&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-460/3-81-021
July 1981
Air
Estimating Mobile Source Pollutants
in Microscale Exposure Situations
-------
EPA-460/3-81-021
Estimating Mobile Source Pollutants in
Microscale Exposure Situations
by
Melvin N. Ingalls
Southwest Research Institute
6220 Culebra Road
San Antonio, Texas 78284
Contract No. 68-03-2884
Task Specification 1
»
EPA Project Officer: Robert J. Garbe
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
July 1981
-------
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 (MD-33), Research Triangle Park, North
Carolina 27711; or, for a fee, from the National Technical Information
Service, 5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Southwest Research Institute, 6220 Culebra Road, San Antonio, Texas, in
fulfillment of Task Specification 1 of Contract No. 68-03-2884. 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 company or product names is not to be considered as
an endorsement by the Environmental Protection Agency.
Publication No. EPA-460/3-81-021
11
-------
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 January, 1980, and completed in February, 1981.
The project was conducted under Task Specification No. 1 of Contract
68-03-2884 and was identified within Southwest Research Institute as Project
11-5830-001.
Mr. Robert J. Garbe of the Emission Control Technology Division, Office
of Mobile Source 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 supervision of Melvin N. Ingalls, Senior Research Engineer, who served
as Project Leader and principal investigator. Key personnel in the Department
of Emissions Research involved in the project were Kathleen Hanna, Senior
Technician, Joyce Winfield, Technician, and Katherine Anderson, Student
Scientist.
111
-------
SUMMARY
The goal of this study was to provide a method of estimating the concen-
tration of any vehicle-generated pollutant in areas- where people are in close
proximity to vehicles. This close proximity exposure is often referred to as
a microscale exposure situation. A list of common microscale exposure situa-
tions was extracted from several hypothetical daily activity routines.
These situations are:
• Enclosed Space
-residential garage
-parking garage
-roadway tunnel
• Street Canyon
• Expressway
-on expressway
-beside expressway
• Localized Area
-parking lots
-industrial sites
For each of the situations, an appropriate dispersion model was selected
for use in obtaining pollutant concentrations. However, no model was found
for the localized area situations. Two models were combined in a new manner
to obtain a single model for the on expressway situation. Section III of the
report contains a mathematical description of each of the models selected.
To determine the exposure level in each case, it was decided to calculate
the pollutant concentrations in typical and severe actual situations. Real
world actual situations were chosen over hypothetical situations to provide
more meaningful and defensible pollutant concentration levels. The concept of
a "severe" situation was used rather than "worst case," since using the real
world worst case requires examining every existing physical situation to
find the one worst case.
The range of physical variables for each situation was examined to define
the typical and severe cases for each situation. The pertinent variables for
each of the situations are summarized below:
• Residential Garages - There are over 34.1 million residential
garages in the United States. Multifamily dwellings were not
included in this category so that only single family garages
are simulated. No data was found on size range. A single car
IV
-------
garage 10 feet wide, 26 feet long, and 8.42 feet high, was
chosen for both the typical and severe situations, with
pollutant generation time being longer for the severe case.
Report Section IV.A, page 22, defines the range of all
physical variables of this category.
Parking Garages - There are over 5,300, possibly as many as
10,000, parking garages in the United States. They vary
in parking capacity from under 100 cars to over 4,000 cars.
The average capacity is about 400 cars. Most parking garages
are above ground with open sides and depend on ambient air
circulation for ventilation. About ten percent of the garages
have mechanical ventilation. Mechanical ventilation rates
vary from four air changes per hour to 16 air changes per hour.
The Convention Center parking garage in San Antonio, Texas, was
chosen as the typical situation. It is a 5 level garage with a
total capacity of 461 vehicles. It is above ground with natural
ventilation. The Music Center garage in Los Angeles, California
was chosen as the severe case. This garage has 8 levels, 6 of
which are underground, with a capacity of 1,582 vehicles. It is
entirely mechanically ventilated at a rate of 4 air changes per
hour (prior to 1979). Section IV.B, page 23, defines the range
of physical variables considered.
Roadway Tunnels - There are approximately 105 roadway tunnels
over 500 feet long in the United States. The longest tunnel is
9,117 feet long. Over half these tunnels are under 2,000 feet
in length. Approximately 59 of these tunnels were classed as
commuter tunnels. The daily traffic in these commuter tunnels
ranged from under 5,000 vehicles per lane per day to over 30,000
vehicles per lane per day.. All but three of the commuter tunnels
over 500 feet long were mechanically ventilated. Ventilation
rates vary from 0.068 m3/sec per lane meter to 0.275 m3/sec per
lane meter. The Lowery Hill tunnel in Minneapolis, Minnesota
was chosen as the typical tunnel exposure. This 1,496 foot long
tunnel has two tubes with three lanes per tube. It has a daily
traffic count of approximately 12,000 vehicles per lane. The
Baltimore Harbor tunnel in Baltimore, Maryland was chosen as the
severe tunnel exposure. This 6,700 foot tunnel has two tubes
with two lanes per tube. It has a daily traffic count of 16,375
vehicles per lane. Section IV.C, page 31, covers the range of
physical variables considered.
Street Canyons - There are estimated to be almost 850 miles of
street canyons in the United States. In general, the streets are
less than seven lanes wide and the buildings under 26 stories.
The minimum canyon height to width ratio for a building-lined
street to be considered a street canyon is approximately 0.3.
Since the street canyon model used was limited to a height to
width ratio of 2.0, very narrow, deep canyons were not considered
-------
in this study. Houston Street in San Antonio, Texas, between
the cross streets of Navarro and St. Mary's was chosen as the
typical street canyon exposure. The street is four lanes wide
with seven foot wide sidewalks and a height to width ratio of
1.81. Main Street in Houston, Texas, between the cross streets
of Capitol and Rusk, was chosen as the severe exposure situa-
tion. Main Street is six lanes wide with a 15 foot sidewalk
on each side. It has a height to width ratio of 1.39. Section
IV.E, page 50 discusses the physical variables of street canyons.
• Expressways - There are approximately 17,000 miles of urban
expressways in the United States. Over half of this mileage
consists of highways of four or less lanes. Less than one
percent of the highway mileage consists of 9 or more lanes.
The average daily traffic (ADT) varies from under 10,000 to
over 200,000 vehicles per day. The median ADT is between
30,000 and 40,000 vehicles per day. An examination of wind
direction relative to expressway orientation revealed that
all relative wind directions were equally alike on a nation-
wide basis. The average wind speed was approximately 7 mph.
A section of 1-410 on the west side of San Antonio, Texas
between U.S. 90 and 1-35 was chosen as the typical exposure
situation. This four lane urban expressway runs in a NNW to
SSE direction and has an ADT of approximately 28,000 vehicles.
A portion of the Santa Monica Freeway (1-10) in Los Angeles
between Washington and La Brea was chosen as the severe express-
way situation. At this section, the freeway is 10 lanes wide
with an ADT of over 200,000 vehicles. Section IV,D, page 40
of the report discusses expressway variables.and prevailing wind
speeds and directions.
Pollutant concentrations were calculated for each typical and severe exposure
situation using the chosen dispersion models. Concentrations were calculated using
one gram per mile per vehicle for tunnel, street canyon and expressway situations,
and one gram per minute per vehicle for the residential and parking garages. The
major data and assemptions used for each situation are summarized by category in
the following paragraphs:
• Residential Garage - The typical and severe residential garage
exposure situations use the same physical setting, but different
pollutant generation times. The physical situation chosen was a
single car garage with a volume of 2189 cubic feet (10 feet by 26
feet by 8.42 feet). The only ventilation is natural air. circula-
tion at a rate of 615 cubic feet per minute. The ventilation
factor, as discussed on pages 13 and 66 of the report, was assumed
to be 0.25 for the vehicle exhaust equation.
• Parking Garage - The typical exposure situation was chosen as the
fourth level of the San Antonio, Texas, Convention Center parking
garage following an event at the adjacent Convention Center. This
garage is naturally ventilated, With an assumed wind speed of 7
VI
-------
mph, the ventilation rate for the fourth level is 308,000 cfm.
The total volume of this level is 356,000 cubic feet. It was
assumed that there were 17 active cars on the fourth level at
all times during the emptying process. A ventilation factor of
0.4 was used for the vehicle exhaust equation. The severe expo-
sure situation was chosen as the fifth level of the Music Center
parking garage in Los Angeles, California. This garage is
mechanically ventilated. The ventilation scheme modeled was the
pre-1979 ventilation system. The level five volume is 737,316
cubic feet with a ventilation rate of 38,100 cfm. It was assumed
that there were 56 active cars on the fifth level at all times
during the emptying process. A ventilation factor of 0.3 was
used for the vehicle exhaust equation.
Roadway Tunnels - The typical exposure situation was chosen as
the morning rush hour in the 1500 foot long Lowery Hill tunnel
in Minneapolis, Minnesota. This two tube tunnel has a semi-
transverse ventilation scheme with a maximum ventilation rate
of 536,000 cfm per tube. The maximum ventilation rate was
increased 10 percent to account for the piston effect of the
cars. It was assumed that there were 25 vehicles traveling
at 45 mph in each tube at all times during the morning rush hour.
A ventilation factor of 1.0 was used. The severe tunnel exposure
situation used was rush hour in the Baltimore Harbor Tunnel.
This 6700 foot long tunnel has a fully transverse ventilation
system with a maximum ventilation rate of 800,000 cfm per tube.
To this rate is added 50,000 cfm for the piston effect of cars
entering the tube. It was assumed that there were 165 vehicles
traveling at 25 mph in the tube at all times during rush hour.
A ventilation factor of 1.0 was used for this situation.
Street Canyons - As explained in Section IV, the street canyon
conditions that constitute typical and severe exposures depend
on the emission type being considered. For some emissions, a
severe case is low speed, high traffic density; for others a
higher speed, lower density. Street width then becomes the
controlling parameter in defining typical and severe situations.
A four lane street was used for the typical case, and a six
lane street, the severe case.
The typical case, Houston Street in San Antonio, Texas, is 18.6
meters wide including sidewalks. The street runs in a West to
East direction. The average building height is 33.7 meters in
the city block used for this situation. The receptor was located
on the sidewalk, 8,8 meters from the center of the street, 1.5
meter above street level. Wind direction used was SE to SW at
8 miles per hour. Two traffic conditions were used: 800 vehicles
per hour, at 5 mph and 1600 vehicles per hour at 20 mph. The
severe case, Main Street in Houston, Texas is 27 meters wide.
The average building height in the city block chosen for this
situation is 38 meters. The street is orientated NE to SW. The
vii
-------
receptor was located on the sidewalk, 11.4 meters from the
center of the street, 1.5 meters above street level. The
wind conditions used were 0.9 miles per hour, perpendicular
to the street. Two traffic conditions were used: 1200
vehicles per hour at 5 mph and 2400 vehicles per hour at
20 mph.
Expressways - Two types of expressway exposures were con-
sidered: the receptor in a vehicle on the expressway and
a receptor beside the expressway. The typical on-expressway
situation was chosen as a vehicle occupant in the outside
southbound lane of IH410 between Valley High Drive and U.S.
90 in San Antonio, Texas during the morning peak traffic
period. The complete set of typical situation input values
for the despersion model is listed below.
Highway dimensions
-4 lanes, each 3.66 meters (12 feet) wide
-median, 13.41 meters (44 feet) wide
Vehicle volume
-900 vehicles per hour per lane on southbound (receptor) side
-500 vehicles per hour per lane on northbound side
Point source (vehicles on receptor side) information
-emission strength, 0.0153 g/sec, (1 g/mile at 55 mph)
-average of four vehicle lengths between vehicles
-average vehicle length, 6.7 miles (22 feet) assumes 80 percent
cars, 20 percent trucks
-exhaust height assumed to be 0.3 meters (one foot)
-10 vehicles in each lane ahead of receptor vehicle
Line source (vehicles in opposite lanes)
-line source strength 0.0862 x 10~3 g/sec m lane (1 g/mile per
vehicle, 500 vehicles per hour)
-bouyancy flux 0.0191 m3/sec3
Receptor
-located at center of outside downwind lane
-height 1.22 meters (4 feet)
Meteorology
-wind direction, from 357.5 to 270 degrees relative to
vehicle direction
-wind speed, from 1 m/sec to 6 m/sec (2.2 to 13.4 mph)
-stability, Pasquill Class C (slightly unstable)
The severe on-expressway situation was chosen as a vehicle
occupant in the middle eastbound lane of the Santa Monica
Freeway (IH10) between Washington Boulevard and La Brea
viii
-------
Boulevard on the west side of Los Angeles during the morning
peak traffic period. The complete set of severe situation
input values for the dispersion model is listed below:
Highway dimensions
-10 lanes, each 3.35 meters (11 feet) wide
-median, 6.71 meters (22 feet) wide
• Vehicle Volume
-2000 vehicles per hour per lane on eastbound (receptor) side
-1675 vehicles per hour per lane on westbound side
• Point Source (vehicles on receptor side) information
-emission strength, 0.01389 (.1 g/mile at 50 miles/hr)
-average of two vehicle lengths between vehicles
-average vehicle length 6.7 meters (22 feet) assumes 80 percent
cars, 20 percent trucks
-exhaust height, 0.3 meters (one foot)
-10 vehicles per lane ahead of receptor vehicle
• Line Source (vehicles in opposite lanes)
-line source strength, 0,2982 x 10~3 g/sec m lane (1 g/mile
per vehicle, 1675 vehicles/hr per lane)
-bouyancy flux 0.0638 rn /sec3
• Receptor
-located at center of center lane
-height 1.22 meters (4 feet)
• Meteorology
-wind direction, from 357.5 to 270 degrees relative to
receptor vehicle direction
-wind speed, from 1 m/sec to 6 m/sec (2.2 to 13.4 mph)
-stability, Pasquill Class C (slightly unstable)
1-10 at Silber Road on the west side of Houston was chosen as the severe
beside-expressway exposure location. At this location., 1-10 runs due east
and west, with four lanes in each direction. The receptor location was on
the north side of the expressway. The 1977 traffic count was 167,860 vehicles
per hour per lane. Wind direction was 90 degrees relative to the road (180
degrees true), and wind speed of 1 m/sec. Neutral atmospheric stability was
used. For the long term exposure, Houston wind data indicated that vehicle
emissions were dispersed to the north side of the expressway, approximately
60 percent of the time.
A summary of the pollutant concentrations for all the exposure situations
considered is given in the following table. Also shown is the section of the
report which contains that data and assumptions used in calculating the con-
centrations. The reader is cautioned to understand the assumptions before
using information from this table.
ix
-------
SUMMARY OF MICROSCALE SITUATION CONCENTRATIONS
Situation
1.
2.
3.
4.
Residential Garage
Typical (30 second run tijne)
Severe (5 minute run time)
Parking Garage
Typical (parking level)
Severe a) inlet air component
b) exhaust emission component
Roadway Tunnel
Typical
Severe
Street Canyon (sidewalk receptor,
includes background)
Typical a) 800 vehicles per hr
b) 1600 vehicles per hr
Severe c) 1200 vehicles per hr
d) 2400 vehicles per hr
Concentration
yg/m3 Ca)
7,900
67,000
3,900
9,600
46,100
1,123
2,856
42
85
141
282
5.
On Expressway (Wind: 315 deg. relative, 2.2 mph)
Typical 124
Severe 506
Report
Section(bV
V.A page 64
V.B page 71
V.C page 83
V.D page 86
V.E page 96
6. Beside Expressway
Severe 1 meter
10 meters
100 meters
1000 meters
Short Term Annual V.E page 102
397
334
105
13.6
61
48
14
1.6
(a)For 1 g/vehicle mile (1 g/vehicle minute for garages). Assumes no
background concentrations except as noted.
(k^The report section shown contains the data and assumption used to
determine pollutant concentrations.
To use these values with emission factors other than one gram per mile or one
gram per minute, multiply the concentration in the table by the desired emission
factor.
x
-------
TABLE OF CONTENTS
FOREWORD ii;L
SUMMARY iv
LIST OF FIGURES
LIST OF TABLES
LIST OF SYMBOLS
I . INTRODUCTION 1
A. Objectives 1
B . Approach 2
II. SELECTION OF EXPOSURE SITUATIONS 3
A. Activity Patterns of Population Groups 3
B. Types of Exposure Situations Derived from Activity Patterns 5
III. MODELS FOR DETERMINATION OF AMBIENT CONCENTRATION 9
A. Literature Search 9
B. Model Selection 10
C. Model Use 21
IV. RANGE OF PHYSICAL VARIABLES FOR EXPOSURE SITUATIONS 23
A. Residential Garage 23
B. Parking Garages 24
C. Roadway Tunnels 32
D. Urban Expressways 41
E. Street Canyons 51
F. Localized Area Sources 54
V. AMBIENT CONCENTRATIONS FOR SELECTED EXPOSURE SITUATIONS 65
A. Residential Garage Concentrations 55
B. Parking Garage Concentrations 72
C. Roadway Tunnel Concentrations .34
D. Street Canyon Concentrations 87
E. Urban Expressway Concentrations 97
F. Localized Area Concentrations
VI . RECOMMENDATIONS 113
XI
-------
TABLE OF CONTENTS (Cont'd)
REFERENCES
APPENDICES
A. Verification of ONEX Computer Program
B. Tunnels in the United States
C. Bibliography
xii
-------
LIST OF FIGURES
Figure Page
1 Typical Residential Garage 12
2 Typical Roadway Tunnel
3 Typical Parking Garage
•I C
4 Street Canyon Exposure Situation
?n
5 Expressway Exposure from a Relative Motion Aspect ^u
6 Size Distribution of Parking Garage Projects - 1972 and 1979 29
7 Road Tunnel Distribution in the United States 34
8 Tunnel Length Distribution ^5
9 Average Daily Traffic Distribution 36
10 Relationships Between Traffic Rate, Vehicle Speed and
Traffic Density 37
11 Types of Tunnel Ventilation 3^
12 Urban Interstate Highway Distribution by Number of Lanes ^4
13 Average Daily Traffic Distribution for the Urban Interstate
System 45
14 Wind Speed Distribution for 25 cities at 7 a.m. 52
15 Aerial View of Central Business District Houston, Texas ^6
16 Minimum Building Height for Helical Air Circulation in a
Street Canyon 58
17 Example of Traffic Speed Relationship to Traffic Flow 6^
18 Effect of Driving Cycle Speed on NOX and CO Emissions 62
19 Effect of Driving Cycle Speed on Ammonia and Sulfate
Emissions 63
20 Single Car Garage with Storage Space 65
Xlll
-------
LIST OF FIGURES (Conf d)
Figure page
21 Residential Garage Multipliers for C^ and C2 68
22 Residential Garage Multiplier for C^ 69
23 Sketch of Parking Level Arrangement, Music Center Garage 73
24 Floor Plan of Parking Level 5 of Music Center Garage 74
25 Parking Level Arrangement, San Antonio Convention Center
Parking Garage 79
26 Typical Parking Level Floor Plan for Convention Center Garage 80
27 Houston Street Between Navarro and St. Mary's, San Antonio,
Texas 88
28 Main Street Between Capitol and Rusk, Houston, Texas 92
29 Aerial View of a Portion of Downtown Houston, Texas 93
30 Sketch of Expressway Section for Typical Expressway Exposure 98
31 Ambient Pollutant Concentration as a Function of Wind
Direction. Typical Expressway Case. 101
32 Concentrations Experienced by Receptor Traveling on
Expressway. Severe Expressway Case. 105
33 1-10 at Silber Road, Houston, Texas 106
34 Concentrations Downwind of Expressway 109
xiv
-------
LIST OF TABLES
Table Page
1 Types of Exposure Situations 5
2 Average Number of Parking Garages by Urbanized Area Size "
3 Total Parking Garages by Urban Area Size Circa 1965 25
4 Total Parking Garages for Standard Metropolitan Statistical
Areas - Circa 1972 27
5 Parking Garage Construction Between 1972 and 1979 27
6 Ventilation Rates for Some Mechanically Ventilated
Parking Garages 31
•3Q
7 Commuter Tunnel Ventilation Type by Tunnel Length
8 Maximum Ventilation Rates for Some U.S. Roadway Tunnels ^0
9 Urban Interstate Highway System Mileage ^2
10 Estimated Capacity of Urban Interstate Expressways by Number
of Lanes 47
11 Prevailing Wind Direction for SMSA Over 400,000 Population 48
12 Distribution of Relative Wind Direction for San Antonio,
Texas, Urban Interstate Expressways
50
13 Central Business District Areas for Standard Metropolitan
Statistical Areas Over 400,000 Population (1970 Census) 54
14 Average Daily Traffic for Some Central Business District
Streets in 'Selected Cities 59
15 Ambient Concentrations for Receptor in Residential Garage-
Typical and Severe Cases 71
16 Data for Music Center Garage 75
17 Ambient Concentrations for Receptor in a Parking Garage
Severe Case-Los Angeles Music Center Garage(Pre 1979) 78
18 Ambient Concentrations for Receptor in a Parking Garage
Typical Case-San Antonio, Convention Center Garage 83
xv
-------
LIST OF TABLES (Cont'd)
Table Page
19 Ambient Concentrations for Receptor in Roadway Tunnel -
Typical and Severe Cases 86
20 Street Canyon Ambient Air Concentrations 96
21 Ambient Concentration for Receptor on Expressway-Typical
Expressway Exposure Situation
22 Ambient Concentration for Receptor on Expressway-Severe
Expressway Exposure Situation
23 Ambient Concentrations for a Receptor Downwind of an
Epxressway 1°7
24 Annual Average Concentrations for a Receptor on the
Northside of 1-10 at Silber Road, Houston, Texas
xvi
-------
LIST OF SYMBOLS
C = concentration of pollutant mass/volume
CQ = concentration of pollutant in enclosed space at time t = 0.
mass/volume
Ci = Initial concentration contribution to enclosed space concentration
C2 = ventilation contribution to enclosed space concentration
03 = vehicle contribution to enclosed space concentration
C^ = concentration of pollutant in ventilation air
E = efficiency of pollutant filter, if any (0 to 1.0)
F-^ = bouyancy flux, lengths/time^
H = building height, length
h0 = plume center height, length
hc = effective height of emission source, length
K = turbulent diffusivity of momentum "V 1 m^/sec^
KC = concentration coefficient, dimensionless
L = block length
m = mixing factor
Q = pollutant generation rate, mass/time
QL = pollutant generation rate per meter
R^ = volume rate of ventilation, volume/time
Rf = volume rate of air through filter, volume/time
S = slant distance from source to receptor
t = time
U = wind speed, length/time
V = enclosed space volume length^
W = street canyon width, length
Y = horizontal distance from plume center line to receptor
Z = vertical distance from plume center line to receptor
Greek Letters
6 = penetration depth of rooftop wind into street canyon, meters
Oy = horizontal dispersion parameter, meters
0Z = vertical dispersion parameter, meters
xvii
-------
I. INTRODUCTION
The exhaust from vehicles powered by combustion engines contains many
more compounds than the four exhaust constituents (HC, CO, NOX/ and parti-
culates) currently regulated.(1/2)* jn addition, systems designed to control
the four regulated emissions can alter the chemical composition of the
exhaust.(3,4) Malfunctions of engine and emission control systems can also
change the concentrations of the various chemical species in the exhaust.(5,6,7)
The Clean Mr Act Amendments of 1977 require vehicle manufacturers to
show that each emission control system will not "... cause or contribute to
an unreasonable risk to public health, welfare or safety in its operation or
function."(8) Thus, in evaluating any proposed emission control system, a
manufacturer must determine if any chemical compounds in the exhaust are
increased as a result of that system. If such increases occur, the manu-
facturer must determine whether or not they contribute to an unreasonable
risk to public health or welfare.
The EPA, in OMSAPC Advisory Circular 76^' , presented a list of chemical
compounds which might occur in engine exhaust pollutants and which the Agency
felt might contribute to a public health risk. The list was not intended to be
a complete list of possible deleterious compounds. Sampling methodologies and
analytical techniques have since been developed for most compounds on the
list.(l°fH) Thus, it is now possible to determine the exhaust emissions, in
mass per distance or mass per time, of many unregulated emissions which have a
possibility of "causing or contributing to an unreasonable risk to public health."
Two steps remain in ascertaining if the unregulated emission levels pro-
duced by prototype emission control systems meet the requirements of the law.
The first is to determine, for a variety of exposure situations, what ambient
air concentrations results from the measured exhaust emission levels. The
second step is to determine if these ambient air levels pose a public health
or welfare problem. This study addresses the first step: the determination
of ambient air concentrations for a variety of situations.
A. Objectives
In order to provide ambient air concentrations for enough situations to
determine (when combined with health effects) if a given pollutant emission
level is a public health problem, this project has four objectives. They are:
1. to identify and classify the situations in which a person is exposed
to mobile source pollutants
*Superscript numbers in parentheses designate references at the end of this
report.
-------
2. to choose and make ready for use, mathematical models capable
of determining ambient air concentrations from vehicle emission
rates for each exposure situation
3. to determine the range of physical variables for each expsoure
situation so that "typical" and "severe" exposure cases may be
defined
4. to apply the models to determine ambient air concentrations for
each exposure situation using the real world "typical" and
"severe" situations
B. Approach
Fulfillment of the first objective of this project required a considera-
tion of sites where highway vehicles are used, and the location of people in
relationship to the vehicles. This item was accomplished considering the
daily routine of individuals in various age groups, occupations, and geographic
locations. Similar situations were then grouped together to obtain a minimum
number of truly different situations.
The mathematical models for each different situation were chosen from
the literature. No attempt was made to develop new models, although existing
models sometimes required modification or use in a new way to most accurately
define the ambient air concentrations.
The defining of real world typical and severe conditions for each
situation was accomplished by searching the literature for physical descrip-
tions of situations, such as the sizes of tunnels and street canyons, deter-
mination of time-averaged atmospheric conditions, and on-site investigations
of real physical situations. The aim was to define actual places and actual
atmospheric conditions occurring in these places that could be classified as
typical and severe from consideration of the physical and atmospheric
characteristics of each exposure situation. Thus, the need to prove that
some arbitrary, hypothetical situation could actually occur was avoided.
Ambient air concentrations for each of the real world exposure
situations chosen was obtained by using the parameters from these situations
together.with the appropriate model. Where the vehicle emission rate is a
simple multiplier to'the rest of the terms in the model equation, the model
was run with a unity emission factor in the correct units for the equation
(e.g., one gram/min). Where the vehicle emission factor is not a simple
multiplier, curves of ambient air concentration as a function of emission
factor for the given physical situation were produced.
-------
II. SELECTION OF EXPOSURE SITUATIONS
For this project, no definitive investigation of population activity
patterns was undertaken, nor is it known if such studies exist in the
literature. EPA is currently studying population activity patterns,1
but the results of that study are not yet available. However, by envisioning
possible activity patterns for several different population groups, it should
be possible to arrive at most of the common exposure situations.
A. Activity Patterns of Population Groups
Obviously,all people in the United States do not follow the same daily
activity pattern. Just as obvious, however, is the fact that large groups
of the population do have similar activity patterns. Since the desired result
of this investigation is exposure situations involving emissions from mobile
sources, the situations investigated should involve areas where there are
large concentrations of vehicles. This means that the population groups of
interest are in an urban or suburban, rather than rural, setting.
There are many ways of dividing the population by activity groups.
Depending on the criteria, a large number of groups could be identified.
For purposes of this study, four groups of people were chosen with obviously
different activity patterns. These groups are:
• Urban commuter
• Suburban to central city commuter
• Suburban homemaker
• Suburban school child
By following a member of each of these groups through a hypothetical
daily activity cycle, those situations involving exposure to mobile source
emissions can be identified.
1. The Urban Commuter
The urban commuter is one whose residence is in the urban area of a
city. The residence is assumed to be multistoried (four or more stories). The
urban commuter commutes by public transportation, with his commuting trip
lasting under one-half hour. It is assumed that he both lives and works in an
area of multistoried buildings. A hypothetical urban commuter might have a
daily activity pattern as follows. In the morning he leaves his residence
and walks, in a shallow street canyon, one block to a bus stop; waits 10
-------
minutes at edge of the traffic lane for a bus; rides the bus for 10 minutes
through street canyons with morning rush hour traffic; gets off the bus; then
walks one block to his work place. His work location is on the first floor
of a multistoried, air-conditioned building. At lunch time, he walks one
block to a restaurant (air-conditioned) on the first floor of a multistoried
building; eats, and returns to his work place. After work, he reverses his
morning commuting routine. He spends the evening and night in his residence.
From the preceding scenario, it can be seen that an urban commuter's exposure
to vehicle emissions is almost entirely in street canyons of various sizes and
orientations to the wind. His main exposure is during high traffic hours.
2. The Suburban Commuter
The hypothetical suburban commuter lives in a single family dwelling
in a residential area. He commutes to work in a private vehicle. The first
encounter with vehicle emissions during his assumed daily activity pattern is
when he starts his car in an enclosed garage. His commuting trip to a city
central business district (CBD) exposes him to vehicle emissions in the
following situations: suburban through streets, expressway, a highway tunnel,
and the CBD street canyons. The total trip takes one-half hour and ends in an
underground parking garage. He walks one block through a street canyon to his
work place. This work place is several floors up in a multistory, air-conditioned
building. At lunch time, he walks one block through a street canyon to a res-
taurant, eats, and walks back. After work he reverses his morning commuting
routine. In the evening he attends a professional sports event at a large
stadium, using his private vehicle to go to and from the event over roadways
that are similar to his commuting route. He returns to his suburban residence
following the sports event. From the preceding scenario, it can be seen that
the suburban commuter encounters a variety of microscale exposure situations.
However, his main exposure is also during high traffic hours.
3. The Suburban Homemaker
The hypothetical suburban homemaker is a woman with two children,
whose activity pattern is a great deal more varied than the commuter patterns
described above. The homemaker's vehicle emission exposure begins in an enclosed
residential garage. The remainder of her exposure during the day results from
numerous short, nonexpressway trips in her private car, to such places as schools
shopping centers (large and small), etc. Early morning trips, for example to
deliver children to school, are made during the peak traffic period. The re-
mainder of the trips are at nonpeak hours. However, she spends more time waiting
at signaled intersections than the commuter. Assume that the homemaker also
attends a professional sports event in the evening, traveling to and from the
event over a variety of roadways in a private car. She spends the remainder
of the day in her suburban home. From the above scenarios, it can be seen
that the suburban homemaker exposure is primarily from primary and secondary
suburban streets and parking lots. The exposure is generally at nonpeak traffic
hours.
4. The Suburban Schoolchild
The exposure to mobile source emissions of the hypothetical suburban
-------
schoolchild begins while waiting for a school bus beside a suburban secondary
street. Assume this exposure lasts 15 minutes, followed by a 10 minute bus
ride on suburban streets to the school. Assume the school is a single story,
nonair-conditioned building, located within a kilometer of a major expressway.
The child has approximately three quarters of an hour outdoor activity during
the day. At the end of the school day, the bus trip is reversed. The remainder
of the day and all night are spent in the vicinity of the child's suburban home.
Thus, the major exposures to mobile source emissions for this hypothetical
school child occur while waiting for and riding the school bus, and during
outdoor activity near a major expressway.
B. Types of Exposure Situations Derived from Activity Patterns
From the preceeding hypothetical activity patterns, a group of common
exposure situations likely to be encountered can be determined. These exposure
situations are listed in Table 1 and discussed individually below.
TABLE 1. TYPES OF EXPOSURE SITUATIONS
I. Enclosed space
A. Residential garage
B. Parking garage
C. Roadway tunnel
II. Street Canyons (short term)
III. Expressway
A. On expressway
B. Beside expressway
IV. Localized area sources
A. Parking lots, airports
B. Industrial and construction sites
V. Mesoscale area (e.g., a suburban housing area)
1. Enclosed Spaces
Several enclosed space exposure situations are identified above.
They are listed in Table 1 as the residential garage, parking garage, and
roadway tunnel. By their nature, these situations are short term exposures,
lasting probably less than 15 minutes except in rare circumstances. However,
because of their enclosed nature, these situations have the possibility of
producing the highest ambient concentrations of mobile source pollutants.
Therefore, this situation was chosen as one of the situations for which a
model would be required.
-------
2. Street Canyons
Street canyon exposures, both short term and long term, appeared
frequently in the hypothetical activity patterns. Because the buildings
along a street canyon inhibit dispersion of the pollutants, street canyons
have the possibility of producing high ambient pollutant concentrations.
Therefore, street canyons were chosen as one of the situations for which a
model would be required.
3. Expressway
Two expressway situations appear in the hypothetical activity pat-
terns. One situation is the commuter exposure on the expressway, the other is
exposure in close proximity (within a fraction of a kilometer) to the expressway.
The on expressway situation is short term, with a definite possibility of pro-
ducing relatively high ambient pollutant concentrations. The close proximity
situation could be either short term or longer term (on the order of eight
hours). Here, too, there is a possibility of relatively high ambient con-
centrations of mobile source pollutants. Therefore, the expressway situations,
both on and beside an expressway were chosen as situations for which models
would be required.
4. Localized Area Sources
The hypothetical activity patterns included several mobile source
exposure situations in parking lots, such as at a shopping mall or sports stadium.
These locations are normally classed as area sources, since the sources of the
pollutant are spread over a relatively large area. For purposes of this study,
the exposure of interest is within the area source itself. For parking lots,
the exposure is obviously of short duration, probably less than one-half hour.
Note,however, that for a single parking event the exposure occurs twice, once
on entering the parking lot- and once on leaving.
Similar area-type sources, while not mentioned in the activity pattern
scenarios, can be envisioned. Such sources as airports, trucking terminals
and construction sites could all qualify as area sources. Additionally, expo-
sures in these sources could last for a matter of hours, rather than the fraction
of an hour assumed for parking lots. For purposes of this report, with its
emphasis on short term, high level exposure, this exposure situation will be
covered only briefly.
5. Mesoscale Areas
In all of the suburban activity patterns presented above, a large
part of each day was spent at a suburban residence. In most cases, the sub-
urban residence will not be in close proximity (within 1 km) to a major mobile
pollutant source. At approximately one kilometer downwind, the effect of a
single highway is negligible.(13) Therefore, when examining the exposure at
-------
a suburban residence, an area large enough to include all the major mobile
sources within several kilometers must be included, since the sum of many
mobile sources could add up to a value that is not negligible. Areas up to
tens of square kilometers in size are often referred to as mesoscale areas.
It is this size area that is required for the determination of ambient levels
of mobile source emissions at a suburban residence. While the ambient con-
centrations may not be negligible, they are certainly relatively low when
compared to some of the other exposures considered above.
This exposure siutation is of interest when long term, low level
exposure is important, such as in the investigation of carcinogens or criteria
pollutants which have significant public health hazards. Since this report is
concerned with recurring short term exposures, the mesoscale case will not be
covered in this report.
-------
III. MODELS FOR DETERMINATION OF AMBIENT CONCENTRATIONS
A. Literature Search
The purpose of this study was to find the best available dispersion
models to determine ambient concentrations in the situations chosen, not
to develop new models. To obtain the best available models, a literature
search was conducted for available dispersion models covering a spectrum
of mobile source emissions exposure situations.
To obtain the maximum coverage of the literature, a computerized
abstract service, Lockheed Information System's "Dialog," was used. Four
data bases were searched: (1) Engineering Index, (2) Enviroline, (3) Pol-
lution Abstracts, and (4) Dissertation Abstracts International. The key
words used in the search were "air quality" and "model". The Engineering
Index search produced 232 abstracts; Enviroline produced 375; Pollution
Abstracts produced 259 abstracts; and Dissertation Abstracts yielded 8
abstracts. In addition, two NTIS published Bibliographies were obtained:
(1) "Automobile Air Pollution Atmospheric Motion (1970-1979)" and (2) "At-
mospheric Modeling of Air Pollution," Volumes 1, 2 and 3, covering the
period 1964 to June 1979. A total 1248 bibliographic listings were obtained
(including many duplications).
Of these 1248 listings, approximately 375 were chosen for further
screening. Some of these listings were on hand or immediately available
from the Southwest Research Institute (SwRI) Library. From references in
the literature obtained, additional listings were found. Not all of these
were obtained, and of those obtained, not all were used. Nevertheless,
the complete list is contained in Appendix C. Literature actually used
is, of course, shown in the List of References.
At the beginning of the study it was decided to initially consider
only models for nonreactive pollutants. Atmospheric chemical reactions
normally require time frames of hours to days. The highest level mobile
source exposures are normally of short durations (fraction of an hour).
Thus, it was fe.lt that considering atmospheric reactions would add un-
warranted complexity to the study at this time. After the methodology has
been developed, it may be desirable to include models which consider chemical
reactions in the atmosphere for certain long term, low level exposures. Thus
the literature list in Appendix C contains entries relating to reactive
models, but these articles have not been reviewed.
The review of the literature concentrated on actual dispersion models,
so that a model for each situation could be chosen as soon as possible. The
models chosen are defined in the next report section.
-------
B. Model selection
This section covers the analysis performed in selecting a model for each
of the exposure situations defined, and presents the algorithm selected.
1. Model Selection Criteria
As the investigation of available models was begun, it became obvious that
a set of screening criteria was needed based on the intended use of the models.
From the task statement of work and discussion with the EPA Project Officer,
a succinct statement of the intended use of the models is:
"The dispersion models will be used to determine the ambient
concentrations of mobile source unregulated emissions to which
humans are exposed in various situations."
The first screening criterion considered was: Is the model appropriate
for the intended use? It was determined that some general types of models were
inappropriate for use with this study.
Air Quality models can be divided into three general categories as listed
below:
1. Rollback models
2. Statistical models
3. Physical Simulation models
a. Box
b. ' Gaussian
c. Mass conservation
(1). Eulerian (multibox) models
(2). Lagrangian (moving-cell) models
(3) . Particulate-in-cell and others
Rollback or proportional scaling models do not, strictly speaking,
predict ambient concentrations. They are used in the regulatory process to
determine if a given control strategy will result in the desired improvement
in air quality. Therefore, rollback models are not appropriate for this study.
Statistical morlels make use of regression analysis to relate pollutant
emissions to ambient air quality using data from measured ambient concentrations.
The success of a statistical model depends on having a large number of measured
concentrations from areas similar to those where the predicted concentrations
are desired. Since measured ambient concentrations of unregulated pollutants
are not generally available, it is improbable that a statistical model would
be applicable to this study.
10
-------
This leaves what is called here, for lack of a better description,
physical simulation models. Physical simulation models all consider the
physical situation to be modeled and attempt to apply the basic equations
for the movement and dilution of the pollutant emissions in the atmosphere.
In theory, these types of models should be applicable to any pollutant and
any physical setting. Thus, it is physical simulation models that will be
investigated for this study.
The second criterion is that the models initially selected need not
consider atmospheric chemical reactions. A third criterion is that the models
should have been subjected to validation testing. Where two models are avail-
able that have been compared against actual measured concentrations, the one
.which yields concentrations closer to the measured concentrations the larger
percentage of the time will be used. If both models underpredict, this pro-
cedure obviously is best. However, if one or both of the models overpredict
the choice is not so obvious. There are those who argue that the model which
overpredicts should be used to give a "conservative" estimate. However,
this approach more often than not just adds to any controversy that may be
generated by the use of the model. It would appear more straightforward to
use the most accurate model, then apply a clearly defined "factor of safety,"
if a "conservative estimate" is desired.
It should be pointed out that it may not be possible in all exposure •
situations to find a model which has been validated. In such cases, the various
theoretical aspects of each model must be evaluated. The model chosen should
then be the one judged to best fit the physical situation.
2. Enclosed Spaces
Three exposure situations involving enclosed spaces were identified
in Section II. The situations are: (1) private residential garage, (2) roadway
tunnel, and (3) parking garage. Sketches of each of these exposure situations
are shown in Figures 1, 2 and 3. Ideally, one algorithum should be applicable
to all three situations.
Compared to other situations, the literature on enclosed space dis-
persion models is rather limited, with only 12 abstracts found which might be
applicable. Additional articles were obtained from current literature as the
project progressed. All references obtained indicate that some form of Turk's
Equation should be used for the enclosed space dispersion model. (14/ 15, 16)
Other studies derive essentially the same equation independently. (17,18,19,20)
The basic form of Turk's Equation that is generally used has been shown to
be incorrect in two recent articles. (19/20) ^g form of ^he equation used
in this study will be that from Reference 20 as shown below:
C.R.
C = (C0 exp [-(R-L + ERf) Fvlt/V]) + (R. * gR } (1 - exp [(*._ + ERf)FV2t/V])
The equation symbols are defined in the List of Symbols, pagexii.
11
-------
wind
-, Tfc- '
S
weather stripp
door to house
interior
' , '? ,-Ok,,!' ' • V
,,,^^ ,,:
vents
Figure 1. Typical residential garage
1
t
alr
air
out
.^.,..,. .^ ,-.JT"-
0
-^—r-^t-^f-=^r-
-)—/—A
Figure 2. Typical roadway tunnel
12
-------
25% open
RAMP UP |
RAMP DN |
0 o o
"
— „. „ —
— a . —
0
......
. . 0
0 . 0
i RAMP UP
; RAMP DN
O 0 0 b C
* J
LI
. ol o
Floor Plan (one floor)
. .1 s s A .. .##»i»g
, . '. ...... '...v..'' >...._ '....':.,.-. .. ..'....t~*as.
;_ „ ,,, ; V,V,,,,i ^~ — =a=B
• . -'i. '
aii»" •
'~*%"if'Tlg
^PSJ* ^ c
Wtxrf^t-1
••'1*5?*!
isasAr, j
-'-JirgTi
1^1
tofn-^'i' — ."'."..*'... - .
hi~^, . <*.'.;... .-•>',„„)
to- frt.",V
li; - -,4 "- "-.
Elevation
Figure 3. Typical parking garage
13
-------
While the model has been derived independently by several researchers,
little evidence of validation could be found. The studies that did attempt to
validate the model used cigarettes as the pollution source and carbon monoxide
as the pollutant. The carbon monoxide generation rate of the cigarette, in
general, was not accurately known.
For use with this study, it will be assumed that there is no filtering
of the pollutant, therefore E = 0 and Rf = 0. The equation can then be written:
Q
C = C exp (-
V,
/V) + q (1-expt-tRiF^/V]) +
(1-exp [-tRiFv3/v] )
Note that this equation assumes constant Cif Q, and R. . It is possible to handle
step changes in any of these variables by resetting time equal to zero when
the variables change. For this study the equation will be broken down into
its three component parts to permit easy computation of concentrations for dif-
ferent pollutant rates. The first component is the decay of the ambient concen-
tration:
Cl = Co 6XP f-tRi'Fvi/W
The second component is the concentration caused by the pollutant entering
with the ventilation air.
c2 = c.
- exp(-tRjLFV2/V)
The last component is the concentration caused by the vehicle exhaust.
Q
but: Q = nq
where: n = number of vehicles
q = emissions per vehicle, g/min
therefore:
C3S"F^5R- 1 " exP(-tRi'Fv3/V)|
The total pollutant concentration in the enclosed space is:
C = C
For a given situation, each equation can be calculated separately.
However, for each instance in which time is reset to zero with a change of C.,
Q or R^, it must be reset to zero for all equations„
All of the variables in the equations, with the exception of the
ventilation factor, can be obtained or estimated from the physical situation
to be modeled. The variable termed "effective ventilation factor" (FV), in
this study has been called the "mixing factor" in most studies.
14
-------
It is important to realize that the model algorithm was derived
assuming all pollutants are instantaneously and homogeneously mixed with the
enclosed space air. Since this assumption is not true in real situations, a
factor was introduced into the equation to account for the real situation
behavior. The factor was originally called the mixing factor, and was supposed
to be an indication of how well mixed (i.e., homogeneous) the air was. Because
of this definition, the factor was assumed to vary bewteen 0 and 1.0 (no mixing
to completely homogeneous) . The factor, however, did not distinguish between
mixing and ventilation effectiveness. In some real situations the pollutant
can be removed at a faster rate than if the mixture were homogeneous, requiring
a mixing factor greater than 1.0. The term always appears as a multiplier on
the ventilation rate in the equation. Thus, it is more properly considered
as an effective ventilation factor having a value less than 1.0 when less
pollutant mass is removed than would be removed in the pollutant-room air
mixture were homogeneous, and greater than 1.0 when more pollutant mass is
removed than would be if the mixture were homogeneous. It is interesting to
note that although this equation has been in the literature for approximately
20 years, it is only within the past two years that articles have appeared
which show the correct placement of this factor in the equation. (19,20) An
important part of applying the enclosed space equations will be the defining
of the mixing factor for each situation.
3. Alternate Model for Tunnel Exposure Situations
During the course of this project, the Federal Highway Administration
published the results of an extensive study of air quality management in
highway tunnels. (21,22) This study included a well developed dispersion model
called TUNVEN. This model, while starting with the same mass balance as the
general enclosed space model, considers the special aerodynamics of tunnels.
The TUNVEN computer program was written to obtain numerical solutions for the
coupled one-dimensional steady-state aerodynamic and advection equations
developed for highway tunnels. While this model was not obtained in time
for use in this program, it is recommended for use in any future study of
pollutant dispersion in tunnels. Reference 23 contains a listing and user
instructions for the TUNVEN program.
4. Street Canyons
Street canyons have been standard areas for evaluating mobile source
pollutant contributions for some time, yet the literature on street canyon
models is limited. From the twelve articles on hand (24-35 ) only three
algorithms were found. (25,29,31) Of these three, only two^25'31) gave results
at any point within the canyon. The third (29) was for "average" street canyon
concentrations .
Several investigations 2^) of air flow within street canyons have shown
that when the rooftop wind is perpendicular to the street, a helical vortex
circulation pattern is created in the street canyon, as shown in Figure 4.
As would be expected, studies have found that the upwind (referenced to rooftop
wind) side of the street has higher concentrations than the downwind side.^2^)
For the purposes of this study, it is the higher concentrations that are ini-
tially of more interest. Therefore, initially only equations for the upwind
side of the street canyon will be considered.
15
-------
BUILDING
BACKGROUND
CONCENTRATION
Figure 4. Street canyon exposure situation
16
-------
The algorithm for the upwind side of a street canyon developed by
Stanford Research Institute(25,26) was found in two forms:
7 QT
Reference 25: C =
(U + .5) (S + 2)
7
Reference 26: C =
(U + .5)3
The equation in Reference 26 was said to overpredict the high concentra-
tions and underpredict the low concentrations . The addition of the +2 meters in
the equation from Reference 25 may tend to help the correlation somewhat.
A wind tunnel scale model study using an idealized city was performed
by Wedding at Colorado State University. (31) jje defined a dimensionless con-
centration coefficient applicable to both sides of the street and all wind
directions by the equation:
C U H L
c ~ Q
It is not clear from Reference 31 whether U is the free stream velocity
over the city, roof top velocity, or the velocity in the canyon. There were
several illustrations in the reference showing lines of constant Kc over the
face of the upstream buildings with the wind perpendicular to the street. The
values of Kc at midblock ranged from 19 at 2 meters (full scale height) to 9 at
17 meters (full scale height) . The Kc values varied as an inverse exponential
function of height. If the vehicle emissions are divided by the block length
to obtain emissions per unit length, as in the Stanford model, and H is replaced
by the slant height, S, a new concentration coefficient, K^/ can be obtained:
K .S C U S
C- H - QL
The new KC" varies from 7.49 to 8.38 from 2 to 17 meters height at midblock.
If 7.75, the value of Kc' at the median slant height, is taken to represent
the Kc' values for the whole building face, the values calculated for C
would be approximately 3.5 percent high at 2 meters, and 8 percent low at 17
meters up the face of the building. The equation for C would then be:
c = 7'75
comparing this to the Stanford equation, Reference 25,
/"i —
QL
(U + .5) (S + 2)
it can be said that the wind tunnel testing lends credence to the Stanford
model. Obviously for a given U, S, and QL the wind tunnel equation will
give a higher concentration. Since the Stanford model overpredicts somewhat
using the wind tunnel model would not improve prediction ability. It was
therefore decided that the Stanford equation would be used. It should be
pointed out that the model should be limited in its use to canyons within
17
-------
certain height to width ratios. The EPA guidelines for evaluating
indirect sources(33) indicate that the Stanford equation is only good when
the building height exceeds the penetration depth of the rooftop wind.
Reference 33 gives the penetration depth as:
« - 7(5?)*
u
For most street canyon and wind conditions that would actually be
encountered, the height to width ratios (H/W) would always have to be above
0.3 and often above 0.5. For very narrow or very tall canyons with H/W ratios
greater than 2.0, there is evidence that the Stanford model is not applicable,
since there is not a vortex circulation at street level.( ' ) In a tele-
phone conversation with John Sontowski of the New York City Department of
Environmental Protection, he indicated that he had developed a model for deep
street canyons, but that it was not yet available for general circulation.
5. Expressways
There are two exposure situations involving expressways that are of
interest. The first is the exposure close (within one km) to the downwind
side of an expressway. The second situation is the commuter in traffic on
the expressway.
There are a large number of line source dispersion models for calculating
concentrations downwind of an expressway. 157-45) Two comparisons (42,46) of
predicted to actually measured concentrations indicated that the G.M. model
developed by Chock^2) was j^e most accurate. The form of the chock model is:
Q
c =
u a J'TrT
z V
exp |
1-1/2(1
hO2l
Here U is the effective crossroad wind (not ambient wind speed) and
h0 is the plume center height at distance X from the road. While this model is
similar to Gaussian line source models with winds perpendicular to the road, the
Chock model is developed in such a way that it can be' used in oblique wind
conditions without the use of a 0 . For a complete explanation of the Chock
model, see Reference 42. In addition to the usual parameters used in line
source computer programs, the Chock model requires a term called bouyancy
flux, in m3/sec3. The derivation of the bouyancy flux is covered in
Reference 42.- From the G.M. sulfate experiment, the bouyancy flux, Ft is
taken as:(47)
_ _ 0.052 |number of vehicles \
t -, —
1 1365 I hour lane
18
-------
The second exposure situation involving expressways is that of the com-
muter on the expressway. After reviewing the many line source models available,
it was determined that all are for emissions off the highway, and are not
usable for the on -highway situation. The solution to this situation is best
seen as a relative motion problem. Assume the commuting car is stationary,
and the rest of the world is rushing by at the vehicle speed. Assume further
that the position of the other cars traveling in the same direction does not
change relative to the receptor's car. The situation is then as shown in
Figure 5. The wind speed is the relative wind, combining the ambient wind
and vehicle speed. The vehicles ahead of the receptor are then considered
stationary point sources. Some work has been done to show that, even though
there is a turbulent wake behind each vehicle, the pollutant dispersion is
still Gaussian. ^ > Thus, the standard point source Gaussian plume dispersion
model can be used. The form of the equation to be used is:
C =
y ? •> 22
{exp [-0.5 (Z-hJ /a ] + exp [-0.5 (z-hj /a ]}
*-* z C- z
The standard deviations, ay and az, normally used in the Gaussian
dispersion equation are for tall stacks in open country. Thus, for cars on
a highway, new standard deviations should be determined. These could probably
be determined from References 31, 48, 49 and 50.
From Figure 5 it can be seen that the traffic lanes in the opposite
direction are line sources to the receptor. Thus, the contribution from
these line sources must be added to the point source contributions of the
cars ahead of the receptor. The model chosen for evaluation of expressway
concentrations should be one that is capable of handling both multiple
point and line sources. The EPA model "PAL" is just such a model. (51)
However, in its standard form PAL uses the HIWAY line source model, not the
Chock model desired. It also does not have the capability to calculate the
required relative wind direction and speed. Therefore, a new computer pro-
gram for the on-expressway model was developed. The program, called ONEX,
uses the basic input and structure of the EPA PAL computer program. The
following are the major changes:
a. The horizontal line source subroutine was removed and
replaced with the Chock line source model.
b. Relative wind speed and direction are calculated based
on ambient wind and vehicle speed and direction.
c. Relative wind is used in the calculation of point source
concentrations.
d. PAL subroutines for area sources and curved line sources
are not included.
19
-------
U ambient
U relative
Line sources from
opposite traffic lanes
I
Cars ahead of receptor
car appear as stationar
sources.
Figure 5.
* - receptor car |Plume
Expressway exposure from a relative motion aspect
20
-------
One of the major concerns with the method of determining the on-
expressway concentrations was that the point source dispersion coefficients
used in PAL (and most other point source Gaussian models) would not be correct
for a line of cars on an expressway. Data from three dispersion studies were
used in an effort to determine dispersion coefficients for vehicles in
traffic.(31,49,50) Preliminary evaluation of the data indicated that the
dispersion coefficients obtained from these studies fall between the standard
dispersion coefficients for "neutral" and "unstable" atmospheric stability
classes. Therefore, the standard dispersion coefficients were retained for
the present. Appendix A contains a study to validate the ONEX model using
sulfate levels taken inside vehicles during the G.M. sulfate experiment.
6. Localized Area
Localized area exposure can occur in a variety of situations such
as shopping mall parking lots or sports stadium parking lots, trucking termi-
nals, and construction sites. A search was conducted for models that pre-
dicted concentrations within these areas. Several models were found that
predicted concentrations downwind of area sources, but none were found that
predicted concentrations within the area source itself. Since it was not the
purpose of this project to develop models for the situations studied, no
attempt was made to derive a model for the localized area exposure situation.
C. Model Use
Using the models chosen in the preceding section, ambient concentra-
tions of vehicle emissions can be calculated for a variety of everyday
exposure situations if the emission rate from the vehicles and the physical
situation are defined. The ultimate end use of these ambient air concen-
trations will be as a step in determining if any unregulated mobile source
pollutant poses a risk to public health or welfare. The other steps in the
process are: (1) from knowledge of population activity patterns determine
the amount of time spent in each exposure situation (per hour, day, week,
year or whatever timeframe desired); (2) from average respiratory rates,
exposure time and ambient concentration, determine the total mass of pollu-
tant inhaled for each exposure situation; (3) sum the inhaled mass from each
exposure situation to obtain the total dose of pollutant during the time-
frame chosen; (4) make a judgement as to whether or not the total dose is
a risk to health or welfare.
For an accurate assessment of the health risks, the physical exposure
situations that determine the ambient air concentrations, must be as repre-
sentative as possible. Defining representative situations, whether "typical"
or "worst case," requires gathering a large amount of information concerning
actual physical settings for each situation. In fact, to define the actual
worst case (as opposed to some hypothetical worst case) all physical situa-
tions would have to be investigated. Even if the worst case were found, it
is possible that it would affect so few people that it would be of no use in
defining a health hazard to the general population.
21
-------
Rather than use a "worst case," the concept of a "severe case" is used
in this study. Consideration of meteorology, traffic and physical location
can lead to choices of actual physical situations which tend to maximize
ambient concentrations, and which affect large numbers of people. While
the chosen actual physical location may not be the absolute worst case,
it will certainly produce ambient concentrations approaching a worst case.
Many different vehicle emission species are to be investigated using
this methodology. It would therefore be convenient if the modeling results
for each of the exposure situations were expressed in a form which would
allow determination of the ambient concentrations for each different emission
specie by a simple calculation involving only the emission rate for that
specie. Fortunately, the models chosen in the previous section have the
emission rate as a multiplier to the rest of the equation terms. Therefore,
each exposure situation can be modeled using a unity value for the emission
rate (one gram per mile or one gram per minute, as appropriate). The re-
sulting ambient concentration need only be multiplied by the correct emission
rate to obtain the correct ambient concentration for the desired exhaust
pollutant.
22
-------
IV. RANGE OF PHYSICAL VARIABLES FOR EXPOSURE SITUATIONS
The key to the success of this study is to adequately define the range
of physical parameters of real exposure situations. With the range defined,
typical and severe situations can be properly chosen. As work on the pro-
ject progressed, it became evident that definition of the range of physical
parameters for the various situations was not a trivial task. In no case
was the information readily and completely available in the literature.
For each exposure case, several agencies or or organizations had to be
consulted. Often the information obtained was incomplete. In some cases
the information required was not available at all. Where necessary, judgement
was used to reconcile apparently conflicting statistics. The descriptions
presented here are the most complete that could be assembled after a
diligent search and evaluation of the information sources.
A. Residential Garage
The residential garage is a common situation for exposure to automotive
pollutants. There were approximately 56.8 million single-family homes in
the U.S. in 1980.(52'53) Approximately 60 percent of these homes had
garages,(53,54) for a total of 34.1 million residential garages in the U.S.
in 1980. Note that this figure does not include garages for multi-family
dwellings such as apartments or condominiums. Statistics on the number of
garages with multi-family dwellings could not be found. Therefore, the total
number of garages given above should be considered a minimum estimate.
The number of persons encountering a residential garage exposure each
day is difficult to estimate. Not all garages are used to house cars,
having been converted after construction to additional living space. In
multi-car families, the most used car may not be housed in the garage.
However, if it is assumed that 50 percent of the garages house a car that
is used daily, the number of persons exposed to this situation each day
is at least 17 million.
No nationwide description of residential garage characteristics was
found. Therefore characteristics of residential garages built in the San
Antonio area will be assumed to be typical of garage construction nationwide.
According to local building codes, a single .car garage must be at
least 10 ft wide, 20 ft deep and have at least 8.42 ft ceiling height.
Typically, an additional 6 ft in depth is provided for storage space and for
appliances, such as washer, drier, and water heater. Nominal auto-access
doorways or garage door must be at least 8 ft wide by 7 ft high. The type
of door used varies with individual contractors or architectural preference.
Since the garage is specifically for auto storage at least 288 square inches
of "free air vent" must be provided for either a single or double car garage.
Most garages are joined to the house structure with at least one wall common
to the interior of the house, through which access to the garage from the
interior is made.
23
-------
Locally, most residential garages incorporate an "overhead door" for
the garage door. This door is made of hinged sections on guide tracks which
allow it to be lifted vertically and "rolled" up and back, parallel with the
ceiling. Some garages are not finished except for walls which are common to
interior living space. These common walls are generally insulated and some-
what fire resistant, as prescribed by building codes. If other walls are
finished, insulation is not generally used. Some garages also have an open
ceiling in which the structural members of the roof are exposed, while others
have finished ceilings. Among appliances mentioned, the water heater is
generally located in the garage space. The gas water heater is enclosed in
a "sealed closet" with three vents, one to supply air for combustion, one to
prevent buildup of natural gas in case of malfunction and one which serves
as a flue for combustion products. (55)
The residential garage exposure is, in general, extremely short term.
A typical exposure scenario might be as follows: receptor enters garage,
opens garage door, starts car and allows 30 seconds to warm-up at fast
idle, backs car out of garage, returns, closes garage door and leaves. A
severe exposure might be as follows: on a cold winter morning, receptor
enters garage, opens garage door, starts car and returns to inside of home
while car warms up at idle. After approximately five minutes receptor
returns to garage, backs out car, closes garage door and leaves. The
garage used for both these situations will be a single car garage.
B. Parking Garages
Another source of exposure to automotive pollutants is the parking garage.
Parking garages are growing in popularity in crowded downtown areas, high rise
office and apartment buildings, and shopping malls. Despite this growing
popularity, statistics on parking garages are difficult to locate.
Two primary sources of older information are available. The first is
the Highway Research Board Special Report 125 "Parking Principles," dated
1971.(56)' This report analyzes several studies done between 1960 and 1968.
The information presented in Table 2, showing the average number of garage
spaces and garage spaces as a percent of total spaces, is from this publi-
cation. To obtain the average number of garages, it was assumed that for
all "urban places" over 250,000 people, the average garage size is 500 spaces
and below 250,000 people, the average garage size is 300 spaces. The total
number of garages in the United States using the 1970 census listing of
"urban places," circa 1965 was approximately 1300, as shown in Table 3.
The second source is a parking study conducted by the National League of
Cities in the early seventies, and published in 1972 under the title "National
Parking Facility Study.(57) This study relied on a mail questionnaire for
information. Thus, not all needed information was collected. For instance,
the data on number of parking garages was collected by city. However,
there was no data for New York, Chicago, Los Angeles, or Philadelphia -
the four most populous cities in the country according to the 1970 census.
The study reported that there were an estimated 2040 public (either muni-
cipally or privately owned) parking garages in the country.
24
-------
TABLE 2. AVERAGE NUMBER OF PARKING GARAGES BY URBANIZED AREA SIZE
Population Group
of Urban Area
10,000-25,000
25,000-50,000
50,000-100,000
100,000-250,000
250,000-500,000
500,000-1,000,000
Over 1,000,000
TABLE 3.
No. of
Cities
Studied
6
6
30
33
16
15
5 3,
Garage
Average Space Avg. No.
Pop. of as % of of Garage
Sample Total Spaces
17,000 0 10
37,000 3 140
68,000 5 260
160,000 11 820
360,000 16 1,940
720,000 30 6,900
700,000 31 18,600
Average
No. of
Garages
0
0.4
0.8 '
2.8
4.0
13.6
36.5
TOTAL PARKING GARAGES BY URBAN AREA SIZE
CIRCA 1965
Population No. of Urban
Group of Places in
Urban Places Group a
25,000-50,000
50,000-100,000
100,000-250,000
250,000-500,000
500,000-1,000,000
Over 1,000,000
520
240
100
30
20
6
Average No.
of Garages
0.4
0.8
2.8
4.0
13.6
36.5
Total U.S.
Total
Garages
208
192
280
120
272
219
1291
per 1970 Census
25
-------
The raw data from the study^57), which was included in an appendix
of the study, was reanalyzed for this project using Standard Metropolitan
Statistical Areas (SMSA's) rather than cities. Included in these data
were the number of parking garages in New York-, Los Angeles, Chicago,
Philadelphia and Houston, taken from counts of listings in the Yellow
Pages of the various telephone directories. The results of this analysis
are shown in Table 4. The total of 2653 garages, should be taken as
representing the public parking garage population in 1972. This value
would represent an increase of 194 garages per year between 1965 and 1972.
To bring the garage population estimate up to 1980, it is necessary
to examine the construction levels since 1972. From a search of "Parking"
magazine issues between 1972 and 1970, construction totals reported by the
National Parking Association were obtained for 1972, 1974, 1978 and
1979.(58,60) These construction totals are shown in Table 5.
If the annual construction for the years when data are missing is
estimated to be equal to the lowest annual total found (i.e. 302 in 1978),
then the total number of parking garages built between 1972 and 1979 was
2647.
In a discussion with engineers at Allright Auto Parks, Inc., one of
the largest operators and parking garage consultants in the country, it
was learned that relatively few parking garages are taken out of service
each year. Expressed as a percentage of new garage construction, the number
would be much less than ten percent. Therefore, for all practical purposes
the number of garages constructed each year can be taken as the garage
population increase for the year. Adding the number of garages built since
1972 to the number of parking garages at the start of 1972, (2653, as presen-
ted in the last progress report) the estimated 1980 garage population of
5300 is obtained.
An effort was made to obtain parking garage construction figures from
the F.W. Dodge Division of McGraw-Hill. "Parking garage" is one of the
categories for which they compile construction reports. However, the cost
to obtain reports of parking garage construction over the past eight years
was considered too large an expenditure for the information obtained, and
the matter was not pursued further. However, the Dodge estimate of nation-
wide parking garage construction in 1979 was obtained. Dodge lists 720
total parking garage projects in 1979 with a total floor area of 55 x 106
square feet. Note that this is more than twice the projects reported by
the National Parking Association (NPA) in "Parking" magazine.
This discrepancy in the number of parking garages is reconcilable when
the sources of information are considered. The NPA is an association of
parking garage designers and operators. The NPA construction estimates are
dependent on reports from their members. These companies would likely be
involved in the larger projects intended for use by a large segment of the
general public. Small projects associated with banks, office building,
26
-------
TABLE 4. TOTAL PARKING GARAGES FOR STANDARD
METROPOLITAN STATISTICAL AREAS-CIRCA 1972
SMSA Size Group
50,000-100,000
100,000-250,000
250,000-500,000
500,000-1,000,000
1 , 000 , 000-2 , 500 , 000
Over 2,500,000
Total
No. of SMSA
in Group ( a)
27
111
60
33
25
8
264
Average No.
of Garages
4.0
10.6
8.1
32.8
60.8
Total
Garages
0
444
636
267
820
486
2653
TABLE 5. PARKING GARAGE CONSTRUCTION
BETWEEN 1972 AND 1979
Number of
Year Projects Reported
1972
1973
1974
1975
1976
1977
1978
1979
Total
451
h
(302)
363
(302)
(302)
(302)
302
323
2647
Reported the National Parking Association as
.listed in Parking magazine.
Numbers in parentheses are estimates using the
lowest number from the 1972, '74, '78, and '79
reported projects.
27
-------
apartments or universities would not always be reported. F. W. Dodge on the
other hand, through its network of field reporters, endeavors to report on
all construction. Dodge is therefore more likely to include smaller pro-
jects not reported to the NPA. This likelihood is supported by examining
the total floor area of the 720 projects reported by Dodge. If the
floor area per car for a self-park garage is taken as 300 sq. ft/"-'-',
the 55 X 106 sq. ft. total floor area yields an average of 244 spaces
per garage. This is approximately half the size of the average garage
reported by the NPA. Apparently, the Dodge estimate includes a large number
of smaller private parking garages, not reported by the NPA. Assuming
this discrepancy has always existed, it is entirely possible that the
number of parking garages in the country is actually twice the 5300 estimated
from the NPA data. The conclusion to the analysis of this potpouri of
information is that there are currently at least 5300 parking garages in
the country and quite possibly as many as 10,000 parking garages.
Most literature on parking garages quotes, without supporting data,
400 to 500 spaces as the average size of a parking garage. (56,61) As can j-,e
seen from the 1979 Dodge data quoted above there is some evidence that when
all types of parking garages are considered the average size may be between
200 and 300 spaces. The only complete size distribution data that could
be found was the July 1973 issue of "Parking" magazine.(58) of the 451
parking garage projects reported to the NPA in 1972, 381 included number of
spaces. The size distribution of these garages is shown in Figure 6. A
partial distribution for 1979 construction was contained in April 1980 issue
of "Parking".(6°) This partial distribution is also shown in Figure 6. Note
that the 1979 distribution is similar to the 1972 distribution. The median
garage size for this distribution is approximately 600 spaces. If the
garages were assumed clustered at the midpoint of each interval, the mean
would be approximately 740 spaces. Thus, depending on which figures are
being used, and how "average" is defined, an average parking garage could
contain anywhere from about 250 to 740 spaces. Given this variation, using
400 to 500 spaces as an average parking garage size does not seem unreasonable.
If the figures derived from NPA data are used for number of garages
(5300) and average size is taken as 500 spaces, then the number of national
daily parking garage exposures would be a minimum of 3.7 million people;
assuming the number of cars using a garage during a day is equal to the
number of spaces, and that there are 1.4 persons per car.
References 56 and 61 state that most garages are above ground, and that
most of these are of open side design. Thus, ventilation for the majority
of garages is natural and affected by local microscale climatic conditions.
Reference 62, which modeled diesel particulate in parking garages, states
that 0.5 mph wind through an open side garage is approximately equal to one
CFM per square foot of floor area.
Mechanical ventilation is required in enclosed parking garages which
are mainly underground garages. Only 46 of 451 (10.2 percent) garage
construction projects reported by the NPA in 1972 were wholly underground,(58)
requiring mechanical ventilation. If this fraction were extrapolated to the
entire garage population, there would be between 500 and 1000 mechanically
28
-------
80
16
o
0)
g
0) 1 9
& J-^
-------
ventilated garages. The amount of ventilation used depends mainly on local
building codes. Most building codes require four to six air changes per hour.
The New York City Building Code required a minimum ventilation rate of one CFM
per square foot of garage floor area. Using the recommended floor-to-ceiling
height(56) of seven feet, this rate produces approximately eight air changes
per hour.
Actual ventilation rates are difficult to obtain. A 1961 article in the
American Industrial Hygiene Journal(63) and a 1973 article in "Parking"
magazine(64) did list some ventilation rates for underground garages. These
rates are listed in Table 6. The Los Angeles Music Center Garage ventilation
rates were obtained from Hayakawa Associates of Los Angeles, who were con-
tracted to revamp the Music Center garage ventilation. This work was com-
pleted in 1980.
Operational factors can also have effect on emission concentrations in
a parking garage. In an article which discussed CO emergency situations in
parking garages(64) f the main cause of these situations was identified as a
backup at the exit lane (or lanes). These backups occurred when, .for one
reason or another, all cars tried to exit at the same time (end of a parade,
work day, etc.). If there were insufficient exit lanes, and if the streets
onto which the cars exited could not handle the traffic, a backup occurred.
In addition, situations which require parking payment upon exiting slowed down
the exit process considerably. To avoid this, many garages charge a flat
rate per event, collected upon entrance to the garage.
The information discussed above can be used to define the typical and
severe exposure cases. The typical garage is an above ground garage, with
open sides for natural ventilation and a capacity of 400 to 500 cars. The
Convention Center parking garage in San Antonio, Texas meets these criteria
and was chosen as the typical parking garage exposure case. A complete de-
scription of' the garage is included in Section V, where the ambient concen-
tration in the garage is estimated using the enclosed space model.
The severe exposure case would be a garage with a capacity of 1500 or
more cars, mechanically ventilated, with a ventilation rate of under 300 cfm/
space (or 4 air changes per hour), and often used for single events. The
parking space-to-outbound lane ratio should be greater than 250 spaces per
lane, and it should exit onto a crowded street.
During the investigation of possible severe case parking garage exposures,
it was discovered that the Music Center Garage in Los Angeles was undergoing
extensive modification because of poor ventilation. A comprehensive study
was done on the garage in its original configuration to determine the necessary
improvements. The original garage ventilation, at four air changes per hour,
was within, but at the low end of, the building codes at the time of its
construction. As can be seen from Table 6, the garage has a capacity of
1582 spaces and is often used for event type parking. Since considerable
study has already been done on the garage, there was the possibility of
obtaining much more comprehenisve data than would usually be available.
After Mr. Robert Magusin, Project Manager in the County Engineer's Office
and Mr. H. Endo of Hayakawa Associates, the County's Engineering Consultant
30
-------
TABLE 6. VENTILATION RATES FOR SOME MECHANICALLY VENTILATED PARKING GARAGES
Name of Garage Total
and No. of Total Capacity Ventilation Air Changes
Location
Grant Park
Chicago, Illinois
Pershing Square
Los Angeles, Calif.
Union Square
San Francisco, Calif.
Grand Circus Park
Detroit, Mich.
Military Park
Newark, New Jersey
Mellon Square
Pittsburgh, PA.
Brooklyn Civic Ctr.
New York City, N. Y.
Lincoln Center
New York City, N. Y.
Music Center
Levels
2
3'
4
2-3
3
6
3
2
8
No. of Spaces a
2,359
2,000
1,700
1,100
1,031
897
700
610
1582
CFM
1,766,400
488,000
390,000
540,000
400,000
245,340
240,000
277,000
340,000
CFM/Space
750
244
230
490
388
273
343
454
215
Per/Hr.
16.8
6.3
6.0
11.3
10.0
7.8
8.3
8.1
4.0
CFM/FT2
2.2
0.9
0.9
1.5
1.15
1.0
1.14
1.11
0.6
Los Angeles, California
Floor area per space ranges from 270 to 410 sw. ft./space
Source: (1) Journal of the American Industrial Hygiene Ass'n., Dec. 1961
(2) Parking magazine, Oct. 1973
-------
for the work, were contacted to ensure that the data, was available, the Los
Angeles Music Center Parking Garage was chosen as the severe exposure case.
A complete description of the garage is included in Section V, where the
ambient concentrations in the garage are estimated using the enclosed space
model.
C. Roadway Tunnels
Roadway tunnels are another enclosed space where people are exposed to
automotive pollutants. It was decided to limit the investigation to tunnels
over 500 feet (152 m) in length; since even at a dead crawl of five miles per
hour, it would take only 68 seconds to traverse a 500 foot tunnel. While length
is an indication of exposure time, vehicle speed through the tunnel also
determines exposure time. The pollutant concentration obviously increases
as ventilation rate decreases and number of cars in the tunnel at any given
time increases. Thus tunnel length, vehicle speed, traffic density and
tunnel ventilation rate must also be considered when chosing typical and
severe exposure situations.
Several sources were used to compile a list of tunnels over 500 feet
long. "Highway Tunnel Operations" by the Transportation Research Board(65)
contains a list of mechanically ventilated tunnels over 500 feet long. A
list of toll tunnels(66) Was obtained from Mr. Tom Benedict of the FHWA in
Washintgon, and an "unofficial" list of road tunnels was obtained from
Mr. Jim Washington of the FHWA Bridge Division, also in Washington.
Additionally, a list of road tunnels in Colorado was obtained from
Mr. Burrell Gerhardt of SwRI, who was previously with the Planning and
Research Division, Colorado Division of Highways. Mr. Bill Rush of the
International, Bridge Tunnel and Turnpike Association in Washington, D.C.
was also contacted regarding toll tunnels.
From this information, a list of 105 road tunnels was compiled.
Parallel tubes serving the same road were counted as one tunnel even
if they were of different lengths or were built at different times or
had different names. The list, included in this report as Appendix B,
should not be taken as a complete, official list. It probably contains
at least ninety percent of the tunnels in the country. Thus, a good
estimate of the number of road tunnels over 500 feet long in the U.S.
would be under 120.
The number of tunnels in the U.S. will probably not increase greatly
over the next 20 years. Tunnels are more expensive to build and more
expensive to maintain than other highway elements.(65) Thus, they tend
to be used only at locations where thorough study indicates such costs
can be justified. Currently, several (four to six), tunnels are in the
planning stage for 1-70 east of Glenwood Springs in Colorado.(67,68) In
addition, Reference 66 lists seven possible toll tunnels under consideration.
Of these only two are new locations; the remaining five are parallel tubes
for existing tunnels.
32
-------
Figure 7 shows the distribution of tunnels by state. Note that the
majority of the tunnels are concentrated in a few states. The tunnels
were divided into two classes, intercity and commuter, depending on their
probable use. Intercity tunnels were those carrying primarily highway
traffic (noncommuter) between cities. Commuter tunnels were those used
for daily commuting to and from areas of high population density. Of the
105 tunnels, 59 were classed as commuter tunnels and 46 as intercity
tunnels. The tunnels were classified by length categories for both
tunnel types. Figure 8 is a bar graph of the number of tunnels in each
length category for both commuter and intercity tunnels. Note that for
both tunnel types, over half of the tunnels are less than 2000 feet long.
The average daily traffic (ADT) was available for 47 of the tunnels
(29 commuter, 18 intercity). The tunnels were classified in terms of
ADT/lane for both commuter and intercity tunnels. Figure 9 is a bar
graph of the number of tunnels in each ADT/lane interval. For all
the intercity tunnels, the ADT/lane was 10,000 or less. By contrast,
over half of the commuter tunnels had an ADT/lane of over 10,000. From
this examination, it can be seen that not only are commuter type tunnels
more numerous, they also represent the majority of the exposure
"events" (considering one vehicle going through a tunnel one time as an
exposure "event"). Thus,only commuter type tunnels will be considered
in chosing typical and severe tunnel situations.
An estimate of the number of persons encountering a tunnel exposure
situation per day can be obtained from the number of tunnels and the ADT.
Assume that the list of 59 commuter tunnels compiled for this project
contains all the commuter type tunnels in the country, that the average
commuter tunnel ADT is approximately 52,000, and finally that the average
car carries 1.4 persons.(69) The total number of tunnel exposures .is
then approximately 4.3 million per day. If it is assumed that persons going.
one way through the tunnel in the morning return through the tunnel in the
opposite direction in the evening, then 2.15 million people encounter tunnel
exposure twice a day.
An excellent summary of the relationship between traffic speed,
density and rate (vehicle/hr) for tunnels is contained in Reference 21.
Figure 10, which is Figure 10.2 in Reference 21, shows the relationships
based on Lincoln Tunnel data.(22) The Lincoln Tunnel data plus data
taken during the study reported in Reference 21 indicate a practical peak
flow limit of about 1500 to 1600 vehicles/hr per lane, with a corresponding
traffic density of 125 vehicles per mile per lane.
Depending on the traffic control of the particular tunnel, during
rush hour the traffic density may or may not increase above peak flow
toward congested flow. Obviously, from an exposure standpoint, congested
flow causes the most severe exposure, both because it increases the
traffic density and because it decreases the vehicle speed. It should
be pointed out that it is necessary to know both traffic rate and speed
to determine if a tunnel is in the congested mode. The tunnel with the highest
ADT/lane will not necessarily provide the most severe case from a traffic con-
gestion standpoint.
33
-------
no tunneIs
1-3 tunnels
4-6 tunnels
10-12 tunnels
> 12 tunnels
Figure 7. Road tunnel distribution in the United
-------
22-i
U)
Ln
20H
18 H
16H
14H
(U
C
10-
8-1
o-i
Commuter Tunnels
Intercity Tunnels
500 1000 2000 3000 4000 5000
to to to to to to
999 1999 2999 3999 4999 5999
Tunnel length, feet
6000
to
6999
7000
to
7999
8000
to
8999
9000
to
9999
Figure 8. Tunnel length distribution
-------
Commuter Tunnels
12H
0)
c
0
fc
0)
20,001
to
25,000
Average daily traffic per lane
Intercity Tunnels
25,001
to
30,000
Figure 9. Average daily traffic distribution
-------
0)
ft
CO
0)
H
o
•H
I
Jam
Density
101
100 200 300
Density (Veh/Mile For 2 Lanes)
(a) Average Speed Versus Density
500
3000 L
Congested
Flow
100 200 300
Density (Veh/Mile For 2 Lanes)
(b) Flow Rate Versus Density
Uoo
500
Figure 10. Relationships between traffic rate, vehicle speed
and traffic density
37
-------
Tunnel ventilation has been studied extensively/ with Reference 22
providing an excellent survey of the entire subject. There are four
principle tunnel ventilation systems used in the United States. They
are:
• Natural
• Longitudinal
• Semi-transverse
• Transverse
Figure 11 shows examples of each of these systems.
From Reference 22 and other sources, the distribution of commuter
tunnels by length and type of ventilation was obtained for 41 of the 59
commuter tunnels. This distribution is shown in Table 7. In general,
natural ventilation is used in tunnels under 1000 feet long, longitudinal
ventilation for tunnels up to about 2000 feet long, semi-transverse in
medium length tunnels up to 5000 feet, and transverse for tunnels longer
than 5000 feet. Maximum ventilation rates for some U.S. tunnels are given
in Table 8. The best way to compare ventilation between tunnels is to look
at ventilation volume per second per meter of lane length. On this basis,
the tunnels in Table 8 vary from a low of 0.073 m3/sec per lane meter for
the Brooklyn-Battery Tunnel to a high of 0.275 m-^/sec per lane meter for
the Eisenhower tunnel in the Colorado Rocky Mountains. It should be pointed
out that the Eisenhower tunnel was purposely designed with higher than usual
ventilation capacity to account for the higher than normal CO emissions of
vehicles at high altitude.
TABLE 7. COMMUTER TUNNEL VENTILATION TYPE BY TUNNEL LENGTH
Type of Ventilation
Semi-
Length Natural Longitudinal Transverse
500-999 151
1000-1999 244
2000^2999 1 2
3000-3999 5
4000-4999 1
5000-5999 1
6000-6999
7000-7999
8000-8999
9000-9999
Total 3 11 13
Transverse
1
1
3
1
2
1
2
2
1
14
Total
7
11
4
8
2
3
1
2
2
^
41
38
-------
Natural
Longitudinal
1
nq
1/72
-r-—/•—=A
:;S_;_
Transverse
Semi-Transverse
Figure 11. Types of Tunnel Ventilation
39
-------
TABLE 8. MAXIMUM VENTILATION RATES FOR SOME U. S. ROADWAY TUNNELS
Vent. Vent./lane
Lanes/ Type of per tube meter (m3/sec)/
Length, No. of
Tubes
2
2
2
1
1
1
2
1
1
2
2
1
2
2
3
2
1
1
S = semi-transverse; L = longitudinal; T = transverse
second number is for downhill tube
°excluding Eisenhower, which is greater because of tunnel altitude (11,000 ft)
Tunnel
Lowry Hill
Squirrel Hill
Fort Pitt
Bankhead
George Wallace
Baytown
Allegheny Mt.
Caldecott #3
Sumner
Queens/Mi dtown
Brooklyn-Battery
Detroit-Windsor
Baltimore Harbor
Holland
Lincoln
Eisenhower
Thimble Shoals
Hampton Roads
meters
456
1288
1097
948
914
917
1850
1027
1724
1933
2779
1564
2332
2580
2408
2725
1749
2280
Tube Vent .
3
2
2
2
2
2
Average
2
2
2
2
2
2
2
2
2
2
2
2
Ave rage
S
S
S
S
S
S
for
L
T
T
T
T
T
T
T
T
T
T
T
for
,a m3/sec
252
275/175b 0
380
260
345
425
semi-tr an verse
566
232
289
330
405
380
357
884
839
1500
420
400
transverse0
lane meter
0.184
.10 7/0. 06 8b
0.173
0.126
0.189
0.232
0.154
0.153
0.113
0.084
0.085
0.073
0.121
0.077
0.171
0 . 174
0.275
0.120
0.088
0.111
Sources: (1)
(2)
(3)
Schlang, R. N., et al. "Management of Air Quality In and Near
Highway Tunnels, Interim Report," Federal Highway Administration
Report FHWA-RD-78-184 January 1980.
Rodgers, S. J., et al, "Tunnel Ventilation and Air Pollution
Treatment," Federal Highway Administration Report FHWA-RD-72-15,
June 1970.
Personal Communication. P. L. Chandler, District Maintenance
Engineer, District 5, Minnesota Department of Transportation
to Melvin Ingalls, Southwest Research Institute, San Antonio, TX,
dated August 28, 1980.
40
-------
With the information presented above, the range of physical variables
for the tunnel exposure situation are well defined. The typical tunnel
exposure situation would be a 1000 to 2000 foot tunnel, with average daily
traffic of between 10,000 and 15,000 vehicles per day. Ideally the venti-
lation system would be longitudinal or semi-transverse, with a ventilation
rate close to 0.17 m^/sec per lane meter. The severe exposure situation
would be a tunnel over 5000 feet long, with ADT in excess of 15,000 vehicles
per day. Tunnels of this length are generally transversely ventilated. If
this is the case, it should have a ventilation rate less than 0.11 m /sec
per lane meter.
The Lowry Hill tunnel in Minneapolis, Minnesota was chosen as the
typical exposure situation. This tunnel has two tubes with three lanes
per tube and is 456 meters (1496 ft) long. It has an ADT per lane of
12,000 vehicles. For commuter type tunnels this ADT per lane is about
average. The ventilation is semi-transverse with a ventilation rate of
0.18 (m3/sec)/lane meter. The ventilation rate and ADT were obtained from
the District 5 Office of the Minnesota Department of Transportation. As
shown in Table 7, the type of ventilation (semi-transverse) used in the
Lowry Hill Tunnel is typical. The ventilation rate of 0.18 (m3/sec)/lane
meter is slightly higher than average for semi-transverse tunnels, as can
be seen from examination of Table 8.
For the severe case, the Baltimore Harbor Tunnel was chosen, There
are only three of the 59 commuter-type tunnels longer than the Baltimore
Harbor Tunnel, making it one of the longest tunnels. Only five of the 29
commuter tunnels for which traffic counts were available had higher ADT/lane
than this tunnel's ADT/lane of 16375 vehicles. The ventilation rate is
also one of the lower rates for transversely ventilated tunnels, as shown
in Table 8. Complete descriptions of both tunnels are included in Section
V where the ambient concentrations for the tunnels are calculated.
D. Urban Expressways
The "MVMA Motor Vehicle Facts and Figures '80," lists the completed
urban miles of the Interstate Highway System as 9678 miles.(7°) Thus, there is
no doubt that the expressway commuter exposure situation is important. To
define the typical and severe exposure situations, the top 76 SMSA's in the
1970 census were investigated for expressway characteristics, ADT on
various expressways, and climatic characteristics of the area. These 76
areas, with a 1970 population of 107.5 million, contained 53% of the 1970
U.S. population. The climatic data for the U.S. and most of the 76 SMSA's
were obtained from the National Climatic Center in Asheville, NC.
The first step in defining typical and severe urban expressway exposures,
is to define the number of miles of urban expressways. Table 9 lists the 1978
urban interstate mileage for each state. This information was taken from
Table INT-11 of Reference 71. However, interstate highways are not the only
urban expressways. In 1975 there were 8671 miles of urban interstate expressways
and 6480 miles of other urban expressways. (7.2) in 1978 there were 9678 miles
of urban interstate expressways.(71) Assuming the same percentage increase in
other urban expressways, the 1978 mileage for this category would be 7232.
This would bring the total estimated urban expressway mileage in the country
in 1978 to 16,910 miles.
41
-------
TABLE 9. URBAN INTERSTATE HIGHWAY SYSTEM MILEAGE
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Dist. of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Miles of
Urban Interstate
202
151
93
864
111
220
23
30
444
238
34
45
457
263
138
139
132
151
28
199
332
426
165
88
291
State
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Miles of
Urban Interstate
37
33
32
45
258
40
534
194
29
656
138
76
339
67
79
21
188
870
106
7
241
229
71
94
30
Total 9,678 Miles
42
-------
A joint frequency distribution of number of lanes and average daily
traffic (ADT) would be ideal for determining expressway characteristics for
the typical and severe exposure situation. However, after studying publi-
cations of the Federal Highway Administration (FHWA), and talking with
personnel at FHWA, it was determined that useable joint frequency distri-
butions-do not exist. Here "useable" means one.that defines number of lanes
up to 10 or more total traffic lanes and ADT up to 200,000 vehicles or more.
Joint frequency distributions listed in "Highway Statistics, 1978"(7D have
the highest categories as: "4 or more lanes" and "40,000 or more vehicles-",-
thus placing almost all expressway miles in one category. Since joint fre-
quency information was not available, separate distributions .of expressway
miles by number of lanes and by ADT were sought.
Information on miles of urban expressway by number of lanes was not
available in sufficient detail from any of the FHWA publications. However,
this information was retrieved from computer tapes of state highway
statistics by the FHWA Highway Statistics Division for this project.(73)
The printout was by state for 31 of the 50 states. Not all states were
included because of differences in definitions and format of some reported
statistics from some states. Unfortunately, many of the states for which
information was not provided are states with large urban interstate mileages.
To improve the completeness of the table, this information was obtained
directly from the states of California and New York. Figure 12 is a
histogram of the percent of urban interstate highway mileage by number of
lanes for the 33 states for which information is available. Note that
over half of the urban interstate highway mileage consists of four lane
roadway.
Using just the interstate urban arterial figures from "Highway Statis-
tics, 1978," the mean 1978 ADT for urban interstate arterials can be cal-
culated as 47,664 vehicles. A complete distribution of the traffic volume
is not available. However, a partial distribution for the urban interstate
highway system is listed in Table INT-15 of "Highway Statistics, 1978."
Figure 13 is a histogram of ADT taken from that table. As can be seen from
the figure, the median ADT for the urban interstate system is between
30,000 and 40,000 vehicles. It is estimated that rush hour traffic counts
are approximately nine percent of ADT counts.(74) The mean rush hour
traffic is then about 4766 vehicles/hour.
From the above information on total miles of expressway and rush hour
ADT, an estimate of the national total number of persons encountering a
freeway exposure situation can be calculated. The average trip length is
taken as 8.9 miles(70) with 5 miles of the trip on an expressway, and the
average vehicle occupancy as 1.4 persons per vehicle for home-to-work
traffic.(74) The number of persons exposed daily to rush hour expressway
conditions in 1978 was approximately 22.6 million persons.
43
-------
NOTE:
For 33 states only. Does not include: Arizona, Connecticut,
Dist. of Columbia, Florida, Georgia, Illinois, Kansas, Mary-
land, Massachusetts, Michigan, Missouri, New Hampshire, New
Jersey, North Dakota, Oklahoma, Rhode Island and Wyoming.
60-,
•H
S
+J
£
W
c
H
I
+J
c
o>
u
50-J
40 H
30 J
20 H
10 H
o-1
2 3S4 5s6 7&8 9& over
Number of Lanes
Figure 12. Urban interstate highway distribution
by number of lanes
44
-------
0)
en
a
.
CO
rt
4-1
M
^
0)
4-1
C
•s
D
ifl
4J
4-1
C
0)
o
20
18
16
14
12
10
8
6
4
2
0
_
-
-
•
-
-
44% less than 30,000
59% less than 40,000
over 40,000 = 41%
10,000 20,000 30,000 40,000
Average Daily Traffic (ADT), vehicles
Figure 13. Average daily traffic distribution
for the urban interstate system
45
-------
No additional information on traffic count distribution was found. In
an attempt to generate a better traffic count distribution, a computer listing
of the annual average ADT from the FHWA permanent traffic count locations in
all states was obtained from FHWA. Of the approximately 3500 locations, 362
were on urban interstate highways. How representative these locations are is
unknown. In fact, of the 362 urban interstate locations, only 3 had ADT levels
above 100,000 vehicles. A check of the 1978 Texas Traffic County Map, however, shows
numerous locations in Houston, Dallas and San Antonio with ADT in excess of
100,000 vehicles. Since it appears that the permanent FHWA counter locations
do not adequately represent the higher levels of ADT, this information was not
used to determine the ADT distribution.
Some idea of maximum ADT for various numbers of lanes can be obtained
by considering highway capacity. Expressway capacity is usually taken to
be 1800 vehicles per lane per hour.^74^ Rush hour traffic is generally
around 9 percent of total daily traffic.^ ' Using these two facts the
capacity of expressways of various numbers of lanes are shown in Table 10.
This information, together with the ADT distribution presented in Figure 13,
provides enough information to determine the expressway characteristics for
the typical and severe exposure situations.
Using Figures 12 and 13 and Table 10, a severe exposure situation would
be one with an expressway of 10 lanes (5 in each direction) with an ADT of
approximately 180,000 vehicles. The typical exposure siutation would be on
a four lane expressway. The ADT for a typical exposure siutation is some-
what harder to define. However, if it is assumed that the distribution
shown in Figure 13 is primarily for four lane expressways up to 40,000
vehicles, then Figure 13 can be used to determine a typical four lane
expressway 'ADT.
The validity of the assumption that the ADT distribution for ADT less
than 40,000 vehicles shown in Figure 13 can be used as the distribution on
four lane expressway rests on several considerations. The maximum ADT for
a four lane expressway is taken as 72,000 vehicles. Expressways greater than
four lanes are not generally built unless the ADT for the desired "level of
service" exceeds that of a four lane expressway. The highest level of service
normally used is an hourly vehicle count equal to 0.35 times the highway
capacity.(74) This means that few expressways of six or more lanes will have
ADT levels below 38,000 vehicles. Thus, the 59 percent of Interstate mileage
with ADT below 40,000 vehicles, shown in Figure 13, must include only a small
percentage of expressways with more than four lanes. With this assumption,
and with 55 percent of the total urban expressway mileage being four lanes,
and 59 percent of the total urban expressway ADT being less than 40,000 vehicles.
there can only be a small percentage of four lane expressway with ADT greater
than 40,000 vehicles. Thus, the part of the total ADT distribution between 0
and 40,000 vehicles per day shown in Figure 13 can be used as the ADT distri-
bution for four lane expressways. It appears then, that a typical ADT for a
four lane expressway would be between 20,000 and 30,000 vehicles.
To completely define the expressway exposure situations, information on
46
-------
wind speed and direction is also required. Wind direction and speed distri-
butions were investigated for 76 of the most populous standard metropolitan
statistical areas (SMSA). While these distributions were sought for the rush
hour time period of 7 to 8 a.m., only the wind speed distribution was found
by time of day; wind direction distributions found included all hours of the
day.(75,76)
TABLE 10. ESTIMATED CAPACITY OF URBAN INTERSTATE
EXPRESSWAYS BY NUMBER OF LANES
Estimated Maximum
Number of Lanes Average Daily Traffic5
4 72,000
6 108,000
8 144,000
10 180,000
12 216,000
Estimated from data in Table 34 and Table 50, "Quick-
Response Urban Travel Estimation Techniques and Trans-
ferable Parameters, Users Guide." National Cooperative
Highway Research Program Report 187.
Annual wind direction distributions for all hours was available
in the 16 compass point increments from either Reference 75 or Reference 76
for 62 of 76 SMSA. The prevailing wind direction and percent of time the
wind is from that direction is shown in Table 11 for the 62 SMSA. There
are 16 compass directions plus the category "calm" for a total of 17
possible wind direction categories. If there were an even distribution,
each category would occur 5.9 percent of the time. In general, the
prevailing wind direction occurs less than 20 percent of the time and,
with the exception of Honolulu, never exceeds 25 percent.
In expressway exposure situations, it is the relative wind direction
that is important. Thus, if a city had a prevailing wind a large percentage
of the time, and had a large percentage of its expressway mileage in one
direction, it could be possible to have a particular relative wind direction
occuring on the expressway much more frequently than others. If this
happened in a number of cities, it might be possible that there would be
a relative wind direction on a national scale that occurs much more
frequently than other directions.
To determine if this might happen, for those cities with a prevailing
wind direction occuring a high percentage of the time, city expressway maps
were visually inspected for a predominant expressway direction. No cities
were observed to have a predominant direction for expressways. As a check
to insure that the relative wind direction would be approximately equally
distributed, the percent of total expressway miles for each relative wind
direction was calculated in detail for San Antonio, Texas.
47
-------
TABLE 11. PREVAILING WIND DIRECTION FOR SMSA
OVER 400,000 POPULATION
Akron, Ohio
Albany, N.Y.
Atlanta, Ga.
Baltimore, Md.
Birmingham, Ala.
Boston, Mass.
Buffalo, N.Yi
Charlotte, N.C.
Chicago, 111.
Cincinnati, Ohio
Cleveland, Ohio
Columbus, Ohio
Dallas, Tex. (Love Field)
Dayton, Ohio
Denver, Colo.
Detroit, Mich.
Fort Worth, Tex. (Dallas-Ft. Worth Regional)
Fresno, Calif.
Grand Rapids, Mich.
Greensboro-Winston Salem-High Point, N. C.
Harrisburg, Pa.
Honolulu, Hawaii
Houston, Tex.
Indianapolis, Ind.
Jacksonville, Fla.
Kansas City, Mo-Kan.
Knoxville, Tenn.
Los Angeles-Long Beach, Calif.
Louisville, Ky-Ind.
Memphi s, Tenn.-Ark.
Miami, Fla.
Milwaukee, Wise.
Minneapolis-St. Paul, Minn.
Nashville, Tenn.
,Newark, N. J..
New Orleans, La.
New York, N.Y.
Norfolk-Va. Beach-Portsmouth, Va.
Oklahoma City, Ok.
Omaha, Nebr-Iowa
Philadelphia, Pa.
Phoenix, Ariz.
Pittsburgh, Pa.
Portland, Ore-Wash.
Providence-Warwick-Pawtucket. R.I.-Mass.
Richmond, Va.
Direction
S
WNW
NW
WNW
S
SW
SW
SW
SW
SSW
S
SSW
S
S
s
N
S
NW
SSW
SW
WNW
ENE
SSE
SW
NW/SW
SSW
NE
WSW
S
S
ESE
WNW
SE
S
SW
S
S
SW
SSE
SSE
WSW
E
WSW
NW
SW
S
Percent
Time
12.4
15.2
14.7
11.0
8.2
12.0
15.0
9.3
9.0
13.3
15.0
9.1
16.0
12.3
16.9
10.0
21.3
21.1
12.8
17.4
14.9
38.0
14.0
10.3
9.0
12.3
15.1
20.0
10.8
13
11.0
10.0
9.6
14.5
10.0
9.0
10.0
11.6
21.9
18.8
11.0
15.2
15.5
14.0
10.8
10.5
48
-------
TABLE ll(Cont'd). PREVAILING WIND DIRECTION FOR SMSA
OVER 400,000 POPULATION
Rochester, N. Y.
Sacramento, Calif.
St. Louis, Mo-Ill.
Salt Lake City, Utah
San Antonio, Texas
San Diego, Calif.
San Francisco-Oakland, Calif.
Seattle-Everett, Wash.
Syracuse, N. Y.
Tacoma, Wash.
Tampa-St. Petersburg, Fla.
Toledo, Ohio
Tulsa, Okla.
Washington, D.c.
Wilmington, Del.
Youngstown-Warren, Ohio
Direction
WSW
SW
S
SSE
SSE
WNW
WNW/WNW
SW
WNW
SSW
E
WSW
S
S
NW
SW
Percent
Time
15.4
15.0
10.8
18.9
15.0
10.0
24.0/14.6
13.0
11.
14.
10.0
11.2
18.6
12.0
11.2
12.2
.7
.0
49
-------
For San Antonio, the complete wind direction distribution was used
with the expressway direction distribution to obtain the percent of
.expressway miles with each relative wind direction. The expressway
directions were expressed to the nearest compass point. If there were
equal mileage in all directions, each direction would contain 6.25 percent
of the expressway miles. For San Antonio, there is no one predominant
expressway direction, although NE to SW and E to W expressways each have
20 percent of the total expressway mileage. The predominant wind direction
in San Antonio is SSE occurring 15 percent of the time. The percent of
expressway miles with each relative wind direction is shown in Table 12.
TABLE 12. DISTRIBUTION OF RELATIVE WIND DIRECTION FOR
SAN ANTONIO, TEXAS, URBAN INTERSTATE EXPRESSWAYS
Wind Direction Percent Of
Relative To Expressway, Degrees Expressway Miles
0 and 180 12.30
22.5 and 202.5 9.87
45 and 225 11.19
67.5 and 247.5 11.59
90 and 270 13.29
112.5 and 292.5 12.60
135 and 315 13.50
157.5 and 337.5 12.56
Calm 3.00
If the expressway miles were equally distributed with respect to the
relative wind direction, each direction (including "calm") would contain
11.1 percent of the mileage. The actual percentages shown in Table 12
do not vary enough from the 11.1 percent to use any particular relative
wind direction as predominant. This conclusion has been extended to
include the other cities in the U.S., by a comparison to San Antonio
of their wind direction distributions and on a visual estimation of their
expressway direction distribution. Thus, there is no one relative wind
direction that can be called typical.
Wind speed is also needed to define the typical and severe expressway
exposure situations. Since there is no one predominant relative wind
direction, the wind speed distribution for all wind directions was
used. The wind speed distribution with the largest number of speed
intervals was found in the NOAA publication "Airport Climatological
Summary" ,(76) This publication was available for 42 cities. Wind speed
distribution is presented at three hour intervals. Unfortunately the
data is taken at the same Greenwich Time for all stations, regardless
of time zone. Thus, Eastern time zone cities have speed distributions
at 7 a.m., Central time cities at 6 a.m., Mountain time cities at 8 a.m.,
and Pacific time cities again at 7 a.m.
50
-------
The 7 a.m. data is probably the most representative of morning rush
hour wind conditions. There were 25 cities with 7 a.m. data available;
20 Eastern cities and five West coast cities. This data was used to deter-
mine the average wind speed distribution. To obtain a wind speed distri-
bution based on individual exposure, the speed distribution for each city
was weighed by that city's percent of the total population for all 25 cities.
This population weighted wind distribition is shown in Figure 14
together with the unweighted minimum frequency, maximum frequency, and
means for the various wind speed intervals. A wind speed of 6 knots
(6.9 mph) would appear to be a typical speed, both from a median and mean
standpoint.
With typical wind conditions defined, the typical expressway exposure
situation can be defined as a four lane expressway with ADT of 20,000 to
30,000 vehicles and a wind speed of seven miles per hour, with all wind
directions equally likely. The severe expressway exposure situation is a
10 lane expressway with ADT in excess of 180,000 vehicles. The most severe
wind conditions are generally considered to be parallel wind, and wind
speed as low as allowed by the dispersion model (about one m/sec., or 2.2
mph) .
The typical exposure situation was chosen as IH 410 on the west side
of San Antonio, Texas, between U.S. 90 and IH 35. This is a relatively
straight, level, six mile section of four lane urban expressway, running
in a NNW-SSE direction. The traffic count averages approximately 28,000
vehicles per day over the section to be modeled. It has the added advantage
of being close to Southwest Research, so that traffic counts, grade, and
dimensions could be easily checked.
The severe exposure situation was chosen as the Santa Monica Freeway
in Los Angeles between the Harbor Freeway and IH 405, west of downtown.
This is a relatively straight, level, 10 mile section of 10 lane freeway,
running approximately in an east-west direction. The traffic count is
approximately 200 ,000 vehicles per day. Complete descriptions of both
section of expressway are included in Section V where ambient concentrations
for both the typical and severe expressway situations are calculated.
E. Street Canyons
Street canyons, streets running between tall building, as occur sometimes
block after block in the downtown section of large metropolitan areas, are
yet another situation in which people are exposed to elevated levels of
automotive pollutants. Ideally, information on the number of miles of
street canyons by street width and canyon height-to-width ratio should be
used to determine typical and severe exposure situation. However, no
compilations of street canyon statistics (number of miles of street
canyons, height-to-width ratios, etc.) were found.
51
-------
o
•H
•P
rt)
M
^28
Wind Speed, Knots
Figure 14. Wind speed distribution for 25 cities at 7 a.m.
52
-------
A procedure to at least estimate the number of miles of street canyons
was developed. It was to: (1) assume that all street canyons were in
central business district (CBD); (2) obtain the CBD size for all SMSA's
over 400,000 population (1970 census); and (3) assume a value for the number
of street miles per square mile of CBD area. From these assumptions, the
miles of streets in all CBD's was calculated. Assuming a certain average
percentage of the CBD street miles were canyons, the total street canyon
miles for all cities was obtained. This procedure gave the approximate national
total, since few SMSA's under 400,000 population have any street canyons at all.
A search was made for information on central business district areas.
It was found that the Bureau of Census has defined a CBD for each city over
100,000 population. The CBD area is generally defined in terms of a whole
census tract or tracts. No compilation exists of the DBD areas(77), however,
maps of each CBD are included in the Major Retail Center Statistics section
of the 1972 Census of Retail Tracts.(78) using these maps, the CBD area for
71 SMSA over 400,000 population was calculated. The total CBD area for the
71 SMSA's was 68.02 square miles. Table 13 lists the SMSA's and CBD areas.
The miles of street per square mile of CBD were calculated for six
CBD of different size and type. The average was 19 miles of street per
square mile of CBD. Using this average, the total miles of street in the
71 CBD's is 1292 miles. For all 76 SMSA over 400,000 pooulation, the total
street mileage is 1383 miles.
To help determine the street canyon portion of this mileage, an aerial
photograph of the CBD of Houston was obtained from a local aerial mapping
company, United Aerial Mapping. (See Figure 15). From the photograph, it
appears that approximately 30 percent of Houston CBD area could be con-
sidered street canyons.
It should be realized that Houston is a new, western style city. Older
eastern cities probably have a greater portion of street canyons in the CBD.
Some large city CBD's may even approach 100 percent street canyons. A 60
percent portion of the CBD mileage in street canyons will be assumed for
this study. If 60 percent of the street mileage is in street canyons, there
are approximately 830 miles of street canyons in the country.
Data on height to width ratios for street canyons are also nonexistent.
Some idea of the possible range of street canyon height to width ratios can
be obtained by looking at extremes of street width, building height, and
wind speed. Wind speed is a necessary variable, since the street canyon
dispersion model, used is not good unless a helical air circulation is
eatablished in the canyon. The establishment of a helical air circulation
is dependent on windspeed, street width and building height.(33)
The wind speed and direction information collected for the expressway
situation is also valid for the street canyon exposure situation. Since
streets, like expressways, run in all directions, all relative wind directions
are equally likely. The wind speed distribution during the morning rush hour
used for expressways, shown in Figure 14, indicates that the wind speed is
generally under 16 knots (18 mph, 8 meters/sec).
53
-------
TABLE 13. CENTRAL BUSINESS DISTRICT AREAS
FOR STANDARD METROPOLITAN STATISTICAL AREAS
OVER 400,000 POPULATION (1970 CENSUS)
SMSA
1970 Population CBD Area sq.mi,
Akron, OH 679~239 0.42
Albany-Schenectedy-Troy, NY 720,786 0.76
Allentown-Bethlehem-Easton, PA 543,551 0.40
Anaheim-Santa Ana-Garden Grove, CA 1,420,386 1.62
Atlanta, GA 1,390,164 1.39
Baltimore, MD 2,070,670 0.58
Birmingham, AL 739,274 0.15
Boston, MA 2,753,700 3.42
Buffalo, NY 1,349,211 0.65
Charlotte, NC 409,370 1.73
Chicago, IL 6,978,947 1.59
Cincinnati, OH 1,384,911 0.71
Cleveland, OH 2,064,194 1.11
Columbus, OH 916,228 0.82
Dallas, TX 1,555,950 1.44
Dayton, OH 850,266 0.81
Denver, CO 1,227,529 0.91
Detroit, MI 4,199,931 1.23
Flint, MI 496,658 1.38
Ft. Lauderdale-Hollywood, FL 620,100 0.18
Fort Worth, TX 762,086 0.93
Fresno, CA 413,053 0.28
Gary-Hammond-East Chicago, IN 633,367 0.94
Grand Rapids, MI 539,225 0.36
Greensboro-Winston Salem-High Point, NC 603,895 1.08
Hartford, CT 663,891 0.38
Honolulu, HI 630,528 0.19
Houston, TX 1,985,031 1.60
Indianapolis, IN 1,109,882 2.15
Jacksonville, FL 528,865 0.29
Jersey City, NJ 609,266 0.20
Kansas City, MO-KS 1,256,649 0.72
Knoxville, TN 400,337 0.61
Los Angeles-Long Beach, CA 7,032,075 2.82
Louisville, KY-IN 826,553 1.72
54
-------
TABLE 13(Cont'd). CENTRAL BUSINESS DISTRICT AREAS
FOR STANDARD METROPOLITAN STATISTICAL AREAS
OVER 400,000 POPULATION (1970 CENSUS)
SMSA 1970 Population CBD Area sq.mi.
Memphis, TN-AR 770,120 0.48
Miami, FL 1,267,792 0.33
Milwaukee, WI 1,403,887 0.88
Minneapolis-St. Paul, MN 1,813,647 1.68
Nashville, TN 540,982 0.82
Newark, NJ 1,856,556 1.14
New Orleans, LA 1,046,470 1.28
New York, NY 11,528,649 3.75
Norfolk-Va. Beach-Portsmouth, VA 680,600 0.90
Oklahoma City, OK 640,889 0.28
Omaha, NB-IA 541,453 1.06
Paterson-Clifton-Passaic, NJ 1,358,794 0.24
Philadelphia, PA 4,817,914 1.65
Phoenix, AZ 968,487 1.00
Pittsburgh, PA 2,401,245 0.54
Portland, OR-WA 1,009,129 0.39
Providence-Warwick-Pawtucket, RI-MA 914,110 0.62
Richmond, VA 518,319 1.08
Riverside-San Bernardino-Ontario, CA 1,143,146 1.68
Rochester, NY 882,667 0.66
Sacramento, CA 800,592 0.22
St. Louis, MO-IL 2,363,017 0.33
Salt Lake City, UT 557,635 0.40
San Antonio, TX 864,014 1.07
San Diego, CA 1,357,854 0.32
San Francisco-Oakland, CA 3,109,519 1.66
San Jose, CA 1,064,714 1.33
Seattle-Everett, WA 1,421,869 0.48
Springfield-Chicopee-Holyoke, MA 529,922 0.22
Syracuse, NY 635,946 0.45
Tacoma, WA 411,027 0.31
Tampa-St. Petersburg, FL 1,012,594 1.36
Toledo, OH 792,571 0.46
Tulsa, OK 475,991 1.36
Washington, DC 2,861,123 1.51
Youngstown-Warren, OH 536,003 0.53
55
-------
mm1 * *'
i?/x, f> •, //'A
,» V -v^ * ->
* » .^,>^-' •> v
,*/- *,>'>, ^y'f«
X
I
N, -4» ^^2/f:J -%">"JS^ ?> *4
*^&#^^^
,^tS*«W*i
lyj^V;». rvA -^, •«
y^fl *-^*4lv " ^'x
Figure 15. Aerial View of Central Business District
Houston, Texas
56
-------
For this study, the minimum street width was assumed to be 9 meters
(two 12 foot lanes, plus two three foot sidewalks). The maximum street
width is assumed to be 63 meters (ten 13-foot lanes, three 12-foot medians
and two 20-foot sidewalks). It is not known if there is actually a street
this wide in existence, but it is judged to be a safe maximum assumption.
Average street widths appear to be on the order of 20 meters.
Using the equation for the wind penetration depth from Reference 33
(See Section III), the minimum building height necessary for helical air
circulation can be calculated for the range of wind speeds and street
widths defined above. Figure 16 shows this minimum building height. Note
from Figure 16 that for usual wind speeds, the minimum height to width
ratio for helical flow must be above 0.3. Thus, 0.3 will be used as the
minimum height-to-width ratio for a building-lined street to be defined
as a street canyon.
For the narrowest street, the building height must be above 6.5 meters.
This height is approximately two stories. Thus, street canyons must have
buildings at least two stories high, even for the narrowest street. The highest
buildings in the country are well over 1000 feet high. However, buildings
over 300 feet (91 meters, about 26 stories) are considered "notable" by
The World Almanac(52), which lists the notable tall buildings in major
cities. Thus, it is probable that in most street canyons, buildings are
under 91 meters in height.
Traffic counts for representative CBD streets were obtained from several
sources.(79,80,81) These traffic counts are shown in Table 14. During peak
traffic periods 1000 vehicles/hr for a CBD street would seem to be a good
average for all SMSA over 400,000 population. Since the average CBD
area is 0.96 square miles, it is not unreasonable to assume that each vehicle
drives 0.5 miles through street canyons. One study has shown that 20 per-
cent of the downtown vehicles are buses.' ' If it is assumed that each bus
has 30 passengers and the remaining 80 percent of the vehicles are cars
carrying 1.4 persons, then there are approximately 7 persons per vehicle
in the CBD. Using these figures, approximately 11.6 million people in
vehicles are exposed to CBD peak hour street canyon pollutant concentrations.
Note that this figure does not include pedestrians who entered the canyon
by another means (subways or on foot, for example).
While average daily traffic was used to obtain the number of persons
exposed to street canyon concentrations, the traffic count is not sufficient
to define typical and severe street canyon exposure situations. This is
because the same number of vehicles per unit time can occur at two different
speeds, as shown in Figure 17. Thus, a CBD street with low traffic counts
can have a high traffic density traveling at a slow speed (congested traffic),
or a low traffic density traveling at a higher speed (uncongested traffic).
With stop and go driving, such as in encountered in city traffic, vehicle
speed becomes an important consideration, since emissions can vary greatly
with average cycle speed.
57
-------
0)
-P
1
170 r
160
150 H
140 \-
130 h
120 h
to no h
100 h
« 90
Cn
C
'•3 80
70 h
60 h
50 h
40 h
30 h
20 h
10 h
Miniinum heights
calculated from:
Street width
(meters)
63
2345678
Wind Speed, meters per second
10
Figure 16. Minimum Building Height for Helical Air
Circulation in a Street Canyon
58
-------
TABLE 14. AVERAGE DAILY TRAFFIC FOR SOME
CENTRAL BUSINESS DISTRICT STREETS
IN SELECTED CITIES
City
Location
Los Angeles, CA
Los Angeles, CA
Los Angeles, CA
Los Angeles, CA
Houston, TX
Houston, TX
Houston, TX
San Antonio, TX
New York City, NY
Notes:
Olympic West of Figueroa
1st Street east of Los Angeles St.
Figueroa south of Pico
Figueroa at llth Street
Milam south of Pease
Main north of Pease
Louisiana north of McKinney
Houston Street, east of St. Mary's
Lexington between 45 and 46th St
Daily Volume
28,662
18,418
20,282
23,394
18,183
15,608
18,965
15,293
-18,000
1. Los Angeles data from "Traffic Counts 1979" City of Los Angeles
Department of Transportation.
2. Houston data from "City of Houston, Traffic and Transportation
Department, Traffic Volume Summary as of 03-23-79."
3. San Antonio data from San Antonio Traffic and Transportation
Department.
4. New York City data taken from: "Effect of New York City Taxi
Strike on CO Concentrations in Midtown Manhattan." APCA Journal
Vol. 29, No. 8. August 1979.
59
-------
O
•H
B
T)
fl)
0)
ft
en
u
•r-l
-------
Just how emissions vary with speed is dependent on the pollutant. Figures
18 and 19 show regulated and unregulated emission levels from four different
cars operated on four different urban driving cycles, as a function of the
average speed of each cycle J83) As can be seen from the figures, there is no
single trend. Some emissions increase with increasing vehicle speed, some
decrease. Thus, it is difficult to say whether 1000 vehicles per hour
traveling at five miles per hour (congested traffic) or 2000 vehicles per
hour traveling at 20 miles per hour (uncongested traffic) would be the more
severe case from an air pollution standpoint.
These facts indicate that two different traffic flows should be con-
sidered for each street canyon situation. One flow should be of congested
traffic (low flow, low speed) another of uncongested traffic (higher flow,
higher speed). While this conclusion is a departure from the strict "real world
situation" philosophy of this project, it is necessary to allow assessment
of the effects of a variety of unregulated pollutants in a street canyon
situation. From an examination of a variety of traffic flow data and
information, it appears that 200 vehicles per hour per lane at five miles
per hour would be a good representation of congested flow; while 400
vehicles per hour per lane at 20 mph would be a good representation of
uncongested flow. These traffic flows will be used with both the typical
and severe street canyon situations.
It should be obvious from the above discussion that it will not be
possible to define typical and severe street canyon exposure situations
with any degree of certainty. In addition, the height to width limitation
of 2.0 imposed on the model means that the truly severe exposure situation
(wind speed lower than 0.5 m/sec, or building height above 91 meters) can
not be modeled. Therefore, the street canyon locations were chosen on a
subjective basis keeping in mind the range of variables possible.
The typical location was chosen as Houston Street between Navarro and
St. Mary's in downtown San Antonio, Texas. This is a four lane, one-way
street with seven foot wide sidewalks on both sides. The total width of the
street and sidewalks is 18.6 m (61 feet). The average leeward side building
height (assuming southerly winds) is 33.7 m (111feet). The height to width
ratio is then 1.81. The typical wind speed of 3.13 m/sec (7 miles/hr) will
be used. The ADT for Houston Street at this location is approximately 15,300
vehicles per day.
Since the model limited the canyon height to width ratio to 2.0, the
severe case was chosen on the basis of number of traffic lanes and traffic
counts, plus the more practical consideration of availability of the neces-
sary information. A local aerial mapping company had stereographic aerial
photographs of the Houston area from which building heights could be
determined. Since downtown Houston streets are in general six-lane
one-way streets with heavy traffic, a street canyon in Houston appeared
to be a good location for the severe exposure case. Care had to be taken
to insure that a street canyon with less than 2.0 height to width ratio was
chosen.
61
-------
B
,X
en
•n
O
5.0
4.0
3.0
2.0
1.0
0 10 20
Cycle
Name
1.2
e
I
0.8
0.6
0.4
0.2
30 40 50 60
Average Speed, km/hr
NYCC
FTP
SET
J I
0 10
20 30 40 50 60
Average Speed, km/hr
O 1978 Calif Pi
Q 1978 Calif Su
£ 1978 Saab
0 1979 Calif Ma
70 80
HFET
J i
70 80 90
Figure 18. Effect of driving cycle speed on NOX and CO Emissions
-------
60
50
I 4°
G
X
-------
The final section was Main Street, which rims NE to SW, between Capitol
and Rusks streets. The building height on the southeast side of the block
(leeward side) average 38 m (125 feet) above street level. The street is
27 m (90 feet) wide. The canyon height to width ratio is 1.39. The average
traffic count is on the order of 20,000 vehicles per day. For the severe
case, the wind speed should be as low as possible. From Figure 16, for a
street 27 meters wide, the building height of 38 meters, the minimum wind
speed to have a helical flow pattern is approximately 0.9.: meters/sec (2.1
miles per hour).. Complete descriptions of the two street canyons are in-
cluded in Section V with the calculation of the ambient pollutant concen-.
trations in each canyon.
F. Localized Area Sources
Since no dispersion model was found for predicting concentrations
within an area source, a much smaller effort was allotted to the inves-
tigation of the range of physical variables of area sources. There have
been a number of studies of individual parking lots(94,95,96,97) , but
none have attempted to define the magnitude of the problem on a national
scale. No studies were found on such area sources as trucking terminals,
warehouses or similar situations. From the data presented in Reference 57,
it is estimated that there are at least 60,000 privately-owned and municipal
parking lots open to the general public. The average lot has 66 parking
spaces. However, parking lots can be many times the average size. For
instance, the Oakbrook Shopping Center in suburban Chicago has a parking
lot capacity of 7,200 spaces.(94)
Since there is no dispersion model for this expsoure situation, no
typical or severe location was chosen. However, the ambient CO concen-
tration measured in three large parking lots will be discussed in Section V.
64
-------
V. AMBIENT CONCENTRATIONS FOR SELECTED EXPOSURE SITUATIONS
In this section, the models selected in Section III are applied to the
real world typical and severe situations selected in Section IV. The emis-
sion factor used was one gram per mile per vehicle, or one gram per minute
per vehicle, as appropriate.
Ambient concentrations for any pollutant may be obtained from the calculated
concentrations presented here, simply by multiplying the concentration by the
new pollutant emission factor in grams per mile or grams per minute as appro-
priate. To demonstrate this procedure, the ambient concentration CO was de-
termined for each situation. Where possible, the CO level was compared to
measured CO concentrations for the same or a similar location. The CO emission
factor used for computation varied with the calendar year for which the com-
parison was made, location (whether in California or another state), and
average vehicle speeds. In every case, an attempt was made to use the most
realistic CO emission factor available.
A. Residential Garage Concentration
The typical and severe residential garage exposure cases selected in
Section IV use the same physical setting, but different pollutant generation
times. An illustration of the garage is shown in Figure 20.
8.42'
Figure 20. Single car garage with storage space
-------
All walls and ceiling are finished. There are two free air vents with
a combined open area of 288 in2. An "overhead door" is provided for auto.
access. A weather stripped door is provided for garage-interior access,
and is assumed to be effectively "sealed." The volume of the garage is
2189 cubic feet (10x26x8.42). The volume of vehicle and appliances have
not been subtracted from the total volume.
Since there is no mechanical ventilation in a residential garage, all
ventilation is the result of infiltration and natural ventilation. Infil-
tration pertains to ventilation due to wall leakage and window leakage.
Natural ventilation pertains to ventilation by intentional openings, such
as windows or vents. From the ASHRAE handbook, a room with no windows or
exterior doors would have an effective room air change every two hours.
The air for the room air change comes from air infiltration through the walls,
caused by a differential pressure between the interior and exterior spaces,
and it depends on ambient conditions of wind and temperature.'°^'
The main source of infiltration for the garage is the garage door.
The door has a sealing perimeter of 30 feet with a loose fit to enable it
to slide. It is assumed that the infiltration is similar to a wooden double
hung window which is not weatherstripped. Assuming a Ap of 0.1 in. J^O,
corresponding to a fairly low wind velocity and only a minor temperature
difference, the infiltration would range from about 27 cfh to 53 cfh per
foot of crack for cracks 0.047 to 0.094 inches wide, respectively. Picking
40 cfh per foot of crack yields 1200 cfh (20 cfm), or a change of garage
air every 1.8 hours.(^9) xhis 20 cfm of air infiltration would pass through
the 2 square feet of free air vents with a velocity about 2 in/sec, which
seems reasonable. The ventilation for the garage with the door closed is
taken to be 20 cfm.
Natural ventilation for the residential garage would be primarily due
to opening the garage door in conjunction with the free air vents. The
garage door represents almost one of four walls and has an opening area of
56 ft2. Assuming an 88 ft/min wind velocity, V, and an effectiveness, E, of
about 0.3 due to the wind direction being skewed to the plane to the door,
the natural ventilation through the 56 ft2 doorway would be 1478 ft^/min.
Flow rate is defined as: cfm - Effectiveness x Wind Velocity (ft/min) x
Area (ft2). Applying these same assumptions to the free open vent area
of 2 ft2, natural ventilation through the vents would be 52.8 ft^/min.
With an "inlet" of 56 ft2 (garage door) and outlet of 2 ft2 (free open
vents), ASHRAE recommends taking the flow rate through the smaller opening
and adding a portion of the flowrate from the larger opening.(84) Based
on the ratio of inlet to outlet areas, this is 52.8 ft^/min from the vents
plus 38 percent of the 1478 ft^/min from the garage door opening. This
gives a ventilation rate of 615 ft3/min which corresponds to 16.8 changes
per hour, or a room change every 3.57 minutes with the garage door open.
The velocity of this ventilation air through the garage doorway would be
11 ft/min. The ventilation for the garage with the door open is thus
taken to be 615 cfm.
66
-------
1. Application of the Enclosed Space Model
Since both the typical and severe exposures occur while the room
concentrations are increasing rapidly it is necessary to consider the concen-
trations as a function of time. Each of the three components of the enclosed
space model set forth in Section II will be calculated separately and plotted
as a function of time. Aside from the physical variables of the garage
described on the previous page, several other parameters are required. To
calculate the contribution of the initial concentration, it is assumed that
sufficient time has passed to allow the room concentraion to reach ambient
concentration. Initial concentration corresponds to Co in the model equation
for C-L. A unit concentration of 1 yg/m3 is assumed. The emission rate is
used in the model equation for €3. A unit emission rate of 1 gram/min is
assumed.
As mentioned in Section II, the effective ventilation factor is difficult
to assess, and little information as to its value has been found. Considering
the definition of the ventilation factor can help to estimate a value for
these situations. Recall from Section III, that the ventilation factor is
less than 1.0 if the ventilation system removes less pollutant than would
be removed if the pollutant were distributed homogeneously. In this situation,
with little air movement within the garage it is probable that much less of
the generated pollutant is removed per unit time than would be removed if it
were distributed homogeneously. Therefore, an effective ventilation factor
for the generated pollutant (C^ equation) of 0.25 has been assumed. For the
equations for initial concentration decay (C^ equation) and for the incoming
ventilation concentration (C2 equation), the effective ventilation factor
has been assumed to be 1.0; since the pollutant is nearly homogeneously dis-
tributed in both cases.
Figures 21 and 22 contain the curves of concentration as a function of
time for the three components of the enclosed space model; initial concen-
tration decay (C^), inlet ventilation concentration (C2), and generated
pollutant concentration (C^).
2. Typical Residential Garage Concentrations
For the typical exposure situation, the receptor enters the garage
from the house, opens'the overhead door, starts the car and allows 30 seconds
to warm-up at fast idle. He then backs the car out of the garage, closes
the overhead door and drives off.
The maximum concentration to which the receptor is exposed occurs at
30 seconds after the engine was started. To use Figures 21 and 22, the engine
starting time is taken as time zero. At time 30 seconds the three model com-
ponents obtained by reading each curve at t = 0.5 minutes are:
G! = 0.870 yg/m3
C2 = 0.132 yg/m3
C3 = 7900 yg/m3
67
-------
ro
e
g
T)
OJ
to
-------
1,000,000
500,000
"t" --j--•-;: }-.-. |-
1,000
0
0.5 1.0 1.5 2.0 5 10
Time, Minutes
Figure 22. Residential garage multiplier for C3
-------
A summary of the physical parameters and concentrations for the typical case
is given in Table 15 following the presentation of the severe case.
3. Severe Residential Garage Concentrations
For the severe exposure situation, the pollutant generation time
is greater, and the receptor spends more time in garage after the car is
started. The severe case assumes a cold winter morning with an unheated
garage. The receptor enters the garage from the house entrance, opens the
overhead door, starts the car, and returns to the inside of the house while
the car warms up at idle. After approximately five minutes, the receptor
returns to the garage, backs out the car, closes the overhead door and
drives off.
The maximum concentration the receptor will be exposed to in
this case occurs five minutes after the engine was started. Prom Figures
21 and 22, at time five minutes, the values of the three components are:
C-L = 0.242 yg/m3
C2 = 0.750 yg/m3
C3 = 67000 yg/m3
A summary of the physical parameters and concentrations for the severe case
is presented in Table 15.
4. Use of Residential Garage Concentrations with Other Emission Factors
To use these values with other pollutants, multiply C^. C2 and C^
by the appropriate concentration or emission factor then sum the three new
concentrations. For example, assume that the cold-start CO generation rate
for the first 30 seconds is 8 g/min, and the CO background is 3000.yg/m3.
For the typical exposure situation, after 30 seconds, the CO con-
centration in the garage is:
Cx = 0.870 x 3000 = 2,610 yg/m3
C2 = 0.132 x 3000 = 396 yg/m3
C3 = 7900 x 7 = 63,200 yg/m3
Total concentration = 66,206 yg/m3 or 57.6 ppm CO
For the severe exposure situation, assume that the CO rate drops
from 8 g/min to 5 g/min at t = 30 seconds. To calculate the CO concentration
for the severe case, it is necessary to redesignate the time at which the CO
generation rate changed as t = 0. In this case,, it^is 30 seconds after
the engine started. C0 is now total concentration in the garage 30 seconds
after the engine started, or 66,206 yg/m3. C-i has not changed. The time
at .which to enter the graph for the severe case (five minutes after engine
start) is then t = 4.5 minutes.
70
-------
TABLE 15. AMBIENT CONCENTRATIONS FOR RECEPTOR IN RESIDENTIAL GARAGE
TYPICAL AND SEVERE CASES
Garage Volume, cubic feet
Ventilation rate (door open) cfm
Initial concentration, yg/m3
Intake air concentration, yg/m3
Emission factor, g/min
Ventilation factor, initial cone.
Ventilation factor, intake air
Ventilation factor, exhaust emissions
Time after start of engine, minutes
Concentrations:
C~L, contribution of initial cone.
yg/m3
C2, contribution of intake cone.
yg/m3
C3, contribution of exhaust
emissions,
Typical
Case
2189
615
1.0
1.0
1.0
1.0
1.0
0.25
0.5
Severe
Case
2189
615
1.0
1.0
1.0
1.0
1.0
0.25
5.0
0.870
0.132
0.242
0.750
7900
67000
71
-------
G! = 0.290 x 66,206 = 19,200 yg/m3
C2 = 0.720 x 3,000 = 2,160 yg/m3
C3 = 60,000 x 5 = 300,000 yg/m3
Total Concentration = 321,360 yg/m3 or 279 ppm CO
Note that in this case, it was not possible just to take the concen-
trations at 5 minutes and multiply by the emission factor. Because the emis-
sion factor changed, it was necessary to use the graphs and work the problem
in two parts.
No CO measurements within a residential garage were found in the
literature. One study did show CO measurements in a room next to an attached
garage.'^5) jn that study, the CO increased from 1 ppm before the car was
removed to 5 ppm after the car was taken from the garage. This situation was
apparently similar to the typical situation presented here. It does not seem
unreasonable that the CO concentration inside should increase by a factor of
5 while the garage CO concentration increased by a factor of 20.
B. Parking Garage Concentrations
The typical and severe parking garage exposure situations were chosen
in Section IV on the basis of size and ventilation capacity. The enclosed
space model explained in Section III B 2 was used to determine the pollutant
concentrations within the two parking garages.
Recall that the enclosed space model contains an effective ventilation
factor. The numerical value of this factor is not well defined in the
literature. However, an unpublished CO measurement studv done by Los Angeles
county in the severe case parking garage^°®' was utilized in this project to
determine the effective ventilation factor. For that reason, the severe ex-
posure case will be presented first.
1. Severe Parking Garage Concentrations
The Music Center parking garage in downtown Los Angeles was
chosen as the severe exposure situation. This is an 8 level underground
parking garage. A sketch of the garage cross section is shown in Figure 23.
A floor plan of a typical level (level 5) is shown in Figure 24. The most
severe exposure occurs at the end of an evening event, when all cars are
trying to exit at the same time.
The situation modeled used the original ventilation system
which existed prior to 1979. In that system, the mechanical ventilation
was on the exhaust side only. Fresh air entered from the street level entrances
(level 2), the pedestrian tunnel on level 4, and the vehicle tunnel on
level 8. Following many complaints about the garage ventilation, a
complete redesign of the ventilation system was begun in 1977. The
current system uses blowers to supply the air through ducts, exhausting
through the extrances. The current ventilation capacity is almost
three times the original capacity. For the purposes of this project,
however, the original ventilation system best fits the severe exposure
condition.
72
-------
Grand Ave.
Entrances
NOTE: Not to Scale
Figure 23. Sketch of parking level arrangement, Music Center Garage
-------
APPROXIMATE
NORTH
Figure 24. Floor plan of parking level 5 of Music Center Garage
74
-------
The air supply reaches the individual parking levels through the
ramps, so the pollutant concentrations in the ramps must be also calculated
to obtain the pollutant concentration of the air entering the parking level.
The air flow in each between-level ramps will be different, since some air
coming down the ramps is exhausted at each level. The vehicle capacity,
ventilation, and volume of each level and interlevel ramp is shown in
Table 16.
TABLE 16. DATA FOR MUSIC CENTER GARAGE
Leve Is
1
2
3
4
5
6
7
8
Total
Ramps
Street to
3 to 4
4 to 5
Parking
Spaces
120
190
195
181
226
216
223
231
1,582
3
-
_
Volume
cubic feet
636,262
1,000,407
701,760
705,546
737,316
795,328
737,316
795,328
6,109,263
18,445
3,400
3,400
Ventilation
cfm
39,820
35,430
28,770
40,610
38,100
48,600
35,400
44,200
310,930
52,500
47,705
40,937
From an analysis of the CO study performed on the garage in 1978,
it appeared to take between 20 and 30 minutes to clear the garage after a
performance. Following other studies of parking garages, it is assumed that
approximately 25 percent of the vehicles are in operation at any instant
during this time.(62)
75
-------
The remaining variable requiring definition is the effective venti-
lation factor for each equation. Since the effective ventilation factor is
a function of how the pollutant is introduced into the garage as well as the
ventilation scheme, there is a different ventilation factor for each of the
three equations. From an analysis of the CO measurements done in 1978, it
appeared that a reasonable value of the ventilation factor for the vehicle
generated pollutant is 0.3. For the initial concentration and background
contributions, the ventilation factor is 1.0.
For the severe situation, the receptor will be assumed to be lo-
cated on level 5 and go up to exit the garage onto Grand Ave. at level 2
after a performance at the Music Center. It is assumed that the receptor
spends 8 minutes on level 5, and 4 minutes in the ramp area. For purposes
of this severe case, it is assumed that the receptor is one of the last
vehicles out of the garage and so experiences the maximum concentration.
This maximum occurs at approximately 20 minutes after the garage starts to empty,
The ramps will approach an equilibrium concentration in less than
three minutes. This means that the initial concentration will have decayed
to essentially zero, the inlet air contribution will be the inlet air concen-
tration and the vehicle contribution will be the exhaust emission rate
divided by the ventilation rate. As the air flows downward through the
ramps, the concentration in each ramp level becomes the inlet air concen-
tration for the next ramp.
To calculate the ramp concentration, it was assumed there were
three active vehicles on each ramp at all times, except for the ramp from
level 3 to the street which contained 15 cars. For an assumed emission rate
of 1 g/min, an initial concentration of 1 yg/m3, and a street air concentration
of 1 yg/m3, the concentrations after 20 minutes in the ramps are:
Ramp Street to 3 Ramp 3 to 4 Ramp 4 to 5
Cj_ = o o o
C2 = 1 yg/m3 1 yg/m3 1 yg/m3
C3 = 15(1 x K)6) 3(1 x lp6) 3(1 x 1Q&)
(1) (52,.500) (0.02832) (1) (47705) (0.02832) (1) (40937) (0.02832)
C3 = 10,100 yg/m3 2,200 yg/m3 2,600 yg/m3
= 10,101 yg/m3 12,300 yg/m3 14,900 yg/m3
Thus, the pollutant concentration in the air entering level 5 is the sum of
the three C3 terms (ignoring the 1 yg/m3 street concentration).
76
-------
While the ramps reach equilibrium quickly, the parking levels do
not. The concentration in level 5 after 20 minutes, assuming a 1 yg/m3
initial concentration, 14,900 ug/m3 inlet air concentration, 1 g/minute
emission rate and 56 active cars is:
C-L = (1) (exp(-(l) (38100) (20)/737316)
C-j_ = 0.36
C2 = 14,900 (1 - exp(-(l) (38100) (205/737316)
C2 = 9,600 yg/m3
C3 " (0.3)3oO)oo2832) (1 ~ e^(~(I) (38100) (205/737316)
C3 = 46,100 yg/m3
Table 17 lists the concentrations in the various areas of the garage together
with the values of the parameters required to calculate those concentrations.
2. Typical Parking Garage Concentrations
The Convention Center Parking Garage in San Antonio, Texas,
was chosen for the typical exposure situation. This parking garage is
an above ground, open structure with natural ventilation only. It is
5 levels high and can store up to 461 vehicles. Its primary use is
during special events held at the nearby Convention Center complex.
This type of useage is referred to as surge type parking and resembles
useage common to large office building areas. A sketch of the garage
is shown in Figure 25. A floor plan of a typical level is shown in
Figure 26 .
The various parking level surfaces are vertically spaced 10 feet
apart, and are staggered with intermediate levels, giving the appearance of
split level construction. For application of the model, a given parking
level will be considered as being isolated from the other parking levels,
and will then be considered as an independent enclosed space. The various
levels are connected by ramps with about a 9 percent grade. These ramps
will be considered separately, due to the higher incidence of traffic queuing
in the relatively confined area. The drive from entrance to uppermost level
is approximately 1640 feet via primary driveway, of which nearly a fourth
is ramped.
The floor area of a parking level is approximately 39,500 ft .
Ceiling clearance is 7 1/2 ft. Since the ceiling is made of exposed
structural shapes, for volume considerations the height is taken to be
9 ft. These dimensions give a volume of about 356,000 ft3 per enclosed parking
level. As indicated in the illustrations, the parking levels are served
by two ramp sections. Each section is approximately 45 ft in height,
25 ft wide and 52 ft long, which represents a volume of 58,500 ft3 for
each "ramp room."
77
-------
TABLE 17_ AMBIENT CONCENTRATIONS FOR RECEPTOR IN A PARKING GARAGE
SEVERE CASE-LOS ANGELES MUSIC CENTER GARAGE
-------
Bldg.
r
-«^r-
I
T
T
---T-
~ T
r~
~—[~ --
\- '
|
^.
i
64
1
Figure 25. Parking level arrangement, San Antonio Convention
Center Parking Garage
-------
Bldg.
co
0
0
Q
0
O
0
Q Q S " " " "J y
RAMP UP
i
RAMP DI
I
I
0 0
I RAMP UP !
! RAMP Dl
1 omnTT
JS
o o o b *" 6" t
^ 221'
a !
i
1
ol o
O
o
o
o
o
j ELEVATOR
te
17
i
;
9'
\
1 North
Figure 26. Typical parking level floor plan for Convention Center Garage
-------
This open parking garage relies solely on natural ventilation,
and is dependent on the ambient conditions. The overall "enclosed" portion
of the garage structure can be considered as a large room with one large
window in each wall. The percentage of open space was calculated using
the total open area divided by the area of the corresponding wall (external
dimensions). The east wall projection is about 7000 ft2, with an open area
of 40 percent corresponding to a 2400 ft2 opening. The west wall is com-
pletely blocked by an adjacent building. Both north and south walls pro-
ject areas of 8800 ft2 each, with an open area of about 24 percent corre-
sponding to 2100 ft2 opening. In an effort to estimate the air flow, a
7 mph wind speed is assumed. This speed is a typical for the country as a
whole (see Figure 14 in Section IV). It is assumed that the wind is from
the southeast, so that the wind is skewed to the south and east walls of
the structure. In San Antonio, the wind is from the southeast, between
3 and 7 mph approximately three percent of the time. The effectiveness
of the open areas is assumed to be about 0.4, resulting in approximately
714,000 cfm through the east opening and 518,000 cfm through the south
opening. This situation results in a total ventilation of the enclosed
portion of the structure of about 1,232,000 cfm. Dividing by the four
enclosed levels yields 308,000 cfm ventilation per level.
The ramp area or "ramp room" is an integral part of the structure.
Natural ventilation is provided through access openings from individual
levels. Ventilation of the ramp room is aided by the chimney effect due
to warm engine air rising. It is assumed that the ramp room would have
a ventilation rate of 61,600 cfm.
This garage has more diffuse sources of ventilation than the Music
Center garage used in the severe case. The effective ventilation factor
should therefore be somewhat closer to one than the severe case. A venti-
lation factor of 0.4 was assumed for the vehicle exhaust equation. The
effective ventilation coefficients for the initial concentration and inlet
air equations were again assumed to be 1.0.
For this typical situation, the receptor will be assumed to be
located on level 4 and exit down to the street level, level 1, following
an event at the Convention Center. It is assumed that the receptor spends
4 minutes on level 4 and 4 minutes in the "ramp room". For purposes of
this exposure case, it is assumed that the receptor starts to exit the
garage approximately -15 minutes after it has started to empty. Level 4
has a capacity of 92 cars. It is assumed that it is 75 percent full and
that 25 percent of the cars are active at all times during the emptying
process. Thus, there are 17 cars active at all times on level 4. It is
also assumed that there are 12 active vehicles in the ramp room at all
times. Assuming an initial concentration of 1 pg/m3, an inlet air con-
centration of 1 jJg/m3, and an exhaust concentration of 1 g/min (IxlO6 yg/
min) the concentrations on level 4 at t = 15 minutes are:
81
-------
C-L = 1.0 x (exp - (1.0 x 303,000 x 15/356,000))
Cj_ = 0.0 yg/m3
C2 = 1.0 x (1 - exp - (1.0 x 308,000 x 15/356,000))
C2 = 1.0 yg/m3
17 x (ixio6)
C3 = 05 x 308000 x QQ282 (1 " &^ (°'5 X 3°8'°00 X W356,000)
0.5 x 308,000 x Q.Q2832
C3 = 3,900 yg/m3
For the ramp room the concentrations are:
Cj_ = 1.0 x (exp - (1.0 x 61,600 x 15/58,500))
Cj_ = 0.0 yg/m3
C2 = 1.0 x (i - exp - (1.0 x 61,600 x 15/58,500))
C2 = 1.0 yg/m3
12 x
C3 = 0.5 x 61,600 x O.Q2832 (1 ~ £XP ' (°'5 X 61'600 X 15/58,500)
C3 = 13,750 yg/m3
Table 18 lists the parameters required by the model equations together with
the concentrations after 15 minutes for the typical exposure case.
3. Use of Parking Garage Concentrations With Other Emission Factors
To demonstrate how these concentrations can be used with other
pollutants, the CO level in each of the two garages will be calculated.
Severe Case (Music Center garage) . A reasonable CO emission rate
for California in 1978 is 7.25 g/min for cars pn a parking level and 10 ' g/min
for cars on the ramp. ^98' A-ssume also that initial concentration and outside
background are negligible. The following CO levels are obtained:
Level 5: C2 = 9,600 x 10 = 96,000 g/m3
C3 = 46,100 x 7.25 = 334,225 g/m3
CTotal = 430,255 g/m3 or 374 ppm CO
Ramp 5 to 4: 14,900 x 10 = 149,000 yg/m3 or 129 ppm CO
Ramp 4 to 3: 12,300 x 10 = 123,000 yg/m3 or 107 ppm CO
Ramp 3 to street: 10,000 x 10 = 101,000 yg/m3 or 88 ppm CO
The level 5 CO concentration can be compared with the CO concen-
trations measured on level 5 on March 21, 1978, when the garage was approxi-
mately three quarters full. (86> On that date, 20 minutes after a perfor-
mance ended at the Music Center, a level 5 CO reading reached a maximum
of 325 ppm. The calculated value at 20 minutes of 374 ppm, is in good
agreement, considering that it was obtained assuming a full garage.
82
-------
TABLE 18. AMBIENT CONCENTRATIONS FOR RECEPTOR IN A PARKING GARAGE
TYPICAL CASE-SAN ANTONIO CONVENTION CENTER GARAGE
Parking Level Ramp Room
Number of active cars 17 12
Volume cubic feet 356,000 58,500
Ventilation rate, cfm 308,000 61,600
Emission factor, g/min 1.0 1.0
Intake air concentration, yg/m3 1 1
Ventilation factor, exhaust emissions 0.5 0.5
Ventilation factor, intake air 1 1
Ventilation factor, initial cone. 1.0 1.0
After 15 minutes: 1.0 1.0
C., contribution of initial cone.,
yg/m3 0 0
C_, contribution of intake cone.,
yg/m3 1 1
C.,, contribution of exhaust
emissions, yg/m3 3,900 13,750
83
-------
While the severe case was taken as a mechanically ventilated garage, it
should be pointed out that concentrations of the same order of magnitude
are possible with open construction, naturally ventilated garages when the
wind speed is below two miles per hour.
Typical Case (Convention Center Garage). A CO emission rate for
1978 on the parking level of 10 g/min and a CO emission rate of 13 g/min
on the ramps were assumed. The CO concentration in the intake air is 3 ppm,
or 3448 yg/m3. The parking garage CO concentrations are:
Parking level: C2 = 1 x 3448 = 3,448 yg/m3
C3 = 3,900 x 10 = 39,000 yg/m3
42,448 yg/m^ or 37 ppm CO
Ramp Room: C2 = ^ * 3448 = 3,448 yg/m3
C3 = 13,750 x 13 = 178,750 yg/m3
182,198 yg/m3 or 159 ppm CO
C. Roadway Tunnel Concentration
The calculations for ambient pollutant concentrations in the typical
and severe tunnel exposure situations also make use of the enclosed space
model. The tunnel exposure is assumed to take place after the tunnel
concentrations have essentially reached equilibrium.
1. Typical Roadway Tunnel Concentrations
As mentioned in Section IV, the Lowery Hill tunnel on 1-94 on the
west side of Minneapolis was chosen for the typical tunnel exposure situation.
Tunnel traffic and ventilation data were obtained from the District 5 Office
of the Minnesota Department of Transportation.(87) This 1500 foot long
tunnel consists of two tubes of rectangular cross section, each with three
lanes of one way traffic.
The average daily traffic count is approximately 60,250 vehicles.
The morning rush hour traffic used for this situation is assumed to be nine
percent of the ADT. The average speed in the east bound tube during the
morning rush hour is 45 miles per hour. At this rate it takes 23 seconds
to traverse the tunnel. From Figure 10, at this speed there are an
average of 25 vehicles in the east bound tube at all times during the
morning rush hour.
.• The tunnel uses a semi-transverse ventilation scheme. The maximum
ventilation rate is 536,000 cfm per tube. .To this rate is added 10 percent of
the total capacity to account for the piston effect of the cars.
Since this situation takes place at the end of the morning rush
hour, the pollutant concentration has essentially reached equilibrium. This
means that the initial concentration has decayed to zero, the ventilation
contribution is equal to the background concentration, and the vehicle con-
tribution is equal to the total vehicular emission rate divided by the
effective ventilation.
84
-------
Since roadway tunnels have high levels of ventilation and the air is thoroughly
mixed by the vehicle wakes, the effective ventilation factor (see Section III)
for the vehicle pollutants is assumed to be 1.0.
For an assumed emission factor of 1 g/mile (1X106 i-ig/mile) the
pollutant concentration in the tunnel is then:
= 25(1 x 106)(45)
3 60(1)(536,000 x 1.10)(0.02832)
, C3 = 1,123 yg/m3
A summary of the typical exposure situation parameters is contained
in Table 19 following the presentation of the severe tunnel exposure case.
2. Severe Roadway Tunnel Concentrations
The Baltimore Harbor Tunnel was chosen for the severe case. This
6700 foot long tunnel is part of the Baltimore Harbor Throughway. It has
two tubes, each with two lanes of one way traffic.
The latest ADT counts were obtained from the International Bridge,
Tunnel and Turnpike Association.(88) The traffic averages about 60,000
vehicles per day or approximately 3000 vehicles per hour per tube during
rush hour. The rush hour traffic speed through the tunnel averages 25 mph.
From Figure 10 (Section IV), it can be calculated that there are an average
of 165 vehicles in a tube at all times during the rush hour.
The tunnel has a fully transverse ventilation system. The total
maximum ventilation rate is 800,000 cfm per tube. To this rate is added
50,000 cfm for the piston effect of the cars entering the tube.(89)
For this tunnel also, the effective ventilation factor for the vehicle
emissions will be assumed to be 1.0. Since the exposure situation occurs at
the end of the rush hour, the pollutant concentrations have essentially reached
equilibrium and are equal to the total vehicle emission rate divided by the
effective ventilation rate. Using 1 g/mile per vehicle (ixio6 ug/mile), the
ambient concentration in the tunnel is:
(1X1Q6)(165)(25) _ ,, . 3
C3 = (i)(850,000)(0.02832)(60) ~ 2856 yg/m
Table 19 summarizes the physical variables and the ambient concentration
for the severe tunnel exposure situation.
3. Use of Tunnel Concentrations with Other Emission Factors
To use the ambient concentrations shown in Table 19 to obtain the
ambient concentrations other pollutants, multiply the ambient concen-
trations by the individual vehicle emission rate in grains per mile. For
example, to obtain the 1969 CO concentration in the typical exposure case
(for comparison with the severe case) multiply the ambient concentration
by the assumed 1969 CO emission factor at 45 mph of 45 g/mile.
85
-------
TABLE 19. AMBIENT CONCENTRATIONS FOR RECEPTOR IN ROADWAY TUNNEL -
TYPICAL AND SEVERE CASES
Tunnel location
Ventilation rate, one tube, cfma
Number of vehicles in one tube
Intake air concentration, g/m3
Emission factor, g/mile
vehicle speed, mph
Ventilation factor, intake air
Ventilation factor, exhaust
emissions
Concentration:
C2, contribution of intake air
Ug/m3
Co, contribution of exhaust
emi s s i on, Ug/m3
Typical Case
Lowery Hill
Minneapolis, MN
589,600
25
1
1.0
45
1.0
1.0
Severe Case
Baltimore Harbor
Baltimore, MD
850,000
165
1
1.0
25
1.0
1.0
1123
2856
including ventilation air from vehicle piston effect
86
-------
C3 = 1123 x 45 = 50,535 Ug/m3 or 44 ppm CO
While no measurements are available for comparison, this CO concentration
does not appear unreasonable.
The Baltimore Harbor Tunnel has been the subject of considerably
more study. Reference 89 reports on a CO measurement and modeling study
done in 1969 and 1970. To compute the CO concentration in the severe case
for 1969, the ambient concentration shown in Table 19 is multiplied by the
25 mph CO emission factor for that era. From CO emissions given in Reference
89, a CO value of 65 g/mile would appear to be appropriate. The tunnel concen-
tration is then:
C3 = 2856 x 65 = 185,640 yg/m3 or 163 ppm CO
This value of 163 ppm can be compared with the statement in Reference 89,
that the CO level in the Baltimore Harbor Tunnel "rarely exceeded 180 ppm
during peak traffic periods."
D. Street Canyon Concentrations
Two street canyons were chosen in Section IV to represent typical and
severe exposure situations. "Typical" and "severe" refer more to street
width than any other parameter. This is because the dispersion model chosen
is limited to a height to width ratio of 2.0. As explained in Section IV,
what constitutes severe traffic depends on the emission type being considered.
For some emissions, a severe case is low speed, high density; for others a
higher speed, lower density. Therefore, street width became the controlling
parameter. A four lane street was chosen as the typical exposure situation,
and a six lane street as the severe exposure situation. Each situation will
be modeled with two different traffic densities, using the model selected in
Section III.
1. Typical Street Canyon Concentrations
For the typical exposure situation, a street canyon in San Antonio,
Texas, was used. The street is Houston Street (one way) between the cross
streets of Navarro and St. Marys. (See Figure 27). The pertinent physical
data for this street canyon are:
•Block length - 128 meters
•"Canyon" width including sidewalks = 18.6 m
•Average building height = 33.7 m
•Height to width ratio: 1.81
•Street orientation - WeAt to East (270 - 090°)
•Distance from street center to sidewalk center = 8.8 m
The street canyon dispersion model chosen is valid for wind
directions within-±45 degrees of perpendicular to the street. Houston
Street runs West to East, therefore wind directions from SE to SW would
fall within these limits. The wind direction in San Antonio is between
SE and SW 47 percent of the time.
87
-------
NORTH
3.8 m
18.6 m
11.0 m
3.8 m
33.7 m
j, _ receptor
location
Navarro
NORTH
t
traffic direction
St. Marys
Limits of SE to SW wind
(45° to 135° relative to street)
Figure 27. Houston Street between Navarro and St. Marys,
San Antonio, Texas
88
-------
On a national basis, a typical (median) wind speed would be
approximately 7 knots (8 mi/hr, 3.6 m/sec). In San Antonio, when the wind
is from SE to SW, the wind speed is 8 to 12 miles/hr (3.6 m/sec to 5.4 m/sec)
approximately 38 percent of the time. Thus, the wind is from SE to SW at
8 to 12 miles/hr approximately 18 percent of the year.
To account for both the low speed, high traffic density case and
the higher speed, lower density case, two assumed traffic flow and speed
combinations were used.. For the four lanes of Houston Street, these
combinations were 800 vehicles per hour (0.222 vehicles/sec) at 5 mph
and 1600 vehicles per hour (0.444 vehicles/sec) at 20 mph. For comparison,
the actual rush hour traffic flow is approximately 1300 vehicles per hour.
Two receptor locations were chosen, The first was a point at the
midpoint of the sidewalk, 1.5 meters above the ground. The second receptor
location was in a vehicle in the street with a slant distance from the
assumed pollution line source of one meter. From data taken during the
development of the model, it appeared that the background concentration from
other vehicles in the area could be assumed to be equal to the street canyon
contribution at a height of three meters above the sidewalk.(25,26) since
the background is caused by other vehicles in the area, and since it is an
appreciable fraction of the total concentration, it will be included in the
calculation of the ambient concentration.
As with all the microscale situations modeled in this study, the
vehicle emission factor was selected to have a unity value. In this case,
one g/mile.
Using an emission factor of 1 g/mile per vehicle at 800 vehicles
per hour, the pollutant generation rate is:
Qr =
vehicle mile I
1 mile
1609 meters
X
0.222 vehicles
sec
-4 '
Q_ = 1.38 x 10 g/sec m
L
Using the data set forth above with the street canyon model from Section III
.yields the following .street contribution to the ambient concentration for a
person on the upwind sidewalk.
_ 7 x IQ6 1.38 x 1Q~4 = 7 xl.38 x 1Q2
CV "(3.6 + 0.5)(8.93 + 2) 4.1 x 10.93
G = 21.56 ug/m
89
-------
The background concentration is assumed equal to the vehicle con-
tribution at 3 meters above the sidewalk. In this case, the slant distance
is 9.3 meters, with a resulting concentration of 20.85 yg/m3. The total
concentration experienced by a person on the sidewalk with a traffic flow of
800 vehicles per hour is then:
CT = Cv + Cb = 21.6 + 20.8 = 42.4 yg/m3
A person in a vehicle in the street canyon will experience a higher
concentration since the applicable slant distance is smaller. For a person
in a vehicle, the slant distance is assumed to be one meter. The concentra-
tions at this point are:
Cv = 78.54 yg/m3
CT = Cy. + Cb = 78.5 + 20,8 = 99.3 yg/m3
For 1600 vehicles per hour (0.444 vehicles/sec), the pollutant
generation rate for a 1 g/mile emission factor is:
QL = 2.76 x icr4 g/sec m
For the sidewalk receptor the concentrations are:
= (7 106) x (2.76 10"4)
^ (3.6+0.5) x (8.93+2)
Cy = 43.11 yg/m3
(7 106) x (2.76 10~4)
T> (3.6+0.5) x (9.3+2)
^ = 41.70 yg/m3
QJ, = Cy + Cb = 43.11 + 41.70 = 84.8 yg/m3
For the receptor in a vehicle (slant distance of one meter) with a traffic
flow of 1600 vehicles per hour is:
Cy = 157.07 yg/m3
Ct = 157.1 + 47.1 = 198.8 yg/m3
The street canyon physical parameters and the ambient concentrations for both
traffic flows are summarized in Table 20, following the presentation of the
severe exposure situation.
90
-------
2. Severe Street Canyon Concentrations
Since the model limited the canyon height to width ratio to 2.0,
the severe case was chosen on the basis of number of traffic lanes, plus
the more practical consideration of availability of the necessary infor-
mation. A local aerial mapping company has stereographic aerial photo-
graphs of the Houston area from which building heights could be determined.
Since downtown Houston streets are in general six-lane one-way streets
with heavy traffic, a street canyon in Houston appeared to be a good
location for the severe exposure case. However, care had to be taken
to insure that a street canyon with less than 2.0 height-to-width ratio was
chosen. The final selection was Main Street (one-way) between the cross
streets of Capitol and Rusk. A sketch of the block showing building
heights and wind orientation is presented as. Figure 28. An aerial photograph
of this block is shown in Figure 29. The pertinent physical data for this
street canyon are:
•Block length = 86 m
•"Canyon" width including sidewalks = 27 m
•Average building height (leeward side) = 38 m
•Height to width ratio = 1.39
•Street orientation •- NE to SW (045 to 225°)
•Distance from street center to sidewalk center = 11.4 m
The assumed wind direction will be southeasterly. Since wind
direction 45 degrees on either side of perpendicular to the street is
considered to be perpendicular by the model, the wind direction can be
between East and South. At the Houston Intercontinental Airport at 6 a.m.
and 9 a.m., the wind is from these directions 32 and 44 percent of the
time, respectively.
For a severe exposure situation, the wind speed should be as
low as possible. There are two considerations. The first is that for a
given street canyon, there is a minimum wind speed that will produce a
helical wind flow pattern in the street which is required for use of the
model. The second consideration is that it should be a wind speed which
actually occurs at the chosen location. Using the equation for wind
penetration, it can be ascertained from Figure 16 in Section IV that for a
street 27 meters wide and a building height of 38 meters, the minimum wind
speed to have a helical flow pattern is approximately 0.9 meters/sec (2.1
miles per hour). From wind distribution at the Houston Intercontinental
Airport, at6 a.m., and 9 aim. the wind is between East and South at 0 to 3 miles
per hour (0 to 6.7 m/sec) 6.9 and 4.3 percent of the time, respectively.
Assuming the wind speed is evenly distributed in the interval, the severe
wind condition of 0.9 meters per second, perpendicular to the street does
indeed occur at the location chosen for the severe exposure situation.
91
-------
71 m
40 m
27 m
40 m
18 m
NORTH
limits of E to S wind
(45° to 135° relative
to street)
Figure 28. Main Street between Capitol and Rusk, Houston, Texas
92
-------
Figure 29. Aerial View of a Portion of Downtown Houston, Texas
-------
Again, two assumed traffic flow and speed combinations were used,
200 vehicles per hour per lane at 5 mph, and 400 vehicles per hour per-lane
at 20 mph. For the six lanes of main street, the two traffic flows are
1200 vehicles per hour and 2400 vehicles per hour.
The receptor locations were the same as for the typical situation.
The first, located 1.5 meters above the midpoint of the sidewalk, and the
second in a vehicle in the outside leeward lane.
Using a 1 g/mile per vehicle emission factor, the total emission
strength for 1200 vehicles is 2.07 x 10~4 g/sec m. The street canyon con-
tribution to the ambient concentration experienced by a receptor at the
sidewalk midpoint is:
(7X1Q6)(2.07X10"4) 7 x 2.07 x io2
C\7 =
"v (0.9+0.5)(11.52+2) 1.4 x 13.52
Cv = 76.55 yg/m3
Again, it is assumed that the background concentration caused by
other vehicles in the area is equal to the street canyon contribution at
3 meters above the canyon edge of the sidewalk. For 3 meters above the
sidewalk, the slant distance is 14.04 meters. The resulting concentration is
64.53 yg/m3. The total concentration experienced by a receptor at the
midpoint of the leeward sidewalk when the traffic flow is 1200 vehicles
per hour is:
Ct = 76.6 yg/m3 +64.5 yg/m3 = 141.1 yg/m3
For a receptor in a vehicle in the outside leeward lane of the
street canyon, a slant distance of 7.72 meters can be assumed. The concentrations
at this point, for a traffic flow of 1200 vehicles per hour are:
Cv = 134.07 yg/m3
Ct = 134.1 + 64.5 = 198.6 yg/m3
For 2400 vehicles per hour the total emission strength is
4.14 x 10~4 g/sec m. For a receptor on the upwind sidewalk, and a traffic
flow of 2400 vehicles per hour, the street canyon contribution to the
ambient concentration is:
Cy = 153.11 yg/m3
The background concentration from other vehicles in the area is
again assumed to be equal to the street contribution at three meters above
the sidewalk. The background concentration for 2400 vehicles per hour is:
Cb = 129.05 yg/m3
94
-------
The total concentration for a receptor on the sidewalk with a traffic
flow of 2400 vehicles per hour is:
Ct = 153.1 + 129.0 = 282.1 yg/m3
For a receptor in a vehicle in the outside leeward lane of the street
canyon, and a traffic flow of 2400 vehicles per hour, the concentrations are:
Cv = 268.14 yg/m3
Ct = 268.1 + 129.0 = 397.1 yg/m3
A summary of the pertinent physical parameters and the ambient
concentrations for both the typical and severe street canyon exposure
situations is presented in Table 20.
3. Use of Street Canyon Concentrations With Other Emission Factors
To obtain the ambient air concentrations for these situations
with other vehicle emission rates in grams/mile, simply multiply the ambient
concentration obtained assuming 1 g/mile by the new emission rate.
For example, if an urban cycle CO emission rate for 1970 is taken as
85.6 g/mile then the ambient concentration on the sidewalk of the typical
street canyon with 1600 vehicles per hour traffic flow would be:
Cv = 43.1 x 85.6 = 3689 yg/m3 or 3.69 mg/m3
C = 41.7 x 85.6 = 3569 yg/m3 or 3.57 mg/m3
b
C = 84.8 x 85.6 = 7258 yg/m3 or 7.26 mg/m3
The total ambient concentration for this location would have been
approximately 7.26 mg/m3, or 6.3 ppm CO. From the measured street canyon
CO values found in the literature, this is a representative concentration
for the 1970 time period.
As another example, consider the severe street canyon situation
using Main Street in Houston with 2400 vehicles per hour. Again, calculate
the CO emission rates for 1970:
C = 153.1 x 85.6 = 13105 yg/m3 or 13.11 mg/m3
Q = 129.0 x 85.6 = 11042 yg/m3 or 11.04 mg/m3
C = 282.1 x 85.6 = 24148 yg/m3 or 24.15 mg/m3
The total ambient CO concentration for this location and traffic
flow would have been 24.15 mg/m3, or 21.0 ppm. This value can be compared
to the maximum one hour average CO reading of 29.5 ppm (occurring at 9 am)
taken by the Texas Air Control Board at Main and Capital during a survey
in October 1968.
95
-------
TABLE 20. STREET CANYON AMBIENT AIR CONCENTRATIONS
Typical Situation
Houston St., San Antonio, TX
Severe Situation
Main St., Houston, TX
VO
800 vehicles/hr
Sidewalk
1.
2.
3.
4.
5.
6.
7.
8.
Rooftop wind speed, m/sec
Rooftop wind direction
relative to steeet, degrees
Traffic volume vehicles/hr
Emission Factor, g/vehicle mi
Receptor slant distance, m
Concentration from vehicles
in street canyon, yg/m3
Estimated background, yg/m3
Total concentration, yg/m3
3.6
45-135
800
1.0
8.9
21.6
20.8
42.4
In-Vehicle
3.6
45-135
800
1.0
1.0
78.5
20.8
99.3
1600 vehicles/hr
Sidewalk
3.6
45-135
1600
1.0
8.9
43.1
41.7
84.8
In-Vehicle
3.6
45-135
1600
1.0
1.0
157.1
41.7
198.8
1200 vehicles/hr
Sidewalk
0.9
45-135
1200
1.0
11.5
76.6
64.5
141.1
In-Vehicle
0.9
45-135
1200
1.0
7.7
134.1
64.5
198.6
2400 vehicles/hr
Sidewalk
0.9
45-135
2400
1.0
11.5
153.1
129.0
282.1
In-Vehicle
0.9
45-135
2400
1.0
7.7
268.1
129.0
397.1
-------
E. Urban Expressway Concentrations
The expressway exposures considered were of two types; the receptor in
a vehicle on the expressway, and the receptor beside (within one kilometer
downwind of) the expressway. For the on-expressway situation, two urban
expressways were chosen to represent the typical and severe exposure situation.
The ambient concentrations for a commuter in traffic were calculated using
the ONEX computer program developed for this project, and explained in
Section III. Since it has been shown in Section IV that all relative wind
directions are equally likely when considering all urban freeways, the
ambient concentrations were calculated over the 90 degree quadrant that
would produce the highest ambient concentration. The concentrations were
also calculated over a range of wind speeds from one to six m/sec (2.2 to
13.4 mph) .
While the expressway commuter exposure is short term (less than one
hour), the downwind exposures of interest are both short term and long
term. Since ambient concentrations decay quickly downwind of a line source,
only a severe exposure situation was investigated. In addition to the
parameters covered in Section IV for a severe case expressway, the long
term severe exposure situation required that the wind be perpendicular to
the expressway a large portion of the year. For this reason, a different
physical situation was chosen for the severe downwind exposure case than was
used for the severe on-expressway situation.
1. Typical On-Expressway Concentrations
The typical exposure situation was chosen at IH 410 on the
west side of San Antonio, between Valley Hi Drive and U.S. 90. This is
a relatively straight, level, four lane section of urban expressway
running in a SSE and NNW direction. Average daily traffic at this point
is approximately 30,000 vehicles per day. The receptor was chosen as a
vehicle occupant in the outside southbound lane during the morning peak
traffic period.
Observation of traffic at this point showed that vehicles tended
to be-bunched in packs of approximately 20 cars, evenly distributed between
lanes, with an average of approximately 1.5 seconds between vehicles in
the same lane. The average speed was approximately 55 miles per hour.
Figure 30 is a sketch of the physical situation. The complete set of input
parameters for the typical exposure situation is listed below:
•Highway dimensions
-4 lanes, each 3.66 meters (12 feet) wide
-median, 13.41 meters (44 feet) wide
•Vehicle volume
-900 vehicles per hour per lane on southbound (receptor) side
-500 vehicles per hour per lane on northbound side
97
-------
TRAFFIC
1800 vehicles
per hour
30.2 m
13.4 m-
3.4 m
1410 west, southbound
lanes are 3.66 m (12 feet) wide
median is 13.41 m (44 feet) wide
= receptor
en
TRAFFIC
1000 vehicles
per hour
LIMITS OF
WIND DIRECTION
CONSIDERED
1410 west, northbound
Figure 30 . Sketch of expressway section for typical expressway exposure
-------
•Point source (vehicles on receptor side) information
-emission strength, 0.0153 g/sec, (1 g/mile at 55 mph)
-average of four vehicle lengths between vehicles
-average vehicle length, 6.7 miles (22 feet assumes 80 percent
cars, 20 percent trucks
-exhaust height assumed to be 0.3 meters (one foot)
-10 vehicles in each lane ahead of receptor vehicle
•Line source (vehicles in opposite lanes)
-line source strength 0.0862 x 10~3 g/sec m lane (1 g/mile per
vehicle, 500 vehicles per hour)
-bouyancy flux 0.0191 m3/sec3
•Receptor
-located at center of outside downwind lane
-height 1.22 meters (4 feet)
•Meteorology
-wind direction, from 357.5 degrees to 270 degrees relative to
receptor vehicle direction
-wind speed, from 1 to 6 m/sec (2.2 to 13.4mph)
-stability, Pasquill Class C (slightly unstable)
The calculated ambient concentrations are shown in Table 21 and
in graphically in Figure 31. For the whole range of wind directions and
speeds, the ambient concentration experienced by a receptor in the outside
downwind lane varies from 61 to!24yg/m3.
The highest ambient concentrations occur when the wind is close
to perpendicular to the road. This is because the vehicle speed is so
much greater than the wind speed, that the wind direction has little
effect on the relative wind speed and direction for the vehicle's traveling
in the same direction as the receptor. Thus the contribution from the
vehicles on the receptors side of the road decreased only slightly as the
wind direction changes to more perpendicular. The contribution from the
vehicles traveling in the opposite direction, increase by a greater
amount as the wind becomes more perpendicular, resulting in a net increase
in ambient concentrations. Note however, that the wind direction for
maximum concentration varies with wind speed.
2. Severe On-Expressway Concentrations
The severe expressway situation was chosen as the Santa Monica
Freeway (IH 10) on the west side of Los Angeles between the cross streets
of Washington and La Brea. This is a relatively, straight, level, 10
lane section of urban freeway running in a WSW to ENE direction. Average
daily traffic at this point is approximately 200,000 vehicles per day.
The receptor chosen was a vehicle occupant on the eastbound side of the
freeway during the morning peak traffic period. Two receptors were
initially used to check whether the middle or outside lane had the largest
concentrations. It was determined that a receptor in the middle lane would
be subject to the highest concentrations. Therefore, the concentrations
presented in this section are for a receptor in the middle eastbound lane.
99
-------
TABLE 21. AMBIENT CONCENTRATION FOR RECEPTOR ON EXPRESSWAY
TYPICAL EXPRESSWAY EXPOSURE SITUATION
Outside downwind lane
Wind Direction Wind Speed Ambient Concentration
Degrees Relative m/sec (mph) yig/m^
357.5 1.0 (2.2) 120
355.0 1.0 (2.2) 120
340.0 1.0 (2.2) 122
315.0 1.0 (2.2) 124
270.0 1.0 (2.2) 122
357.5 2.0 (4.5) 114
355.0 2.0 (4.5) 113
340.0 2.0 (4.5) 109
315.0 2.0 (4.5) 103
270.0 2.0 (4.5) 95
357.5 3.0 (6.7) 109
355.0 3.0 (6.7) 107
340.0 3.0 (6.7) 99
315.0 3.0 (6.7) 85
270.0 3.0 (6.7) 77
357.5 6.0 (13.4) 96
355.0 6.0 (13.4) 92
340.0 6.0 (13.4) 75
315.0 6.0 (13.4) 61
270.0 6.0 (13.4) 72
100
-------
n
B
§
•H
-P
-P
OJ
O
C
8
130 |-
120
110
100
90
80
70
60
50«-i
270
True Wind Speed
O 1 m/sec
D 2 m/sec
A 3 m/sec
O 6 m/sec
I
J_
I
280 290
300
310
320 330
340
350
360
Wind direction, relative to direction of
receptor vehicle, degrees
Figure 31. Ambient pollutant concentration as a function
of wind direction. Typical expressway case.
101
-------
Traffic counts for this section of freeway at 15 minute intervals
were obtained from the District 7 Office of the California Department of
Transportation along with the freeway dimensions <
The complete set of input parameters for the severe exposure situation is
listed below:
•Highway dimensions
-10 lanes, each 3.35 meters (11 feet) wide
-median, 6.71 meters (22 feet) wide
•Vehicle Volume
-2000 vehicles per hour per lane on eastbound (receptor) side
-1675 vehicles per hour per lane on westbound side
•Point Source (vehicles on receptor side) information
-emission strength, 0.01389 (1 g/mile at 50 miles/hr)
-average of two vehicle lengths between vehicles
-average vehicle length 6.7 meters (22 feet) assumes 80 percent
cars , 20 percent trucks
-exhaust height, 0.3 meters (one foot)
-10 vehicles per lane ahead of receptor vehicle
•Line Source (vehicles in opposite lanes)
-line source strength, 0.2982 x 10~3 g/sec m lane (1 g/mile
per vehicle, 1675 vehicles/hr per lane)
-bouyancy flux 0.0638 m
•Receptor
-located at center of center lane
- height 1.22 meters (4 feet)
•Meteorology
-wind direction, from 357.5 degrees to 270 degrees relative to
receptor vehicle direction
-wind speed, 1 to 6 m/sec (2.2 to 13.4 mph)
- stability, Pasquill Class C (slightly unstable)
For a severe exposure situation, the wind conditions should
be those that cause the highest concentrations within the range of
naturally occurring conditions and within the capability of the model.
It has been shown that all wind directions, relative to the expressway
were equally likely when all expressways in the county are considered.
The highest concentrations are also caused by the lowest speed wind.
For the model used, one m/sec (2.2 miles/hr) is the lowest useable speed.
Therefore, for a severe exposure situation the wind will be perpendicular
to the .road at one m/sec. For this situation, the wind would be from the
NNW at 2.2 mph. A check of the wind direction and speed matrix for Los
Angeles shows that the wind is from the NNW at 0 to 3 knots (0 to 3.5 mph)
approximately 0.5 percent of the time. Thus, the severe wind conditions
represent a real, but uncommon, wind condition. As with the typical
exposure situation, other wind direction and speeds were also investigated
for comparison purposes .
102
-------
As with all the microscale situations examined in this study, the
emission factor per vehicle was set to one gram per mile to permit easy
scaling of the results for other emission factors.
The ambient concentrations for this situation as calculated using
the ONEX computer program are shown in Table 22. Note that for this situation
the highest concentrations are obtained with the lowest wind speed in a
direction perpendicular to the road. However, as the wind speed increases,
the maximum ambient concentration shifts toward parallel with the road. This
shift can be seen better in the graphical presentation of the concentrations,
Figure 32. For the severe case, the concentrations at the receptor varied
from 148 yg/m3 to 506 yg/m3.
3. Severe Beside-Expressway Concentrations
To obtain a severe situation for a receptor downwind of an expressway
for both short and long term exposure, it was necessary to have a large
expressway located in such a manner that the prevailing wind was perpendicular
to the expressway a large part of the time.
The actual site chosen was 1-10 at Silber Road on the west side
of the city of Houston. At this location, 1-10 runs due east and west,
with four lanes in each direction. The 1977 traffic count was 167,860
vehicles per day. A sketch of this expressway section is shown in Figure 33.
An examination of the Houston wind direction data indicated that
the wind is from 90 compass degrees to 270 compass degrees approximately 60
percent of the time. This means that the highway vehicle emissions will be
dispersed to the north of the expressway approximately 60 percent of the time.
The severe short term exposure situation was chosen as the morning
rush hour. The traffic count at this time was taken as nine percent of the
ADT count, or 1890 vehicles per hour per lane.
For a severe beside-expressway situation, the wind should be per-
pendicular to the expressway at a speed as low as permitted by the model.
The Houston wind direction data indicated that during the morning rush hour
the wind was from the south between 1 and 1.54 m/sec (0 and 3.4 mph), approxi-
mately one percent of the time. The atmospheric stability was within one
stability class of neutral approximately 90 percent of the time. Therefore,
the severe exposure case represents a situation that does occur.
For the short term exposure, the line source portion of the ONEX
exposure model was used to determine the concentrations at various distances
from the north side of the expressway. The concentrations obtained from
applying the model to this situation are shown in Table 23. A graphical
presentation is shown in Figure 34 following the discussion of the long
term exposure.
103
-------
TABLE 22. AMBIENT CONCENTRATION FOR RECEPTOR ON EXPRESSWAY -
SEVERE EXPRESSWAY EXPOSURE SITUATION
Wind Direction
Degrees Relative
357.5
355.0
340.0
315.0
270.0
357.5
355.0
340.0
315.0
270.0
Wind Speed
m/sec (mph)
1.0
1.0
1.0
1.0
1.0
(2.2)
(2.2)
(2.2)
(2.2)
(2.2)
2.0 (4.5)
2.0 (4.5)
2.0 (4.5)
2.0 (4.5)
2.0 (4.5)
Outside downwind lane
Ambient Concentration
454
467
494
506
495
453
458
428
400
369
357.5
355.0
340.0
315.0
270.0
357.5
355.0
340.0
315.0
270.0
3.0
3.0
6.0
6.0
6.0
(6.7)
(6.7)
3.0 (6.7)
3.0 (6.7)
3.0 (6.7)
6.0 (13.4)
6.0 (13.4)
(13.4)
(13.4)
(13.4)
444
435
375
327
285
399
366
275
200
148
104
-------
600
500
g
Cn
tn
c
o
•H
-P
c
0)
o
c
0
o
.400
300
200
100
Wind Speed
O 1 m/sec
G 2 m/sec
A 3 m/sec
0 6 m/sec
I
I
I
270 280
290
300
310
320
330 340
350
360
Wind direction relative to direction of
receptor vehicle, degrees
Figure 32. Concentrations experienced by receptor traveling
on expressway. Severe expressway case.
105.
-------
downwind receptors
10m
t
5m
1
lanes are 12 feet wide
median is 20 feet wide
1-10 Westbound
Traffic
874 vehicles/hr/lane
average
1890 vehicles/hr/lane
(rush hr)
1-10 Eastbound
Traffic
874 vehicles/hr/lane
average
1890 vehicles/hr/lane
(rush hr)
wind direction: 90° at 1 m/sec for short term case
Figure 33. 1-10 at Silber Road, Houston, Texas
-------
TABLE 23. AMBIENT CONCENTRATIONS FOR A RECEPTOR DOWNWIND OP AN EXPRESSWAY
Distance downwind
from road edge,
meters
1
5
10
25
50
100
500
1000
Ambient Concentration
Ug/m3
397
368
334
248
171
105
26.3
13.6
Site Conditions:
1. Distance from lane center to downwind
road edge in meters:
Lane 1 = 1.8, Lane 2 = 5.5
Lane 3 = "9.1, Lane 4 = 12.8
Lane 5 = 22.6, Lane 6 = 26.2
Lane 7 = 29.9, Lane 8 = 33.5
2. Number of vehicles per lane per hour: 1890
3. Emissions per vehicle, g/mile 1.0
.4. Emissions per lane, yg/sec m: 326
•5. Bouyancy flux,m3/sec3 (per lane): 0.072
6. Wind direction relative to road: 90°
7. Wind speed: 1 m/sec
8. Stability: Neutral
9. Source height: 0.5 meters
10. Receptor height: 1.8 meters
107
-------
The long-term exposure assessment is for persons living and working
close to the expressway. For this study, long-term will be taken to be the
annual average. Most line source dispersion models for highway vehicle
emissions are for short-term (typically one hour) estimation, since wind
and atmospheric stability are constant only for short periods. The problem
is to obtain annual averages from the short-term models. Using a method
suggested by Turner(90), H. E. Sievers of the Texas Air Control Board staff
modified three line source models extensively to give annual arithmetic
means.(91) The method employed involves the computation of the ambient
particulate concentration for each of the 576 defined meteorological com-*
binations (16 wind directions, times 6 stability classes, times 6 wind speed
classes), then multiplying by the frequency of occurrence of each particular
meteorological combination. All non-zero concentrations are then summed to
obtain the annual arithmetic mean.
Fortuitously, Reference 91 contains a plot of the annual arithmetic
concentrations as a function of distance from the roadway on the north side
of an east-west expressway, using Houston meteorological data, assumed
traffic count, and vehicle emission factors. These concentrations were
generated using the Chock-G.M. line source model, the same calculations used
in the ONEX model. In this model, as in all Gaussian plume models, the
emission rate (the product of the grams/km/vehicle and vehicles/hr) is a
multiplier to the dispersion equation. Thus, the results from Sievers1
study need only to be multiplied by the ratio of the emission rate in this
study to the emission rate in that study to obtain the annual arithmetic
average particulate concentrations at the chosen site. The estimated
annual arithmetic average particulate concentrations on the north side of
1-10 at Silber Road are presented in Table 24. Figure 34 presents a
comparison of the severe short-term exposure and the annual average
exposure for all wind conditions.
4. Use of Expressway Concentrations With Other Emission Factors
To illustrate the use of the expressway exposure results and to
put them in the context of a commonly measured pollutant, the CO concen-
tration experienced by the receptor in each of the expressway situations
will be calculated.
For the typical on-expressway exposure case in San Antonio, the current CO
emission factor is assumed to be 50 g/mile at 55 mph. The CO concentration
at the most severe wind condition is then: ~V -
124 yg/m3 x 50 = 6200 yg/m3 or 5.4 ppm CO
For the severe on-expressway exposure case in Los Angeles, the current CO
emission factor is assumed to be 50 g/mile at 50 mph. The CO concentration
at the most severe wind condition is then:
506 yg/m3 x 50 = 25,300 yg/m3 or 22.0 ppm CO
108
-------
400
§
•H
-P
(fl
0)
u
g
u
200
180
160
120
100
80
60
40
20
c
OJ
u
G
O
U
350
g 300
tn
O
•H
250
200
150
0 10 20 30 40 50 60 70
Distance downwind, meters
short term
— — _ annual average
200 400 600 800
Distance downwind, meters
1000
Figure 34. Concentrations downwind of expressway
109
-------
TABLE 24. ANNUAL AVERAGE CONCENTRATIONS FOR A RECEPTOR ON THE NORTHSIDE
OF I-10 AT SILBER ROAD, HOUSTON, TEXAS
Distance downwind
from road edge,
meters
1
5
10
25
50
100
500
1000
Ambient Concentration
yg/m3
61
55
48
35
23
14
4
1.6
Site Conditions:
1. Distance from lane center to downwind
road edge in meters:
Lane 1 = 1.8, Lane 2 = 5.5
Lane 3 = 9.1, Lane 4 = 12.8
Lane 5 = 22.6, Lane 6 = 26.2
Lane 7 = 29.9, Lane 8 = 33.5
2. Number of vehicles per lane per hour: 874
3. Emissions per lane, g/sec m:
4. Bouyancy flux,m3/sec3 (per lane):
5. Wind direction relative to road:
6. Wind Speed:
7. Stability:
8. Source height:
9. Receptor height:
150.9
0.033
10 year average
of observation
by time of day
0 .5 meters
1.8 meters
110
-------
A 1977 study measured the CO level inside a moving vehicle on the
Santa Monica Freeway at 10 am in April 1977.(92) For a 55 mph traffic speed,
the CO level inside the car was approximately 5 ppm. A recent British study
measured CO concentrations in traffic, both inside and outside a moving
vehicle.(93) in expressway type driving, internal vehicle CO levels of
10 ppm were common. The ratio of inside to outside concentrations for 11
cars ranged from 0.35 to 0.75. Applying the minimum ratio to both studies
referenced above would give expressway CO concentrations of between 15 and
30 ppm. Thus, the calculated severe case CO concentration of 22 ppm is in
agreement with the measured values.
To obtain ambient CO concentrations for the short-term beside-
expressway situation, a 1979 CO emission factor of 75 g/mile is assumed. The
concentrations with distance are then:
5 meters: 352 x 75 = 26,400 yg/m3 or 23 ppm CO
10 meters: 334 x 75 = 25,800 yg/m3 or 22 ppm CO
50 meters: 171 x 75 = 12,825 yg/m3 or 11 ppm CO
100 meters: 105 x 75 = 7,875 yg/m3 or 7 ppm CO
500 meters: 26.3 x 75 = 1,972 yg/m3 or 2 ppm CO
The long-term CO concentrations, assuming the same CO emission factors,
are:
5 meters: 55 x 75 = 4,125 yg/m3 or 3.6 ppm CO
10 meters: 48 x 75 = 3,600 yg/m3 or 3.1 ppm CO
50 meters: 27 x 75 = 2,025 yg/m3 or 1.8 ppm CO
100 meters: 14 x 75 = 1,050 yg/m3 or 0.9 ppm CO
500 meters: 4 x 75 = 300 yg/m3 or 0.3 ppm CO
F. Localized Area Concentrations
The one localized area for which some information was available was
parking lots. Sizeable parking lots occur at facilities such as shopping
malls and sports stadiums. Despite the apparent possibility of high ambient
air concentration of automotive pollutants in parking lots, EPA studies
have shown that this is not the case.^96'94) The referenced studies measured
CO in a shopping mall parking lot and in sports stadium parking lots, respec-
tively. The one hour CO concentration in the shopping mall parking lot never
exceeded 13 ppm, and 90 percent of the time it was less than 5 ppm.
Ill
-------
The sports stadiums, which of course were event-oriented, had higher
CO levels. In Reference 95, CO was measured before and after several base-
ball games at Threerivers Stadium (Pittsburgh) and Atlanta Stadium. At both
stadiums the highest CO levels were recorded approximately one-half hour
after.end-of-game. In Pittsburgh, the highest CO measured in 15-minute
bagged samples around the stadium and in the stadium parking lots was 23.5
ppm. This value is about 22 ppm above the usual background.
In Atlanta, the highest 15-minute bagged reading was 50.5 ppm, or about
45 ppm above the post-game background. The Atlanta reading was taken on
the sidewalk next to the main egress street for the stadium. This situation
should probably be classed as a roadway exposure rather than a parking lot
exposure. Other bag samples taken farther away from streets, in the area
of the stadium entrances, never exceeded 9 ppm CO.
Thus, parking lots do not appear to produce levels of automotive pol-
lutants as high as the other microscale situations examined. Should some
pollutant be found that would affect health at these low levels, it would
be necessary to develop a dispersion model for this source type and expend
further effort in defining the magnitude of the exposure occurrence on a
national basis.
112
-------
VI. RECOMMENDATIONS
During the course of this study, a number of deficiencies in available
information were found. If the approach taken in this project for evaluating
the possible health and welfare effectsv of unregulated pollutants is to be
developed further, several areas need additional study. From the experience
gained during this project, the following are recommendations for further
study:
1. Enclosed Space Dispersion Model - The enclosed space model utilized
in this study is not well developed. At the least, a thorough study of the
effective ventilation factor is required. A better solution would be the
development of a new model using the ventilation flow patterns within an
enclosed space, and the fundamental fluid dynamics equation for dispersion
of the pollutant.
2. Tunnel Dispersion Model - A model for pollutant dispersion within
a tunnel considering the fundamental fluid dynamics equaltions has been
developed. This model, TUNVEN, is recommended for all future work involving
tunnel pollutant concentrations.
3. Street Canyon Dispersion Model - Since it is likely that very deep
street canyons will have higher pollutant levels than typical street canyons,
a dispersion model for deep street canyons is needed. It is recommended that
a dispersion model be developed for street canyons having height to width
ratios greater than 2.0.
4. On Expressway Dispersion Model - The dispersion model created for
the on expressway situation in this study needs additional work. In particular,
it is recommended that further study be done on the value of the dispersion
coefficients in traffic.
5. Localized Source Dispersion Model - No model was found for concen-
trations within (as opposed to downwind of) an area source. It is recommended
that such a model be developed.
6. Street Canyon Description - The range of physical variables for
existing street canyons is not well defined. It is recommended that further
information be collected on street canyon statistics. Such information should
include number of miles by height to width ratio and by traffic density.
7. Localized Area Description - In addition to parking lots (downtown,
shopping center, and sports stadiums), other area sources can be envisioned.
These include truck terminals, industrial areas, and construction sites. No
compiled statistics were found on these sites. If it is determined that
these sites do constitute an exposure situation requiring concentration
estimates, considerable effort will be needed to properly define the range
of physical variables needed for a dispersion model.
113
-------
REFERENCES
1. "Report by the Committee on Motor Vehicle Emissions." National Research
Council, National Academy of Sciences, November 1974.
2. Springer, K. J., and Stahman, R. C. , "Diesel Car Emissions—Emphasis on
Particulate and Sulfate." SAE Paper 770254.
3. Gentel, J. E., Manary, O. J., and Valenta, J. C., "Characterization of
Particulates and Other Nonregulated Emissions from Mobile Sources and
the Effect of Exhaust Emissions Control Devices on These Emissions."
EPA OAWP Publication APTD-1567, March 1973.
4. Smith, L. R., "Characterization of Emissions for Motor Vehicles Designed
for Low NOX Emissions." Draft Final Report to the Environmental Protection
Agency under Contract 68-02-2497, February 1980.
5. Urban, C. M., "Regulated and Unregulated Exhaust Emissions from Malfunctioning
Three-Way Catalyst Gasoline Automobiles." Final Report to the Environmental
Protection Agency under Contract No. 68-03-2588, EPA Publication 460/3-80-004,
January 1980.
6. Urban, C. M., "Regulated and Unregulated Exhaust Emissions from Malfunctioning
Non-Catalyst and Oxidation Catalyst Gasoline Automobiles." Final Report
to the Environmental Protection Agency under Contract No. 68-03-2499, EPA
Publication 460/3-80-003, January 1980.
7. 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 No. 68-03-2692, EPA Publica-.
tion No. 460/3-80-005, January 1980.
8. The Clean Air Act as Amended August 1977. 42 U.S.C. 1857 et. seq. Serial
No. 95-11. Section 202(a)(4)(A).
9. OMSAPC Advisory Circular No. 76, "Compliance With the Requirements of
Sections 202(a)(4) and 206(a)(3) of the Clean Air Act." U. S. Environmental
Protection Agency, Office of Air and Waste Management, June 28, 1978.
10. Dietzmann, H. E., et al, "Analytical Procedures for Characterizing Un-
regulated, Pollutant Emissions from Motor Vehicles." Environmental Protec-
tion Agency Report No. 600/2-79-017, February 1979.
114
-------
REFERENCES (Cont'd)
11. Smith, L. R., et al, "Analytical Procedures for Characterizing Unregu-
lated Emissions from Vehicles Using Middle-Distillate Fuels." Interim
Report to the Environmental Protection Agnecy under Contract 68-02-2703.
EPA Publication No. 600/2-80-068, April 1980.
12. "The Diesel Emissions Research Program." Environmental Protection Agency,
Mobile Source Research Committee, Washington, D.C., undated (circa. 1980).
13. Stern, A. C. (editor), Air Pollution, 3rd Edition, Vol. 1, Academic Press,
New York, NY 1976, p. 532.
14. Brief, R. S., "Simple Way to Determine Air Contaminants," Air Engineering,
Vol. 2, No. 4, p. 39, April 1960.
15. Turk, A., "Measurements of Odorous Vapors in Test Chambers: Theoretical."
ASHRAE J. Oct. 1963, p. 55.
16. Jones, R. M. and Pagan, R., "Application of Mathematical Model for the
Buildup of Carbon Monoxide from Cigarette Smoking in Rooms and Houses."
ASHRAE J. August 1974, p. 49.
17. Wang, T. C., "Air Quality in the Conference Room," Chemtech, March 1978,
p. 154.
18. Porstendorfer, J., Wicke, A. and Schraub, A., "The Influence of Exhala-
tion, Ventialation and Deposition Processes Upon the Concentrations of
Radon (222Rn), Thorn (220Rn) and their Decay Products in Room Air."
Health Physics, Vol. 34, No. 5, p. 465, May 1978.
19. Esmen, N. A., "Characterization of Contaminant Concentrations in Enclosed
Spaces." Environmental Science and Technology, Volume 12, No. 3,
March 1978, p. 337.
20. Ishiza, Y., "General Equation for the Estimation of Indoor Pollution."
Environmental Science and Technology, Volume 14, No. 10, October 1980,
p. 1254.
21. 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-78-184, January 1980.
22. Schlaug, R. N., et al., "Management of Air Qulaity in and Near Highway
Tunnels", Interim Report under Contract DOT-FH-11-8538 to Science
Applications, Inc. Report FHWA-RD-78-184, January 1980.
23. Schlaug, R. N., "User Guide for the TUNVEN and DUCT Programs." Final
Report under Contract DOT-FH-11-8538 to Science Applications, Inc.
Report No. FHWA-RD-78-187.
115
-------
REFERENCES (Cont'd)
24. Johnson, W. D., Ludwig, F. L., Moon, A. E., "Development of a Practical,
Multipurpose Urban Diffusion Model for Carbon Monoxide." In: Proceedings
of Symposium on Multiple-Source Urban Diffusion Models, EPA Publication
AP-86, 19:70.
25. Johnson, W. B., et al, "An Urban Diffusion Simulation Model for Carbon
Monoxide." APCA J. Vol. 23, No. 6, p. 490, June 1973.
26. Dabberdt, W. F., Ludwig, F. L., and Johnson, W. B., "Validation and
Application of an Urban Diffusion Model for Vehicular Pollutants."
Atmospheric Environment, Vol. 7, p. 603, 1973.
27. Chang, T. Y., and Weinstock, B., "Urban CO Concentrations and Vehicle
Emissions," APCA J. Vol. 23, No. 8, p. 691, August 1973.
28. Ott, W. and Eliassen, R. , "A Survey Technique for Determining the Re-
presentativeness of Urban Air Monitoring Stations with Respect to Carbon
Monoxide." APCA J. Vol. 23, No. 8, p. 685, August 1973.
29. Nicholson, S. E., "A Pollution Model for Street-Level Air." Atmospheric
Environment, Vol. 9, p. 19, 1975.
30. Latshaw, G. L., et al, "Microscale CO Concentrations and Wind Characteristics
on New York City Streets," p. 195 of Assessing Transportation - Related Air
Quality Impacts, Conference of October 22-24, 1975, Washington, D.C.,
Sponsored by U.S. DOT and EPA.
31. Wedding, J. B., Lombardi, D. J., Cermak, J. E., "Wind Tunnel Study of
Gaseous Pollutants in City Street Canyons." APCA J. Vol. 27, No. 6, p. 557,
June 1977.
32. Kennedy, I. M. and Kent, J. H., "Wind Tunnel Modelling of Carbon Monoxide
Dispersal in City Streets." Atmospheric Environment Vol. 11, p. 541, 1977.
33. "Street Canyon Dispersion Model" Appendix C of Guidelines for Air Quality
Maintenance Planning and Analysis. Vol. 9 (revised): Evaluating Indirect
Sources. EPA 450/4-78-001, September 1978.
34. Norbeck, J. M., Chang, T. Y., Weinstock, B., "Effects of New York City
Taxi' Strike on CO Concentration in Midtown Manhattan." APCA J. Vol. 29,
No. 8, p. 845, August 1979.
35. Rashidi, M. and Massondi, M. S., "A Study of the Relationship of Street
Level Carbon Monoxide Concentrations to Traffic Parameters." Atmos-
pheric Environment, Vol. 14, p. 27, 1980.
36. Sontowski, J., New York City Dept. of Environmental Protection. Telephone
conversation with M. N. Ingalls, Southwest Research Institute, April 28, 1980.
116
-------
REFERENCES (Cont'd)
37. Zimmerman, J. R., and Thompson, R. S. User's Guide for Hiway: A Highway
Air Pollution Model. Prepared by Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, N.C. Publication
No. EPA 650/4-74-008. February 1975.
38. Ward, C. E., Jr., Ranzieri, A. J., and Shirley, E. C. Caline-2 - An
Improved Microscale Model for the Dispersion of Air Pollutants from a Line
Source. Interim report prepared by the California Department of Transpor-
tation, Sacramento, CA, under Contract No. DOT-FH-11-7730 to U.S. Department
of Transportation, Washington, D.C. Publication NO. FHWA-RD-77-74. June 1977.
39. Bullin, J. A., and Polasek, J. C. Traps 52 User's Guide: Analytical and
Experimental Assessment of Highway Impact on Air Quality. Research report
prepared by Texas ASM University, College Station, TX, under Contract
No. 2-8-75-218 to Texas State Department of Highways and Public Transpor-
tation, Austin, TX. Publication No. FHWATX78-218-3. March 1978.
40. Bullin, J. A., and Polasek, J. C. Traps II User's Guide: Analytical
and Experimental Assessment of Highway Impact on Air Quality. Research
report prepared by Texas ASM University, College Station, TX, under
Contract No. 2-8-75-218 to Texas State Department of Highways and Public
Transportation, Austin, TX. Publication No. FHWATX78-218-2. March 1978.
41. DeTar, D., "A New Model for Estimating Concentrations of Substances
Emitted from a Line Source." APCA J. Vol. 30, No. 1, January 1980.
42. Chock, D. P. "A Simple Line-Source Model for Dispersion Near Roadways."
Atmospheric Environment, Vol. 12, p. 823, 1978.
43. Noll, K. E., "Comparison of Three Highway Live Source Dispersion Models."
Atmospheric Environment Vol. 12, p. 1323, 1978.
44. Darling, E. M., Jr., Prerau, D. S., Downey, P. J., and Mengert, P. H.,
Highway Air Pollution Dispersion Modeling: Preliminary Evaluation of
Thirteen Models. Final Report prepared by U.S. Department of Transpor-
tation, Cambridge, MA, DOT-TSC-OST-77-33; PB 271-049, U.S. Department
of Commerce, June 1977.
45. Richardson, B. C., et al, "An Inventory of Selected Mathematical Models
Relating to the Motor Vehicle Transportation System," Highway Safety
Research Institute, University of Michigan, Ann Arbor, MI. Report UM-
HSRI-78-28. September 1978.
46. Rao, S. T., et al, "An Evaluation of Some Commonly Used Highway Dispersion
Models." APCA J. Vol. 30, No. 3, March 1980.
47. Chock, D., General Motors Research Laboratories, Warren, MI. Telephone
conversation with M. N. Ingalls, Southwest Research Institute, San Antonio,
TX, February 7, 1979.
117
-------
REFERENCES (Cont'd)
48. Love, D. D. and Stuckel, J. J., "Dispersion of Pollutants in Automobile
Wakes." J. Environmental Engineering Division, Proc. Am. Soc. Civic
Eng. V102 EE3, p. 571, June 1976.
49. Hall, D. J., Simmonds, A. C., and Carrol, J. D., "Experiments on the
Dispersal of Exhaust Material for Diesel Vehicles." Warren Spring
Laboratory Publication LR197 (AP). Warren Spring Laboratory, Gunnels
Wood Rd., Stevenage, Herts, England, 1974.
50. Springer, K. J., "AN Investigation of Diesel-Powered Vehicle Emissions,
Part V." Final Report to the Environmental Protection Agency under
Contract PH 22-68-23, April 1974.
51. Peterson, W. B., "User's Guide for PAL, A Gaussian-Plume Algorithm
for Point, Area, and Line Sources." U.S. Environmental Protection Agency,
Research Triangle Park, NC Publication No. EPA-600/4-78-013, 1978.
52. Debury, G. E., editor The 1973 World Almanac. Newspaper Enterprise
Association, New York, New York. 1972.
53. U.S. Bureau of the Census. Statistical Abstract of The United States;
1973 (94th edition). U.S. Government Printing Office, Washington, D.C.
1979.
54. U.S. Bureau of the Census. Statistical Abstract of the United States;
1979 (100th edition), U.S. Government Printing Office, Washington, D.C.
1979.
55. Telephone conversation, T. Ullman with homebuilder, Mr. Douglas Cross,
May 16, 1980.
56. "Parking Principles", Special Report 125, Highway Research Board, NRC,
Washington, D.C. 1971.
57. Robertson, W. E., "National Parking Facility Study" NLC Research Report.
National League of Cities, Washington, D.C. 1972.
58. Martin, N. D., "Report of the Executive Vice President-Garage Construction
1972." Parking magazine, July 1973. National Parking Assoc., Washington,
D.C.
59. Ivey, D. L., "Report from the Capitol" Parking magazine, April 1975.
60. Martin, N.D., "Report from the Executive Vice President-Parking Construction
in 1979, Annual Report." Parking magazine, April 1980. National Parking
Assoc., Washington, D.C.
61. Weant, R. A., "Parking Garage Planning and Operation," ENO Foundation for
Transportation, Inc., Westport, CT, 1978.
118
-------
REFERENCES (Cont'd)
62. Forrest, L., "Assessment of Environmental Impacts of Light Duty Vehicle
Dieselization," Aerospace Report No. ATR-79(7740)-1. Draft Final Report
by Aerospace Corporation under Contract DOT-TSC-1530, March 1979.
63. Kama, G. M., et al. "Air Flow Requirement for Underground Parking Garages."
American Industrial Hygiene Journal, December 1961, pgs. 462-470.
64. Lewis, S. W., "Carbon Monoxide (CO) Emergency Procedures." Parking
magazine, October 1973. National Parking Assoc., Washington, D.C.
65. "Highway Tunnel Operations", National Corporative Highway Research
Program Report No. 31. Transportation Research Board, NRC, Washington,
D.C., 1975.
66. "Status of Toll Facilities in the United Stated", FHWA Bulletin dated
June 28, 1976. Federal Highway Administration, Washington, D.C.
67. Gerhardt, B., "Tunnel Ventilation Problems", presented at the Symposium
on Air Quality Predictions Related to Transportation" Los Angeles,
California, September 20, 1973.
68. Donnelly D., Colorado Division on Highways, Denver, Colorado. Telephone
conversation with Melvin Ingalls, Southwest Research Institute, San
Antonio, Texas, July 28, 1980.
69. "MVMA Motor Vehicle Facts and Figures '77," Motor Vehicle Manufacturers
Association of the United States, Inc., Detroit, Michigan.
70. "MVMA Motor Vehicles Facts and Figures '80." Motor Vehicle Manufacturers
Association of the United States, Inc. Detroit, Michigan.
71. "Highway Statistics, 1978." Federal Highway Administration. GPO Stock
No. 050-001-00157-8.
72. "National Functional System Mileage and Travel Summary." Procedural
Development Branch and Special Studies Branch, Program Management
Division, Federal Highway Administration. June 1977.
73. Letter from Mr. David R. McElhaney, Chief Highway Statistics Division,.
Federal Highway Administration, Washington, D.C. to Melvin Ingalls,
Southwest Research Institute, San Antonio, Texas, dated October 17, 1980.
74. "Quick-Response Urban Travel Estimation Techniques and Transfer Parameterd,
Users Guide." National Cooperative Highway Research Program Report 187.
Transportation Research Board. National Research Council. 1978.
75. "Summary of Hourly Observations" from the Decennial Census of United
States Climate. Climatography of the United States No. 82- (number
varies with city). For various cities. National Climatic Center,
National Oceanic and Atmospheric Administration, Asheville, N.C.
119
-------
REFERENCES (Cont'd)
76. "Airport Climatological Summary." Climatography of the United Stated
No. 90. For various airports. National Climatic Center, National
Oceanic and Atmospheric Administration, Asheville, N.C.
77. Milke, A., U.S. Bureau of the Census. Telephone conversation with
M.N. Ingalls, Southwest Research Institute, November 17, 1980.
78. 1972 Census of Retail Trade. Vol. Ill-Major Retail Centers. Bureau of
the Census, Washington, D.C.
79. "Traffic Counts 1979", City of Los Angeles Department of Transportation,
Los Angeles, California.
80. "Traffic Volume Summary as of 03-23-79". City of Houston, Traffic and
Transportation Department, Houston, Texas.
81. Norbeck, J. M., et al., "Effects of New York City Toxic Strike on CO
Concentrations in Midtown Manhattan. APCA Journal, Vol. 29, No. 8,
August 1979.
82. Transportation and Traffic Engineering Handbook. Chapter 5, Urban Travel
Characteristics. Institute of Traffic Engineers, 1976. Published by
Prentice-Hall, Englewood Cliffs, N.J.
83. Smith, L. R., "Characterization of Emissions from Motor Vehicles Designed
for Low NOX Emissions." Final Report for EPA Contract No. 68-02-2497
with Southwest Research Institute, San Antonio, Texas. Environmental
Protection Agency, Research Triangle Park, N.C.
84. "ASHRAE Handbook of Fundamentals" published by the American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Inc., NY, NY.
85. Benson, Ferris, et al., Indoor-Outdoor Relationships: A Literature
Review. EPA AP-112, August 1972.
86. Endo, Hiado of Hayakawa and Associates, Los Angeles, California. Private
communication with Melvin Ingalls of Southwest Research Institute, San
Antonio, Texas. December 11, 1980.
87. Chandler, P. L. , Minnesota Dept, of Transportation, District 5, letter to
Melvin Ingalls, Southwest Research Institute, San Antonio, Texas, dated
8/28/80.
88. "Corporation Traffic Statistics, Annual 1979-1989." International Bridge,
Tunnel and Turnpike Assoc., Washington, D.C.
120
-------
REFERENCES (Cont'd)
89. Rodgers, S. J., et al., "Tunnel Ventilation and Air Pollution Treatment",
Prepared for FHWA, Office of Research by Mine Safety Appliance Research
Corp. Report FHWA-RD-72-15, June 30, 1970. NTIS No. PB210-360.
90. Turner, D. B., "Workbook of Atmospheric Dispersion Estimates", U.S.
Public Health Service Publication 999-AP-26.
91. Sievers, H. E. "Modified line source models for predicting 24-hour and
annual particulate concentrations resulting from reentrainment of roadway
dust." APCA Paper No. 78-14.6. Houston, June 1978.
92. Chaney, R. W., "Carbon Monoxide Automobile Emissions Measured from the
Interior of a Traveling Automobile." Science, Vol. 199, 17 March 1978.
93. Colwill, D. M. and Hickman, A. J., "Exposure of Drivers to Carbon Monoxide",
JAPCA, Vol. 30, No. 12. December 1980.
94. Patterson, R. M. and F. A. Record. Monotoring and Analysis of Carbon
Monoxide and Traffic Characteristics at Oakbrook. Final Report prepared
by GCA Corporation, Bedford, MA, under Contract No. 68-02-1376 to the
U.S. Environmental Protection Agency. Research Triangle Park, N.C.
Publication No. EPA-450/3-74-058, November 1974.
95. Bach, W. D., B. W. Crissman, C. E. Decker, J. W. Minear, P. P. Rasberry,
and J. B. Tommerdahl. Carbon Monoxide Measurements in the Vicinity of
Sports Stadiums. Final Report prepared by Research Triangle Institute,
Research Triangle Park, N.C., under Contract No. 68-02-1096 Task No. 1
to the U.S. Environmental Protection Agency, Monitoring and Data Analysis
Division, Research Triangle Park, N.C. Publication No. EPA-450/3-74-049.
July 1973.
96. Patterson, R. M. , R. M. Bradway, G. A. Gordon, R. G. Orner, R. W. Cass,
and F. A. Record. Validation Study of an Approach for Evaluating the
Impact of a Shopping Center on Ambient Carbon Monoxide Concentrations.
Final Report prepared by GCA Corporation, Bedford, Mass., under Contract
No. 68-02-1376 to the U.S. Environmental Protection Agency, Research
Triangle Park, N.C. Publication No. EPA-450/3-74-059, August 1974.
97. McWaters, J. T. and Pelton, D. J. " Compilation of Indirect Source
Monitoring Studies." Report No. EPA-450/3-74-043, April 1974.
98. Mobile Source Emissions Factors. Final Document. Prepared by Office
of Air and Waste Management, Environmental Protection Agency, Washington,
D.C. Publication No. EPA-400/9-78-005. March 1978.
121
-------
APPENDIX A
VERIFICATION OF ONEX COMPUTER PROGRAM
-------
VERIFICATION OF ONEX COMPUTER PROGRAM
A study was conducted to check the ability of the ONEX model to predict
actual on-the-road concentrations. Data from the G.M. Sulfate Experiment-
were used for this study.(A1) In-car sulfate measurements were taken in
six cars on 15 days of the experiment. The reader is referred to Reference
Al for a complete description of the G.M. Sulfate Experiment. The cars were
not always the same six vehicles nor were they always in the same position
in the pack of cars. These measured in-car sulfate concentrations were
compared with concentrations calculated using the ONEX computer program.
The in-car sulfate concentrations can be greatly influenced by the cars
immediately ahead of the receptor car. Therefore, an attempt was made to
estimate the sulfate emissions on a car by car basis. To accomplish this,
sulfate emissions measured from chassis dynamometer tests run during the
experiment, were used to obtain estimates of the sulfate emissions for each
make and model of test vehicle. A list of the make and model of each vehicle
in the G.M. experiment by pack number and location in the pack was obtained
from Dr. S. Cadle at General Motors Research. The dynamometer sulfate
emissions by make and model were then used for the point source portion of
ONEX. The measured dynamometer sulfate emissions were considered as average
for the vehicle model. The dynamometer emission rate was adjusted for each
day, using the ratio of the daily fleet emission rate to the average fleet
emission rate for all days as calculated from measured ambient sulfates.
The emission rate for line source portion of the program was calculated
from the average fleet emission rate and the number of vehicles and vehicle
speed as given in Reference Al. The meteorological conditions used were the
average of those measured during each day of testing as presented in Reference
Al. The emission source height was set at one half meter and the receptor
height at 1.2 meters.
Since the sulfate emissions for each car were not precisely known, the
calculated concentrations from each vehicle were averaged for each day. In
addition, the calculated sulfate value for each pack position for which
sulfate measurements were available was averaged for all days. Measured
concentrations were not always available for each pack position used
(position 2, 3, 5, 6, 8 and 10) for each of the 14 days of the experiment
used in this evaluation. Where the measured concentration for a vehicle was
not available, the calculated concentration was not used in the average.
Table Al and A2 presents the results of the ONEX program calculations
together with the measured sulfate values for comparison. Figure Al presents
these comparisons graphically.
A-2
-------
Both daily averages of all positions and position averages for all days
are shown. For 10 of the 14 days of data used, the difference between
calculated and measured sulfate concentrations was less than 20 percent.
The agreement between measured and calculated sulfate levels indicate that
the on-expressway model is adequate for the purposes of this project.
A-3
-------
TABLE Al. COMPARISON OF MEASURED AND CALCULATED IN-CAR SULFATE
CONCENTRATIONS BY TEST DAY
Avg. Wind Avg. Wind
Avg. Meas. Avg. Calc.
291.45
335.75
221.7
250.28
59.25
196.65
61.06
271.8
223.35
45.8
183.83
180.45
199.65
350.88
354.92
TABLE
2.81 | 274
1.81
2.55
1.31
1.51
275
276
279
281
2.56 286
2.23 | 290
2.05
1.41
0.51
2.86
2.83
293
294
Avg. 274-294
295
296
297
2.18 300
2.6 302
1.28 I 303
Avg. 295-303
A2. COMPARISON OF MEASURED AND
CONCENTRATION BY PACK
Avg . Meas..
Position Conc.(yg/m3)
2 2.17
3 2.91
5 3.47
6 3.83
8 6.60
10 6 . 06
2.64
3.88
5.23
4.8
3.0
5.53
2.15
2.67
3.04
3.66
4.28
2.00
3.33
5.25
4.25
3.82
CALCULATED IN-CAR
POSITION
Avg. Calc.
Cone. (yg/m3)
2.52
3.02
3.59
3.87
5.00
4.01
2.8
I
4.25
4.30
4.06
3.52
4.62
3.15
2.98
2.65
3.59
2.96
3.14
3.80
3.18 •
4.63
3.54
SULFATE
A-4
-------
O Daily average all positions
Q Position average all days
23456
Average measured sulfate concentrations, yg/m3
Ul
' Al.. Comparison of measured in-car sulf ate concentrations to concentrations
predicted by ONEX computer program for G,M, sulfate experiment data
-------
REFERENCES FOR APPENDIX A
1. Cadle, S. H., Chock, D. P., Heuss, J« M., and Monson, P. R., "Results
of the General Motors Sulfate Dispersion Experiment." Prepared by
General Motors Research Laboratories, Warren, Michigan. Publication
No. GMR-2107, EV-26, March 1976.
A-6
-------
APPENDIX B
TUNNELS IN THE UNITED STATES
-------
TUNNELS IN THE UNITED STATES
STATE
Ala.
Ariz.
Calif.
NAME AND LOCATION
BANKHEAD. U.S. 90 under Mobile Rivers
Mobile.
WALLACE. 1-10 under Mobile Rivers
FtobTle.
MULE PASS, U.S. 80 between Benson and
and Douglas
QUEEN CREEK. 'u.S. 60, 70 between
Superior and Miami
BROADWAY. Broadway between Hyde
and Mason Sts,; San Francisco
CALDECOTT. Cal. Route 24 at Contra
Costa-Alameda Cnty. line; Oakland
COLLIER. U.S. 199 near Oregon state
line; Del Norte County
POSEY. Cal. Route 260 under Oakland
Estuary; Oakland
WEBSTER ST. Cal. Route 260 under
Oakland Estuary (Companion to Posey)
LAKEWOOD BLVD. Cal. 19. under Long
Beach Airport Runway; Long Beach
SPRING ST. Spring Street under Long
Beach Airport Runways Long Beach
SECOND STREET. Second Street between
Figueroa and Hill Sts.; Los Angeles
SEPULVEDA BLVD. Cal. 1 under L. A.
Internal1!. Airport runway; L. A..
THIRD STREET. Third Street between
Flower and H111 Sts., Los Angeles
WALDO. No. of Golden Gate on U.S. 101
Parallel Bore
ELEPHANT BUTTE. Feather River
Canyon near Pluraas
UNDERPASS. U.S. 40 in Newcastle
UNDERPASS. Sepulveda under
OPERATOR
Ala. Highway
Department
Ala. Highway
Department
City of San
Francisco
Calif. Dlv. of
Highways
Calif. Dlv. of
Highways
Calif. Dlv. of
Highways
Calif. Dlv. of
Highways
City of Long
Beach
City of Long
Beach
City of Los
Angeles
city of LOS
Angeles
City of Los
Angeles
DATE
OPENED
1941
1973
1958
1953
1952
1937
1965
1963
1928
1963
1958
1958
1924
1953
1901
1937
1956
1937
19321
VI 930
LENGTH
PORTAL TO PORTAL
Ft M
3109
3000
1400
1.200
1616
3610
3371
1835
3545
3350
908
1080
1502
1S08
1059
1000
1000
1187
531
655
948 •
914
493
1100
1027
559
1081
1021
' 277
329
458
582
323
to
UJ
1
2
1
1
2
2
1
1
1
1
2
2
1
2
1
1
1
1
1
1
^.
11
2
2
3
3
2
2
2
2
2
2
3
2
4
3
2
3
3
2
2
3
10
ADT . 1
10,000 V
22,000 S
28,500 L
50,000 .
50,000 '
S
24,500 T
24,500 T
S
S
L
54,400 T
L
ADT/LANE
5', 000
5,500
7,125
16,667
12,250
12,250
9,067
Mull hoiland 1n Los Angeles
-------
TUNNELS IN THE UNITED STATES (Cont'd)
STATE NAME AND LOCATION
Calif. FIGUEROA ST. Los Angeles
HAWONA. Yosemlte Nat'l Park
BIG OAK FLAT. Yosemlte Nat'l Park
BROADWAY LOW LEVEL. State highway
24 in Oakland
PRESIDIO PARK. San Francisco
YERBA BUENA. Yerba Buena Island
San Francisco Bay -
Colo. EISENHOWER. 1-70 under continental
Divide; 60 mi. W. of Denver
MESA VERDE. Entrance to Mesa
Verde Nat'l Park
to STAPLETON. 1-70 East of Denver
I under Stapleton Airport Runway
LO
CLEAR CREEK #1. U.S. 6 West of
Golden
CLEAR CREEK #2. U.S. 6 West of
Golden
CLEAR CREEK #3. U.S. 6 West of
Golden
CLEAR CREEK K. U.S. 6 East of
Idaho Springs
HORSESHOE (No Name). 1-70 East of
Glenwbod Springs
IDAHO SPRINGS. 1-70 near
Idaho Springs
Conn. WEST ROCK. Wilbur Cross Pkwy.
(Route 15); New Haven
D. :C. CENTER LEG. 1-395 under the Mall,
Washington.
DUPONT CIRCLE. Connecticut Ave. under
Dupont Circle; Washington
9TH STREET. 9th Street under the Mall;
Washington
12TH STREET. 12th Street under the Mall;
Washington
OPERATOR
Colo. Dept. of
Highway
Colo.. Dept. of
Highway
Colo Dept. of
Highway
Colo. Dept. of
Highways
Colo. Dept. of
Highways
Colo. Dept. of
Hi ghways
Colo. Dept. of
Highways
Colo. Dept. of
Highways
Conn. Dept. of
Transportation
D. C. Dept. of
Transportation
D. C. Dept. of
Transportation
D. C. Dept. of
Transportation
D. C. Dept. of
. Transportation
DATE
OPENED
•v.1936
1933
1940
1937
M940
1936
LENGTH
PORTAL TO PORTAL
Ft M
755
4233
2167
2944
,1300
514
1973 1st,, .8941
19792ndTube
1957
1965
1952
'1952
1952
1952
1965
1961
1949
1973
1950
1971
'1964
1470
757
859
1069
726
588
1044
665 EB
725 WB
1200 366
3400 1036
578 176
1610 491
729 222
in
LU
1
1
1
2
1
1
2
1
2
1
1
1
1
2
2
2
2
2
1
1
SB!
3
2
2
2
4
10
2
2
3
2
2
2
2
2
2
2
4
2
3
3
ACT * ADT/LANE
8,200 S
66,000
4.700 U
4,600 U
4,600 U
2,700 U
5,100
14,900
31,000 L 7,750
58,000 T 7,250
(c)
27,000 L
14,000 S 4,667
22.000 L
-------
TUNNELS IN THE UNITED STATED (Cont'd)
STATE NAME AND LOCATION
D. C. ZOO. Rock Creek Park
Fla. NEW RIVER. U.S..1 under New River;
Fort Lauderdale
Hawaii WILSON. Kallht Valley; Honolulu, Oahu
State road 63
NUUANU. Pali 11, state road 61
Oahu
NUUANU. Pall #2, state road 61
OiFu
La. BELLE CHASSE. State Route 31 under Intra-
coastaTTaterway; Algiers
HARVEY. Bus. U.S. 90 under Intracoastal
Waterway; Harvey
HOUHA. State Route 3040 under Intra-
Cfl coastal Waterway; Houraa
I
*> Md. BALTIMORE HARBOR. Under Patapsco
River; Baltimore
Mass. BEWEY SQUARE. John Fitzgerald
Expwy. under Dewey Sq.; Boston
CALLAHAN. U.S. 1 under Boston Inner
Harbor; Boston
SUMNER. U.S. 1 under Boston Inner
Harbor (Companion to Callahan}
PRUDENTIAL CENTER. Mass. Turnpike
under Prudential Center; Boston
Mich. DETROIT-WINDSOR. Under Detroit
River betw. U.S. and Canada
Minn. LOWRY HILL. 1-94 under Lyndale and
Hennepin Aves.; Minneapolis
Nev. CARLIN. 1-80
N. J. G. W. BRIDGE APPROACHES. East and West
tunnels on approaches to G. N. Bridge
OPERATOR
Fla. Dept. of
Transportation
City & County
of Honolulu
La. Dept. of
Highways
La. Dept. of
Highways
La. Dept. of
Highways
Md. Dept. of
Transportation
Mass. Dept. of
Public Works
Mass. Turnpike
Authority
Mass. Turnpike
Authority
.Mass, Turnpike
Authority
Detroit-Canada
Corporation
Minn. Dept. of
Highways
Port Authority of
N.Y. & N.J.
DATE
OPENED
1964
1960
1960
1957
1957
1956
1957
1961
1957
1958
1962
1934
1964
1930
1971
1962
LENGTH
PORTAL TO PORTAL
Ft M .
800
800
2780
1080
500
800
1080
960
7650
2400
5070
5657
1980
5130
1496
1993
631
547
244
847
244
329
293
2332
732
1545
1724
604
1564
456
192
167
UJ
QQ
1
2
2
1
1
1
2
1
2
2
1
1
2
1
2
1
£w
UJ UJ
Z CQ
3£
2
2
2
2
2
2
2
2
2
3
2
2
4
2
3
2
ADT > ADT/LANE
44,000 L
35,000 V 8,750
S
S
L
65,500 T 16,375
125,000 L 20,833
65,000 V 32,500
65,000 T 32,500
L
T
32,000 S 5,333
(c)
V
-------
TUNNELS IN THE UNITED STATES (Cont'd)
00
Ul
STATE
N. V.
N. C.
Ohio
NAME AND LOCATION
LINCOLN. Under Hudson Rvr. betw.
N.Y.C. & Ueehawken. N.J.
BATTERY PARK. Under Battery Park In
Manhattan; New York City
\
FIRST AVENUE. 1st Avenue under U.N.
Plaza; New York City
HUGH GRANT CIRCLE. 1-95 under Hugh Grant
Grant Circle, Bronx; N.Y.C.
PARK AVENUE. Park Avenue between
33rd & 40th Streets; N.Y.C.
BROOKLYN-BATTERY. Under East River
betw. Manhattan & Bklyn.; N.Y.C.
QUEENS-MIDTOWN. Under East Rtver
betw. Manhattan & Queens; N.Y.C.
HOLLAND. Under Hudson Rvr. betw.
NT?7C. & Jersey City, N.O.
CROSS TOWN. 178 St. N.Y.C,
F.D. ROOSEVELT DR. 81-89 St. N.Y.C.
F.D: ROOSEVELT DR. 42-48 St. N.Y.C.
BEAUCATCHER. Through Beucatcher
mountain, Ashevl lie
TUNNEL #1 (Eastboiind). 1-40 along
Pigeon River
TUNNEL j>l (Westbound). Companion
to Tunnel #1 (eastbound )
TUNNEL '«. 1-40 along Pigeon
LYTLE PARK. 1-71 under Lytle Park;
Cincinnati
OPERATOR
Port Authority of
N.Y. S N.J.
New York C11y
New York CUy
New York City
New York City
Tr1 borough Bridge
& Tunnel Authority
Trlborough Bridge
& Tunnel Authority
Port Authority of
N.Y. & N.J.
City of
Cincinnati
OPENED
1937
1945
1957
1951
1953
1955
(b)
1950
1940
1940
1927
1927
1929
1969
1969
1969
1970
PORTAL
Ft
8216
7482
8006
2400
1378
700
1392
9117
6272
6414
8558
8371
2414
2400
1600
1100
1043
1119
1198
850
TO PORTAL
M
2504 -
2281
2440
732
420
213
424
2779
1912
1955
2608
2551
259
M
Ul
CO
1
1
1
2
2
2
1
2
1
1
1
1
2
2
1
1
1
1
2
i
S?
Ul LU
2= CO
3£
2
2
2
2
2
3
2
2
2
2
2
2
3
3
2
2
2
2
3
i
(0
ADT > ADT/LANE
T 15,833
S
S
40,000 T 10,000
30,000 T 15 nnn
30,000 T 15>UUO
30,000 , 15 000
30,000 T lb'uuu
L
-------
TUNNELS IN THE UNITED STATES (Cont'd)
STATE NAME AND LOCATION
Ore. KNOWLS CREEK. Love Co.
ARCH CAPE. U.S. 101, Oregon
Coast Highway
ELK CREEK. State Highway 58
Underpass
VISTA RIDGE. U.S. 26 In Portland
Pa. FORT PITT. 1-79 under Mount
Washington; Pittsburgh
LIBERTY. Under Mt. Washington betw.
W. Liberty Ave. & Liberty Bridge;
Pittsburgh
SQUIRREL HILL. 1-376 under Squirrel
Hill; Pittsburgh
ALLEGHENY. Pa. Turnpike under
w Allegheny Mt.; Somerset Co.
0% BLUE MOUNTAIN. Pa. Turnpike under
Blue Mt. ; Franklin Co.
LEHIGH. Pa. Turnpike. N.E. extension;
Lehigh Co.
KITTATINNY. Pa. Turnpike under
Kittatinny Mt.; Franklin Co.
TUSCARORA. Pa. Turnpike under
•Tuscarora Mt. ; Frriaklln Co.
EVANS. Pa. Turnpike
ARMSTRONG HILL. Pittsburgh
R. I. CONVERTED STREETCAR. Providence
Tenn. HcCOLLIE AVE. Through Missionary
Ridge 1n Chattanoga
BACHHAN. U.S. 41 In Chattanoga
WILCOX. Chattanoga
OPERATOR
Pa. Dept. of
Transportation
Pa. Dept. of
Transportation
Pa. Dept. of
Transportation
Pa. Turnpike
Commission
Pa. Turnpike
Commission
Pa. Turnpike
Coimrisslon
Pa. Turnpike
Commission
Pa. Turnpike
Commission
DATE
OPENED
1937
1932
1970
1960
1924
1953
1940
1965
1940
1968
1957
1940
1968
1940
1968
1957
Unk.
1955
1913
1955
1929
Unk.
PORTAL
Ft
1430
1228
1085
1100
3600
5690
4225
6070
6070
4339
4339
4380
4727
4727
5326
5326
4379
1325
1793
933
1001
1027
1312
TO PORTAL
M
1097
1734
1288
1850
1850
1323
1323
1335
1441
1441
1623
1623
£
1
1
2
2
2
2
2
2
1
2 '
2
1
2
1
1
1
2
1
Ie
2
2
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
£
ADT > ADT/LANE
78,000 S 19,500
50,000 1 12,500
83,000 S 20,750
18,800 S 4,700
11,100 S 2,775
12,000 6,000
11,100 S 2,775
11,200 S 2,800
-------
TUNNELS IN THE UNITED STATES (Cont'd)
STATE NAME AND LOCATION
Tex. BAYTOWN-LA PORTE. SH 146 under Houston
ship Channel betw. Bay town & LaPorte
WASHBURN. Federal Rd. under Houston Ship
Channel betw. Pasadena and Galena Park
AIRPORT UNDERPASS. Dallas, Ft. Worth
Ai rport ~
Utah COPPERFIELD. Blnghara Canyon,
Southwest of Salt Lake City
ZION NAT'L PARK. State Road 15 1n.
Zion Nat'l Park
Va. BIG WALKER. 1-77 under Big Walker Mt.;
near Wytheville
HAMPTON ROADS. 1-64 under Hampton Roads
Harbor betw. Norfolk & Hampton
DOWNTOWN. U.S. 460A under Elizabeth
Rvr. betw. Norfolk & Portsmouth
W MIDTOWN. U.S. 58 under Elizabeth Rvr.
i betw. Norfold & Portsmouth
BALTIMORE 'CHANNEL. Under north ship
channel ; Chesapeake Bay
THIMBLE SHOAL CHANNEL. Under south
ship channel; Chesapeake Bay
EAST RIVER MOUNTAIN. 1-77 under East
River Mt. on Va.-W. Va. border
MARY'S ROCK. Mile 32 of Skyline
Drive, Shenandoah Nat'l Park
Wash. BATTERY ST. Beneath Battery Street
betw. 1st & 7th Aves.; Seattle
MT. BAKER RIDGE. U.S. 10 In Seattle
W. Va. WHEELING. 1-70; Wheeling
MEMORIAL. West Virginia. Turnpike;
Kanawha County
BLAND CO. 1-77
Wyo. GREEN RIVER. 1-80
OPERATOR
Texas Dept. of
Highways
City of Houston
Va. Dept. of
Highways & Transp.
Va. Dept. of
Highways & Transp.
Va. Dept. of
Highways & Transp.
Va. Dept. of
Highways & Transp.
Chesapeake Bay
Brdge. & Tun. Dist.
Chesapeake Bay
Brdge. & Tun. Dist.
Va. Dept. of
Highways & Transp.
Wash. Dept. of
Highways
W. Va. Dept. of
Highways
W. Va. Turnpike
Coimission
DATE
OPENED
1953
1950
1963
1939
1936
1972
1957
1952
1962
1964
1964
1974
1954
1940
1966
1954
1966
PORTAL
Ft
3009
2936
813
7000
5766
4230
7479
3350
4194
5450
5738
5400
660
'2140
1466
1490
2669
5661
1135
TO PORTAL
M
917
895
1289
2280
1021
1278
1661
1749
1646
652
454
814
3
1
1
2
1
1
2
1
1
1
1
1
2
1
2
2
2
1
2
i§
2
2
3
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
ADT > ADT/LANE
20,000 S 10, COO
27,000 S 13,500
1 ,600 T 400
(c)
22,400 T 11,200
25,000 S 12,500
16,000 T 8,000
4,000 T 2,000
4,000 T 2,000
T
U
23,500 T 5,875
29,000 L 7,250
T
-------
APPENDIX C
BIBLIOGRAPHY
-------
BIBLIOGRAPHY
Air Quality Display Model. NTIS Publication No. PB-189 194, November 1969.
Aron, Robert H.; Aron I-Ming. Statistical Forecasting Models: Carbon
Monoxide and Oxidant Concentrations in the Los Angeles Basin. (Central
Michigan UnivJ, and (SouthCoast Air Quality Management District, Calif),
APCA J. Jul 78, V28, N7, P681.
Bachf W. D., B. W. Crissman, C. E. Decker, J. W. Minear, P. P. Rasberry, and
J. B. Tommerdahl. Carbon Monoxide Measurements in the Vicinity of Sports
Stadiums. Final Report prepared by Research Triangle Institute, Research
Triangle Park, NC, under Contract No. 68-02-1096 Task No. 1 to the U.S.
Environmental Protection Agency, Monitoring and Data Analysis Division,
Research Triangle Park, NC. Publication No. EPA-450/3-74-049. July 1973.
Benson, Ferris. Indoor-Outdoor Relationships: A Literature Review. EPA
AP-112, August 1972.
Briggs, T. M., et al. Air Pollution Considerations In Residential Planning,
Volume I: Manual, Prepared for Department of Housing and Urban Devel-
opment and EPA, EPA-450/3-74-046-a, July 1974.
Bullin, J. A. and J. C. Polasek. Analytical and Experimental Assessment of
Highway Impact on Air Quality. Prepared by Texas ASM University
for Texas Department of Highway and Public Transportation, Report No.
TTI-2-8-75-218-1, August 1976.
Cadle, S. H., D. P. Chock, J* A. Heuss, and P. R. Monson. Results of the
General Motors Sulfate.Dispersion Experiment. Prepared by General Motors
Research Laboratories, Warren, Michigan. Publication No. GMR-2107,
EV-26o March 1976.
Camann, David, et al. A Model for Predicting Pathogen Concentrations in
Wastewater Aerosols, March 9, 1978.
Chock, D. P. General Motors Sulfate Dispersion Experiment: An Overview of
the Wind, Temperature, and Concentration Fields. Prepared by General
Motors Research Laboratories, Warren, MI. Publication No. GMR-2231,
ENV-30. June 1977.
C-2
-------
.Chock, D. P. The General Motors Sulfate Dispersion Experiment: Assessment
of the EPA Hiway Model. Prepared by General Motors Research Laboratories,
Warren, MI, to the EPA Symposium of GM Sulfate Dispersion Experiment.
Research Triangle Park, N.C. Publication No. GMR-2126, EV-27. April 1976.
Chu, K. J., and J. H. Seinfeld. Formulation and initial application of a
dynamic model for urban aerosols. Atmos. Environ, 9_: 375-402, 1975.
Clarke, John. A Simple Diffusion Model for Calculating Point^Concentrations
from Multiple Sources, J. Air Poll. Cpnt. Assn., V14, N9, September 1964.
Cole, Henry S. and John E. Summerhays. A Review of Techniques Available for
Estimating Short-Term N02 Concentrations, J. Air. Poll. Cont. Assn.,
Vol. 29, No. 8, August 1979.
De Mandel, R. E., Lewis H. Robinson, James S. Fong, and Ronald Y. Wada.
Comparison of EPA Rollback, Empirical/Kinetic, and Physicochemical
Oxidant Prediction Relationships in the San Francisco Bay Area. Bay
Area Air Quality Management District, Calif. APCA J. V29, N4, P352,
April 1979.
The Diesel Emissions Research Program, Environmental Protection Agency.
Dispersion Model by Stephen Budiansky, Environmental Science and Technology,
Vol. 14, No. 4, April 1980.
Dolan, D. F. and D. B. Kittelson. Roadway Measurements of Diesel Exhaust
Aerosols, SAE Paper No. 790492, February 1979.
Doty, S. R, Climatological Aids in Determining Air Pollution Potential —
Where We are Today. National Climatic Center, Asheville, NC.
October 1978.
Doty, S. R., and B. L. Wallace. A Climatological Analysis of Pasquill
Stability Categories Based on 'Star' Summaries. Prepared for U.S.
Environmental Protection Agency, Environmental Sciences Research
Laboratory. Research Triangle Park, NC, April 1976.
Downay, Paul J., Jeffrey D. Garlitz, and Kevin B. Murphy. A comparison of
Six Highway Air Pollution Dispersion Models Using Synthetic Data.
Transportation Systems Center, Cambridge, Mass. Final Report November
1975-March 1976. 192 p, September 1977.
Forrest, L. Assessment of Environmental Impacts of Light-Duty Vehicle
Dieselization. Interim Report Task 1 prepared by Aerospace Corporation,
Los Angeles, CA, under Contract No. DOT-TSC-1530 to the Department of
Transportation, Cambridge, Mass. November 1978.
Freedman, Sandor J. Assessment of Human Exposure to Mobile Source Pollutants,
Environmental Protection Agency, Contract 68-02-2459, September 1977.
C-3
-------
Green, Nicholas, et al. Dispersion of Carbon Monoxide from Roadways at Low
Wind Speeds. J. Air Poll. Cont. Assn., Vol. 29, No. 10, October 1979.
Guzewich, David and William Pringle. Validation of the EPA-PRMTP Short-Term
Gaussian Dispersion Model. J. Air Poll. Cont. Assn., Vol. 29, No. 6,
June 1977.
Hall, D. J., A. C. Simmons, and J. D. Carrol. Experiments on the Dispersal
of Exhaust Material from Diesel Vehicles. Warren Spring Laboratory
Publication LR187(AP). 1974.
Halpern, Marc. Indoor/Outdoor Air Pollution Exposure Continuity Relation-
ships. J. Air Poll. Cont., Assn., Vol. 28, No. 7, July 1978.
Hameed, S. Modeling Urban Air Pollution. Atmospheric Environment, Vol. 8,
1974.
Hameed, S. A Modified Multi-Cell Method for Simulation of Atmospheric
Transport. March 1974.
Hanna, Steven. Diurnal Variation of the Stability Factor in the Simple
ATDL Urban Dispersion Model. J. Air Poll. Cont. Assn., Vol 29, No. 2,
February 1978.
Hanna, Steven. A Simple Dispersion Model for the Analysis of Chemically
Reactive Pollutants. February 1973.
Hanna, Steven. A Simple Method of Calculating Dispersion from Urban Area
Sources. J. Air Poll. lont. Assn., Vol. 21, No. 12, December 1971.
Hare/ Charles T. Methodology for Estimating Emissions from Off-Highway
Mobile Sources for the RAPS Program. EPA-450/3-75-002, October 1974.
Henderson, John J. et al. Indoor-Outdoor Air Pollution Relationships,
Volume II, An Annotated Bibliography, EPA AP-112b, August 1973.
Holzworth, G. C. Mixing Heights, Wind Speeds, and Potential for Urban Air
Pollution Throughout the Contiguous United States. Prepared by Office
of Air Programs, U.S. Environmental Protection Agency. Research Triangle
Park, N.Co January 1972.
Hoult, D. P., et al. Turbulent Plume in a Turbulent Cross Flow: Comparison
of Wind Tunnel Tests with Field Observations. J. Air Poll. Cont. Assn.,
Vol. 27, No. 1, January 1977.
Jensen, N. O. and E. L. Petersen. The Box Model and the Acoustic Sounder,
A Case Study. December 1978.
C-4
-------
Keenan, M. T., G. Sistla, A. R. Peddada, P. J. Samson, and S. T. Rao. Sulfate
and lead concentrations adjacent to the Long Island Expressway near
Huntington, N.Y. Presented at the APCA meeting on "Question of Sulfates",
Philadelphia, April 1978.
Kircher, D. S. and D. P. Armstrong. An Interim Report on Motor Vehicle
Emission Estimation. Prepared by Office of Air and Water Programs,
U.S. Environmental Protection Agency. Research Triangle Park, NC.
Publication No. EPA-450/2-73-003. October 1973.
Kuethe, Arnold. Investigation of the Turbulent Mixing Regions Formed by
Jets. J. App. Mech.
Kummler, R. H., B. Cho. G. Roginski, R. Sinha, and A. Greenberg. A Comparative
Validation of the RAM. and Modified Sia Models for Short-Term SO2 Concen-
trations in Detroit. Wayne State University. APCA J. V29, N7, P720,
July 1979.
Kunselman, P., H. T. McAdams, C. J. Domke, and M. E. Williams. Automobile
Exhaust Emission Modal Analysis. Prepared by Calspan Corporation,
Buffalo, N.Y., under Contract No. 68-01-0435 to the U.S. Environmental
Protection Agency, Ann Arbor, Michigan. Publication No. EPA-460/3-74-005.
January 1974.
Laird, A. Rachel and R. W. Miksad. An Application of a Pseudo-Second Order
SO2 Reaction Algorithm to Urban Air Pollution Modeling. J. Air Poll.
Cont. Assn., Vol. 29, No. 2, February. 1979.
Lane, Dennis and James Stoukel. Dispersion of Pollutants in Automobile
Wakes. J. Environmental Engr. Div., June 1976.
Larsen, Ralph. A Mathematical Model for Relating Air Quality Measurements
to Air Quality Standards, EPA Publication No. AP-89, November 1971.
Larsen, R. L. A New Mathematical Model of Air Pollutant Concentration
Averaging Time and Frequency. NAPCA, Bureau of Criteria and Standards,
Durham, N.C. Air Pollution Control Association. Journal, 19,
January 1969.
Lawson, T. J. and S. Uk. The Influence of Wind Turbulence, Crop Characteristics
and Flying Height on the Dispersal of Aerial Sprays, December 1978.
Mage, D. T. and W. R. Ott. Refinements, of the Lognormal Probability Model for
Analysis of Aerometric Data. EPA, Population Studies Div., Exposure
Assessment Branch, Health Effects Research Lab., Research Triangle Park,
NC 27711. Air Pollution Control Association, Journal 28, August 1978.
Maxwell, Christine, et al. Development of Lead Emission Factor for Reentrained
Dust from Paved Roadways, 71st Annual Meeting, Air Poll. Cont. Assn.,
June 1978.
C-5
-------
Maybank, John, et al. Spray Drift from Agricultural Pesticide Applications,
J. Air Poll. Cont. Assn., Vol. 28, No. 10, October 1978.
McElroy, James L. and Francis Pooler. St. Louis Dispersion Study, Volume II -
Analysis, National Air Poll. Control Admin. Publication No. AP-53,
December 1968.
McMahon, T. A. and P. J. Denison. Empirical Atmospheric Deposition Parameters
- A Survey, Atmospheric Environment, Vol. 13, pp 571-585.
Miller, C. W. An Application of The ATDL Simple Dispersion Model. Oak
Ridge National Lab., Health and Safety Research Div., P.O. Box X,
Oak Ridge, TN 37830. Air Pollution Control Association, Journal 28.
August 1978.
Miller, Frederick, et al. Size Considerations for Establishing a Standard
for Inhalable Particles. J. Air Poll. Cont. Assn., Vol. 29, No. 6,
June 1979.
Moe, R. D., J. A. Bullin, and J. C. Polasek. Effects of Stability Episodes
on Air Pollutant Levels Along Roadways. APCA Paper No. 78-14.1.
Nagda, N. L.f J. R. Ward, B. J. Mason, et al. Retrospective Modeling of
Ambient Concentrations of Benzo(a)pyrene. Geomet, Inc., 15 Firstfield
Rd., Gaithersburg, MD 20760. 71st APCA annual meeting and exhibition
Houston, Texas, June 25-30, 1978.
Nunge, Richard J. Application of an Analytical Solution for Unsteady,
Advective Diffusion to Dispersion in the Atmosphere-I-II, November 1973.
Pasquill, Frank. Atmospheric Dispersion Modeling, J. Air Poll Cont. Assn.,
Vol. 29, No. 2, February 1979.
Patterson, R. M. and F. A. Record. Monitoring and Analysis of Carbon Monoxide
and Traffic Characteristics at Oakbrook. Final Report prepared by GCA
Corporation, Bedford, MA, under Contract No. 68-02-1376 to the U.S.
Environmental Protection Agency. Research Triangle Park, N.C. Publication
No. EPA-450/3-74-058, November 1974.
Patterson, R. M., R. M. Bradway, G. A. Gordon, R. G* Orner, R. W. Cass, and
F. A. Record. Validation Study of an Approach for Evaluating the Impact
of a Shopping Center on Ambient Carbon Monoxide Concentrations. Final
Report prepared by GCA Corporation, Bedford, Mass., under Contract No.
68-02-1376 to the U.S. Environmental Protection Agency, Research Triangle
Park, N.C. Publication No. EPA-450-3-74-059, August 1974.
Petersen, W. B.p and S. T. Rao. The Long Island Expressway Dispersion
Experiment. Presented at the American Meteorological Society's Con-
ference on Turbulence and Diffusion, Reno, January 1979.
06
-------
Pollack, R. I., and T. Austin. Two Methods for Calculating Light-Duty Vehicle
Emission Factors. Draft report under FHWA Contract No. FH-11-9143.
.September 1978.
Ragland, Kenneth W. Multiple Box Model for Dispersion of Air Pollutants from
Area Sources, May 1973.
Rao, S. T., M. Chen, M. Kennan, G. Sistla, R. Peddada, G. Wotzak, and N. Kolak.
Dispersion of Pollutants near Highways, Experimental Design and Data
Acquisition Procedures. Prepared under Grant No. R-803881-01 by New
York State Department of Environmental Conservation, Albany, N.Y. Interim
Report. U.S. Environmental Protection Agency, Research Triangle Park,
N.C., 1978.
Rao, S. T. and Michael Keenan. Suggestions for Improvement of the EPA-HIWAY
Model. J. Air Poll. Cont. Assn., Vol..30, No. 3, March 1980.
Risk Analysis Offers New Engineering Tool. Automative Engineering, July 1978.
Roberts, E. M. Review of Statistics of Extreme Values With Applications to
Air Quality Data, Part I Review. J. Air Poll. Cont. Assn., Vol. 29,
No. 6, June 1979.
Roberts, E. M. Review of Statistics of Extreme Values with Application to
Air Quality Data, Part II, Applications. Dames & Moore, Denver, APCA J.
V29, N7, P733, July 1979.
Roberts, P. J. et al. Contaminant Emissions in the Los Angeles Basin — Their
Sources, Rates and Distribution, Appendix A of Development of a Simulation
Model for Estimating Ground Level Concentrations of Photochemical Pol-
lutants, Report 71SAI-6 for EPA, March 1971.
Sauter, G. D., and W. R. Ott. A Computer Program for Projects of Vehicular
Pollutant Emissions in Urban Areas. J. Air Poll. Cont. Assn., 24(1);
54-59, 1974.
Scheel, J. W. A Method for Estimating and Graphically Comparing the Amounts of
Air Pollution Emissions Attributable to Automobiles, Buses, Commuter
Trains, and Rail Transit. SAE Paper No. 720166. Detroit, January 1972.
Sheih, C. M. A Puff-Grid Model for Predicting Pollutant Transport over an
Urban Area, J. Air Poll. Cont. Assn., Vol. 27, No. 8, August 1977.
Sievers, Henry E. Modified Line Source Models for Predicting 24-hour and
Annual Particulate Concentrations Resulting from Reentrainment of Roadway
Dust. Tex Air Control Board. Austin. Proc. Air. Pollut.Contol.Assoc.,
71st Annu. Meet. Houston, Texas. Published by Air Pollut. Control Assoc.
Pittsburgh, PA 1978 V 1 Pap 14, 6 16 p.
The Simple Box Model Revisited, Short Communication, Atmospheric Environment,
Vol. 12, 1978.
C-7
-------
Sistla, Gopal, et al. A Study of Pollutant Dispersion Near Highways, November
1978.
Smith, D. G., E. A. Taylor, S. M. Doucette, B. A. Egan, Peter S. Daley, and
Dennis F. Naugle. Validation Studies of Air Quality Models at Dulles
Airport: Analysis of Airport Emissions. Env Research & Technology, Mass.
APCA J., V29, N2, P110 February 1979.
Solomon, R. L., et al. Spatial Variation of Airborne Lead Concentration in
an Urban Area. JAPCA, Vol. 27, No. 11, Page 1095, November 1977.
Springer, Karl J. Investigation of Diesel-Powered Vehicle Emissions. Final
Report to EPA, April 1974.
Supplementary Guidelines for Lead Implementation Plans — Revised Section 4.3
(Projecting Automotive Lead Emissions). EPA-450/2-78-038a, OAQPS No. 1.2-
104a, July 1979.
Technical Note - Modeling Urban Air Pollution, Atmospheric Environment, Vol. 7,
pp 131-136, 1973.
Texas Line Source Models. State Department of Highways and Public Transportation.
August 1978.
Thayer, S. D. and J. D. Cook. Vehicle Behavior In and Around Complex Sources
and Related Complex Sources Characteristics, Volume VI - Major Highways.
Final Report prepared by Geomet, Inc., RQckville, MD, under Contract No.
68-02-1094 to the U.S. Environmental Protection Agency. Research Triangle
Park, NC, Publication No. EPA-450/3-74-003f. November 1973.
Tiao, G. C. amd S. C. Hillmer. Statistical Models for Ambient Concentrations
of Carbon Monoxide, Lead, and Sulfate based on the LACS Data. Environ.
Sci. Technol., 12_(7) :820-827, 1978.
Venkatram, A. Examination of Box Models for Air Quality Simulation. Ont
Minist of the Environ. Toronto. Atmos Environ. V12 N12, p 2243-2249, 1978.
Wu , Daniel. Rapid Prediction of Ambient Air Pollutant Concentration at a
Small Isolated Area Source. J. Air Poll. Cont. Assn., Vol. 27, No. 12,
December 1977.
Workbook for Comparison of Air Quality Models. U.S. EPA, Research Triangle
Park, NC. U.S. Environ Prot Agency, OSQPS Tech Rep EPA-450/2-78-028a,
112p, May 1978.
Zib, Pavel. Seasonal Variability of the Simple Urban Dispersion Model. J.
Air Poll. Cont. Assn., Vol. 39, No. 1, January 1980.
C-8
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-460/3-81-021
I. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Estimating Mobile Source Pollutants in Microscale
Exposure Situations
5. REPORT DATE
July 1QR1
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Melvin N. Ingalls
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southwest Research Institute
6220 Culebra Road
San Antonio, Texas 78284
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-03-2884
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Mobile Source Air Pollution Control
2565 Plymouth Road
Ann Arbor, Michigan 48105
13. TYPE OF REPORT AND PERIOD COVERED
Task Final Report 1/80-2/81,
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The goal of this study was to provide a method of estimating the concentration of any
vehicle generated pollutant in areas where people are in close proximity to vehicles
(microscale exposure situations). A list of common exposure situations was ex-
tracted from several hypothetical daily activity routines. These situations were:
residential garages, parking garages, tunnels, street canyons and expressways. For
each of the situations an appropriate dispersion model was selected from the litera-
ture for use in obtaining pollutant concentrations. To determine the exposure level
for each case, the pollutant concentrations were calculated for typical and severe
actual situations. The range of physical variables for each type of situationi.was
obtained from a search of the literature, to define the typical and severe cases for
each situation. Using the physical variables describing the typical and severe cases,
actual locations were chosen to represent the typical and severe exposure for each situ-
ation. Pollutant concentrations were calculated for these actual locations using the
chosen dispersion models. Concentrations were calculated using one gram per mile for
tunnel, street canyon, and expressway situations, and one gram per minute for resi-
dential and parking garages. To use the calculated concentrations with emission
factors other than one gram per mile or one .gram per minute, the concentrations giver
in the report are multiplied by the desired emission factor in the correct units.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATl Field/Group
Air Pollution
Exhaust Emissions
Motor Vehicles
Dispersion Models
Parking Garages.
Tunnels
Street Canyons
Expressways
Exposure Estimates
13. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
182
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
|