, _ + D = D . + D .
t ht s ct s
The Federal Test Procedure covers many areas of vehicle testing including
vehicle preconditioning, test data recording procedures, fuel specifications,
instrumentation required, conduct of instrumentation and apparatus calibration,
emission computations, testing procedures, etc. For details regarding these
areas of the FTP, the reader should consult reference 1.
60
-------
REFERENCES
1. Federal Register. Vol. 42, No. 124, Part III, pp. 32906-33004.
28 June 1977.
61
-------
APPENDIX B
OVERVIEW OF THE COMPREHENSIVE PLANNING PROCESS
INTRODUCTION
To fully understand the procedures used in this study, it is necessary to
have a general understanding of the comprehensive planning process since data
developed for and used in the comprehensive planning process were used exten-
sively in the analysis of cold mode operation. The following provides an over-
view of the planning process in general terms, which will satisfy the requirement
stated above. It is strongly recommended, however, that the reader who is
totally unfamiliar with the techniques used in the comprehensive planning pro-
cess refers to a more detailed treatise of the various techniques used in order
to fully undestand the implications of applying this type of data in the anal-
ysis described in the main body of this report.
BACKGROUND
The Federal-Aid Highway Act of 1944 first provided Federal-aid funds on
a regular basis for use in urban areas. These funding programs included pro-
visions for conducting transportation planning activities as well as for de-
signing and implementing specific projects. The Federal-Aid Highway Act of
1962, however, required that programs for Federal-aid highway projects approved
subsequent to 1 July 1965, in urban areas with populations greater than . .50, 000., ....
must be based on a continuing, comprehensive, and cooperative transportation
planning process. The planning process that evolved from this concept is known
as the 3-C planning process.
Several basic features of the 3-C process are important to consider.
First, the continuing nature of the process focuses on developing and contin-
uously evaluating both short- and long-range highway and transportation plans
that are soundly conceived to meet the objectives and basic requirements of
the state and the urban communities. The comprehensive nature of the process
requires that (1) economic, demographic, and land-use parameters be included in
the planning activities, (2) that forecasts be made of future demands for all
transportation modes, both private and public, for the movement of both per-
sons and goods; (3) that the process include consideration of terminal and
transfer facilities, and traffic control systems; and (4) that the entire area
within which the forces of development are interrelated, and which is expected
to be urbanized within the forecast period, be included. To satisfy this re-
quirement the planning process must characterize several specific elements for
the present (with respect to the time when the planning procedure is implemented)
condition; the characterization include inventorying and analyzing:
62
-------
the economic factors that affect regional development
population
land-use patterns
transportation facilities (including mass transportation)
terminal and transfer facilities (goods movement)
traffic control devices and systems
zoning ordinances, subdivision regulations, building codes, etc.
financial resources
social and community-value factors
and more recently:
air and water quality
Finally, the cooperative aspect ensures that the final product of the process
reflects the various needs and desires of all segments of the population, and
accounts for the plans and policies of all relevant local, state, and federal
agencies.
To define the primary objective of the comprehensive planning process,
one could say that it is a mechanism for (1) identifying both the short- and
long-range regional transportation requirements, (2) defining various methods
for satisfying these requirements, and (3) evaluating alternative transpor-
tation and land-use scenarios in order to optimize future regional growth
patterns.
Several organizational structures have been used to carry out the com-
prehensive planning process. Among the most widely used are:
Centralized state staff generally drawn from the State's
Department of Transportation;
Semi-independent organizations established as an ad hoc joint
effort and responsive to a multiagency board;
Council of Governments comprised of elected or appointed represen-
tatives of communities within the region;
Regional Planning Commission; and
Contract study organizations such as private consulting firms
retained by a state or regional agency holding overall responsi-
bility for the planning effort.
The organization responsible for conducting the planning effort generally re-
ceives input from several committees. These include technical committees, pol-
icy committees, and citizens advisory groups. These committees provide guidance
63
-------
on administrative matters, set technical objectives and standards, and gen-
erally serve to ensure that the planning effort is conducted in accordance
with the needs and desires of the region. Generally, a study director or-
chestrates the entire process.
Obviously, there are wide differences in the scope of planning projects
from city to city. The size of study areas ranges from less than a few hundred
square miles to several thousand square miles. The cost of conducting these
studies is directly proportional to the size (in terms of population and land
area) of the area being studied. It has been indicated that the costs asso-
ciated with conducting a comprehensive transportation planning study are in
the range of $1.00 to $1.50 per capita.1 The nature of these studies is such
that the duration of the study period may be 2 to 3 years or longer.
A discussion of every aspect of the comprehensive planning process is
clearly beyond the scope of this report. Detailed discussions of the entire
process can be found in several texts and reference material (for instance,
References 1 through 4). The discussion here is limited to only the specific
portions of the planning process that relate to the highway transportation
system, specifically, as it concerns the movement of persons within a region.
TRAVEL FORECASTING
One of the most important concepts in the transportation planning process
is travel forecasting. Travel forecasting is the process from which detailed
descriptions of regional travel activity are developed. Urban travel patterns
are a function of several observable and quantifiable elements, viz:
Land-use patterns including location and intensity
Social and economic factors particular to the area
The availability of various types of transportation facilities
in the area.
Travel forecsting, then, actually is a process that derives relation-
ships between regional travel patterns and the aggregate of the interrelation-
ships that exist among the three elements listed above.
The general process of travel forecasting consists of four phases. First,
inventories are conducted in order to define economic activity, population,
land use, travel patterns, transportation facilities, etc. for the region. In
the second phase, the inventory data are analyzed and the relationships between
travel patterns and the socioeconomic, land use, and transportation facility
characteristics are defined. The third phase involves redefining the regional
travel patterns based on various scenarios of future economic, social, land
use, and transportation facility characteristics. Finally, the fourth phase
provides for the evaluation of alternative transportation systems and land use
plans based on the forecasted travel patterns defined in the third phase.
Each of the phases associated with travel forecasting warrants additional
discussion.
64
-------
Transportation Inventories and Analyses
Inventories
Various inventories provide the foundation for the entire planning process.
Inventories are conducted to define characteristics that generally relate to:
land use
population
economic factors
transportation facilities
travel patterns
legislation (laws, bylaws, ordinances)
financial resources
environmental aspects
Many of the analyses and inventorying tasks are conducted at a subregional
level. Planning regions are generally divided into analysis zones, which can
range in size from a fraction of a square mile, to areas that encompass entire
townships. Certain inventories and analyses are conducted at levels that in-
clude several individual analysis zones; these larger areas are referred to
as analysis districts.
For the purposes of this discussion, the most important inventories are
those that concern transportation facilities and travel patterns. Inventories
of regional transportation facilities include physical highway inventories,
highway use and operating inventories, and mass transit inventories. Physical
inventories of the regional highway network are conducted in order to define
.certain functional and physical characteristics of each roadway in the net-
work. Functional characteristics include the type of service that a roadway
provides. Several general categories are typically used to describe the
functional classification of each street in the network; examples of these
categories include:
principal arterial system streets
major arterial system streets
minor arterial system streets
collector system streets
local streets
Additional categories may be used to designate each section according to
its location within the region in terms of being in the central business
district, in the urbanized portion of the region, in a rural area, or others.
Other portions of the physical highway inventory identify features of
each roadway section that affect capacity. Included are elements such as the
65
-------
number of lanes, lane width, lane use, Intersectional traffic control, speed and
and parking regulations, etc.
Travel time or speed data are also included in the inventorying process.
Studies are usually conducted on portions of the highway network to determine
speed characteristics during both peak and off-peak travel periods.
Data regarding regional transit systems are included in the inventory.
The data required for a transit inventory must provide good definition of a
number of aspects including:
Service area
- routes and type of service
- locations of transfer points, terminals, etc.
- nature of the areas through which the routes pass
- potential new service areas
System data
- route mileage
- physical resources (number and condition of transit vehicles
and facilities)
- franchise limitations or restrictions
- fare structure
- organizational structure and financial condition of transit
companies
- methods of financing
Revenue data
- summaries of annual, average, weekday, and average Saturday
or Sunday revenue for each route
- revenue per vehicle-mile for each route
Cost data
- salaries, facility operating costs, vehicle operating and
replacement costs, etc.
Operating characteristics
- annual and average weekday vehicle-miles and vehicle-hours
of operation
- route data including streets utilized, hours of operation,
average headways, etc.
66
-------
Passenger volumes
- annual and average weekday user volumes by system and route
- locations of maximum activity (loadings and unloadings)
- counts at terminals and major loading points, etc.
Obviously, the transportation facilities inventories are quite extensive.
The items described above are generally the types that are included in the
regional inventory. It is possible that the inventory for a specific region
would contain additional elements as well.
Transportation Analyses
The second major portion of the regional inventory concerns identifying
travel patterns. The interest is in identifying the number of trips that
occur between all pairs of analysis zones (internal-internal trips), between a
point located outside the region and each analysis zone (external or internal-
external trips), and between points located outside the region but where a
portion of the trip occurs within the region (external-external trips); the
modal distribution of these trips; and the travel path taken. The primary
method used to compile these data is to conduct origin^destination surveys
throughout the region.
Four types of origin-destination (0-D) surveys are used to determine
regional vehicle and person travel; viz:
Home interview surveys
Truck and taxi surveys
Cordon line surveys*
On-board surveys of transit vehicles
Home interview surveys provide a measure of the travel undertaken by re-
sidents of the region. It is generally agreed that these trips account for
between 80 and 90 percent of the total number of trips that occur within an
urban area. Sampling techniques are used in the conduct of the home inter-
view surveys. The overall sampling rate is a function of the population of
the area being analyzed. General guidelines indicate that a sample rate of
about 12 percent should be used in urban areas with populations in the range
of 50,000 to 150,000 while a sample rate of about 4 percent would be adequate
in areas where the population exceeds 1 million.2 The basic unit used in the
selection of a sample is the housing unit. Specific housing units are se-
lected at random from each analysis zone and interviews are conducted through
personal visits or by telephone.
Cordon line surveys involve determining travel characteristics of persons en-
tering an area defined by a cordon line around the area. Studies are made at
the cordon line and can include vehicle counts, classification counts, inter-
views, etc.
67
-------
The data obtained using the home interview technique generally include
information regarding (1) the number of persons by age and sex occupying the
dwelling, (2) family income, (3) vehicle ownership, (4) various physical charac-
teristics of the dwelling, (5) employment categories of household members, and
(6) trip characteristics. Trip characteristics include a record of every trip
made during a specified day by each individual (usually over 5 years old) in
the household, indicating the exact points of origin and destination, the mode
used, time that the trip began and ended, purpose, parking characteristics
(if appropriate), and others.
The second type of 0-D survey characterizes the movement of trucks and
taxis through the region. Interviews are conducted at truck terminals and
taxi dispatch points to determine the routes and frequency of truck service,
and the extent of the demand for taxi service.
Cordon line surveys are conducted on the major highways radiating from
the study region. Interview stations are established along these highways at
the edge of the study region. Information on trip origins and destinations,
purpose, frequency, vehicle occupancy, etc., is obtained for trips crossing
the cordon line. Several techniques are used to obtain the data including
direct interviews, distributing mail-back questionnaires, and mailing ques-
tionnaires to motorists observed passing the survey station using their
license plate number for identification. The sampling rate used in these
studies ranges from about 25 percent for conducting direct interviews, to 80
or 90 percent for distributing questionnaires.
On-board surveys are conducted to identify the travel patterns of transit
patrons or airline passengers. Data are obtained regarding the trip origin
and destination, waiting time at the bus stop or terminal, modes used to get
to the bus stop or terminal and to get from the bus stop or terminal to the
final destination, frequency of travel, reason for using the bus (or for
flying), etc.
The data samples obtained from the travel pattern inventory are expanded
based on the actual usable sample size obtained and the expanded data is
transcribed onto a trip record file (usually on a computer readable medium).
The expanded data, then, represents regional travel patterns at the analysis
zone level. An important process that is undertaken at this point is to
evaluate the travel data obtained. Several techniques can be used to assess
the quality of the travel data. The most basic test involves comparing ex-
panded dwelling unit data such as total population or number of vehicles avail-
able, to similar information found in other independent sources. This par-
ticular test merely indicates whether or not the relatively easily obtained
dwelling unit data have been collected and expanded correctly; it does not
necessarily indicate whether or not the trip data are acceptable. Other tests
do provide this indication. The most widely used test is the screenline com-
parison. A screenline is an imaginary line dividing the study area into two
parts. Manual classification counts are taken on all roadways intercepted by
the screenline. These counts are compared to the number of trips having, an
origin and destination on different sides of the screenline as determined by
the expanded interview data. Comparisons made in this fashion on an hourly
basis provide a good check of the accuracy and completeness of the expanded
68
-------
trip data. If the comparison shows that the number of screenline crossings
from the expanded trip data are at least 85 to 90 percent of the total ground
count crossings, the trip data are considered adequate. Where the expanded
trips are less than about 85 percent of the ground count trips, adjustments
must be made to account for the underreporting. Several methods are suggested
in the literature for deriving adjustment factors. These generally involve
determining where the underreporting error is likely to occur and using this
determination as the basis for adjusting trips. For instance, the number of
peak-hour trips may compare favorably, therefore the off-peak hours could be
adjusted. Also, only internal trips are adjusted owing to the much higher
probability that external trips are more fully-reported since much larger
sample rates are used to define external trip patterns.
Once the analyst is satisfied that the trip data are adequate, trip tables
are generated. These trip tables show the number of trip interchanges that
occur between all analysis zone pairs. Numerous tables can be generated to
show any particular type of trip pattern. For instance, separate tables could
be generated showing total person trips, auto-driver trips only, auto-driver
work trips only, etc. These tables are usually computer generated.
Other inventories are also conducted during the first phase of the travel
forecasting process. These inventories are concerned with defining the land-
use and population characteristics of each analysis zone (which, by the way,
is typically comprised of a standard unit of area used by the U.S. Bureau of
Census, such as a census tract or enumeration district). Other inventories
that are required concern the economic factors and financial resources of the
region, and laws and ordinances regulating land use. These inventories will
not be discussed in detail here, but are mentioned for completeness.
The first phase, then, of the travel forecasting process involves de-
fining the existing nature of travel in the region. It is interesting to
note that about half of the total cost of a comprehensive study is reflected
in this first phase.5 v .--..,-.
Regional Transportation Analysis
Trip Generation Analysis
The second phase of the travel forecasting process focuses on analyzing
the data developed in the first phase in order to define and quantify the
functional relationships that exist between travel and socioeconomic charac-
teristics. The first of several specific analyses that are undertaken as part
of this phase is trip generation analysis. Trip generation analysis attempts
to identify causal relationships between the number and types of trips
Apparently, underreporting is the dominant problem in the survey process.
As was indicated previously, external trip characteristics are identified
from cordon line studies where sampling rates are generally 20 to 25 percent
whereas internal trip data is obtained from home interviews using sampling
rates of 2 to 10 percent.
69
-------
generated in an area such as an analysis zone, and certain features of that
zone that relate to its population, land use, employment, and general social
and economic nature. The general process used is to develop correlations
between observed or measured trip-making and the socioeconomic, demographic,
and land use features of the area. These correlations are made using mathe-
matical techniques such as regression or factor analysis with either single
or multivariate approaches. The outcome of the analysis is an equation that
expresses the intensity of trips generated from a zone as a function of a set
of variables such as land use, employment, etc. The significance of this
analysis is that it provides a means of assessing the impact of changes in,
say, land use of a zone in terms of travel activity; this capability is es-
sential in travel forecasting.
Trip generation analyses identify these functional relationships for
various types of trips (such as truck trips, auto-driver trips) and by trip
purpose. In separate but related analyses, relationships are derived that
define the probable modal split between transit and private auto trips. The
discussion to this point has merely skimmed the surface of the topic of trip
generation analysis. Detailed discussions of the topic can be found in the
literature (see References 1, 2, 3, 4, 6, and 7, for instance).
Trip Distribution Analyses
The second type of analysis undertaken concerns trip distribution. Trip
distribution analyses focus on defining and quantifying the relationships that
determine how trips generated in one zone are distributed to other zones in
the region. Two general types of trip distribution models have received much
use. The first type uses growth factor techniques to project future trip dis-
tribution patterns. The most widely used method of this type is the Frater
Growth-Factor Method. This method was developed using the basic premise that
the distribution of future year trips generated from a zone is proportional
to the distribution pattern identified in the base year travel surveys, modi-
fied by both a growth factor for the zone pair being considered, and an adjust-
ment factor for the relative attractiveness of other competing 'zones. This
technique is used primarily'for estimating future external trips or for inter-
urban trip interchanges in small areas.
A second type of procedure referred to is the Gravity Model, which has
been the most widely-used model in trip distribution analysis work. This
model provides a totally synthetic approach to the problem of trip distribu-
tion. The concept is that zone-to-zone interchanges are directly proportional
to the total trip production at the origin and the total trip attraction at
the destination, and inversely proportional to some function of the spatial
separation of the two zones. In this relationship, the spatial separation
factor provides a friction factor that reflects the relative travel time
between any zone pair. This process does not utilize observed (base year)
trip distribution data directly; hence, it is entirely a synthetic process,
Since it is a synthetic model, calibration is required. In the calibration
process the factor describing the spatial separation is adjusted until several
criteria are met regarding agreement between observed and simulated trip-
length frequency distributions and the need for satisfying a trip-end
70
-------
constraint equation that basically states that the total number of trips attrac-
ted to a zone must be equal to the number of trips generated from the zone.
Several other models have been developed and used successfully. Two of
these are the Intervening Opportunities Model and the Competing Opportunities
Model. These are also synthetic in the sense that the distributions derived
are based entirely on mathematical relationships.
Traffic Assignment
The third element involves a process for synthesizing the interzonal trip
movements over the regional highway network. This process, known as traffic
assignment modeling, involves the computer selection of the minimum impedance
path between a zone pair; the trips that occur between the zones are assigned
to the roadway comprising the minimum impedance path. A basic requirement is
that the regional highway network be described in a format suitable for com-
puter processing. The first step in the traffic assignment process, then,
involves selecting and coding the network to be used. All streets and highways
in the region cannot be included in the network. An important consideration in
the process, then, is the selection of streets that will be included. The
principal criterion for selecting the network is judgment. It is important
that a balance be achieved between having too few or too many streets in the
network. Generally, all streets and roadways that carry substantial volumes
(obviously, this is a relative term; substantial volume might be 5,000 vehicles
per day in a large city, while in a small city it might imply 1,000 vehicles
per day) and have functional classifications of collector or higher; local
streets are generally not included in the network. The basic network is des-
cribed in terms of nodes and links. A node is an intersection of two or more
network roadways, the intersection of a network roadway and a State or municipal
boundary, or an arbitrary location along a network street that serves to limit
distances between nodes. Each node is assigned a number for identification.
The roadway section between nodes is referred to as a link. All interzonal
trip activity is assumed to occur from a single point in each analysis zone.
This poirit'is the zone centroid which theoretically is the center of the trip
generating activity for the zone. Zone centroids are connected to the street
network by a centroid connector, which actually is a synthetic street link.
All centroids are assigned an identification number.
The next procedure involves describing the operating parameters of the
network. Data obtained during the first phase is used to describe the physical
and operational aspects of each link in the network in terms of functional
classification, travel speed, capacity, etc. At each intersection on the
network, turn penalties and prohibitions are defined. Turn penalties reflect
the additional impedance to traffic flow that occurs at intersections, and
also reflects the relative impedance for each possible movement (for instance,
higher impedance would be reflected for left turns than for through movements
at most intersections). A sample network map is shown in Figure B-l.
The link data and intersection data are coded using a standard format
and transcribed onto a medium such as a computer tape. The entire file
describing all elements of the network is called the Historical Record file.
71
-------
ruiiMntn
zmti i-u
CXTCMIM. COODON STATIOm M-tO
1»%I KOOC HO W*
LOHOC9T «LLOV OI9T. «90MI
LOfMEST ALLOW TIME tOONM.
TU*N PCIULTT 03O«I«.
l£S£Np.
' M , LCytTN or tmn m «*
« 9MCOOM L«MI MHPH
LOCM CONHCCTIOM TO ZOWt CtMTIKW
COHMM tmt
Figure B-l. Sample network map (Source: Reference 2).
72
-------
The actual assignment process is accomplished using a standard computer
program that uses the historical record and the trip distribution results as
input. The program computes the minimum impedance path between each zone pair
As the assignment process takes place, the number of trips assigned to each
link is accumulated in storage. The primary output of the assignment process
is a listing of the assigned volume for each link. A satisfactory assignement
is one where the total assigned volume within a corridor approximates the
total measured volume within the corridor. Calibration of the assignment
model is undertaken by adjusting the impedance factors (travel speed and turn
penalties) on various links. Usually, several iterations are required to
obtain satisfactory results.
Several types of assignment models are used. The most straightforward of
these is the all-or-nothing model where all trips from an origin zone to a
specific destination zone are assigned to one travel path only. A much more
sophisticated technique is capacity restraint. This is an iterative process
whereby the accumulated volumes on each link are continually compared with the
links capacity. As the assigned volume increases the impedance factor is in-
creased at a predetermined rate based on the ratio of the volume to capacity.
Analyzing Future Scenarios (Forecast)
The third phase of the travel forecasting process involves forecsting
changes in regional population, employment, land-use, etc. based on perceived
trends and alternative development scenarios for the region. The impact on
travel demand at the analysis zone level is defined using the relationships
derived in the trip generation analyses conducted during the second phase. The
impact on trip generation for all zones in the region is analyzed and regional
trip distribution characteristics are redefined, again using the procedures
in Phase II. The highway network is updated to reflect expected new facilities
that will be available by the forecast year, and traffic assignment modeling
is conducted. Assignment modeling using forecast year travel and the base year
network is also conducted to identify where new facilities are needed.
Systems Analysis
This phase involves additional analysis of the transportation system to
determine specific requirements to meet the forecast year travel demands.
Alternative scenarios of highway and transit improvements (or changes) are
considered. An outline of the elements involved in the urban travel forecasting
process is provided in Figure B-2.
An important tool used in large scale planning is a battery of standard
computer programs developed by the Federal Highway Administration. This battery
of programs, referred to as the Urban Planning System 360 battery, provides the
capability for performing an extremely wide range of analyses using the basic
data inputs described in the Phase II portion, above. Included are basic pro-
grams for building files, performing trip generation and trip distribution,
and utility programs for performing many types of special analyses. A complete
description of the UPS 360 battery is provided in Reference 2.
73
-------
INVENTORIES
ECONOMIC Acnvmr
AND POPULATION
ANALYSIS OF
EXISTING
CONDITIONS AND
CALIBRATION OF
FORECASTING
TECHNIQUES
FORECAST <
ECONOMIC ACTIVITY
AND POPULATION
PROJECTION
TECHNIQUES
TRAVEL
CHARACTERISTICS
(Highway-Transit)
FUTURE
ECONOMIC ACTIVITY
AND POPULATION
SYSTEMS
ANALYSIS
ACCURACY CHECKS
AND
TRIP TABLES
INITIAL ASSIGNMENT
AND
NETWORK ADJUSTMENT
CALIBRATION OF
TRIP DISTRIBUTION
MODEL
FUTURE
TRIP GENERATION
FEEDBACK
RECOMMENDED
SYSTEM
(HIghway-TranlU)
S
( IMPLEMENTATION
Figure B-2. General perspective of the urban travel forecasting process
(Source: Reference 2).
74
-------
REFERENCES
1. Hutchinson, B. G. Principles of Urban Transportation Systems Planning.
McGraw-Hill Publishing Company. 1974.
2. Urban Transportation Planning - General Information. U.S. Department of
Transportation. Federal Highway Administration. March 1972.
3. Transportation and Traffic Engineering Handbook. Institute of Transporta-
tion Engineers. 1976.
4. Creighton, Roger L. Urban Transportation Planning. University of Illinois
Press. 1970.
5. Hillegass, Thomas J. Urban Transportation Planning - A Question of Em-
phasis. Traffic Engineering. XIX, No. 7, 1969.
6. Guidelines for Trip Generation Analysis. U.S. Bureau of Public Roads.
June 1967.
7. Fleet, Christopher R. and Sydney R. Robertson. Trip Generation in the
Transportation Planning Process. Highway Research Record No. 240. High-
way Research Board, Washington, B.C. 1968.
8.. FHWA Computer Programs for Urban Transportation Planning. U.S. Depart-
ment of Transportation. Federal Highway Administration. July 1974.
75
-------
APPENDIX C
EXAMPLE OF THE METHOD USED TO IDENTIFY THE PERCENTAGE OF
VEHICLES OPERATING IN THE COLD MODE
Presented here is an example of the computations required to determine the
percentage of vehicles operating in the cold mode for any study link, The
example shows the computations for a link located on Forbes Avenue in Pittsburgh;
this link is identified as location 17 on the Study Link Locations Map shown
in Figure 11.
The first step in the analysis is to identify the origin-destination dis-
tributions for all trips using the link. This is accomplished through the use
of the LINKUSE computer program (described in Appendix B) and the trip genera-
tion, trip distribution, and traffic assignment data for the region. This step
was accomplished by the Southwestern Pennsylvania Regional Planning Commission
(SPRPC). The result is a computer printout showing the number of trips entering
and leaving each analyses zone that used the designated study link. Table C-l
shows this printout for the study link.
The next step involves identifying the travel times from all zones to the
study link. The FMTSKIM computer program, which generated a printout for each
zone showing the travel time to all other zones, was utilized. The printout
for zone 039 (see Table C-2), which is adjacent to the study link, was used
for the example study link. The travel times shown in Table C-2 represent
travel on (1) a centroid connector between the origin zone and the street net-
work, (2) the various highway links, and (3) a centroid connector between the
street network and the destination zone. For the purposes of this analysis,
the desired travel time is from each analysis zone to the study link, therefore,
to use the skimtree data, the travel time from the study link to the destination
zone had to be subtracted from the total zone-to-zone travel time. Since a
criterion for selecting study links was that one end-point of the link had to
be located at a centroid connector, travel times from all zones to the study
link could be determined by subtracting the travel time for the adjacent zone
centroid connector from the total travel time between the connected zone and
all other zones. Travel times on links and centroid connectors were determined
from the Historical Record File.
At the time that these computations were being made, detailed information
regarding the relationship between the cold soak duration and the cold mode
operating time was not available. Therefore, it was estimated that the upper
boundary for the time that a vehicle would be operating in the cold mode would
be 10 minutes. From this point, then, two groups of vehicle trips were con-
sidered - trips originating within 10 minutes travel time from the study link,
76
-------
and trips originating further than 10 minutes away from the study link; trips
in the first category were studied in detail while trips in the second cate-
gory were considered in general terms.
The next step in analyzing the example study link was to identify the
analysis zones that were within 10 minutes travel time from the study link,
and to determine the number of trips generated by these zones that were routed
over the study link. A tabulation of the zones and number of trips is presented
in Table C-3; these were developed from Tables C-l and C-2.
The detailed analysis of trips with origins within 10 minutes of the study
link focused on identifying various trip characteristics from the Trip Record
File, described in Section 3. This file provided information such as trip ori-
gin zone, time parked, and the (clock) time that the trip began. Aggregations
of these types of data were made for groups of zones, referred to as analysis
districts. These aggregations provided the basis for defining several critical
attributes of trips occurring on the study links as will be demonstrated. The
aggregations were made for groups of analysis zones rather than individual
analysis zones primarily because the accuracy of the data did not warrant anal-
yses at a level as fine as the zone level. The analysis districts were estab-
lished by combining zones that provide similar land-use functions; for instance,
11 individual analysis zones comprising the Pittsburgh central business dis-
trict were combined to form one analysis district. In this manner, a total of
12 analysis districts were defined by combining 138 analysis zones in and sur-
rounding the City of Pittsburgh; the individual analysis zones included in each
district are listed in Table C-4.
An important objective of the project was to identify the diurnal patterns
of vehicles operating in the cold mode. In this connection, four time group-
ings were selected, viz.:
Group I - 0500 to 1000 hours
Group II - 1000 to 1500 hours
Group III - 1500 to 2000 hours
Group IV - 2000 to 0500 hours
Using these time groupings, the Trip Record File was computer-sorted to de-
termine the fractional distribution of the trips generated in each analysis
district by time group; the result of this is shown in Table C-5.
Based on the data provided in Tables C-3 and C-4, a table was constructed
showing the distribution of trips originating within 10 minutes of the study
link by the origin (analysis district). This distribution is shown in
Table C-6.
Using the data provided in Tables C-5 and C-6, a determination was made
of the fraction of trips (originating within 10 minutes of the study link)
generated in each analysis district that occurred during each time group; these
fractions are shown in Table C-7.
77
-------
Tables C-3 and C-4 were used to develop a table showing the fraction of
total link trips originating in each analysis district as a function of the
travel time from the origin zone to the study link; these fractions are shown
in Table C-8.
The data in Tables C-7 and C-8 were used to develop Table C-9, which shows
the fraction of trips (originating within 10 minutes of the study link) by
analysis district, by time group, as a function of the time from trip origin.
The next step involved analyzing the parking characteristics of each
analysis district. This was accomplished utilizing the trip record file. A
computer-sorting routine was developed that identified the parking duration
distribution for each analysis district as a function of the time group during
which the trip began. This distribution is shown in Table C-10.
Tables C-9 (column C) and C-10 were then used to develop Table C-ll,
which shows the distribution of trips (originating within 10 minutes of the
study link) by time from trip origin, and by time group, as a function of the
parking duration. Development of this table completed the characterization
of link trips that was necessary to identify the percentages of cold operating
vehicles for each time group.
Data provided by the U.S. Environmental Protection Agency,Office of Mobile
Sources were used to define a functional relationship between cold mode opera-
tion and cold soak duration and operating time. Table C-12 provides a summary
of this relationship; indicated in Table C-12 is the duration (in minutes) of
the cold operating cycle as a function of the cold soak duration (in hours) for
ambient temperatures of 75°F and 20°F.
Tables C-ll and C-12 provided the data required to identify the cold
mode percentages during .each .time group, for the study link. Table C-12. shows,
for example, that the duration of the cold operating cycle is at least 2.4 (say
2.5) minutes regardless of the cold soak duration; therefore, all vehicles
that are within the first 2.5 minutes of operation are in the cold mode. From
Table C-ll, then, all trips that are in the 0.5 to 2.5 minute range of the Time
From Trip Origin column, are in the cold mode. For other ranges of Time from
Trip Origin, the fraction of cold operating vehicles is found by interpolation.
As an example, consider the 2.5 to 4.5 minute range. Here it is first assumed
that the mid-point time (3.5 minutes) will adequately represent the trips in
this group. Therefore, from Table C-12 it is indicated that (at 75°F) the cold
operating duration of 3.5 minutes occurs when the cold soak duration is almost
1.5 hours; hence, all trips that are within the range of 2.5 to 4.5 minutes
from their origin are assumed to be operating in the cold mode if the cold
soak duration is 1.5 hours or more. The results of applying this technique
to the Forbes Avenue study link are shown in line A of Table C-13; lines A
through F provide an overall summary of the analysis of the link.
78
-------
TABLE C-l. NUMBER OF TRIPS TRAVERSING ANALYSIS LINK, FORBES AVE. NODE 03901 TO NODE 04106
1972 NETWORK - SELECT LINK - FOR SCA L1NKUSE 03/15/77 le.5».?3 c»GC-
27
TRIP END SUMMARY FOR TRIPS MEETING SELECTED LIN* CRITERIA FOR TABGRP 4
ZONE
2
6
10
14
18
22
26
30
34
36
42
46
50
54
58
62
06
70
74
78
82
86
90
94
98
102
106
110
114
118
122
126
130
134
136
142
146
150
154
15b
162
166
170
174
178
182
186
190
194
TRIPS TRIPS
IOME
5
9
17
21
25
29
33
37
41
45
49
53
57
45
69
73
77
Bl
85
9
93
97
101
105
109
113
117
121
125
129
133
137
141
145
149
153
1S7
161
165
169
173
177
181
ias
189
193
OUT
23
73
3
2
108
38
192
227
231
72
1
~!~
__
285
25
1
47
2
1
27
27
2
1
1
1
1
1
7
236
S3
7
3
30
5
93
67
30
71
81
2
6
IN
-^
23
73
43
137
78
290
165
266
33
51
76
__
242
7
13
47
8
1
1
62
1
1
.
172
60
1
26
1
93
31
1
1
1
31
TRIP
ENDS
23
96
76
45
245
116
482
392
497
105
1
51
76
__
527
32
14
94
10
2
28
89
3
1
1
1
1
1
1
-
7
408
113
8
3
56
6
186
98
31
72
82
2
37
UPS
WT
52
55
77
65
91
2
34
269
6
65
36
154
386
77
a
i
i
26
6
50
5
3
2
67
3
69
272
2
7
23
33
5
14
9
83
27
4
4
TRIPS
IN
1
i
71
6«,
31
,3
29
290
3
131
35
81
165
439
71
'5
13
.!_
26
; 3
68
5
23
1
; 1
28
__
1
?9
278
29
41
_:..
1
3
49
2
41
22
1
1
TRIP
ENDS
53
56
148
129
127
5
63
559
9
196
71
81
319
825
148
13
4
1
52
9
118
10
26
3
68
26
3
1
98
550
31
48
23
34
8
63
11
124
49
5
5
ZCNE
3
7
11
15
19
23
27
31
35
39
43
47
51
55
59
63
67
71
75
79
83
87
91
95
99
103
107
111
115
119
123
127
131
135
139
143
147
151
155
159
163
167
171
175
179
163
137
191
195
TRIPS
OUT
42
197
70
63
106
172
1302
5
«
SOB
78
85
3
1
2
2
11
24
44
4<>
25
1
1
2
49
28
56
60
3
53
5
31
67
9
S
^
5
30
2£&o
208
;
9v5
A
5
301
104
6
101
60
29
2
3
1
1
42
1
1
1
54
117
27
1
1
37
13
24
44
I
1.
26
154
4*
5
3!)
4
61
3378
546
1V59
11
5
513
683
24
215
68
22
58
3
14
3
1
3
45
1
24
.2
-
1
T
111
221
95
84
7
62
98
29
82
117
33
30
3
83
-------
TABLE C-l (continued). NUMBER OF TRIPS TRAVERSING ANALYSIS LINK, FORBES AVE. NODE 03901 TO
NODE 04106
19T2 NfTttOftK - SELECT LINK - FO* GC* LINKUSE 03/15/77 IE.04.23 PAGE 26
TRIP END SUMMARY FOR TRIPS MEETING SELECTED LINK CRITERIA FOR TABGRP *
TRIPS TRIPS TRIP
UME OUT IN ENDS
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
TRIPS TRIPS TKI?
ZONE OUT IN tNOS
197 3
201
209 6 2
209 5 1
213
217 ~
221
225
229 ~
233
237
249
293
297
00 2*1
0 2*9
2*9
273
277
2*1
2*9
2*9
293
297
301
309
309
313
317
321
323
329
333
397
341
345
349
393
337
3*1
3*5
3*9
373
377
3*1
3«5
389
3
65
2
3
2
2
2
1
1
3
1
0
2
23
4
-
1
2
1
2
1
1
3
3
1
--
--
1
--
1
5
88
*
2
-
1
5
3
2
*
2
2
3
6
1
__
2
--
1
198
202
206
210
214
218
222
226
230
234
238
242
2*6
250
25*
258
262
266
270
27*
278
282
286
290
29*
298
302
306
310
314
318
322
326
330
33*
338
3*2
3*6
350
35*
35b
362
366
370
37*
378
382
386
*
2
78
__
-
-
*
I
-
15
23
*
-
22
2
22
14
1
1
3
1
-
1
1
r
-
-._
1
1
.
-
--
4
'
1
~
-
_
--
3
1
1
41
1
i
i
i
32
--
2
2
1
-
2
2
'
--
5
2
79
8
-
2
--
-
-
.
18
2*
5
*1
1
23
1
~
3
22
46
1
3
5
2
3
3
--
199
, 203
207
211
215
219
223
227
231
235
239
2*3
2*7
251
255
259
263
267
271
275
279
283
287
291
295
299
303
307
311
315
319
323
327
331
335
339
343
3*7
351
355
359
363
367
371
375
379
383
387
391
1
4
3
27
1
)
1
23
'
«
~
37
-
2
1
2
1
--
24
2
1
1
A
1
2
2
--
1
1
1
29
--
1
1
' --
--
1
--
5
-
«
-
-
24
__
-
--
--
1
1
~
--
1
24
1
Z
--
1
2
1
2
--
1
--
1
-
1
5
3
28
1
to
-
1
47
.
37
r
'
_-
3
2
2
1
1
43
3
3
2
3
1
3
4
1
1
2
29
--
1
*
1
20C
.?0*
208
21?
J16
220
224
223
232
236
240
244
246
252
256
260
264
268
272
276
280
284
288
292
296
300
30*
308
312
316
320
32*
32b
332
336
340
344
343
352
356
360
364.
372
376
380
38*
388
392
.
3
4
2
3
34
--
1
'
3
1
«,
23
22
1
3
1
1
1
53
34
1
1
1
1
1
1
1
1
'I
2V
3j
11
23
1
1
22
3
1
1
1
2
14
73
1
1
1
__
1
4
4
5
3
32
67
~
-
1
. --
«
14
24
5
24
44
1
6
2
2
1
1
2
~
67
57
1
1
2
2
1
2
--
-------
TABLE C-l (continued). NUMBER OF TRIPS TRAVERSING ANALYSIS LINK, FORBES AVE. NODE 03901 TO
NODE 04106
1*72 NETWORK - SELECT LINK - FOR SCA
LINKUSE 03/15/77 IR.S^.PJ
TRIP END SUMMARY F01 TRIPS MEETING SELECTED -INK CRITERIA FOR'TABGRO 4
TRIPS TRIPS TRIP
ZONE,*: OUT IN ENDS
00
3*3
397
401
409
409
413
417
421
425
429
433
437
441
443
449
453
457
461
465
469
473
477
401
485
489
493
497
501
505
509
513
517
521
525
529
533
537
541
545
549
553
557
561
565
569
573
577
581
585
1
1
2
~
-
26
2
1
1
-_
1
1
--
-
1
'
--
I
1
-_
1
1
~
1
1
2
1
28
1
1
-
I
_
-
-
~
1
.
.
.
2
2
4
--
. -
1
54
3
I
1
1
1
~
1
~
1
--
-
--
~
1
-
1
~
.
1
1
1
1
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
TRIPS TR1?S TRIP
ZONt OUT IN tNDS
394
396
402
406
410
414
418
422
426
430
434
438
442
446
450
454
458
462
466
470
474
478
482
486
490
494
493
502
506
510
514
518
522
526
530
534
538
542
546
550
554
558
562
566
570
574
578
582
586
3
22
3
1
3
22
25
44
28
1
1 -- 1
3 3
1 3
2
395
399
403
407
411
415
419
423
427
431
435
439
443
4<,7
451
455
459
463
4A7
471
475
479
483
487
491
495
499
503
507
511
515
519
523
527
531
535
539
543
547
551
555
559
563
567
571
575
579
583
587
1
1
10
1
3*6
40C
403
420
424
"25
432
436
440
444
446
456
460
464
468
472
476
430
484
488
492
496
500
504
508
512
520
524
52S
532
536
540
54 b
552
556.
560
564
563
572
376
580
584
568
11
1
1
27
1
37
1
1
_
:
1
-
1
__
1
1
. 1
24
1
26
1
25
1
1
1
~
1
_
--
--
35
1
2
55
2
t>2
1
1
1
1
2
-
-
2
__
--
--
--
__
1
. _
1
1
--
-------
TABLE C-l (continued). NUMBER OF TRIPS TRAVERSING ANALYSIS LINK, FORBES AVE. NODE 03901 TO
NODE 04106
1972 NETWORK - SELECT LINK - FOR GCA
LINKUSt 03/15/77 It.54.23 PAGf 3.C
TRIP END SUMMARY FOR TRIPS MEETING SELECTED LIN* CRITERIA FOR TA9GRP 4
TRIPS TRIPS
ZONE OUT IN
M9
TRIP
ENDS
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
oo
601
605
609
613
617
621
625
629
633
637
641
645
6*9
653
697
661
645
6*9
673
677
6*1
6tS
6*9
6*3
6*7
701
709
709
713
717
721
.729
729
733
737
7*1
. 745
7*9
753
757
761
765
769
',773
777
1
-
2
2
1
1
1
1
1
22
2*
1
2
2
28
3
3
2
2
2
2
2
1
*
26
2
1
_
_ '
1
1
1
_
~
i
i
i
i
i
i
i
i
i
i
i
i
i
23
1
1
1
--
2
1
3
1
1
1
2
1
2
22
2*
2
3
3
29
*
*
3
3
2
3
-
3
1
5
49
3
2
. --
-
_
--
590
594
598
602
606
610
614
618
622
626
630
63*
638
6*2
6*6
650
65*
658
662
666
670
67*
678
682
686
690
694
698
702
706
710
71*
718
722
726
730
734
738
7*2
746
750
754
758
762
766
770
77*
778
782
1
--
-
1
1
20
1
2
3
2
1
1
1
1
1
1
3
1
4
25
5
2
1
2
1
1
1
2
2
2
1
1
«
--
-
. ~
-''
-
~
'
~
'
~
-
~
--
-
1
T
I
-i
. i
T.
i-i
1
"T .
-
~
-.-
1
-
-
--
- '
-
--
. 4
-
i
-
--
-
-
-
l
i
20
1
2
3
2
1
1
1
1
1
1
*
1
5
25
6
2
1
3
1
1
1
2
3
2
1
1
TRIPS TRIPS
ZONE OUT IN
591
595
59V
603
607
611
615
619
623
627
631
635
639
643
6*7
651
655
659
663
667
671
675
679
683
687
691
695
6V9
703
707
711
715
719
723
727
731
735
739
743
747
751
755
759
763
767
771
775
779
783
1
'
-
1
2
1
I
1
1
1
1
2
2
1
2
1
3
3
25
27
23
2
1
29
-- '
1
3
1
2
-- .
--
-
-
--
'
~
-
' -
_-
1
-
-
-
1
-
-
1
-
-
1
1
1
2*
-
--
-
--
1
1
«
-
--
--
-
-.
1
~
TRIP
ENDS
~
-
1
1
3
1
1
1
1
~
2
1
2
2
2
2
1
4
4
26
51
23
2
1
29
2
4
1
4.
-
-
-
1
TRIPS TRIPS
ZONb OUT l!i
1R1P
59:
596
600
604
60b
612
616
620
624
628
632
636
640
644
64 a
652
656
660
664
663
672
676
680
684
638
692
696
700
704
706
712
716
720
724
72 &
732
736
740
74,
748
752
756
760
76,
76 1>
772
776
760
7S4
.
. 2
2
2
.
I
-
1
1
2
1
2
2
3
26
3
2
1
1
2
2
1
1
1
2
7
1
1
2
--
--
-
1
1
23
1
1
1
1
-
23
2
-
2
--
2
.
1
2
1
2
2
2
2
3
26
3
25
1
1
3
2
2
1
1
2
S
1
2
~
2t>
~
-------
TABLE C-l (continued). NUMBER OF TRIPS TRAVERSING ANALYSIS LINK, FORBES AVE. NODE 03901 TO
NODE 04106
1972 NETWORK - SELECT LINK - FUR GCA . . LINKU'SE 03/1S/77 lh.14.2J PAGE 31
TRIP END SUMMARY FOR TRIPS MEETING SELECTED LINK CRITERIA FOR TABGRP
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
09
785
789
793
797
801
805
809
813
817
821
825
629
833
837
841
845
849
853
657
861
865
869
873
877
881
885
889
893
89T
901
905
909
913
917
921
925
929
933
937
941
945
949
953
957
961
965
969
973
977
~
»-
1
1
1
2
1
~
1
-
1
'
-
1
__
1
1
1
1
.
--
--
1
1
1
9
_
-
-^
__
-
1
2
2
1
2
2
--
«
--
-
1
1
1
1
2
1
2
1
9
-
~
1
~
-
~
-
-
--
1
TRIPS TRIPS TRIP
ZONE OUT IN ; ENDS
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
786
7*0
794
796
802
806
810
814
813
822
826
830
834
638
842
846
850
854
858
862
866
870
874
878
882
886
890
894
898
902
906
910
914
918
922
926
930
934
938
942
946
950
954
956
962
966
970
974
976
41
1
40
9
81
10
10
1
19
1
17
IB
787
791
795
799
803
807
811
815
819
823
827
831
835
839
843
847
851
855
659
863
867
871
875
879
883
887
891
895
899
903
907
911
915
919
923
927
931
935
939
943
947
951
955
959
963
967
971
975
979
23
1
22
2
1
--
1
.
-
-
1
-
--
--
- .
.
1
.
2
1
12
11
~
--
1
--
-
--
15
1
1
-
__
-
1
1
-
--
«
--
' -
--
:
2
6
26
7
24
-
'
1
22
-
15
3
1
1
1
1
1
'
1
--
--
' 1
4
7
36
IS
TRIPS TRIPS TRIP
ZONE OUT IN ENDS
786
7*2
7">6
300
804
608
812
816
b?0
1
1
1
I
2
828
822
836
54C
344
846
852
856
860
864
&6b
872
876
&80
884
888
692
896
900
904
906
912
916
920
924
92B
932
936
940
94*
946
V5i
956
960
96
-------
TABLE C-l (continued). NUMBER OF TRIPS TRAVERSING ANALYSIS LINK, FORBES AVE. NODE 03901 TO
NODE 04106
1972 NETWORK - SELECT LIN* - FOR GCA L1NKUSE 0J/15/77 1S.S4.23 PAGE1 3?
TRIP END SUMMARY FOR TRIPS MEETING SELECTED LINK. CRITERIA FC,"* TABGRP 4 . '
TRIPS TRIPS TRIP TRIPS TRIPS TRIP TRIPS TRIPS TRIP TRIPS WnS TRIP
tONE OUT IN ENDS . ZONE OUT ; IN ENDS ZONE CUT IN ENOS ZONE CUT I'i E^OS
981 51 3 54 982 983 16 3 19
-------
TABLE C-2. TRAVEL TIMES FROM PITTSBURGH ZONES TO ZONE 039
(FORBES AVE. LINK 03901 TO 04106)
..... 1932-IRAMEU-»IID-TERMlJaL_TlM£i_£QR..SEi.ECt£0 ZONES ._
03/15/77
TABLE 1 FROM ZONE 39 CONTINUED.- -
ZONE
00
10
20
JO
40
50
60
70
80
JW*
100
110
120
130
140
150
160
170
180
190
200
no
220
230
24O
25O
26O
zm
280
290
300
310
32O
330
34O
350
360
370
380
39O
400
410
_ -420
430
440
. . .450- _
46O
470
480
490
500
_. 51ft. _
520
13
16
6_ .,
4
15
6
10
20
22
21
14
23
23
31
12
27
id
16
19
29
24
44
IS
2O
25
22
26
31
23
24
35
28
31
36
35
27
34
42
5i
25
39
. _!«.__
30
42
.50.
48
63
56
65
66
...SA.
74
11
IS
7
2
13
4
9
21
25
23
18
26
24
25
14
24
12
16
20
24
26
42
15
24
27
24
25
30
26
27
33
32
35
34
38
29
37
4O
47
27
34
-24._
3O
32
_49 .
41
61
49.
58
57
Al_
79
12
16
5
6
14
5
11
20
24
22
18
26
23
21
16
if
23
22
25
25
43
18
25
29
28
29
25
3D
37
27
39_.
29
i;38
28
37
41
28
44
._24__
38
33
.... 53_
45
56
53
62
61
63
78
18
13
.2
5
13
- 3
12
19
26
22
20
26
26
17
16
27
16
21
19
25
29
44
17
21
28
31
26
31
21
-29
39
26
3ft.
26
35
3O
34
41
22
31
47
32
36
_ -53
45
69
52
69
63
__-i8
78
18
9
JJL
a
11
3
11
19
16
19
18
26
27
14
-2Q
22
15
22
19
23
31
14
17
28
3O
25.
36
25
31
38
25
35
24
35
3O
42
43
21
31
46
.22.- _
31
37
59_ _
45
70
-50.
72
61
jfc5_ ...
74
12
8
J3
10
9
4-
13
20
17
22
19
18
26
7
-22
20
13
20
21
26 .
33__
10
19
27
25
31
26
36
33
31
39
25
4O_
23
38
31
38
45
_22
35
43
_23__
33
40
55-
49
66
5_7
76
59
__64. .
70
10
7
1ft... _
11
8
1O
12
22
18
23
20
19
17
7
25
It,
19 :
26
25
33
13
20
22
34
28
36
30
_-3A_
24
29
43
;*,,
30
39
47
_20_
35
40
_Jtt_,
39
47
52-.
53
64
57
72
65
. 67
67
10
7
5
12
a
. 5
12
18
20
'24
22
20
18
1O
-33. . .
23
14
23
27
28
34-
11
22
23
35
34
37
29
33
23
27
3.8
31
43
_3J_
41
39
22
29
42
_.27
41
42
.50 ...
56
58
65
72
63
-i.71. .
72
8
6
...3
16
9
17
21
18
27
21
22
21
10
..41
9
14
20
23
26
37_
14
21
3*
22
33
35
20
28
34
22
28
43
31
43
34
43
41
_26
30
38
..28. _
35
45
56__
49
66
59_
71
60
49-
7O
11
6
13
6
16
20
21
17
23
21
27
12
9
16
lt_
25
27
38
14
2O
33
17
30
36
19
26
25
31
34_
32
44
31
49
52
26
34
21
2SL._
42
47
54- ...
57
6O
61
65
6O
_76.
66
Note: Times shown are minutes.
85
-------
TABLE C-2 (continued) . TRAVEL TIMES FROM PITTSBURGH ZONES TO ZONE 039
(FORBES AVE. LINK 03901 TO 04106)
1972-_TRAJfet- *MO -TERMINAL -TIMES JrOR SELECTED ZONES
03/15/77
TABLE . 1 FROM ZONE 39_CONT1NUEQ_ ...,_.
ZONE
00
77 72 75
80 86 91
_90_- -._«!_ __ .9.7_
65 56 52
75 79 78
at 96. aa
74 83
98 89
._ 102
56 64
74 80
al __ 75 ___ a.1 ____
82 82 95 95
92 105 1O6 56
61 ____ 50 ______ 52
58 60 62 60
71 74 65 63
S3 54 50.
41 4O 38 4O
46 45 44 46
41 ____ 4O
49 43 46 35
28 26 . 29 32
88 87 84
101 116 104
47 4«» S2
61 54 57
81 ai 74
49 48 47
47 48 SO
4? 45
36 42
31 32
650 50
66O 46
67O __ 48
MO 3«
69O 32
48 56 64
6O 59 87
32 ______ 30
43 40 37
47 55 50
-55_-__56
56 55
52 52
5
4O 42 36 45
43 51 48 57
5O ---- 46- -43 ---- -4O
46 55 55 49
51 -55 53 45
1 -- -32, ____ 5J __ - 55
64 68 65 61
72 74 58 6O
57 58 67
69 71 68
- 56 59-
71 67 69
69 58 59
ft, 74 73
82 80 87
76 75 69
9X ft 79-
108 92 131
U7
Note: Times shown are minutes.
86
-------
TABLE C-3. RECORD OF TRAVEL TIME AND NUMBER OF TRIPS FROM ALL ZONES
TO ANALYSIS LINK, FORBES AVENUE (NODE 03901 TO 04106)
Analysis
zone number
16
17
18
24
25
26
27
28
29
30
31
32
33
36
37
38
39
40
41
42
43
44
45
55
56
57
58
59
60
61
62
63
64
65
66
67
68
,69
70
71
145
146
147 .
148
168
169
170
225
Subtotal
All other
zones
Total
Travel time,
minutes
10
10
8
9
8
7
7
6
6
6
7
5
7
10
5
3
2
4
2
6
5
8
10
. 9
8
8
9
6
6
4
5
3
3
4
10
5
6
7
10
9
7
7
10
10
9
9
10
10
>10
Number Fraction of
of trips total trips
31
3
91
31
108
34
106
1092
38
269
172
378
192
0
227
65
1302
964
231
36
0
0
72
0
212
0
154
0
579
0
386
508
0
0
0
78
18-
285
77
85
236
272
28
104
61
93
5
0
8623 0.646
4731 0.354
13,354 1.000
87
-------
TABLE C-4. ASSIGNMENT OF PITTSBURGH ANALYSIS ZONES
TO ANALYSIS DISTRICTS
Analysis district Analysis zones Analysis district Analysis zones
I 001, 002, 003, IX 075, 076, 077,
004, 005, 006, 078, 079, 080,
007, 008, 009, 083, 084, 085,
010, 968 086, 089, 094,
095, 096, 097,
II Oil, 012, 015, 098, 104, 106,
016, 017, 018, 289
019
X 109, 110, 111,
III 013, 014, 020, 112, 113, 114,
021, 125, 128 115, 116, 117,
126, 127, 129,
IV 022, 023, 024, 130, 131, 132,
025, 026, 027, 133
028, 029, 040
XI 118, 119, 120,
V 031, 034, 035 121, 122, 123,
124, 134, 135
VI 030, 032, 033,
036, 037, 038, XII 343, 344, 345,
039, 064, 065, 360, 361, 362,
067, 146, 148, 363, 366, 393,
168, 169, 170 399, 967
VII 041, 042, 043,
044, 045, 055,
056, 057, 058,
059, 060, 061,
062, 063, 068,
069, 070, 071,
074, 145, 147
VIII 046, 047, 049,
050, 051, 052,
053, 054, 072,
073, 136
88
-------
TABLE C-5. DISTRIBUTION OF DAILY TRIPS OCCURRING BY TIME PERIOD FOR
PITTSBURGH ANALYSIS DISTRICTS
Analysis Time period
district lt IX** IIZtt ivi Total
I 0.064 0.235 0.543 0.158 1.000
II 0.072 0.281 0.402 0.245 1.000
III 0.105 0.211 0.505 0.179 1.000
IV 0.086 0.232 0.387 0.295 1.000
V 0.106 0.140 0.458 0.296 1.000
VI 0.157 0.222 0.321 0.300 1.000
VII 0.120 0.229 0.357 0.294 1.000
VIII 0.157 0.200 0.330 0.313 1.000
IX 0.120 0.199 0.332 0.349 1.000
X 0.082 0.245 0.409 0.264 1.000
XI 0.155 0.213 0.280 0.352 1.000
XII 0.116 0.231 0.351 0.302 1.000
* Derived from Trip Record File.
t Time period I: 0500 to 1000 hours.
** Time period II: 1000 to 1500 hours.
tt Time period III: 1500 to 2000 hours.
§ Time period IV: 2000 to 0500 hours.
89
-------
TABLE C-6. DISTRIBUTION OF TRIPS ORIGINATING WITHIN 10 MINUTES OF THE STUDY
LINK, BY ORIGIN (ANALYSIS DISTRICT); FORBES AVENUE LINK
(NODE 03901 TO 04106)
Analysis district
I
II
III
IV
V
VI
VII
VIII
IX
X
XI
XII
Total
Number of trips
0
125
0
2373
172
3046
2907
0
0
0
0
0
8623
Fraction of total
0.000
0.009
0.000
0.178
0.012
0.228
0.218
0.000
0.000
0.000
0.000
0.000
0.646
90
-------
TABLE C-7. DISTRIBUTION OF TRIPS ORIGINATING WITHIN 10 MINUTES OF THE STUDY LINK BY
ORIGIN ANALYSIS DISTRICT AS A FUNCTION OF TIME GROUP, FORBES AVENUE LINK
(NODE 03901 TO 04106)
Time Origin
group district
I II
IV
V
VI
VII
II II
IV ,
V
VI
.VII
III II
IV
V
VI
VII
IV II
IV
V
VI
VII
/traction of trips generatedl
1 from district by time 1
V group (from Table C-5) /
0.072
0.086
0.106
0.157
0.120
0.281
0.232
0.140
0.222
0.229
0.402
0.387
0.458
0.321
0.35,7
-
0.245
0.295
0.296
0.300
0.294
i
(Fraction of link trips \
generated from origin I
district (from Table C-6)/
0.009
0.178
0.012
0.228
0.218
E = 0.646
0.009
0.178
0.012
0.228
0.218
I =0.646
0.009
0.178
0.012
0.228
0.218
E = 0.646
0.009
0.178
0.012
0.228
0.218
I = 0.646
Fraction of total link
trips originating within
= 10 minutes of study link
by analysis district and
time group
0.001
0.015
0.001
0.036
0.026
0.079
0.003
0.041
0.002
0.051
0.050
E = 0.167
0.004
0.069
0.006
0.073
0.078
I = 0.230
0.002
0.052
0.004
0.068
0.064
E =0.190
EE = 0.646
-------
TABLE C-8. FRACTION OF TRIPS ORIGINATING IN EACH ANALYSIS
DISTRICT AS A FUNCTION OF THE TRAVEL TIME FROM
ORIGIN TO STUDY LINK, FORBES AVENUE LINK
(NODE 03901 TO 04106)
Analysis
district
II
Subtotal
IV
Subtotal
V
Subtotal
VI
Subtotal
VII
Subtotal
Travel time
analysis zone
to study link
0.5 -
2.5 -
4.5 -
6.5 -
8.5 -
0.5 -
2.5 -
4.5 -
6.5 -
8.5 -
0.5 -
2.5 -
4.5 -
6.5 -
8.5 -
0.5 -
2.5 -
4.5 -
6.5 -
8.5 -
0.5 -
2.5 -
4.5 -
6.5 -
8.5 -
2.5
4.5
6.5
8.5
10.5
2.5
4.5
6.5
8.5
10.5
2.5
4.5
6.5
8.5
10.5
2.5
4.5
6.5
8.5
10.5
2.5
4.5
6.5
8.5
10.5
Number
of trips
0
a
0
91
34
125
0
964
1130
248
31
2373
0
0
0
172
0
172
1302
65
952
464
263
3046
231
508
1019
733
416
2907
Fraction of
subtotal
0.000
0.000
0.000
0.728
0.272
1.000
0.000
0.406
0.476
0.105
0.013
1.000
0.000
0.000
0.000
1.000
0.000
1.000
0.427
0.021
0.313
0.152
0.086
1.000
0.079
0.175
0.351
0.252
0.143
1.000
92
-------
TABLE C-9. DISTRIBUTION OF LINK TRIPS ORIGINATING AT ANALYSIS ZONES BY TIME FROM TRIP ORIGIN -
FORBES AVENUE LINK (NODE 03901 - 04106)
U)
Column A
Distribution
of total daily
Origin link trips
?* . analysis during time
period dl8trlct periods orig-
it
I II
IV
V
VI
VII
II II
IV
V
VI
VII
III II
IV
V
VI
VII
IV II
IV
V
VI
VII
*
From Table C-7.
"""From Table C-8.
^Column C - (Column
'"analysis'11 °'5 to 2>
district*
0.001
0.015 -
0.001
0.036 0.427
0.026 0.079
0.003 -
0.041
0.002 -
0.051 0.427
0.050 0.079
0.004
0.069
0.006
0.073 0.427
0.078 0.079
0.002 -
0.052 -
0.004;
0.068 0.427
0.064 0.079
A) x (Cplumn B).
Column B
Time from trip origin distribution
for link trips originating at
analysis district*
> 2.5 to 4.5 r
-
0.406
-
0.021
0.175
-
0.406
-
0.021
0.175
0.406
-
0.021
0.175
-
0.406 :
-
0.021
0.175
4.5 to 6.5
-
0.476
-
0.312
0.351
-
0.476
0.312
0.351
0.476
-
0.312
0.351
-
0.476
-
0.312
0.351
6.5 to 8.5
0.728
0.105
1.000
0.152
0.252
0.728
0.105
1.000
0.152
0.252
0.728
0.105
1.000
0.152
0.252
0.728
0.105
1.000
0.152
0.252
8.5 to 10.5
0.272
0.013
-
0.087
0.143
0.272
0.013
-
0.087
0.143
0.272
0.013
-
0.087
0.143
0.272
0.013
-
0.087
0.143
Column C
Distribution of link trips
originating at analysis
zone by time from trip
origin*.
0.5 to 2.5 2.5 to 4.5
0.006
- ' -
0.016 0.001
0.002 0.005
- -
- 0.016
-
0.022 0.001
0.004 0.009
' - -
0.028
-
. 0.031 0.001
0.006 0.014
- -'
0.021
-
0.029 0.001
0.005 0.011
4.5 to 6.5
-
0.007
-
0.011
0.009
-
0.020
-
0.016
0.017
-
0.033
-
0.023
0.027
-
0.025
-
0.021
0.022
6.5 to 8.5
0:001
0.002
0.001
0.005
0.006
0.002
0.004
0.002
0.008
0.013
0.003
0.007
.0.006
0.011
0.020
0.002
0.005
0.004
0.010
0.016
8.5 to 10.5
-
-
-
0.003
0.004
0.001
0.001
-
0.004
.0.007
0.001
0.001
-
0.007
0.011
-
0.001
-
0.007
0.010
-------
TABLE C-10. PARKING DURATION DISTRIBUTION FOR PITTSBURGH
Analysis
district
I
II
III
IV
V
VI
VII
VIII
IX
X
XI
XII
Time
period
I
II
III
IV
I
II
III
IV
I
II
III
IV
I
II
III
IV
I
II
III
IV
I
II
ITI
IV
I
II
III
IV
I
II .
Ill
IV
I
II
III
IV
i
ii
in
IV
i
ii
in
IV
I
II
III
IV
Parking duration distribution, hours
<1
0.423
0.306
0.109
0.177
0.417
0.530
0.259
0.155
0.314
0.332
0.095
0.179
0.314
0.310
0.220
0.163
0.189
0.500
0.087
0.166
0.140
0.158
0.411
0.187
0.300
0.489
0.308
0.163
0.178
0.485
0.288-
0.176
0.233
0.563
0.292
0.149
0.256
0.449
0.225
0.167
0.103
0.596
0.389
0.165
0.181
0.472
0.365
0.172
1-2
0.094
0.226
0.057
0.114
0.056
0.174
0.090
0.159
0.059
0.100
0.024
0.067
0.070
0.216
0.098
0.115
0.047
0.179
0.017
0.017
0.025
0.171
0.193
0.120
0.017
0.192
0.154
0.122
0.032
0.130
0.173-
0.073
0.025
0.151
0.099
0.092
0.073
0.229
0.103
0.133
0.019
0.080
0.175
0.108
0.034
0.152
0.165
0.142
2-3
0.031
0.119
0.057
0.133
0.014
0.068
0.085
0.180
0.029
0.062
0.038
0.028
0.012
0.121
0.121
0.217
0.000
0.000
0.000
0.017
0.006
0.054
0.093
0.120
0.008
0.061
0.070
0.112
0.000
0.075
0.091-
0.089
0.025
0.045
0.075
0.089
0.012
0.073
0.071
0.076
0.000
0.056
0.136
0.102
0.009
0.074
0.077
0.103
3-4
0.016
0.157
0.040
0.089
0.014
0.050
0.057
0.106
0.029
0.043
0.030
0.022
0.012
0.112
0.070
0.156
0.000
0.000
0.000
0.034
0.006
0.041
0.037
0.067
0.008
0.048
0.056
0.085
0.000
0.055
0.045
0.093
0.017
0.035
0.042
0.069
0.012
0.045
0.042
0.095
0.000
0.019
0.046
0.082
0.000
0.030
0.048
0.063
4-5
0.016
0.064
0.028
0.063
0.014
0.032
0.022
0.078
0.000
0.038
0.032
0.050
0.012
0.047
0.034
0.071
0.000
0.057
0.017
0.105
0.000
0.032
0.034
0.047
0.000
0.044
0.039
0.061
0.000
0.030
0.039-
0.054
0.017
0.040
0.024
0.069
0.012
0.045
0.015
0.049
0.000
.0.019
0.025
0.043
0.009
0.043
0.031
0.043
5-6
0.016
0.034
0.029
0.082
0.014
0.046
0.015
0.020
0.000
0.033
0.014
0.022
0.012
0.043
0.016
0.037
0.000
0.000
0.000
0.051
0.000
0.005
0.016
0.053
0.008
0.026
0.014
0.048
0.013
0.020
0.021
0.038
0.008
0.015
0.009
0.052
0.012
0.020
0.017
0.042
0.019
0.000
0.007
0.085
0.000
0.026
0.014
0.043
6-7
0.016
0.034
0.031
0.038
0.028
0.014
0.047
0.065
0.000
0.038
0.010
0.017
0.012
0.009
0.026
0.041
0.000
0.000
0.000
0.027
0.070
0.014
0.019
0.037
0.008
0.009
0.017
0.054
0.013
0.005
-0.027-
0.038
0.017
0.005
0.021
0.046
6.012
0.016
0.022
0.042
0.019
0.009
0.007
0.043
0.017
0.004
0.017
0.043
7-8
0.016
0.013
0.063
0.020
0.014
0.004
0.057
0.045
0.067
0.028
0.016
0.056
0.047
0.009
0.070
0.031
0.000
0.000
0.035
0.034
0.025
0.005
0.031 '
0.053
0.017
0.009
0.028
0.037
0.032
0.015
0.027
0.029
0.008
0.000
0.030
0.052
0.037
0.029
0.046
0.030
0.045
0.005
0.039
0.026
0.017
0.009
0.020
0.053
8-9
0.250
0.030
0.322
0.095
0.014
0.021
0.209
0.024
0.171
0.242
0.396
0.246
0.058
0.039
0.238
0.041
0.292
0.264
0.635
0.483
0.057
0.023
0.078
0.040
0.083
0.017
0.165
0.061
0.089
0.010
0.130
0.058
0.117
0.050
0.253
0.080
0.049
0.020
0.284
0.106
0.090
0.005
0.107
0.034
0.078
0.048
0.160
0.070
>9
0.125
0.017
0.263
0.190
0.417
0.060
0.159
0.167
0.295
0.085
0.347
0.313
0.453
0.095
0.109
0.129
0.472
0.000
0.207
0.068
0.669
0.140
0.087
0.277
0.550
0.105
0.148
0.255
0.643
0.175
0.158
0.351
0.533
0.095
0.154
0.304
0.524
0.073
0.176
0.261
0.703
0.211
0.079
0.313
0.655
0.143
0.103
0.268
94
-------
TABLE C-ll. DISTRIBUTION OF TOTAL LINK TRIPS BY COLD SOAK DURATION
FOR TRIPS WITHIN 10.5 MINUTES OF STUDY LINK
Tt-e
I
II
III
IV
Tloe
from Analyile
origin
0.5 to 2.5
2.5 to 4.5
4.5 to 6.5
6.5 to 8.5
8.5 to 10.5
0.5 to 2.5
2.5 to 4.5
4.5 to 6.5
6.5 to 8.5
8.5 to 10.5
0.5 to 2.5
2.5 to 4.5
4.S to 6.5
6.5 to 8.5
8.5 to 10.5
0.5 to 2.5
2.5 to 4.5
4.5 to 6.5
6.5 to 8.5
8.5 to 10.5
VI
VII
IV
VI
VII
IV
VI
VII
II
IV
V '
VI
VII
VI
VII
VI
VII
IV
VI
VII
IV
VI
VII
11
IV
V
VI
VII
II
. IV
VI
VII
VI
VII
IV
VI
VII
IV
VI
VII
II
IV
V
VI
VII
II
IV
.vr
VII
VI
VII
IV
VI
VII
IV
VI
VII
II
IV
V
VI
VII
IV
VI
VII
Distribution
of total
link trips
0.016
0.002
0.006
0.001
0.005
0.007
0.011
0.009
0.001
0.002
0.001
0.005
0.006
0.003
0.004
0.022
0 004
0.016
0.001
0.009
0.020
0.016
0.017
0.002
0.004
0.002
0.008
0.013
0.006
0.001
0.004
0.007
0.031
0.006
0.028'
0.001
0.014
0.033
0.023
0.027
0.003
0.007 .
0.006
0.011
0.020
0.001
0.001
0.007
0.011
0.029
0.005
0.021
0.001
0.011
0.025
0.021
0.022
0.002
0.005
0.004
0.010
0.016
0.001
0.007
0.010
- Distribution of total link trips by cold
sosk duration, hours
<1
0.002
0.001
0.002
_
0.002
0.002-
0.002
0.003
_
0.001
0.001
0.002
_
0.001
0.003
0.002
0.005
0.004
0.004
0.003
0.008
0.001
0.001
0.001
0.001
0.006
0.003
_
0.001
0.003
0.013
0.002
0.006
0.004
0.007
o.oos
0.008
0.001
0.002
0.001
o.oos
0.006
_
-0.003
0.003
0.005
0.001
0.003
0.002
0.004
0.004
0.004
_
0.001
0.001
0.002
0.003
_
0.001
0.002
1-2
_
-
_
_
-
_
-
_
-
_
-
0.004
0.001
0.003
0.002
0.006
0.003
0.003
-
0.001
o.ooi
0.002
0.001
_
0.001
0.001
0.006
0.001
0.003
0.002
0.003
' 0.004
0.004
_
0.001
0.002
0.003
_
0.001
0.002
0.003
0.001
0.002
0.001
0.003
0.003
0.003
_
0.001
0.001
0.002
_
0.001
0.001
2-3
_
-
_
_
-
_
_
-
_
_
-
_
. -
0.001
-
0.002
0.001'
0.002
0.001
0.001
-
0.001
-
_
-
0.003
0.003
0.001
0.004
0.002
0.002
_
0.001
0.001
0.001
_
0.001
0.001
0.003
0.001
0.005
0.001
0.005
0.003
0.002
_
0.001
0.001
0.002
_
0.001
0.001
3-4
_
-
_
_
-
_
_
-
_
_
-
_
-
0.001
-
0.002
-
0.002
0.001
0.001
-
0.001
-
-
0.001
0.002
0.001
0.002
0.001
0.002
-
0.001
_
..
0.001
0.002
0.003
0.001
0.004
0.001
0.002
_
0.001
0.001
0.001
_
0.001
4-5
_
-
-
_
-
_
-
-
-
-
o.ooi
-
0.001
-
0.001
0.001
0.001
-
o.ooi
-
_
-
0.001
-
0.001
0.001
0.001
0.001
0.001
-
o.ooi
-
. .
-
0.001
-
0.001
0.001
0.002
o.ooi
o.ooi
_
0.001
_
o.ooi
5-6
_
-
~
-
_
-
-
-
-
_
-
0.001
-
0.001
-
-
-
-
-
_
-
_
-
0.001
-
-
-
-
-
-
0.002
0.001
0.001
0.001
0.001
0.001
_
0.001
0.001
-
~
6-7
0.001
-
-
-
0.001
-
-
-
-
-
-
_
-
-
-
-
-
-
-
-
-
_
-
0.001
. -
0.001
-
-
-
-
. -
-
0.001
0.001
0.001
o.ooi
0.001
0.001
_
0.001
_
0.001
7-8
_
-
-
_
-
_
-
-
.*
-
-
-~
_
-
-
-
-
-
-
-
-
-
0.001
0.002
-
0.002
0.001
0.001
-
0.001
-
-
-
0.002
o.ooi
-
0.001
0.001
0.001
_
0.001
0.001
_
~
8-9
0.001
-
-
-
_
0.001
0.001
-
-~
-
_
0.001
-
0.001
-
0.001
-
-
0.001
-
-
-
0.002
0.001
0.007
0.002
o.oos
0.002
0.004
0.001
0.002
0.004
0.001
0.003
-
0.001
0.002
0.001
-
0.001
0.001
0.001
0.001
0.001
_
0.002
_
o.ooi
_
o.ooi
>9
0.011
0.001
0.003
0.001
0.003
0.003
0.007
0.005
-
0.001
0.001
0.003
0.003
0.002
0.002
0.003
-
0.002
0.001
0.002
0.002
0.002
-
0.001
0.001
-
o.ooi
0.001
0.003
0.001
0.003
0.002
0.004
0.002
0.004
-
0.001
0.001
0.001
0.003
-
- 0.001
0.002
0.008
0.001
0.003
0.003
0.003
0.006
0.006
_
0.001
0.003
0.004
_
0.002
0.003
Bacord.nl froo T.bla C-9.
95
-------
TABLE C-12. COLD MODE CYCLE LENGTH (IN MINUTES) AS A FUNCTION OF COLD
SOAK DURATION^(IN HOURS) FOR AMBIENT TEMPERATURES OF
75°F AND 20°F
... Cold soak period (mid-point), hours
temperature ~~~
0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5
75°F 2.4 3.6 4.3 4.9 5.3 5.8 6.1 6.4 6.7 7.0
20°F 3.3 4.4 5.0 5.4 5.8 6.1 6.4 6.8 7.0 7.2
TABLE C-13. SUMMARY OF THE COLD MODE ANALYSIS FOR THE FORBES AVENUE STUDY
LINK (NODES 03901 TO 04106) .
Time period
II III IV Total
A. Cold mode operation as
a fraction of daily
link trips. 0.004 0.050 0.096 0.088 0.278
B. Stabilized mode opera-
tion as a fraction of
daily link trips. 0.181 0.153 0.290 0.098 0.722
C. Total - distribution
of daily vehicle
operation. 0.225 0.203 0.386 0.186 1.000
D. Number of vehicles in
cold mode operation
per time period. 590 670 1286 1180 3,726
E. Total number of link
trips per time
period. 3018 2724 5209 2509 13,460
F. Percent of vehicles
operating in cold
mode by time period. 19.5 29.6 24.7 47.0 27.7
96
-------
GLOSSARY
Analysis District: A geographic area defined by a group of two or more anal-
ysis zones (see below).
Analysis Zone: A subdivision of a study or survey area that comprises the
smallest geographic area for which data are aggregated and basic analyses
are made.
Arterial Street: A major street or highway that provides regional service for
through traffic; arterials generally have at-grade intersections and pro-
vide access to abutting land development.
Centroid: A point within an analysis zone or district that is assumed to repre-
sent the center of trip generation and attraction activity; all travel
to and from the zone or district is assumed to occur through the centroid.
Centroid Connector: A synthetic roadway link connecting a zone centroid to
the actual highway network. Generally, up to four connectors may be used
to connect a zone centroid to the highway network.
Cold Mode Operation: The initial stage of vehicle operation when engine and
coolant temperatures have not stabilized and usually the choke is in use.
Cold mode operation is characterized by very high emission rates of car-
bon monoxide and hydrocarbons relative to stabilized operation.
Collector Street: A street that collects or distributes traffic between higher
type arterial streets and either less important local streets or desti-
nations. The functions of traffic movement and access to abbuting pro-
perties are equally important for collector streets.
Cordon Line: An imaginary line enclosing a study area along which external
interviews are conducted.
Federal Test Procedure: A detailed procedure prescribed by the U.S. Environ-
mental Protection Agency for certifying the performance of motor vehicles
with regard to compliance with federal emission regulations. (Also, see
Appendix A of this report.)
Forecasting: The process of estimating future values of land-use, socioecono-
mic, and trip-making variables within an area based on observed (measured)
values for a base year case.
Link: A section of highway designated by nodes at each end-point.
97
-------
GLOSSARY (continued)
Modal Split: The distribution of total person trips by mode of travel.
Mode of Travel: Means of travel such as auto driver, auto passenger, transit
passenger, walk, etc.
Network: A system of links and nodes describing a highway system.
Node: A designated point on a street or highway that delineates the end-point
of a street or highway link.
Skim Trees: Computer-generated listing of the travel time between zone pairs
reflecting the assigned travel path between each zone pair.
Terminal and Transfer Facilities: Refers to trucking or railroad terminals.
These facilities are included in studies of goods movement within a
region.
Traffic Assignment: The process of determining probable travel paths over
the highway network between analysis zones. The assignment process is
based on the assumption that trips between zone pairs will occur over the
path (combination of links) that minimizes travel time.
Trip Distribution: A general term describing the ultimate distribution of
trips between each zone and all other zones, both internal and external
with respect to the study area. Analytical techniques employed in de-
termining trip distribution generally include defining trip production
and attraction characteristics for each zone based on land-use patterns
so that relationships between the relative attracting power of all zones
for trips generated in each zone.
Trip Generation: A general term describing the analysis of numerous charac-
teristics of an area (e.g., an analysis zone) to define the relationships
that exist between the area's trip-making characteristics, and parameters
such as land-use, socioeconomic structure, demographics, etc. within the
area.
Turn Penalties: A feature of the traffic assignment process that accounts for
delays typically experienced in making turning movements at intersections
by adding an increment of time (i.e., a penalty) for each turn in an
assigned routing. Generally, left-turn (across traffic) penalties are
more severe than right-turn penalties.
98
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-77-023
I. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
DETERMINATION OF PERCENTAGES OF VEHICLES OPERATING
IN THE COLD START MODE
5. REPORT DATE
August 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Theodore P. Midurski and Alan H. Castaline
8. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-77-19-G
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA Corporation
GCA/Technology 'Division
Burlington Road
Bedford, Massachusetts 01730
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-1376, TO 29
12. SPONSORING AGENCY NAME AND ADDRESS
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
200/4
15. SUPPLEMENTARY NOTES
16.ABSTRACT Estimates 'of the percentages of vehicles operating in less than the
stabilized condition (i.e., cold mode) have been made for 60 locations in the Pitts-
burgh and Providence regions. These estimates were derived using detailed regional
travel data for both regions.
:^!curate knowledge of the"p"ercentages of vehicles operating In the cold m6de"at*a
particular study location is a key factor in estimating carbon monoxide emissions.
That carbon monoxide emission rates are severely affected by operation in the pre-
stabilized condition has been well documented through numerous laboratory studies.
While these laboratory studies have served to characterize carbon monoxide emissions
from cold operating vehicles with respect to variables such as ambient temperatures an<
'cold" soak-duration; the- study being-reported here- examines specific-traffic streams to
identify the actual percentages of vehicles that are operating in the cold mode.
Cold start percentages are presented for the morning commuter hours, the mid-day
period, the evening commuter hours, and the evening and early morning off-peak period
for each of the 60 traffic streams analyzed. Results indicate that the percentage of
cold operating vehicles is highly variable both among locations within an urban area,
and during different time periods at any specific location. It is also indicated that
the actual percentages of vehicles operating in the cold mode may be somewhat different
from the percentages assumed in the Federal Test Procedure.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/GlOUp
Air Pollution
Exhaust Gases, Automobile
Carbon Monoxide
Traffic Characteristics
Emission Rate Variations
Automobile Exhaust
Light Duty Vehicles
13B
13. DISTRIBUTION STATEMENT
Distribution Unlimited
19. SECURITY CLASS {This Report)'
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (This page)
UNCLASSIFIED
1IL
22. PRICE
EPA Form 2220-1 (9-73)
99
------- |