-------
10
Average Route Speed, KPH
30 50
70.
I
1.5
o
+j
o
c
o
•r—
•«J
O
OJ
5
Q.
CO
1.0
0.5
I
f
I
15 30
Average Route Speed, MPH
45
FIGURE 2. Speed Correction Factors for
1968 Model Year Vehicles in Denver
-------
10
Average Route Speed, KPH
30 50
70
1.5
o
«O
c
o
O
O
O
•o
o>
(U
CL
to
1.0
CO
0.5
1
I
15 30
Average Route Speed, MPH
45
FIGURE 3. Speed Correction Factors for
1971 Model Year Vehicles in Denver
- 6 -
-------
- 7 -
A further finding is that complex highway design configurations,
unique localized meterology, presence of topography and rough terrain
and downtown or high-rise street canyons play readily identifiable,
but relatively less understood roles in air quality. In the simplest
sense, average speed and vehicle miles of travel on a link are relevant
inputs as indicators of CO, EC and NO emissions. The realistic design,
Jt
environmental and neighborhood attributes modify the impact of these
two basic stimuli on air quality, and the mechanics of these modifica-
tions are not developed in depth in the basic research to date. Figures
4, 5 and 6 are offered as visual examples of research output of the
effect of geometric configuration on air quality. A further behavioral
input, closely related to average speed, is the smoothness of traffic
flow and capability of avoiding traffic congestion effects. Figure 7
shows the concentrations for smooth, uninterrupted flow of 50 nrph of
typical freeway movement versus higher emissions induced by poor signal
timing, pedestrian and parking interference which increases delay on a
typical interrupted flow arterial street operating poorly. As such,
from a traffic engineering point of view, V/C ratios and acceleration
noise parameters of the traffic stream are relevant to emissions levels.
In conclusion, the literature search revealed:
1.) The basic stimuli of emissions to be well documented
against inputs of average speed and VKT for CO and HC,
with more questionable data and relationships with respect to NO .
2.) Further complexities in air quality overlaid on the above when
cut, fill, and complex cross section and geometric design
configurations are included, as illustrated in Figures 4»
5 and 6,
-------
-5
I I
c t
0.0$
..0.01.
0.03
SO
30
20
60
do
JO
0.02 . 10 -
Level 3 70
Level 2 25
Levef I 0
/ W
25 Feet
I
FIGURE 4. Pollution Levels Along Transverse Street Cross Section of
Centered Expressway with Joint Development Structures
- 8 -
-------
0.05
0.04
0.03
0.02- 10-
Level 3 601 1
Level 2 25^ \-
Level 1 oL
I
FIGURE 5. Pollution Levels Along Transverse Street Cross Section of
Centered Expressway Without Joint Development Structures
- 9 -
-------
K F.«t
FIGURE 6. Pollution Levels Along Transverse Street Cross Section of
Centered Expressway-Boulevard
- 10 -
-------
8
17H n.p.h.,
7H m.p.
200 400 600 800 1000 1200 1400
Vehicle Flow Per Hour
FIGURE 7. Comparison of City Street and Freeway Conditions
- 11 -
-------
- 12 -
3.) Further complexity in emissions when system delay from
congestion is introduced, as illustrated in Figure 7.
4.) Further complexity in air quality when overlaid by local,
uniquely complex topography and terrain, and localized
unique meterologic conditions.
5.) The apparent need for further basic research on line
sources where such complexities in 2-4 above are introduced.
That is, further research into locating them, stratifying
their attributes, classifying them in an orderly data
system, and relating their attributes to their resulting
link emissions and the air quality of their locale,
C. FORMAL RESEARCH WORK PLAN
1.) Emissions and Air Quality-Related Traffic Flow Research Aspects
In light of the previous literature search, it has been categorically
found that volume, or vehicle miles of travel, and average speed are
critical inputs which relate to emissions of nitrogen oxides, carbon
monoxides and hydrocarbons. However, exhaustive use of typical traffic
flow data related to the above two inputs should allow a refined and
meaningful statement of flow phenomena and traffic systems design which
pollution output is sensitive to. Thus, a short discussion of such
flow-related parameters is warranted, to demonstrate their categorization
as inputs:
Volumes - Average Daily Traffic and peak hour volumes, historically
as indicators of use of the facility.
-------
— 13 —
Average Travel Speed - As an indicator of efficiency of the
facility and adequacy of design.
General Interzonal Origin-Destination Patterns - As a regional
mapping of incidence of travel, and proximity of
travel paths to other regional activities.
Functional Classification of Highways - Classed as whether freeway,
arterial, collector or local, as an indication of importance
and frequency of use and level of design standards employed.
Delay Information - which modifies or refines information on
average speed, above, through studies of volume to
capacity ratios, travel time profiles, travel time
contours, or waiting time or delay contours. Locations
are detected in the system where speeds are radically
altered due to delay and congestion.
Locations of^Design Related Phenomena - such as extremely
complicated route or interchange configurations, and
areas of cut or fill, or frontage roads with struc-
tures, v/hich induce localized alterations in air quality
and emissions, the latter when average speed is modified
due to the design phenomena.
Unique Areas of Progression - in addition to areas which can be
reviewed from speed and delay information as stated
above, these are unique in the network, in that some
engineering or planning alteration exists to eliminate
congestion by specific means with highly predictable
results, such as one-way street flows, or progressive
signaliaation, thus allowing atypical consistency in
volumes or average speeds, with stable emissions output.
-------
- 14 -
Areas of Critical Land TJse Adjacent to or Within the Network -
The first type is an area adjacent to the traffic system
which is a highly sensitive land use to emissions output
and local air quality, or a land use type such as industrial,
which supplements and confounds the emissions level and
air quality in the vicinity of the corridor. The second
type is the downtown or core area, or like areas of
high-rise, high-density buildings. The building heights
or "street canyons" affect air quality in the vicinity of
the grid and corridor sources.
Vehicle Mix - the composition of traffic, in terms of percentage
of autos, intermediate size trucks and large size trucks
is relevant, due to differing emissions from vehicle type,
and the impact of the traffic composition on average
speed and traffic flow throughout a link. The composition
of vehicles by age also determines the level of emissions
from the traffic stream.
Frequency of Monitoring - all links under study will encompass some
or many of the above flow related phenomena which have
an impact on emissions and/or air quality. The frequency
of observing such network components with respect to
adequate characterization of CO, NO , and HC information
Jv
is critical. Typical choices of duration of traffic
volume counting periods include 1, 8, 12, 24 hours,
weekly, and peak-off peak combinations.
-------
Thus, the above considerations are relevant to comprehensively
using traffic flow information to develop the "optimal" network inputs
for specifying emissions line sources. Use of this information may be
characterized by the following three dimensional array in Figure 8.
The research proceeds by essentially categorizing each apparently
relevant link component with respect to the above array, then testing
and trading off combinations of these with respect to emissions levels
to ultimately produce a set of line sources. The process pro-
ceeds interactively, making full use of local traffic engineering
knowledge about network components, yet developing the taxonomy in a
rigorous and consistent logic.
2.) Specific WorkPlan of Research
To accomplish the stated objectives, and develop the output dis-
cussed above, a six-phase work program has been pursued, as shown in
Figure 9«
Phase 1 has four tasks, which were carried on simultaneously.
Task 1.1 consisted of review of the flow aspects discussed above with
respect to their impact on emissions. Tasks 1.2 and 1.5 included the
procurement of St. Louis Air Quality Control Regional Traffic data, and
review of important links and their traffic operating attributes (Figure 8) with
local professionals. Through the principal investigator's local knowledge and
contacts, use was made of East-West Gateway Regional Coordinating Council profes-
sionals, and engineers with Missouri State Highway Department, Illinois
Department of Transportation, and St. Louis, St. Clair, Madison and
other appropriate county engineering personnel. The results of 1.2 and
1.5 yielded realistic insights into the operation of corridors and
arterial highway grid components. In TasV 1.4, the traffic input and
emissions output aspects of the model SAPOLLUT was reviewed, and the
model was procured and put on the Washington University Software Library.
-------
%
3
o
•p
•H >-.
C 0
0 C
SO
CO D
(H 0)
3£
w
CO
X
-p
C
o
3
rH
0)
0)
^
•^
•=*
CM
CM
I
fl>
0)
O
-p
O
CD
O
O
O
3
o
- 16 -
-------
Review of
Traffic Flow
Parameters
(1.1)
Review
St. Louis Area
Traffic Data
(1.2)
V
Categorization of
Appropriate Corridor,
Link and Grid
Components
(2.1)
t
^L
Develop Combinations of
Network Components, Estimate
Emissions with SAPOLLUT
(3.D
A/
Review
Link Attributes
with Local
Professionals
(1.3)
Procure and
Study SAPOLLTJT
Model
(1.4)
Review Mechanics
of Sensitivity
Analysis
(2.2)
1
1
1
A
1
1
L-? -
x__
\
/
Documentation of
Emissions Estimates vs.
Network Combinations vs.
Monitoring Frequency
(4.0)
Sensitivity Analysis
of Different Network
Combinations
(3.2)
\
\
X
/
Final
Specification
of St. Louis
Air Quality
Control Region
Line Sources
(5.0)
Write
Final
' (6?0)
FIGURE 9. Research Work Plan
- 17 -
-------
- 18 -
Phase 2 included two overlapping tasks. The initial task, 2.1
categorized appropriate corridors, links and street grid components
with respect to the flow phenomena illustrated in the previous three-
dimensional array. The coding was developed so that a complete set of
descriptors exists for each corridor, link or arterial street grid level con-
ceptualized and developed as input. Overlapping with this, task 2.2 familiarized
the study team with the sensitivity analysis, and the mechanics of
its performance in order to assure continuity between Phase 3 and
later phases.
Phase 3 had two tasks. Task 3.1 developed levels and combinations
of network hierarchial components as inputs to studying emissions.
The following sequence of components were developed for use with
SAPOLLUT and sensitivity testing:
Round 1; All freeway corridors, link components broken up by
average speed differentials and volume differentials.
Round 2; All freeway corridors, additional breakdown by sites
of specifically complex configuration.
Round 3: Addition of the arterial street grid to above, broken
down by speed, delay and volume differentials.
Round 4: All the above, additional breakdown by arterial sites
of specifically complex configuration.
Round 5t Addition of refined locations adjacent to sensitive
land use areas, and areas of exceptional progression.
Round 6: All of the above, cross classified by vehicle mix.
-------
- 19 -
Thus, the approach is to ultimately yield a descriptive network of
line sources N, which is composed of
5~ 9
N ~ <~~~ A, ijkmnopq t
i,3,k,m,n,o,p,q, t
where JL is a specific link or network component and i, j, k, m, n, o, p, q .....
are its flow parameters shown in the three-dimensional array, and t is
the monitoring frequency. At each of the above rounds, the network
could be input to SAPOLLUT, and emissions information estimated. The
network is synthesized in a sequential manner through the several rounds,
to develop clarity about its composition and emissions output, allowing
for a more orderly sensitivity analysis.
Task J>.2 articulated changes in the above synthesized network
and site monitoring frequency. Prom knowledge of the local area,
specific link and network components developed through all of the above
rounds were removed of added, noting through the use of SAPOLLUT in a
sensitivity analysis, the change in resulting emissions information.
The sensitivity analysis of the coded link and network components was
developed in a batch software format, to allow a certain programmed
sequence of network alterations to occur. This process was hooked
with SAPOLLUT, to allow network input operations and emissions estima-
tion to occur in one continuous software run.
Phase 4 documented the results of the use of SAPOLLUT and the
sensitivity analysis carried on in J>.2. Reference to Figure 9 shows
feedback and interaction across 3.1, 3.2 and 4 to adequately formulate
test, converge and document the above processes.
-------
- 20 -
Phase 5 utilizes the above array of sensitivity and convergence
information and specifically states network component locations, their
frequency of emissions monitoring, and the required traffic flow-related
data. This specification defines the emission line sources for the
St. Louis Air Quality Control Region.
Phase 6 combines previous conceptual analysis, the computer software,
associated network synthesis, use of SAPOLLUT and the sensitivity analysis
and its results, with specifications for refinement and further research
into a final written report contained herein. The following text will
elaborate on data collection, software development, use of SAPOLLUT,
sensitivity analysis and definition line sources, and provide appropriate
appendices on software documentation and data formats.
-------
Footnotes Chapter I
Federal Register, Volume 36, No. 84, Friday, April 30, 1971.
2
Scott Research Laboratories, Inc., Development of Representative Driving
Patterns at Various Averse Poute Speeds, EPA rlo. 68-02-1301 (6-1 J>).
February 11, 1974, San Sernadino, California.
., pp. 3-1 through 4-1S.
^Sturman, G. M. , The Effect of Highways on the Environment, May, 1970.
Steering Group and Working Group Appointed by the Minister of Transport,
Great Britain, Cars for Cities, 1967.
- 21 -
-------
CHAPTER II
DATA DEVELOPMENT
-------
CHAPTER II
DATA DEVELOPMENT
An assessment of highway line sources with respect to emissions requires that
the road network under study be classified according to a set of para-
meters that allows appropriate hierarchial analysis, since some links
will be much more critical from an emissions standpoint than others.
The data collection effort focused on those areas designated
as urban and urbanizing. The East-West Gateway Coordinating Council
defines these areas to include the City of St. Louis, St. Louis County,
and parts of St. Charles and Jefferson Counties in Missouri. In
Illinois part of Monroe, and Madison and St. Glair Counties are included.
Roadway data was collected within this area in light of emissions and
their sensitivity to highway functional class, volume and composition of
vehicles present, and the operational characteristics of vehicles related
to both traffic volume and average speed alterations due to roadway align-
ment and profile. Other data collected, such as the intensity and type of
the adjacent land use is relevant, along with certain highway design
characteristics, to the localized air quality.
A number of agencies were contacted to compile the inventory.
They included the East-Vest Gateway Coordinating Council, Missouri
State Highway Department Jefferson City Office and St. Louis Office,
the Region VII Office of the Illinois Department of Transportation,
the Office of the Deputy Commissioner of the City of St. Louis Street
Department, St. Louis County Division of Highways and Traffic, and
the City of St. Charles.
A working set of highway links was established through the use
of East-West Gateway's 7001 link-node map for the entire region under
- 22 -
-------
study. Actual street names for all links were recorded from maps
provided by the Auto Club of Missouri.
The initial major data collection effort was for the freeway
functional class. Volume and vehicle attributes sought for each link
were average daily traffic, peak hour traffic, hourly distribution of
traffic, percent of heavy duty vehicles, and the directional distribu-
tion of traffic. All the above were obtained for the Missouri counties
except percent heavy duty vehicles. However, in Illinois, only average
daily traffic and percent heavy duty vehicle data was available.
Link attributes sought to describe vehicle operating characteristics
were the volume to capacity ratio (V/C), peak hour speed by direction,
off peak speed by direction, frequency of complex interchanges, lane
drops and existing freeway bottleneck sections. The capacity informa-
tion obtained from East-West Gateway was a representative daily capacity
factored down to obtain a peak hour V/C ratio at level of service E. An
hourly V/C ratio table was given in the SAPOLLUT users manual based on
an hourly speed distribution. Current peak hour speed data by direc-
tion was available for 75?° of the area freeway links from East-West
Gateway. Through the cooperation of the Illinois Department of
Transportation and local agencies in St. Louis and St. Louis County,
a complete set of links with complex interchanges, lane drops, and
major freeway bottlenecks was compiled. The study team drove the
network to collect information on roadway topography as a freeway
link descriptor. Four terrain types were categorized: high fill,
deep cut, rolling terrain, and level terrain.
-------
- 24-
As a final set of link descriptors for the freeway functional
class, land use by type and intensity was sought. East-West Gateway
provided an area map delineating high density regionally oriented
land uses. Types included commercial, educational, medical, recrea-
tional, and airport.
The next major data collection effort was for principal arterial
roadways. These roadways are characterized "by a minimum number of
intersections at grade. Such intersections are typically designated
to provide channelization and signal synchronization to enhance the
traffic movement along the arterial. The link attributes sought to
describe volume and vehicle composition were as before, average daily
traffic, peak hour traffic, hourly distribution of traffic, percent
of heavy duty vehicles, and the directional distribution of travel.
The same data gaps that existed for freeways exist for principal
arterials in this attribute set. In addition to volume to capacity
ratios, peak hour directional speed, off peak directional speed, lane
drops, and general roadway bottlenecks were collected as link attributes
for vehicle operating characteristics. Farther, information concerning
the degree of progressive movement was sought. Progressive movement
is typified by a continuous flow of a platoon of vehicles over long
stretches of highway. Such movement can be induced by the- type of
signal systems employed at intersections and the distribution of
one-way streets. It is desirable to separate out links that have
progressive movement since the degree of vehicle delay is much less
than on arterials without it. Attributes for progression included
links with pre-timed signal systems, physically interconnected signal
-------
- 25 -
systems, and one-way street flows. The City of St. Louis, St. Louis
County, Missouri State Highway Department, and the Illinois Department
of Transportation provided us with complete information on this attri-
bute set.
As in the freeway case, sensitive land uses were included as link
descriptors. In addition to those already listed, the central business
district was included as an area type in the inventory. This was done
to aid in the delineation of an additional topography attribute, the
categorization of street canyons.
The final data collection report was for minor arterial roadways.
Such roadways provide for both traffic movement and land access. The
data sought was the same for principal arterials and had the same data
gaps.
In total, 28 link descriptors across three functional classes of
highway were assembled. This data was compiled on an individual link
basis, coded for keypunching and readied for input into the software
system to be described in Chapters III and IV. The collection and
coding format is illustrated in Appendix B.
-------
Footnotes Chapter II
Level of Service E represents operation of the system with volume at or
near capacity. Operating speed is relatively low, flow is unstable
and momentary stoppage occurs; the system is on the verge of complete
jam and saturation with attendant congestion effects.
- 26 -
-------
CHAPTER III
MODEL DEVELOPMENT
-------
CHAPTER III
MODEL DEVELOPMENT
A. OVERVIEW OF MODEL LOGIC
This chapter describes the logic construct of the network sensitivity
model NETSEN, and its interface with the emissions estimation model
SAPOLLTJT. The flow chart of the master logic for the model NETSEN is
shown in Figure 10. In general, the model works by defining a series
of sequential tests of presence of network related attributes shown in
steps 2.0-13.0. These are presence of the link within the Central Business
District (i.e., the Downtown Core Commercial Area), functional class of the
link, its ADT, presence of special topography, capacity alterations, pre-
sence of sensitive land uses, presence of progressive movement, speed
difference, truck volumes, and V/C ratio. After reading the coded link data
records in, with step 2.0, each link is tested to sort and classify it ac-
cording to combinations of attributes present in it. Those groups of links
thus containing certain combinations specified in the control card in step
1.0 are then output to SAPOLLUT for use in estimating emissions. A complete
description of the software documentation is given in Appendix A. Two
specific points are important in overviewing the logic at this point:
1.) The network and any link subset component of it can be
tested at any level of data attributes relevant, from
very gross descriptions containing only ADT information,
to very refined descriptions of the network, classifying
and locating all of the attributes shown in steps 2.0-13.0
on the network.
- 2? -
-------
- 28 -
,' Read in Control Card (l.O) /
^7 Read in Link REG (2.0)
Modify by CBD?
(3.1)
Modify by
Functional Class
(4.1)
Test for
Functional Class
(4.2)
Yes
'<-
la
-¥-
V'
(Continued)
FIGURE 10. Master Logic of NETSEN Model
-------
Modify by ADT?
(5.1)
Test for ADT
(5.2)
Modify by
Special Topography
(6.1)
Test for
Topography
Modify by
Capacity Alterations
(7.1)
Test for
apacity Alteration
(Continued)
-------
Test for
Sensitive Land Use
(8.2)
No
Modify by
Progressive Movement
„ (9.1)
Test for
Progressive Movement
(9.2)
Modify by
Speed Difference
(10.1)
(Continued)
-------
3a
. No
No
No
Difference
Modify by
Truck Volumes
(11.1)
Test for
Truck Volumes
(11.2)
Modify by
V over C
12.1)
over
(12.2)
(Continued)
-------
4
OUTPUT;
LINKS MEETING
ASSIGNED CHARACTERISTICS
(13.1)
ARE ALL LI1
IS FILE TESTED
GO TO 2
Yes
TO EMISSIONS MODEL
WITH TAGGED
LINK DESCRIPTORS
(14.0)
-------
- 33-
2. The level of attribute refinement chosen to be tested for
may be varied with the refinement of detail of data present
on the network the user has access to, or the level of re-
finement deemed necessary for the user to study emissions.
Thus, complete flexibility exists in describing the traffic
related behavioral aspects of the network as related to
emissions estimation.
B. SOFTWARE SYSTEM DEVELOPMENT
A discussion of the formal software system and interface with
SAPOLLUT should be prefaced with brief mention of the traditional
Urban Transportation Planning (UTP) background of SAPOLLUT. The
implication is that the model was designed to link to the FHWA pro-
gramming battery which performs the UTP process of trip generation,
trip distribution and traffic assignment, with the attendant problem
of using a loaded traffic assignment network versus actual ground
counts. The design of the UTP process makes it necessary to modify
some software usage in order to use realistic ground counts in
SAPOLLUT.
The completed software system correctly employing these modifica-
tions is shown in Figure 11. The system begins with the sequential
file of the network link records of data attributes (l.l). The
complete documentation of the format of this file is in Appendix A.
The system branch containing program modules 1.2, 1.3 and 1.4 is an
initialization routine executed only one time to set up the system's
-------
-34-
LINK RECORD FILE
(1.1)
FORMAT REVISION PROGRAM
(1.2)
NETWORK SENSITIVITY PROGRAM
(2.1)
NODE PAIR OUTPUT
(2.2)
INTERFACE PROGRAM
(3.0)
DATA MANAGEMENT PROGRAM
(1.3)
MODIFIED LINK FILE;
(1.4)
BUILDHR (FHWA)
(4.0)
\t
(5.0)
(FHWA)
HRMOD (FHWA)
(6.0)
SAPOLLUT
(7.1)
EMISSIONS OUTPUT SUMMARY/
(7.2)
FIGURE 11. Traffic Emissions Software System
-------
operation. The format revision program HRCVHT (1.2), takes the link
record file as input and transforms it into output which will eventually
be in a usable format for input into the FHWA program BTJILDHR in step 4.0.
As such, module 1.3» referred to as the Data Management Program,
accepts as input the link records from the Format Revision Program.
The program creates what is called an indexed sequential data set.
Essentially, this implies attaching a key to each link record so that
it may later be retrieved with a single command, obviating a search
procedure. The output of the program, termed a modified link (1.4)
file, is essentially the same as its input except for the above keyed
reorganization to speed access, it is important to emphasize that this
program does not make any functional changes in the link records and
its function could be performed elsewhere in the system. As noted
earlier, this program is only executed once for all runs of NETSEN.
The next program' in the software is the Interface Program (3«0)»
and as its name implies, it is the heart of the interfacing procedure.
It accepts input from two places. First, it accepts a node pair output
(2.2) from the NETSEN Program (2.1) which has passed all logic tests in
that program. It then uses this node pair as a key to retrieve with a
single statement the link record from 1.4 which is identified by the
node pair. The program then processes the link by producing a dummy
link to connect to its A-node if the previous link already processed
does not have a B-node which is the same as the current A-node. The
program then checks to see if the A-node is numerically less than the
B-node, and if so the two are reversed. Further action takes the sum
-------
of the A-B volume count and the B-A volume count and places this sum
minus 1000 in an A-B count location. The program then makes a series
of edit checks to assure that valid links or dummy links do not violate
any of the traffic assignment coding conventions for leg numbers and
to assure that the proper count volumes are passed to SAPOLLUT. The
output is a network compatible to the historical record building pro-
gram, BUILDHR, in 4.0.
The next program in the software system is the FHWA BUILDHR program.
The program accepts link records as input, and functionally, the program
performs edit checking on them for consistency in coding, ultimately
outputting a binary historical record for each link and also one for
each node. The types of edit checking done by BUILDHR include checking
for unusually long links, excessively large volume-to-capacity ratios
as well as duplicate node and leg numbers.
Another FHWA program, PRIKTHR (5«0), follows in the software system.
The program accepts the binary historical records as input from BUILDHR
and prepares a printed summary of information in the records which is
useful in checking for proper operation of preceding programs as well
as for interpreting the output of SAPOLLUT on the basis of a specific
set of links. Although this program is not functionally necessary
for operation of the system, it provides useful information at a small
cost.
The final interface program preceding SAPOLLUT is the FHWA program,
HRMOD (6.0). This program is necessary because of SAPOLLUT's orienta-
tion toward traffic assignment loaded networks. Although the basic
historical record has ground count information included in it, SAPOLLUT
-------
- 37 -
can not use it in its storage location. It is necessary to relocate
actual ground count data from their storage locations in the historical
record to those storage locations where traffic assignment 'oads would
normally be. The program KRMOD is used to shift the ground counts to
the locations where the loads are normally situated. It then outputs
this modified historical record to SAPOLLTJT.
The last program in the software system is the emissions model
SAPOLLUT (?.l). It receives the modified historical record from HBMOD
and several control cards as input. It then proceeds to compute three
types of emissions (HCt CO, 110 ) for three different area types (CBD,
jC
r)
Central City, Suburb) and two different functional classes, freeways and
arterials, using vehicle miles of travel and average speed input, the latter
developed by one of three alternative methods available to the program.
The emissions output is currently available only in the aggregate, broken
down by area type, hour of the day, and functional class across each emission
type. By further dividing gross kilograms of emissions by vehicle-miles
traveled it provides emissions in grams per vehicle-mile and grams per
p'assenger-mile, given an average auto occupancy level.
In its current state the software system is fully automated in
batch mode. Thus, when the network inventory is loaded onto a tape
or disk data set and the initialization programs (Format Revision
Program (1.2) and Data Management Program (1.3)) are run once, the
system can execute several runs of WETSEN, examining the network at
several levels of refinement, with one submittal to the computer. By
providing a series of control cards to NETSSN, separate member data
-------
- 38 -
sets for each run are created, saving the individual sets of node pairs
to "be processed all the way through SAPOLLUT on successive runs through
the system. This enables the user to rapidly analyze the network with
a set of pre-determined runs. It should be emphasized that no manual
interface is necessary during this process. The following chapter will
demonstrate the flexibility and solution properties of the system opera-
tion in documenting line sources for the St. Louis Air Quality Control
Region.
-------
Footnote? Chapter III
Program Documentation, Urban Transportation Planning, March 1972,
Federal Highway Administration. A loaded network is defined
as a transportation planning network in which the traffic volumes
on various links are the result of theoretical computation. A
ground count is defined as the actual recorded count of vehicles
on a roadway link, generally made by mechanical counting devices.
2
Central Business District is the downtown commercial core; the Central
City is the non-commercial downtown core area, and the suburban
area is the outlying area of lesser density.
-------
CHAPTER IV
DOCUMENTATION OF EMISSION LINE SOURCES
-------
CHAPTER IV
DOCUMBKTATION OP MISSION LUTE SOURCES
A. Introduction-Issues in Defining Line Sources
As previously discussed, useful definition of line sources hinges
on the capability to analyze the highway network and its traffic and
design attributes at varying levels of detail, depending on the data
availability and the level of spatial refinement sought in emissions
information from SAPOLLUT. As such, this chapter demonstrates the deve-
lopment of a very unrefined definition of line sources, termed gross
line sources, and the sequential refinement of such to a set of descriptors
termed ultimate line sources. Each of these extremes of definition are
consistent with the basic definition of a line source, given immediately
below in Section B.
B. Basic Definitions-Delineation of Gross Line Sources
The following basic definition of a line source was employed in
documenting emissions for the St. Louis Regional network:
"the smallest sentient of inventoried roadway depictable with a
given specific set of attributes for the roadway."
At the grossest level, the line sources were broken down in the categories
depicted by Table 1. The X's indicate which ADT ranges were used for
functional classes of roadway. In addition, separate baseline runs were
made to select all freeways, all principal arterials, and all minor
arterials within these ADT classes. The emissions produced by this
gross ADT breakdown of line sources are summarized in the following set
of figures and tables for 24-hour periods:
- 40 -
-------
- 41 -
1. Figure 12 represents those freeway links in each of the ADT
ranges from Table 1. Table 2 presents vehicle-miles travelled (VMT),
total emissions, and emission rates for freeways in each of these ADT
ranges.
2. Figure 13 represents those principal arterial links in each
of the ADT ranges from Table 1. Table $ presents WIT, total
emissions, and rates for principal arterials in each of these ADT ranges.
3. Figure 14 represents those minor arterial links in each of
the ADT ranges from Table 1. Table 4 presents VMT, total emissions
and rates for minor arterials in each of these ADT ranges.
A composite analysis of these tables leads to conclusions generally
consistent with the literature, subject to subtle interpretation which
must be employed when aggregating emissions information over all links
under study in a particular functional class. Total kg of emissions
over all categories (CO, NO , and EC) rank lowest for minor arterials,
• J\.
reflecting lowest VMT exposure. Both the freeways and principal
arterials have very similar VMT totals. However, the freeways exhibit
lower totals for CO and HC, and higher NO totals than the principal
-A.
arterials, reflecting consistently lower average speeds on the arterials.
The CO rates (grams/vehicle mile) show the most noticeable change across ADT
ranges for all functional classes, with the most sharp changes associated
with the last three ADT ranges in each class. This is apparently due to
the distinct reduction in average speed associated with links at these
particular ADT ranges nearing or exceeding saturation for their functional
class. HC and NO , however, show generally stable emissions totals across
.X.
all functional classes. The HC rate rises slightly, reflecting the aggre-
gate impact across all links in the network of reduced average speed
-------
ADT
Ranges
(Thousands)
Freeways
Principal Arterials
Minor Arterials
1-30
30-40
40-50
X
50-60
60-70
70-200
5-10
10-15
15-20
20-25
25-30
X
30-35
35-40
40-100
20-40
TABLE 1
ADT RANGES USED FOR GROSS
LEVEL LIKE SOURCE DEFINITION
-------
- 45-
GROSS LEVEL LINE SOURCES
(JTCCTIOIIAL CLASS-FREEWAYS)
SYSTEM 7001
IDT Ranges (Thousands)
1-50 IU4UI
30-40 ^=*
40-50 ca»
50-60 • « * •
60-70
70-200
-------
-44-
1-30
ADT Ranges (Thousands)
30-40 40-50 50-60
60-70 70-200 Total
VMT
(24 hours)
Emissions (kilc
grains /24-hr, pe
CO
NO
X
HC
Emissions
Rates
(grams per
vehicle-mile)
CO
N0x
^ HC
624,407
riod)
10,906
5,809
2,257
17
9
4
563,007
10,221
5,133
2,073
18
9
4
889,163
16,876
7,929
3,340
19
9
4
1,015,38^
20,849
8,701
3,956
21
Q
4
1,728,011
35,901
14,747
6,773
21
Q
4
2,774,162
60,37^
23,16^
11,126
22
8
4
7,594,134
155,128
65,482
29,525
TABLE 2
GROSS LEVEL LIKE SOURCE EMISSIONS
(FOECTIOML CLASS-FREEWAYS)
-------
-45-
T*
O*AP*»C SCAtf IN Miff
VfHsa.
SYSTEM 7001
„..„„„ LINE SOTIRCSS
(;"TOcrio:uL CIASS-PHIIICIPAL ARTERIALS) 10-15
15-20
20-25
25-50
30-35
35-40
40-100
tiff Ranges (Thousands)
?"~
-------
ro
ei
6
GH
O
C
^
ft!
4-^
O
E-i
O
O
rH
I
O
*3
o
1
10,
LO
to
/O
O
to
1
If"1-
CM
10
CM
1
O
CM
O
CM
rH
to
*~l
rH
O
rH
LO
ON
CO
C—
00
KJ~
•»
o
VO
to
Vl
LO
o
CM
VO
ON
•v?
to
rH
rH
CO
CM
CM
CM
VO
VO
O
KN
cT
ON
rH
#«
rH
LO,
^3-
vk
^-
vo
rH
rH
00
C—
CM
CO
rH
ON
O
f*^^
[^
LO
rH
CO
CM
LO
•s
CM
C-
vo
/*— s
m
l\
o
,r!
r ^ "^'
r ' i CM
[;. v_x
, — s
i
o
•H
!H
1 0
rH
•H fa
^W £3
^^ O
r^H
W |
r* ^J"
O CM
"ft ^***N»
03 CO
w E3
•H n?
f-^ ?H
w ^
ITN
•^j-
CO
^p
VO
CM
to
H
VO
ft
CO
CM
^
LO
rH
O
CM
VO
CM ,
C--
CM
LO
*^-
0
ho,
KN
«.
K>
l/^
00
^,
H
LO
CM
O
CO
CM
LO
LO
*,
H
CM
O
0
rH
r—
ON
rH
•^
\
CM
!O
H
ft
rH
CO
*^+-
£•—
O
^
rX;
1
VO
VO
CO
co"
0
00
ro
«^
CO
rH
ON
06"
CO
CO
•^
N^
W
o
r i
|
^j~ \
CM
00*"
K"\
J
LO
LO
H
•t
rH
O
VO
C~-
00
ON
KN
CO
vo
£
ON
pt
r—
o
rH
VO
f —
CM
CO
ON
r~
LO
CM
to
v>>
to
r.i
t
I
x«s
Ol
rH
W ;-\
n P
f3 W 1
u' '•:! .it,
':" 1 ^-^ t,!j
'•' ' ^
U ">
. , -J > J-l
• . (-1 l'l
l-i '!< Ol
VO
^d"
CM
^"
CM
CO
to
LO
•sf
IO
CM
to
O
C j
VO
vo
VO
VO
VO
.
VO
VO
VO
X
p
? 1
VO
VO
VO
LO
LO
LO
LO
LO
ff]
i
CO
o
PH
[
00
o
o
-46-
-------
- 47-
Banges (ThousandB)
SYSTEM 7001
GROSS LEVEL LIKE SOURCSS '
(RJNCTIONAL CLASS-MINOR ARTERIALS) 15_2Q
20-40
(
-------
- 48 -
ADT Ranges (Thousands)
5-10
-] 5
15-20
20-40 Total
VMT
(24 hours)
Emissions (kill
grams/24-hour ;
CO
NO
X
HC
Emissions
Rat GO (c/voh,-
mile)
CO
NOX
HC
612,870
period)
20,330
3,496
3,076
35
6
5
594,058
19,713
3,369
2,994
33
6
5
319,406
12,611
1,779
1,744
39
6
5
191,680
9,695
1,012
1,193
51
5
6
1,718,014
62,349
9,656
9,007
TABLE 4
GROSS LEVEL LIME SOURCE EMISSIONS
(.FUNCTIONAL GLASS-MIHOR ARTKRIALS)
-------
- 49 -
due to increased flow. The NO rate shows some reductions in the latter
JL
ADT ranges of each class. This is apparently due to reductions in the
average speed on these higher volume sets of links which operate at or
near saturated levels of congestion.
C. Refined Level - Ultimate Definition of Line Sources
The most refined level of line source definition used involved
classifying the links "by narrow ranges of ADT and combinations of
special characteristics defining their attributes. These attributes
have been noted in depth in Chapters II and III. The following para-
graphs show some examples of typical 24-hour graphic and tabular informa-
tion which form the ultimate definition of line sources for the St. Louis
Air Quality Control Region. Detailed commentary will be reserved until
synthesis of information occurs in Table 17.
The initial presentation in Table 5 exhibits the basic .ADT
ranges used in conjunction with the various combinations of attributes.
Figure 15 represents those freeway links in the various ADT ranges
for which none of the attributes of special topography, capacity
alterations, or sensitive land use were present. Table 6 is an
emissions summary of these links from Figure 15. Further, Figure 16
depicts those freeway links in various ADT ranges for which capacity
alterations were present but not special topography or sensitive land
use attributes. Their emissions are summarized in Table 7. Likewise,
Figure 17 depicts those freeway links in various ADT ranges for which
both special topography and capacity alterations were present, but
sensitive land use attributes were not. Table 8 summarizes emissions
for these particular links. Figure 18 r:presents those freeway links,
-------
in various ADT ranges, for which the combination of attributes of
special topography, capacity alterations and sensitive land use were
present. Their emissions are summarized in Table 9 .
The next component series of figures and tables illustrates the
most refined level of line source definition for principal arterials,
using ADT ranges as noted in Table 5 . Figure 19 exhibits those
principal arterial links in various ADT ranges which do not have any
of the attributes of capacity alteration, progressive movement, or
sensitive land use present. Table 10 summarizes emissions for
these links. Figure 20 depicts those principal arterial links in
various ADT ranges which had attributes of capacity alteration, but no
attributes of progressive movement, or sensitive land use. Their emis-
sions are summarized in Table 11. Further, Figure 21 illustrates
those principal arterial links in various ADT ranges which have attributes
of capacity alteration and progressive movement,but not sensitive land
use. The emissions summary for these links is shown in Table 12.
Figure 22 exhibits those principal arterial links in various ADT
ranges which have all the attributes of capacity alteration, progres-
sive movement, and sensitive land use. Their emissions are summarized
in Table 1?.
The final example of line source definitions is composed of
minor arterials in the ADT ranges previously noted in Table 5. Figure
2J illustrates those minor arterial links in various ADT ranges which
did not have any attributes of capacity alterations, progressive move-
ment, or sensitive land use. Their emissions are summarized in
Table 14« Figure 24 shows those minor arterial links in various
-------
-51-
ADT
Ranges
(Thousands)
Freeways
Principal Arterials
Minor Arterials
5-10
10-15
15-20
20-25
X
20-40
25-30
30-35
X
35-40
40-45
45-50
50-55
55-60
60-65
65-70
X
70-200
X
40-100
TABLE 5
APT RANGES USED FOR ULTIMATE
LINE SOURCE DEFINITION
-------
-52-
1OT Ranges (Thousands)
30-55
55-40
Attributes! Special Topography i not present 40-45 *
Capacity Alterations: not present 45-50
Sensitive Land Use: not present 50-55
65-70 .'.V
70-200 w/m
-------
•XJ
CO
0)
to
i-H
-P
O
EH
O
0
CM
ci
f-
o
*!"
in
vo
VO
vo
o
VO
in
in
in
1
o
in
o
in
£
in
Vl
^
O
1
10
I
Cl)
N^
i>-
vo
*
rH
VO
CO
a\
in*
CM
in.
^^*
C7\
«^
CM
rH
O
m
CM
KN
KN
IfN
rH
O
0
O
O
CM
VO
•*^*
•»
O
ON
^^*N
n
jL|
r-j
b
TJ
O
•H
I q>
O P
H
'H ^
v_x 0
CO |
S ^~
O CM
•H ~\
0 CO
H
CM
*
CM
CM
CM
rH
00
ON
in
VO
o
^
CO
ON
CM
0
o
o
0
t—
ON
VO
•k
CM
O
ON
rH
*vT
y*
o
r—
^
rH
ON
O
00
O
O
in
en
rH
O
O
O
O
0
ON
VO
M
p
f •*
00
CM
«.
rH
VO
r-T
ON
O
00
O
CM
00
m
o
0
0
o
CM
^*
O
1
•
'p
«""C
f,\ ^0
O ' — '
•1-1
CO ^} Ci
"-J -P rH
m
KN
m
o
ON
rH
O
O
O
o
o
KN
O
O
VO
VO
O
ON
o
o
o
o
00
w
0
-
VO
o
^
o
o
o
o
, in
p
c
0)
CO
• o
M
CO
VO
H
W
s
•8
0) -P
M C
_ ft £
, . ft rt ct
EH CO O K
O
ca
0)
-p
- 53 -
-------
-54-
..,*'
iOT Ranges (Thousands)
ULTIMATE UHE SOTjnCSS JO-J5 W8
(fUTCTIOIIAL CLASS-?SSEtfAYS) 35-40 »=«=•
Attributes: Special Topographyi not present 40-45 •«•«
Capacity Alterations: all categories 45-50 ^5$s
Sensitive Land Use: not present 50-55 t-i~j
55-60 -v.v
60-65 35-CQr
65-70 ..V.'
70-200 '/////M
-------
n
t5
§
M
§
£
CO
p
fca
rH
CVJ
-P
O
EH
0
0
CM
ci
C-
O
1
IT
O
VD
O
1
in
in
in
in
0
if-
0
in
i
in
3
o
in
in
o
if*
CO
CO
in
vo
CO
rH
O
0
O
O
CO
CO
••
in
0
0
CM
KN
KN
rH
O
0
O
d
(J
£J
O
p"* ^^"
L-4 CM
r" v~-'
s~~.
0
•H
fH
\
rH
•H !4
N~ ' O
W |
£j -3-
O CM
t: Vl
C'J S
•/-( CD
rj H
M ci3
in
CO
CM
O
0
O
O
in
in
O
CM
in
. CM
o
O
o
o
0
^
i-H
in
rH
0
o
0
0
CM
CM
ON
CO
rH
0
O
0
W
p
V£>
^
O
O
0
0
VO
o
o
in
o
0
0
0
u:
it
Q)
Wv.
f? fcO
o > —
•rl
'Jl 0) O
•H -<•> rH
ti d -H
H a; u
o
o
o
o
3
CJN
rH
0
0
O
O
U
O
O
o
o
c—
ON
o
o
o
M
*- ,
o
0
o
o
VO
^.
o
o
o
a
•H
I-
s
CO
O
M
CO
CO
O
O
CO
M
PH
-p HO-P
C CD C!
W -P Q>
w d M
o o
ftHft
rH
-p d -P
o o
c .. c
CO
£> O Q>
,C -H W
ft-P !=>
d d
H fn ti
Cl> S
ftrH »3
O «J
EH co
rH -P -H
Ci iH +"
•H O --H
O d CO
0) ft C
ft d o
co o co
62
0>
-P
- 55 -
-------
- 56-
/
IDT Ranges (Thousands)
ULTIMATE LIKE SOURCES JO-J5 «W
(r,~T:io!:AL CLASS-FHSJ-WAYS) 55-40 *=s
Attributes I Special Topography: all categories 40-45 •»••
Capacity Alterations: all categories 45-50
Sensitive Land Use: not present 50-55
55-60
60-65
65-70
70-200
-------
§
W
g
VI
C.I
to
EH
rH
ni
•P
o
EH
0
o
CM
£
o
f-
lA
vo
LT>
VO
|
O
vo
o
vo
1
l/N
LfN
0
LO
O
tr-
lO
,
VO
rH
0
C—
• v^
/^^N
rrj
O
•rH
rH
1 G)
O Pi
H
•rH ?H
0
43
« 1
O CM
"rH ^^^
11 W
I'J £
•rH Cu
H F
|JH ^3
rH
00
f-
KN
KN
rH
ON
o
VO*
rH
IfN
ON
CM
cT
rH
O
H
^-
00
•<=)-
vo
KN
O
O
00
LfN
t-
CM
CM
CM
LfN
O
O
ON
ON
rH
C—
00
O
•v
CM
VO
f—
00
rH
^
CM
H
CM
VO
IfN
rH
o .
O
CM
rH
CM
H
H
i-T
ON
VO
CM
W
0
JL^
o
CO
"*
VO
rH
0
CM
0
O
•«*•
rH
O
t—
^
ON
vo
o
o
vo
ON
KN
LfN
CO
0
rH
CJ
|
^
"^1
O "^
•rH
n ti ^-— ,
w o O
•r-l -I' rH
i ' ri •'!
!u (-: f '
o
IfN
o
IfN
CM
o
CM
o
o
LfN
CM
00
rH
CO
i-H
O
0
vo
f-
00
ON
o
o
CO
ON
ON
.,
(y
K-i
vo
IfN
^
o
0
-
»
f 5
fl'l
00
EH
CO
O
M
CO
CO
CO
W
CO 'H rt
SS-F
0
to
CD
>3 O
X! -H
PH -P
c3 co
fH rH 'O
tD O C
<; O -P rt
3 p
l
t
3
a
H
•rH -P
O -H
ri M
CQ
•H
H
- 57 -
-------
- 58-
JtDT Ranges (Thousands)
E LINE SOURCES 30- J5
(/;:.CI'IC::AL CLASS- FRIEV/AYS) 35-40 =====
Attribute3i Special Topography: all categories 40-45 ••»»
Capacity Alterations: all categories 45-50 ^t^Z
Sensitive Land Use: all categories 50-55 "=» , - *..»v\
55-60 «v.v 60-65 3.-W33
65-70 .-.V
70-200