&EPA
United States
Environmental Protection
Agency
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
EPA 460/3-84-004
March 1984
Air
Mobile Source Exposure
Estimation
-------
EPA 460/3-84-004
Mobile Source Exposure Estimation
by
MelvinN. Ingalls
Southwest Research Institute
6220 Culebra Road
San Antonio, Texas 78284
Contract No. 68-03-3073
EPA Project Officer: Robert J. Garbe
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
March 1984
-------
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, Environmental Protection Agency, 2565 Plymouth
Road, Ann Arbor, Michigan 48105.
This report was furnished to the Environmental Protection Agency by
Southwest Research Institute, 6220 Culebra Road, San Antonio, Texas,
in fulfillment of Work Assignment 6 of Contract No. 68-03-3073. The
contents of this report are reproduced herein as received from
Southwest Research Institute. The opinions, findings, and conclusions
expressed are those of the author and not necessarily those of the
Environmental Protection Agency. Mention of compnay product names is
not to be considered as a endorsement by the Environmental Protection
Agency.
-------
FOREWORD
This project was conducted for the U.S. Environmental Protection Agency
by the Department of Emissions Research of Southwest Research Institute. The
project was begun in June 1982 and completed in May 1983. The project was
conducted under Work Assignment 6 of Contract 68-03-3073, and was identified
within Southwest Research Institute as Project 05-6619-006.
Mr. Robert J. Garbe of the Emission Control Technology Division, Office
of Mobile Source Air Pollution Control, Environmental Protection Agency, Ann
Arbor, Michigan, served as EPA Project Officer. Mr. Charles T. Hare, Manager,
Advanced Technology, Department of Emissions Research, Southwest Research
Institute, served as the Project Manager. The project was under the super-
vision of Melvin N. Ingalls, Senior Research Engineer, who served as Project
Leader and principal investigator. The assistance of Mr. Thomas McCurdy and
Mr. George Duggan of the EPA, Office of Air Quality Planning and Standards,
and Mr. Roy Paul of PEDCo, Inc., in running the NAAQS Exposure Model (NEM)
is gratefully acknowledged.
111
-------
SUMMARY
This project was conducted to provide the nationwide annual person
hours of exposure to non-reactive mobile source pollutants. The first
activity of the project was to determine the suitability of the National
Ambient Air Quality Standards (NAAQS) Exposure Model (NEM), as used in a
study of alternative standards for CO to provide the exposure estimate.
It was determined that, by itself, the NEM CO study did not provide a
sufficiently accurate estimate of mobile source exposure for the follow-
ing reasons:
• The CO monitor data used were rolled back to meet
the ambient standard being studied.
There was only one mobile source microenvironment
included in the NEM.
• Additional nonautomotive CO sources, such as smoking
and gas stoves were included in the NEM microenviron-
ments.
The NEM for CO, with modified inputs, could be used in conjunction
with a mobile source microenvironment exposure model to produce the desired
exposure estimates. The NEM can be thought of as a "people specific" model,
since it follows groups of people through their daily activities. The
mobile source microenvironment exposure model developed for this project is
a "place specific" model in that it calculates exposure for a given place
and time, and is not concerned with where the people in the microenvironment
are before or after their stay in the microenvironment. Exposure in four
separate microenvironments was examined: parking garages, street canyons,
on-expressways, and roadway tunnels.
For these microenvironments, measured CO concentrations were used as
the indicator of mobile source pollutant concentrations. CO concentrations
were obtained from the published literature for parking garages, on express-
ways, and tunnels. Street canyon CO concentrations were determined by
averaging 23 CO monitors from the EPA SAROAD data base which were identified
as being in street canyons. The nationwide population in these microenviron-
ments for each hour of the day was obtained from published literature.
Using the CO concentrations and the hourly population for each micro-
environment, the nationwide annual person hour exposure to CO was calculated
using the mobile source microenvironment model for parking garages, street
canyons and tunnels. The on-expressway exposure was calculated as part of
the NEM rerun. The exposure estimates obtained were in the form of person hours
of exposure as a function of CO concentrations. To convert these exposure
distributions to exposures that could be used for any mobile source pollutant,
v
-------
the CO concentrations were divided by an emission factor appropriate to
the microenvironment. This produced the exposure distribution that would
be experienced from a 1.0 gram per minute emission factor. To obtain exposure
for any pollutant, the 1.0 gram per minute exposure distribution is multiplied
by the emission factor for that pollutant.
To obtain the exposure everywhere else but in the three microenvironments,
the NEM for CO was rerun with the following changes from the published NEM CO
study:
• Air Quality monitor data were used as measured, not
"rolled back."
• No indoor sources, such as smoking or gas stoves, were
used.
• Additional concentration intervals below 7 ppm CO were
added to the printout.
The exposure distributions produced by the rerun of NEM were converted
to exposure distributions for a 1.0 gram per minute emission factor in the
same manner as the microenvironment exposure distributions,using the nation-
wide urban CO emission factor for 1978, which was the median year of the NEM
air quality data base.
The microenvironment and NEM exposure distributions have not been combined
at this time. This is because the microenvironment and NEM exposure distri-
butions for 1.0 gram per minute must each be multiplied by a different emission
factor to obtain the nationwide exposure estimate for a given pollutant. Once
each distribution has been multiplied by the appropriate emission factor, the
distributions can be added together to obtain the nationwide exposure estimate
for that pollutant.
VI
-------
TABLE OF CONTENTS
Page
FOREWORD ill
SUMMARY v
LIST OF FIGURES ix
LIST OF TABLES xi
I. INTRODUCTION 1
II. INVESTIGATION OF METHODOLOGY 3
III. MICROENVIRONMENT POLLUTANT CONCENTRATIONS
FROM MOBILE SOURCES 7
IV. NUMBER OF PERSONS IN MICROENVIRONMENTS 29
V. EXPOSURE IN MOBILE SOURCE MICROENVIRONMENTS 59
VI. MOBILE SOURCE NEM EXPOSURE ESTIMATE 73
VII. NATIONWIDE EXPOSURE TO MOBILE SOURCE POLLUTANTS 85
VIII. CONCLUSIONS AND RECOMMENDATIONS 89
REFERENCES 91
APPENDICES
A. DEVELOPMENT OF LOGNORMAL POLLUTANT DISTRIBUTIONS
FOR PARKING GARAGES AND ROADWAY TUNNELS
B. FORTRAN LISTING OF SAROAD FILE EDITING PROGRAM FOR
STREET CANYON MONITORS
C. FORTRAN LISTING OF MICROENVIRQNMENT EXPOSURE MODEL
FOR PARKING GARAGES
D. FORTRAN LISTING OF MICROENVIRONMENT EXPOSURE MODEL
FOR STREET CANYONS
E. FORTRAN LISTING OF MICROENVIRONMENT EXPOSURE MODEL
FOR TUNNELS
F. UNIVAC 1100 RUNSTREAMS FOR NEM RERUNS
vi i
-------
LIST OF FIGURES
Figure Page
1 Parking garage CO concentration distribution,
25 percent active cars, average wind speed 11
2 Wind speed distribution, average of seven U.S. cities 13
3 Frequency Distribution of CO concentrations outside
three cars on Los Angeles Freeways 22
4 CO concentration as a function of Hourly Percent ADT
for the Sumner Tunnel (1961) 24
5 CO concentration as a function of Hourly Percent ADT
for an average Roadway Tunnel 26
6 Hourly average cars in motion for weekdays in Parking
Garages 32
7 Hourly average cars in motion for Saturdays in Parking
Garages 33
8 Hourly average cars in motion for Sundays in Parking
Garages 34
9 Number of person trips to CBD per person in urban areas 36
10 Hourly traffic distribution in the CBD for an average
weekday 39
11 Hourly traffic distribution in the CBD for Saturday 42
12 Hourly traffic distribution in the CBD for Sunday 43
13 Pedestrians for individual days of the week as a
percent of total weekly pedestrians 44
14 Hourly pedestrian distribution in the CBD for weekdays 46
15 Hourly pedestrian distribution in the CBD for Saturdays 47
16 Expressway traffic by day of the week 48
17 Hourly expressway traffic for weekdays 52
18 Hourly expressway traffic for Sundays 53
19 Hourly distribution of people in the NEM "transport
vehicle" environment in four cities for Saturdays 54
ix
-------
LIST OF FIGURES (Cont'd)
Figure Page
20 Hourly tunnel traffic for weekdays 56
21 Hourly tunnel traffic for weekends 57
22 Nationwide cumulative exposure distribution in
parking garages 65
23 Nationwide cumulative person hour exposure
distribution in street canyons 68
24 Nationwide cumulative person hour exposure
distribution in roadway tunnels 72
x
-------
LIST OF TABLES
Table Page
1 CO Levels Found in Parking Garages 9
2 Probability Distribution Parameters for Parking
Garage Pollutant Distributions 14
3 Discrete Pollutant Distributions for Parking Garages 15
4 SARQAD CO Monitors Used in Street Canyon Analysis 16
5 Descriptive Statistics for Street Canyon CO Readings 18
6 Distribution of Hourly Average CO Levels in Street
Canyons 19
7 Pollutant Concentration Intervals for Street Canyons 20
8 Measured CO on Expressways 21
9 CO Levels Found in Roadway Tunnels 23
10 Pollutant Concentration Intervals for Roadway Tunnels 27
11 Distributions of Hourly Average CO Levels in Roadway
Tunnels 28
12 Parking Garage Construction in the U.S., 1967-1982 30
13 Estimated CBD Daily Traffic as a Percent of Weekly Traffic 37
14 Daily Person Trips and Vehicles in Street Canyons 38
15 Distribution of People Assigned to the "Transport Vehicle"
Mode in the NEM CO Report 50
16 Total Person Hours of Exposure, on-Expressway Situation 50
17 Traffic Distribution by Day of the Week for Several
Situations 55
18 Values of Variables Used in Determination of Parking
Garage Person Hour Exposure Estimate ^3
19 Person Hour Exposure Distribution for Parking Garages &4
20 Values of Variables Used in Determination of Street
Canyon Person Hour Exposure 66
xi
-------
LIST OF TABLES (Cont'd)
Table Page
21 Person Hour Exposure Distribution for Street Canyons 67
22 Values of Variables Used in Determination of Roadway
Tunnel Person Hour Exposure 69
23 Person Hour Exposure Distribution in Roadway Tunnels 71
24 NEM Input Modifications Required 74
25 Person Hours of Exposure to Mobile Source CO for Chicago 75
26 Person Hours of Exposure to Mobile Source CO for
Los Angeles 76
27 Person Hours of Exposure to Mobile Source CO for
Philadelphia 77
28 Person Hours of Exposure to Mobile Source CO for St. Louis 78
29 1980 Person Hours of Exposure to CO in Four Cities 80
30 Values of Variables Used to Extrapolate NEM Exposure in
Four Cities to Nationwide Exposure 81
31 1980 Nationwide Urban Mobile Source CO Exposure from NEM 82
32 1980 NEM Nationwide Urban Exposure for Mobile Source
Pollutants 83
33 Total Person Hours of Exposure in Parking Garage, Street
Canyon and Tunnel Microenvironments 83
34 1980 Rural Exposure to Mobile Source Pollutants 86
35 1980 Total Nationwide (Urban and Rural) Exposure to
Mobile Source Pollutants Exclusive of Three Micro-
en vi ronments 86
XI1
-------
I. INTRODUCTION
Internal combustion engines used in motor vehicles produce exhaust
gases that contain a multitude of chemical compounds. Four of these
pollutants are regulated by standards (hydrocarbons, CO, NOX and particulate
matter). In devising means to control these four regulated pollutants,
undesirable chemical compounds may inadvertently be produced. Malfunction
of the engine and emission control systems can also change the concentrations
of the various chemical species in the exhaust. Additionally, alternative
fuels proposed for use in motor vehicles can produce greatly different
compounds and concentrations of compounds than are currently produced by
gasoline and diesel fuels. The EPA has instituted a program to determine
if any of these unregulated mobile source emissions cause or contribute
to a risk to public health, welfare or safety.
Previous Work
The work reported on here is the latest in a series of projects
on the subject of unregulated emissions conducted for the EPA by Southwest
Research Institute. Previous projects have developed methodologies for
measurement of unregulated pollutants in vehicle exhaust,(1/2,3)* determined
the magnitude of these unregulated emissions in a variety of vehicles,(1/4-8)
evaluated the effects of engine and emission control system malfunctions,(9-12)
and measured unregulated emissions using a. variety of alternative fuels.(13-17)
Thus, the exhaust emission rates, in terms of mass per distance or mass per
time, of many unregulated pollutants are known.
To determine effects on health and welfare, emission rates of
these unregulated pollutants must be transformed into the ambient pollutant
concentrations to which people are exposed. Another of the previous projects
at SwRI examined localized situations in which the dispersal of mobile
source pollutants is hindered, causing higher than usual concentrations.
Several situations involving small areas, called microscale areas, were
identified. Mathematical dispersion models of these situations were selected
and validated to allow the prediction of ambient concentrations in these
microscale areas, based on knowledge of vehicle exhaust emission rates.
These models allowed the identification of areas of high mobile source pol-
lution and permitted the evaluation of the health and welfare effects of
short term, high level exposure to these unregulated emissions.(19)
It was not possible from that study to ascertain long term effects
due to chronic exposure, such as cancer risk, for unregulated pollutants.
This present project was conducted to provide annual exposure information
*Superscript numbers in parentheses refer to references at end of this report.
-------
resulting from mobile source pollutants which can then be used to evaluate
health effects from long term exposure. The exposure is expressed in terms
of person hours. Person hours of exposure are simply the number of people
in a situation multiplied by exposure time in that situation in hours. Thus,
1000 person hours of exposure can be one person exposed for 1000 hours,
1000 people exposed for one hour, or any other values of persons and hours
whose product is 1000.
Objective
The objective of this project was to develop the methodology to
obtain the nationwide, annual person hours of exposure to any mobile source
pollutant. Then, using that methodology, the annual exposure in person
hours can be determined as a function of ambient concentration for any
mobile source pollutant.
Approach
To determine the annual exposure in person hours to any mobile
source pollutant, the number of persons exposed to the pollutant in various
places must be known as a function of time. The ambient concentration in
these places must also be known as a function of time. The exposure in
person hours can then be ascertained, expressed either as the number of
person hours in various concentration intervals or as an average concentration
for the entire population examined.
The ambient concentrations of most unregulated pollutants are not
available from direct measurements. The concentrations of these pollutants
must be inferred from measurement of some other pollutant. For this project
CO was used as a surrogate, since urban CO is entirely mobile source related.
If the CO concentration in an urban area is known, then the concentration
of other mobile source pollutants can be estimated by multiplying the CO
concentration by the ratio of the new pollutant emission rate to the CO
emission rate:
,_ _ J_J_. __ j_ /Pollutant emission rate\
Pollutant Concentration = CO concentration — : : )
y CO emission rate /
This approach assumes that the desired pollutant and CO have equivalent
dispersion and reaction characteristics in the ambient air.
At the start of the projeect, there was no method available to
combine the concentrations from mobile sources, including the important
microenvironments, with the person in the various environments. The project
approach was to first determine if methods existed that could be used either
directly or with modification. If methods did exist, it was planned to become
familiar with their use, then use them to determine exposure. If no satis-
factory method existed, a methodology would be developed. After collecting
the necessary data, the methodology would be applied to yield the desired
relationship between person hours of exposure and pollutant concentration.
-------
II. INVESTIGATION OF METHODOLOGY
Determining person hours of exposure requires that the pollutant con-
centrations for the areas of mobile source exposure be known, and that the
number of people in these areas be known. There are two basic approaches
to determining person hours of exposure. One approach is to follow people
through their various activities during the day, determining the amount
of time spent in various locations. In this report, this approach is
referred to as "people specific." The second approach is to look at the
places where people encounter mobile source pollutants, then determine the
population in those places for each hour of the day. This approach is
referred to as "place specific." If only total person hours of exposure
is desired, and not an individual person's exposure pattern, then the place
specific approach appears to be the easiest and most accurate.
While all locations in urban environments are exposed to some concen-
tration of pollutants from mobile sources, the highest concentrations occur
in small, confined areas where people and vehicles are in close proximity.
These areas are referred to as microscale areas or microenvironments, and
include such areas as personal garages, parking garages, street caynons,
expressways, and tunnels. Thus, any method must be able to include these
microenvironments.
At the start of the project, it was learned that the EPA Office of Air
Quality Planning and Standards (OAQPS) had developed an exposure model to
evaluate the exposure profile for various levels of pollutant ambient
standards. A study had just been completed for CO using this model.(33)
If it were possible to use the OAQPS model CO results, the project would
be spared the expense and time of developing a new model for mobile source
exposure. The OAQPS-developed model had been given the acronym "NEM",
from "NAAQS Exposure Model."
Evaluation of the NEM CO Report
The NEM basically traces the movement of people in an individual
city, determining the location of similar groups of people for each hour
of the day, each day for a year. It is a "people specific" model. A
pollutant concentration in each location is estimated for each hour of
the year. As used in the CO study, it determined the hourly exposure of
56 different groups of people moving through six different neighborhood
types. In any neighborhood, the groups can be in one of six micro-
environments.
The groups of people, called "Activity-Occupation" (A-O) groups,
had been determined from studies of people's activities. The number of
people assigned to each group for a given city was determined from census
-------
information. The A-O groups have titles such as: sales workers, laborers,
housewives, and children under five.
The six different neighborhood types provide the basis for the
air quality level. For each city, one CO monitor was chosen to represent
each neighborhood type. Thus, for each hour of the year, there is a
different CO concentration associated with each neighborhood type.
The actual exposure level takes into account the fact that people
can be exposed to more or less CO than the monitor level because of their
immediate surroundings—their microenvironment. Six microenvironments are
used: indoors (home), indoors (work), transport vehicle, roadside, outdoors,
and kitchen. For each microenvironment there is a single multiplication
factor and an additive factor applied to the hourly CO monitor values.
The NEM produces a person-hour exposure distribution in various
pollutant concentration intervals for a single city. Currently, NEM CO
results are available for four cities. These four city results are then
extrapolated to the entire country.
The NEM study was reviewed in detail for its applicability to
mobile source exposure estimation. It was concluded that the CO study, as
published, was not satisfactory for use in determining exposure estimates
from mobile sources for the following reasons:
1. Since the purpose of the NEM analysis was to study the
effect of ambient CO standards, the representative CO
monitor data was "rolled back," (i.e., reduced) so that
all areas meet the three different CO standards being
investigated. Thus, the exposure distributions obtained
did not represent an actual distribution which could be
used with a calendar year CO mobile source emission factor,
but rather a distribution that would exist if certain
ambient CO standards were met.
2. The microenvironments considered were only:
• indoors (home)
• indoors (work)
• in a vehicle
• roadside
• outdoors
• kitchen
A number of microscale environments with the possibility
of high ambient concentrations from mobile sources that
were identified under EPA Contract 68-03-2884, Task
Specification 1, were not included in the NEM study.
-------
3. The microenvironments used in the NEM CO study included
CO sources other than vehicles, such as smoking and gas
stoves. To properly use the CO data to indicate mobile
source emissions there must be no other known CO sources.
4. The use of multipliers on the CO monitor data to obtain
microenvironment CO levels is not sufficient in mirco-
environments that contain large numbers of mobile sources.
In these cases, there is probably no relationship between
the neighborhood monitor CO level and the CO level in the
microenvironment. The work done at SwRI under Contract
68-03-2884 indicates that the number of vehicles and the
microenvironment ventilation rate are the controlling
factors for the CO level in the microenvironment. These
factors have little effect on outdoor CO monitors, nor are
they constant from one microenvironment to anohter in the
same neighborhood.
None of the preceding is intended as criticism of the NEM CO study. The
remarks are only intended to indicate that the NEM study was not, by itself,
a satisfactory method to meet the objective of the present project.
While the NEM model does not .adequately cover the microenvironments
of concern in mobile source exposure, it could be satisfactory for the
mesoscale exposure (e.g., in a suburban housing development), if the monitor
CO values were used "as is," (not rolled back) and the microenvironment
additive factors (which were used to account for sources such as smoking,
etc.) were set to zero.
Final Methodology
Since the NEM could provide an exposure estimate of mesoscale
exposure, which accounts for most of the yearly person hours of exposure,
it was decided that the NEM computer program should be rerun with the
following changes to the input instructions:
• CO monitor data would be used "as is"
• No additive sources, such as smoking, would be used in the
NEM microenvironments
To determine the person hours of exposure to those mobile source
microenvironments not accounted for by the NEM, a new model was developed.
Four microenvironments were investigated: parking garages, street canyons,
expressways and roadway tunnels. The NEM appears to satisfactorily cover
the microenvironments where people are exposed to lower levels of mobile
source pollutants, such as non-street caynon, non-expressway streets, in
the NEM "outdoor" and "beside roadway" microenvironments. Thus, non-street
canyon, non-expressway streets are not examined separately, since to do so
would not make an appreciable change in the NEM-generated exposure
distribution.
-------
The mobile source microenvironment model is a "place specific"
model. To obtain annual person hours of exposure in each of the important
mobile source microenvironments, the model uses the population in a micro-
environment for each hour of the day. Several discrete pollutant concen-
tration frequency distributions were developed for each microenvironment.
For each hour of the day, the frequency of occurrence of each concentration
interval is multiplied by the microenvironment population for that hour to
obtain person hour exposure distribution for that hour. The total exposure
is obtained by multiplying the single hour distribution by the number of
days in a year, them summing the distribution from each hour of the day.
To obtain the range of CO concentrations for the various micro-
environments considered in this project, information from literature searches
done under Contract 68-03-2884, and the EPA SAROAD air quality monitor data
base, was used. The number of persons nationwide in each microenvironment
for each hour of the day was determined from information in the literature.
The NEM results representing only mobile source emissions can be combined
with person-hour exposure distributions from the four microenvironments
(parking garages, street canyons, expressways, and tunnels) to obtain a
single nationwide person-hour exposure distribution to mobile source
pollutants.
-------
III. MICROENVIRONMENT POLLUTANT CONCENTRATIONS FROM MOBILE SOURCES
In order to estimate the exposure of persons to mobile source pollu-
tants, information is needed on the ambient concentrations of these pol-
lutants in each microenvironment. Since it is not possible to know the
pollutant concentration in every occurrence of the microenvironment for all
hours of the year, some method of estimating the distribution of pollutant
concentration values is required.
From an examination of a number of individual examples of each micro-
environment, an estimate of the average concentration and range of concen-
trations within that microenvironment can be obtained. If an assumption
is made about the shape of the distribution, then a mathematical description
of the distribution can be obtained.
Pollutant concentrations within a microenvironment change with the hour
of the day. The change is, in general, dependent on number of vehicles and
ventilation rate. Thus, several concentration distributions for each micro-
environment may be necessary, either as a function of time of day, number of
vehicles, or ventilation rate.
Use of Carbon Monoxide Concentrations
Nationwide, mobile sources produce approximately 76 percent of
the total carbon monoxide emitted into the atmosphere.(2^) For the micro-
environments considered in this study, mobile sources are the only signifi-
cant source of CO. Thus, CO measurements in these microenvironments can
be used to determine the level of any mobile source pollutant.
Pollutant concentrations are proportional to the exhaust emission
rate ("emission factor"). For example, the concentration of a pollutant
emitted from vehicles at the rate of 4 g/min will be twice as high as the
concentration of pollutant emitted at 2 g/min in the same situation. Thus,
CO concentration measurements can be used to determine the concentration
of any unregulated mobile source emission by multiplying the CO concentration
by the ratio of the unregulated emission factor to the CO emission factor.
In this project, the CO concentration is divided by the CO emission factor
in g/min, producing a concentration equivalent to a 1.0 g/min emission
factor. To obtain the pollutant concentration for any unregulated emission,
the concentration at 1.0 g/min is multiplied by the appropriate unregulated
emission factor expressed in grams per minute.
The CO emission factors were obtained from the EPA publication
"Compilation of Air Pollutant Emission Factors: Highway Mobile Sources,"
EPA 460/3-81-005, (21) or directly from the MOBILE 2 computer program which
was used to generate the emission factors in the published compilation.
-------
Different emission factors were used for the different microenvironments
since average vehicle speed varies with the situation. Also, the calendar
year, or years, when measured CO levels were available varied with each
microenvironment.
Use of Frequency Distributions of Pollutant Concentrations
Any investigation of actual pollutant concentrations in any
environment reveals a rather wide range of pollutant levels. Thus, persons
in these environments at different times are exposed not just to one pollu-
tant concentration, but to a range of concentrations. The concentration
level depends mainly on the number of vehicles present and the amount of
ventilation (natural or artificial). The concentrations vary not only with
time at a single location, but also from location to location at a given
time. By accounting for the occurrence of pollutant concentration ranges
for all locations of a microenvironment for a year, a single pollutant con-
centration frequency distribution can be developed which takes into account
the concentration variation with time and location.
In most of the microenvionments investigated in this study,
insufficient CO measurements were available to define the concentration
frequency distribution for the measured data. In those cases where measured
data were not available, some assumption has to be made about the mathe-
matical form of the distribution. The lognormal distribution has been
used for about 15 years to describe the distribution of ambient pollutant
measurements both with time and location. '22) Tne lognormal distribution
gets its name from the fact that the logarithms of the independent variable
are normally distributed. Over the years, a lively debate has been conducted
in the literature on whether the lognormal distribution was the best repre-
sentation of the concentration distribution. While other distributions,
such as the Weibull distribution, have been suggested, the lognormal appears
to be the most widely used. For this study, the lognormal distribution
has the additional virtue of being definable from available concentration
values. Therefore, the lognormal distribution was chosen as the mathema-
tical form to use when there were insufficient data to define the distri-
bution from measured values.
The expression for the two-parameter lognormal distrubution is:' '
dF(x) = exp
\/2T
2CJ2
(In x - p)'
dx
where:
x = pollutant concentration
y = mean
02= variance
-------
2
The distribution can be completely defined if y and a , are known. Appendix
A presents the various relationships between y and a and the mean, median
and mode which were used in this project to define the lognormal distri-
butions for each microenvironment from measured concentration values.
Parking Garage Pollutant Concentrations
The first microenvironment for which ambient air concentrations
of pollutants were developed was the parking garage. While CO concentrations
in parking garages have been a concern for years, few quantitative data
are available. What is available, is often only as a maximum CO or average
CO concentration. These values are not sufficient for this project. A
distribution of concentrations is needed, since using an average concen-
tration would eliminate any high-level exposures. However, there was not
sufficient information in the literature to determine this distribution.
Table 1 is a list of measured parking garage CO concentrations found in
the literature. Additionally, the modeling study done under Contract
68-03-2884 calculated a CO concentration of 37 ppm in a "typical" (mode
average) garage and 374 ppm for a "severe case" parking garage. These CO
values all indicate that pollutant concentrations are not normally distri-
buted, but rather are skewed, with a long "tail" at the higher concentrations.
TABLE 1. CO LEVELS FOUND IN PARKING GARAGES
CO Levels Reference No. Study Location Study Date
20 ppm to over 100 ppm (off 27 Detroit 1961
scale 100). Several peaks
appear as though they would
exceed 150 ppm
0-200 ppm Car Park A (Sat.) 28 England 1976
0-20 ppm Car Park A (Tues.)
0-200 ppm Car Park Bl
0-30 ppm Car Park B2
33 ppm @ 200 car/hr Car Park B4
Max. 110-130 ppm Car Park Cl
50 ppm Car Park C2
105 ppm Car Park Dl
Max. 400-450 ppm Car Park D2
(mech. vent.)
30-87 ppm eight hr avg 29 Philadelphia 1977
peaks often above 200 ppm 30 Los Angeles 1975
max 365 ppm
-------
Since adequate data to define the distribution do not exist, it
was decided to choose a distribution equation, then use the CO values for
the typical and severe parking garage from Contract 68-03-2884 to define
the distribution. The concentration distribution is then in a mathematical
form which could easily be modified if measured data become available in
the future. As explained earlier, the lognormal distribution was chosen
to represent the pollutant concentration distributions where there were
insufficient data to define the distribution experimentally. As shown in
Appendix A, a lognormal distribution can be completely defined if the median
and mode of the distribution are known. For this project, the typical
garage CO value developed under Contract 68-03-2884 (18) was used as the
mode. The median was adjusted so that the frequency of occurrence in the
"severe" concentration range was approximately equal to one garage.
Assuming 10,000 parking garages in the country, this is equivalent to a
frequency of occurrence of 0.01 percent.
While ultimately distributions in yg/m3 based on an emission
factor of 1 g/min are required, a distribution in ppm CO, using the calcu-
lated typical and severe levels was first computed for a better visuali-
zation of the distribution and for comparison to the data in Table 1. The
mode was taken as 37 ppm^18) and the median adjusted until the frequency of
occurrence in the 300 to 400 ppm interval was approximately 0.01 percent.
This resulted in a median value of 48. This distribution is shown in
Figure 1 . Note that the largest number of occurrences are in the range
of the CO values shown in Table 1 .
This study requires ambient concentrations in yg/m3 based on a
1.0 g/min emission rate. The typical concentrations obtained from the
modeling work done under Contract 68-03-2884 were actually calculated using
a one gram per minute emission factor. The typical garage has an ambient
air level of 3900 yg/m3 for 1.0 g/min pollutant emission rate; the severe
garage, a level of 46,100 yg/m3.^8) For both these situations, the active
cars represented 25 percent of parking capacity. For this distribution,
the mode was 3900 yg/m3 and the median 5500 yg/m3, giving a y of 8.6125
and a a of 0.58632 (see Appendix A).
The model used for parking garage ambient concentrations in
Contract 68-03-2884 indicates that after a short period, for instance
10 minutes, the concentrations are essentially proportional to the number
of cars divided by the effective ventilation rate. Thus, concentrations
will vary with number of active cars. The active cars vary from 0 to
21.3 percent of parking capacity (see Section IV). To keep the number of
concentration distributions reasonable, it was decided to develop three
distributions, one each for 3, 9 and 19 percent active cars. These dis-
tributions were developed by scaling the 25 percent mode and median by
the ratio of active cars to 25 percent active cars, since concentration
varies linearly with number of cars.
As mentioned above, pollutant concentration varies not only with
number of active cars, but also with effective ventilation rate. The yg/m3
10
-------
24
o
•H
X
ff
0)
O
0)
3
O
O
O
C
O
•H
20
16
12
50
100
150
200
CO, ppm
250
300
350
400
Figure 1. Parking garage CO concentration distribution,
25 percent active cars, average wind speed
-------
pollutant levels for the typical garage were calculated assuming a naturally
ventilated garage with an "avergae" value for wind speed. Ninety percent
of garages in the country are naturally ventilated,' ' with ventilation
rate depending on wind speed. Since the person hour exposure for this
study is to be for an entire year, the ventilation rate can be expected
to vary during the year as wind speed varies. It has been shown in a
number of studies that wind speed at a given measuring station varies
lognormally.(31)
To obtain an estimate of how wind speed varies through the
country, the annual wind speed distributions for seven cities were averaged.
The wind speed distributions for each city were obtained from the NOAA
publication "Airport Climatological Summary."(32) r^he average of the seven
distributions is shown in Figure 2. Note that approximately 65 percent of
the time the average wind speed is between 4 and 10 knots, 25.5 percent of
the time it is between 11 and 27 knots, and 9.5 percent of the time it
is 3 knots or less.
The pollutant distributions above were calculated using approxi-
mately a seven knot wind speed for the typical (mode) garage. To account
for the concentration variation with wind speed, two other wind speeds,
1.5 knots and 14 knots, were chosen to represent ventilation rates lower
and higher than the mode. Pollutant concentration distributions were cal-
culated using three wind speeds by multiplying the mean and mode of each
of the three distributions calculated at 7 knots by the ratio: 7.0/new
wind speed, since pollutant concentration is inversely proportional to
ventilation rate.
These calculations resulted in nine pollutant distributions;
three distributions depending on ventilation rate of each of three levels
of active cars. The mode, median, y and 0 of each of the distributions
are shown in Table 2.
The continuous pollutant distributions generated must be converted
into discrete distributions to obtain the person hour exposure for parking
garages. A computer program was written to evaluate the expression for the
lognormal distribution, producing frequences of occurrence for each pollutant
interval. Table 3 shows the relative frequencies for each pollutant inter-
val in all nine of the pollutant frequency distributions. The pollutant
intervals were arbitrarily chosen. The values shown in Table 3 are one of
the inputs used to calculate the person hours of exposure in parking garages
as explained in Section V of this report.
Street Canyon Pollutant Concentrations
The street canyon exposure calculated in this project is for people
outside on the sidewalk or in vehicles in the street canyon. It does not
include people in buildings adjacent to the street canyon. Unlike parking
garages, where there have been few measurements of vehicle pollutant levels,
CO samples have been taken in downtown street canyons by fixed monitors for
12
-------
u>
0)
Cr>
(0
4-1
C
0)
o
so r
40
i
-H 30
M-l
O
20
10
9.5
0-3
29.8
m
35.2
21.2
3.2
4-6
0.7
22-27
7-10 11-16 17-21
Wind speed, knots
Figure 2. wind speed distribution, average of seven U.S. cities
-------
TABLE 20 PROBABILITY DISTRIBUTION PARAMETERS FOR
PARKING GARAGE POLLUTANT DISTRIBUTIONS
Percent
Active
Cars
19
Wind
Speed,
Knots
1.5
7
14
Probability Distribution Parameters,
Mode Median y o
19500
4200
2100
9.8782
8.3428
7.6496
0.5756
0.5801
0.5801
1.5
7
14
6500
1400
700
9300 9.1378 0.5985
2000 7.6009 0.5972
1000 6.9078 0.5972
1.5
7
14
2200
470
235
3000 8.0064 0.5569
660 6.4922 0.5827
330 5.7991 0.5827
a number of years. The National Air Data Branch of the EPA was contacted
regarding the information that could be retrieved from the data bank of
monitor readings using the SAROAD (Storage and Retrieval of Aerometrics
Data) system. From the information received, it appeared that there was
the potential for obtaining a great deal of information regarding CO levels
in street canyons. However, the data that were easily available were not
sufficient for the needs of the project. Since the raw data for CO is
stored by hour of the day, it is possible to get summaries of specific
monitors by hour of the day for each month of a year. These data could
be used to obtain frequency distributions for weekday, Saturday and
Sundays.
To use the stored monitor data, it was necessary to know which
monitors are in street canyons close to street level. There are currently
over 4000 active sites, which makes it a time-consuming task to locate CBD
sites manually. Discussions with personnel in EPA's Office of Air Quality
Planning and Standards led to the EPA group that is charged with inspection
of all monitors that are part of the National Air Monitoring (NAM) network.
This group was able to furnish a list of central city monitors which are
known to be in street canyons. The list is contained in Table 4.
14
-------
TABLE 3. DISCRETE POLLUTANT DISTRIBUTIONS FOR PARKING GARAGES
Pollutant
Concentration
Interval , yg/m3
0-360
360-463
463-618
618-773
773-1030
1030-1288
1288-1546
1546-1804
1804-2061
2061-2319
2319-2577
2577-3000
3000-4000
4000-5000
5000-6000
6000-8000
8000-10000
10000-15000
15000-20000
20000-25000
25000-30000
30000-40000
40000-50000
Pet. Active cars: 19
Wind Speed, kts: 1.5
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0001
0.0001
0.0004
0.0026
0.0062
0.0113
0.0403
O.D615
0.2008
0.1929
0.1494
0.1060
0.1226
0.0579
19
7
0.0000
0.0001
0.0004
0.0013
0.0062
0.0132
0.0215
0.0299
0.0370
0.0428
0.0466
0.0804
0.1844
0.1513
0.1125
0.1371
0.0666
0.0567
0.0111
0.0029
0.0008
0.0004
0.0001
19
14
0.0034
0.0035
0.0131
0.0248
0.0661
0.0885
0.0981
0.0971
0.0898
0.0840
0.0698
0.0929
0.1370
0.0665
0.0325
0.0255
0.0072
0.0037
0.0003
0.0000
0.0000
0.0000
0.0000
Fraction
9
1.5
0.0000
0.0000
0.0000
0.0000
0.0001
0.0004
0.0009
0.0017
0.0028
0.0043
0.0058
0.0133
0.0401
0.0703
0.0818
0.1678
0.1472
0.2410
0.1132
0.0538
0.0252
0.0206
0.0056
of Time in Interval
9
7
0.0054
0.0052
0.0176
0.0308
0.0760
0.0958
0.1016
0.0975
0.0881
0.0775
0.0663
0.0873
0.1267
0.0610
0.0298
0.0236
0.0068
0.0036
0.0003
0.0001
0.0000
0.0000
0.0000
9
14
0.0435
0.0540
0 . 1096
0.1211
0.1845
0.1433
0.1024
0.0708
0.0490
0.0337
0.0232
0.0242
0.0269
0.0074
0.0024
0.0014
0.0002
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
3
1.5
0.0003
0.0004
0.0020
0.0053
0.0203
0.0368
0.0520
0.0630
0 . 0690
0.0713
0.0701
0.1069
0.1969
0.1235
0.0731
0.0691
0.0243
0.0149
0.0017
0.0003
0.0001
0.0000
0.0000
3
7
0.1481
0.1196
0.1805
0.1497
0.1698
0.0967
0.0535
0.0298
0.0172
0.0100
0.0059
0.0053
0.0046
0.0009
0.0002
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
3
14
0.5656
0.1569
0.1389
0.0684
0.0475
0.0159
0.0058
0.0023
0.0010
0.0004
0.0002
0.0001
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
-------
TABLE 4. SAROAD CO MONITORS USED IN STREET CANYON ANALYSIS
SAROAD No.
10 1960
10 2700
10 4360
11 0200
14 1220
15 2040
18 2380
19 2020
21 0120
22 0240
22 2160
23 1180
33 4680
33 4680
34 0700
36 1220
39 7140
39 7260
41 0300
44 2540
45 1310
45 4570
49 1840
09 0020
26 4280
085H01
018G01
056G01
04 3F01
040F01
034F01
026G01
017F01
034H02
022F01
007F01
021G01
058F01
062F01
029G01
021G01
045H01
005G01
009F01
021G01
053H01
046F01
077F01
022112
079H01
Hourly Readings
Available5 State
7848 FL
8144 FL
7782 FL
6309 GA
7616 IL
7912 IN
7782 KY
5814 LA
7768 MD
6512 MA
1084 MA
8030 MI
8090 NY
8023 NY
8002 NC
7945 OH
8336 PA
6716 PA
7948 RI
8659 TN
7899 TX
2320 TX
8529 WA
N.A.b DC
N.A. MO
City
Jacksonville
Miami
Tampa
Atlanta
Chicago
Indianapolis
Louisville
New Orleans
Baltimore
Boston
Springfield
Detroit
New York
New York
Charlotte
Cincinnati
Philadelphia
Pittsburgh
Providence
Nashville
Dallas
San Antonio
Seattle
Washington
St. Louis
. 8760 hours in one year
Data from these monitors requested, but not available
The purpose in obtaining these data was to provide as large a
measured CO base as possible from which to generate a street canyon
pollutant frequency distribution. Data from all the cities were merged
in one set of CO readings regardless of city size or other variables.
The time and effort allotted to this project were not sufficient to weight
the individual monitor values by city size, street size, or traffic density,
or to compare individual monitor distributions. The data, as received,
were formatted to have 12 hours of data from one station on each line.
This means that one day's data from one station required two lines, if
all 12 hours of data were missing, the whole line was skipped. Thus, some
days would have two lines of data and some would have only one line. The
statistical computer program used to analyze the information (Statistical
Package for the Social Sciences, SPSS) could not handle this type of
variable format. The data from the monitoring stations were also not all
16
-------
in the same units. Some data were in ppm and some in milligrams per cubic
meter. The readings themselves consisted of four integers, with the decimal
point location for the data indicated elsewhere on the line. Additionally,
it was desired to separate out weekdays, Saturdays and Sundays. This
required checking the date of each line and assigning the CO readings to
the correct category. With approximately 14,000 lines of data, the editing
process obviously had to be computerized. A FORTRAN computer program was
written to provide the necessary editing. Appendix B contains a listing of
the editing program.
Once edited, the data were processed using the SPSS program to
obtain frequency distributions, the mean, median, and range of the CO data
in ppm for each hour of the day for weekdays, Saturdays and Sundays. The
minimum, maximum, mean and median for each hour are presented in Table 5.
Next, the CO distributions for each hour of the day were investi-
gated. CO distributions were generated using the SPSS computer program
for each hour of the day. The ppm CO concentrations were divided into 19
intervals. To reduce the number of distributions to a more manageable
level, the individual hourly distributions were examined and similar dis-
tributions combined. This resulted in six distributions for each day type.
These CO distributions are shown in Table 6. These values are presented
in terms of ppm CO, because it is a familiar unit. For use in the project,
the ambient concentrations should be in yg/m3 for an emission factor of
1.0 g/min.
From the published "Mobile 2" data, Table G-9 of Reference 21, a
1981 CO emission factor of 50.51 g/mile was obtained, using a 75°F day.
At 19.6 miles/hr, this emission factor is equivalent to 16.50 g/min.
Using this emission factor, the CO ppm intervals were converted to "Vlg/n\3
of pollutant" intervals by multiplying the ppm CO values by the appropriate
conversion factor (1157/16.50 = 70.12). The new intervals are shown in
Table 7. The distributions using the yg/m3 intervals were used to deter-
mine the person hour exposure in street canyons, as explained in Section V.
On-Expressway Pollutant Concentrations
The approach to this project was to use the NEM CO study results
wherever possible. Since the NEM contains a "transport vehicle" micro-
environment, it was decided to use this microenvironment for the on-
expressway situation, if at all possible. The NEM computer program
attempts to estimate the CO concentrations in cars by multiplying a CO
monitor reading by a constant. As part of the NEM CO study, a literature
review was conducted to determine what this multiplier should be. The
investigation indicated that the "transport vehicle" microenvironment
multiplier should be between 1.3 and 4.7. A value of 2.1 was chosen as
the best estimate.(33)
17
-------
TABLE 5. DESCRIPTIVE STATISTICS FOR STREET CANYON CO READINGS
Weekdays
Saturdays
Sundays
Hour
Ending
1 am
2
3
4
5
6
7
8
9
10
'-' 11
oo J--L
Noon
1 pm
2
3
4
5
6
7
8
9
10
11
Midnight
Concentration
Range
0-12
0-12.5
0-12.0
0-12.5
0-15.0
0-11.3
0-17.1
0-24.5
0-28.5
0-17.6
0-18.0
0-18.5
0-22.3
0-19.9
0-16.0
0-19.2
0-20.8
0-22.7
0-20.5
0-18.3
0.17.0
0-19.6
0-16.0
9-17.0
Mean
1.63
1.37
1.16
0.99
0.98
1.38
2.57
3.97
4.41
3.96
3.92
3.97
4.06
3.94
3.97
4.32
4.89
4.33
3.07
2.57
2.35
2.32
2.28
2.15
, ppm
Median
1.09
0.99
0.83
0.75
0.79
1.02
2.04
3.47
3.81
3.36
3.37
3.45
3.49
3.41
3.51
3.96
4.42
3.65
2.46
2.02
1.98
1.98
1.96
1.73
Concentration
Range
0-11
0-13
0-12
0-9
0-9.5
0-7.3
0-8.8
0-8.5
0-10.2
0-19.0
0-9.5
0-11.0
0-17.0
0-17.0
0-17.5
0-13.0
0-13.3
0-14.0
0-14.5
0-11.5
0-11.6
0-17.0
0-16.4
0-11.5
Mean
2.26
2.10
1.85
1.42
1.20
1.21
1.49
1.78
1.99
2.12
2.28
2.40
2.45
2.53
2.62
2.62
2.57
2.48
2.30
2.35
2.34
2.45
2.59
2.50
, ppm
Median
1.95
1.49
1.18
1.00
0.96
0.98
1.06
1.49
1.70
1.87
1.99
2.01
2.01
2.02
2.03
2.04
2.02
1.99
1.94
1.97
1.96
1.99
2.04
2.00
Concentration
Range
0-19.4
0-16.0
0.15.3
0-14.3
0-13.6
0-12,. 4
0.11.6
0-10.0
0-10.4
0-12.2
0-10.2
0.9.0
0.10.2
0-9.2
0-11.2
0-11.5
0-11.0
0-11.2
0-12.9
0-11.1
0-13.5
0-10.3
0-10.4
0-10.7
Mean
2.26
2.03
1.78
1.39
1.14
1.04
1.12
1.22
1.21
1.30
1.36
1.46
1.57
1.64
1.80
1.85
1.94
1.99
1.98
1.97
1.94
1.87
1.81
1.65
, ppm
Median
1.72
1.48
1.13
0.99
0.88
0.84
0.96
0.99
1.00
1.02
1.04
1.04
1.16
1.22
1.34
1.39
1.45
1.50
1.53
1.60
1.55
1.48
1.32
1.18
-------
TABLE 6• DISTRIBUTION OF HOURLY AVERAGE CO LEVELS
IN STREET CANYONS
PPM CO
Interval
0.0-0.5
0.6-1.5
1.6-2.5
2.6-3.5
3.6-4.5
4.6-5.5
5.6-6.5
6.6-7.5
7.6-8.5
8.6-9.5
9.6-10.5
10.6-11.5
11.6-12.5
12.6-13.5
13.6-15.5
15.6-18.5
18.6-21.5
21.6-24.5
24.6-29.5
0.0-0.5
0.6-1.5
1.6-2.5
2.6-3.5
3.6-4.5
4.6-5.5
5.6-6.5
6.6-7.5
7.6-8.5
8.6-9.5
9.6-10.5
10.6-11.5
11.6-12.5
12.6-13.5
13.6-15.5
15.6-18.5
18.6-21.5
Percent Frequency for Time Interval Shown
l-6am
33.58
40.51
14.35
6.31
2.71
1.30
0.60
0.34
0.15
0.07
0.04
0.02
0.02
—
<0.01
—
—
—
l-3am
19.87
32.46
19.46
11.62
6.50
3.72
2.61
1.68
1.01
0.50
0.34
0.13
0.07
0.03
--
--
—
7am
6.21
28.58
25.68
17.50
10.00
5.41
3.00
1.39
0.98
0.57
0.35
0.10
0.12
0.04
0.04
0.02
—
—
4-6am
31.37
40.96
16.04
6.18
3.31
1.14
0.57
0.30
0.07
0.07
—
--
—
—
--
—
—
8-9am
1.91
11.37
18.62
18.18
15.05
10.49
8.11
6.03
3.26
2.39
1.63
1.04
0.66
0.35
0.47
0.30
0.07
0.03
0.02
7-8am
16.47
42.22
23.39
10.79
4.57
1.41
0.70
0.30
0.10
0.05
—
—
—
—
—
—
—
Weekdays
10 am- 3pm
2.23
12.92
19.88
18.00
14.27
10.39
7.57
5.49
3.50
2.36
1.42
0.76
0.51
0.32
0.29
0.09
0.01
<0.01
Saturday
9-noon
8.39
33.20
27.34
15.64
8.29
4.24
1.49
0.78
0.38
0.15
0.03
0.03
0.03
—
—
—
0.03
4 -6pm
1.82
10.05
16 . 16
16.73
14.21
11.33
9.34
6.75
4.70
3.24
2.16
1.28
0.84
0.53
0.50
0.24
0.09
0.01
l-6pm
6.13
29.98
26.22
15.69
9.47
5.48
3.25
1.67
0.98
0.56
0.27
0.10
0.08
0.05
0.02
0.05
—
7-Mid.
9.23
29.54
25.37
15.50
8.85
4.77
2.75
1.66
0.95
0.56
0.26
0.18
0.10
0.09
0.12
0.04
0.02
—
7-mid.
9.92
30.70
23.11
15.72
9.55
4.62
2.38
1.57
1.16
0.62
0.34
0.19
0.05
0.02
0.02
0.05
—
19
-------
TABLE 6 (CONT'D). DISTRIBUTION OF HOURLY AVERAGE CO LEVELS
FOR STREET CANYONS
PPM CO
Interval
0.0-0.5
0.6-1.5
1.6-2.5
2.6-3.5
3.6-4.5
4.6-5.5
5.6-6.5
6.6-7.5
7.6-8.5
8.6-9.5
9.6-10.5
10.6-11.5
11.6-12.5
12.6-13.5
13.6-15.5
15.6-18.5
18.6-21.5
Percent
Frequency
for Time Interval Shown
Sunday
l-2am
18.51
31.90
20.31
11.49
7.64
3.08
3.38
1.44
1.08
0.41
0.15
0.21
0.21
—
0.05
0.10
0.05
3-4am
27.18
38.85
16.18
8.30
3.84
1.97
1.66
0.93
0.41
0.31
—
0.16
—
0.10
0.10
—
—
5-10am
28.73
48.47
15.09
4.89
1.43
0.91
0.17
0.12
0.05
0.03
0.03
0.02
0.05
0.02
—
—
11am- 2pm
18.47
46.72
19.76
9.30
2.99
1.67
0.48
0.25
0.20
0.10
0.05
—
—
—
—
—
—
3- llpm
14.84
39.46
21.86
11.60
5.74
2.96
1.54
0.89
0.43
0.28
0.26
0.11
0.01
0.02
—
—
—
Midnight
20.64
40.87
18.15
11.20
4.77
2.49
0.73
0.62
0.10
0.21
0.10
0.10
—
—
—
—
—
TABLE 7. POLLUTANT CONCENTRATION INTERVALS FOR STREET CANYONS
PPM CO Rg/m3 at 1. Og/min a
0-.5 0-35
.6-1.5 36-105
1.6-2.5 106-175
2.6-3.5 176-245
3.6-4.5 246-315
4.6-5.5 316-386
5.6-6.5 387-456
6.6-7.5 457-526
7.6-8.5 527-596
8.6-9.5 597-666
9.6-10.5 667-736
10.6-11.5 737-806
11.6-12.5 807-876
12.6-13.5 877-947
13.6-15.5 948-1087
15.6-18.5 1088-1297
18.6-21.5 1298-1508
21.6-24.5 1509-1718
24.6-29.5 1719-2069
aBased on a CO emission factor of 50.51 g/mile for the
FTP for 1981. At 19.6 miles per hour this is equivalent
to 16.50 g/min. One ppm CO = 1157 yg/m3.
20
-------
Using the 2.1 multiplier and the maximum one hour CO reading
occurring in each of the four cities used in the NEM study, the highest CO
values used for the NEM transport microenvironment are 52 ppm for Chicago,
66 ppm for Los Angeles, 40 ppm for Philadelphia and 48 ppm for St. Louis.
The maximum geometric means for the transport microenvironment from the
six monitors used in.each city are: 6 ppm, 8 ppm, 5 ppm, and 11 ppm for
Chicago, Los Angeles, Philadelphia, and St. Louis, respectively.(33)
The work done at SwRI in developing an on-expressway model,
under Contract 68-03-2884, indicates that using a multiplier on a CO value
from a fixed monitor is not the best way to determine CO levels on express-
ways. Nevertheless, the CO distribution for the transport microenvironment
resulting from the NEM calculations is probably as good as any alternate
distribution that could be developed with reasonable effort. To check the
reasonableness of the NEM distributions, they were compared to the results
of the only studies of on-expressway CO measurements found in the literature.
(34-48) A summary of the CO ranges found i/n these studies is presented
in Table 8. Concentration distributions from Reference 38, for three cars,
are shown in Figure 3.
TABLE 8. MEASURED CO ON EXPRESSWAYS
CO PPM
Cumulative
Frequency %
25
50
75
90
95
1966*
Chicago St.
24
30
46
50
— —
Louis
28
38
44
51
56
est .
Chicago
12
15
22
24
— *~*
1979b
St. Louis
14
19
21
25
27
1982
Car 2
12
20
21
22
30
Los Angelesc
Cars 1 & 3
10
12
15
20
25
From Reference 34
Reference 34 values multiplied by ratio of 1979 to 1966
CO emission factors (30.57/62.06)
From Reference 38
Notes: Reference 35 gives values of 15-20 ppm CO with only one
reading (at 45 ppm) higher than 25 ppm in 1974.
Reference 36 gives values of 15-45 ppm CO in 1977
Comparing the maximum NEM transport microenvironment CO concen-
trations with the 95 percent levels in Table 8, the NEM maximums do not
appear unreasonable, considering that maximum values correspond to the
99.99 percentile of a distribution. The geometric mean values from the
NEM study appear to be reasonable, considering that the distributions shown
in Figure 3 are one minute averages, while the NEM values are one hour
averages. Since there are so few measured on-expressway CO data, and the
NEM CO values do not appear to be grossly out of line, the NEM concentration
21
-------
10
•P
o
EH
M-i
O
-P
C
8
Car 3
10 _
10 20 30
CD concentrations, ppm
40
50
O
EH
4-1
C
01
O
M
0)
P4
Car 2
10 _
10 20 30
CO concentrations, ppm
40
50
4J
g
M-4
O
0)
Car 1
10 20 30
CO concentrations, ppm
40
50
Figure 3. Frequency Distribution of CO concentrations outside
three cars on Los Angeles Freeways
22
-------
will be used for the on-expressway microenviixmment. The NEM program uses
actual hourly CO measurements (or estimated, if actual value missing) from
the SAROAD data base, so no frequency distribution information is produced
by the NEM. While frequency distributions could conceivably be developed
from the hourly measurements, the time and effort allotted to this study
did not permit the development of such distributions.
Roadway Tunnel Pollutant Concentrations
The literature on air pollution in tunnels had been extensively
investigated under Contract 68-03-2884. While CO is monitored in almost
all mechanically ventilated tunnels, there are few published CO data.
A list of the information found is presented in Table 9. The most complete
study found was in Reference 39, which had CO levels for the Sumner Tunnel
in Boston. From the information presented on the Sumner Tunnel in
Reference 39, plots of both average and maximum CO levels as functions off
percent of average daily traffic (ADT) per hour for weekdays were developed.
These plots are shown in Figure 4.
TABLE 9. CO LEVELS FOUND IN ROADWAY TUNNELS
CO Concentrations
Reference
No.
Study Location
110 ppm (7:30-8:00 A.M., No. Tube) 39
190 ppm (7:30-8:00 A.M., So. Tube)
140 ppm ( 5:00-5:45 P.M., West Tube) 39
75 ppm (avg over 3 days) 39
rarely exceeded 180 ppm
12-144 ppm mean 39
30-238 ppm peak
40-200 ppm (North tube) 39
10-60 ppm (Center tube)
10-100 ppm avg (West Tube) 39
250 max
40-250 ppm 39
42-122 ppm (Brooklyn bound tube) 40
Squirrel Hill Tunnel
Pittsburgh, PA
Liberty Tunnel
Pittsburgh, PA
Baltimore Harbor Tunnel
Baltimore, MD
Sumner Tunnel
Boston, MA
Lincoln Tunnel
New York, NY
Fort Pitt Tunnel
Pittsburgh, PA
Armstrong Tunnel
Pittsburgh, PA
Brooklyn Battery Tunnel
New York, NY
Study
Date
1969
1969
~1971
1961
est. 1970
1971
-1971
~1971
23
-------
D
220
200
180
160
Q Maximum
O Mean
Max CO * a + b (% ADT)
a - 52.896
b = 27.674
r2 = 0.7904
a
G
/
/a
Q
a
a.
o
4-1
c
01
o
0
o
u
140
120
100
80
GO
40
Mean CO = a + b (% ADT)
a = 0.78840
b = 19.267
r2 = 0.9724
1234567
Percent ADT
Figure 4. CO concentration as a. function of hourly percent ADT
for the Sumner Tunnel (1961)
24
-------
All tunnels have ventilation systems designed for some maximum CO
rate at the maximum expected traffic levels. This means that for all
tunnels, the CO relationship with percent of average daily traffic (ADT)
should be similar. The maximum design CO level has generally been in the
200 to 250 ppm CO range. Maximum hourly traffic is rarely over eight percent
of ADT. Thus, the Sumner Tunnel data can be used as the basis for develop-
ment of pollutant concentration distributions for roadway tunnels.
A linear regression was performed on the average CO concentrations
and the percent ADT values shown in Figure 4. The results of that analysis
are shown on the figure, together with the results of the regression analysis
of the maximum values. As might be expected, the maximum CO values did not
produce as good a linear fit as the mean CO values. The mean CO level is
obviously a function of the percent ADT.
For this study, a pollutant distribution representing the con-
centrations over all tunnels, for all days of the year, is required. Since
the concentrations are a function of percent ADT, seven different distri-
butions were developed, one for each of seven different levels of percent
ADT. If the pollutant concentration is assumed to be lognormally distri-
buted for a given percent ADT as a result of tunnel^tor;tunnel variability,
fleet composition, weather and traffic flow variability, then, with the
mean concentration taken as the distribution mean and the maximum concen-
tration as an indication of the range, the distributions could be defined
from the data on hand. However, from a summary of tunnel ventilation in
Reference 18, the Sumner Tunnel appears to have a worse than average venti-
lation rate in terms of cubic meters per meter of lane, while having
higher than average ADT. Thus, the Sumner Tunnel is likely to have higher
CO levels than an average tunnel.
The Sumner Tunnel ventilation rate is approximately 25 percent
less than the average rate. This can be taken into account by reducing
the mean CO value shown in Figure 4 by 25 percent at a given percent ADT.
The maximum values will not be changed since they are values that are found
in tunnels and must be included in the distribution. The ADT in the Sumner
Tunnel is approximately 1.6 times higher than the mean ADT for all tunnels.
(18,39) To account for this, the percent ADT can be increased 1.6 times
for a given CO concentration. Again the maximum values are not adjusted.
When these two adjustments are made, the relationships between hourly
percent ADT and CO concentrations for an average tunnel are as shown in
Figure 5.
Assuming that the CO concentration in tunnels has a lognormal
distribution, the CO distribution can be determined from the mean and
maximum CO values at any percent ADT. See Appendix A for the method used
to obtain these distributions. CO distributions were determined for roadway
tunnels at one percent ADT intervals starting with 0.5 percent ADT and ending
at 6.5 percent ADT. The equations obtained from the calculations in
Appendix A are continuous ppm CO distributions. This project requires
discrete distributions in terms of yg/m3 for a 1.0 g/min emission factor.
However, it is also desirable to have the discrete distributions in terms
of ppm CO for comparison with measured data.
25
-------
60 -
I 40
8
20 -
o S
250
230
210
190
170
150
eu
130 -
110
90
70
50
1 23 4 5 6 7
Percent ADT
Figure 5. CO concentration as a function of Hourly Percent ADT
for an average Roadway Tunnel
26
-------
To convert from ppm CO to yg/m3 at 1.0 g/min, the emission factor
for vehicles in tunnels is required. The Sumner Tunnel data were taken
in 1961. Therefore, a 1961 CO emission factor should be used to convert
the data in Figure 5 to a 1.0 g/min emission factor. A national average
fleet emission factor for 1961 is not available from published EPA emission
factors. To obtain the 1961 emission factor, Ms. Lois Platte at the EPA
Mobile Source Laboratory in Ann Arbor, MI was contacted. At her direction,
the EPA emission factor computer program, MOBILE 2, was run to generate a
1961 national fleet CO emission at 35 mph. The resulting emission factor
was 62.06 g/mile, which converts to 36.2 g/min at 35 mph. The conversion
factor for ppm CO at 36.2 g/min to yg/m3 at 1.0 g/min is: 31.96=- 1157/36.2.
The concentration intervals are shown in Table 10. The discrete pollutant
distributions in terms of ppm CO are shown in Table 11 for various levels
of hourly percent ADT. These distributions, with the yg/m3 interval values
replacing the ppm CO values, were used to determine the person hour exposure
in tunnels, as explained in Section V.
TABLE 10. POLLUTANT CONCENTRATION INTERVALS FOR ROADWAY TUNNELS
PPM CO yg/m3 at 1.0 g/mina
0 0
6.26 200
12.52 400
18.77 600
21.90 700
25.03 800
28.16 900
31.29 1000
37.55 1200
43.80 1400
50.06 1600
56.32 1800
62.58 2000
68.83 2200
75.09 2400
81.35 2600
93.86 3000
125.15 4000
187.73 6000
250.30 8000
Based on a 1961 CO emission factor of 62.06 g/mile
at 35 mph. This is equivlaent to 36.20 g/min.
One ppm CO = 1157 yg/m3 CO.
27
-------
TABLE 11. DISTRIBUTIONS OF HOURLY AVERAGE CO
LEVELS IN ROADWAY TUNNELS
PPM
CO Interval
0-6.26
2.26-12.52
12.52-18.77
18.77-21.90
21.90-25.03
25.03-28.16
28.16-31.29
31.29-37.55
37.55-43.80
43.80-50.06
50.06-56.32
56.32-62.58
62.58-68.83
68.83-75.09
75.09-81.35
81.35-93.86
93.86-125.15
125.15-187.73
187.73-250.30
Frequency f
0.5
0.777
0.129
0.044
0.011
0.008
0.006
0.004
0.006
0.004
0.003
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.000
1.5
0.288
0.317
0.169
0.052
0.038
0.029
0.022
0.030
0.019
0.012
0.008
0.005
0.004
0.003
0.002
0.003
0.003
0.001
0.000
2.5
0.067
0.230
0.218
0.086
0.071
0.058
0.047
0.069
0.046
0.032
0.022
0.015
0.011
0.008
0.006
0.007
0.008
0.004
0.001
or Percent
3.5
Weekdays
0.013
0.109
0.175
0.088
0.081
0.073
0.065
0.105
0.077
0.056
0.041
0.030
0.022
0.016
0.012
0.016
0.019
0.009
0.001
ADT Shown
4.5
0.006
0.045
0.105
0.068
0.071
0.071
0.068
0.120
0.099
0.078
0.061
0.047
0.036
0.028
0.021
0.029
0.033
0.019
0.002
5.5
0.001
0.018
0.057
0.045
0.052
0.057
0.059
0.116
0.107
0.092
0.077
0.063
0.051
0.041
0.033
0.047
0.057
0.035
0.005
6.5
0.000
0.007
0.030
0.027
0.035
0.042
0.047
0.100
0.102
0.095
0.085
0.074
0.062
0.052
0.043
0.065
0.085
0.056
0.009
28
-------
IV. NUMBER OF PERSONS IN MICROENVIRONMENTS
This study is concerned with the total, nationwide, annual person hours
of exposure. As such, it is not necessary to know which particular people
are exposed, or how the microenvironment is related to other exposures. It
is only necessary to know how many people are in each microenvironment for
each hour of the day. To determine the microenvironment hourly population,
information on the number of locations, size range, and daily usage of each
microenvironment were required. The published literature, as well as special
sources, were used to obtain the best possible estimates of these parameters.
Number of persons in Vehicles
To obtain the hourly population for all microenvironments, an esti-
mate of the number of persons in each vehicle is required. There are a vari-
ety of estimates given in the literature. Early in this project, it was
realized that obtaining an accurate estimate of person hour exposure would
require defining the microenvironment populations on Saturday and Sunday, as
well as during the work week (called "weekdays" in this report). Thus, any
differences in vehicle occupancy between weekdays and weekends should be
included in the calculations. The vehicle occupancies used in this study,
taken from References 41 and 42, are shown below.
Weekday Saturday Sunday
a a a
Cars 1.4 2.3 2.3
b b b
Buses 26 23 10.6
From Reference 41
b
From Reference 42
Population of Parking Garages
From the literature search conducted under Contract 68-03-2884,
Task Specification 1, approximately 70 abstracts on parking garages were
reviewed for information on usage and pollutant levels. Of these, ten had
some useful information. The earlier study of parking garages had obtained
information from several different sources on the number of parking garages
in the country. The result was an estimate that varied from 5300 to 10,000
parking garages in the country. This spread in the estimated number of
garages was too large to provide the estimate of the population in parking
garages needed for this project.
29
-------
For this project, a list of parking garage construction projects
since 1967 was obtained from Data Resources, Inc. This company provides
construction statistics based on the F. W. Dodge data bank of construction
projects. F. W. Dodge is widely recognized as the authoritative source for
statistics on all types of construction projects. The parking garage con-
struction projects in the U. S. from 1967 to 1982 are listed in Table 12.
The total number of projects from 1967 to 1980 is 8499. .There were approxi-
mately 1291 public parking garages in existence in 1965. In Reference 18,
construction data found in the magazine, "Parking," indicated that there
were an average of 330 parking garages built per year between 1971 and 1980.
Apparently, the construction reported was primarily public parking garages.
The F. W. Dodye data is for all construction, and shows approximately 592
garages constructed per year between 1971 and 1980. The difference between
the two construction estimates is apparently a number of private garages
included in the Dodge statistics. If the same ratio of public to private
garages has always existed, then there would be almost twice as many parking
garages in 1965 as the 1291 estimated garages. Assuming 2500 garages in 1965,
and assuming 600 built in 1966, the number of parking garages in 1980 would
be 11,600.
TABLE 12. PARKING GARAGE CONSTRUCTION
IN THE U.S., 1967-1982
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Number of
Projects
682
664
646
587
482
535
642
546
503
493
544
837
717
621
633
596
Source: F.W. Dodge/Data Resources, Inc.
30
-------
The total number of parking spaces in these 11,600 parking garages
was calculated from the information in Reference 18 and from the information
supplied by Data Resources, Inc. The garages were divided into two groups,
one with 6200 large garages averaging 740 spaces per garage, and the other
with 5400 smaller garages averaging 150 spaces per garage. The total number
of spaces is then 5,398,000.
Only three references could be found that had.information on cars
in motion by time of day in a parking garage. ' ' Since there are
over 10,000 parking garages in the country, this is an extremely small sample.
Nevertheless, this information will be used to represent the nationwide average,
since it is all that is available. From the three references, a composite of
weekday vehicles in motion as a percent of garage parking capacity was obtained
for each hour of the day. Similar information for Saturday was derived from
data in Reference 43. The percent of vehicles in motion by hour for Sunday
was estimated considering the values for weekdays and Saturday together with
the fact that for most garages, Sunday would be a day of greatly reduced activ-
ity. The average percent of cars in motion by hour for each of the three types
of days is shown in Figures 6, 7, and 80
Two important points about the fraction of active cars need to be
emphasized. The curves presented are on a per hour basis. However, the actual
time any one vehicle is in motion is much less than one hour, generally on the
order of five minutes. Thus, the vehicles in motion at any instant would equal
the hourly fraction divided by twelve. For the maximum fraction of active cars
per hour (21.34 percent), this gave 1.78 percent active cars at any instant.
This figure agrees well with an Aerospace Corp. study which found an
average of 1.5 percent active cars at all times in Los Angeles garages. The
percent of cars in motion is important since it will be used to select the
proper pollutant concentration distribution in the calculation of person hours
of exposure. The second point is that the people in the cars are, in general,
not exposed to the garage pollution levels for a full hour. A fifteen minute
exposure has been used in this study.
The total nationwide person hours of exposure for any hour of the
day was obtained by multiplying the hourly fraction of cars in motion by the
total number of garage parking spaces available nationwide, then multiplying
by 1.4 persons per car for weekdays or 2.3 persons per car for weekends, then
dividing by 4.0. This calculation is carried out internally within the mobile
source microenvironment computer program, so there is no need to calculate
hourly person hours separately. However, an example may help in visualizing
the number of people involved in the parking garage microenvironment. For
the weekday hour ending at 11 a.m., the active cars are 0.213 of the total
spaces, thus the number of people in the parking garage microenvironment
during that hour in 1980 was:
0.213x(5.398 x 106) x 1.4 = 1,609,684 persons
Then: 1,609,684/4 = 402,421 person hours of exposure
31
-------
range of data
>1
4J
•H
O
tf
&
O
tP
•H
-------
0.30
u>
M-l
o
C >i
O -P
•H -H
-P 0
U CO
m a
i-l tO
M-l O
01 -H
as X
^
(A It!
U rH
ID
0) -P
> o
•H 4-1
-P
U
0.20
0.10
M
8 10 N 2 4 6
Hour of the Day
8 10 M
Figure 7. Hourly average cars in motion for Saturdays in Parking Garages
-------
0.30 r
-------
Number of People in Street Canyons
Recall that the street canyon exposure in this project is for pedes-
trians and motorists within the canyon; it does not include persons in build-
ings adjacent to the street canyon. Therefore, the number of people and
exposure times are those outside in the canyon. Because of sparsity of infor-
mation on street canyon populations on hand, a computerized literature search
was run to assist in locating additional information. The search generated
a listing of 994 abstracts. An examination of the abstracts revealed only eight
references with information useful to the project. In addition, three other
references were located by other means. Only four references had information
useful in determining the total number of people per day traveling to, from,
or within the central business district (CBD) in vehicles.
It was clear from the information obtained that the street canyon
population would have to be obtained in two parts: vehicular and pedestrian.
After examining the data found in'the references, it appeared that the best
way to arrive at CBD population in vehicles was to use CBD cordon counts,
which were available for the peak traffic hour for a number of cities in Refer-
ence 45. Table 10-40, in Reference 45, lists the peak hour person cordon
count for the top 20 urbanized areas in the U.S. Based on these counts, there
were 2,245,000 people passing the CBD cordon limits during the peak traffic
hour in the 20 most populous urban areas.
Peak hour traffic generally averages about eight percent of total
daily traffic. ' Using this value, there would be a total of 28,062,500
person trips whose origin and destination are within the CBD. Approximately 2.5
percent of all trips in an urban area are within the CBD and 27.5 percent of
all trips (urban and rural) either originate or end in the CBD, with other
than a CBD origin or destination. Thus, total trips in the CBD are approxi-
mately nine percent higher (2.5/27.5 = 0.09) than measured at the cordon. To
account for the intra-CBD trips, which are not measured at the cordon, the
cordon counts were multiplied by 1.09, giving 30,616,188 person trips, into,
out of, or within the CBD. The total population of these urban areas was
given as 64,920,646. Using these figures, an average of 0.472 CBD trips/person
per day was obtained.
Two other studies of individual cities (Rochester, NY, and San
Antonio, TX) yielded 0.578 and 0.737 CBD trips per person. (48, 49) From
Reference 45, the urban population averages 2.43 trips per person per day
for all purposes. Reference 47 indicates that 24.5 percent of all urban
trips are into, out of, or within the CBD. These two figures give 0.60
trips per person per day for the CBD. The available data on CBD person
trips per day is plotted as a function of urban population in Figure 9.
Some studies have indicated that there is an increasing number of CBD trips
per person as the size of the urban area decreases. In Figure 9, this trend
appears very weak. In any case, the 0.60 CBD trips per person appears high.
A more reasonable average, obtained from population weighting the data in
Figure 9, is 0.473 CBD trips per day per person.
35
-------
w
CTi
i.Q
0.8
o
U)
0.6
m
Q
M
o;
ft
•H 0.4
O
in
0)
CM
§ °'2
o
o
O
©
O
o
0
O
O Most populous 20 urban areas, Ref. 45
O Providence, R.I., Ref. 45
D San Antonio, Texas, Ref. 49
A Rochester, N.Y., Ref. 48
O
0
O
I
1
I
J_
4 6 8 10
Urban Area Population in Millions
12
14
16
Figure 9.. Number of person trips to CBD per person in urban areas
-------
Using the MEM CO study value for 1980 urban population of 132,023,885
and 0.473 CBD trips per person, there are approximately 62,500,000 CBD trips
per day in the U.S. In earlier studies, it was assumed that street canyons
comprised 60 percent of the CBD streets. (18) If CBD trips are considered
equally distributed over the CBD, then there are 37,500,000 (62,500,000x.600)
person exposures in vehicles to the street canyon environment each day on
weekdays.
For estimates of number of persons exposed on Saturday and Sunday,
three weekly traffic distributions from two different references were used.
(46,47) The average weekly urban traffic distribution obtained from these
references is shown in Table 13.
TABLE 13. ESTIMATED CBD DAILY TRAFFIC AS A PERCENT OF WEEKLY TRAFFIC
Nashville
Urbana CBDa Urbanb Average Used
Sunday 13.2 5.8 10.5 9..9 10
Monday 14.5 15.5 15.2 15.1 15
Tuesday 13.7 15.3 14.7 14.6 15
Wednesday 13.7 15.0 14.7 14.5 15
Thursday 14.0 14.8 14.7 14.5 15
Friday 15.0 15.5 15.9 15.5 15
Saturday 16.2 17.8 14.8 16.3 15
Reference 47
Reference 46
In Table 13, the values in the column headed "Used," are the daily traffic
percentages used in this study for street canyons. The weekday values were
adjusted to give the same percentage for all weekdays, the Sunday value
rounded off to the nearest whole percent, and the Saturday values adjusted
so that the total was 100 percent.
The total number of person trips and vehicles in street canyons
for each day of the week was calculated starting with the base figure of
37,500,000 persons trips in vehicles on weekdays. Of the person trips to
and from the CBD, 47.4 percent are by car and 52.6 percent by transit„<45)
Of the transit trips, 74 percent are bus passengers and 26 percent are rail
passengers. (58) These facts can be used to determine the total person trips
into the CBD as follows:
37,500,000= 0.474y +0.74 (0.526y)
Where y = total number of person trips into the CBD
37
-------
Solving this equation gives 43,441,000 person-trips in the CBD each
weekday, with 20,590,000 (0.474 times 43,441,000) person trips by auto and
16,903,000 (0.74 times 0.526 times 43,441,000) person trips by bus. Using
these values and the values of 1.4 persons per car and 26 persons per bus on
weekdays as shown earlier, the weekday vehicle trips in the CBD can be calcu-
lated as shown below:
Weekday vehicle trips = 20,590,000 + "16,903,000
1.4 26 = 15,358,000
The daily vehicle trips for Saturday and Sunday can be calculated
from the weekday vehicle trips and the daily vehicle trip relationships in
Table 13. The Saturday and Sunday vehicle trips are 15,358,000 and 10,240,000
respectively. Using these vehicle trip values, the persons per vehicle shown
earlier and the weekday ratio between car person trips and bus person trips
(20,590,000/16,903,000 = 1.22) the Saturday and Sunday person trips can be
calculated using the following equation:
BV
person trips = (1.22 + 1)
=: + 1
Where: B = persons per bus
C = persons per car
V = vehicle trips
The average persons per vehicle can then be obtained by dividing the person
trips by the vehicle trips. The results of these calculations are presented
in Table 14.
TABLE 14. DAILY PERSON TRIPS AND VEHICLES IN STREET CANYONS
Weekdays Saturday Sunday
Daily person trips 37,500,000 59,410,000 36,380,000
Daily vehicle trips 15,358,000 15,358,000 10,240,000
Persons per vehicle 2.44 3.87 3.55
The hourly traffic flow as a fraction of daily traffic for weekdays
was obtained from Reference 41. A population weighed average for the data
from various city sizes was used. The hourly traffic distribution is shown
in Figure 10.
38
-------
Data from Reference
- ~- Computed from SAROAD CO data
0.09
0.08
-H °-07
ifl
^ 0.06
•H
Q 0.05
i— i
4-1
^ 0.04
0
_
o 0.03
o
(3
fe 0.02
0.01
—
—
_
-
™
-
•
. 1C
w ^ •
—
—
..-.
M « .
• " •
-, - -
M
8 10 N
6 8 10
P.M.
M
Hour of the Day
Figure 10. Hourly traffic distribution in the CBD for an average weekday
-------
Saturday and Sunday traffic distributions were not found in the
literature for the CBD. However, if the assumption that the CO is almost
entirely from mobile sources is correct, then the hourly traffic distribution
should be able to be deduced from the measured hourly CO distributions for
street canyons given in Section III. To check the assumption that CO level is
related to traffic, the weekday traffic distribution presented in Figure 10
was paired with the SAROAD concentration data in Section III by hour of the
day. A linear regression was performed using the mean hourly CO level and the
hourly percent ADT. The regression analysis produced a coefficient of deter-
mination, r2 , of 0.9598. The equation and its coefficients are:
p = a + bx
Where: p = hourly percent ADT
x = hourly mean ppm CO
a = -1.4972
b = 1.9276
To see the physical meaning of this equation, the linear form can be
rearranged as follows:
P = b (x - c)
In this form, "c" can be thought of as the background concentration, (assumed
constant for the entire day), and "b" as reciprocal of the contribution of one
percent ADT to the ambient CO PPM.
Since:
P = b (x - c ) = bx - be
then: a = -be
and: c = - 7—
b
Thus, from the regression analysis, the background concentration is
- (-1.4972/1.9276) = 0.777 ppm for weekdays, and "b" (the reciprocal of the
ADT "emission factor") is 1.9276 percent ADT/PPM.
The percent ADT for each hour of the day for weekdays was calculated
using the above equation. These values are also shown in Figure 10 for com-
parison with traffic count data. As would be expected from the high coefficient
of determination, the two values of ADT agree closely, demonstrating that the
hourly CO pattern can be used to develop an hourly traffic pattern.
40
-------
Implicit in the equation is some actual value for the weekday ADT.
To apply this equation to Saturday and Sunday, the "b" coefficient must be
adjusted for the traffic count difference between weekdays and Saturdays and
Sundays. The "b" coefficient was changed by the ratio of weekday vehicles to
Saturday and Sunday vehicles. The ratio was 1.0 for Saturday and 1.5 for
Sunday. However, when the weekday equation, adjusted for traffic differences,
was applied to the Saturday and Sunday CO distribution, the calculated hourly
percent ADT values did not total 100 percent.
After some thought, it was realized that the weekend on-the-road
fleet had a smaller percentage of commercial vehicles (mostly trucks) than
the weekday fleet. Since truck CO emissions can be several times car emissions,
this change would reduce the fleet CO emission factor on weekends. With lower
mobile source CO emissions and reduced industrial activity, the weekend back-
ground will also probably be lower.
If the emission factor .and background are assumed to be equal on
Saturday and Sunday, but the Sunday traffic is two- thirds of the Saturday
traffic, then two equations can be written for the percent ADT as a function
of PPM with only two unknowns; the background and the "b" coefficient. When
these two equations are solved, the resulting background level is 0.589 PPM.
The "b" coefficient for Saturday, which has the same total traffic as weekdays,
is 2.650. This is equivalent to a 28 percent reduction in the weekday fleet
emission factor. The "b" coefficient for Sunday is 1.5 times the Saturday
coefficient (or 3.975) to account for the reduced Sunday traffic. The percent
of ADT distributions which result from applying these equations to the hourly
PPM readings in Table 5 of Section III, are shown in Figures 11 and 12 for
Saturday and Sunday, respectively.
Three references for pedestrian street canyon population were all
that could be found in the literature. (50, 51, 52) The total number of
pedestrian trips was calculated from Reference 52 , which indicated that there
were 5.8 weekday pedestrian trips for each weekday vehicle trip into the
Chicago CBD. Since no other information could be found, this number was
used for all CBD's in the country. At 5.8 pedestrian trips per trip in a
vehicle and 37,500,000 person trips in vehicles (Table 14), there are
89,076,000 pedestrian trips in the U. S. each weekday.
From a study of Seattle pedestrians, the daily pedestrians as
a percent of total weekly pedestrians were calculated. This distribution by
day of the week is shown in Figure 13. The average weekday pedestrian trips
are 16.54 percent of the total weekly trips. The total weekly trips are then
538,552,000. Saturday trips are 13.0 percent of the weekly trips, giving
70,012,000 pedestrian trips on Saturday in the U. S. Sunday pedestrian trips
are 4.3 percent of the total weekly trips, or 23,158,000 Sunday pedestrian
trips.
41
-------
Fraction of total Saturday Traff
3 O O O O O O
5 O O O O O O
3 H to OJ *=• <-n CT>
U . UU
_—w
••••••
—
—
—
—
M 2 4 6 8 10 N
A.M.
2 4 6 8 10
P.M.
M
hour of the day
Figure 11. Hourly traffic distribution in the CBD for Saturday
-------
o.o?r
o 0.06
••H
fl3
EH 0.05
m
Ti
§ 0.04
en
|6
0 0.03
m
0
g 0.02
u
(0
£ 0-01
0.00
M
6
A.M.
10
1
2
4
6
P.M.
8
hour of the day
10
M
Figure 12. Hourly traffic distribution in the CBD for Sunday
-------
20
18
§ 16
14
w
0)
•o
01
10
g 6
o
M
0)
Sun
Mon
Tue
Wed
Thu
Fri
Sat
Figure 13. Pedestrians for individual days of the week as
a percent of total weekly pedestrians
-------
From the three references, hourly distributions of pedestrians in
street canyons for weekdays and Saturdays were developed. These distributions
are shown in Figures 14 and 15 as percent of total daily pedestrians. There
were no distribution data available for Sundays. Therefore, the Saturday
distribution in percentage terms was used for Sunday also. There were no data
in the literature as to how long people spend for each pedestrian trip. There-
fore, the exposure period was assumed to be 15 minutes.
Number of Persons on Expressways
The total number of people on expressways and their hourly distribu-
tion were determined from traffic count information in the literature. Since
the NEM has a "transport vehicle," microenvironment, which could be used for
the expressway exposure, the. number of persons and hourly distribution of
people used in the NEM "transport vehicle" microenvironment were also compiled
and compared with the values from traffic counts to determine how well the NEM
values agreed with the expressway literature.
The total number of persons on expressways can be determined if the
total miles of expressway, the average daily expressway traffic, and the
average trip length are known, as shown below.
miles of expressway , . , persons
_ _ x vehicles x _ = person trips
miles of expressway/trip vehicles
From the work done at SwRI under Contract 68-03-2884, it was deter-
mined that there were 16,910 miles of urban expressway, and that the average
daily traffic was 47,664. (18) ^ thorough search of the literature was con-
ducted to define the expressway trip distance. Almost no data were found.
Average urban trip distances (including nonexpressway trips) were found, but
were obviously dominated by short nonexpressway trips, since the mean values
of trips for all purposes were all between 1 and 4 miles. Three studies were
found that could be used in combination to define expressway trip length.(53,54,55)
Unfortunately, they all dealt with the Los Angeles area.
From Reference 55, using results from the 1967 "LART" study, the
average time on freeways in the L. A. area, for persons using the freeway,
can be computed as 13.3 minutes. Reference 53 gives the average speed at
peak hour L. A. freeway traffic as 31.46 mph. Using this average speed and
13.3 minutes as the average time, gives an average of 7.0 miles of expressway
travel per trip. Los Angeles, however, is infamous for its long commuting
trips. In older eastern cities, the average expressway trip has the possi-
bility of being much shorter. Five miles per expressway trip nationwide
appears to be a reasonable estimate.
The average daily traffic (ADT) of 47,664 given in Reference 18
is for a seven day week. Using the weekly expressway distribution from
Reference 49, as shown in Figure 16, the weekday ADT is 50,714, the Saturday
ADT is 45,043, and the Sunday ADT is 35,033.
45
-------
M
Max of available data
Average
Min of available data
M
Figure 14. Hourly pedestrian distribution in the CBD for weekdays
-------
M -14
c
W *
13
0)
* -10
o
^ .08
H
irt
5
0 .06
14-1
O
g .04
-H
O
n)
^ .02
0
^
.
-
-
-
-
-
1 * ,
M 2 4 6 8
A.M.
1
3
^
I
l
i
\
(.
\
3 8 10 M
P.M.
Time of Day
Figure 15. Hourly pedestrian distribution in the CBD for Saturdays
-------
00
O
-H
M-l
(0
20 r-
18 -
16 -
14 _
12 -
U) .-I
(0 X
0)
U 0)
•H S 10
M-l
ffl (^
^ S 8
•H O
(fl
Q -M
Q)
U
0)
Sun
Mon
Tues
Wed
Thur
Fri
Sat
Figure 16. Expressway traffic by day of the week
-------
Computing the total person exposures, using 16,910 miles of urban
expressway, average weekday daily traffic of 50,714, average trip length of
five miles and 1.4 persons per car, gives 240,120,647 person exposures per
weekday. However, as shown above, trip times are on the order of one quarter
hour. Using the 240,120,647 weekday person exposures calculated above, at
one quarter hour per exposure gives 60,030,162 person hours of weekday express-
way exposures. Similar calculations can be made for Saturday and Sunday using
the ADT given above, five miles per expressway trip, and 2.3 persons per vehicle.
The total person trips and person hours of expressway exposure for the three
day types are:
Person trips Person hours
Weekdays 240,120,647 60,030,162
Saturday 350,371,480 87,592,870
Sunday 272,507,692 68,126,923
While the NEM has a "transport vehicle" microenvironment as part
of the model, the total number of persons exposed to this situation was not
readily available from the NEM CO report. It was necessary to tally the
people and hours from the activity files of the age-occupation (A-O) groups.
The numbers of persons assigned to the NEM "transport vehicle" category for
weekdays, Saturday and Sunday were calculated using the A-O assignments in
the April 1982 draft of the NEM CO study. The NEM study was actually run
for four cities, then scaled up to a nationwide estimate. The "transport
vehicle" population from the four cities must then also be scaled. Table 15
shows the NEM "transport vehicle" hourly distribution and four city totals,
as well as a scaled nationwide estimate. The scaling factor for the nation-
wide estimate is the total national urban population given in Chapter 8 of
the April 1982 draft NEM CO report, divided by the total non-farm population
of the four urban areas studied (132,023,885 7 15,190,177 = 9.30). The NEM
CO study uses a one-hour exposure time for all environments. Thus, using the
NEM study, a nationwide expressway exposure estimate of 100,075,843 person
hours per weekday was calculated. The NEM person hours in the "transport
vehicle" microenvironment and the person hours derived from traffic counts
are compared in Table 16.
As can be seem from Table 16, the NEM person hours are approximately
65 to 70 percent higher than the person hours calculated in this study.
This is not to imply that the NEM "transport vehicle" microenvironment
figures are incorrect. The NEM "transport vehicle" category represents
all vehicular travel, of which automotive expressway travel is obviously
just a subset. It is reasonable, therefore, that the NEM "transport
vehicle" category should contain more person hours of exposure.
49
-------
TABLE 15. DISTRIBUTION OF PEOPLE ASSIGNED TO THE
"TRANSPORT VEHICLE" MODE IN THE NEM CO REPORT3
Number of People in all Four Cities
Hour of day Weekday Saturday Sunday
Mid - 1 am
1
2
3
4
5
6
7
8
9
10
11
noon
1
2
3
4
5
6
7
8
9
10
11
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- noon
- 1 pm
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- mid.
—
30,041
—
—
—
487,539
1,538,501
1,081,739
727,154
884,372
217,109
198,317
497,537
1,436,532
773,521
1,473,064
1,288,102
122,815
—
—
—
—
—
—
—
—
—
—
—
—
—
2,109,330
2,032,865
308,943
614,568
2,332,526
2,383,733
358,224
1,014,843
619,711
13,411
2,221,929
1,227,262
156,333
159,299
—
— —
—
—
—
—
—
—
1,165,518
3,451,698
2,387,678
1,705,864
351,393
579,666
281,118
250,811
807,800
—
237,141
1,120,290
174,942
—
—
—
4 City Total 10,756,341 15,552,977 12,510,919
National Estimateb 100,075,843 144,703,230 116,400,249
Calculated from data contained in the April 1982 draft of "The NAAQS
Exposure Model (NEM) Applied to Carbon Monoxide"
Four city total multiplied by 9.30
TABLE 16. TOTAL PERSON HOURS OF EXPOSURE ON-EXPRESSWAY SITUATION
Person Hours of Exposure
Third Studya NEM CO Study,
Weekday 60,030,162 100,075,843
Saturday 87,592,870 144,642,686
Sunday 68,126,923 116,351,547
.on-expressway exposure only
"transport vehicle" microenvironment
50
-------
The hourly distribution of people in the on-expressway microenviron-
ment was also investigated. The hourly weekday expressway traffic distribution
was obtained from Reference 41 as a percent of daily traffic. This distribu-
tion is shown in Figure 17. There are few weekend expressway traffic
distributions in the literature. A Sunday expressway distribution for an urban
Chicago expressway was found in the "Highway Capacity Manual," Reference 46.
This distribution is shown in Figure 18. No information was found on Saturday
expressway traffic distribution.
An examination of the microenvironment assignments in the NEM
activity pattern subgroups shows that an effort was made to distribute people
in the "transport vehicle" microenvironment throughout the day and on weekends.
These assignments were, in general, made on an intuitive basis.^3) The NEM
weekday hourly distribution presented in Table 15, was also calculated as a
percent of daily traffic. This distribution is also shown in Figure 17. While
the two distributions are somewhat similar, in that they peak at the same hours,
the peaks are much higher in the NEM distribution. In addition, the NEM dis-
tribution has an anomalous peak at 2 to 3 p.m. Similar calculations were done
for the Saturday and Sunday NEM person hour distributions.
The Sunday distribution is plotted on Figure 18, together with the
distribution from this study. The Saturday distribution is shown on Figure
19. The Sunday distributions do not compare well at all. The NEM distribution
has extremely high percentages between 9 and 11 a.m., which are not confirmed
by traffic count data. The Saturday NEM distribution also has some very high
percentages for the same hours. It is apparent from the weekday and Sunday distri-
butions found in the literature, that hourly traffic does not really reach
these levels.
From this analysis, it is clear that for mobile source exposure, the
NEM model needs to be rerun, changing the number and distribution of cohorts
in the transportation vehicle microenvironment to more accurately reflect the
person hour exposure in the on-expressway situation. However, the effort
allotted to this project was not sufficient to permit the reprogramming of the
NEM A-O activity file to change the number and distribution of people in the
"transport vehicle" microenvironment. Nevertheless, the "transport vehicle"
microenvironment was used for the on-expressway exposure because it contained
reasonable CO concentrations. It should be kept in mind that a more accurate
person hour exposure distribution is possible from the NEM with changes in the
cohort assignments to the "transport vehicle" microenvironment.
Persons in Urban Roadway Tunnels
(59)
From previous studies at SwRI, the total number of urban tunnelsv '
is known, as well as the average daily traffic for all days of the week
(52,000).(18) If the weekly traffic distribution is known, then the weekday
and Saturday and Sunday average tunnel ADT can be calculated. Since tunnels
occur on all types of roadways, expressways, arterials, etc., it was decided
to average the weekly traffic distribution for several different road types to
obtain a weekly distribution for urban tunnels. These distributions and the
average are shown in Table 17. It was not possible to weight them by popula-
tion, traffic, road type, etc., since these values were not available in the
references.
51
-------
NEM "Transport Vehicle"
Expressways
0.16
0.14
Q
-i-t
tO
£ 0.12
S n in
5C U . XU
CN
r-J
« 0.08
0
H
"o 0.06
c
0
•H
o 0.04
m
fa
0.02
n
\j
*™
^*
-
-, .
| •
i i'tr"
1 1
M 2 <•
1
1
n
t «
AI
1
1
i
i
•^«
3
••""•
li
._.
D
1
^
^^BV
...
2
^
1
i
Pt
1— •
I 1 L J
i i !
i ! •
i i
n i
i i
5 8 10 M
1
Time of Day
Figure 17. Hourly expressway traffic for weekdays
-------
0.276
U>
0.20
0.18
0.16
o
•H
M-l
2 o-14
H
M
o 0.12
K
(N
^ 0.10
10
O
EH
g., 0.08
0
O
'$ 0.06
o
(0
0.0<
O.Oi
c
^^
•»
^
_
••
_
•
_
i 1 ~T~~~i i-—^
i \ ' i — "»-— -r~"^ i
ti 1 i i i i I
ill
».«•
"^fe
—
—
M 2 4 6 8 10 N
AM
i — i
i
•
f — 1
J
l
1
2
N3M "Transport Vehicle"
Expressways
^^^_
!
r— T"~j
f
J
1
1
1
i
•
*
r—
i
—
i
"1
H
i
1 — i
i — -i
i
•
i
i
i
j
i
4 6 8 10 M
PM
Time of Day
Figure 18. Hourly expressway traffic for Sundays
-------
0.18
0.16
-P
§ °-14
o
o 0.12
CN
•H 0.10
-P
O
EH
>H 0.08
o
Fraction
0
•
o
(^
0.04
0.02
0
—
^™
-
-
—
-
™"
-
1 1 1 1 1 1
M 2 4 6 8 10
AM
___.
1 i
N 2 4
—
1
i i
5 8 10 M
PM
Time of Day
Figure 19. Hourly Distribution of people in the NEM "transport vehicle" environment
in four cities for Saturdays
-------
TABLE 17. TRAFFIC DISTRIBUTION BY DAY OF THE
WEEK FOR SEVERAL SITUATIONS
Daily Traffic as a Percent of Weekly Traffic
a
Urban
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Reference
b Reference
c Reference
13.
14.
13.
13.
14.
15.
16.
47
46
49
5
5
7
7
0
0
2
CBD
5.
15.
15.
15.
14.
15.
17.
a
8
5
3
0
8
5
8
Nashville
Urbanb
10.
15.
14.
14.
14.
15.
14.
5
2
7
7
7
9
8
San Antonio
Expressway
10.5
15.
15.
15.
15.
16.
13.
0
0
0
0
0
5
Unweighted
Average
10.
15.
14.
14.
14.
15.
15.
1
1
7
6
6
6
6
From the seven day average ADT of 52,000 and the information in
Table 17, the total traffic for all 59 urban commuter tunnels can be calcu-
lated. Using 1.4 persons per car on weekdays and 2.3 persons per car on
weekends, the person exposures for each day type can also be calculated.
The results of these calculations are shown below.
Day Type Total Vehicles/Day Person Exposure/Day
Weekday
Saturday
Sunday
3,199,924
3,350,256
2,169,076
4,479,894
7,705,589
4,988,875
The hourly traffic distribution for two tunnels for weekdays and
weekends was found in the literature. (39,56) one tunnel was identified as
the Surnner tunnel in Boston, the other tunnel simply as an "urban tunnel."
The hourly traffic distributions for these two tunnels are shown in Figure 20
for weekdays, and Figure 21 for weekends. For each day type, the distribu-
tions from the two tunnels look very similar. The data were taken from
different studies conducted during different years, but there is possibility
that the tunnel identified in Reference 56 as an "urban tunnel," is, in fact,
the Sumner tunnel. It is used nevertheless, since there is so little infor-
mation in the literature. No data were found on the individual period of
exposure. Considering that a large number of the exposures occur during
peak traffic periods when traffic is moving at its slowest rate, a five minute
period of exposure was chosen for tunnels.
55
-------
Sumner Tunnel, ref. 39
Urban Tunnel, ref. 56
C 'i i
EH <4-l
<0
O H
C M
O 3
•H O
O.OSi-
0.06
0.04
m -* 0.02
M tM
tl
en
1 1
4 2
1 1 1 1 1
4 6 8 10 N
AM
1 1 1
246
PM
1 1 1
8 10 M
Time of Day
Figure 20. Hourly tunnel traffic for weekdays
-------
Sumner Tunnel, ref. 39
"Urban Tunnel, ref. 56
u. uo
1— 1
(C U
£ :H o.oe
O <4-l
f-t 4_|
(0
4-4 >-l
0 EH
n n/i
c n °'04
0 3
•H O
•P ffi
2^; °-°2
CM
0
ni .
-r --•-j^— j
™P "^Q'l
r ~ '
m
•-i m
nJt_r_^j
h i i i i i i i i i i i i
M246810N246810M
AM PM
Time of Day
Figure 21. Hourly tunnel traffic for weekends
-------
V. EXPOSURE IN MOBILE SOURCE MICROENVIRONMENTS
There are several microenvironments that contribute to mobile source
pollutant exposure. Of these environments, four are considered the most
significant in terms of total person hours of exposure. These four are:
parking garages, street canyons, on-expressways, and roadway tunnels. The
NEM can be used for only one of these four microenvironments: The on-
expressway microenvironment. The person hours of exposure from the other
three microenvironments must be determined independently of the NEM using
a model specifically developed for that purpose.
Microenvironment Exposure Model
To calculate only person hours of exposure in a specific micro-
environment, no information about the movement of people from place to place
is required. All that is required is the population and concentration dis-
tributions, nationwide, for each microenvironment. This approach to calcu-
lating person hours of exposure is referred to as "place specific" as con-
trasted to the NEM which is considered "people specific" since it follows
a group of people through the day.
For a place specific model, the number of people in the micro-
environment multiplied by the hours per year each pollutant concentration
interval occurs gives the person hours of exposure in the microenvironment
to that concentration interval. Because pollutant concentration is a
direct function of the number of vehicles, the most accurate estimates are
obtained by considering hourly populations, with pollutant concentration
ranges appropriate to that population. Since hourly populations are
different for weekdays, Saturdays and Sundays, three sets of hourly popu-
lation figures are required. The person hours of exposure are summed for
all hours of the day and all day types for the entire year to give the
annual person hours of exposure. The mathematical expression for the model
is:
3 24
P = Z I a. T..F..
k j=l i=l 3 13
Where:
P^ = population exposures to concentration interval "k"
TJ_-: = population during hour "i" in the microenvironment on
day type "j"
F^-j, = frequency of occurrence of concentrations in interval
"k" during hour "i" of day type "j"
a-; = number of type " j" days in a year
59
-------
After calculating the person hours within a given pollutant interval, the
intervals are summed to yield the cumulative frequency distribution. The
number of weekdays per year was taken as (52x5) + 1 = 261. From this was
subtracted 13 holidays, for a total of 248 weekdays. The holidays were
further divided into 10 "Saturday type" holidays and 3 "Sunday type."
Thus, there were 52 + 10 = 62 "Saturdays," and 52 + 3 = 55 "Sundays."
The pollutant concentration distributions were generally not
available by hour. For each microenvironment, the model was modified to
account for the way in which the pollutant concentration distribution was
related to the microenvironment population. While it would be possible to
develop one computer program to handle all the variations, it was con-
sidered to be more cost effective to use a separate computer program for
each microenvironment. The computer programs are listed in Appendices C,
D and E for parking garages, street canyons and roadway tunnels, respectively.
For each of these three mobile source microenvironments, hourly
populations and pollutant concentration distributions were developed and
have been presented in previous sections of this report. The remainder of
this section covers the results of applying the microscale exposure model
to parking garages, street canyons and tunnels.
Exposure in Parking Garages
The actual model algorithm used for the parking garage situation
was somewhat different from the general form shown above. This change was
necessary since the pollutant concentrations for parking garages given in
Section III of this report were not functions of time of day, but rather,
functions of the number of active cars and wind speed. Also, the hourly
populations, as shown in Section IV of this report, were expressed as the
fraction of total parking capacity in motion, not actual people.
The computer program considers each of the day types (weekday,
Saturday, and Sunday) separately. For each hour of the day, for each day
type, the program obtains the fraction of active cars in the garage per
hour, then divides by 12 to obtain the fraction of active cars at any
instant. Using this value, a set of concentration distributions corres-
sponding to that fraction of active cars is chosen for that day type and
hour. The set of concentration distributions consists of a distribution
for each of three wind speed ranges. Starting with the fraction of the
total people exposed for each of the selected pollutant concentration
intervals at the lowest wind speed, the total number of person hours in
each interval was obtained by multiplying the total number of active cars
(total spaces times fraction of active cars), the number of people per
car, the number of days per year of the particulate day type, and the
fraction of time the wind was in that particular speed range.
60
-------
The person hours in each concentration interval are summed as each
successive wind speed, hour of the day, and day type is considered. Expressed
mathematically the program processed the equation:
3 24 3
P.= Z I Z a. (SNC. ,/4) F. W
k .;_-, ;_! __! D iJ km m
3=1 1=1 m=l
where:
P. = person hours of exposure in concentration interval "k"
S = total number of parking garage parking spaces available
nationwide •
N = number of persons per car
C.. = hourly fraction of total parking capacity in motion
F = frequency of occurrence of concentration interval k.
A function of concentration "k", fraction of active
cars at any instant, and wind speed "m".
a. = number of type "j" days per year
W = fraction of time wind is within speed range "m"
.m . ^
i = hours
j = day type
m = wind speed interval
People generally are not exposed to the garage pollutant levels for a full
hour. A fifteen minute exposure was used in this study. Thus, for a given
hour, the computer program divides the hourly population by four. After
the person hours in each concentration interval were calculated, the inter-
vals were summed to give the cumulative frequency distribution.
Since cars in motion at any instant obtained from the garage
population curves are always less than two percent of capacity (see Section
IV), the high extremes, such as can be found when a garage is emptying at
the end of an entertainment event or workday, would not be considered.
These situations do occur regularly and should be included in the study.
There are data which show that the cars in motion at any instant in these
situations can be as high as 25 percent of capacity. (!•&) These situations
normally last about one-half hour, for either filling or emptying.
For work related peaks, it was estimated that 25 percent of the
garages experience this type of use at a rate of five times per week. For
entertainment events, it was estimated that 25 percent of the garages expe-
rience this type of usage at a rate of one per week. To account for the
first of the situations, the 5:00 P.M. calculation was split into two parts.
For 75 percent of the population, the concentration frequencies were from
the distributions for 3 percent active cars. For 25 percent of the popu-
lation, the concentration frequencies were from the distribution for 19
percent active cars (see Table 3). The second situation was accounted for
by modifying the 10:00 P.M. Saturday calculation in a like manner.
61
-------
Table 18 contains a list of the variables that were used in the
parking garage exposure model. Where it was necessary to estimate a value,
the estimates were made from impressions and inferences gleaned from the
information on parking garages collected under EPA Contract 68-03-2884,
Task Specification 1. d8)
The nationwide annual person hours of exposure in parking garages
above selected pollutant concentration values calculated from the model
are listed in Table 19 and presented in graphic form in Figure 22. The
figure indicates that the concentration intervals are sufficiently close
to allow linear interpolation to obtain person hours at concentrations
other than those given in Table 19.
This information can be used to determine the person hours of
exposure to various levels of any mobile source pollutant. If the complete
relationship is desired, then the concentrations at a given person hours of
exposure should be multiplied by the parking garage emission factor for the
particular pollutant. If all that is desired is the person hours of
exposure above some concentration of the pollutant, then the actual
pollutant concentration, in yg/m3 is divided by the emission factor to
obtain a concentration at 1.0 g/min emission factor. The person hours of
exposure above this level are then obtained by linear interpolation of
Table 19.
As an example , suppose the person hours of exposure in parking
garages to CO above the one hour NAAQS of 35 ppm are desired. The 35 ppm
is converted to yg/m3 by multiplying by 1157, then divided by the 1980
parking garage CO emission factor of 5 g/min. This gives 8099 yg/m3 at
1.0 g/min emission factor. Interpolating between 8000 and 10,000 yg/m3
in Table 19 gives 26.821x10° annual person hours of exposure above 35 ppm
CO in parking garages.
Exposure in Street Canyons
The previous report sections presented the concentration distri-
butions, total persons in vehicles, total pedestrians, and hourly distri-
butions of vehicles and pedestrians in street canyons for weekdays,
Saturdays and Sundays. The values of these parameters are summarized in
Table 20. These values were used in the computer model to obtain a nationwide
person hour exposure distribution for street canyons. The computer program
used was similar to the program used to calculate the parking garage exposure
distribution. For the street canyon case, the time of an individual exposure
was also taken as 15 minutes. The concentration distribution used was a
function of time of day and day type. The computer program processes the
expression:
3 24
62
-------
TABLE 18. VALUES OF VARIABLES USED IN DETERMINATION OF
PARKING GARAGE PERSON HOUR EXPOSURE ESTIMATE
Variable
Value
1. Total Spaces
2. Persons per car
3. Active time per vehicle
4. Exposure time per person
5. Percent time in
wind speed range
6. Number of weekdays
7. Number of holidays
8. Number of Saturdays
9. Number of Sundays
10. Percent of garages
experiencing peak
active cars
11. Number of hours of peak
activity
12. Fraction of active cars
per hour (exclusive of
peak activity)
13. Distribution of exposure
with concentration
5.398 x 10° =
(6200x740) + (5400x150)
1.4 on weekdays, 2.3 on weekends
5 minutes
15 minutes
9.5%, 0 to 3 kts
65.0%, 4 to 10 kts
25.5%, >10 kts
248
13
62
(52 + 10 holidays)
55
(52 + 3 holidays)
25%
310/year
(248 + 62)
varies with type of day and
hour of day. Minimum = 0,
maximum = .2134
lognormal distribution with
separate distribution for
ranges of active cars and wind
speed. Concentrations vary from
0 to 50,000 yg/m3
63
-------
TABLE 19. PERSON HOUR EXPOSURE DISTRIBUTION FOR PARKING GARAGES
Concentration q
Exceeded, Ug/m"
0
360
463
618
773
1030
1288
1546
1804
2061
2319
2577
3000
4000
5000
6000
8000
10000
15000
20000
25000
30000
40000
Person Hours
(in millions)
1520.380
1182.417
1016.489
802.033
639.181
461.891
358.797
296.807
254.810
224.255
199.532
178.714
150.486
100.445
68.766
48.830
27.293
17.753
8.667
5.375
3.379
2.131
0.683
64
-------
10000?;
1000.
(0
8
§
u
id
o
O
a
w
14-1
o
s
en
M
0)
100
10
0.1
J_
10000 20000 3
Pollutant Concentration, yg/m
30000
40000
Figure 22. Nationwide Cumulative Exposure Distribution in Parking Garages
65
-------
TABLE 20. VALUES OF VARIABLES USED IN DETERMINATION OF STREET CANYON
PERSON HOUR EXPOSURE
Variable
Value
Total nationwide persons in vehicles
in street canyons per day
Total pedestrians in street canyons
Exposure time per person
Number of weekdays
Number of Holidays
Number of Saturdays
Number of Sundays
Fraction of street canyons average
daily traffic
Fraction of total daily pedestrians
in street canyons
Distribution of pollutant
concentrations
37,500,000 on weekdays
59,410,000 on Saturdays
36,380,000 on Sundays
89,076,000 on weekdays
70,012,600 on Saturdays
23,158,000 on Sundays
15 minutes
248
13
62 (52 plus 10 holidays)
55 (52 plus 3 holidays)
Varies by hour of the day and day
type from 0.5 to 8.5 percent ADT
Varies by hour of the day and day
type from 0 to 13 percent of total
daily pedestrians
Determined from street canyon CO
monitors in SAROAD data base. Con-
centrations vary by hour of the day
and day type from 0 to approximately
2100
66
-------
Where:
pk = person hours of exposure in concentration interval "k"
per year
V = total number of persons in vehicles in street canyons per
day for each day type
P = total number of pedestrians in streets canyons per day
for each day type
Cv = hourly fraction of total people in vehicles in street canyons
C = hourly fraction of pedestrians in street canyons
= fraction of time pollutant in concentration interval k. A
function of concentration interval, time of day and day type
aj = number of type " j" days per year
j = day type: weekday, Saturday and Sunday
i = hour of the day
After calculating the person hours in a given pollutant interval, the intervals
were summed to give the cumulative frequency distribution, as shown in Table 21.
Figure 23 is a plot of this distribution. The figure indicates that the
concentration intervals are sufficiently close to allow linear interpolation
to obtain person hours at concentrations other than those given in Table 21.
TABLE 21. PERSON HOUR EXPOSURE DISTRIBUTION FOR STREET CANYONS
Concentration
Exceeded Millions of
yg/m3 a Person Hours
0 9907.003
35 9324.053
105 7332.058
175 5295.949
245 3691.592
315 2520.097
386 1709.217
456 1124.188
526 713.121
596 451.391
666 275.330
736 165.660
806 103.723
876 63.166
947 38.209
1087 14.141
1297 3.332
1508 0.577
1718 0.164
for 1.0 g/min emission factor
67
-------
10,000
1,000
O
o
tn
tn
O
100
s 10
\
N
3
2x
\
^m
:._.__ :>_:|.-
1^=-Y.T'
200 400 600 ~^800 1000 1200 1400 1600 1800 2000
Concentrations
Figure 23. Nationwide cumulative person hour exposure distribution
in street canyons
68
-------
The table can be used to determine the street canyon exposure for
any pollutant in the same manner as for parking garages. For example, to
obtain the person hours of CO exposure in street canyons above the CO one
hour NAAQS of 35 ppm, first convert the 35 ppm to yg/m3 by multiplying by
1157. The 40495 yg/m3 obtained is divided by the 1980 street canyon CO
emission factor of 17.9 g/min, to give 2263 yg/m3 at 1.0 g/min emission
factor. This is above the highest concentration shown in Table 21. There-
fore all that can be ascertained is that, in street canyons, less than
0.164 million person hours of exposure occur annually at CO levels above
35 ppm.
Exposure in Roadway Tunnels
The roadway tunnel person hour exposure distribution can be cal-
culated using the total tunnel traffic and hourly traffic distribution from
Section IV, together with the pollutant concentration distributions from
Section III. A summary of the exposure model input values is given in
Table 22. The person hour distribution for roadway tunnels was calculated
TABLE 22. VALUES OF VARIABLES USED IN DETERMINATION OF
ROADWAY TUNNEL PERSON HOUR EXPOSURE
Variable
Total nationwide person
exposures in roadway
tunnels
Exposure time per person
Number of weekdays
Number of Saturdays
Number of Sundays
Fraction of tunnel average
daily traffic
Distribution of pollutant
concentrations
Weekdays - 4,479,894
Saturdays - 7,705,589
Sundays - 4,988,875
5 minutes
248
62 (52 plus 10 holidays)
55 (52 plus 3 holidays)
Varies by hour of the day
and day type from 0.5 to
6.9 percent ADT
Lognormal distributions, with
six separate distributions,
one for each of six different
values of ADT. Concentrations
vary from 0 to approximately
8000 yg/m3.
69
-------
in the same manner as the street canyon distribution, except that the
exposure time was taken as 5 minutes and the pollutant concentration was
a function of ADT rather than time of day. The computer algorithm for
roadway tunnels is:
3 24
Pu = £ £ [(V.C . J/12] P
where :
P]c = person hours of exposure in concentration interval "k"
per year
V = total number of persons in vehicles in tunnels per day
for each day type
Cv = hourly fraction of total people in vehicles in tunnels
= fraction of time pollutant in concentration interval k.
A function of concentration interval, and Cv
a j = number of type " j " days per year
j = day type : weekday , Saturday or Sunday
i = hour of the day
After calculating the person hours in a given pollutant interval , the inter-
vals were summed to give the cumulative frequency distribution. The cumu-
lative frequency distribution for person hours of exposure in roadway
tunnels is given in Table 23. A plot of the distribution is shown in
Figure 24. The figure indicates that the concentration intervals are
sufficiently close to allow linear interpolation to obtain person hours
at concentrations other than those given in Table 23.
The table can be used to determine the tunnel exposure for any
pollutant in the same manner as for parking garages and street canyons.
For example, to obtain the person hours of CO exposure in tunnels above the
one hour CO NAAQS of 35 ppm, convert 35 ppm to yg/m3 by multiplying by
1157 to give 40495 yg/m^. The 1980 tunnel CO emission factor is 26.92
g/mile(21) at 35 mph or 15.7 g/min. (21) The 40495 yg/m3 concentration is
divided by 15.7 to give 2579 yg/m3 at one gram/min emission factor. Inter-
polating between 2400 and 2600 yg/m3 in Table 23, gives 22.583xl06 person
hours of exposure in tunnels above 35 ppm CO.
70
-------
TABLE 23. PERSON HOUR EXPOSURE DISTRIBUTION IN ROADWAY TUNNELS
Concentration
Exceeded Millions of
yg/m3(a) Person Hours
0 148.859
200 144.970
400 138.681
600 128.618
700 122.167
800 115.082
900 107.536
1000 99.815
1200 84.812
1400 70.884
1600 58.722
1800 48.353
2000 39.686
2200 32.598
2400 26.794
2600 22.089
3000 15.171
4000 6.418
6000 0.847
for 1.0 g/min emission factor
71
-------
140
120
1000
2000
5000
3000 4000
Concentration, ^g/m3
Figure 24. Nationwide cumulative person hour exposure distribution
in roadway tunnels
72
6000
-------
VI. MOBILE SOURCE NEM EXPOSURE ESTIMATE
The methodology for this project was to use the NEM computer model to
provide the bulk of the exposure estimate, then add to that estimate
exposure from several mobile source microenvironments not included in the
NEM.
As explained in Section II, it was found that the input used in the NEM
CO study was not structured to provide a satisfactory estimate of mobile
source pollutant exposure. It could be used, however, if rerun with modified
inputs. As the project progressed, it was also determined that a better
estimate of mobile source exposure could be obtained if, for the NEM "transport
vehicle" microenvironments, the number of people and their hourly distribution
could be modified. Thus, the NEM computer program was rerun to provide a
more useful estimation of mobile source pollutant exposure.
New NEM Input
Several of the input parameters used in the NEM CO study required
changes in the input values. The modifications required are listed in
Table 24. After investigating the NEM input files and the structure of the
NEM program, it was found that changes to the activity patterns would require
more effort than was available for this project. Therefore, the new NEM run
did not include changes to the activity patterns(Item 3 in Table 24).
The NEM computer program is stored on the UNIVAC computer at the
EPA's National Computer Center (NCC) in Research Triangle Park, N.C. The
EPA Office of Air Quality Planning and Standards (OAQPS) has used the NEM
program extensively for a variety of pollutants over the past several years.
It was learned from personnel in the OAQPS that most of the files required
to modify the air quality data and microenvironment factors, while not used
in the published NEM CO study, did exist. A search of the input files
stored on the computer at the NCC located the necessary files.
New Exposure Distribution for the Four Cities
Using the modified input files, the NEM model was rerun for CO for
the four cities used in the NEM CO report. For record purposes, the UNIVAC
computer runstreams for these computer runs are included in Appendix F. The
cumulative exposure distributions are shown in Tables 25 to 28 for Chicago,
Los Angeles, Philadelphia and St. Louis, respectively.
It is emphasized that the published CO study was done for the
purposes of regulatory analysis. The exposure distributions listed in the
draft versions of that report represent exposures that would occur if certain
CO standards were met. The exposure distributions shown in Tables 25 to 28
73
-------
TABLE 24. NEM INPUT MODIFICATIONS REQUIRED
Input Area
Air Quality Data (CO monitors)
Microenvironment Factors
Modification required
EPA study used a variety of rollback
factors. Rollback factors must be set
equal to 1.0 so that CO data is used
"as is."
The EPA study used multiplicative factors
for all, and additive factors for some,
microenvironments. All additive factors
must be set to zero.
3. Activity Patterns
4. Concentration Levels
EPA CO study has nationwide estimates
for only a portion of the U.S. population.
This study requires using all people
(all ages and both sexes). The hour by
hour assignment of people to environments
will require extensive modification.
These modifications are necessary to
correctly calculate the expressway
microenvironment and account for people
in microenvironments not considered by
the NEM.
This input defines the CO intervals in
the distribution. It must be modified
to give more intervals in the lower ppm
range.
74
-------
TABLE 25. PERSON HOURS OF EXPOSURE TO MOBILE SOURCE CO FOR CHICAGO
Total Population, One Hour Averaging Time
Concentrat Ion
Exceeded
(ppm)
60.0
50.0
40.0
35.0
30.0
25.0
20.0
15.0
12.0
9.0
7.0
6.0
5.0
4.0
3.0
2.0
1.5
1.0
0.5
0.0
Max. Concentration
Encounters at Max.
Low Exercise
2,250
164,000
359,000
1,470,000
3,090,000
10,200,000
32,900,000
72,200,000
181,000,000
385,000,000
566,000,000
915,000,000
1,410,000,000
2,460,000,000
5,470,000,000
8,850,000,000
13,100,000,000
17,500,000,000
18,900,000,000
51.1
2,260
Medium Exercise
4,610
25,300
125,000
444,000
1,600,000
5,560,000
16,000,000
29,700,000
60,800,000
118,000,000
260,000,000
586,000,000
867,000,000
1,180,000,000
1,470,000,000
1,560,000,000
32.2
4,620
High Exercise
57,900
344,000
1,360,000
2,570,000
5,930,000
12,900,000
33, 100,000
83,700,000
130,000,000
176,000,000
222,000,000
234,000,000
14.6
15,000
Any Exercise
2,250
164,000
359,000
1,480,000
3,110,000
10,300,000
33,400,000
73,800,000
187,000,000
403,000,000
598,000,000
982,000,000
1,540,000,000
2,760,000,000
6,140,000,000
9,850,000,000
14,500,000,000
19,200,000,000
20,700,000,000
51.1
2,260
tn
-------
TABLE 26. PERSON HOURS OF EXPOSURE TO MOBILE SOURCE CO FOR LOS ANGELES
Total Population, One Hour Averaging Time
Concentration
Exceeded
(ppm)
60.0
50.0
40.0
35.0
30.0
25.0
20.0
15.0
12.0
9.0
7.0
6.0
5.0
4.0
3.0
2.0
1.5
1.0
0.5
0.0
Max. Concentration
Encounters at Max.
Low Exercise
297,000
881,000
4,900,000
10,300,000
33,500,000
79,100,000
187,000,000
509,000,000
1,240,000,000
2,940,000,000
4,860,000,000
7,780,000,000
12,200,000,000
19,700,000,000
39,400,000,000
51,200,000,000
54,300,000,000
60, 100,000,000
61,500,000,000
58.8
66,400
Medium Exercise
188,000
1,530,000
9,010,000
29,100,000
104,000,000
243,000,000
357,000,000
629,000,000
1,000,000,000
1,750,000,000
3,290,000,000
4,310,000,000
4,550,000,000
5,020,000,000
5,100,000,000
28.8
107,000
High Exercise
145,000
568,000
2,540,000
13,100,000
32,500,000
57,300,000
109,000,000
180,000,000
330,000,000
597,000,000
866,000,000
875,000,000
1,010,000,000
1,030,000,000
22.8
43,400
Any Exercise
297,000
881,000
4,900,000
10,300,000
33,700,000
80,800,000
197,000,000
540,000,000
1,350,000,000
3,210,000,000
5,270,000,000
8,520,000,000
13,400,000,000
21,800,000,000
43,300,000,000
56,400,000,000
59,700,000,000
66,200,000,000
67,600,000,000
58.8
66,400
•f-
+-
-------
TABLE 27. PERSON HOURS OF EXPOSURE TO MOBILE SOURCE CO FOR PHILADELPHIA
Total Population, One Hour Averaging Time
Concentration
Exceeded
(ppm)
60.0
50.0
40.0
35.0
30.0
25.0
20.0
15.0
12.0
9.0
7.0
6.0
5.0
4.0
3.0
2.0
1.5
1.0
0.5
0.0
Max. Concentration
Encounters at Max.
Low Exercise
35,600
43,300
88,800
283,000
438,000
2,310,000
5, 180,000
10, 100,000
30,000,000
73,900,000
169,000,000
317,000,000
480,000,000
980,000,000
1,660,000,000
4,890,000,000
7,380,000,000
9,470,000,000
16,400,000,000
23,300,000,000
"M.4
35,600
Medium Exercise
10,500
10,500
10,500
64 , 700
70,200
28 1 , 000
1,110,000
5,040,000
11,400,000
17,400,000
34,800,000
75,000,000
160,000,000
445,000,000
715,000,000
917,000,000
1,480,000,000
1,990,000,000
42.0
10,500
High Exercise
*
215,000
854,000
1,750,000
2,180,000
4,570,000
10,200,000
23,100,000
67,700,000
134,000,000
144,000,000
285,000,000
380,000,000
14.3
58,600
Any Exercise
35,600
43,300
99,300
294,000
449,000
2,380,000
5,250,000
10,400,000
31,300,000
79,800,000
182,000,000
336,000,000
520,000,000
1,060,000,000
1,850,000,000
5,410,000,000
8,230,000,000
10,500,000,000
18,200,000,000
25,700,000,000
71.4
35,600
-------
TABLE 28. PERSON HOURS OF EXPOSURE TO MOBILE SOURCE CO FOR ST. LOUIS
Total Population, One Hour Averaging Time
-j
oo
Concentration
Exceeded
(ppw)
60.0
50.0
40.0
35.0
30.0
25.0
20.0
15.0
12.0
9.0
7.0
6.0
5.0
4.0
3.0
2.0
1.5
1.0
0.5
0.0
Max. Concentration
Encounters at Max.
Low Exercise
79,200
135,000
181,000
326,000
631,000
3,280,000
8,340,000
19,400,000
61,000,000
131,000,000
221,000,000
389,000,000
715,000,000
1,570,000,000
3,750,000,000
5,380,000,000
7,390,000,000
9,000,000,000
9,700,000,000
55.9
29,500
Medium Exercise
6,520
6,520
121,000
405,000
987,000
3,550,000
7,970,000
13,700,000
26,200,000
58,300,000
144,000,000
334,000,000
481,000,000
646,000,000
771,000,000
826,000,000
31.4
6.520
High Exercise
17,800
120,000
591.000
1,110,000
1,840,000
3,150,000
7,880,000
25,700,000
61,500,000
90,700,000
123,000,000
145,000,000
158,000,000
19.8
3,260
Any Exercise
79,200
135,000
181,000
332,000
638,000
3,400,000
8,760,000
20,500,000
65,200,000
141,000,000
236,000,000
418,000,000
782,000,000
1,740,000,000
4, 140,000,000
5,950,000,000
8, 160,000,000
9,910,000,000
10,700,000,000
55.9
29, 500
-------
are estimates of exposures that did occur in the calendar year of the air
quality data used. Thus, it is not possible to directly compare the exposures
shown in Tables 25 to 28 with those in the draft versions of the NEM CO report.
Nationwide Exposure Estimate
The nationwide CO exposure distribution was calculated from the
distribution for the four cities following the procedure used in the April
and December 1982 draft NEM CO reports. (33,57) Before estimating the nation-
wide exposure, it was necessary to obtain the distribution for each city for
1980. The calendar year for the air quality data used for each city varied
by city. The years were: 1979 for Chicago, 1977 for Los Angeles and 1978
for Philadelphia and St. Louis.
To adjust the distribution from each city, the CO levels were
multiplied by the ratio of the 1980 FTP CO emission factor to the city
base year FTP CO emission factor.
n n .^, 1980 FTP CO g/mile
1980 ppm = city base year ppm —
city base year FTP CO g/mile
These calculations produced different CO intervals for each city. Linear
interpolation between the new CO values was used to produce person hour
distributions with the original CO intervals for each city. These person
hour distributions for all four cities are shown in Table 29.
To extrapolate the four cities to a nationwide exposure estimate,
the NEM CO study divided the 105 urban areas in the country with a population
of 200,000 or more (1970 census) into four categories. Each category
corresponded to one of the four cities investigated in the NEM CO study.
The relative CO distribution obtained for the study city was assumed to
represent all urban areas in that category. The urban areas were assigned
to one of the four cities based on such considerations as proximity to
the base area, average wind speed, observed peak CO concentration, climate,
and general character of the area.^57' The nationwide CO exposure estimate
for 1980 was calculated using the following relationships.
, ,_ Total Population 1980 Total Urban Population in 1970
(C)~ Total Population 1970 X Total Population >200,000 in 1970 *
E e. (c) f .
i i
where:
E(c) total nationwide CO exposure distribution
ei(c) = CO exposure distribution for city "i"
Total population in city "i" type areas
fi ~ Population of city "i"
79
-------
TABLE 29. 1980 PERSON HOURS OF EXPOSURE TO CO IN FOUR CITIES
Concentration
Exceeded
Person Hours (Millions)
60.0
50.0
40.0
35.0
30.0
25.0
20.0
15.0
12.0
.0
0
0
0
.0
.0
2.0
1.5
1.0
0.5
0
9.
7.
6.
5.
4
3.
Chicago
0.000
0.000
0.118
0.262
1.000
2.530
8.270
28.500
62.400
163.000
350.000
515.000
846.000
1,368.000
2,501.000
5,662.000
9.064.000
13,843.000
18,868.000
20,700.000
Los Angeles
0.000
0.000
0.389
0.742
3.840
10.000
41.100
124.000
251.000
837.000
1,830.000
2,960.000
5,720.000
9.284,000
16,500.000
34,200.000
48,100.000
53,300.000
64,800.000
67,600.000
Philadelphia
0.036
0.038
0.068
0.103
0.319
1.030
3.640
8.240
19.600
59.300
132.000
207.000
391.000
759.000
1,520.000
4,400.000
7,050.000
9,870.000
17,100.000
25,700.000
St. Louis
0.000
0.000
0.104
0.136
0.205
0.424
1.850
6.510
13.900
46.300
104.000
156.000
290.000
579.000
1,340.000
3,460.000
5,190.000
7,550.000
9,670.000
10,700.000
The value used for each of the variables is shown in Table 30.
When all the various factors are multiplied together, for each of the
values of CO in Table 29, the equation becomes:
National Person Hours =21.051 (Chicago person hours)!
+4.368 (Los Angeles person hours)
r 1
+J9.963 (Philadelphia person hours)
»- J
+;18.211 (St. Louis person hours)
The nationwide Urban CO exposure distribution resulting from these
calculations is presented in Table 31. Again, since the NEM CO study was
conducted to study the effects of various levels of ambient CO standards,
there is no table in the published NEM CO study comparable to Table 31.
80
-------
TABLE 30. VALUES OF VARIABLES USED TO EXTRAPOLATE NEM
EXPOSURE IN FOUR CITIES TO NATIONWIDE EXPOSURE
Variable Value
From References 33 and 57
Total urban population >200,000 (1970) 103,137,849
Total urban population in 1970 118,447,000
Total population 1970 203,212,000
Total population 1980 226,505,000
Total population of Chicago (1970) 2,364,970
Total population of Chicago-like urban
areas >200,000 (1970) 38,894,365
Total population of Los Angeles (1970) 7,719,108
Total population of Los Angeles-like
urban areas >200,000 (1970) 26,339,249
Total population of Philadelphia (1970) 2,935,244
Total population of Philadelphia-like
urban areas >200,000 (1970) 20,553,523
Total population of St. Louis (1970) 1,219,561
Total population of St. Louis-like
urban areas >200,000 (1970) 17,350,712
81
-------
TABLE 31. 1980 NATIONWIDE URBAN MOBILE SOURCE CO EXPOSURE FROM NEM
One Hour Average
Concentration
Exceeded
(CO PPM) Person Hours (Millions)
60.0 0.323
50.0 0.341
40.0 6.687
35.0 12.160
30.0 44.420
25.0 113.900
20.0 419.900
15.0 1,334.000
12.0 2,839.000
9.0 8,462.000
7.0 18,440.000
6.0 28,470.000
5.0 51,580.000
4.0 86,700.000
3.0 162,700.000
2.0 371,000.000
1.5 558,600.000
1.0 772,000.000
0.5 1,010,000.000
0.0 1,156,000.000
NEM Exposure Estimate for Mobile Sources
For the evaluation of unregulated pollutants, the person hour
exposure distribution is needed in terms of JJg/m3 at a 1.0 g/min emission
factor rather than in ppm CO. To convert Table 31 to the required distri-
bution, the CO values are converted to yg/m^ by multiplying by 1157, then
adjusted to one gram/min by dividing by the 1980 FTP CO emission factor
of 17.9 g/min (54.65 g/mile at 19.6 mph).(18) The nationwide person hour
exposure distribution in yg/m3 for a 1.0 g/min emission factor is presented
in Table 32.
To properly combine the NEM results with the results from the
parking garage, street canyons, and roadway tunnel, the total person hours
represented by these three microenvironments must be subtracted from the
NEM distribution.
Table 33 lists the person hours of exposure for each of the three
microenvironments together with the total annual person hours of exposure
for all three situations. Ideally, these person hours would be subtracted, in
82
-------
TABLE 32. 1980 NEM NATIONWIDE URBAN EXPOSURE
FOR MOBILE SOURCE POLLUTANTS
Mobile Source
Pollutant .
Concentrations/ yg/m3
0
32
65
97
129
194
259
323
387
452
582
776
970
1293
1616
1939
2262
2585
3232
3878
Person Hours
from NEM
(Millions)
1,156,000.000
1,010,000.000
772,000.000
558,600.000
371,000.000
162,700.000
86,700.000
51,580.000
28,470.000
18,440.000
8,462.000
2,839.000
1,334.000
419.900
113.900
44.420
12.160
6.681
0.341
0.323
NEM Person Hours
Minus Microenvironment
Person Hours (Millions)
1,144,424.000
999,886.000
764,269.000
553,006.000
367,285.000
161,071.000
85,832.000
51,063.000
28,184.000
18,255.000
8,377.300
2,810.600
1,320.600
415.700
112.760
43.975
12.038
6.614
0.338
0.320
For a 1.0 g/min emission factor
TABLE 33. TOTAL PERSON HOURS OF EXPOSURE IN PARKING GARAGE,
STREET CANYON AND TUNNEL MICROENVIRONMENTS
Microenvironment
Parking Garage
Street Canyons
Tunnels
Total
Yearly person hours
of exposure
(millions)
1,520.380
9,970.003
148.859
11,576.240
83
-------
the form of numbers of people, from the NEM A-O groups in the central city that
would logically be in these microenvironments. The A-O groups would include:
students over 18, professional and administrative, sales workers, and clerical
workers. One method would be to have,the NEM exposure concentrations set at zero
for the number of persons involved in the three microenvironments. The total
person hours could then be subtracted from 0 ppm in the NEM person hour
distribution in Table 32. To do this the NEM "kitchen" microenvironment
multiplicative factor would be set equal to zero, then, each hour, the
number of people equal to those exposed to the three microenvironments
would be assigned to the kitchen microenvironment. Since the ambient
pollutant level would then be multiplied by zero, these people would be
put in the interval containing zero ppm for that hour. Any persons pre-
sently assigned to the kitchen microenvironments would be reassigned to
the "indoor-home" microenvironment, which has the same multiplicative
factor as the kitchen has currently.
, *
Unfortunately, the time and effort allotted to this study did not
permit this adjustment to the A-O group population and activity patterns.
A less exact, but more expeditious method of subtracting the required person
hours from the NEM is to proportionally remove them from each of the NEM
concentration intervals. If 20 percent of the total NEM person hours are
in the interval between 65 and 97 Mg/m3, then 20 percent of the person hours
to be subtracted would be taken from this interval. Using the total person
hours in the three microenvironments as 11,576,240, the NEM exposure distri-
bution was adjusted to give the exposure without the microenvironments.
This adjusted exposure is also shown in Table 32.
84
-------
VII. NATIONWIDE EXPOSURE TO MOBILE SOURCE POLLUTANTS
To obtain the total urban person hour exposure to mobile source pol-
lutants, the microscale exposure distribution from Section V (Tables 19,
21 and 23) must be combined with NEM exposure distribution from Section
VI (Table 32). It should again be emphasized that the exposure estimates
for mobile sources in this report are based on CO. The use of these CO
based exposures as surrogates for other mobile source pollutants must be
approached with reasoned caution. While CO is probably the best surrogate
to use on a national basis for a mobile source surrogate, especially for
the NEM model, other mobile source pollutants may not have the same ends-:
sion rates under the same vehicle operating conditions or chemical reactivity
as CO (e.g., the spatial and temporal distributions may be different).
Adjustments or corrections to these exposure estimates may be appropriate
if the mobile source pollutant under study has characteristics which differ
markedly from CO.. However, in most intended uses of this methodology, a
rough assessment of exposures to a mobile source pollutant which has not
been adequately monitored in the ambient air is desired, and these estimates
may be entirely adequate. Both the microscale exposures and the NEM exposure
are for urban situations. The total nationwide exposure should contain
rural exposure as well. Therefore, an estimate of rural mobile source
exposure is required before a total nationwide exposure estimate can be made.
Rural Exposure
The urban exposure represented by the NEM and microscale models
accounts for the exposure of approximately 132 million of the 226.5 million
people in the country (1980 census).(33) The remaining 94.5 million people
live in rural areas. In general, persons living in rural areas do not
experience high concentrations of mobile source pollutants. To estimate
the magnitude of rural exposure, CO was again used as the indication of
mobile source emissions. Background levels of CO range from 0.03 to 0.22
ppm.(20) However, air masses that have recently traversed urban areas
show levels as high as 1.0 ppm in rural areas.^^'
A detailed examination of rural exposure to mobile source pollutants
was not part of the scope of this study. However, from the data presented
above, it is estimated that all rural exposure to CO is below 2 ppm. Using
the 1980 mobile source FTP emission factor for CO of 17.9 g/min, this 2 ppm
CO converts to a mobile source exposure upper limit of 129 yg/m3 at 1.0 g/min
emission factor. For purposes of this study, it was assumed that 50 percent
of the person hour exposure was between 0 and 32 yg/m3, 30 percent between
32 and 65 yg/m3, 15 percent between 65 and 97 yg/m3, and 5 percent between
97 and 129 yg/m3. For the 94.5 million people in rural areas, the person
hour exposure distribution is then as shown in Table 34.
85
-------
TABLE 34, 1980 RURAL EXPOSURE TO MOBILE SOURCE POLLUTANTS
FOR ONE GRAM PER MINUTE EMISSION FACTOR
Concentration
Exceeded Person Hours
yg/m (Millions)
0 827,820.000
32 413,910.000
65 165,564.000
97 41,391.000
129 0.000
Total National Exposure
Since the rural exposure estimate and the NEM urban exposure
estimate should always use the same mobile source emission factor, the two
distributions can be combined as shown in Table 35. This table gives the
TABLE 35. 1980 TOTAL NATIONWIDE (URBAN AND RURAL) EXPOSURE TO
MOBILE SOURCE POLLUTANTS EXCLUSIVE OF THREE MICROENVIRONMENTS
FOR ONE GRAM PER MINUTE EMISSION FACTOR
Concentration
Exceeded Person Hours
yg/n\3(a) (Millions)
0 1,972,244.000
32 1,413,796.000
65 929,833.000
97 594,397.000
129 367,285.000
194 161,071.000
259 85,832.000
323 51,063.000
387 28,184.000
452 18,255.000
582 8,377.300
776 2,810.600
970 1,320.600
1293 415.700
1616 112.760
1939 43.975
2262 12.038
2585 6.614
3232 0.338
3878 0.320
86
-------
nationwide (urban and rural) mobile source exposure distribution exclusive
of the three mobile source microenvironments, for a one gram per minute
emission factor. While it is possible to combine this distribution with
the microenvironment distributions to produce a single exposure distribution
for a 1.0 g/min emission factor, such a distribution would have little use.
This is because the pollutant emission factors for each individual environ-
ment are different. Therefore, each individual exposure distribution must
be multiplied by a different emission factor to obtain the exposure in
that environment for a given pollutant.
Two different uses are examined for the exposure distributions
developed from this project. The first use is to determine the person
hours of exposure above a specified concentration of a specific pollutant.
The second use is to develop new sets of exposure distribution tables for
a specific pollutant. Ideally, a computer program should be written for
both of these problems. However, the time and effort available for this
project did not permit the development of the necessary computer program.
The paragraphs that follow explain how to manually calculate the solutions
to these problems.
Person Hours of Exposure above a Specific Concentration
The total nationwide annual person hours of exposure above a
specified mobile source pollutant concentration can be calculated using
Tables 19, 21, 23, and 35. The steps are as follows:
1. Divide the specified pollutant concentration, expressed
in yg/m3, by the 1980 FTP emission factor converted to
grams/min for the pollutant being studied. This gives
the concentration at a 1.0 g/min emission factor.
2. Enter Table 35, the urban and rural exposure, with the
new concentration from Step 1. Read the person hours
of exposure exceeding the concentration from the table;
linearly interpolating between concentration values,
if required.
3. For each of the three microenvironments, divide the
specified pollutant concentration by the 1980 emission
factor. The emission factor will probably be different
for each different microenvironment.
4. Enter the appropriate Table for each of the microenviron-
ments with the new pollutant concentration for each one
of the microenvironments. Table 19 is used for parking
garages, Table 21 for street canyons, and Table 23 for
tunnels. From each of the tables obtain a value for
person hours of exposure above the concentration. Inter-
polate, if necessary.
87
-------
5. Add the four values of person hours of exposure together
to obtain the nationwide number of person hours exposure
to the pollutant above the specified concentration.
Tables of Exposure for a Specific Pollutant
Exposure tables showing person hours of exposure above given
concentrations of a pollutant can be constructed for any mobile source
pollutant for which emission factors are known, using Tables 19, 21, 23
and 35. In fact, for a given pollutant, a single exposure table can be
constructed combining the urban plus rural exposure and the microenvironment
exposures. The procedure is as follows:
1. For Tables 19, 21, 23, and 35, multiply each concentration
reading in each table by the emission factor appropriate
to the environment for which the individual table was
constructed. This will produce four tables with four
different sets of concentration values in terms of yg/m3.
2. If units other than yg/m3 are desired (for instance, ppm)
multiply each concentration in the tables by the appro-
priate conversion factor at this time.
3. Choose a convenient set of concentration intervals that can
be used for all tables. For each table, interpolate to
find the person hours of exposure at the chosen concentra-
tion values. This will produce four tables with the same
concentration values.
4. For each concentration value, add the person hours of
exposure for each table to obtain a single table showing
the nationwide person hours of exposure for the given
pollutant.
88
-------
VIII. CONCLUSIONS AND RECOMMENDATIONS
The goal of this project was to obtain nationwide annual person hours
of exposure to any mobile source pollutant. This goal was accomplished
within the appropriate limitations and caveats mentioned previously, including
the caution which should be used when assessing pollutants with emissions
distributions markedly different from CO. This goal was accomplished.
Tables 19, 21, 23, and 35 should be used for this purpose, following the
steps shown in Section VII. Additionally, as the result of the work done
for this project, several important facts concerning the estimation of
exposure to mobile source pollutants were revealed, prompting the conclu-
sions and recommendations below.
Conclusions
Conclusions from this study are:
1. The results of NAAQS Exposure Model (NEM) study of CO
exposure do not provide a sufficient estimation of
mobile source pollutant exposure.
2. The NEM, with inputs modified from the published CO
study inputs, and a mobile source microenvironment
exposure model used together were able to provide a
reasonable estimate of exposure to any mobile source
pollutant.
3. The place specific approach used in the mobile source
microscale exposure model developed for this project
is an efficient and accurate method for exposure deter-
mination when only person hours of exposure is desired.
4. In mobile source microscale situations, CO concentra-
tions are an excellent indicator of the number of
vehicles present.
5. Neighborhood monitor concentrations will not adequately
predict mobile source pollutant concentrations within
microenvironments with a large number of mobile sources.
Recommendations
During the course of this study, a number of data deficiencies,
methodology improvements, and needed additional work were identified.
From these, a number of recommendations for future action have been
developed.
89
-------
1. To make the results of this study easier to use, a
computer program should be written to provide a single,
nationwide exposure distribution for any pollutant.
2. To improve the exposure estimates for the on-expressway
situation, the number of people in the NEM A-O groups
needs to be adjusted, as do the activity patterns of
the groups.
3. On a longer range basis, the NEM computer program should
be rewritten for nationwide mobile source exposure esti-
mates. It should be structured to utilize a nationwide
data base of CO monitor data for the mesoenvironments,
as well as national populations. The time frame would be
reduced to one quarter hour. Activity patterns would be
adjusted to account for the various mobile source micro-
environments. The exposure within the microenvironments
themselves could possibly be part of the program.
4. Additional mobile source microenvironments need to be
included in the exposure estimate. The two most significant
areas needed are the personal garage and area sources,
such as parking lots and trucking terminals.
5. The best estimate of on-expressway exposure would be
obtained by removing this microenvironment from the NEM.
It should be possible to identify some of the SAROAD
data base monitors as beside expressway monitors. These
monitors could be used to better estimate expressway
exposure. The on-expressway situation could then be
treated as any other microenvironment.
6. Additional measured CO data should be collected for parking
garages and tunnels.
7. Additional data are needed on the range and average time of
individual exposure periods for each of the microenvironments.
90
-------
REFERENCES
1. Springer, K. J., and Ingalls, M.N., "Measurement of Sulfate and Sulfur
Dioxide in Automobile Exhaust." Prepared for the Environmental Pro-
tection Agency under Contract No. 68-03-2118, EPA 460/3-76-015.
August 1976.
2. Dietzmann, H. E., Smith, L. R., Parness, M. A., and Fanick, E. R.,
"Analytical Procedures for Characterizing Unregulated Pollutant Emis-
sions from Motor Vehicles." Interim Report under Contract No. 68-02-2497
to the Environmental Protection Agency, Publication No. EPA 600/2-79-017,
February 1979.
3. Smith L. R., Parness, M. A., Fanick, E. R., and Dietzmann, H. E.,
"Analytical Procedures for Characterizing Unregulated Emissions from
Vehicles Using Middle-Distillate Fuels." Interim Report to the Environ-
mental Protection Agency under Contract No. 68-02-2497, Report No.
EPA 600/2-80-068, April 1980.
4. Springer, K. J., "Investigation of Diesel-Powered Vehicle Emissions
VII." Final Report to the Environmental Protection Agency under
Contract 68-03-2116, EPA 460/3-76-034, February 1977.
5. Springer, K. J., "Characterization of Sulfates, Odor, Smoke, POM, and
Particulates from Light and Heavy Duty Engines - Part IX." Final Report
to the Environmental Protection Agency under Contract No. 68-03-2417,
EPA 460/3-79-007, June 1979.
6. Hare, C. T., "Characterization of Gaseous and Particulate Emissions from
Light-Duty Diesels Operated on Various Fuels." Final Report to the
Environmental Protection Agency under Contract 68-03-2440, EPA 460/3-79-008,
July 1979.
7. Smith, L. R., "Characterization of Exhaust Emissions from High Mileage
Catalyst-Equipped Automobiles." Final Report to Environmental Protection
Agency under Contract 68-03-2884, Task Specifications 7 and 10,
EPA 460/3-81-024, September 1981.
8. Smith, L. R., "Unregulated Emissions for Vehicles Operated under Low
Speed Conditions." Final Report to the Environmental Protection Agency
under Contract 68-03-3073, Work Assignment 4, EPA 460/3-83-006,
May 1983.
9. Urban, C. M., "Regulated and Unregulated Exhaust Emission from Malfunctioning
Non-Catalyst and Oxidation Catalyst Gasoline Automobiles." Final Report
to the Environmental Protection Agency under Contract No. 68-03-2499.
Report No. EPA 460/3-80-003, January 1980.
91
-------
REFERENCES (Cont'd).
10. Urban, C. M., "Regulated and Unregulated Exhaust Emissions from Malfunc-
tioning Three-way Catalyst Gaseoline Automobiles." Final Report to the
Environmental Protection Agency under Contract 68-03-2588, Report No.
EPA 460/3-80-004, January 1980.
11. Urban, C. M., "Regulated and Unregulated Exhaust Emissions from A Malfunc-
tioning Three-Way Catalyst Gasoline Automobile." Final Report to the
Environmental Protection Agency under Contract 68-03-2692. Report No.
EPA 460/3-80-005, January 1980.
12. Urban, C. M., "Unregulated Exhaust Emissions from Non-Catalyst Baseline
Cars Under Malfunction Conditions." Final Report to Environmental
Protection Agency under Contract 68-03-2884, Task Specifications 4 and 5.
Report No. EPA 460/3-81-020, May 1981.
13. Hare, C. T., "Characterization of Gaseous and Particulate Emissions from
Light-Duty Diesels Operated on Various Fuels." Final Report to the
Environmental Protection Agency under Contract 68-03-2440, EPA 460/3-79-008
July 1979. '
14. Ullman, T. L., and Hare, C. T., "Emission Characterization of an Alcohol/
Diesel-Pilot Fueled Compression-Ignition Engine and Its Heavy-Duty
Diesel Counterpart." Final Report to Environmental Protection Agency
under Contract 68-03-2884, Task Specification 6. Report No. EPA 460/3-81-023,
August 1981.
15. Bykowski, B. B., "Characterization of Diesel Emissions from Operation of
a Light-Duty Diesel Vehicle on Alternate Source Diesel Fuels." Final
Report to the Environmental Protection Agency under Contract 68-03-2884,
Task Specification 3, Report No. EPA 460/3-82-002, November 1981.
16. Ullman, T. L., and Hare, C. T., "Emission Characterization of a Spark-
Ignited, Heavy-Duty, Direct-Injected Methanol Engine." Final Report to
the Enviornmental Protection Agency under Contract 68-03-3073, Work
Assignment 2, Report No. EPA 460/3-83-003, November 1982.
17. Smith, L. R., Urban. C. R., and Baines, T. M., "Characterization of
Exhuast Emissions from Methanol- and Gasoline-Fueled Automobiles."
Final Report to Environmental Protection Agency under Contracts 68-03-2884
and 68-03-3073, Task Specifications 11 and 12, and Work Assignments 1 and
3, Report EPA 460/3-82-004, August 1982.
18. Ingalls, M. N., "Estimating Mobile Source Pollutants in Microscale
Exposure Situation." Final Report to Environmental Protection Agency
under Contract 68-03-2884, Task Specification 1. Report No. EPA 460/
3-81-021, July 1981.
92
-------
REFERENCES (Cont'd).
19. Harvey/ C. A., et al "A Study of the Potential Impact of Some Unregulated
Motor Vehicle Emissions." SAE Paper 830937 to be presented at the 1983
SAE Passenger Car Meeting, Dearborn, Michigan, June 1983.
20. "Carbon Monoxide", National Research Council, Committee on Medical and
Biologic Effect of Environmental Pollutants, National Academy of Sciences,
Washington, D.C., 1977.
21. "Compilation of Air Pollutant Emission Factors: Highway Mobile Sources"
EPA 460/3-81-005, dated March 1981, U.S., EPA, Office of Mobile Source
Air Pollution Control, Emission Control Technology Division, Ann Arbor,
Michigan.
22. Larsen, R. I., "A New Mathematical Model of Air Pollutant Concentration
Averaging Time and Frequency." Journal of the Air Pollution Control
Association, Vol. 19, No. 1, January 1969.
23. Larsen, R. I., "A Mathematical Model for Relating Air Quality Measurements
to Air Quality Standards," AP-89, Environmental Protection Agency,
Research Triangle Park, N.C., November 1971.
24. Mage, D. T., and Ott, W. E., "Refinements of the Lognormal Probability
Model for Analysis of Aerometric Data," APCA Journal Vol. 28, No. 8,
August 1978.
25. Johnson, T., "A Comparison of the Two-Parameter Weibull and Lognormal
Distributions Fitted to Ambient Ozone Data." Proc. of Specialty
Conference on Quality Assurance in Air Pollution Measurement, Air Pollution
Control Association, 1979.
26. Aitchison, J., and Brown, J.A.C., The Lognormal Distribution, Cambridge
University Press, New Youk, 1957.
27. Kama, G. M., et al, "Air Flow Requirement for Underground Parking Garages,"
American Industrial Hygiene Journal, December 1961, pp 462-470.
28. Barker, I. W., and Fox, M. F., "Vehicular Pollution in Car Parks," Royal
Society of Health Hournal, Vol. 96.
29. Glazer, N., "The Regulation and Control of Carbon Monoxide in Enclosed
Parking Garages." APCA Paper 78-4.6, presented at the 71st Annual
Meeting of APCA, June 1978.
30. Ayres & Hayakawa, Consulting Engineers, "Project Program Plan and Statement
of Probable Costs. Music Center and Mall Garages. Phase I and III."
Submitted to County and Los Angeles for Capitol Projects No. 7085.09,
7065.21 amd 7065.22, May 1975.
93
-------
REFERENCES (Cont'd).
31. Mage, D. T., "Frequency Distributions of Hourly Wind Speed Measurements,"
Atmospheric Environment, Vol. 14, pp 367-374, 1980.
32. "Airport Climatological Summary." Climatography of United States No. 90,
for various airports. National Climatic Center, National Oceanic and
Atmospheric Administration, Asheville, N.C.
33. Johnson, T., and Paul, R. A., "The NAAQS Exposure Model (NEM) Applied to
Carbon Monoxide." Draft Final Report by PEDCo Environmental, Inc. for
EPA, Contract 68-02-3390, Work Assignments 13 and 16, April 12, 1982.
34. Brice, R. M., and Roesler, J. F., "The Exposure to Carbon Monoxide of
Occupants of Vehicles Moving in Heavy Traffic." JAPCA Vol. 16,
No. 11,.November 1966.
35. Peterson, G. A., and Sabersky, "Measurement of Pollutants Inside an
Automobile", JAPCA, Vol025, No. 10, October 1975.
36. Chaney, L. W., "Carbon Monoxide Automobile Emissions Measured from the
Interior of a Traveling Automobile." Science, Vol. 199, 17 March 1978.
37. Colwill, D. M., and Hickman, A. J., "Exposure of Drivers to Carbon Monoxide "
JAPCA, Vol. 30, No. 12, December 1980. '
38. Peterson, W. B., and Allen, R., "Carbon Monoxide Exposures to Los Angeles
Area Commuters," JAPCA, Vol. 32, No. 8, August 1982.
39. Rodgers, S. J., et al, "Tunnel Ventilation and Air Pollution Treatment,"
Prepared for FHWA, Office of Research by Mine Safety Appliance Research
Crop. Report FHWA-RD-72-15, NTIS No. PB210-360, June 30, 1970.
40. "Study of Air Pollution Aspects of Various Roadway Configurations," Final
Report to the New York City Department of Air Resources under Contract
• 209624 by the General Electric Company, 3198 Chestnut Street, Philadepphia,
PA., dated September 1, 1971.
41. Sosslau, A. B., and Hassom, A. B., "Quick-Repsonse Urban Travel Estimation
Techniques and Transferrable Parameters,',1 National Cooperative Highway
Research Program Report 187. Transportation Research Board, National
Research Council, Washington, D.C., 1978.
42. "Short-Range Urban Transit Study, San Antonio, Texas," prepared for the
San Antonio Transit System and City of San Antonio by Wilbur Smith and
Associates, Houston, Texas, 1972.
94
-------
REFERENCES (Cont'd).
43. Thayer, S. D., "Vehicle Behavior In and Around Complex Sources and
Related Complex Sources Characteristics, Vol IV—Parking Facilities,"
Final Report prepared by Geomet, Inc., Rockville, MD, under Contract
No. 68-02-1094 Task Orders to the U.S. Environmental Protection Agency,
Research Triangle Park, N.C., Publication No. EPA-460/3-74-003d,
October 1973.
44. Forrest, L. "Assessment of Environmental Impacts of Light Duty Vehicle
Dieselization," Aerospace Report No. ATR-79(7740)-!. Draft Final Report
by Aerospace Corporation under Contract DOT-TSC-1530, March 1979.
45. Homburger, W. S., editor, Transportation and Traffic Engineering Handbook,
second edition, Prentice Hall, Inc. Englewood Cliffs, N.J.
46. "Highway Capacity Manual," Highway Research Board Special Report 87.
Highway Research Board, National Research Council, Washington, D.C. 1965.
47. Matson, T. M., Smith, W. S., and Kurd, F. W., Traffic Engineering,
McGraw-Hill Book Company, Inc., New York, N.Y., 1955.
48. Curtin, J. F., "Traffic-Transit-Parking in Downtown Rochester: How to
1975." Contained in Highway Research Board Bulletin 293, Highway
Research Board, National Research Council, 1961.
49. "San Antonio-Bexar County Urban Transportation Study Report 6A and 6B -
Origin-Destination Survey 1968," Texas Highway Department, Austin, Texas.
50. Pushkarev, B. and Zupan, J. "Pedestrian Travel Demand;" Highway Research
Record No. 355, Highway Research Board, National Academy of Sciences,
Washington, D.C.,1971.
51. Cameron, R. M., "Mechanical Measurements of Pedestrian Volumes." Trans-
portation Research Record No. 498. Transportation Research Board,
National Academy of Sciences, Washington, D.C., 1974.
52. Rutherford, G. S., and Schofer, J. L., "Analysis of Some Characteristics
of Pedestrian Travel," Transportation Research Record No. 605. Transpor-
tation Research Board, National Academy of Sciences, Washington, D.C., 1976.
53. Teague, D. M., "Los Angeles Traffic Pattern Survey" SAE Paper 171 pub-
lished in "Vehicle Emissions, Part I", SAE Progress in Technology
Series, Vol. 6, Society of Automotive Engineers, Inc. pp 17-38, 1964.
54. Kruse, R. E., and Huls, L. A., "Development of the Federal Urban Driving
Schedule," SAE Paper 730553 presented at Detroit, MI, Society of Automotive
Engineer, Warrensdale, PA, May 1973.
95
-------
REFERENCES (Cont'd).
55. Papette, B. and Horowitz, J. "Stochastic Model of Worst Case Exposures
to Sulfuric Acid from Catalyst-Equipped Vehicles." U.S. Enviornmental
Protection Agency in-house (unpublished) report, November 1975.
56. Schlaug, R. N., and Carlin, T. J., "Aerodynamics and Air Quality Management
of Highway Tunnels." Final Report for Contract DOT-FH-11-8538 to Science
Applications, Inc. Report FHWA-RD-78-185, January 1980.
57. Johnson, T., and Paul, R. A., "The NAAQS Model (NEM) Applied to Carbon
Monoxide." Draft Final Report by PEDCo Environmental, Inc. for EPA
Contract 68-02-3390, Work Assignments 13 and 16, December 1982.
58. Meyer, J.R., Kain, J.F., and Wohl, M., The Urban Transportation Problem.
Harvard University Press, Cambridge, MA., 1965.
96
-------
APPENDIX A
Development of Lognormal Pollutant Distributions for Parking Garages
and Roadway Tunnels
-------
The Lognormal Distribution
The lognormal distribution has historically been used to describe
ambient air concentration distributions. For areas where the pollutant
concentration distribution is not completely defined by measured-values,
but values such as the mean, median, mode or range are known, the properties
of the lognormal distribution can be used to define the complete pollutant
distribution.
The lognormal distribution is defined as:
df (x) =
exp
2 a
2
(in x - y)
dx
Obviously, if a and y are known, the distribution can be defined for any
value of X.
Additionally for a lognormal distribution:
Mean, X = e (y + 1/2 a2)
Median = e^1
Mode = eM~a
The coefficient of variation n, defined as the standard deviation divided
by the mean is:
For the parking garage, the concentration distributions were developed
by using a CO concentration value known to be typical of parking garages
as the distribution mode, then adjusting the median so that the frequency
of occurence in the 300 to 400 ppm CO range was 0.01 percent. If the median
and mode are known, then y and a are solved for as follows:
In (median)= In (e^)
y = In (median)
In (mode) = In (ey~a'')
y - a2 = In (mode)
a = yln (mode) - y
With y and a defined, the entire lognormal distirbution can be defined.
For roadway tunnels a slightly different approach was used. The
measured data on concentrations in roadway tunnels produced a mean value
and a maximum value of CO concentration as a function of average daily
traffic (ADT).
A-2
-------
Using the proper CO emission factor, these CO levels were converted to
pollutant concentrations for a 1.0 g/minute emission factor (see text).
The CO concentrations and pollutant concentrations at 1.0 g/minute for
several ADT values are shown in the table below.
Pollutant Concentration
yq/m3 at 1.0 g/min(a)
ADT Mean Max. Mean Max.
0.5 5.2 66.7 166.2 2131.8
1.5 14.3 94.4 457.0 3017.2
2.5 23.4 122.1 747.9 3902.5
3.5 32.5 149.8 1038.7 4787.8
4.5 41.6 177.4 1329.6 5669.9
5.5 50.7 205.1 1620.4 6555.3
6.5 59.9 232.8 1914.5 7440.6
based on a CO emission factor of 62.06 g/mile at 35 mph
Assuming that the pollutant concentration is lognormally distributed,
the problem is to define the distribution for each value of ADT, given only
the mean and maximum concentrations. Note that if the minimum concentration
is assumed to be 0 then the range (maximum - minimum) is also known. The
expression for the mean of a lognormal distribution is shown above. It can
also be shown that the standard deviation, S, can be approximated by the
following relationship, assuming that the range available is the result of
a large number of observations(2)
S =
Thus the standard deviation for the tunnel distribution can be estimated
from the range. If the standard deviation and mean are known the coeffi-
cient of variation is known and a can be obtained as follows:
n2 = e°2 - 1
e°2 = n2 + 1
a2 = In (n2 + 1)
= Vln(n2
a = Vln (n" + 1)
A-3
-------
Once a is known, y can be obtained from the equation for the
mean:
X = e
(y + 1/2
M + 1/2 a2 = In X
y = in x - 1/2 a2
With both y and O known the distribution is fully defined. The
table below shows the y and a for the tunnel pollutant concentration dis-
tributions at seven different values of ADT.
APT
0.5
1.5
2.5
3.5
4.5
5.5
6.5
range
(max-min)
2131.8
2017.2
3902.5
4787.8
5660.0
6555.3
7440.6
std. dev.
range
S = 6.5
327.97
464.18
600.38
736.58
872.29
1008.51
1144.71
mean
= X
166.2
457.0
747.0
1038.7
1329.6
1620.4
1914.5
n
= S
X
1.97
1.02
0.80
0.71
0.66
0.62
0.60
a
1.26
0.84
0.71
0.64
0.60
0.57
0.55
y
4.32
5.77
6.37
6.74
7.01
7.23
7.40
References:
1. Aitchison, J. and Brown, J.A.C., The Lognormal Distribution,
Cambridge University Press, New York, 1957.
2. Sokal, R. and Rohof, T.J., Biometry, W. H. Freeman and Co., 1981,
A-4
-------
APPENDIX B
Fortran Listing of SAROAD File Editing Program for
Street Canyon Monitors
(written for CDC Cyber 173)
-------
GO TO 75
190 PRINT 2040, I STACK),IMON(K),IDAY(K)
200 60 TO 75
220 PRINT 2050, ISTA(1),IMON(1),IDAY(1),IMON(2),IOAY(2),ISTA(2)
CALL DUMREC( OUM,IDUML,IDUM4.K)
K=1
WRITE(2, 3000) NARI(K), ISTA(K),NAR2(K),NAR3(K),IMON(K), |OAY(K),
1 IDUML,IDUM4,ITYPE(K),IUNT(K),IDEC(K),(DUM(M),M=1,12)
NAR1(1)=NAR1(2)
NAR2O)=NAR2(2)
NAR3(1)=NAR3(2)
NAR4(1)»NAR4(2)
IMON(1)=IMON(2)
IDAY(1)=IDAY(2)
ITYPE(1)=ITYPE(2)
ISTA(1)*ISTA(2)
IUNT(1)=IUNT(2)
IDEC(1)=IDEC(2)
ILINEd ) = l LINE(2)
00 225 M-1,12
225 RDG(M,1)=RDG(M,2)
I =IMON(K)
J»IDAY(K)
GO TO 70
1000 FORMAT(I 3,I 7,A3,I 3,I 2,I 2,I 1,16,I 2,I 2,I1,12F4.0)
2000 FORMAT(1X,«FOR STATION ",17," SOMETHING IS WRONG AT »,I2,M/»,I2,
i 1 "LINE 2")
f° 2010 FORMAT(IX,"FOR STATION ",I7,« DATA IS MISSING FOR MONTH »,I2,
1 ". NEXT MONTH WITH DATA IS MONTH ",12)
2020 FORMAT(IX,"FOR STATION ",17," DATA IS MISSING FOR MONTH »,I2,
1 " DAY ",12,". NEXT DAY WITH DATA IS DAY " 12)
2030 FORMAT(1X,«FOR STATION ",17," THERE IS NO FIRST LINE OF DATA FOR "
1 ,I2,"/",I2)
2040 FORMAT(1X,«FOR STATION ",17," SOMETHING IS WRONG AT ",I2,"/",I2,
1 "LINE 1")
2050 FORMAT(IX,"FOR STATION ",17," THERE IS NO SECOND LINE OF DATA FOR»
1 1X,I2,"/"I2,". NEXT DATA IS ",I 2,"/",I 2," STATION ",17)
3000 FORMAT(I3,I7,A3,I 3,I 2,I 2,I 1,16,I 2,I 2,I1,12F4.1)
999 STOP
END
SUBROUTINE DAYWEK(IMON,(DAY,ITYPE)
COMMON/SATOA /ISAT(52)/SUNDA /ISUN(52)
IDATE «(IMON »100) + IDAY
ITYPE " 1
00 100 L - 1,52
IFdDATE.EQ.ISAT(L)) ITYPE * 2
IF(IDATE.EQ.ISUN(L)) ITYPE = 3
100 CONTINUE
RETURN
END
SUBROUTINE DUMREC( DUM,IDUML,IDUM4.K)
DIMENSION DUH(12)
DO 200 I» 1,12
DUM(I) * 99.9
200 CONTINUE
IFCK.EQ.1) IDUML - 0
-------
PROGRAM EOTEPA( INPUT. OUTPUT, EDDAT,TAPE2»EDDAT,TAPE60=I NPUT )
DIMENSION IMON(2),IOAY(2),ILINE(2),ITYPE(2), RDG( 12,2) , I UNT<2)
DIMENSION NAR1(2),NAR2(2),NAR3(2),NAR4(2), I STA(2),DUM( 12) . IDEC(2)
ISTA(2) = 1960085
50 READ 1000,NAR1(K), I STA(K),NAR2(K),NAR3(K), IMON(K), IDAY(K) ,
II LINE (K), NAR4
-------
IF(K.EQ.I) IDUM4 » 042101
IF(K.EQ.2) IDUML « 1
IF(K.EQ.2) IDUM4 « 242101
RETURN
END
SLOCK DATA SATDATE
COMMON/SATDA/1 SAT(52)
DATA (ISAT(I),l»1,52)/
1 0103,0110,0117.0124,0131,0207,0214,
2 0221,0228,0307,0314,0321,0328,0404,
3 0411,0418,0425,0502,0509,0516,0523,
4 0530,0606,0613,0620,0627,0704,0711,
5 0718,0725,0801,0808,0815,0822,0829,
6 0905,0912,0919,0926,1003,1010,1017,
7 1024,1031,1107,1114,1121,1128,1205,
8 1212,1219,1226/
END
BLOCK DATA SUNOATE
COMMON/SUNDA/1 SUN(52)
DATA (ISUNO ),l*1,52)/
1 0104,0111,0118,0125,0201,0208,0215,
2 0222,0301,0308,0315,0322,0329,0405,
3 0412,0419,0426,0503,0510,0517,0524,
4 0531,0607,0614,0621,0628,0705,0712,
5 0719,0726,0802,0809,0816,0823,0830,
6 0906,0913,0920,0927,1004,1011,1018,
7 1025,1101,1108,1115,1122,1129,1206,
8 1213.1220.1227/
END
-------
APPENDIX C
Fortran Listing of Microenvironment Exposure
Model for Parking Garages
(written for CDC Cyber 173)
-------
PROGRAM PHD 1ST 73/74 OPT»1 FTN 4.8+552 83/05/21. 16.57.26 PAGE
1 PROGRAM PHD I ST( INPUT, OUTPUT)
REAL NPC
COMMON/POP/ P(24,5) /CONC / FRAC(23.9)
DIMENSION R(2),FTT(3),SUM(23), ICON 1 (23) , ICON2C23) , ITI TLE (8) ,X<23)
5 DATA ICON1 /O, 361 ,464, 61 9, 774, 1031 , 1289, 1 547, 1805, 2062. 2320,2578,
1 3001,4001,5001,6001,8001,10001,15001,20001,25001,
2 3000 1,4000 1/
DATA ICON2 /360, 463, 618, 773, 1030, 1288, 1 546, 1804, 206 1 , 231 9,2577,
1 3000,4000,5000,6000,8000,10000,15000,20000,25000,
10 2 30000, 40000, 50000/
C**»* READ INPUT VALUES
READ 1000, ITITLE
READ * , TOTCAR
15 DO 50 1-1,23
SUM (I )*0.0
50 CONTINUE
DO 900 JDAY - 1,5
GO TO ( 100,120,140,145,150) JDAY
20 C*«* JDAY*1 IS WEEKDAYS
100 H»248
NPC-1.4
GO TO 160
C**» JDAY-2 IS SAT. + SOME HOLIDAYS
25 120 H=62
NPC-2.3
GO TO 160
O C* JDAY=3 IS SUN. + 3 HOLIDAYS
' 140 H=55
30 NPC=2.3
GO TO 160
C**«» JDAYS= 4 IS FOR WORKER RELATED PEAK HOUR
145 H* 248.
NPC-1.4
35 TOTCAR«.25»TOTCAR
GO TO 160
C**»* JDAYS= 5 IS FOR ENTERTAINMENT RELATED PEAK
150 H=» 62
NPC«2 3
40 160 CONTINUE
DO 800 IHR= 1,24
PTC- 1.0
IFUDAY.EQ. 1.AND.IHR.E0.17) PTC=0.75
IF(JDAY.EQ.3.AND.IHR.Ep.22) PTC=0.75
45 IF(JDAY.GT.3.AND.IHR.GT.1) GO TO 800
170 IFUDAY.LE.3) GO TO 180
N-7
GO TO 240
180 IF«P(|HR,JDAY)/12.).GT.6) GO TO 200
50 N-1
GO TO 240
200 IF((P(IHR,JDAY)/12.).GT.14) GO TO 220
N-4
GO TO 240
55 220 N-7
240 CONTINUE
M-N+2
-------
PROGRAM PHDIST 73/74 OPT-1 FTN 4.8+552 83/05/21. 16.57.26 PAGE
DO 700 LWS = N,M
IF(LMS.EQ.N) WF-0.095
60 IF(LWS.EQ.N-H) WF-0.65
IFUWS.EQ.N+2) WF=0.255
DO 600 KCON =1,23
SUM(KCON) = SUM(KCON) + (FRAC(KCON,LWS) *PTC* TOTCAR»(P(IHR,JDAY)
1 /4.) *NPC * H * WF)
65 600 CONTINUE
700 CONTINUE
800 CONTINUE
900 CONTINUE
PRINT 2000 , ITITLE
70 DO 940 I* 1,23
940 PRINT 2010 ,ICON 1(I),ICON2(I),SUM(I)
X(23)=SUM(23)/1E6
DO 960 1=2,23
J = l-1
75 XC24-I )=X(24-J)-KSUM(24-I )/1E6)
960 CONTINUE
PRINT 2020,ITITLE
PRINT 2040,ICON1(1),X(1)
DO 980 1=2,23
80 980 PRINT 2040,ICON2(I-1),X(I)
2020 FORMAT(1H1,//,26X,8A10,////,17X,"CONCENTRATION EXCEEDED*,
I 5X,*PERSON HOURS*,/)
2040 FORMAT(20X,6X,I5,5X,10X,F9.3)
1000 FORMAT (8A10)
85 2000 FORMAT (1H1,//,26X,8A10,////,17X,"CONCENTRATION INTERVAL*,
1 5X,*PERSON HOURS*,/)
2010 FORMAT (20X,I5,* TO *,I 5,10X,F11.0)
999 STOP
END
-------
BLOCK DATA CONCEN 73/74 OPT=1 FTN 4.8+552 83/05/21. 16.57.26 PAGE
1 BLOCK DATA CONCEN
COMMON/CONC/P6FRAC(23,9)
C»*» ORDER OF DATA IS 3 PCT,FOR 1.5,7,14 KTS WIND,THEN 9+19 PCT.
DATA((PSFRAC(I
10
15
20
25
n
i
A
B
C
D
E
F
6
H
1
J
K
.000,
.000,
.002,
.005,
.020,
.037,
.052,
.063,
.069,
.071,
.070,
.148,
.120,
.180,
.150,
.170,
.097,
.053,
.030,
.017,
.010,
.006,
DATA((P6FRAC(I,
L
M
N
0
P
0
R
S
T
U
V
w
.107,
.197,
.123,
.073,
.069,
.024,
.015,
.001,
.000,
.000,
.000,
.000,
END
.005,
.005,
.001,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.566,
.157,
.138,
.068,
.047,
.016,
.005,
.002,
.000,
.000,
.000,
J),J=1,9)
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.001,
.002,
.003,
.004,
.006,
,1=12,
.013,
.040,
.070,
.081,
.168,
.147,
.241,
.113,
.054,
.025,
.021,
.006,
.005,
.005,
.018,
.031,
.076,
.096,
.102,
.097,
.088,
.077,
.066,
23)/
.087,
.127,
.061,
.030,
.024,
.007,
.003,
.000,
.000,
.000,
.000,
.000,
.044,
.054,
.110,
.121,
.185,
.143,
.102,
.071,
.049,
.034,
.023,
.024,
.027,
.007,
.002,
.001,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.003,
.006,
.011,
.040,
.061,
.201,
.193,
.149,
.106,
.123,
.058,
.000,
.000,
.000,
.001,
.006,
.013,
.022,
.030,
.037,
.043,
.047,
.080,
.184,
.151,
.112,
.137,
.067,
.057,
.011,
.003,
.000,
.000,
.000,
.003,
.004,
.013,
.025,
.066,
.089,
.098,
.097,
.090,
.084,
.070/
.093,
.137,
.067,
.033,
.026,
.007,
.004,
.000,
.000,
.000,
.000,
.OOO/
-------
BLOCK DATA POPPG
73/74 OPT=1
FTN 4.8+552
83/05/21. 16.57.26
PAGE
n
i
Ul
10
15
20
25
30
BLO K DATA POPPG
COMMON/POP/PGP(24,5)
C««» DATA ORDER IS WEEKDAY,
DATA ((P6P(I,J),J=1,5),
SAT., SUN
A
B
C
D
E
F
G
H
1
J
K
L
DATA
M
N
0
P
0
R
S
T
U
V
w
X
END
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0099,
.0693,
.1284,
.1943,
.2134,
.1750,
( (PGPd.J)
.1729,
.1819,
.1678,
.1902,
.1960,
.1836,
.1033,
.0829,
.0723,
.0409,
.0449,
.0251,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0040,
. 0060 ,
.0340,
.0600,
.0820,
.1060,
,J=1,5)
.1340,
.1610,
.1800,
.1930,
.1980,
.2000,
.2000,
.1980,
.1750,
.1200,
.0540,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0125,
.0375,
.0500,
.0500,
.0500,
0.750,
0.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
0.750,
0.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.ooo/
,l=12,24)/
.0500,
.0500,
.0500,
.0500,
.0375,
.0125,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.000,
.ooo/
M TO
TO
TO
TO
TO
TO
TO
1
2
3
4
5
6
7 TO 8
8 TO 9
9 TO 10
10 TO 11
11 TO 12
12 TO
13 TO
TO
14
15 TO
16 TO
17 TO
18 TO
19 TO
20 TO
21 TO 22
22 TO 23
23 TO 24
13
14
15
16
17
18
19
20
21
-------
APPENDIX D
Fortran Listing of Microenvironment Exposure Model
for Street Canyons
(written for CDC Cyber 173)
-------
PROGRAM SCOIST(INPUT,OUTPUT)
COMMON/SPOP/ P(24.6) /SCONC / FRACd9,6,3)
DIMENSION TOTP(3),TOTPD(3),SUM(19),ICONK19>,ICON2M9),mTLE<8),
1 X(19)
DATA ICON1 /O,36,106,176,246,3 16,387,457,527,597,667,737,807,877,
1 948,1088,1298,1509,1719/
DATA ICON2 /35,105,175,245,315,386,456,526,596,666,736,806,876,
1 947, 1087,1297,1508,1718,20697
C«*»« READ INPUT VALUES
READ 1000, ITITLE
READ » ,(TOTP(I ), 1=1,3), (TOTPDd ), 1=1,3)
C*»»»
DO 50 1=1,19
SUM(I)=0.0
50 CONTINUE
DO 900 JDAY =1,3
60 TO ( 100,120,140) JDAY
C*«» JDAY=1 IS WEEKDAYS
100 H=248
GO TO 160
C**» JDAY=2 IS SAT. + SOME HOLIDAYS
120 H»62
GO TO 160
C» JDAY=3 IS SUN. + 3 HOLIDAYS
140 H=55
GO TO 160
160 CONTINUE
00 800 IHR= 1,24
GO TO (200,220,240) JDAY
200 IF(IHR.LE.6) L-1
IFdHR.EQ.7) L-2
IF(IHR.GE.8.AND.IHR.LE.9) L-3
IF(IHR.GE.10.AND.IHR.LE.15) L=4
IF(IHR.GE.16.AND.IHR.LE.18) L-5
IFdHR.6E.19) L«6
GO TO 280
220 IFdHR.LE.3) L-1
IFdHR.GE.4.AND.IHR.LE.6) L-2
IF(IHR.GE.7.AND.IHR.LE.8) L-3
IFdHR.GE.9.AND. IHR.LE. 12) L-4
IFdHR.GE.13.AND.IHR.LE.18) L=5
IFdHR.GE.19) L-6
GO TO 280
240 IFdHR.LE.2) L-1
IFdHR.GE.3.AND. IHR.LE.4) L*2
IFdHR.GE.5.AND.IHR.LE.10) L-3
IFdHR.GE.11.AND.IHR.LE.14) L«4
IFdHR.GE.15.AND.IHR.LE.23) L=5
IFdHR.GE.24) L«6
280 CONTINUE
DO 600 KCON =1,19
SUM(KCON) » SUM(KCON) + (FRAC(KCON,L,JDAY) »(((TOTP(JDAY)*P(IHR,
1 JDAY)) +(TOTPD(JDAY)*P(IHR,JDAY+3)))/4) * H)
600 CONTINUE
800 CONTINUE
900 CONTINUE
-------
•1,3), (TOTPD(I),I-1.3)
PRINT 2000 , ITITLE
PRINT 2005, (TOTP(I).I
PRINT 2007
DO 940 1-1,19
940 PRINT 2010 ,I CON 1(I),ICON2(I>,SUM(I )
X(19)»SUM(19)/1E6
00 960 1-2,19
J = l-1
X(20-l )-X(20-J)-HSUM(20-l )/1E6)
960 CONTINUE
PRINT 2000,I TITLE
PRINT 2005, (TOTP(I),1-1.3), (TOTPDCI),I-1,3)
PRINT 2020
PRINT 2040,ICON1(1),X(1)
DO 980 1=2.19
980 PRINT 2040,ICON2(I-1),X(I)
2020 FORMAT(17X,"CONCENTRATION EXCEEDED",
1 5X,"PERSON HOURS",/)
2040 FORMAT(20X,6X,I5,5X,10X.F9.3)
1000 FORMAT (8A10)
2000 FORMAT (1H1.//.26X.8A10,////)
2005 FORMATOOX.6F12.0,///)
2007 FORMAT(17X,"CONCENTRATION INTERVAL",5X,"PERSON HOURS*,/)
2010 FORMAT (20X.I5,* TO *,I 5,10X,F11.0)
999 STOP
END
BLOCK DATA SCCONC
COMMON/SCONC/SCFRAC(19,6,3)
C""» ORDER OF DATA IS WEEKDAYS BY HOUR GROUP THEN SAT. AND SUN.
DATAC
A
B
C
0
E
f
G
H
1
J
K
L
M
N
0
P
Q
R
S
(SCFRACd.J,
.3358,
.4051,
.1435,
.0631.
.0271,
.0130,
.0060,
.0034.
.0015,
.0007,
.0004,
.0002,
.0002,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
D.J-1.
.0621,
.2858,
.2568,
.1750,
.1000,
.0541,
.0300,
.0139,
.0098,
.0057,
.0035,
.0010,
.0012,
.0004,
.0004.
.0002,
.0000,
.0000,
.0000,
6), 1-1, 19)/
.0191,
.1137,
.1862,
.1818,
.1505,
.1049,
.0811,
.0603,
.0326,
.0239,
.0163,
.0104,
.0066,
.0035,
.0045,
.0030,
.0007,
.0003,
.0002,
.0223,
.1292,
.1988,
.1800,
.1427,
.1039,
.0757,
.0549,
.0350,
.0236,
.0142,
.0076,
.0051,
.0032,
.0029,
.0009,
.0001,
.0000,
.0000,
.0182,
.1005,
.1616,
.1673,
.1421,
.1133,
.0934,
.0675,
.0470,
.0324,
.0216,
.0128,
.0084,
.0053,
.0050,
.0024,
.0009,
.0001,
.0000,
.0923,
.2954,
.2537,
.1550,
.0885,
.0477,
.0275,
.0166,
.0095,
.0056,
.0026,
.0018.
.0010,
.0009,
.0012,
.0004,
.0002,
.0000,
.oooo/
DATA((SCFRAC(I,J,2),J-1,6), 1-1
A
B
C
D
E
F
.1987,
.3246,
.1946,
.1162,
.0650,
.0372.
.3137,
.4096,
.1604,
.0618,
.0331,
.0114,
.1647,
.4222,
.2339,
.1079,
.0457,
.0141,
.0839,
.3320,
.2734,
.1564,
.0829,
.0424,
.0613.
.2998,
.2622,
.1569,
.0947,
.0548,
.0992,
.3070,
.2311,
.1572,
.0955,
.0462,
WKD
NKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
WKD
SAT
SAT
SAT
SAT
SAT
SAT
-------
G
H
1
J
K
L
M
N
0
P
P
R
S
DATA(
A
B
C
D
E
F
G
H
1
J
K
D L
l M
*" N
0
P
Q
R
S
END
BLOCK
.0261,
.0168,
.0101,
.0050,
.0034,
.0013,
.0007,
.0003,
.0000,
.0000,
.0000,
.0000,
.0000,
(SCFRAC(I,J,
.1851,
.3190,
.2031,
.1149,
.0764,
.0308,
.0338,
.0144,
.0108,
.0041,
.0015,
.0021,
.0021,
.0000,
.0005,
.0010,
.0005,
.0000,
.0000,
DATA POPSC
.0057,
.0030,
.0007,
.0007,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
. 0000 ,
.0000,
.0000,
3),J=1,6>
.2718,
.3885,
.1618,
.0830,
.0384,
.0197,
.0166,
.0093,
.0041,
. 003 1 ,
.0000,
.0016,
.0000,
.0010,
.0010,
.0000,
.0000,
.0000,
.0000,
.0070,
.0030,
.0010,
.0005,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
.0000,
,1=1,19
.2873,
.4847,
.1509,
.0489,
.0143,
.0091,
.0017,
.0012,
.0005,
.0003,
.0003,
.0002,
.0005,
.0000,
.0002,
.0000,
.0000,
.0000,
.0000,
.0149,
.0078,
.0038,
.0015,
.0003,
.0003,
.0003,
.0000,
.0000,
.0000,
.0003,
.0000,
.0000,
)/
.1847,
.4672,
.1976,
.0930,
.0299,
.0167,
.0048,
.0025,
.0020,
.0010,
.0005,
.0000,
.0000,
.0000,
.0000,
.0325,
.0167,
.0098,
.0056,
.0027,
.0010,
.0008,
.0005,
.0002,
.0005,
.0000,
.0000,
.0000,
.1484,
.3946,
.2186,
.1160,
.0574,
.0296,
.0154,
.0089,
.0043,
.0028,
.0026,
.0011,
.0001,
.0002,
.0000,
.0238,
.0157,
.0116,
.0062,
.0034,
.0019,
.0005,
.0002,
.0002,
.0005,
.0000,
.0000,
.00007
.2064,
.4087,
.1815,
.1120,
.0477,
.0249,
.0073,
.0062,
.0010,
.0021,
.0010,
.0010,
.0000,
.0000,
.0000,
.0000, .0000, .0000,
.0000, .0000, .0000,
.0000,
.0000, .0000,
.0000, .0000, .00007
COMMON/SPOP/PSC(24.6)
C«*» DATA
DATA
A
B
C
0
E
F
G
H
1
J
K
L
DATA
M
N
0
P
0
ORDER 1 S WEEKDAY,
((PSC(I,J),J
.0117
.0079
.0050
.0050
.0067
.0158
.0388
.0765
.0650
.0565
.0571
.0598
«PSC(I,J),
.0580
.0580
.0601
.0674
•1,6), 1
.0444
.0399
.0334
.0222
.0161
.0165
.0238
.0315
.0370
.0407
.0447
.0479
•1,6). 1
.0492
.0513
.0538
.0538
.0840 , .0524
SAT., SUN
•
t
t
9
»
t
9
9
9
t
9
9
9
9
9
9
9
9
1 , 12)/
.0663
.0572
.0472
.0319
.0219
.0179
.0211
.0249
.0247
.0281
.0305
.0345
12, 24)/
.0390
.0419
.0481
.0501
.0536
.004
.002
.001
.001
.001
.003
.008
.030
.059
.054
.058
.075
.126
.115
.090
.075
.085
.0046
.0023
.0012
.0000
.0000
.0012
.0023
.0046
.0185
.0323
.0611
.0934
.1130
.1280
.1349
.1153
.1038
.0046
.0023
.0012
.0000
.0000
.0012
.0023
.0046
.0185
.0323
.0611
.0934 /
.1 130 ,
.1280 ,
.1349 ,
.1153 ,
.1038 ,
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SAT
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
SUN
M
1
2
3
4
5
&
7
8
9
10
1 1
12
13
14
15
16
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
-------
R
S
T
U
V
w
X
END
0675
0472
0417
0338
0317
0246
0190
.0500
.0455
.0466
.0463
.0494
.0530
.0506
.0555
.0554
.0550
.0538
.0509
.0484
.0422
.090
.044
.026
.018
.014
.010
.008
.0669
.0369
.0219
.0185
.0150
.0150
.0092
.0669
.0369
.0219
.0185
.0150
.0150
.0092 /
17 TO 18
18 TO 19
19 TO 20
20 TO 21
21 TO 22
22 TO 23
23 TO 24
o
I
Ln
-------
APPENDIX E
Fortran Listing of Microenvironment Exposure Model
for Tunnels
(written for CDC Cyber 173)
-------
PROGRAM TUDIST(INPUT,OUTPUT)
COMMON/TPOP/ P(24,3) /TCONC / FRAC(19,7)
DIMENSION TOTP(3),AP(24,3),SUM(19),ICON1(19),ICON2(19),ITlTLE(8),
1 X(19)
DATA ICON1 70,201,401,601,701,801,901,1001,1201, 1401,
1 1601,1801,2001,220t,2401,2601,3001,4001,60017
DATA ICON2 7200,300,600,700,800,900,1000,1200,1400,1600,1800,
1 2000,2200,2400,2600,3000,4000,6000,80007
C»**» READ INPUT VALUES
READ 1000, ITITLE
READ * ,(TOTP(I),I=1,3)
C**»»
DO 50 1=1,19
SUM(I)=0.0
50 CONTINUE
DO 900 JDAY =1,3
GO TO ( 100,120,140) JDAY
C«*» JDAY=1 IS WEEKDAYS
100 H=248
GO TO 160
C*»« JDAY=2 IS SAT. + SOME HOLIDAYS
120 H=62
GO TO 160
C* JDAY=3 IS SUN. + 3 HOLIDAYS
140 H=55
w GO TO 160
I 160 CONTINUE
10 DO 800 IHR* 1,24
IF(JDAY.EQ.I) AP(IHR.I) = P(IHR,1)
IFUDAY.EQ.2) AP(IHR,2) = P(IHR,2)*(TOTP(2)/TOTP(1))
IFUDAY.EQ.3) AP(IHR,3) = P( I HR,3)»(TOTP(3)/TOTP( 1 ) )
IF(AP(IHR,JDAY).LE.0.01) L=1
IF(AP(IHR,JDAY).GT.0.01.AND.AP(IHR,JDAY).LE.0.02) L=2
IF(AP(IHR,JDAY).GT.0.02.AND.AP(IHR,JDAY).LE.0.03) L=3
IF(AP
-------
980 PRINT 2040,ICON2(I-1),X
-------
R
S
T
U
V
w
X
END
.0634
.0599
.0576
.0461
.0417
.0375
.0299
.0560
.0583
.0635
.0562
.0497
.0456
.0424
.0560
.0583
.0635
.0562
.0497
.0456
.0424 /
17 TO 18
18 TO 19
19 TO 20
20 TO 21
21 TO 22
22 TO 23
23 TO 24
M
-------
APPENDIX F
UNIVAC 1100 Runstreams for NEM Reruns
-------
Runstream to Produce the NEM Air Quality Input
File for Chicaco
«RUN,R/R AIRDAT/80,l«iMMiM,EXES/MMSAP, 120,9999
«DELETE,C EXES*PRTAg.
|DELETE,C EXES«AQCHIC.
ICOND
ICAT.P PRTAQ.,F33///140
«US*ER.BK1,A PRTAQ.
IASG.CP EXES*AQCHIC.
fASG.T AQFILE.
fASG.A SASO*CO-TRACK.
|COPY,A SASD*EXP-ABS.TRACK-11
|SETC,0
fX(JT SASD*EXP-ABS.TRACK-H
7 1.31 11 SASD*CO-TRACK.CHICAGO
•COPY AQFILE.,EXES«AQCHIC.
«SETC,X
fUS»ER.BK2,N
Runstream to Produce the NEM Air Quality Input
File for Los Angeles
0RUN,R/R AQDAT2/80,•••••*!•••,EXES/MMSAP,120,9999
IDELETE,C EXES»PRTAQLA.
iDELETE.C EXES»AQLA.
ICOND
«CAT,P PRTA(JLA.,F33///140
fUS*ER.BKI,A PRTAQLA.
«ASG,CP EXESMQLA.
|ASG,T AQFILE.
fASG.A SASO*CO-TRACK.
iCOPY.A SASD*EXP-ABS.TRACK-11
«SETC.O
«XQT SASD*EXP-ABS.TRACK-11
7 1.75 II SASO*CO-TRACK.LOS-ANGELES
fCOPY AQFILE.,EXES*AQLA.
iSETC.X
fUS»ER.BK2,N
F-2
-------
Runstream to Produce the NEM Air Quality Input
File for Philadelphia
0RUN.R/R AQDAT3/80,Mt*IMim,EXES/MMSAP. 120, 9999
«OELETE,C EXES»PRTAQPHIL.
§DELETE,C EXES»AQPHIL.
8CONO
ICAT,P PRTAOPHIL.,F33///140
iUS»ER.8KI,A PRTAQPHIL.
«ASG,CP £XES»AQPHIL.
«ASG,T AQFILE.
tASG.A SASD*CO-TRACK.
ICOPY.A SASO«EXP-ABS.TRACK-11
«SETC,0
«XQT SASO*EXP-ABS.TRACK-11
7 0.96 11 SASD»CO-TRACK.PHILADELPHIA
iCOPY AQFILE.,EXES»AQPHIL.
«SETC,X
SUS*ER.BK2,N
EOF:16
Runstream to Produce the NEM Air Quality Input
File for St. Louis
IRUN.R/R A
-------
NEM Runstream for Chicago
IRUN.R/R NEMRN2/80,ilfi|iim,EXES/MMSAP,30,50
IDELETE.C EXES*PRTNEM1.
ICO NO
«CAT,P PRTNEM1..F33///150
fUS*ER.BK],A PRTNEM1.
IASG.T SUMFILE.
IASG.A EXES»AQCHIC.
IUSE AQFILE.,EXES»AQCHIC.
ICOPY.A SASD*EXP-ABS.NEM
IXQT NEM
1
iADD SASD*TRACK-DATA.YEAR-START/MONDAY
CHICAGO 8 CHIC,ALL PERS, B.E.MICRCXNO SOURCESJAS IS AQ
IADD SASD"CO-TRACK.BE/BASE
IADD EXES*MOBILE-DATA.CONCS/20
IAOO SASD*CO-TRACK.CONCS-COHB/U
8ADD SASD*CO-TRACK.CONSTANTS/M
IADD SASD*TRACK-DATA.CHICAGO/NT-6
IEOF
IUS»ER.BK2,N
NEM Runstream for Los Angeles
, R/R NEMRN3/80, ••«»••••, EXES/MMSAP, 30, 50
IDELETE.C EXES»PRTNEMLA.
ICOND
ICAT.P PRTNEMLA..F33///150
«US*ER.BKJ,A PRTNEMLA.
|ASG,T SUMFILE.
IASG.A EXES»AOLA.
IUSE A9FILE.,EXES»AQLA.
ICOPY,A SASD*EXP-ABS.NEM
IXQT NEM
1
IADD SASD*TRACK-OATA.YEAR-START/SATUROAY
LOS-ANGELES 8 ALL PERSONS, B.E. MICRO(NO SOURCES),AS IS AQ
IADD SASO»CO-TRACK.BE/BASE
IADD EXES»MOBILE-DATA.CONCS/20
IADD SASD*CO-TRACK.CONCS-COHB/14
IAOD SASD*CO-TRACK.CONSTANTS/M
IADO SASD*TRACK-DATA.LOS-ANGELES/NT-6
IEOF
IUS»ER.BK2,N
F-4
-------
NEM Runstream for Philadelphia
iRUN,R/R NEMRN4/80,M«lMt*M,EXES/MMSAP,30,50
iDELETE,C EXES«PRTNEMPHIL.
fCOND
iCAT.P PRTNEMPHIL..F33///150
iUS*ER.BK1,A PRTNEMPHIL.
iASG,T SUMFILE.
iASG.A EXES*AQPHIL.
IUSE A
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA 460/3-84-004
2.
3. RECIPIENT'S ACCESSION-NO.
TITLE AND SUBTITLE
Mobile Source Exposure Estimation
5. REPORT DATE
March 1984
6. PERFORMING ORGANIZATION CODE
AUTHORIS)
Melvin N. Ingalls
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Southwest Research Institute
6220 Culebra
San Antonio, Texas 78284
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-03-3073
2. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
13. TYPE OF REPORT AND PERIOD COVERED
Final (.Tung \ Qft? - May 1
14. SPONSORING AGENCY CODE
S. SUPPLEMENTARY NOTES
This project was conducted to provide a national exposure, in terms of person
hours, to non-reactive mobile source pollutants. The basis for the estimate was
the EPA "NAAQS Exposure Model" (NEM) as applied to carbon monoxide, supplemented
by four mobile source microenvironments: parking garages, street canyons, on-
expressways, and roadway tunnels. From previous studies, both published and
unpublished, CO concentration distributions and national population estimates,
by hour of the day, for each of these mobile source microenvironments were
developed. That information was combined to determine national exposure in the
microenvironments. By using the mobile source CO emission factor, exposure to
mobile source pollutants based on a pollutant emission rate of one gram per
minute was determined for each of the microenvironments and the environments
covered by the NEM. The methodology for using this information to determine
exposure to any mobile source pollutant, regulated or unregulated was explained.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Exhaust Emissions
Motor Vehicles
Exposure Estimates
Parking Garages
Tunnels
Street Canyons
Expressways
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
20. SECURITY CLASS (This page)
Unclassified
21. NO. OF PAGES
131
22. PRICE
EPA Form 2220-1 (9-73)
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
6. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of Issue, date of
approvcl, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR (S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hireaichy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Leave blank.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented at conference of,
To be published in. Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 196S COSATI Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. & 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reversal
------- |