EPA-450/3-76-024
July 1976
NATIONAL ASSESSMENT
OF THE URBAN
PARTICIPATE PROBLEM
VOLUME I
NATIONAL ASSESSMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-76-024
NATIONAL ASSESSMENT OF THE
URBAN PARTICIPATE PROBLEM
VOLUME I
SUMMARY OF NATIONAL ASSESSMENT
by
David A. Lynn
Gordon L. Deane
Rebecca C. Galkiewicz
Robert M. Bradway
Frank A. Record, Project Director
GCA/Technology Division
Burlington Road
Bedford, Massachusetts 01730
Contract No. 68-02-1376, Task Order No. 18
EPA Project Officer: Thompson G. Pace
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
July 1976
-------
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 - as supplies permit - from the Air Pollution Technical Informa-
tion Center, Environmental Protection Agency, Research Triangle Park, North
Carolina 27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by GCA/
Technology Division, Bedford, Massachusetts 01730, in fulfillment of Contract
No. 68-02-1376, Task Order No. 18. The contents of this report are reproduced
herein as received from GCA/Technology Division. The opinions, findings, and
conclusions expressed are those of the author and not necessarily those of the
Environmental Protection Agency. Mention of company or product names is not to
be considered as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-76-024
ii
-------
CONTENTS
Page
List of Figures v
List of Tables vii
Acknowledgments viii
Executive Summary ix
Sections
I Introduction to Study 1
Background 1
Study Approach 4
Organization of the Study Report 5
II General Findings of the Study 7
Study City Analysis 7
Summary of Factors Affecting Attainment 27
III Assessment of Factors Affecting Attainment 33
Background and Large-Scale Considerations 35
Particulates From Traditional Somces 52
Particulates From Nontraditional Sources 63
Monitoring Considerations 78
Meteorology and Climatology 93
Relative Contributions of Various Factors 104
References 109
iii
-------
CONTENTS (continued)
Sections Page
IV Summary and Conclusions 110
Attainment Factors Identified in City Studies 110
Control Strategy Options 115
Framework for Control Strategy Prioritization 120
V Recommendations- 124
Recommendations for Emission Control Efforts 125
Recommendations Concerning Air Quality Management
Planning 130
Other Aspects NAAQS Review 133
Appendices
A Comparative Data on the 14 Study Cities 135
B Microscopic Filter Analysis 146
C A Case Study of Long-Range Transport 160
D Assessment of Traditional Source Impact on TSP Standards
Attainment I67
E Assessment of Particulates From Nontraditional Sources 213
F Influence of Meteorology and Climatology on TSP Levels 265
G Preliminary Assessment of Source Contributions 324
IV
-------
LIST OF FIGURES
No. Page
1 Trend in National Average TSP Levels (Average of 1014
Urban Sites) 3
2 Geographical Distribution of 14 Study Cities 8
3 Schematic Relationship Among Five Major Factors 32
4 Average TSP Levels by Neighborhood Type 34
5 Composite Annual Geometric Mean TSP Levels at Nonurban NASN
Sites From 1970 Through 1973 (yg/m3) 49
6 Annual Geometric Mean Sulfate and Nitrate Levels at Nonurban
NASN Sites - 1974 51
7 Average Estimated Contributions to Nonurban Levels in the
East, Midwest, West 53
8 Relationship Between City Wide Average TSP Levels and
Traditional Source Emission Density 56
9 TSP Increments Due to Traditional Sources at Different Site
Types by City Category 64
10 The Range and Average Lead Concentrations Found at Monitor-
ing Sites 72
11 Average and Range of TSP Loadings Due to Tire Wear at
Different Monitoring Site Classifications 74
12 Nontraditional Source Increments at Different Site Types 79
13 Range of Heights in Typical Hi-Vol Installations 86
14 Duration of Rainfall Effectiveness in Reducing TSP Levels
at Two Birmingham Sites 9;>
-------
LIST OF FIGURES (continued)
No. Page
15 Relationship Between TSP Concentrations and Wind Speed at
Three Selected Sites in Birmingham on Days With 48-Hour
Precipitation Amounts <_ 0.02 Inches 98
16 Relative Effect of Annual Precipitation on Annual TSP
Level in a Hypothetical Urban Area 102
17 Summary of Average Impact of Major Contributors to TSP
Levels ' 106
VI
-------
LIST OF TABLES
No^ Page
1 Population and Physical Setting of Case Study Cities 10
2 Study Cities by Their Dispersion and Industrialization
Characteristics 11
3 Summary of Sites Exceeding Air Quality Standards in 1974 12
4 Source and Emissions Characterization of Study Cities 13
5 Composite Summary of Microscopic Analysis in 14 Cities,
yg/m3 25
6 Estimates of Average Filter Loadings by Site
Classification 26
7 Composite Summary of Particle Size by Components 26
8 Estimates of Particles Smaller Than 20 ym Radius Emitted
Into or Formed in the Atmosphere (106 metric tons/year) 38
9 Nonurban Levels in the 14 Study Cities 44
10 Comparison of Urban and Nonurban Levels of Sulfates and
Nitrates in the 14 Study Cities 47
3
11 Approximate Impact (In yg/m ) of Vehicular Traffic on
Nearby Hi-Vol Sites 68
12 General Monitoring Objectives 80
13 Average Annual TSP Concentrations by Neighborhood Type 84
14 Variation in Monitor Height Among the 14 Cities 87
15 Number of Sites With an Estimated Impact of Local Influences,
by Neighborhood Classification 89
16 Control Priorities 123
vii
-------
ACKNOWLEDGMENTS
Numerous persons and organizations have made significant contributions to
this overall study effort, and GCA/Technology Division wishes to sincerely
acknowledge their participation. On-going project supervision has been
received from Thompson G. Pace, Project Officer, of EPA's Control Programs
Development Division. Professional staff members of the EPA Regional
Offices and of the state and local agencies responsible for the case study
cities have been uniformly cooperative and helpful.
Specific thanks are due to Dr. William E. Wilson and Ronald K. Patterson
of the Environmental Sciences Research Laboratory, for sharing the results
of their field research in several cities; to Dr. David S. Shearer and
Dr. Richard J. Thompson of the Environmental Monitoring and Support
Laboratory, for their analytical support; to Gerald Gipson of the Monitoring
and Data Analysis Division for providing modeling results in several cities;
and to David Dunbar, Edward J. Lillis, John D. Bachmann, and the staff of
the Monitoring and Reports Branch for their on-going review and advice.
Vlll
-------
EXECUTIVE SUMMARY
This report presents the results of a study"conducted to assess the national
particulate problem in urban areas, with emphasis on identifying the factors
involved in the attainment or nonattainment of the National Ambient Air
Quality Standards (NAAQS) for total suspended particulates (TSP), and to
develop recommendations for future action and research needs.
The overall approach to the study involved a series of case studies in 14
major urban areas, including a field visit to each area to conduct detailed
inspections of the TSP monitoring sites. No major field measurement
programs were undertaken, but selected hi-vol filters from the case study
cities were analyzed by microscopic techniques to provide information on
the types and sources of the particles collected. Some of the filters were
also analyzed by EPA for metals and nonmetallic inorganic ions.
Although this study approached the national particulate problem with many
of the same tools an individual air quality planner would use to assess
the problem in a particular urban area (air quality data analysis, emis-
sions data analysis and modeling, and analytical particle identification),
the intent of the study was to provide comparable data for extrapolation
to the national scale rather than a detailed study of each urban area.
Though the results and conclusions of this study are primarily applicable
on a national scale, the technical information can be useful to local air
quality planners through comparisons of the more general data to the
specific problems in their urban areas.
IX
-------
GENERAL FINDINGS
Factors Affecting Attainment
The study grouped the various factors that affect attainment and maintenance
of the TSP standards into five conceptual categories as follows: three
general categories of sources that contribute to the TSP loading at any
given point, and two factors that act to modify the ambient levels
measured. The three general categories are emissions from traditional
sources, emissions from nontraditional sources, and natural and transported
particulates. The two modifying factors are meteorology and monitoring
network configuration and siting.
Traditional sources are those sources that have historically been of concern
to the air pollution control community. Fuel combustion sources, solid
waste disposal operations, and industrial process emissions, including
both stack and fugitive emissions, comprise this segment. Particulate emis-
sions from traditional sources were found to have decreased significantly in
most areas since the State Implementation Plans (SIPs) were applied, with re-
sulting improvement in TSP air quality. However, these sources are still
the dominant problem in some urban areas with much heavy industry; in these
Q
cities, traditional sources may be contributing anywhere from 15 yg/m-5 in
residential areas to over 60 yg/m^ in industrial neighborhoods. These
levels can be reduced significantly if sources are brought into compliance
with stringent regulations. The traditional sources that remain the
greatest problem are the primary metals and mineral products industries,
as well as fugitive emissions from all traditional sources. While the
stack emissions are usually controlled through standard regulations, the
control of fugitive emissions requires new regulations such as those
which control operating practices or concentrations crossing the boundary
line.
Nontraditional sources include reentrainment of road dust, fugitive dust
emissions from construction and demolition operations, particulate auto
-------
exhaust emissions, dust from unpaved areas, and emissions from other urban
activities. These sources have not been controlled adequately under
existing SIPs, and in many cases were not even considered. A principal
finding of the study was that particulates from these sources contribute
25 to 35 yg/m-* to the citywide TSP levels, thereby j?reventing many urban
areas from attaining the ambient standards; unless controlled in the
future, they will continue to do so. The control measures that must be
applied are also nontraditional: controlling dust deposition or cleaning
streets to reduce reentrainment from paved roads; chemical stabilization,
cover, or other means of lowering surface disruption at construction/
demolition sites or on unpaved areas; reducing lead in gasoline or re-
quiring particulate control devices for motor vehicles.
Natural and transported particulates consist of those contributions to the
large-scale TSP level frequently called "background." The long-range
transport of both primary and secondary particulates, as well as naturally
occurring particulates, are included in this portion. Normally, the dif-
ferent contributions to these particulate levels have not been considered
in air quality planning, although their reduction may be essential for
standards attainment. While natural particulates are not amenable to
control, the nonurban levels of sulfates and nitrates (in the range of
5 to 10 yg/nr* in the northeast) and transported primary particulates (up
to 30 yg/m^ in dense metropolitan areas) can be controlled through
regional planning. In addition, the portion of urban secondary particulates
(as much as 10 yg/m-^) that is formed from gaseous emissions within the urban
area should be considered in regulatory programs.
Meteorological factors were analyzed to determine the extent of their
effect on the magnitude of ambient particulate levels. For example, pre-
cipitation was found to change annual average TSP levels as much as
3
20 yg/m over the normal range of precipitation values. The location of
the area with respect to movement of air masses from other regions can
affect the nonurban levels coming into the area; the number of heating
degree days, along with the fuel available for heating, can also affect
XI
-------
the TSP levels. Although not a factor subject to control, the variations
in meteorological effects over time and in different locations must be con-
sidered in air quality management planning. The results of this study's
analyses will help in doing this more quantitatively.
Network configuration and hi-vol siting affect the perception of particu-
late levels. Variations in siting practice and network configuration
make comparisons of values between cities and neighborhoods difficult.
These variations frequently distort the overall picture and constrain
problem identification and subsequent strategy development which is
essential for attainment/maintenance. Often, differences in the monitor's
location of only 100 meters on the horizontal or 15 meters on the vertical
can change the measured TSP levels as much as 50 yg/nr*. Nevertheless, such
variations in monitor siting are within the general guidelines prescribed
by EPA.
SPECIFIC FINDINGS
Underlying the general conclusions outlined above are a variety of specific
findings in the areas investigated. The evidence developed and findings
reached are described below.
Microscopic Analysis of Filters
In the course of the study, 300 hi-vol filters from the 14 study cities
were examined by polarizing microscope to determine the mass percent con-
tribution of the major generic categories of particles. The average mineral
content of the particulates was found to be 65 percent of the mass, while
combustion products averaged 25 percent. The cities with the highest per-
centage of minerals were Denver and Oklahoma City, both areas of dry
climate with acknowledged fugitive dust problems. Both of these cities
averaged over 80 percent minerals, with most reported as quartz and
calcite. In only two cities, Cincinnati and Cleveland, did the contribution
from combustion products exceed 40 percent. Biological material was not a
Xll
-------
major component on the average, though some filters collected in the
spring showed as much as 90 percent biologicals. The contribution of
rubber was also highly variable, ranging from 16 percent in San Francisco
to 2 percent in Philadelphia and Washington, and averaging 7 percent.
Aside from the particle identification, valuable information was also
collected on the technique of microscopy as a routine analytical method by
using two laboratories, multiple analysts, and blind replicate examination
of filters. The comparison of different analyses of the same filter indi-
cates that some very sizable differences can occur between analysts;
statistical analysis of the results of the blind replicate examination
of samples warns against relying heavily on any single result. However,
composited results can be used to point out important differences in the
composition of suspended particulate matter as a functioii of time and
location.
Traditional Source Emission Density
By comparing the density of emissions from traditional sources (fuel com-
bustion, industrial processes, solid waste incineration) and the citywide
TSP levels, it was possible to develop a relationship for estimating the
impact of increases or decreases in the emission density. While the
actual impact will vary with the completeness of the emission inventory,
the meteorological and climatological factors, and the topography of the
area, a rule of thumb is that the citywide TSP level will decrease ap- A
proximately 8 yg/m-^ for every 100 tons/year/square mile decrease in the '
emission inventory.
This same analysis indicated that even the cleanest cities (with respect
O
to traditional sources) will be 30 to 40 yg/m above the local nonurban
levels due to nontraditional sources. Therefore, all cities of the size
considered in this study can be expected to have at least one site ex-
ceeding the secondary annual standard. Cities with traditional source
emission densities greater than 100 to 200 tons/year/square mile will
xiii
-------
need further control of traditional sources to be able to meet the primary
annual standard.
Fuel Combustion Emissions
In the 14 cities studied, the fuel combustion contribution to inventoried
emissions ranged from less than 20 percent in clean-fuel, industrialized
areas to well over 90 percent in totally nonindustrial areas. The total
emissions due to fuel combustion vary with the type of fuel available,
the degree of industrialization, population, and heating degree days. In
cities where fuel oil was the predominant energy sources (used in more
than one third of the residences), filters selected for microscopic analysis
x^/ contained from 9. to 13 percent oil soot. Where coal combustion accounted
for over 20 percent of the industrial BTU usage, the filters contained from
3 to 30 percent coal combustion products.
Fugitive Emissions
Particulates from fugitive emissions at industrial facilities were found
to cause significant TSP levels. The actual impact of fugitive emissions
varies greatly with the distance from the source, the windspeed and
direction, other climatological factors, and the type of emission. On
an average basis, sites in an industrial area that are exposed to fugitive
emissions can be expected to have annual mean TSP concentrations 25 yg/m3
higher than similar sites not affected by fugitive emissions. Of all the
monitoring sites in industrial areas that were reviewed under this study,
only those with a fugitive emission influence had annual means above
150 yg/m3; all of the monitors with a fugitive emission influence violated
o
the primary annual standard of 75 yg/mj.
The control of fugitive emissions is usually pursued under either general
nuisance provisions in the regulations, a property line concentration or
visibility restriction, or a specific requirement for control of an opera-
tion. The effectiveness of each of these types of regulations varies with
xiv
-------
the authority and enforcement program of the air pollution control agency.
Specific requirements are felt to be most effective for major sources where
the principal contributors can be identified.
Vehicular Contribution to Ambient TSP
Measurable contributions to the ambient TSP loadings were found to result
from motor vehicle activity (tailpipe emissions, tire wear, reentrained
dust); in some cases, these contributions were quite high and caused viola-
tions of the standards. By far the largest contribution in urban areas was
due to reentrained particulates from paved roads. The wheels of vehicles
grind up the large, nonsuspendible particles, break up their cohesive
bonds, and impart kinetic energy to the particles. The amount of particulate
reentrained is directly related to the amount of dirt on the road, the speed
of the vehicles, and the level of traffic.
The impact of reentrained dust on the measured air quality depends upon
the monitor placement. Between 10 and 100 meters back from streets and
arterials, the TSP concentration was found to fall off with the slant dis-
tance of the monitor from the road. The contribution from the reentrained
dust (in yg/m^) can be roughly estimated by 0.1 ADT/SD where ACT is the
average daily traffic and SD is the slant distance (in feet). The average
impact on monitors reviewed in this study was about 10 to 15 yg/rn-^ in
residential areas and 15 to 20 yg/m-^ in commercial and industrial areas;
o
however, at an individual site the impact may be as much as 50 yg/m .
Other contributions to TSP arising from motor vehicle activity that were
studied include tailpipe emissions and tirewear. The impact of tailpipe
emissions was quantified using elemental analysis with lead as the tracer
element. Commensurate with the level of motor vehicle activity, TSP con-
centrations due to exhaust averaged about 3 yg/m^ in residential areas and
4 to 5 yg/m-* in commercial and industrial areas, based on 20 to 25 percent
of exhaust particulate being lead. A special study of particle sizes by
elemental composition indicated that most lead containing particles have
xv
-------
an effective aerodynamic diameter of 0.25 ym; no major difference in lead
levels with distance from the road was noticed.
Tirewear was studied using microscopic analysis of selected hi-vol filters.
This component of participates due to vehicular activity had the same
neighborhood pattern as the other components - highest in commercial and
industrial areas (5 to 10 yg/m3 and 3 to 7 ug/m3, respectively) and lowest
in residential areas (2 to 5 ug/m3). The levels of rubber on the filters
from monitors located particularly near heavy traffic average twice the
levels at other sites. Despite this relationship, significant concentra-
tions of tubber were often found at monitors not obviously influenced by
traffic.
A final contribution from motor vehicle activity, which was considered only
indirectly because of the lack of appropriate monitor sites in the study
cities, was reentrained dust from unpaved roads. While significant TSP
concentrations can be expected at monitoring sites near unpaved roads,
the current emission factors given in AP-42 provide emission estimates that
can be up to 30 times the emissions from traditional sources. Because of
the different dispersion characteristics of ground level sources versus
stacks and the different particle size characteristics, care must be taken
in using these numbers. If these fugitive dust emissions were treated the
same as traditional emissions and used in rollback calculations, excessive,
undeserved emphasis would be placed on these sources at the expense of
control of traditional sources. The use of these numbers should be limited
to inputs to models that adequately reflect the deposition and other removal
of the particulates.
Construction/Demolition
In nine of the study cities, construction activity (urban renewal, highway,
subway, individual buildings) near monitoring sites was analyzed for its
ambient air impact. The analyses indicated that construction activity
does have an impact on very local TSP levels, but that the effect is not
xvi
-------
readily predictable. Construction will generally elevate concentrations
downwind from the site for distances up to a mile with the amount of in-
crease related to the level of activity, type of activity, distance from
the activity, and control measures employed. Monitors within half a mile
of construction may have annual geometric means 10 to 15 yg/m^ higher than
normal. On^an average basis, the effect on the citywide TSP level is ex-
pected to be only 1 to 3 yg/m^, unless major, widespread construction
activities such as urban renewal are underway.
As with reentrained dust from unpaved roads, the currently available emis-
sion factors for construction activity provide estimates of fugitive dust
emissions well above that would appear logical in terms of the TSP levels
actually observed. When combined with the calculated emissions from un-
paved roads, the fugitive dust emissions in two of the major counties
studied amounted to over 60 times the traditionally inventoried emissions.
Monitor Siting
A major consideration in the understanding of the ambient TSP levels in an
area is the actual siting of the hi-vols used for assessing the attainment
of the standards. The network configuration and the placement of the monitor
affects the extent to which the levels reported are representative. The
visits to over 150 hi-vols in 14 cities provided a unique opportunity to
determine how monitoring networks vary within and between cities. The
primary difference noted was the height of the monitor. Typically, hi-vol
heights range from ground level to the top of many-story buildings. The
median height of hi-vols in the 14 cities ranged from 6 feet in Birmingham
to almost 50 feet in Providence. Despite this order of magnitude range,
most sites were still within EPA's guidelines for monitor siting.
Local influences are also important when trying to determine the repre-
sentativeness of the monitor sites. Almost half of the sites visited were
felt to have at least one local influence with over one-quarter of the
sites judged to have major influences. Because of the frequency of
xvii
-------
occurrence and range of local influences, some sites were further classified
as having "undue" impact; typical TSP loadings due to the undue impact were
around 20 yf/m .
Meteorology^
The day-to-day, season-to-season, and year-to-year fluctuations in meteo-
rology often confound analyses conducted to determine the cause of TSP
levels. In an attempt to quantify these confounding effects, relationships
between a number of meteorological parameters and TSP concentrations were
examined in several cities. The most detailed analyses centered around the
effects of precipitation amounts, heating degree days, windspeed, and the
rate at which TSP concentrations return to normal levels following periods
of precipitation.
Precipitation was found to be very effective in reducing TSP levels in
areas where high concentrations resulted from either industrial or fugitive
dust sources. Average concentrations in these areas decreased steadily
with increasing 48-hour precipitation amounts. The effect of precipitation
was greatest on the day it occurred and lasted on the average about 2 days.
In high concentration areas, concentrations measured during the last half
of a 48-hour period with precipitation > 0.25 in. averaged approximately
half of concentrations measured during 48-hour periods with negligible pre-
cipitation. Concentrations at typical urban sites on days with measurable
precipitation were about 75 to 85 percent of average values for the site
and time of year.
The effect of windspeed on TSP concentrations was found to vary with moni-
tor siting and neighborhood characteristics. Very high concentrations of
TSP were typically associated with light winds and poor ventilation. In
industrial areas, where major contributions were made from point sources,
concentrations decreased with increasing windspeed up to about 10 miles
per hour; at higher windspeeds, TSP levels appeared to be invariant with
xviii
-------
average speed, perhaps as a result of a balance achieved between dilution
and an increased rate of reentrainment of fugitive dust. In the nonin-
dustrial urban areas studied, no significant relationship between TSP
level and average windspeed was found.
A multiple regression analysis was applied to 5-year sets of data from 10
of the study cities. This analysis indicated that a 1-inch increase in
the annual precipitation level would decrease the annual TSP concentration
by 0.4 yg/m^ and that for every 1000 degree-day increase in the annual
o
heating demand the annual TSP concentration would increase 2.5 ug/m .
RECOMMENDATIONS
This study identified several areas requiring increased control emphasis
and program redirections, a number of problems needing guidance, and several
major research needs. The guidance and research needs can primarily be
resolved by EPA, while the control and program recommendations are more ap-
propriately directed to state and local air pollution control agencies.
State and local agencies should concentrate on better defining the cause
of the TSP standards nonattainment and should be pursuing more stringent
regulations and enforcement in those heavily industrialized areas where
traditional sources are major contributors to nonattainment. The agencies'
efforts should also be directed at quantifying the contributions from non-
traditional sources and developing and enforcing regulations appropriate
to these problems; this is particularly important where traditional sources
are a minor problem or where they are already in compliance with stringent
regulations.
EPA needs to provide more guidance in the following areas: monitor siting;
the problem of natural reentrained dust and large particles on short-term
standards attainment; the analytical methods available for particulate
studies; the interpretation of emission factors currently being used for
fugitive dust emissions.
xix
-------
Two major national control and research priorities emerge from this study.
A broadly supported effort directed at developing controls for urban re-
entrained dust and similar nontraditional particulate sources will be re-
quired in order to meet the present particulate standard. In addition,
reduction of the regional-scale burden of transported and secondary par-
ticulates, especially sulfates, should also be pursued if standards are to
be attained in the northeast.
xx
-------
SECTION I
INTRODUCTION TO STUDY
This report presents the results of a study conducted under the auspices
of the Control Programs Development Division of EPA's Office of Air Quality
Planning and Standards. Its overall purpose was to assess the national
particulate problem, based on case studies in 14 major urban areas, with
emphasis on identifying the factors involved in the attainment or non-
attainment of the National Ambient Air Quality Standards (NAAQS) for total
suspended particulates (TSP). The study was intended to improve technical
understanding of the TSP problem, to provide specific guidance to the states
in TSP problem analysis and control strategy formulation, and to develop
recommendations for EPA concerning future program direction and research
needs. Consequently; this document is directed primarily at those managers
and air quality planning specialists at EPA and the various state and local
air pollution control agencies who are concerned with the development of
TSP control programs.
BACKGROUND
The Clean Air Act, as amended in 1970, provides for National Ambient Air
Quality Standards (NAAQS) for major air pollutants, for the definition of
Air Quality Control Regions (AQCRs) and for a joint federal-state control
program to ensure that the air quality in the various AQCRs throughout
the country attains the NAAQS. The Act also provides for State Implemen-
tation Plans (SIPs) , to be prepared by the states and approved by EPA,
which present- each state's plan to implement the national standards within
the periods specified in the Act 3 years after approval for attaining
-------
the primary or health-related standards and a "reasonable time" for at-
taining the more stringent (welfare-related) secondary standards.
In April 1971 EPA promulgated NAAQS for five major pollutants, including
total suspended particulates (TSP). Subsequently, the states prepared
and submitted implementation plans, and the bulk of the provisions of the
SIPs relating to TSP control were approved by EPA on May 31, 1972. Thus
the date of May 31, 1975, became the anticipated date for most AQCRs to
attain the primary standard for TSP, and it is on the occasion of the
passing of this date that the present assessment is undertaken.
On the whole, most aspects of the federal-state TSP control effort have
been successful. The average annual geometric mean concentration of TSP
3
at urban sites has declined, as seen in Figure 1, from 81 to 68 ug/m over
the 5 year period. This represents a decrease of over 16 percent or, con-
sidering only that portion above nonurban background, a 25 percent reduc-
3
tion in the levels above a typical background estimate of 30 ug/m . None-
theless, by the criterion of actually attaining the standards, there is
still a significant number of AQCRs where the control programs have not
yet completely succeeded. In 1974, 136 of the 247 AQCRs in the country
violated either the annual or 24-hour primary standard, or both, as com-
pared to 112 of the 182 AQCRs in 1970 which reported at least minimal data,
It is likely that at least some of the 65 AQCRs which did not report suf-
ficient data in 1970 were violating either or both of the primary stan-
dards that year. Nevertheless, the general picture of the attainment
status of the AQCRs is that better than half are still violating one or
both of the primary standards.
This picture of widespread nonattainment of the standards seems to contra-
dict the picture of decreasing TSP concentrations already mentioned. The
two pictures, however, are not really inconsistent, for several reasons.
A single site exceeding the specified ambient air quality standards any-
where in the AQCR is sufficient to state that the AQCR is not attaining
-------
100
fO
E
z
o
<
o:
o
o
90
80
4-QUARTER RUNNING AVERAGE
\
\ /
ANNUAL MEAN
PRIMARY ANNUAL STANDARD
70
60
\
I
SECONDARY ANNUAL GUIDE
1970
1971
1972
YEAR
1973
1974
Figure 1. Trend in national average TSP levels (average of 1014 urban sites)
-------
the standards. High ambient TSP levels in an AQCR may have been reduced
significantly but not quite enough to attain the standards; or perhaps
the majority of sites in a given area have attained the standards, while
one isolated site or neighborhood has for some reason failed to do so.
It is clearly very difficult to obtain an accurate nationwide picture of
attainment status based on simple data summaries, and essentially impos-
sible to obtain any knowledge of the significant factors affecting attain-
ment , the problems encountered, or the successful approaches utilized,
without a more intensive study of the distribution of air quality levels
and emission sources in specific geographical areas. Thus the present
study was undertaken, to provide a more thorough and detailed assessment
of the national TSP problem.
STUDY APPROACH
The overall approach to the study involved the identification of those
factors influencing the attainment or nonattainment of the standards, the
assessment of their significance on a national basis, and the development
of a recommended action program for EPA and the state and local control
agencies.
The primary source of new information for this national assessment was a
series of case studies in 14 major urban AQCRs. These studies involved
analyzing, for each urban area, the air quality (1974) and emission data
(circa 1974) on TSP, meteorological data from the National Weather Service,
the control regulations in the State Implementation Plans, planning materials
maintained by EPA, and published technical articles or reference materials.
Each city case study also involved a field visit to the city to conduct
detailed inspections of most of the TSP monitoring sites in the area.
Although the primary purpose of the study was not to produce new analytical
data, several specific analyses were performed. Selected hi-vol filters
from each of the 14 case study cities were analyzed by optical microscopic
-------
techniques to provide information on the types and sources of the particles
collected. Some of these filters were reanalyzed for quality control by
the same or other microscopists and laboratories; others were analyzed by
EPA for metals and nonmetallic inorganic ions by standard National Air
Sampling Network (NASN) analytical procedures. EPA provided information
from special field studies in two of the study cities, including wind
directional monitoring, diurnal variations in elemental composition, and
particle sizing data. Data resulting from all these analyses are compiled
in Volume II of this report; discussion of the results is included in this
volume or in the individual city volumes as appropriate.
ORGANIZATION OF THE STUDY REPORT
The final study report consists overall of 16 separately bound volumes.
This document, Volume I, summarizes the data from the other volumes and
should be considered the primary product of the study. Volume II summa-
rizes the analytical data developed during the study, and Volumes III
through XVI are the working documents compiled for discussion purposes
for each individual city. The subject of each volume is listed below:
Volume I - National Assessment
Volume II - Particle Characterization
Volume III - Denver
Volume IV - Birmingham
Volume V - Baltimore
Volume VI - Philadelphia
Volume VII - Washington
Volume VIII - Chattanooga
Volume IX - Oklahoma City
Volume X - Seattle
Volume XI - Cincinnati
Volume XII - Cleveland
Volume XIII - San Francisco
Volume XIV - Miami
Volume XV - St. Louis
Volume XVI - Providence
-------
The remainder of Volume I is organized as follows:
Section II - presents the general findings of the study,
summarizes the findings in each individual case study city,
and identifies the components of the five major factors
influencing TSP levels.
Section III - presents the overall national assessment of
the significance of each of the five factors. It places
in perspective the components of each of these factors and
quantifies their relative contributions to air quality.
Section IV.- summarizes the attainment factors identified
in the city studies and the implications of the study for
control strategy development.
Section V - provides the detailed recommendations.
Appendices - include various supportive and reference materials
and detailed discussions of data and analyses used in Section III.
-------
SECTION II
GENERAL FINDINGS OF THE STUDY
The individual reviews of the TSP attainment situation in the 14 study
cities provided a wide range of general findings. Some of these find-
ings are specific to individual cities, and others are generally
applicable throughout the study cities. The analysis of these findings
among the different cities serves as the basis for an interpretation
of TSP attainment nationwide. This section summarizes the general
findings from the city case studies (reported in Volumes III through XVI)
and identifies the major factors affecting the TSP attainment problem.
STUDY CITY ANALYSIS
Selection Process
The 14 cities (AQCRs) used in the case studies were selected to repre-
sent a cross section of urban areas. Such factors as proximity to
bodies of water, topography, meteorology, degree and type of industri-
alization, fuel usage, and air quality levels were considered. As
shown in Figure 2, the study cities are dispersed around the contiguous
48 states, with many located in the East where a large portion of the
population and industry of the United States is concentrated. In nature
and density, they range from the old industrial cities of the east
coast to the newer, less congested cities of the plains and west coast,
and they include a cross section of the climatological regimes of the
nation. The'cities range in size from Philadelphia, the nation's
-------
00
-Sp^-PROVIDENCE
4-PHILADELPHIA
<4^
;YS-8ALTIMORE
Figure
2. Geographical distribution of 14 study cities
-------
fourth largest urban area, to Chattanooga, a small industrial city in
the southern mountains. Basic statistics on population, topography,
and employment for each of the 14 cities are presented in Table 1.
Two of the major factors that are important in understanding the po-
tential for a TSP problem are the industrial nature and dispersion
characteristics of an area. The industrial nature indicates the general
level of emissions from major point sources, while the dispersion
characteristics dictate the degree to which these emissions affect am-
bient levels of particulates; Table 2 lists the 14 cities classified
according to the possible combinations of these two characteristics.
Considered under the category of dispersion characteristics are non-
urban particulate levels, extremes in rainfall, sea breezes, and topo-
graphy. Therefore, cities such as Miami, San Francisco, and Providence,
which are dominated by their proximity to the ocean, offer good disper-
sion qualities; cities with ventilation dominated by valley topography
(Chattanooga, Birmingham, and Denver) are considered to have adverse
dispersion characteristics. Industrialization was a judgmental catego-
rization based upon the level of employment in the manufacturing sector,
the nature of the manufacturing, and the density of emissions.
Table 3 summarizes air quality for the study year (1974), and Table 4
characterizes sources and emissions for the 14 cities, which are grouped
by industrialization categories. Other tables and figures giving charac-
teristics of the cities are included in Appendix A.
Methodology of City Visits
The core of the city case studies was one or more field visits to each
city. Prior to visiting each study city, a variety of data summaries
-------
Table 1. POPULATION AND PHYSICAL SETTING OF CASE STUDY CITIES
Urban area
Philadelphia
San Francisco
Washington, D.C.
Cleveland
St. Louis
Baltimore
Seattle
Miami
Cincinnati
Denver
Providence
Oklahoma City
Birmingham
Chattanooga
1970 population
AQCR
5,635,406
4,639,949
2,862,912
3,383,879
2,476,757
2,078,379
1,937,371
2,435,089
1,660,495
1,252,007
1,502,801
783,403
1,045,599
689,494
Rank8
in U.S.
4
6
6
9
10
14
17
18
21
24
30
42
43
96
Central
city/county
1,948,609
715,674
756,510
1,721,300
622,236
905,759
1,156,633
1,267,792
924,018
514,678
580,261
526,805
644,991
255,064
Population density
per sq . ml .
SMSA
1357
1254
429
1329
547
917
337
621
644
366
1347
300
272
307
Central
city /county
15,116
15,904
12,402
9,893
10,201
11 ,613
6,350
9,763
5,780
5,418
9,896
579
3,785
2,284
Physical setting
Large bodies
of water
Delaware River
Ocean; S.F. Bay
Potomac River
Lake Erie
Mississipi River
Patapsco River,
Baltimore Harbor
Puget Sound
Ocean; Everglades
Ohio River
Narragansett Bay
-
-
-
Topography
Slightly rolling
Significant hills
Slightly rolling
Flat; river valley
Slightly rolling
Slightly rolling
Significant hills
and valleys
Flat
River valley, sig-
nificant hills
Mountains to west;
rolling to east;
river valley
Slightly rolling
Flat
Valley between sig-
nificant hills
Valley between
sharp ridges
Manufacturing
employment (AQCR)
Magnitude,
thousands
603
347
53
484
279
184
139
144
195
99
201
44
92
128
Percent
of total
34.1
24.4
7.2
36.1
35.1
29.3
26.7
10.5
37.7
20.8
44.2
19.0
31.1
52.6
eased on Urbanized Area population.
-------
Table 2. STUDY CITIES BY THEIR DISPERSION AND
INDUSTRIALIZATION CHARACTERISTICS
Industrialization
Light
Moderate
Heavy
Dispersion characteristics
Favorable
Miami
San Francisco
Providence
Moderate
Oklahoma City
Washington, D.C.
Seattle
Baltimore
Cleveland
Cincinnati
Philadelphia
St. Louis
Adverse
Chattanooga
Denver
Birmingham
-------
Table 3. SUMMARY OF SITES EXCEEDING AIR QUALITY STANDARDS IN 1974
N)
City
Heavily
Industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence0
Lightly
industrialized
Washington, D.C
Oklahoma City
Miami
San Francisco
Total no.
of sites
with complete
1974 data
25
13
10
29
31
25
12
22
10
21
9
14
17
17
Annual s tandard
No. sites exceeding
standard
Primary,
75 ng/m3
12
11
7
9
15
8
5
14
2
1
2
5
2
0
Secondary,3
60 ug/m3
21
11
9
14
26
21
5
21
4
5
5
6
7
2
Highest
geometric mean,
Hg/m3
175
144
122
134
158
130
101
131
105
88
102
107
86
74
24-hour standard
No. sites exceeding
standard
Primary,
260 ug/m3
6
7
4
6
3
0
1
13
2
0
9
5
0
3
Secondary,
150 ug/m3
15
11
8
14
13
7
7
22
5
1
9
13
3
10
7. total obs.
> standard
Primary
4.7
NAb
IMS'
0.9
1.1
0.1
0.9
NAb
0.3
0
2.2
0.7
0.2
0.2
Secondary
17.6
K&
~8
7.6
8.7
4.2
6.9
NAb
0.8
0.8
NAb
5.5
1.8
1.3
Highest
value ,
Hg/-3
534
499
624
NAb
NAb
296
434
565
320
173
527
NAb
NAb
286
aThe secondary annual is a guide, not a standard.
NA - Raw data not available.
CTotals are for statewide monitoring network.
-------
Table 4. SOURCE AND EMISSIONS CHARACTERIZATION OF STUDY CITIES
City
Heavily industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Moderately industrialized
Chattanooga
Denver
Seattle
Providence
Lightly Industrialized
Washington, D.C.
Oklahoma City
Miami
San Francisco
Major point
sources
Steel mills, utilities,
chemicals
Steel mills, cement
Refineries, coking,
smelting, chemicals
Steel mills, incinerators
Coking, steel mills,
grain handling
Utilities, manufacturing
Found r ies , c emeat ,
minerals
Utilities
Manufacturing
Utility, foundries
UtillCy
Utilities, grain, asphalt
Minerals
Minerals , chemicals
Predominant fuels*
Residential
Gas
Gas
Oil
Gas, oil
Gas
Gas
Electricity
Gas
Oil
Oil
GaS
Gas
Electricity
Gas
Industrial
Gas , coal
Gas
Oil, gas
Oil
Gas
Gas, coal
Gas
Gas
Oil
Oil
Gas, oil
Gas
Gas, oil
Gas
Traditional
sources,
emission
density
TPY/sq. mile
335
488
245
240
411
135
88
60
37
30
92
i
30
60
Compliance comments
About 1/3 of major sources in compliance;
smaller sources (including incinerators)
unknown
Most sources under compliance schedules;
compliance about half complete
Most sources in compliance
Most sources In compliance, a few under
plans for compliance
Most sources In compliance
Most sources In compliance; rest (Includ-
ing several large sources) expected withl
2 years
Most sources In compliance
Most sources In compliance
Most sources In compliance
Compliance status not certain - most
seem to be In compliance
Most sources In compliance
Most sources In compliance
Moat sources la compliance
Most sources In compliance
"Those fuels whose usage is greater than 1/3 of total Btu's.
-------
and analyses were prepared and studied, using data on air quality,
emissions, and compliance that were provided by the Project Officer
from EPA's AEROS data banks; local agencies were allowed to review,
comment on, or revise the data as appropriate. Air quality and emis-
sion data were studied for patterns and trends, both temporal and
geographic. Emission and compliance data were used to consider pat-
terns of enforcement and compliance.
The primary purpose of the field visits was to study in detail the
hi-vol monitoring network and to gain a degree of understanding about
the overall nature of the study city and the general patterns of land
use, such as the relationships among industrial and residential areas.
During the field visit, the study team also consulted with technical
staff members from the appropriate pollution control agency concerning
those areas where the EPA data base was incomplete or anomalous, and a
member of the agency staff usually accompanied the visiting team on
the monitoring site study visits. Additional efforts during the study
visits in some cities involved seeking data on traffic volumes, street
cleaning practices, etc., as appropriate to the particulate air pol-
lution problem in the city. In most cases, historical hi-vol filters
from agency files were also selected and brought back for analysis.
Following the field visit, the monitoring site information was reviewed
and compiled into an overall site classification analysis. Other in-
formation and data obtained were integrated with that previously avail-
able, and an overall assessment concerning the factors affecting attain-
ment was formed. The data summaries, analyses, and conclusions relat-
ing primarily to a single city were then compiled into a working docu-
ment for each of the study cities; these documents comprise Volumes III
through XVI of this report.
14
-------
City Study Findings
The results and findings of the city case studies are quite varied in
detail and in breadth of applicability. Those specifically relevant to
only the individual city are presented in the separate volume for that
city. Those that are of significance for the overall conclusions of
the study are summarized here. Appendix A presents tabular summaries
of other data relevant to the various study cities.
Baltimore - A heavily industrialized city, Baltimore has had an inten-
sive pollution control program since the mid-1960s. This has contri-
3
buted to a 50 p.g/m decrease in the annual average at the center-city
NASN site over this time. Nonetheless, the annual primary standard is
still exceeded at nine sites, mostly in the center city and harbor in-
dustrial areas. Under vigorous enforcement of very stringent regulations,
most of the city's major industries and the municipal incinerators have
reduced their emissions an average 70 percent; on the other hand, emis-
sions from the iron and steel industry, which comprise 85 percent of
the total point source emissions in the metropolitan area, have been
reduced only 15 percent. One steel mill clearly contributes to TSP
problems at several industrial sites; however, at the industrial site
3
with the highest annual average (134 ug/m ), local fugitive dust emis-
sions also make a significant contribution to the elevated levels.
Three or four sites located in the urban center appear to be above the
standard without any particular local source influence. Part of this
urban increment is probably due to the use of oil for residential space
heating, one category of source which is not closely regulated; however,
much of the TSP measured at these sites mv--,t simply be attributed to
the dense level of activity in the urban area. TSP levels at the center
city site are further elevated by immediately adjacent expressway
construction.
15
-------
Birmingham - The improvement in air quality has been substantial in
Birmingham, but still only two suburban sampling sites met the annual
primary standard in 1974. Values have fallen from around 300 to 145
3
ug/m at the most polluted industrial site, and commercial and other
industrial sites have shown similar decreases of roughly 50 percent.
Four industrial sites reported 1974 annual geometric means over 125
3
ug/m because of proximity to the predominant steel industry and the
associated coking and foundry operations, which account for the bulk of
point source emissions. Stringent regulations and vigorous enforcement
have resulted through 1974 in about half the emission reduction ultimate-
ly expected, with the balance anticipated over the next 2 years as com-
pliance plans are completed. The general effect of dense urban activity
on TSP levels is not seen clearly in Birmingham because of the much
larger industrial process contribution, but may well be a problem for
future consideration.
Chattanooga - A moderately industrialized city, Chattanooga has also
experienced a significant decline in TSP levels due to a trend away
from the use of coal and, more recently, to vigorous control of indus-
trial emissions. However, five of twelve sites, all in industrial or
commercial locations, continue to exceed the annual primary standard
3
with annual means between 80 and 101 |ig/m . Of these, two are affected
significantly by local industrial sources - a quarry and a cement
plant - and the others by general downtown commercial activity or major
traffic arteries. An overall difficulty relates to adverse topography
and meteorology; particularly stringent emission control for both in-
dustrial and fugitive sources will be needed to meet the standards
under the adverse dispersion conditions prevalent in Chattanooga.
Cincinnati - The TSP trend has been downward since the mid-1960s, with
3
the NASN center-city site experiencing a 70 |_ig/m decrease in its annual
average. The major industries - transportation equipment, fabricated
metals, paper, chemicals, and a small power plant - have made large
16
-------
reductions in particulate emissions, primarily by switching from coal to
natural gas or otherwise reducing fuel combustion emissions. Nonetheless,
seven sites still exceeded the annual primary standard in 1974 at loca-
tions in the central business district (CBD) and the industrial valley,
3
with one site measuring 130 |ag/m as an annual average. All sources are
expected to come into compliance over the next 2 years, during which
time further reductions are expected. However, full compliance may not
result in standards attainment because fugitive emissions and fugitive
dust are likely to remain a significant source of particulates at several
sites in the CBD and industrial areas unless control measures are taken.
Cleveland - Although there has been a fairly steady decrease in ambient
TSP levels since the early 1960s, 1974 levels at 12 of 25 sites in the
county violate the annual primary standard. The city is large and heavi-
ly industrialized, with an emission density of 335 tons per year per
square mile, and has the highest annual average TSP levels of the cities
3
studied (175 ug/m ). Primary metals, fabricated metal products, ma-
chinery and transportation equipment are the predominant industries and,
along with the utilities, the largest sources of particulate emissions.
Control efforts by a few problem sources have resulted in substantial
emission reductions, but overall control efforts have not been effective.
While other factors, such as fugitive dust, may very well become apparent
as the massive industrial contribution is reduced, the most important
factor influencing nonattainment in Cleveland is the level of industrial
emissions. These include not only major point sources but also industrial
fugitive emissions and the vehicle-entrained fugitive dust emissions asso-
ciated with industrial areas.
Denver - In contrast to most other areas studied, the Denver AQCR has
shown no definitive trends in TSP levels during the past 6 years. Des-
pite its low level of industrialization, Denver has had TSP concentrations
17
-------
well above the annual standards since levels were initially monitored
in 1957. In the long term, Denver County has shown approximately a 20
percent improvement in its air quality since 1965. However, in 1974
only one out of 22 sites, a site located in a rural area, met the secon-
dary standard; 14 of the sites exceeded the primary standard. This
general lack of attainment of the standards has been commonly attributed,
in part, to the arid climate which allows easy reentrainment of fugitive
3
dust. The highest annual mean (131 ug/m ) was recorded at a site ob-
viously influenced by fugitive dust, where 36 percent of the observations
were above the secondary 24-hour standard and 14 percent were above the
primary 24-hour standard. The impact of both fugitive dust and tradi-
tional industrial source emissions is spread throughout the region be-
cause of the poor ventilation and topographic characteristics of the
region. In addition, an inadequate data base, from which the initial
implementation planning was done, has contributed to the problem o^
attainment in the region. Several emission inventories have been com-
piled for the region over the vears, but the continuing iterations re-
sulted in a lack of consistency, especially between area source inven-
tories, and has prevented the determination of emission trends. Despite
the problem of appropriate emission data, major sources are assumed to
be generally in compliance with the regulations and the state agency is
now pursuing fugitive dust sources.
Miami - The highest ambient TSP levels in Miami have been fluctuating
near the standards for several years. The area is generally free of
major TSP point sources; the largest emitters are stone and gravel
quarrying operations. During 1974, two sites failed to meet the primary
3
standard. The higher, with an annual geometric mean of 86 (ig/m , was
located in a light industrial-commercial area, on a major arterial high-
way, with an auto junkyard and a variety of other unpaved fugitive dust
sources in the immediate vicinity. The other site was at a highway in-
tersection in a rural area, near a major aggregation of cement plant
operations. A special study conducted by the Dade County agency indi-
cated, however, that the cause of the high levels was not the cement
18
-------
plants themselves, but rather the reentraintnent by traffic of material
spilled on the highway by the sizable number of trucks turning at the
intersection.
Because of the general lack of point source emissions, the relative
homogeneity of the area, and particularly the consistency of the monitor
heights, Miami provided a good opportunity to study that portion of
urban TSP levels that appears to result from aggregate urban activity.
Two measures of urban activity were found to correlate well with the TSP
values at the various sites: traffic volumes and the proportion of ad-
jacent land used for streets and parking.
Oklahoma City - An institutional, light-industrial city where gas is
the predominant fuel, Oklahoma City nonetheless has five of 14 sites
where the annual primary standard is exceeded. TSP levels at the NASN
site show only a slight downward trend over time, reflecting the low
density of readily-controlled industrial sources and the inability to
comprehensively control fugitive dust. The high levels at the sites
over the standard can be attributed in part to either significant traffic
exposure, adjacent construction, or (at three of the sites in the cen-
ter city) a combination of central business district traffic and urban
renewal activity. The dry climate and high winds tend to maximize
natural entrainment of dusts, but the urban pollution problem is not
primarily due to particulates from the surrounding rural area.
Philadelphia - Air quality has been steadily improving in the city;
annual average TSP concentrations at the NASN site are down over 100
3
|j,g/m since 1957. However, the annual primary standard was met at only
three of the 10 monitors operating throughout 1974. Of these three
monitors, all in residential areas of the city, only one met the secon-
3
dary annual standard with a value of 59 p.g/m . The large improvements
in air quality have been directly paralleled by reductions in the inven-
toried emissions due to stringent controls on industry (fabricated metals,
19
-------
machinery, electrical equipment, petroleum) and large incinerators,
phasing out of small incinerators and coal burning, and fuel switching
in power plants. The lack of attainment of the standards, despite
stringent regulations and effective enforcement, reflects the generally
high level of TSP entering the city from outlying industrial activity
3
(40 to 50 ng/m ) and the activity associated with an urban environ-
ment, including space heating and vehicular traffic. Small residen-
tial boilers have not been under any control other than visible emis-
sion regulations. Vehicular traffic was shown to contribute up to
3
50 ug/m to the measured levels of TSP for monitors close to the street.
In addition, fugitive emissions from stockpiled materials and a grain-
handling operation were believed to be major influences on the monitor
3
with the highest annual mean (122 |_ig/m ) .
Providence - A downward trend in TSP concentrations has been occurring
since the mid-1960s, and only one sampling site exceeded the annual
primary standard in 1974. A utility, municipal incinerators, and some
industrial processes - primary metals, fabricated metal products and
electrical equipment - are the largest point sources of particulate
emissions. Fuel switching and the closing of incinerators are apparent-
ly responsible for the emission reductions and the corresponding air
quality improvement, although the compliance status of a number of sources
is unknown. The one site which exceeded the annual primary standard
3
with an average of 88 |_ig/m is excessively influenced by a major express-
way immediately adjacent. Oil-fired space heating is a significant
source category which has not received much attention in light of the
general standards attainment. Some portion of the overall favorable
picture may be due to a generally high average sampler height compared
to other cities and to relatively good dispersion characteristics.
St. Louis - The TSP trend has been downward since the mid-1960s with
3
the NASN site experiencing a decrease of 80 ^ig/m in its annual geometi
mean. Fifteen sites, however, violated the annual primary standard in
20
-------
1974. Transportation equipment, primary metal, fabricated metal pro-
ducts, and machinery industries are the largest sources of particulate
emissions, and fuel switching and industrial process controls have
accounted for much of the emission reductions. The majority of the
sources in the Missouri portion of the AQCR are believed by the local
agencies to be in compliance; however, the St. Louis City and 'County
regulations are among the least stringent of those studied. The
Illinois portion is less advanced in terms of degree of compliance, be-
cause control efforts started later, but it should become comparable to
Missouri by maintaining its program of stringent regulations and
strict enforcement. The fact that high TSP concentrations are still
being experienced despite the presumed general compliance indicates that
other factors are important. These include the relatively weak regu-
lations and the lack of source testing as a method for compliance de-
termination in Missouri, as well as the general tendency of dense
center-city sites to be systematically higher due to urban activity.
San Francisco - San Francisco is the one study city that met the annual
primary standards during 1974. Despite its large population and gen-
erally dense urbanization, the San Francisco AQCR has never had a
serious problem with particulates, due largely to the very clean back-
ground air it receives from over the Pacific Ocean and the low level of
heavy industrial activity. Air quality trends at the San Francisco
3 3
NASN site have shown a decrease of about 20 (ag/m from a high of 73 |_ig/m
in 1957. This decrease of over 30 percent in above background levels
is comparable to the trends reported for emission reductions. These
emission reductions occurred in part (pre-1970) because of extensive
fuel switching by the residential sector from coal and oil to gas and
electricity and (since 1969) because of controls on industrial processes
and burning of materials. These controls are no more stringent than the
average found in the cities studied, but their success is maximized by
an extensive, computerized enforcement program conducted by the Bay Area
Pollution Control District. In addition, regulations have been revised
21
-------
over time as data indicate that more stringent controls are feasible
and warranted. The combination of the above factors and the low level
of emissions (170 tons /day for the entire AQCR) meant that only two
monitors out of 17 exceeded the annual secondary standard, and no mo-
nitors violated either the 24-hour or annual primary standards. The
3
highest annual geometric mean of 74 ug/m was measured at a station
highly influenced by fugitive dust. This monitor and many others in
the network are often subject to poor ventilation conditions because
of the valley topography; even so, the 24-hour secondary standard was
exceeded less than 1 percent of the time in 1974.
Seattle - While it also has the advantage of being a west coast city,
unencumbered with TSP transport from other areas, Seattle is further
inland than San Francisco and significant industry is concentrated in
a valley adjoining the city. However, the higher level of annual pre-
cipitation helps keep the TSP concentrations down around those measured
in the San Francisco area. Air quality levels have shown fairly steady
3
downward trends over the years with a 50 ^ig/rn decrease in the annual
average at the NASN site since 1957. Of the 30 sites in the AQCR, the
annual primary air quality standard was exceeded at two stations and
the annual secondary standard was exceeded at another three sites. The
3
highest concentrations (60 to 105 (ag/m ) occur in the industrial valley
3
and the lowest levels (35 ug/m ) in the residential areas out of the
valley. Estimates of emissions indicate that most traditional source
sectors - industrial processes, fuel combustion, solid waste disposal -
have had some reductions contributing to a 40 percent overall reduction
in inventoried emissions since 1969. These changes in emissions since
1969 have not been reflected in the air quality, which has seen increas-
ing TSP levels in 1973 and 1974. The general process weight rate regu-
lation under which process sources are controlled was found to be much
less stringent than those normally applied; however, th6 Puget Sound Air
Pollution Control Agency has the additional restriction on emissions of
0.03 grains/scf, one of the most stringent requirements in the country.
Fugitive dust due to vehicular traffic has also been cited as one of
22
-------
the problems in the industrial valley. The monitor with the highest
3
annual mean (105 |o.g/m ) is located approximately 25 feet back from a
heavily traveled road; a location 100 yards back from the road has
levels about 40 percent lower.
Washington, D.C. - While Washington has never had the major TSP problem
of other cities because of its nonindustrial nature, two sites in the
city exceeded the annual primary standard in 1974. There has been a
slight downward trend since the early 1960s as coal use decreased and
higher grade fuels were substituted, but two center-city sites con-
tinued to exceed the primary standard in 1974. Control regulations are
stringent, and large emission reductions have resulted from the closing
of incinerators and fuel switching; most sources seem to be in com-
pliance. The cause of nonattainment at the two center-city sites is a
mixture of urban activities. Demolition and construction associated
with urban renewal was prevalent for several years, and has been more
recently supplemented by construction of the METRO transit system.
In comparison with other cities, traffic would appear to be an expected
problem. However, any clear demonstration of that influence is pre-
cluded by the extreme heterogeneity of monitoring site locations, with
almost all monitors located very high, very remote from the traffic,
or both.
Particle Characterization
As mentioned in Section I, optical microscopic examination was undertaken
on selected hi-vol filters from each city. A total of 300 filters were
selected, from several representative sites in each city, and the results
were subjected to quality control checks. A summary of the filter analy-
ses and the results of the quality control are presented in Appendix B of
this volume, and a more detailed presentation is included in Volume II of
this report. The reader is advised, however, to consider the results of
the quality control program which suggests that the data may not be con-
strued to be more than semi-quantitative in nature.
23
-------
In Table 5, the microscopy results for each city have been averaged by
the generic type of material present. These composite results show the
highest percentage component to be mineral matter. Sources of mineral
matter include windblown soil, reentrained dust from streets, fugitive
dust from construction and demolition, and such industrial sources as
primary metals and mineral products industries and material storage piles.
Appendix B gives a more detailed breakout of these components. The higher
t
values for mineral matter in Denver and Oklahoma City support the theory
that fugitive dust sources are particularly important in arid areas.
It should be emphasized tha.t the optical microscope does not allow identifica-
tion of particles smaller than about 1 ym, so the analytical data is repre-
sentative of only the supermicron portion of the particulate. Based upon the
experience of the microscopists involved, an average of about 15 percent of
the mass is submicrometer in size and invisible to the analyst. Therefore, in
applying the component percentages, it was assumed that 15 percent of the mass
is invisible to the microscopist. Table 6 shows the results of this procedure
as applied to a categorization of the components by site type for all sites
studied in the 14 cities. Although the results are listed as average load-
ings, it should be pointed out that the filters selected for analysis were not
necessarily average or even typical. Within each study area, the selection of
samples was normally limited to filters from the 3 consecutive months of 1974
with the highest average TSP level and the 3 consecutive months of 1974 with
the lowest average TSP level; the filters selected from these periods were
usually from days with average to high TSP levels. This was done in an at-
tempt to discern differences in particle type at specific sites as a function
of TSP loading. The composited results may or may not be representative of
the annual average.
Particle Size The average particle size for each of the major visible
components is presented in Table 7. In general, mineral constituents
had the smallest particle size, and biological materials and rubber the
largest. The differences between sizes reported for the aggregate cate-
gories and the subcategories is not significant. Although the average
size of the mineral fraction, 8 ym, is consistent with the average sizes
24
-------
Table 5. COMPOSITE SUMMARY OF MICROSCOPIC ANALYSIS IN 14 CITIES, [ig/m
ho
City
Heavily
industrialized
Cleveland
Birmingham3
Philadelphia3
Baltimore3
St. Louis
Cincinnati
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miamib
San Francisco
All cities
Minerals
Average
51
66
64
69
75
51
36
81
60
64
70
88
79
52
65
Range
28-85
14-90
6-93
52-88
21-99
24-88
3-96
62-97
30-96
28-92
39-87
63-99
75-83
29-73
3-99
Combustion
products
Average
40
22
33
25
21
44
35
7
27
22
23
8
9
29
25
Range
10-70
2-86
6-89
11-61
1-79
9-84
8-78
1-19
1-62
4-68
5-49
1-31
7-12
10-50
1-89
Biological
material
Average
1
2
1
3
<1
1
16
1
3
1
5
3
3
Range
-------
Table 6. ESTIMATES OF AVERAGE FILTER LOADINGS BY SITE CLASSIFICATION3
Components
Mineral
Combustion products
Biological material
Misc. (mostly rubber)
Assumed < 1 (jjn
Total
3
Average loading, |ig/m
Commercial
64
27
2
9
. 19
120
Residential
51
19
3
5
14
92
Industrial
87
42
3
9
25
166
Undeveloped
66
6
<1
<1
13
86
Based on a total of 300 filters analyzed.
Table 7. COMPOSITE SUMMARY OF PARTICLE SIZE BY COMPONENTS3
Component
Minerals
Quartz
Calcite
Hematite
Combustion Products
Oil soot
Coal soot
Glassy fly ash
Biological Material
Pollen
Rubber
Average
size, jam
( 8)
11
9
3
( 5)
13
30
12
(24)
35
(43)
Average size
range , |im
-------
of the principal components, the average size reported for combustion
products, 5 jam, is noticeably lower than the average size of the individ-
ual components within that group. This is because the filters on which
sizing of the individual components was done are not necessarily the
same filters that comprise the aggregate combustion products group. For
some filters the size range was reported only for the combustion products
group as a whole because the individual components comprised less than
5 percent of the observed particulate. The average size of the biologi-
cal material is quite large but understandable in terms of its source
and aerodynamic shape. The very large average particle size reported
for rubber, however, is somewhat harder to understand. The generation
of large rubber particles by mechanical abrasion is easily understood,
but it is difficult to explain how such large particles can be trans-
ported over substantial horizontal or vertical distances.
Quality Control The reader is cautioned to review Appendix B regarding
the results of quality control procedures used in the microscopic examin-
ation. Briefly, it was found that the replicability of the results of
analyses of individual samples varied considerably with some results
quite far apart. Although the compositing of results from many filters
to obtain average results minimizes any systematic bias among microsco-
pists and laboratories, the results should not be construed to be more
than semi-quantitative in nature.
SUMMARY OF FACTORS AFFECTING ATTAINMENT
The purpose of the city case studies was to identify and study the various
factors, problems, and issues concerned with attaining the TSP standards
as they were experienced in each city. Since the 14 cities cover a broad
range of city characteristics and hence represent a variety of situations
with respect to TSP air quality and its determinants, analyses of the
factors in the various cities can be drawn together for an overall assess-
ment of the TSP attainment situation in the study cities and, by extrapo-
lation, throughout the nation.
27
-------
Following the analyses of the study cities, a number of factors were
identified as significant for standards attainment. Many of these had
been first identified in the preliminary literature review and were then
followed up in the city studies; a few others were identified in the
course of one or more of the city studies. The principal issues are
listed below, grouped into five major categories that will subsequently
provide a framework for the more detailed discussions in Section III.
Large Scale Influences
Large scale influences include those factors that dominate an area much
larger than the urban areas being studied. They include natural, trans-
ported, and secondary particulates. The differing influences of these
factors in various urban areas can cause significant differences in the
ability to control the local TSP problem. Their effect on urban levels
is generally estimated by measuring air quality in nonurban areas. The
average nonurban particulate level for the 14 study cities is between
3 3
25 and 30 ug/m ; however, values ranged from less than 15 p.g/m on the
3
west coast to roughly 35 |ig/m in the metropolitan northeast. The three
major large scale factors are described below:
Natural particulates ~ A major factor that can have signifi-
cant impact on standards attainment is the magnitude of the
natural TSP level in the incoming air masses. The west coast
cities benefit from having a very low (global) particulate
level, while cities in the central plains or the east have
additional continental contributions.
Transported primary particulates Cities in the eastern
metropolitan complex have the further problem of manmade
particulates being transported from neighboring urban areas
without space for adequate dilution. Therefore, these cities
have an impaired capability for managing their own air
quality.
Secondary particulates Levels of particulates such as sul-
fates, nitrates, ammonium, and some organic compounds, which
are generally believed to be formed as secondary particulates,
indicate again that cities in the east are receiving increased
TSP levels from other areas. These particulates are formed
both in transport and locally from gaseous emissions not tra-
ditionally controlled in TSP standard attainment strategies.
28
-------
Traditional Source Factors
Much of the problem of standards attainment in several of the 14 study
cities is attributable to emissions from those sources industrial pro-
cesses, fuel combustion, incineration that have traditionally been
considered subject to pollution control efforts. Cities with heavy in-
dustrial activity were found to have citywide TSP levels averaging from
3
10 to 50 yg/m above the levels in cities with little or no industry;
sites particularly close to heavy industrial activity averaged up to
3
25 [J.g/m higher than other industrial sites. Specific nonattainment
factors related to traditional source emissions include the following:
Industrial emissions In several cities, widespread in-
dustrial emissions were obviously the major share of the
overall urban problem; in other cities more isolated in-
dustrial emission problems were responsible for local non-
attainment. The most apparent problems were the steel
industry, with associated coking and foundry operations,
and the various minerals handling industries, such as
cement and asphalt plants and stone and gravel quarries.
Fugitive emissions Several sites, generally near indus-
trial areas, were significantly affected by fugitive emis-
sions from such sources as materials stockpiles, coal load-
ing operations, blast furnace slips, rock crushing and
loading, and similar uncontained industrial processes.
Fuel oil combustion In several cities there was concern
over the degree of impact from oil combustion. In coastal
cities, oil is frequently used for space heating, and the
smaller residential oil burners are typically not controlled
as pollution sources. In midwestern cities where coal is
used significantly, even very major oil-fired combustion units,
such as utility boilers, have sometimes been ignored as sources
because they are cleaner than equivalent coal-fired units.
Fuel use trends One positive influence on standards attain-
ment in several cities is the on-going trend toward cleaner
fuels, especially the shift from coal to gas in small units.
This is a continuation of a trend spanning many years, fueled
by factors of convenience and economic affluence.
Ineffective control programs - In at least one instance, a major
factor in failing to attain the standards is the overall lack
of any effective control program or enforcement effort.
29
-------
Lack of adequate time for control In some instances, failure
to meet the standards is due to having insufficient time since
the inception of control efforts for even an outstanding con-
trol program to cope with a major TSP problem.
Inadequate compliance determination In most of the study
areas, the process of verifying that a source is in com-
pliance and remains so appears to be somewhat haphazard.
Most agencies have less firm knowledge on such matters than
seems desirable
Inadequate regulations In some cases, relatively nonstringent
regulations inhibit meeting the standards by requiring less re-
duction in emissions than is necessary to meet the standards.
Inadequate data base In some cases, planning for air quality
management and standards attainment is inhibited by the lack
of an adequate data base. Usually involving emissions rather
than air quality data, this lack can be so extreme in some
cases that it must be construed as a major misunderstanding
as to the nature of the TSP problem being faced.
Nontraditional Factors
Even with the nonurban, large scale particulate concentrations and the
emissions from traditional sources taken into consideration, analysis of
many of the monitoring sites in the 14 study cities indicated that other
factors, not traditionally considered, were producing TSP levels that wert
3
typically 25 to 30 |_ig/m higher than expected. Such levels could often
be attributed to specific sources such as construction activity or local-
ized fugitive dust emissions, but in many cases the elevated concentra-
tions were simply the result of many uninventoried activities in urban
areas. Specific factors identified include:
Localized fugitive dust emissions A number of sites in
various cities were prevented from attaining the standard
at least in part by emissions from bare unvegetated lots,
unpaved parking areas and roadways, heavily traveled ex-
pressways, and similar local sources of entrained dust.
General urban activity - In most cities, ambient TSP levels
in the densest part of the city are higher than expected,
influencing attainment at a number of commercial and dense
residential sites in almost every city. The best immediate
presumption is that this excess concentration represents the
30
-------
combined, well-mixed influence of higher levels of traffic
and similar types of urban activity that tend to be greater
in the more dense central part of the city.
Construction activity - Several sites, often center-city
commercial sites, were hindered in meeting the standard
by dust entrained from construction sites of various types,
including urban renewal, small building construction, and
highway and subway construction.
In addition to the above factors, which are all related to sources of
particulate emissions, other factors were found to affect the real or
apparent TSP problem. As discussed in the selection of the cities, the
meteorology and climatology of a region can help to aggravate or amelio-
rate the TSP problem; the dispersion characteristics and precipitation
levels are the most prominent influences. The design of the monitoring
network configuration and the actual placement of monitors are also im-
portant in conceptualizing what the TSP situation is. The general find-
ings from the city studies for these factors are summarized below.
Meteorology and Climatology
Dispersion conditions The overall pollutant dispersion char-
acteristics can have significant effects in either direction;
in the southern mountain area, the attainment is clearly more
difficult because of adverse meteorology and topography, whereas
in the coastal and great plains areas the opposite is true.
Precipitation Frequent, significant precipitation is appar-
ently a help in attaining standards, while arid conditions are
a detriment.
Monitoring Considerations
Inappropriate sampler heights One of the more common prob-
lems with hi-vol network design is the question of consistent,
appropriate sampler heights. If the hi-vols are strikingly
higher or lower than typical, the recorded TSP levels will be
artificially decreased or elevated in comparison to other cities,
Similarly, if there are striking height differences within one
31
-------
urban area, there will be difficulties in adequately planning
for standards attainment and in accurately assessing progress.
Network design As the TSP levels vary from monitor to moni-
tor, and generally around the city, it is important to ensure
that those areas of maximum TSP concentration are monitored.
Some cities located monitors with the help of extensive model-
ing efforts, while others picked convenient locations.
Figure 3 is a schematic sketch of the interrelationship of these five
major groups of factors. It is intended as a minemonic device to em-
phasize that three of the factors are actual components of the particu-
late matter, while the other two are distorting influences.
TRADITIONAL
SOURCES
NONTRADITIONAL
SOURCES
LARGE-SCALE
INFLUENCES
MONITOR SITING
Figure 3. Schematic relationship among five major factors
32
-------
SECTION III
ASSESSMENT OF FACTORS AFFECTING ATTAINMENT
This section assesses the significance of the five major factors identified
with respect to the attainment or nonattainment of the ambient standards.
For ease of readership, much of the more extensive analytical and discus-
sion material has been collected into appendices, of which this section
can be considered a summary.
Since the air quality standard makes no distinction among particulates
from various sources, any source of particulates must be considered to be
contributing to standards nonattainment. The extent that the source or
source category contributes to the measured TSP levels at the point in
question is the same as its contribution to standards nonattainment.
Therefore, the assessment of the typical contributions from the various
factors can be seen as the determination of the contributions of the various
factors to an overall typical TSP concentration. This viewpoint, along
with a determination of how the various factors affect variations in TSP
levels, provides a useful discussion framework, which is used throughout
this section. The large number of monitoring sites (154) visited and
studied throughout the country provided a unique opportunity for developing
an understanding of the individual factors affecting the measured TSP le-
vels. From an analysis of the data, the city characteristics, and the mo-
nitoring site locations, the components of the TSP at different types of
sites were estimated.
The most obvious difference among sites is the nature of the neighborhood
in which they are located. Figure 4 shows the average of the annual
33
-------
ISO
1C
-x
o»
4.
§100
cc
»-
z
UJ
o
o
o
50
RESIDENTIAL
COMMERCIAL
INDUSTRIAL
Figure 4. Average TSP levels by neighborhood type
34
-------
concentrations at all the sites in the 14 study cities varied by the pri-
mary neighborhood classification of residential, commercial, and industrial.
Residential neighborhoods generally had TSP concentrations in the range
3
of 50 to 70 |o.g/m ; monitors at commercial sites recorded a wider range of
3
values, principally between 60 and 110 (ig/m ; and industrial neighborhoods
3
had the highest TSP levels , averaging between 80 and 150 |j.g/m .
To explain not only why these TSP levels are what they are but also why
they vary as they do, the analyses conducted in this study drew upon the
general findings from the city studies discussed in Section II. These
findings identified major topics of concern for assessment of the TSP
standards attainment problem on a nationwide basis: large scale consider-
ations, traditional sources, nontraditional sources, monitoring considera-
tions, and meteorology/climatology. This categorization structures the
following discussion, which summarizes the findings of the cross-city
analyses presented in the appendices.
BACKGROUND AND LARGE-SCALE CONSIDERATIONS
In reviewing the literature, and in considering the wide range of typical
air quality levels over different parts of the country. it is apparent
that there is a significant portion of TSP levels that varies over a geo-
graphical scale much larger than any one or even a few AQCRs. On an over-
all average basis, the levels of TSP measured in nonurban areas represent
the concentration of particulates in the air masses before they arrive in
an urban area; hence the wide range of nonurban values implies that com-
parable cities, located in different portions of the country, would require
different degrees of control to attain and maintain the ambient standards.
Therefore, quantitative air quality planning requires an understanding of
these levels and how and why they occur.
Terminology
Variously, particulates either measured in or presumed to originate in
nonurban areas beyond the jurisdiction of local or state agencies have
35
-------
previously been labelled in one way or another under the heading "back-
ground." The usage of the term, however, when examined carefully, is found
to entail a number of somewhat different concepts. The official concept
of background is defined in the Requirements for Preparation, Adoption,
and Submittal of Implementation Plans (40 CFR 51.13):
For purposes of developing a control strategy, background con-
centration shall be taken into consideration with respect to
particulate matter. As used in this subpart, 'background con-
centration1 is that portion of the measured ambient levels of
particulate matter that cannot be reduced by controlling emis-
sions from manmade sources; 'background concentration' shall be
determined by reference to measured ambient levels of partic-
ulate matter in nonurban areas.
Unfortunately, this definition reflects to some extent the prevalent varia-
tion in usage of the word "background," and hence contradictory usage of
the term continues. Under the first part of the last sentence, in which
background is defined as the uncontrollable portion of TSP, EPA regulations
provide for rollback calculations to ascertain the degree of emission
control necessary for the attainment of the air quality standards. The
rollback formula:
, , ambient - standard
reduction required = : r
^ ambient - background
tacitly assumes that the background level is a lower limit below which the
ambient concentration cannot be reduced. On the other hand, in air quality
modeling efforts, the background is frequently defined as the difference
between the measured concentrations and the calculated concentrations
which includes into background any source not included in the inventory
used. In still other circumstances, an agency may choose to regard as
"background" any TSP coming across the boundary into their jurisdiction,
regardless of whether that jurisdiction extends appropriately into nonurban
areas.
36
-------
Contradiction in the usage of the term background also arises from the
latter part of the above citation referring to the actual measurement of
background. Following that concept, the background levels most often used
in air quality reflect measurements of the ambient air quality in some re-
mote rural area; this may possibly be within the county, AQCR, or state,
depending upon the jurisdiction of the planning agency; or may be very far
away. Where possible, these measurements are made upwind of the prevailing
flow of air so as not to sample the particulate contribution of the area.
However, what these remote monitors are actually measuring is not neces-
sarily an uncontrollable or nonmanmade level of TSP. Rather, these mea-
surements simply reflect the particulate concentrations coming into the
urban area, including not only the natural, uncontrollable particulates
but also man's contribution to the nonurban particulate levels; these lat-
ter include emissions in rural areas, particles transported from distant
urban areas, and secondary particulates.
To avoid confusion with previous use of the term background, the term
nonurban particulate will be used in this report to refer to the latter
part of the EPA definition presented above; i.e., the TSP concentration
determined by measuring the ambient levels of particulate matter in non-
urban areas. The major components of nonurban particulates are listed
below:
Natural particulate TSP contributed solely by natural pro-
cesses and thus truly uncontrollable; includes a global con-
tribution, which includes both primary and secondary par-
ticles, and a continental contribution, primarily from wind
erosion of soil.
Transported particulate TSP levels that arise due to emis-
sions from man's activities in "upwind" urban and industrial
areas; includes both primary and secondary particulates trans-
ported from one area to another.
Local influences TSP measured at nonurban monitors that is
contributed by emissions in rural areas (dirt roads, agri-
cultural tilling, space heating, rural industrial sources)
and affected by the actual placement of the monitor (reen-
trained dust, small town activity).
37
-------
Because the term background has also connoted TSP that cannot be explained
through modeling efforts and TSP that cannot be controlled through control
of primary particulate sources, the excess of secondary particulates mea-
sured in urban areas over the levels measured in nonurban areas will also
be considered under this topic.
Natural Particulates
The emission and formation of particulates from natural sources result in
low concentrations of ambient particulates that have always existed, re-
gardless of man's influence. The most important of the natural sources of
particulates are soil and rock debris, forest fires, plants, volcanoes and
ocean salt spray; in addition, natural sources can emit gaseous pollutants
which can react to form particulates. As shown by the estimates in Table 8,
particulate emissions from natural sources and particulate formation from
naturally occurring precursors far outweigh the contribution of man-made
sources on a global scale; however, man-made emissions are more important
for standards attainment because they are gathered into small areas.
Table 8. ESTIMATES OF PARTICLES SMALLER THAN 20 [w RADIUS EMITTED INTO
OR FORMED IN THE ATMOSPHERE (106 metric tons/year)1
Man-made
Particles from direct emissions
Particles formed from gaseous emissions
Sulfate from S02
Nitrate from NOX
Organics from hydrocarbons
Natural
Soil and rock debris (wind erosion)
Forest fires and slash -burning debris
Sea salt
Volcanic debris
Particles formed from gaseous emissions
Sulfate from H2S
Ammonium salts from NH^
Nitrate from NOX
Organics from hydrocarbons
Total
185 - 415
773 - 2200
958 - 2615
10 - 90
130 - 200
30 - 35
15 - 90
100 - 500
3 - 150
300
25 - 150
130 - 200
80 - 270
60 - 430
75 - 200
38
-------
Of the natural sources of particulate emissions, sea salt is probably the
largest emission source, but its greatest effect on TSP concentration
occurs over the oceans; its contribution to TSP levels extends over land
only a short distance. Over land, wind-entrained soil dust is the largest
direct source of particulate emissions. Gaseous emissions from natural
sources are scavenged through various chemical reactions and result in the
production of significant quantities of aerosol materials over a broad area.
Volcanic emissions can vary greatly from year to year but usually do not
contribute a large proportion of the natural particulate emissions. The
contribution of forest fires can only be estimated roughly at present;
though it appears small in Table 8, it may be considerably more important
with respect to air quality.since such fires are frequently adjacent to
urban areas. Pollens, spores, and bacteria are an insignificant fraction
of the total emissions.
The primary distinction between the sources of natural particulates with
respect to TSP levels is simply their location and to some extent the
effective emission height and particle size. Generally, the natural par-
ticulate levels can be thought of as contributions from two types of
sources: global sources and continental sources. Global particulates
arise from the heated emissions from volcanoes and, to a lesser extent,
forest fires, and from secondary particulates formed from natural gaseous
emissions; these emissions are characteristically in the submicron range.
Sea salt is also often referred to as a global particulate. Continental
sources are wind erosion of rocks and soil; pollens and spores; and (over-
lapping global particulates) forest fires and secondary particulates,
especially hydrocarbons from plant exudations.
The contribution from global sources is considered relatively constant across
the North American continent in the range of 1 to 5 ug/m3; for TSP strategy
planning, the contribution from continental sources is much more important
than that from global sources. Continental particulate levels are much higher
and they vary across the continent. For example, in the Great Plains region
of the United States, wind erosion of soil is estimated to produce a larger
mass of particulates than other sources in the region and can result in high
dust concentrations over large areas. Most duststorms occur in the spring
39
-------
but air pollution from duststonns can be a problem in other seasons as well.
The rainfall and soil erosiveness of an area are also influential with
respect to the frequency and severity of duststorms.
Transported Particulates
The term transport refers to the movement of particulates over a greater
distance than normally considered for dispersion modeling used for air
quality planning. Transported particulates include primary particles,
which are emitted directly into the air and secondary particles, which are
formed from reactions of gases in the atmosphere. While natural particu-
lates can also be transported considerable distances, this discussion of
transported particulates is meant to center on particulates that originate
from man's activities.
Transported Primary Particulates The transport of primary particulates
may be divided into two classes short-range and long-range. Long-range
transport occurs when the particulates are mixed into an air mass and
travel several hundred kilometers or more without any removal mechanisms
such as washout or rainout. This phenomenon of transport is directly re-
lated to meteorology because it requires a stable air mass moving across
the country. An interesting case study of the transport of a particular
air mass is described in Appendix C. This study demonstrates how long-
range transport can produce abnormally high values of particulates over
short periods, causing violations of the 24-hour standards. The degree
to which it affects annual means is related to the frequency of such
occurrences.
Short-range transport refers to the transport of particulates over less
distance, ranging from a few to about 100 kilometers. It is primarily
concerned with the movement of particulates across planning area bound-
aries; i.e., from the jurisdiction of one air pollution control agency
into the jurisdiction of another.
40
-------
Obviously, short-range transport of primary particulates is of principal
concern in areas which have adjoining urban areas with insufficient rural
areas between to allow for removal or dispersion of the pollutants. Such
is the situation in the northeast where the air may sweep up past Balti-
more, Washington, and Philadelphia into New York and up to Boston. In
the study of Providence under this effort, high values of particulates,
even in the less developed areas of the AQCR, were most often associated
with winds from the direction of New York City. Meteorology is a com-
plicating factor in this analysis because of variations in windspeed and
rainfall with changing wind direction. However, it is likely that such
conditions do exist.
Over even a smaller scale, transported primary particulates are important
whenever air crosses from one region that is completely autonomous into
another area that has separate control. Such a situation was found in
Philadelphia in the course of this study. The Philadelphia Air Manage-
ment Services has complete responsibility in Philadelphia County, while
the Commonwealth of Pennsylvania Department of Environmental Resources
has responsibility for the counties surrounding Philadelphia. Since
the entire County of Philadelphia can be considered urbanized, the most
remote site in the network is one located in a residential neighborhood
near the border of Philadelphia and Montgomery Counties. This site had
3
an annual mean of 59 (ig/m in 1974, implying that there were virtually
no means of avoiding violating at least the secondary annual standard as
the air passed over the city. Another site, in a more industrialized
corner of the county but also near the industrialized areas of Delaware
County in Pennsylvania and Gloucester County in New Jersey, had an annual
3
mean of 94 |ag/m . The value at this site is especially important since
it is in the southwest corner of Philadelphia County, the direction from
which most of the air crossing the county would come.
41
-------
Transported Secondary Particulate - As mentioned previously, secondary
particulates are the products of chemical reactions occurring in the at-
mosphere. They can initiate in the gas phase or as a result of reactions
between gases and already existing particles. They are a major source of
the ubiquitous Aitken nuclei, or homogenous nucleation centers, that are
essential for most of the condensation processes that take place in the
atmosphere. They are also a prime component of urban smog.
Composition The main ingredients in the formation of secondary particu-
lates are sunlight and gases such as sulfur dioxide, ammonia, nitric oxide,
water vapor, and hydrocarbons, which enter the atmosphere from both natural
and manmade sources. Secondary particulates range in size from molecular
clusters with diameters on the order of 0.005 urn to particles with diam-
2-4
eters as large as several micrometers. Field studies of urban aerosols
have shown that the highest concentration of secondary particulates is
usually in the range 0.01 to 1.0 ym. The concentration of particles in this
size range can vary directly with intensity of sunlight and concentration
of ozone.
The principal factors governing distribution by size are the respective
rates of particulate formation and removal. The smallest particles,
which are created constantly during the daylight hours, coagulate into
larger particles. The overall life cycle of secondary particulates is
difficult to determine; estimates range from 1 week to 40 days. 5 In the
end, the particles are either removed from the atmosphere by precipita-
tion or dry deposition. During this period, however, they may be
transported vast distances from the source of the gaseous precursors.
Found in both urban and rural areas, secondary particulates are in general
composed of three types of chemical compounds sulfates, organics, and
nitrates which are briefly discussed below.
Sulfates. Sulfates are ubiquitous. A large fraction of the
global aerosol is ammonium sulfate (NH^SO^.. Sulfates such
as sulfuric acid (l^SO) , which is found in most urban aerosols,
42
-------
result from the reaction of 863 and water. The sulfate salts,
such as PbS04, in turn, derive from the reactions of compounds
such as ammonia or metallic oxides with sulfuric acid droplets.
Qrganics. The second major constituent of secondary participates
is produced by the reaction of hydrocarbons with oxidants (e.g.,
N02> 03) in the atmosphere to produce peroxide radicals. Through
a series of chain reactions, these radicals eventually form large
organic molecules which condense to form droplets or solid particles.
Primary sources of hydrocarbons in urban areas are automobile ex-
haust and industrial effluents. In some rural areas, hydrocarbon
emissions from natural sources may be significant.
Nitrates. Nitrogen oxides emitted into the air can be oxidized
and react with water to form nitric acid in either vapor or drop-
let form. From this form, nitrates are created through reactions
with gaseous or solid species.
Air quality impact Secondary particulates can occur both over a long
period of transport and apparently also fairly quickly in an urban area.
As with primary particulates, secondary particulates can be important
over both short-range and long-range transport. The long-range transport
study in Appendix C found sulfate levels three times higher than normal;
shorter transport of secondary particulates is known to contribute to
high levels in the northeast.
Annual levels of sulfates and nitrates for nonurban areas near each of
the 14 study cities are given in Table 9 along with the estimated total
nonurban levels. Average values for each of the industrialization clas-
sifications for the cities in the east (of the Mississippi River) versus
the west, and for the cities in the north (heating degree days more than
4000/year) versus the south, have been calculated. While the heavily in-
dustrialized cities have the highest nonurban levels of secondary partic-
ulates, this division apparently results from the geographic location of
the cities. The h.ighest levels within each category of industrialization
occur in cities east of the Mississippi, where the density of industrialization
43
-------
Table 9. NONURBAN LEVELS IN THE 14 STUDY
CITIES3
Cities
Heavily
industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Average
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Average
Lightly
industrialized
Washington , D . C .
Oklahoma City
Miami
San Francisco
Average
East
West
North
South
Nonurban levels,
ug/m3
504
10
7
10
8
6
11
8.7
6
2
3
7
4.5
8
3
5
2
4.5
8
3.2
7.2
4.6
N03
1
1
1
1
1.
1
1.0
2
0
0
1
0.8
1
1
1
0
0.76
1.11
0.4
0.78
1.0
Total
30
30
35
35
25
35
31.7
35
20
15
30
25.0
30
25
25
15
23.8
27.8
20.0
28.3
26.0
aBest estimate averages based upon
composite values of available NASN data,
44
-------
is greatest, and also in the northern cities, where space heating is more
important. In the two dichotomous breakdowns - north versus south, east
versus west - the impact of geographic location on the nonurban values is
obvious, A logical extrapolation of these data would be that nonurban
secondary particulates are highest in the northeast and lowest in the south-
west; such findings have been noted in other studies.
Other Factors
As mentioned previously, two other factors are important when considering
nonurban TSP levels and the use of such levels in air quality planning.
Nonurban levels can reflect local influences that are not of concern when
planning control strategies for an urban area because the particulate mat-
ter is not actually carried into the urban area. At the same time, se-
condary particulates formed locally in an urban area are not accounted
for in nonurban monitoring, and therefore, these urban secondary excess
levels are not appropriately incorporated in traditional air quality
planning.
Local Rural Influences - Local influences on nonurban levels can be tradi-
tional emissions such as from local space heating and rural industrial
activity; but, more likely, these local influences are artifacts of the
monitor placement so that the TSP levels are subject to reentrainment
from dirt roads, agricultural tilling, or natural wind erosion. Similarly,
TSP levels reported as nonurban levels may be from monitors located in
small, rural towns; these monitors would also be measuring the particulates
generated by man's activities (traditional and nontraditional sources) in
the town.
Emissions from dirt roads in the major counties in the study AQCRs are
discussed later under the topic of Nontraditional Sources. That analysis
suggests that the current inventories for dirt roads provides an exaggerated
picture of the importance of unpaved rural roads to the air quality in an
urban area. The current inventories suggest that emission levels due to
45
-------
dirt roads in urban counties are usually much higher than the total tradi-
tional source emission levels in urban areas. Obviously the air quality
impact from these sources is not equivalent to that from traditional sources;
otherwise, the TSP concentrations in rural areas would be expected to reach
or exceed those in urban areas.
Similar findings apply to the inventoried emissions for agricultural tillings.
While the emissions from tilling were never as great as those from dirt roads
in the 14 AQCR's studied, in some areas the levels are quite high; e.g., in
the San Francisco Bay Area AQCR, emissions due to tilling are inventoried at
almost 190,000 tons per year. Since tilling does not occur throughout the
year but only at certain seasons, any impact from tilling would be expected
to be short-term, perhaps causing elevated levels for a month at a time.
Urban Secondary Excess - Secondary particulates formed in the urban area
as a result of local emission sources have been ignored in control strate-
gies applied to air quality planning; i.e., control of TSP concentrations
has been approached solely by reducing directly emitted particulates. Pre-
sumably, the amount of secondary particulate formed is related to the
amount of precursors, so that higher levels of secondary particulates may
be expected to be found in the more industrialized regions. For sulfates,
the amount of space heating with fuels high in sulfur may be important in
winter.
Because of the different control approaches that are open to an individual
air pollution control agency, it is helpful to separate out the secondary
particulates formed within the jurisdiction of the agency from those formed
in transport to the region. The annual levels of sulfates and nitrates for
urban and nonurban areas in each of the 14 study cities are given in Table
10 and averages have been calculated for the same breakdowns given in
Table 9. The data in Table 10 demonstrate that air entering the city is
the predominant factor in determining the sulfate levels in a city but that
nitrate levels can increase by a factor of three in the city. Since
46
-------
Table 10. COMPARISON OF URBAN AND NONURBAN LEVELS OF SULFATES
AND NITRATES IN THE 14 STUDY CITIES3
Cities
Heavily
industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Average
Moderately
industrialized
Cincinnati
Chattanooga
Denver
Seattle
Providence
Average
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miami
San Francisco
Average
East
West
North
South
Urban levels
804
10
14
14
10
12
12
12
11
5
7
9
8.8
12
3
5
5
6.25
10.78
6.4
10.1
7.6
NC-3
3
3
4
3
3
3.2
3
2
3
2
2
2.4
3
2
1
2
2
2.67
2.4
2.89
2.0
Total
13
17
18
13
15
15.2
15
13
8
9
11
11.2
15
5
6
7
8.25
13.44
8.8
13.0
9.6
Nonurban levels
504
10
7
10
8
6
8.2
11
6
2
3
7
5.8
8
3
5
2
4.5
8
3.2
7.2
4.6
N03
1
1
1
1
1
1
1
2
0
0
1
0.8
1
1
1
0
0.76
1.11
0.4
),78
1.0
Total
11
8
11
9
7
9.2
12
8
2
3
8
6.6
9
4
6
2
5.25
9.11
3.6
8.0
5.6
Urban excess
S04
0
7
4
2
6
3.8
1
5
3
4
2
3
4
0
0
3
1.75
2.78
3.2
2.89
3.0
NC-3
2
2
3
2
2
2.2
2
0
3
2
1
1.6
2
1
0
2
1.25
1.56
2.0
2.11
1.0
Total
2
9
7
4
8
6.0
3
5
6
6
3
4.6
6
1
0
5
3.0
4.33
5.2
5.0
4.0
Best estimate averages based upon composite values of available NASN data
47
-------
sulfates are the predominant secondary pollutant, the average increase in
secondary particulates is on the order of 60 to 70 percent. As may be
expected, the largest increases (urban excesses) are seen in the highly
industrialized areas where secondary particulate levels inside the city
o
average 6 ug/m above those in the nonurban setting; lightly industrialized
areas have half of this increase. On the whole, cities may have concentra-
3
tions elevated 5 to 15 yg/m due solely to sulfates and nitrates. Two
important points concerning Tables 9 and 10 are that some sulfates may
be directly emitted as primary particulate. and secondary organics may also
be significant, but adequate measurements were not available.
National Assessment
As has been shown above, particulate levels exist which are beyond the
control of individual state and local air pollution control agencies.
These levels of TSP enter the jurisdiction of an agency along, with the
air mass that is carrying them. The particulates entering a region may
be the result of a facility a few miles upwind of the jurisdiction, of
another city 50 km away, of sources generating precursors to secondary
pollutants hundreds of miles from the region, or of natural nonurban sources
such as sea salt, pollen, and wind-driven dust. While the smaller scale
transport problems may often be adequately handled by cooperation between
adjoining state and local agencies or by EPA regional office intervention,
the larger scale influences either cannot be controlled or need direction
and planning on the national level. These larger scale problems are
addressed below.
The concentration of particulates in rural areas of the country has been
monitored for many years as part of the National Air Surveillance Network
(NASN). Figure 5 presents composites of values reported at various nonurban
NASN sites from 1970 through 1973. The data in this figure indicate the
wide range of annual means being reported. In addition to the problem of
transported primary and secondary particulates, some of the range in TSP
values is expected to be due to inconsistent and, in some cases, poor
48
-------
-p-
vo
Figure 5. Composite annual geometric mean TSP levels at nonurban NASN
sites from 1970 through 1973 (|ag/m3)
-------
monitor siting. For instance, the high values reported at the monitors
on the west coast (California and Oregon), which would be expected to have
values half of those reported based upon global and continental particulate
levels, are recorded by monitors located on the ground. As discussed later
under reentrainment, such low siting is expected to result in highly exag-
o
gerated values. The high value of 58 ug/m reported in Indiana is due to
a nearby power plant influencing the levels in the rural area.
The impact that transported secondary particulates have on these nonurban
sites can be seen in Figure 6, which presents the nonurban sulfate and
nitrate levels for 1974. Almost no nitrates and very little sulfates are
found in the western part of the country, yet secondary particulates con-
3
sistently add more than 5 yg/m to the TSP loading at nonurban sites east
of the Mississippi River. These data reiterate the findings of the 14 city
case studies: secondary particulates in nonurban areas can add significantly
to the TSP burden on cities in the east arid northeast.
The transported primary particulates cannot be so easily addressed on the
national scale for several reasons. One reason is that the density of non-
urban NASN monitors is too low to provide adequate information on the change
of TSP levels between major urban areas. Such an analysis would have to
include monitors operated by state and local agencies in order to be close
to the density of monitors needed, and these data were not available on a
nationwide or even regionwide basis. Therefore, the conclusions on trans-
ported primary particulates must stand solely on the earlier discussion and
the analysis provided in Appendix C.
Another problem already mentioned, monitor height, not only means that
transport cannot be accurately determined but also brings many of the
reported nonurban values into question. Most of the monitors in the non-
urban network apparently are located at heights below 10 feet, with many
of them in the 3 to 6 foot range. Such monitors are as likely to be in-
fluenced by nearby disturbances as by any particulates being transported
into the region.
50
-------
NORTH DAKOTA I ~
I UINNCSOTA
HAWAII
O-SULFATE
Q-NITRATE (>lp«/m3)
Figure 6. Annual geometric mean sulfate and nitrate levels at
nonurban NA.SN sites 1974
-------
Despite these problems, some estimate of the variations in the nonurban
levels can be made from the findings of this study. Figure 7 provides a
conceptualized diagram of the contributions to nonurban levels as one moves
across the country. Global particulates are assumed to be constant across
the country. Continental particulate is lowest on the west coast, where
the land area has not had a chance to contribute significantly, and highest
in the midwest due to the high winds and more arid conditions. The
occurrence of some off-the-ocean air masses is felt to bring down the con-
tinental contribution a little in the east.
Transported secondary particulates contribute only a few micrograms west of
the Mississippi River. However, in the east, nonurban levels of secondary
3
particulates are averaging around 10 |_ig/m . Although the long-range trans-
port case study presented in Appendix C is for Oklahoma City, the impact of
transported primary particulates on annual TSP levels in the west and mid-
west is felt to be minimal. In the east where cities are more concentrated,
3
transported primary particulates are more serious, averaging about 5 yg/m
on an annual basis throughout the east but potentially much higher in the
congested northeast.
On top of all these contributions, another few micrograms have been added
to reflect the local influences occurring on a large scale basis in the
rural areas (space heating, traffic, agricultural activity- etc.). This
additional level does not include immediate impacts from nearby sources
(power plants, roads) or from possible reentrained dust due to low monitor
height.
PARTICULATES FROM TRADITIONAL SOURCES
The most obvious of the many factors affecting attainment of the TS.P
standards are the particulate emissions from three major categories of
pollution sources fuel combustion, industrial processes, and solid waste
disposal operations. These three source categories have long been con-
sidered significant pollution problems, and have traditionally been the
52
-------
Ui
u>
Q.
CO
40
ro 30
e
^v
0>
2 20
o
z
o
o
10
LOCAL-
TRANSPORTED
SECONDARY
CONTINENTAL
GLOBAL
WEST
LOCAL
TRANSPORTED
SECONDARY
CONTINENTAL
GLOBAL
MIDWEST
LOCAL
TRANSPORTED
PRIMARY
TRANSPORTED
SECONDARY
CONTINENTAL
GLOBAL
EAST
Figure 7. Average estimated contributions to nonurban
levels in the East, Midwest, West
-------
primary concern of air pollution control efforts; consequently, they have
been labeled "traditional sources" for purposes of this study. This sec-
tion summarizes the assessment made of the significance of traditional
source emissions as a factor in the nonattainment of the TSP standards;
Appendix D presents summaries of the data assembled and the analyses made
to develop the assessment.
In general, the impact of traditional sources on nonattainment depends
very heavily on the nature of the urban area in question. The 14 case
study areas included both cities where traditional sources totally dominate
the picture and cities where they are not now, and probably never were, a
major share of the problem. Based on analysis of aggregate emission in-
ventories, emission densities, and compliance trends, it is possible to
summarize the impact of traditional sources in the 14 case study areas
as follows:
I. Three areas, all heavily industrialized, still have a
major problem with traditional sources;
II. Three areas have reduced emissions from traditional
sources to the point where they are no longer totally
dominant, although continuing further reduction and
on-going surveillance is still required;
III. Four areas have reduced formerly moderate levels of
traditional source emissions (mostly from fuel use
for heating and light industry) to near insignificance;
IV. In four areas, traditional sources probably never
were a serious problem.
While these specific proportions are not necessarily reflected in the
overall national picture, there are certainly a number of urban areas
throughout the country in each of these categories.
A comparison was also made of the emission parameters leading.to this
classification with air quality levels in the various urban areas, expressed
as city wide average TSP concentrations. With adjustments made for differ-
ing nonurban levels and secondary particulates, essentially no difference
54
-------
in typical air quality between the third and fourth categories listed above
3
was indicated, with the average for both groups being 35 ug/m above non-
urban levels (Denver was excluded as an anomaly). This approximates the
contribution of nontraditional sources as discussed later. The three heavily-
industrialized cities still dominated by traditional source emissions (Cate-
3
gory I) had an average city wide TSP level of 66 |jig/m above nonurban, sug-
3
gesting that about 30 ^g/m is the maximum potential city wide reduction,
even with very stringent traditional source control. The three cities
(Category II) that have made significant but as yet incomplete traditional
o
source reductions averaged 48 yg/m above nonurban levels, suggesting that
o
there is still 10 to 15 yg/m of traditional source influence on city wide
averages which could be reduced somewhat with further control of traditional
sources. Figure 8 displays the relationship between emission density and
air quality, and illustrates the clustering of the study cities into the
categories. It is important to remind the reader that these results must
be interpreted as only semi-quantitative and extrapolated with care. Al-
though it is believed they provide a good national aggregate assessment of
the role of traditional sources, this analysis does average over different
neighborhoods and over vastly different cities with significaly different
sources and control programs, and meteorology.
With respect to the three major categories of traditional sources - fuel
combustion, industrial processes, and solid waste disposal - the relative
contributions also varied significantly with the nature of the different
urban areas. The fuel combustion contribution to inventoried emissions
ranged from less than 20 percent in clean-fuel, industrialized areas to
well over 90 percent in totally nonindustrial areas, with industrial
processes accounting for most of the balance. Solid waste disposal emis-
sions were generally less than 5 percent.
Fuel Combustion
The magnitude of fuel combustion emissions depends primarily on the amounts
and types of fuels burned, but the degree to which they are or can be con-
trolled primarily depends on the installation size point sources and
55
-------
0 100 200 300 400 500 600
TRADITIONAL SOURCE EMISSION DENSITY, tons/year/sq. mile
Figure 8. Relationship between city wide average TSP levels and
traditional source emission density
56
-------
area sources - and sometimes on the type of source - electric power, in-
dustrial or residential. The overall pattern of the fuel combustion emis-
sions was generally as would be expected, with greater emissions in areas
where coal and heavy oil are more prevalent, and in areas where major elec-
tric power or industrial combustion sources remain uncontrolled. Substan-
tial further reductions in particulates from these sources are expected
under present regulatory plans. However, if sources are required to con-
tinue burning coal or are switched back to coal, these reductions may not
take place. There is a significant range in the stringency of emission
regulations applied to combustion sources, so that in many areas there is
room for further reductions by tightening the standards regardless of the
fuel being used.
One particular aspect of fuel combustion emissions control to which the
study sought an answer is the question of residential space heating. The
use of coal in residential units has declined almost to the point of in-
significance, but in coastal cities the use of oil (rather than gas) is
common, and is likely to remain so. Since small oil burners are generally
controlled only through visible emissions enforcement, if at all, the
degree to which they might contribute to the TSP problem is an important
open question. In the cities selected for the study, it proved impossible
to separate the extensive residential use of oil from other fuel use and
industrial sources by means of the air quality and emission data analysis
techniques primarily used. However, it did prove possible to make a rough
estimate of the impact from heating oil based on the microscopic analysis
of hi-vol filters from the various cities. Composited results for each
city, while subject to significant caution in interpretation, did indicate
elevated levels of oil soot in those cities Providence, Washington,
Seattle, and Baltimore where they would logically be expected. An
approximate comparison of these results with those in the other cities
3
suggests a contribution to TSP levels of 5 to 10 |j.g/m . This is not a
3
small portion of the typical urban levels of 75 to 100 |ig/m ; it is a
3
significantly larger portion of the 30 to 35 p.g/m "working range" between
typical nonurban levels and the secondary standard.
57
-------
Industrial Processes
The second major category of traditionally considered pollution sources
are industrial process losses, as distinguished from emissions from in-
dustrial fuel combustion emissions. Process emissions are divided into
stack emissions and fugitive emissions, the latter being those indirect
emissions from doors, windows, etc., material storage piles, or other
outside activity on the plant property. Historically, primary concern
has been directed at stack emissions, which are more easily identified,
quantified, and controlled, and have, in the past, been the major sources
of particulate emissions from industry.
Stack Emissions The degree of the process emission problem is dependent
on the industrialization of the area in question. In the heavy industrial
cities where control of process sources is still being pursued, these
sources, along with industrial fuel use, dominate the air pollution picture.
In those industrial cities where emissions from process sources have
been generally controlled, they tend to remain roughly half the total
inventory. The process weight regulations concerning stack emissions from
process losses are not amenable to significant tightening, and further
reductions of inventoried emissions will generally need to come from en-
forcement of existing regulations, adoption of tighter regulations for
specific types of sources, and control of fugitive emissions. The industry
categories that continue to pose the greatest stack emissions problem are
the primary metals and minerals processing industries.
Fugitive Emissions - Fugitive industrial emissions have been traditionally
recognized, but only minor control efforts have been pursued to date. In
many operations, especially where dry materials handling is prominent,
emissions are generated in processes both inside and outside of plant
facilities, but on plant property, which are either ignored or insuffi-
ciently controlled so that significant particulate emissions are generated.
For industrial processes that operate outdoors, such as coke ovens and rock
crushing operations at quarries, these pollutants are directly emitted to
58
-------
the ambient air. Even when such processes are enclosed, the pollutants
emitted into the working environment may escape to the atmosphere through
windows, doors, roof ventilators, or even unsealed cracks in walls. In
either case, those pollutants that enter the outside ambient air have been
defined as fugitive emissions.
Fugitive emissions result from a wide variety of circumstances, including
poor operation or maintenance of process equipment. For example, fugitive
emissions can be the result of leakage from warped doors on coke ovens as
well as the oven charging operation itself. Storage piles and handling
operations for sand and gravel, coal, grain, and other materials that are
kept in the open can become fugitive sources when a strong wind blows over
them. Similarly, dirt and gravel parking lots and roadways on industrial
property can become major sources of particulates due to either wind
erosion or traffic.
Fugitive emissions have generally been assumed to be small in comparison
to stack emissions. However, with stack emissions coming under controls
that may provide up to 99 percent reduction, the relative importance of
fugitive emissions has been growing, and they may now comprise
a significant portion of nationwide emissions. For example, EPA has
estimated that total fugitive emissions of particulate from electric arc
furnace charging can be 5 to 50 times the amount of the stack emissions
emitted downstream of the control device.
Even if the quantity of fugitive emissions from a process is small in
comparison with the stack emissions, the low height at which they are
typically emitted means that very'little dilution occurs and fairly high
ambient levels are created. Consequently, even though adequate emission
estimates are lacking, fugitive emissions appear to be a significant and
increasing problem. A very rough estimate based on comparing TSP levels
at various monitoring sites (see Appendix D) suggests their aggregate
o
impact in industrial neighborhoods is typically on the order of 25 yg/m .
59
-------
Reduction of fugitive emissions will likely require a new approach to
regulatory control of such emissions. Currently, regulations for control
of fugitive emissions are of three general types: nonspecific nuisance
regulations, quantitative property-line regulations, and regulations that
prescribe specific control measures in specific circumstances. The
majority of regulations in the country are of the first type, defining
dust as a nuisance and often requiring "reasonable precautions" to pre-
vent emissions. While flexible and capable of being strong enforcement
tools, such regulations have not in fact proven effective on an overall
basis. The other two types can be more effective, though clearly not
without serious enforcement efforts. Property-line regulations in par-
ticular require enforcement and measurement techniques that are even
more difficult than those required for stack emission sources. Although
both alternative types are apparently somewhat better than nuisance reg-
ulations, there were only limited areas where they are used, and no areas
where an extensive, effective control effort was underway.
Solid Waste Disposal
Of the several methods of solid waste disposal, incineration and open
burning have traditionally been the most common in urban areas and, there-
fore, the most significant sources of particulate emissions. Under pres-
sure from pollution regulations, however, the larger point sources of
solid waste disposal emissions, municipal incinerators and large indus-
trial installations have been controlled or replaced, while the smaller
residential and commercial incinerators and open burning in dumps are
typically tightly regulated and often banned. With the continuing trend
toward landfills, recycling, and the use of combustible rubbish as a
fuel supplement, solid waste disposal is expected to continue to decline
in significance as a factor in attaining the NAAQS.
60
-------
Surveillance. Compliance and Enforcement Programs
An integral part of the impact that emissions from traditional sources
may have on standards attainment is involved with the nature and effec-
tiveness of the pollution control effort applied to them. Since fuel
combustion, industrial processes and solid waste disposal have long been
viewed as important sources of particulate emissions, maintaining sur-
veillance over these traditional sources and enforcing regulations concern-
ing them have been major activities of many control agencies since their
inception. The nature of the control programs varies significantly, in-
volving various combinations of source registration, permit systems, in-
spections, and so on.
The achievements of surveillance and enforcement programs depend on the
matching of enforcement activities to the nature of the particulate emis-
sion problem. Among the 14 cities, the largest actual reductions in
emissions from traditional sources have been achieved in those cities
where surveillance and enforcement programs are comprehensive and vigorous.
The activities of such programs generally included the following: constant
surveillance and patrols, frequent inspections of problem sources, a gen-
eral knowledge of all of the traditional sources, rigorous compliance
determination, prompt action when a violation or upset occurs, issuance
of compliance orders that are strict yet reasonably attainable in the
opinion of an appeals board, and strict enforcement of compliance schedules,
The stringency of the regulations being applied is obviously important in
determining the reductions in particulate emissions actually obtained.
Study findings indicate, however, that somewhat more important are the
enforceability of the regulations, the strictness of the enforcement, and
especially the manner in which compliance is determined. The enforce-
ability of regulations affects the ease and speed with which emissions
are controlled, and is influenced by the types of regulations in effect.
61
-------
For example, in dealing with numerous small incinerators, a standard
specifying certain types of equipment is much easier to enforce than an
emission standard which requires monitoring; in dealing with coke ovens,
an efficient standard is one that specifies maintenance and operating
conditions; in dealing with fugitive emissions, a source-specific regula-
tion is more effective than a general nuisance regulation. The enforce-
ability of regulations is also related to the institutional channels through
which any hearings and appeals proceed. Enforcement is more effective and
more efficient when control activities and enforcement proceedings are
conducted by the same governmental level or at least by well-coordinated
and geographically proximate agencies.
Another important concern is the matter of compliance determination. It
is not at all clear that reports of full or near compliance actually mean
that all or most traditional sources are in compliance with the regulations.
While it was not a major purpose of the present study, some understanding
of the methods used for compliance determination was obtained during dis-
cussions of overall compliance status. Only a few of the agencies conduct
or require actual stack tests and then not on a routine basis. More
commonly, compliance determination is done on the basis of walk-through
inspections and theoretical calculations based on process loads, emission
factors, control efficiency specifications, and similar data. While this
type of compliance determination is appropriate for some sources and control
measures when done by well-trained agency personnel, it is equally inap-
propriate for other, more complex sources with untried control technology,
particularly when performed by relatively inexperienced agency personnel.
Air Quality Impact
A primary objective of the study was to develop an understanding of the
impact that traditional sources have on TSP levels. Specifically, it is
important to place traditional sources in a proper perspective with res-
pect to the problem of standards attainment. The type of analysis based
62
-------
on citywide levels that was presented in Figure 8 is adequate for a broad,
general perspective; however, a more careful analysis involving individual
monitoring sites was also undertaken in order to provide a more detailed
and comprehensive picture. This effort, involving over 150 hi-vol sites,
considered in detail the nature of the site neighborhoods and the air quality
levels recorded, and has provided an overall perspective which is used
throughout this section to structure the summary discussion. Figure 9
indicates the quantitative impacts in various types of neighborhoods
estimated to result from traditional source emissions. The figure estimates
the impact at residential, commercial, and industrial sites in cities with
two different levels of traditional source prominence, following the
grouping presented in Figure 8. The higher portion of the bars represent
the three cities where traditional sources are still dominant (Category I) ,
and the lower portion the cities where significant control has taken place
(Category II) ; in the other two categories, where traditional sources are
largely absent or controlled, no apparent impact was seen except in the
industrial area itself. With respect to other parameters, such as meteorol-
ogy, the estimates should be viewed as representing a hypothetical average
city. The traditional sources in a heavily industrialized, not yet con-
3 3
trolled city add 10 jag/m at a typical residential site, 20 ug/m at a com-
3
merical site, and about 60 ^g/m at an industrial site. Roughly about
3
20 ng/m of the latter may be attributed to fugitive sources.
PARTICULATES FROM NONTRADITIONAL SOURCES
The above discussion focused on sources traditionally considered for con
trol of ambient levels of particulate matter; i.e., sources which are
generally stationary point sources, or fugitive emissions. These tradi-
tional sources were shown to cause levels of TSP that were far in excess
of the national ambient air quality standards. However, even in cities
where TSP emissions from traditional sources are relatively small, city-
o
wide averages are 30 yg/m or more above nonurban levels, and the secon-
dary annual standard is being violated. Monitors in apparently clean
63
-------
75
CITY CATEGORIES (SEE TEXT):
I TRADITIONAL SOURCES ARE THE DOMINANT PROBLEM
IT TRADITIONAL SOURCES ARE A MODERATE PROBLEM
3H TRADITIONAL SOURCES ARE A MINOR PROBLEM
50
4.
l-~
UJ
UJ
oc.
a.
CO
25
HI
RESIDENTIAL
COMMERCIAL
INDUSTRIAL
Figure 9. TSP increments due to traditional sources at different
site types by city category
64
-------
areas of a city or in smaller, nonindustrial cities have measured high
TSP concentrations which cannot be explained by modeling with tradi-
tional source emissions or which fail to decrease as expected under
controls of the State Implementation Plans. These findings indicate
that a certain level of particulate in cities is caused by the concen-
trated activity in an urban area. These activities are here collec-
tively designated "nontraditional" sources; i.e., those sources not
traditionally considered in air pollution control strategies.
Nontraditional sources of particulates may be divided into two categories.
One category consists of obvious, distinct sources of emissions that have
not been normally considered as sources; these include construction
and demolition activities, emissions from tailpipes, and tire wear. The
other category refers to the more general problem of activity in the
city and the characteristics of the urban setting that allow particulates
to become entrained or reentrained.
Before the discussion of these categories, two terms merit differen-
tiation: fugitive emissions and fugitive dust emissions. Both refer
to general, nonstack emissions of particulates. However, fugitive emis-
sions (included under traditional sources) result from industrial-related
operations and escape to the atmosphere through windows, doors, and vents
rather than through a primary exhaust system. Fugitive dust emissions,
on the other hand, are generally related to natural or man-associated
dusts (particulate only) that become airborne due to the forces of wind,
man's activity, or both. Fugitive dust emissions include windblown par-
ticulate matter from paved and unpaved roads, tilled farm lands, and ex-
posed surface areas at construction sites. Natural dusts that become
airborne during dust storms are also included as fugitive dusts.
Estimates of emissions from nontraditional sources in several comprehen-
sive reports tend to indicate that the total level of particulates from
several of these activities are significantly larger than those from
traditional sources. This study did not include any major effort
65
-------
to develop additional emission factors; rather, it concentrated on a re-
view of the previously published literature plus analysis of air quality
data collected to determine the impact of such sources on ambient TSP-
Appendix E provides the detailed results of much of this analysis, the
following discussion summarizes the findings.
Reentrained Particulates
In an urban area, particulate matter accumulates on the various city sur-
faces due to fallout, and especially heavy loads on streets can result
from dirt and mud carryout from unpaved parking lots and roads, spillage
from trucks, and sand and salt applied for snow control. This particulate
matter can then become entrained and at least temporarily suspended in the
ambient air due to wind erosion or man's activities disturbing the surface.
Natural Reentrainment - Natural reentrainment of particulates occurs when
wind is strong enough to lift particulates from the surface. Due to the
mechanics of wind erosion, such movement is more likely to be initiated
in an urban area where hard, flat surfaces are exposed to the sweeping
action of the wind. Based upon the analysis presented in Appendix F of
this report and other literature, winds above 10 to 12 miles per hour are
likely to be contributing to the TSP levels. Above this speed, the re-
entrained dust maintains the TSP level above what would be projected based
upon the balanced dilution effect of the ventilation accompanying the wind,
Cleaner surfaces (determined by comparisons after rainfall) did allow
the dilution effect to reduce levels further than when the wind was blow-
ing over dirtier surfaces (measured on days before which there was no
rainfall).
Vehicular-induced Reentrainment - The most important contribution to re-
entrainment caused by man in an urban area is the disruption of surface
66
-------
dust by motor vehicle activity. The wheels of the vehicles not only
impart kinetic energy to particles on the road but also grind up the
larger, nonsuspendible particles into smaller ones and break up the co-
hesive bonds of the .dust. Such activity in an urban area occurs primarily
on paved roads with additional local impacts expected due to dirt and
gravel roads and parking lots. The amount of particulate reentrained by
motor vehicles is directly related to the amount of dirt on the road,
its suspendibility (sand versus dust), the speed of the vehicles, and
the level of activity (often expressed as ADT - average daily traffic).
Vehicular activity on paved roads - The data gathered in the course of
this study provided several opportunities for making estimates of the
impact of vehicular-induced reentrainment by comparisons of comparable
monitoring sites. The best data were the result of special studies that
had been or were being conducted by the local agencies to determine for
themselves the impact of traffic on the measured TSP levels. Generally,
this was done by monitoring in one location but at either different
heights or distances from the road, or both. The analyses of these data
for each city are given in the individual city reports and the cross-city
analysis is provided in Appendix E.
The findings from these data indicated that, there was a direct relationship
between the daily TSP concentrations and average daily level of traffic (ADT)
and an inverse relationship between TSP and the distance of the monitor from
the traffic, measured by the slant distance I SD = J(height) + (distance) ).
A comparison of the ADT/SD to the TSP concentration implied that a linear
relationship could be assumed to exist with good correlation.
As is discussed more extensively in Appendix E, this relationship between
ADT and slant distance has a significant impact on the interpretation of
TSP levels measured by a hi-vol anywhere near a street with significant traf
fie. The data assembled in this study were not quite adequate to support
development of a quantitative relationship suitable for accurate calculations-
67
-------
however, it is possible to provide rough estimates of the impact of ve-
hicular traffic on nearby hi-vols. The data in Table 11 are meant to
provide approximate values suitable for identifying sites with potential
problems and roughly judging the magnitude of the problem.
Table 11. APPROXIMATE IMPACT (IN ug/m3) OF
VEHICULAR TRAFFIC ON NEARBY
HI-VOL SITES
Traffic
volume ,
(ADT)
1,000
5,000
10,000
30,000
Slant distance of hi-vol from street
(feet)
*
20
5
25
50
100
50
2
10
20
50
100
5
10
25
150
3
7
15
Vehicular activity on unpaved areas - In many urban areas dirt or gravel
roads and parking lots are used by individual establishments or in indus-
trial areas because of the expense of adequate paving. These areas can
be sources of dirt for carryout to paved areas and may also serve as areas
for naturally reentrained dust. In addition, vehicular activity on these
areas can bring about man-induced reentrainment.
A comparison of the published emissions from unpaved roads with those of
traditional emissions is given in Appendix E for those central counties
analyzed in the course of this study. That analysis implies that the
unpaved road emissions in counties which are not totally urbanized can
be 10 to over 30 times the emissions from traditional sources. Even in
urbanized areas, where unpaved roads are not common, the fugitive dust
from unpaved roads may be over 10 percent of the traditional emissions
in the county,
68
-------
Based on the discussion on reentrainment from paved roads, which indicates
that the impact of fugitive dust from vehicular activity decreases quickly
with distance, it is felt that these numbers are inappropriate for direct
use in air quality planning. If these fugitive dust emissions are treated
the same as traditional emissions and used in rollback or dispersion model-
ing calculations, there would be excessive, undeserved emphasis placed on
these sources and a potential deemphasis of the control of traditional
sources. While these amounts of particulate may be temporarily reentrained
due to vehicular activity, they are not suspended for any length of time.
If they were, the rural areas of counties would be expected to have TSP
levels as high as those found in the cities; such is obviously not the
case. Therefore, the use, if any, of these numbers would have to be li-
mited to inputs to models which adequately reflect the deposition and
other removal of the particulates.
Specific Urban Sources
Certain activities in urban areas have not been considered major contribu-
tors to the TSP levels and therefore have received little attention in
the formulation of control strategies for particulates. Yet these sources
may be considered true emission sources because the particulates arise
directly as a result of the individual activity rather than as a by-
product, and several recent studies have suggested that these sources
may be having more of an impact than previously thought. Of particular
interest in a crowded urban area are transportation sources - the tailpipe
emissions from automobiles and the emission of rubber due to tire wear.
In addition, construction/demolition activities that are constantly
occurring in cities add to the total TSP levels measured. Each of these
sources is discussed below.
Transportation Sources - Although particulates from the transportation
sector have been 'inventoried, they have seldom been regulated except
through ordinances prohibiting smoking vehicles and Federal restrictions
69
-------
on aircraft. Controls on motor vehicles have centered around emissions
of carbon monoxide, nitrogen oxides, and hydrocarbons, which are one to
two orders of magnitude greater than emissions of particulates. In ad-
dition, partiallates from the transportation sector have generally been
assumed to be insignificant when compared with emissions from traditional
sources. However, as emissions from traditional sources have been reduced
under implementation planning, their proportionate contribution to the TSP
problem has been reduced so that particulates from the transportation
sector have become increasingly more important.
Tailpipe emissions Attempts to separate out the contribution of motor
vehicles to the total TSP measured have centered around the use of lead
as a tracer element. In most urban environments where lead, copper, and
zinc smelters, grey iron foundries, or other major point sources of lead
are not prevalent, ambient lead levels are assumed to be due almost en-
tirely to vehicular activity. Therefore, if the ratio of TSP emissions
from tailpipes to the suspended lead emissions is known, the ambient levels
of lead can be multiplied by this ratio to provide the ambient TSP con-
tribution due to total tailpipe emissions. Based on several studies re-
viewed in Appendix E, this ratio may be assumed to range from 3 to 5 de-
pending upon the vehicle type and age mix.
Ambient lead levels are routinely measured in major cities through anal-
ysis of NASN filters, and many state and local agencies also perform their
own studies for lead. (California has an air quality standard for lead of
1.50 yg/m^ for a monthly average.) Ambient lead levels contained in the
National Aerometric Data Bank (MADE) indicate average annual concentra-
3 3
tions ranging from 0.5 to 2 [ig/m with a few cities measuring 3 to 4 ug/m .
These lead data suggest that tailpipe emissions are contributing from 1
3
to 20 ng/m to the total particulate levels measured. Data collected from
the individual cities studied under this effort indicate lead values in the
3
middle range of those reported above; i.e., around 1 |j.g/m . Therefore,
it may be assumed that tailpipe emissions are generally contributing 3 to
3
5 (ig/m to the ambient levels.
70
-------
Special sampling studies conducted by EPA in Miami and St. Louis as part
of this study provided data on particle sizes by elemental composition.
Results in both cities indicated that the lead particles being sampled
are extremely small. Only a small percentage are greater than 4 um in
diameter, and the largest percentage was collected in the last impactor
stage, implying the particles had an effective aerodynamic diameter of
less than 0.25 jam. This small diameter means that the lead would be dis-
persed and transported much as a gas with very little fallout with distance.
Therefore, it may be expected to be measured at rooftop levels or even in
more remote areas.
As part of this special study conducted by EPA in Miami, 2-hour elemental
concentrations were compared with the hourly traffic counts for 1 week.
These data illustrated the expected relationship of increasing lead con-
3
centration (as high as 4.6 [ig/m for a 2-hour average) with increasing
traffic, especially in the early morning when rush-hour traffic started and
before mixing height and wind speed increases caused a drop in lead
concentrat ion.
A compilation of all the sites for which some lead data were available
provided 49 monitors from six cities (Baltimore, Miami, Oklahoma City,
Philadelphia, San Francisco, Washington) with annual average lead concen-
trations as well as individual filter analyses from several sites in the
other cities discussed above. Those monitors with annual data were grouped
according to their site classifications and then averaged to provide a
mean concentration. Because the monitoring sites had a wide range of
local influences affecting the measured lead levels, Figure 10 presents
not only the mean values for each of the site classifications but also
the range of values found. Since there were only four monitoring sites
each for the classification of rural and industrial, these averages and
ranges may not be representative of situations found in other cities.
71
-------
£,.*
2.2
2.0
» '.8
^ 1.6
1 1.4
H 1-2
1 1.0
0
o 0.8
UJ
J 0.6
0.4
0.2
n
v
-
v-INDUSTRIAL
- \ (4 SITES)
- \j
-
i
1
«
»
^
COMMERCIAL
(28 SITES)
_
RESIDENTIAL
(13 SITES)
<
RURAL
J_ (4 SITES)
»
t L
Figure 10. The range and average lead concentrations
found at monitoring sites
72
-------
Tire wear - Aside from direct tailpipe emissions and reentrainment dust,
automobiles are known to generate particulates simply from the deteriora-
tion of the body and parts. Rust, corrosion, and friction of one part on
another are all known sources. However, their magnitude is small compared
with the wear that is seen on tires. In the U.S., an estimated 660,000
tons of tire-tread are worn away each year. Since over half of all vehicle-
miles traveled (VMT) is in urban areas, approximately 350,000 tons of rub-
ber are added to the urban environment every year. Considering the size
and widespread nature of this source, its impact on air quality warrants
study.
Although filters were selected from each city for microscopic analysis,
neither the selection of a few filters nor the accuracy of the microscopy
were believed to be sufficient to characterize the cities. However, the
numerous filters from among the cities were considered to be adequate for
averaging contributions according to the various classifications for the
monitoring sites.
By using the percentage contribution of rubber tire fragments to the total
visible loading on the hi-vol filter (diameter > 1 ^tm) and the assumption
presented in Section II that approximately 85 percent of the loading was
visible, average rubber loadings can be calculated for each site type.
These average values, along with the range of values observed, are plotted
in Figure 11. This figure shows that commercial sites, generally most ex-
posed to traffic, have the highest contribution of rubber while undeveloped
or rural sites barely measure any rubber. (Note cautions in derivation
given in Appendix E.)
In the course of the careful evaluation of the monitoring network in each
city, monitoring sites were also rated on the basis of local influences,
including paved .roads. Sites with an expected paved road influence (10
in all) had rubber concentrations twice as high as sites for which no such
influence had been noted - 9.9 and 4.9 pg/m^ respectively.
73
-------
ov
10
1 28
-------
The values found for rubber in the course of this study are several times
what were expected based on the literature reviewed. The reason for this
discrepancy is not apparent. This experiment may have been better formu-
lated than previous ones to give a good cross-section of values, or the
high levels may be an artifact of the monitoring site and filter selection.
Currently, however, there is no reason to doubt the validity of these
data.
Of some interest in planning for control is the particle size distribution
of the rubber. As shown in Table 7, the average size range of the rubber
tire fragments (13 to 135 urn) is much larger than that of any other par-
ticulate identified. Some particles were found to be 200 ym in length.
Normally, such large particles are not considered suspendible for any
length of time and are too large to be of concern for respiratory effects.
Despite their size, however, no difference was discernible in average con-
centrations of rubber by monitor height. Several monitors 50 to 100 feet
above ground level measured levels of rubber in the 5 to 15 yg/m range
while other, lower monitors recorded no rubber.
Construction/Demolition - The movement of materials associated with con-
struction and demolition activities usually results in the emission of
particulates into the ambient air. Major demolition programs, whether
using a ball and crane or blasting (low-yield), will emit particulates
up to a height equal to that of the building being removed. Construction
involves much more movement of materials continuously for periods of
several months to over a year. Emissions are generated by a wide variety
of operations over the duration of the construction, including land
clearing, blasting, ground excavation, and on-site traffice, as well as
the construction of the facility itself.
The study findings in Appendix E illustrate that construction activity
does have an impact on very local TSP levels but that the effect is not
readily predictable. Construction will generally elevate concentrations
downwind from the site for distances up to a mile; the amount of increase
75
-------
is related to the level of activity, type of activity, distance from the
activity, and control measures employed. Monitors within half a mile of
construction may have annual geometric means 10 to 15 ug/m3 higher than
normal. Therefore, if 10 percent of the monitors in an urban area are
near construction activity, the calculated citywide average TSP level
would be 1 to 2 yg/m3 higher than otherwise expected.
These measured impacts are much less than would be expected based on a simple
interpretation of the emission levels developed for each county (see Appendix
E). Obviously, the use of those emission levels must be restricted to input
into modeling programs which adequately account for the fallout and depo-
sition of particles. Development of new emission factors to reflect the
type and degree of activity and any control measures would be more ap-
propriate than the use of the existing factors.
National Assessment
While the above discussions have not covered all topics possible under
the heading of nontraditional sources, they did center on those sources
that have been identified as probable major influences. From these
sources alone it is evident that there are contributions to the total
TSP levels simply from man's activity and that the contributions are the
highest in an urban area where man's activity is greatest. Similarly,
the closer to the activity, the larger the impact. These variations
have been addressed above for each of the sources, but an overall
combined assessment is needed to indicate the extent of the total
impact of nontraditional sources on TSP levels, and thereby on the
problem of attaining standards.
Because of the range of TSP levels that may be contributed by the various
nontraditional sources, it is not possible to identify at this stage either
the exact impact at any monitoring site or even the average impact in any
one city. Such a determination would require extensive data and modeling,
76
-------
most of which are not available. Rather, the intent is to provide a mea-
sure of the range of impacts that may reasonably be expected in most sit-
uations and an understanding of the relative importance of nontraditional
sources for standards attainment on a national basis. Therefore, this
conclusion should not be taken as sufficient to preclude detailed analysis
in each city but as guidance to the development of national priorities for
further planning measures.
The average and range of TSP levels attributed to tailpipe emissions and
tire wear were given in the above analyses by site type. Recognizing
that ranges of values varied in different cities, average contributions
can still be calculated. Tailpipe emissions provided an average level
3
of TSP in industrial and commercial areas of 4 to 5 yg/m and approximate-
3
ly 3 yg/m in residential areas. Tire wear added rubber concentrations
3O o
of 6 yg/m at industrial sites, 9 yg/mj at commercial sites, and 3 yg/mj
at residential sites.
Construction activity is more difficult to present on an average basis
because of the wide range of possibilities that may occur. Some cities
have construction underway at individual, widely dispersed locations
which are not close to monitoring sites, while others may have similar
activity but close to one or more monitoring sites. Levels of TSP due to
3
construction are expected to range between 0 and 15 yg/m ; the closer
the monitor, the greater the impact. If only one or two monitors out of
a network of 20 are near construction activity, a citywide average will
o
only be affected by 1 to 2 yg/m annual geometric mean. However, in some
cities major construction programs such as urban renewal and subways are
going on in concentrated areas of the city. These activities are apparently
causing higher-than-normal values at a large number of nearby monitors and
will provide elevated average values in the commercial section of the city.
While construction is obviously a localized source, the reentrainment
problem exists wherever there are roads and traffic. Since the level
77
-------
of reentrained matter could not be exactly determined through microscopy
or elemental analyses, as with rubber and tailpipe emissions respectively,
other measures were necessary. By calculating the excess levels at resi-
dential, commercial, and industrial sites that could not be explained
after accounting for nonurban levels and traditional sources, the total
nontraditional impact on the annual geometric mean TSP levels averaged
3 3
around 20 to 25 yg/m at residential sites and 30 to 35 yg/m at commer-
cial and industrial sites. By subtracting the above levels estimated to
be due to tirewear, tailpipe emissions, and construction, the average
contribution due to reentrainment and any other nontraditional sources
3 3
would be approximately 20 (ig/m at industrial monitors, 18 (ig/m at com-
3
mercial monitors, and 14 |_ig/m at residential monitors. (The higher le-
vels at industrial monitors are likely the result of dirtier roads in the
area.) These levels of TSP were compared with those calculated in speci-
fic studies in Miami and Providence, by using the distribution of monitor
siting situations and the expected impact of reentrainment at each site,
and also with projected reentrainment levels based on published reports
of the ratio of tailpipe TSP to reentrainment TSP. The same order of
magnitude was found in all cases.
Figure 12 indicates the quantitative impacts on TSP levels of each of the
nontraditional sources considered above in various types of neighborhoods.
As has been stressed throughout this discussion, these are only average
values and a wide range of values can be expected when comparing particular
situations.
MONITORING CONSIDERATIONS
The monitoring of ambient TSP levels is not a causative factor in the at-
tainment of standards in the same sense as high nonurban levels or emis-
sions from either traditional or nontraditional sources. However, network
configuration and station siting do affect the extent to which measured
levels are representative. These factors also affect the overall quality
and usefulness of the data base needed for both air quality planning and
verifying attainment.
78
-------
40
30
z
LJ
(T
O
Z
Q.
CO
t-
20
REENTRAINED
+
OTHER
TIRE WEAR
TAILPIPE
CONSTRUCT! Oh
REENTRAINED
+
OTHER
TIRE WEAR
TAILPIPE
CONSTRUCTION
REENTRAINED
+
OTHER
TIRE WEAR
TAILPIPE
CONSTRUCTION
RESIDENTIAL
COMMERCIAL
INDUSTRIAL
Figure 12. Nontraditional source increments at different
site types
79
-------
This section compares current EPA guidance on monitoring, as found in
OAQPS Guideline No. 1.2-012, with the actual network configurations and
siting practices found during visits to more than 150 monitoring sites.
It contains an analysis of the effects of the variations in networks and
siting on measured levels and thus on standards attainment, and an anal-
ysis of the impact of deviations from EPA guidelines.
Monitoring Objectives
Table 12 from the EPA monitoring guideline document lists the objectives
that have been commonly used in designing current networks. Basically
these are the outgrowth of the original NASN objective of surveying typical
pollutant levels, with some recent additions in the areas of planning and
enforcement.
Table 12. GENERAL MONITORING OBJECTIVES
Provide data for research
Provide data for air quality planning efforts
Provide data for emergency episode prevention
Monitor time trends and patterns
Monitor source compliance with regulations
Ascertain attainment and maintenance of NAAQS (population exposure)
Determine impact of specific proposed or constructed facilities on
ambient concentration
Provide data to support enforcement actions
Monitoring Guidelines
The current EPA monitoring guidelines define a general structure for
discussion and planning network monitoring objectives, emphasizing the
need to design the network to meet well-defined objectives and data needs.
They prescribe the general size of monitoring effort required, emphasizing
80
-------
that this must be adjusted to suit local conditions. More specifically
concerning hi-vol placement, the guidelines recommend a horizontal clear-
ance of at least 2 meters and a height range of 2 to 15 meters; this lat-
ter permits placement at any height from essentially ground level to
about 50 feet.
Network Configuration
The concept of network configuration involves the number of monitoring
sites and their geographic distribution over the area of concern. It in-
cludes both the concept of selecting patterns of sites and areas of cities
over distances of several miles and the concept of selecting neighborhoods
over distances of a few city blocks.
Configuration Problems - In general, the monitoring networks studied in
the 14 cities did not have major problems with overall configuration.
However, two problems of some concern in several of the cities do warrant
further discussion. These problems were the general lack of stations to
measure incoming air mass concentrations and the lack of clearly defined
industrial area monitors in several cities.
The widespread lack of relatively remote stations is to some extent a
matter of policy and agency jursidiction as well as a matter of network
design. For only a very few of the urban areas studied was there an ap-
propriate station to measure the TSP loadings of incoming air masses.
Previously operated nonurban sites have been abandoned in two areas,
while in many they never existed. When necessary for air quality planning,
a TSP concentration in incoming air is typically just assumed, usually
based on the NASN nonurban sites, which are frequently not appropriately
near.
This lack of adequate nonurban data is not yet a serious problem in major
industrialized urban areas, where ambient levels typically exceed the
standards substantially, so that precise knowledge of incoming levels is
81
-------
not yet necessary. However, it will no doubt become increasingly proble-
matic as ambient levels approach the standards, and as improvements in
air quality require increasingly precise planning. In those study areas
where levels are nearer the standards, particularly those areas
where traditional sources are not dominant, there is already a planning
problem resulting from the lack of precise knowledge of the transition in
levels from remote through suburban into urban areas. This problem is
of additional concern in regions of the country where we find significant
levels of secondary pollutants of generally unknown origin.
The second area in which there were some problems with network configura-
tion is the matter of sites in industrial areas. In some of the heavily
industrialized urban areas studied, there were clearly defined industrial
sites, located either within the industrial area or along the margins be-
tween industrial and residential areas. At these sites, there was no
real question about the air quality influence of the industrial areas
and operations, and the trend or lack of trend in industrial emissions
was clear. In other areas, however, the sites best described as industrial
were not in fact located in or representative of the most uniformly dense
industrial areas, but rather were often influenced primarily by one nearby
industrial source. Consequently, the air quality impact of the city's in-
dustrial areas is not clearly monitored, and the effects of control efforts
are not readily seen. The reason for these problems is primarily the
nature of the industrial areas themselves. In the areas that are well
monitored, the industrial activities are typically iron and steel mills
and associated metallurgical operations, usually located in large, con-
tiguous, readily identifiable areas. In contrast, the poorly monitored
areas are in long, narrow riverside flood plains and contain much more
heterogeneous industrial activity, with extensive warehousing, truck and
rail terminals mixed in. Another somewhat similar problem is the matter
of agency jurisdiction. Not accidentally, major industries are frequently
located just beyond city boundaries, either in smaller satellite munici-
palities or in unincorporated areas, and hence tend to escape thorough
coverage in a network focused upon the responsibilities of a city agency.
82
-------
This is clearly a difficulty occurring in Philadelphia, where a major
heavy industrial area extends from the edge of the city along the river-
front in adjacent counties and into adjacent states.
Air Quality Patterns - Based on site visits, over 150 hi-vol sites were
classified as nonurban, residential, commercial, or industrial on the
basis of the principal impact on air quality, rather than strictly on
location; that is, a site in a residential neighborhood that received a
major impact from an adjoining industrial area was categorized as indus-
trial rather than residential. A consistent pattern appeared when the
sites were grouped by the four neighborhood types. In each urban area,
the average air quality levels were lowest in the residential neighbor-
hoods and highest at the industrial sites. Table 13 compares average
concentrations by site type after those sites that are unduly affected by
nearby sources (in addition to an areawide influence) were removed from
the data base. Not surprisingly, industrial sites were systematically
higher; traditional sources have long been recognized as a major source
of air pollution. However, it is also important to note that commercial
sites were similarly higher. This is because the potential for automotive-
related pollutants (such as exhaust particles, rubber tire particles, and
other types of reentrained street dust) is greater in commercial areas
than in residential areas due to increased vehicle miles traveled (VMT).
Because ambient air quality levels in different types of neighborhoods
show such distinct differences, it is important to consider network con-
figuration, especially the representation of industrial neighborhoods, in
any comparison of air quality between cities. Failure to do so might very
well result in faulty conclusions about the effectiveness of regulations,
the relative contributions of source categories, or any other objective
of the comparison.
83
-------
Table 13. AVERAGE ANNUAL TSP CONCENTRATIONS
BY NEIGHBORHOOD TYPE
Cities
Category I
Cleveland
Birmingham
St. Louis
Average
Category II
Philadelphia
Baltimore
Cincinnati
Average
Category III
Chattanooga
Denver
Seattle
Providence
Average
w/o Denver
Category IV
Washington, D.C.
Oklahoma City
Miami
San Francisco
Average
Ho. of
sites
studied
11
13
22
10
9
12
8
6
7
8
9
14
13
11
Geometric moan TSP concentration
above nonurban levels,
Mg/m3
Residential
45
35
29
36
26
30
24
27
20
48
21
18
27
20
17
26
24
24
23
Commercial
86
56
41
61
48
40
36
41
42
81
33
29
46
35
30
49
3fi
31
37
Industrial
113
88
74
92
58
71
66
65
54
96
62
None
71
58
None
None
None
None
None
Station Siting
The selection of the actual site for placement of the hi-vol is a major
consideration. As a matter of historical practice, siting decisions have
usually been made on the basis of more practical matters, such as building
access, security, power availability, and so on. However, the principal
factors to consider on a technical basis are the effects of height and
distance from the street, and any nearby sources of particulates.
Height Effects - The primary problem with siting is the height of the hi-
vol above the ground or street level. The height issue is difficult be-
cause the traditional practice is to put hi-vols on rooftops, while the
84
-------
more recent view of some is that the health-oriented spirit of the Clean
Air Act should mandate placement near the breathing zone.
The original decision in favor of placing hi-vols on rooftops rather than
at street level, first made in establishing the NASN in the early 1950s,
considered the fact that such siting would minimize the measurement of
particulate matter reentrained from the ground. (The other significant
reasons were concern over vandalism and the fact that ground-level sites
are hard to find in the CBD.) The fact that these original monitors were
deliberately placed where measurement of reentrained particles would be
minimized emphasizes the changing concept of the TSP problem. Today
low-level, reentrainment-type emission sources are seen by many cities
as the reason that they cannot meet the ambient standards, but in the
1950s they were considered by most as extraneous interferences, not pol-
lution sources; the pollution sources were the heavy industries and the
large fuel combustion operations. The EPA monitoring site guidelines
(OAQPS No. 1.2-012) state that the priority area for sampler location is
the zone of highest pollutant concentration, yet it also states that
"only rooftop sampling is recommended to avoid the influence of possible
reentrainment of particulates close to the ground." Consequently, appro-
priate siting of hi-vols is still an issue today, and the inconsistencies,
both among sites within an urban area and among networks in different
urban areas, continue to hamper not only careful data analysis but also
problem definition.
Typically, hi-vol heights range from ground level to the top of many-story
buildings; Figure 13 illustrates three points on this range: a
low-level (6 foot) site type routinely used in Birmingham; an unusually
high site the "Food Circus" building in Seattle at 70 feet; and a more
typical site a one-story fire station in Oklahoma City. In general,
sites of all these varieties are present in each network, but on an overall
basis there is a significant difference among cities in the average height
of the hi-vols. Table 14 shows the variations in monitor height among
the 14 cities. Clearly, variation is significant.
85
-------
(a) Low-level site (Birmingham)
Elevated site (Seattle)
(a) Sites of typical height (Oklahoma City)
Figure 13. Range of heights in typical hi-vol installations
86
-------
The distance of the hi-vol back from the nearest street is often re-
lated to the vertical height because of the relationship of general build-
ing size and neighborhood type. The typical central business district
(CBD) site on a tall building will be both vertically higher and horizon-
tally nearer to the street than a residential site, which may well be on
a one-story building but is more likely to be set back from the street a
significant distance. Distance back from the street is clearly a some-
what more flexible parameter than height, as the hi-vols can usually be
moved about on a rooftop. The analyses of dust reentrainment from streets
in Appendix E found that there was some evidence that TSP levels were more
affected by variations in height than by distance back from the street.
Table 14. VARIATION IN MONITOR HEIGHT AMONG THE 14 CITIES
Cities
Heavily
industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Moderately
industrialized
Cincinnati
Chattanooga
Denver
Seattle
Providence
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miami
San Francisco
Height of
monitors, feet
Mean
37
10
14
31
38
29
21
30
34
56
33
17
20
30
Median
25
6
13
30
19
25
19
23
20
49
28
15
18
18
87
-------
Impact of Nearby Sources - The exact placement of a hi-vol with respect to
other types of sources can also have a significant impact on the levels
measured. This is particularly true at sites that are at all proximate
to low-level dust entrainment sources, such as unpaved parking lots or
roads, construction activity, sources of fugitive emissions, and indus-
trial sources with much settleable particulate or low-level emission
points. The sites visited in the study were classified as to whether any
nearby source had an undue influence on the measured levels; Table 15
summarizes the number and type of sites with nearby sources. It is ap-
parent from the table that a significant fraction of the hi-vols in most
of the cities are influenced by nearby sources, and significant changes
in concentrations at these sites are dependent on controlling the nearby
source. Tn many cases, the source is nontraditional or is related to
fugitive dust.
Only three residential sites had special local impacts, but in the commer-
cial and industrial categories such influences were common. Ten of the 60
commercial sites had local impacts, even when a fairly stringent definition
of undue influence was used. For example, since so many commercial sites
had an obvious impact from traffic, such an .influence was labeled an undue
effect only if the monitor was either unusually low and close to the street
(less than about 15 feet off the ground or 25 feet from the road as in
Figure 13(a)) or affected by a major street, as in the case of samplers
adjacent to expressways. Construction, a cause of special local impacts
at three sites in 1974, was identified only if it was immediately ad-
jacent to the site (within one or two blocks) and if the impact on air
quality levels was apparent.
The 41 industrial sites presented a different situation. As initially
classified, essentially all the industrial sites had impacts from nearby
unpaved roadways, parking areas, trucking terminals, and other fugitive
dust sources. Consequently, it was deemed appropriate to include these
impacts in the basic concept of an industrial neighborhood, and no in-
dustrial sites were classified as having undue impacts. Some sites in
88
-------
00
Table 15. NUMBER OF SITES WITH AN ESTIMATED IMPACT OF LOCAL INFLUENCES,
BY NEIGHBORHOOD CLASSIFICATION
Total number of sites visited
Number with some apparent degree of
local impact
Number with degree of impact judged
major
Number with "undue" impact in context
of neighborhood definition
Average TSP levels at sites without
"undue" impact
Typical increment at sites with
"undue" impact
Residential
39
10
3
3
60 yg/m3
15 yg/m3
Commercial
60
36
10
10
78 yg/m3
25 yg/m3
Industrial
41
24
21
-
0
110 yg/m3
Undeveloped
14
0
0
0
_
-
Total
154
70
34
13
_
-------
commercial neighborhoods, however, were so classified on the basis of
undue impacts from isolated industrial sources.
Air Quality Patterns Approximate impacts of nearby sources of various
types were estimated and are summarized in Appendix G. Also determined
were the impacts of sources of fugitive emissions, generally identified
as being in the immediate vicinity of the monitor; these impacts were
discussed in the section on traditional sources.
The typical effect on air quality levels at sites with significant impact
a
from nearby sources can easily be 20 to 25 |ag/m . This represents an in-
crease of at least 30 percent over the typical levels recorded at commer-
cial neighborhood sites in the 14 study cities. Data at specific sites
can vary substantially from these average values, depending on the prox-
imity of the source to the site.
Comparison of Actual Siting to Guidelines - The general conclusion regarding
siting of the monitors visited is that they are altogether too loosely
placed with respect to both height and horizontal placement. However, they
are generally within the height range specified in the EPA guidance ma-
terial, which recommends sites less than 50 feet high, and provides only
qualitative cautions concerning horizontal placement. Thus, it is con-
cluded that more definitive guidance is needed, particularly on height and
proximity to nearby sources, jsuch as paved roads.
Operating Frequency and Schedule
Designing the frequency and scheduling aspects of TSP network operations
is generally a matter of balancing the precision desired or required in
the resultant annual mean with any necessary resource restraints. These
decisions are not generally a major problem in designing a network and
the matter of operating schedules was not found to be, in and of itself,
a significant factor relative to standards attainment. Common practice
90
-------
in the networks studied was to sample on a systematic schedule every 6th
day, following recommended EPA guidance; this is generally adequate to
produce acceptable estimates of the annual mean and the frequency of vio-
lations of the 24-hour standard.
The only significant interaction between hi-vol operating frequency and
the problems of standards attainment is a matter of having enough data to
provide adequate knowledge of the air pollution problem to be dealt with.
As an example, the analyses of meteorological parameters and TSP levels
in Section III and Appendix F, conducted to investigate the impact of
urban fugitive dust influences, were dependent on having essentially daily
hi-vol data available; in other urban areas where similar detailed anal-
yses would have been important for judging the cause of the TSP problem,
data gathered on the standard every-6-days schedule proved to be completely
inadequate for a proper analysis. This is believed to be a fairly common
failing because the very nature of particulates from urban activity makes
the problem closely interrelated with meteorology. Since the problems of
fugitive dust emissions, resuspension of material from the roadway, etc.,
are increasingly being blamed for failure to attain the standards, in-
creased sampling frequency should appropriately be a part of agencies'
attempts to define this problem adequately for planning purposes.
Summary
The degree to which monitoring considerations influence the attainment of
standards is philosophically difficult to aasess. The matter of network
configuration, or neighborhood selection, is difficult because there are
no clear criteria for defining proper configuration; it would necessarily
be a function of the purpose of the monitoring. The matter of individual
site placement is philosophically difficult because it gets into the vague
area of defining the dividing point between "ambient" air, where the
standards should be met, and source-oriented monitoring sites.
91
-------
Network Configuration The selection of neighborhoods for monitoring re-
flects a difficult tradeoff between the various monitoring objectives.
Selecting residential neighborhoods will provide better population-exposure
coverage but will likely result in lower values than selecting commercial
neighborhoods. Industrial sites, in turn, will have higher values than
commercial sites, but will provide necessary information on the progress
of traditional source control. One obvious conclusion relating to net-
work configuration is that the latitude air pollution control agencies
necessarily have in the design of their networks can easily affect the
number of sites in the jurisdiction that attain the standards. However,
this is a concern over the relative balance between residential, commer-
cial and industrial sites in-the network, which doesn't necessarily relate
to the overall effect of the control program on the aggregate exposure of
the population. One obvious approach to this, which is simple at least in
theory, is to designate smaller subnetworks for various purposes, each of
which might have a different balance of sites.
Hi-vol Siting - For certain specific sites, the difference between attain-
ing and exceeding the standard is very clearly attributable to the special
influence of some nearby source. Philosophically, this situation could be
viewed either of two ways. It could be called a sampler siting anomaly,
which should be resolved by moving the hi-vol or redesignating it a source-
oriented research station. Alternatively, it could be viewed as a site
receiving an impact from a pollution source which, though possibly temporary
or of a fugitive dust nature, is still something that should be controlled.
Which of these two interpretations is appropriate depends on the precise
nature of the site in question. Since public access has been presumed by
EPA to be the criterion for defining ambient air, the key parameter is the
extent to which the public has access to the site in question. The inter-
pretation of public access may prove difficult, however. For instance, the
Dyer Street site in Providence, which exceeds the standard because of the
impact of an expressway, is located on a parcel of right-of-way land imme-
diately adjacent to the expressway. While the public technically and
92
-------
legally has access to the parcel, as a practical matter the general public
has no reason to go there, and in fact there appears to be no significant
pedestrian volume.
In assessing the overall impact of undue influences from local sources,
the important factor is the proportion of sites falling in each of these
two categories. Based on the site visits and the above philosophy on
interpretation of public access, the significant majority of cases appear
to be situations where the source should be controlled, and the instances
of true monitoring anomalies which should be corrected are a small minority.
In a position somewhat between these latter two situations is the group
of sites that have significant but temporary local influences, usually
construction. These are to a certain degree clearly controllable fugitive
dust problems; because of their transitory nature, however, some portion
of the impact must be written off as an anomaly.
Another obvious conclusion is that, because nearby sources do affect mea-
sured air quality considerably,-the siting practices of the agency can
significantly affect the number of sites indicating violations. Consid-
eration of the siting practices of the agency is thus a necessary pre-
requisite to comparing air quality values between cities.
METEOROLOGY AND CLIMATOLOGY
General Considerations
In determining which AQCRs have met or are likely to meet the national
particulate standards and which, if any, are not, it is important to keep
in perspective the role of meteorology and climatology. As part of this
goal, the following discussion summarizes the impact of certain meteorological
variables and meteorological conditions on TSP levels in as quantitative a
fashion as appears reasonable at the present time.
93
-------
The principal effects to be considered are discussed at length in Appen-
dix F where the range of conditions experienced by the 14 study cities is
analyzed. A brief discussion of these findings follows.
Precipitation - The effect of precipitation is twofold: (1) it cleanses
the atmosphere by capturing particles within the cloud (rainout) and by
the washout of particles below the clouds; and (2) it suppresses fugitive
dust. Precipitation is very effective in reducing TSP levels in areas
with high concentrations which have resulted from either industrial or
fugitive dust sources, and average concentrations decrease steadily with
increasing 48-hour precipitation amounts in these areas. The effect of
precipitation is greatest on the day it occurs and lasts an average
of about 2 days. In the city case studies of high-concentration areas,
concentrations measured during the last half of a 48-hour period with
precipitation ^0.25 inch averaged approximately half of concentrations
measured during 48-hour periods with negligible precipitation. Concentra-
tions at typical urban sites (excluding clean residential areas) on days
with measurable precipitation were about 75 to 85 -percent of average
values for the site and time of year. In arriving at these estimates, no
attempt was made to distinguish between the effects of precipitation
itself on TSP levels and the effects of associated changes in other me-
teorological conditions.
Figure 14 presents a graphical summary of the findings at two sites in
one of the study cities when different levels of 24-hour precipitation are
used to classify a day as one with precipitation. Under one analysis, any
day with measurable precipitation (0.01 inch) was considered to have had
precipitation; in a second analysis, only days with at least 0.25 inch of
precipitation were considered as days with precipitation. Figure 14 shows
that, on the average, TSP levels remain depressed the day after rain at
North Birmingham; at Downtown Birmingham the levels return to near normal
more quickly. The effect of rainfall of different intensities is shown by
94
-------
1.50
1.40
O 1.20
t-
<
* 1.10
ui
S 1.00
oe
0.90
0.80
0.70
0.60
(29)
T46)
I 2 3 4 5 OR MORE
NUMBER OF DAYS AFTER PRECIPITATION
a) NORTH BIRMINGHAM
0.01 IN.
0-60
I 2 3 4 5 OR MORE
NUMBER OF DAYS AFTER PRECIPITATION
b) DOWNTOWN BIRMINGHAM
Figure 14. Duration of rainfall effectiveness in reducing TSP
levels at two Birmingham sites. Number of observations
is shown in parentheses. Ratio is ratio of 24-hour
concentration to 5-week running mean concentration
95
-------
the difference between the two curves for each site. The occurrence of
the peak average ratio at 4 days after precipitation at North Birmingham
reflects a few very high ratios apparently associated with periods of
dry, light-wind, poor-dispersion conditions lasting for several days.
Windspeed The effect of windspeed is also complex. As the speed of
the wind increases, the effective volume of air available for dilution
increases and, for constant source strengths, downwind concentrations
are inversely proportional to windspeed. However, in the case of par-
ticulates, total emissions are not invariant with windspeed since the
wind is the agent by which soil and dust particles are naturally en-
trained. The amount of fugitive dust entrained depends on the moisture
content and nature of the soil (or dust) and the wind speed. At speeds
below 10 to 15 miles per hour, however, the amount is basically negligible
even under dry conditions. At greater average speeds, and particularly
under gusty conditions, fugitive contributions can be substantial. Wind
speed also indirectly contributes to fugitive emissions by increasing
the rate of evaporation, and hence speeds the drying of the soil and dust
particles.
Speci: c analyses on the effect of wind speed were conducted in four of the
study cities using 24-hour average TSP concentrations and daily average
airport wind speeds. One of these (Birmingham) is heavily industrialized,
two (Chattanooga and Denver) are moderately industrialized, and one
(Oklahoma City) is lightly industrialized. Birmingham and Chattanooga
have above-average precipitation and low average wind speeds, Denver has
little precipitation and below-average wind speeds, and Oklahoma City
has nearly average amounts of precipitation and high average wind speeds.
The results of these studies showed that the dilution effect of wind
speed was noticeable below speeds of about 10 miles per hour in industrial
areas where major contributions were made from point sources. At higher
wind speeds and in nonindustrial urban areas, average TSP levels did not
appear to be related to wind speed. It was not possible to
96
-------
discover to what extent this invariance with wind speed resulted from an
interplay between dilution and wind-induced fugitive dust contributions.
However, the findings in Birmingham apparently indicate some reentrainment
of particulates at high wind speeds. As shown in Figure 15, the average
concentration decreases with increasing wind speed for speeds up to about
8 knots and then remains essentially constant. Furthermore, the decrease
of average concentration with wind speed, as shown by the dashed line, is
approximately linear from 2 to 8 knots, as would be expected if dilution
were the controlling influence. The change in slope shown at 2.5 knots
can be attributed to stagnation conditions at very low wind speeds. While
the curve for North Birmingham reflects the same relationship as that
for Downtown Birmingham, the average concentration at Mountain Brook ap-
pears to be invariant with wind speed. This difference may be explained
by the availability of particles for reentrainment being very low in the
residential areas. The industrial site, being in a dirtier area, has a
larger amount of particulate that can be reentrained.
Stability and Stagnation The stability of the air, as measured by the
change of temperature with height, controls the rate of vertical turbulent
diffusion and the mixing depth. It therefore is a measure of the dilution
power of the atmosphere in the vertical. The diurnal variation of wind
speed caused by the vertical transfer of momentum is closely related to
the diurnal variation in stability. Stagnating air masses permit the
accumulation of pollutants and the development of stable transport patterns,
While no special attempt was made during this study to isolate the effects
of stability or stagnation periods on TSP levels, several general state-
ments can be made. First, the highest 24-hour concentrations are observed
during stagnating conditions since by definition these are periods with
very low wind speeds and inversion conditions over at least a 24-hour
period. Under these conditions, concentrations are typically about two
or two and a half times the average values for the site and time of year,
and the 24-hour standards are most likely to be exceeded. In the more
polluted areas, levels may be increased 100 yg/m^ or more. For that
97
-------
450
400
350
NORTH BIRMINGHAM
(INDUSTRIAL)
x » NORTH BIRMINGHAM
os DOWNTOWN BIRMINGHAM
=MT. BROOK
IO
E
300
VO
CO
O
z
IU
o
O
250
200
ISO
100
DOWNTOWN BIRMINGHAM ^ -^
(COMMERCIAL) *
50
MT. BROOK (BACKGROUND/RESIDENTIAL)
I I I I | I I
I
Figure 15.
5678
WIND SPEED, knot*
9
10
II
12
13 14
Relationship between TSP concentrations and
wind speed at three selected sites in
Birmingham on days with 48-hour precipita-
tion amounts < 0.02 inches
-------
part of the country with the maximum number of stagnations as defined and
determined by Korshover (see Reference 10, Appendix F), stagnations
occur on an average of 3.8 percent of the days in a year. If on these days
the average concentration is 2.3 times the annual mean, the increase in
the annual mean due to the stagnation days is approximately 5 percent.
Temperature Temperature, and seasonal temperature patterns, are the best
single indicators of space heating requirements, and hence correlate with
particulate emissions from these sources. Emissions from city to city
for the same number of degree days will vary with the type of fuel burned.
Temperature also plays a part in fugitive dust emissions by affecting the
evaporation rate of water and by its influence on the growth of vegetation.
The impact of temperature (using heating and cooling degree days) was in-
vestigated in each city and the results are reported in the individual
volume for the city. Comparisons among the various cities were made
using a regression analysis, reported later in this section. (Also see
Appendix F.)
Wind Direction On a local scale, the wind direction determines the polar
distribution of pollutants around their sources, and hence is requisite to
the understanding of source-receptor relationships and the design of source-
specific sampling networks. On a regional scale, the TSP concentration
in the air mass entering an urban area is determined by the past history
of the air mass, which is best estimated by trajectory calculations;
wind direction roses can also be helpful. While wind direction-
pollution roses were calculated and reported in many of the city volumes,
trajectory analysis was done for only one case; that analysis is given in
Appendix C.
Solar Radiation In addition to being the driving energy source for
weather systems, solar radiation relates to particulate emissions thro gh
temperature, evaporation, and plant growth. Solar radiation can also
99
-------
increase the ambient TSP levels by providing the requisite energy for the
conversion of gases to secondary particulates.
Combined Impact of Meteorology
Multiple regression analysis was used to estimate the effects of annual
variations in meteorological parameters over the 5-year period from 1970
to 1974 on the annual mean TSP levels in the study cities. The three
meteorological parameters initially considered as independent variables
were precipitation, temperature (heating degree days), and wind speed.
After initial calculations , plus reflection on the apparent lack of
short-term correlation between wind speed and TSP level except in in-
dustrial areas, it was decided to exclude wind speed from the analysis.
Allowance for a linear trend was made by designating 1970 as year 1,
1971 as year 2, and so on.
In the analysis, each site type average in each city was treated as a
separate observation. Dummy variables were used to permit different
intercepts for the several cities and the three site types, while the
meteorological effects were estimated based on data from all cities.
This approach in effect assumes that the meteorological parameters operate
in roughly the same manner throughout the country, which is clearly neither
an obvious nor a trivial assumption. Previous analyses that permitted the
meteorological effects to differ from city to city did indicate that the
effects found in the various cities were quite similar in magnitude, and
it is on this basis that the assumption was made.
The resulting equation was:
TSP = C - 2.9Y - 0.43P + 2.5T
3
where TSP = annual geometric mean concentration in
C = constant in
Y = year, 1 to 5
P = annual precipitation in inches
T = heating degree days in thousands.
100
-------
The implication of the general regression equation is that within the
3
study cities concentrations have been lowering at the rate of 2.9 ug/m
per year over the last 5 years, that an increase of 1 inch in annual pre-
3
cipitation decreases the mean concentration by 0.43 (ig/m , and that an in-
crease in heating degree days of 1000 increases the mean concentration by
3
2.5 |ig/m .
This result can be used to get some feel for the magnitude of the effects
of annual changes in precipitation and temperature on TSP levels. Examin-
ation of the variations in precipitation and temperature over the 5-year
period in each of the 14 study cities showed that the smallest range in
precipitation (6 inches) occurred in St. Louis while the greatest range
(27 inches) occurred in both Chattanooga and Providence. The minimum
range for heating degree days was 188 in Miami and the maximum range was
1189 in Cleveland. The implied differences in TSP levels resulting from
3
these annual variations in precipitation range from 2.6 |j.g/m in St. Louis
3
to 11.6 |j.g/m in Chattanooga and Providence, and the differences resulting
3
from variations in heating requirements range from 0.5 ng/m in Miami to
3
3.0 ng/m in Cleveland.
Although conclusions based on this equation must be considered tentative,
the relationship provides a ready means for comparing precipitation and
heating demand effects throughout the country. For example, the results
of applying the equation to the climatological precipitation pattern
(Figure F-20) can be displayed as the relative effect of differences in
total annual precipitation on the annual TSP level, as has been done in
Figure 16. In this figure a precipitation rate of about 35 inches a year
corresponds to the "0" relative effect isopleth. The maximum geographical
3
difference in the annual mean shown by the figure is approximately 35 (ig/m
(from -25 to +10 |_ig/m ).
The use of the relationship between heating degree days and TSP level in
conjunction with the geographical distribution of degree days shown in
Figure F-27 suggests a maximum contribution to the annual mean from space
101
-------
o
NJ
+ 10
Figure 16. Relative effect of annual precipitation on annual TSP level
in a hypothetical, urban area. (Estimated from regression
equation, p. 100)
-------
heating of 25 yg/m3. The spatial variation in degree days found in the
western United States has been drastically smoothed out in Figure F-27,
so attention should really be focused on the area east of the Rockies.
The effect of space heating on TSP levels ranges from about 5 yg/m3 in
the southern tier of states to 22 yg/m3 in the most norther states. Again,
this is an attempt to generalize the effect of space heating using data
from a mix of cities with widely different fuel usage characteristics, and
the results therefore are not necessarily appropriate for any specific city.
Conclusions
Low TSP levels are to be expected in a "standard" urban area if it is
located in a generally flat, well-exposed topographical setting where:
Entering air is clean (i.e., nonurban levels are low).
Annual precipitation is high and distributed throughout
all months of the year. (Moderate amounts of precipita-
tion roughly every third day are very effective in cleans-
ing the atmosphere and in suppressing fugitive dust.)
Wind speeds are not extreme. (Light winds minimize dilu-
tion, and strong gusty winds generate fugitive dust under
dry conditions).
Daytime mixing depths are high and the frequency of low-
level inversions is low. (This combination maximizes
periods of good vertical daytime dispersion and minimizes
periods of poor nighttime dispersion.)
Space heating demands are minimal. (With a standard mix
of fuels, space heating emissions are proportional to
degree days.)
High-pressure weather systems do not stagnate. (Stagnant
anticyclones with accompanying light winds, a persistent
subsidence inversion, and nocturnal surface-based inver-
sions lead to accumulated high TSP concentrations and vio-
lations of the 24-hour standards.)
The frequency and character of winter precipitation is such
that street sanding is rarely, if ever, required. (On the
other hand, snow cover is highly effective in eliminating
fugitive dust.)
103
-------
Conversely, high TSP concentrations are to be expected in a "standard"
urban area if air entering the city already contains a substantial par-
ticulate loading, and if the location is sheltered with resulting poor
ventilation, has a cold, dry climate, and experiences frequent stagnant
high-pressure systems.
Because of complex interrelationships among meteorological parameters and
the generation, transport, dispersion, and depletion of airborne particu-
lates, however, it must be recognized that a single parameter may bring
about opposing effects and, further, that certain meteorological param-
eters tend to be linked by the nature of typical weather systems: fast
moving air dilutes and transports pollutants readily but also increases
evaporation, thus hastening the drying of soil and settled particulates,
and as a consequence the air stream may entrain particulates from the
dried surfaces; high temperatures increase evaporation, but also encourage
the growth of vegetation and eliminate the need for space heating; snow
and ice eliminate fugitive dust emissions from most surfaces, but may re-
sult in the need for extensive street sanding and hence ultimately be
responsible for an increase in urban fugitive dust emissions.
While it is possible to make certain judgments based on what are believed
to be the major effects of precipitation, heating degree days, and stag-
nation periods as done above, only a rigorous statistical analysis can
properly assess the effects of meteorological and climatological param-
eters on nationwide urban TSP levels. Such a study was beyond the scope
of the present effort, and any serious attempt to assign a TSP pollution
potential rating to individual AQCRs should await the completion of such
an analysis.
RELATIVE CONTRIBUTIONS OF VARIOUS FACTORS
By way of summary, the following discussion considers the overall average
ambient TSP level and divides it into portions considered representative
of the influences of the various major factors discussed in this section.
104
-------
Such an analysis is essentially a mnemonic device, averaging over many
important variables. Consequently, it should not be interpreted quanti-
tatively with respect to any particular site or urban area; a discussion
concerning how these results may be particularized to specific conditions
is presented in Appendix G.
The contributions to the overall particulate level at any point have been
classified as coming from three major categories emissions from tradi-
tional sources, emissions from nontraditional sources, and nonurban
particulates. In addition, two major categories of modifying influences
were identified - meteorological factors and monitoring considerations.
The rough qualitative impact of these factors was suggested in the sketch
in Figure 3 in Section II. The following quantification of those factors
is built around Figure 17-
Nonurban Levels The most basic contribution to the total is roughly
3
30 yg/m of natural and transported particulates. As discussed above,
and as seen in the small bar chart to the side in Figure 17, this level
varies significantly among various regions of the continent. The low west
3
coast level of about 15 yg/m represents the favorable situation of being
located on the ocean and receiving air with essentially globally averaged
natural TSP levels. The mid-continent level of about 25 yg/m3 includes a
10 yg/m3 increments, representing a slight increase in secondary par-
ticulates (1 to 2 yg/m3) and a major contribution from the natural par-
ticulates accumulated as the air mass passes over the broad, relatively
dry, western portion of the continent.
3
The northeastern nonurban levels then contain a further 10 ug/m incre-
ment, which consists of two major pieces. A significant increase in trans-
ported secondary particulates, the precise source of which is not thoroughly
proven, is a large share of the increment. The other share is attributed
to transported primary particulates from the major urban concentrations
of the midwest and northeast.
105
-------
n
E
Z
O
<
a:
UJ
o
o
o
120
110
100
90
80
70
60
HI
CITY CATEGORIES (SEE TEXT)'
I TRADITIONAL SOURCES ARE THE DOMINANT PROBLEM
XT TRADITIONAL SOURCES ARE A MODERATE PROBLEM
Jg TRADITIONAL (OURCIS ARK A MINOR PROBLEM
INDUST COMM
RES
50
40
30
10 -
UNDUE
,LM°P%T
i s
E g
O 3
< O
a:
<
CO
_ UJ
t £
o 3
S
(-
a:
j o
< a-
5 I
-
RES
COMM
INDUST
I-AST MIDWEST WEST
Figure 17.
Summary of average impact of major contributors
to TSP levels
106
-------
Emissions From Nontraditional Sources - The general types of nontraditional
sources identified as important are essentially the substance of human ac-
tivity in urban areas. They occur in every urban area and hence are seen
as more basic than the traditional sources, which are of only minor concern
in at least some urban areas. The overall contribution from such sources,
however, differs significantly among different neighborhoods within a single
urban area, as seen in the small bar chart in Figure 17. This contribution,
which consists primarily of the effects of tailpipe emissions, tire wear,
3
and road dust reentrainment, is about 20 ugM in residential areas and
o
about 30 yg/m in commercial and industrial areas, the difference being
the greater level of traffic. The proportion of residential versus
other sites considered in averaging over the other two categories is a
matter of abstract philosophy; an average of roughly 25 yg/m3 has been
selected for use in the figure.
Emissions From Traditional Sources - The third major contributor to am-
bient levels traditional sources varies dramatically not only
with neighborhood type but also with the general industrial nature of
the city. As was seen in Figure 9, the impact of traditional sources
3 3
can range up to about 10 |ig/m and 30 |ig/m at residential and commercial
3
sites respectively, and up to 50 |ig/m at industrial sites. This latter
figure reflects the inclusion of fugitive emissions from industrial
sources, but the contribution of truck traffic and similar industry-
related emissions has been included with nontraditional sources. Roughly
averaging over various neighborhood types and over the spectrum of non-
industrial to heavil;
eluded in Figure 16.
3
industrial to heavily industrial cities, a value of 25 |ig/m has been in-
3
This addition of 25 |ag/m is the same as the contribution from nontra-
ditional sources; thus on a nationwide basis, the two categories are
roughly equivalent. However, it must be kept in mind that the differences
among cities affects this balance significantly.
107
-------
Modifying Factors - The two major categories of modifying factors are not
so readily represented graphically. Meteorological and climatological
factors primarily can increase or decrease the average levels depending on
the various regions of the country. Quantitatively, the effect is fairly
commonly about 5 ug/m^, and in some areas 10
The influence of monitoring considerations comes primarily in distorting
comparisons of measured levels among neighborhoods and cities. However,
the aggregate nationwide average of undue influence from nearby sources
o
can add a very few yg/m to the total, as indicated in Figure 17.
108
-------
REFERENCES
1. Inadvertent Climate Modification. Report of the Study of Man's
Impact on Climate (SMIC). MIT Press, Cambridge, Massachusetts,
1971.
2. Whitby, K. T., R. B. Husar, and B. Y. H. Liu. The Aerosol Size Dis-
tribution of Los Angeles Smog. J Aero Atmos Chem. (ed. G. M. Hidy)
N.Y. Academic Press, pp. 237-264. 1971.
3. Junge, C. E. Air Chemistry and Radioactivity. N.Y.: Academic Press,
1963.
4. Clark, W. E. and K. T. Whitby. Concentration and Size Distribution
Measurements of Atmospheric Aerosols and a Test of the Theory of
Self-Preserving Size Distributions. J Atmos Sci. 24:677. 1967.
5. Cadle, R. D. Particulate Matter in the Lower Stratosphere.
J Chem Lower Stratosphere, (ed. S. I. Rasool. N.Y. Plenum Press.)
1973.
6. Position Paper on Regulation of Atmospheric Sulfates. Strategies
and Air Standards Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency. EPA Report No.
EPA-450/2-75-007. September 1975.
109
-------
SECTION IV
SUMMARY AND CONCLUSIONS
This section summarizes the findings of the study, including the study
approach to general problem assessment, the attainment factors identified,
the assessment of these factors, the control options available, and
their applicability and priority for control.
This study approached the national particulate problem with many of the
same tools an individual air quality planner would use to approach an
assessment of the problem in a particular urban area. It utilized air
quality data analysis, emissions data and modeling, and analytical particle
identification; these, along with special monitoring studies, are the
techniques with which a TSP problem at any level is studied. However, the
results of this study are intended to provide general information on a
national scale rather than urban-scale planning information for the
14 study cities or any others. While the individual air quality
planner will hopefully find the results useful, they are not intended to
solve any specific problems in specific urban areas. Rather, the results
provide an additional component of technical information which will need
to be integrated and assessed in conjunction with local knowledge.
ATTAINMENT FACTORS IDENTIFIED IN CITY STUDIES
The purpose of the city case studies was to identify and study the various
factors, problems, and issues concerned with attaining the TSP standards
as they were experienced in each city. Since the 14 cities cover a broad
110
-------
range of city characteristics and hence represent a variety of situations
with respect to TSP air quality and its determinants, analyses of the
factors in the various cities can be drawn together for an overall assess-
ment of the TSP attainment situation in the study cities and, by extrap-
olation, throughout the nation.
Following the analyses of the study cities, a number of factors were
identified as significant for standards attainment. Many of these had
been first identified in the preliminary literature review and were then
followed up in the city studies; a few others were identified in the course
of one or more of the city studies. These issues are listed below, grouped
into categories that provide a framework for discussions.
Traditional Factors
Air pollution programs have traditionally been oriented toward the control
of fuel combustion, process emissions and incineration. Such control
programs have resulted in substantial reductions in emissions and corres-
ponding improvements in air quality in many of the 14 study cities.
While all of the cities studied have ongoing control programs for tradi-
tional sources, there were differences both in the amount of traditional
source influence and in the success of the program:
Three cities still have significant problems with traditional
sources. Citywide air quality averages are typically 30 (ig/m
higher than they are in similar cities with light emissions from
industry and fuel combustion.
Three cities have had heavy industrial sources and problems, but
have made major improvements in air quality. They still, however,
have citywide averages typically 10 to 15 jag/m^ higher than
similar cities with light emissions. There is some potential
for improvement in air quality in these cities by further reduc-
tion of traditional sources.
Four cities have had moderate emissions problems and have them
well under control. These cities must control nontraditional
sources to make further, substantial improvements in air quality.
Ill
-------
Four cities never had problems with traditional sources, except
fuel combustion for heating and power generation. Further im-
provements must come from control of nontraditional sources.
The major sources still presenting problems with attainment of standards
are in the primary metals and minerals industries. Of primary con-
cern are the fugitive emissions which are not confined and do not come
from a stack or vent. These emissions have substantial impact on air
quality; sites which were influenced by fugitive emissions averaged
3
25 ug/m higher than industrial sites affected by stack emissions only.
This is partly due to the typically low level emission point and poor
potential for dispersion. Another problem is that of area source fuel
combustion; fuel oil is estimated to contribute on the order of 5 to 10 yg/nr*
to citywide averages in cities where oil is a primary fuel. A similar order
of impact is realized from the many small (minor) sources if these sources
are generally uncontrolled.
Major program considerations which were determined to be a factor in
traditional source impact are the assurance of compliance with existing
regulations and, in some cases, the stringency of those regulations. This
will be discussed further under control alternatives.
Nontraditional Factors
While many cities have realized substantial improvements in air quality,
few cities have all sites below the primary standard and fewer still have
all sites below the secondary standard. This is due to the impact of
factors which have not been traditionally addressed by the air pollution
programs, such as reentrained dust, tire wear particles, dust from con-
struction, and automotive exhaust emissions. The impact of all such
general urban activity varies from a typical impact of 20 to 25 yg/m in
residential areas to a typical impact of 30 to 35 yg/m3 in commercial and
industrial areas; the citywide contribution is about 30 yg/m3. The com-
ponents of this contribution are discussed below.
112
-------
Reentrained dust The impact was found to vary with the
traffic flow and inversely to the height and distance
of the monitor from the street. The average impact on
residential monitors is estimated at 10 to 15 ug/m3
and 15 to 20 jag/m? at commercial and industrial sites
respectively. The composition of this component is
mostly mineral matter.
Tire rubber particles This component of nontraditional
particulates is even more variable with neighbor-
hoods. Typical impacts of 2 to 5, 5 to 10 and 3 to
7 yg/m3 were found at residential, commercial and in-
dustrial sites, respectively. Sites located par-
ticularly near heavy traffic averaged twice the levels
at other sites.
3
Construction Impacts of 15 |ig/m are common only if the
construction is close to the monitoring site. Cities with
typical levels of construction have citywide impacts on
the order of 1 to 3 ug/m? -
Automotive exhaust This varies somewhat with neighborhoods,_
with concentrations of 3 |ag/m3 in residential and 4 to 5 |ag/m
in commercial and industrial areas. This estimate is for the
primary particulate only and is generally about 20 to 25 percent
lead.
Large Scale Factors
Large scale factors are the combined influences of particulate matter
that dominate an area much larger than the urban areas being studied.
They include natural particulates and transported primary and secondary
particulates. These factors, in affecting the TSP levels in an urban area,
can cause significant differences in the controllability of the TSP
problem. Their effect on urban levels is generally estimated by measuring
air quality in nonurban areas. The average nonurban particulate level
3
for the 14 study cities is between 25 and 30 ng/m ; however, values ranged
3 3
from less than 15 ng/m on the west coast to over 35 ug/m in the densely
metropolitan east.
Transported secondary particulates make up a significant portion of non-
urban levels. Nonurban sulfate and nitrate levels range from very low
113
-------
3
in the midwest and west to around 10 |j.g/m in the northeast. Total
3
urban secondary levels vary up to around 15 p.g/m in the north and east.
Other Related Factors
In addition to the above factors, which are to varying degrees related to
sources of emissions, other factors were found to affect the real or
apparent TSP problem. As discussed in the selection of the cities, the
meteorology and climatology of a region can help to aggravate or ameliorate
the TSP problem; the dispersion characteristics and precipitation levels
are the most prominent influences. The design of the monitoring network
and the actual placement of monitors is also important in conceptualizing
what the TSP situation is. The general findings from the city studies
for these factors are summarized below.
Monitoring considerations The siting practice of the control
agency has considerable impact on the air quality levels and
number of violations recorded. For example, 10 of the 60
commercial sites visited had local impacts from nearby sources.
With one exception, all these sites violated the primary annual
standard. Also, since industrial neighborhoods were shown to
have substantially higher TSP levels, the proximity of monitors
to industrial neighborhoods is an important variable. A specific
relationship was found in the study between average daily traffic
(ADT) at a site, the slant distance of the monitor from the
street, and air quality levels. For example, based on approximate
calculations, a monitor 100 feet (slant distance) from a busy
street (10,000 ADT) might be influenced by the reentrained dust,
tire wear and exhaust by 10 |ag/m3. This same relationship shows
an impact of 40 to 45 i-ig/m^ if the monitor is only 25 feet from
the street.
Meteorology The study found precipitation to be an extremely
important variable; a yearly increase in rainfall of 1 inch can
cause decreases in citywide annual averages of 0.4 jig/m^. It
also concluded that average windspeeds above 10 miles per hour
could cause some resuspension of dust. The effect on air quality
would depend upon the gustiness of the wind and the condition of
the soil.
114
-------
CONTROL STRATEGY OPTIONS
The development of a control strategy for the attainment and maintenance
of the ambient particulate standards must depend on accurate identifica-
tion of the sources responsible for the particulate problem. Once this
identification has been established, the available control measures can
be defined, evaluated, and combined into an aggregate control strategy
which is appropriate and effective. The major single conclusion of the
study is that the control measures traditionally directed at pollution
sources are not going to be adequate in many cases, and that other, more
novel control measures directed at nontraditional sources and sources con-
tributing to nonurban levels must also be considered.
Proposing detailed control strategies or priorities for specific situa-
tions is beyond the scope of this study. Such a process must necessarily
involve the addition of significant quantities of local knowledge. None-
theless, two points should be mentioned briefly. The first involves the
relative emphasis on the three main contributions to urban TSP- If
traditional sources still contribute the major share of TSP in an area,
they are clearly the most appropriate segment to attack. Potential over-
all reductions in emissions of 75 percent or more may be expected in those
cases where sources are not yet stringently controlled. In every city,
industrial or otherwise, the nontraditional sources present a more dif-
ficult target, with reductions of 50 percent or less probable even from
truly extensive control efforts. The nonurban contributions, which are a
major influence in at least the nonindustrial cities, are the worst in the
sense of being essentially intractable at a state and local scale. The
priorities for control must consider these general differences. The
second point to remember when selecting among strategies is that there are
a number of considerations other than simply mass particulate reduction.
Keeping-in mind the need for individual consideration in each area
violating the TSP standards, the following discussion lists the types of
control measures generally considered appropriate for each of the various
115
-------
sources of participates included in the three major categories previously
outlined.
Control Measures Directed at Traditional Sources
The general pattern of control technology applicable to traditional
sources is well established; the open questions are primarily those of
application. In this sense, it is appropriate to subdivide the category
of traditional sources into four subcategories:
Major point sources
Smaller point sources
Fugitive emission sources
Fuel combustion area sources.
Stack Emissions - With some differences in emphasis, the control measures
available for both major and smaller point sources are similar.
Obtain compliance with regulations - existing efforts to
bring sources into compliance must continue, and must be
extended in those areas where current efforts are
incomplete.
Tighten compliance determination and surveillance pro-
cedures - one area of significant difference by size of
source; major sources should be stack tested at least
annually, smaller sources inspected frequently under a
tight surveillance system.
Tighten regulation stringency - after considering carefully
the local need for tighter traditional source control:
Upgrade regulations for selected major sources
and smaller sources to at least require the best
control technology reasonably available (not
necessarily RACT as currently promulgated).
Incineration can be completely banned in favor
of landfilling or shredding for power generation.
Process losses covered under a general process
weight regulation can be further restricted by
adopting regulations specific to individual
problem industries.
116
-------
Fugitive Emissions - Aside from the major concerns of improved surveil-
lance and enforcement, specific measures for reducing fugitive emissions
from traditional sources include special regulations beyond simple
nuisance or "reasonable precaution" regulations:
Quantitative or visible emission standards at
property lines
Operating and maintenance standards for specific
processes
Covering storage piles
Enclosing materials handling equipment
Paving roadways and loading and parking areas
Area Source Fuel Combustion - Though difficult to effectively control be-
cause of their large numbers, there is a real possibility of reducing
emissions from such sources, if the need warrants, through measures
such as the following:
Design standards for new boilers
Maintenance of oil boilers
Prohibition of coal use
Inspection and maintenance of small oil
burners
Control Measures Directed at Nontraditional Sources
Concurrent with taking the appropriate steps to control traditional
sources, and in areas where traditional sources have never been the major
problem, control strategies for the attainment of the TSP standards may
have to consider control of those nontraditional sources discussed in
Section III. Implicit in the very label "nontraditional" is the concept
that these sources have not had control measures directed at them. How-
ever, control of these sources is not so much a matter that is beyond
reach technically as it is a matter of using measures that are costly
117
-------
or hard to justify or, in many cases, simply measures that are not normally
seen as pollution control measures. Examples of these types of measures
are listed below for the three major categories of such sources:
Motor vehicle exhaust
Fugitive dust from construction and demolition operations
Reentrainment of particles from roadway surfaces.
Motor Vehicle Exhaust - In urban areas where tailpipe emissions are found
to be contributing 5 to 10 ug/m3 to the total suspended particulate load-
ing, it may be necessary to consider their control. The available control
measures for lowering tailpipe emissions are:
Reduce lead in gasoline
Reduce VMT totals in area
Reduce emission per mile through inspection and maintenance
Particulate control devices on cars.
Construction/Demolition and Other Fugitive Dust Sources - The control of
dust from construction and demolition activities and similar activities is
most often approached through the use of nuisance or reasonable precaution
regulations much as is the case with fugitive emissions. More specific
control measures that can be applied include:
Watering construction site and demolition rubble
Chemical soil stabilization
Use of sequential blasting demolition.
Reentrained particulates - The majority of such particulates are street
dust entrained by traffic; control measures include both preventive
measures and street cleaning:
Control of dust deposition - A viable approach to controlling
reentrained particulates is reducing the amount of particulate
matter available for reentrainment; control measures include:
Fallout - The fallout level will decrease automatically
with control of traditional source emissions.
118
-------
Carryout - Dirt and mud carryout from unpaved
roads and parking lots can be reduced by require-
ments for paving or stabilization of these
areas.
Spillage - The loading on streets due to spillage from
trucks is easily regulated against, but enforcement may
be a problem. Regulations that require specific equip-
ment on trucks would probably be an improvement.
Tirewear Particulates - Reduction in the amount of
rubber tire particles could be effected by VMT re-
duction or by designing and requiring tires with better
wear characteristics.
Sanding - Because of the obvious hazard of slippery
roads, sanding and salting operations will obviously
continue. Analysis of the efficiency of sanding may
result in procedures that apply less sand more ef-
fectively; however, systematic road cleaning after
sanding operations would be more appropriate.
Street cleaning to remove deposited material - Since it is
not possible to prevent all deposition of particles on a
paved surface, an alternative or auxiliary approach to con-
trol of particle reentrainment is to remove the particles
from the surface. Control measures include:
Street cleaning
Rotating broom sweepers
Regenerative air blast sweepers
Vacuum street cleaners
Street flushing with water.
Control Measures Directed at Transported Primary and Secondary Particulates
A complicating factor in any standards attainment strategy development is
the concentration of TSP in the incoming air mass. When such concentra-
tions approach or exceed the standards, there is little if anything that
an agency can do to meet the standards unless the controls are also pur-
sued for this incoming (nonurban) TSP. Obviously, there are really no
appropriate measures for the contribution of natural sources to these TSP
119
-------
levels; however, those particulates directly attributed to man's activi-
ties are, to varying degrees, controllable. In certain cases attention
directed toward control of these sources may be more profitable than non-
traditional control techniques.
Transported Primary Particulates - Those particulates that are emitted
directly as primary particulates and are then transported from one area
to another can be controlled through conventional traditional source
regulations. The difficulty arises in the development of regionwide con-
trol strategies in which one area may need to limit its emissions more
severely than necessary to simply meet the standards in that area because
of the impact on a neighboring area. Such regionwide planning may be
immediately possible on a statewide basis or may require interstate and
inter-EPA region cooperation.
Secondary Particulates - Much the same thing can be said for secondary
particulates except that many of the secondary particulates are formed in
transport over much longer distances than would normally be of concern
for primary particulates. Therefore, these control strategies must seek
national direction rather than state or interstate cooperation. However,
secondary particulates that are locally formed can presumably be locally
controlled. As indicated in Section III, urban excess secondary particu-
3
lates may add 5 ^g/m to levels in the city and these particulates could
be controlled once the appropriate precursor relationships are known.
FRAMEWORK FOR CONTROL STRATEGY PRIORITIZATION
The institution of control measures must be preceded by the careful
prioritization of available options. This section suggests a framework
for evaluating control options as a function of the scale of the problem.
There are significant differences in the general prevalence of the three
major factors traditional sources, nontraditional .sources, and large-
scale consideration - and thus the applicability of their associated con-
trol options, when dealing with problems of different geographical scales,
120
-------
Consequently, priorities for instituting control options are best made
within this context of differing scales.
Concept of Scale
A general assessment of the relative scale of impact of the factors affect-
ing attainment can be made. It is helpful to think of TSP problems as
affecting either a relatively small area (neighborhood), an entire urban
area, or many urban areas in a general region (intercity). These are dis-
cussed below.
Neighborhood scale - TSP problems over areas a few blocks
in size occur even in the urban areas with relatively low
citywide levels; typically measured by only one hi-vol,
these problems are often caused by a relatively local
source, frequently a small industry or fugitive dust source.
Major sources are seen as less of a problem on the neighbor-
hood scale due to the tall stacks providing good dispersion.
Urban scale - In some urban areas, the TSP problem is to a
large extent citywide (in addition to neighborhood hot spots),-,
and is less likely to be due to a relatively few indentifiable
sources as on the neighborhood scale.
Intercity scale - In some portions of the country, there is
a general regional TSP problem that transcends any individual
urban area or AQCR.
Within any particular area, it is very possible to have problems in all
three of these scales simultaneously. Nonetheless, the overall control
approach to each can be generally independent and in fact, problems of
certain types typically occur primarily in one geographic scale. For
example, nontraditional sources typically cause mostly neighborhood
problems, while traditional sources may more likely cause urban-scale
problems.
121
-------
Priorities of Instituting Control Options
The previous sections have delineated various control measures that could
be instituted by agencies and suggested the general scale of applicability
of different factors affecting attainment and their associated control
measures. This section suggests the specific categories of sources for
which control measures could be applied to reduce particulate concentra-
tions at each level or scale. The source categories into which the many
types of air quality problems have been generalized are as follows:
Traditional.
Major sources
Small sources
Fugitive emissions
Area source fuel combustion
Nontraditional
Resuspended dust from roadways
Fugitive dust from construction and demolition
Auto tailpipe emissions
Large scale
Transported primary particulates
Transported secondary particulate
Urban secondary excess
The following matrix (Table 16) summarizes the priorities for adopting
control measures for these categories of sources as a function of the
scale of the problem.
122
-------
Table 16. CONTROL PRIORITIES
Order of
priority
ro
u>
Regional scale of problem
Neighborhood
Fugitive emissions
Reentrained dust
Fugitive dust/
construction
Small sources
Area source fuel
combustion
Auto tailpipe
Urban secondary
excess
Major sources
Urban
Lightly
industrialized
Reentrainment dust
Auto tailpipe
Area source fuel
combustion
Urban secondary
Small excess
sources
Fugitive dust/
construction
Major sources
Heavily
industrialized
Major sources
Urban secondary
Small excess
sources
Fugitive emis-
sions
Reentrained dust
Auto tailpipe
Area source fuel
combustion
Fugitive dust/
construction
Intercity
Major sources
Transported primary
Transported
secondary
Small sources
Area source fuel
combustion
Fugitive emissions
Reentrained dust
Fugitive dust
Auto tailpipe
-------
SECTION V
RECOMMENDATIONS
The broad scope of this study provided an opportunity for an evaluation
of the control practices to date, as assessment of the controls that are
needed in the future, and an understanding of the obstacles that are pre-
venting the attainment and maintenance of the NAAQS for TSP. Based on
these findings, numerous recommendations have been formulated. The rec-
ommendations cover a wide range of topics and are directed at various
audiences both inside and outside of EPA. Some of these recommendations
are readily apparent from this study and are, in fact, called out sepa-
rately, as in Appendix G. Others are the result of integrating all the
findings and analyses conducted over the course of the study with the
literature reviewed and the discussions with federal, state, and local
control officials.
The recommendations provided below run the full gamut of considerations.
They are organized by first presenting specific recommendations for emis-
sion control efforts, then more general recommendations concerning the
major upgrading of quantitative air quality management planning, and
finally recommendations regarding the current review and further devel-
opment of the NAAQS for TSP.
In general, all of these recommendations recognize the need to provide
appropriate justification for any new major control approach that must
be adopted. Equally important is the demonstration that proposed con-
trol programs will result in the necessary improvements in air quality.
124
-------
RECOMMENDATIONS FOR EMISSION CONTROL EFFORTS
The following recommendations are grouped for convenience into three
groups corresponding to the three major components of urban TSP levels as
discussed previously traditional particulate sources, nontraditional
sources, and nonurban particulate levels.
Control of Traditional Sources
All urban areas do not require an equivalent degree of traditional source
control. The first set of recommendations presents control efforts con-
sidered appropriate for major industrialized urban areas having an apparent
TSP problem primarily associated with emissions from major fuel combustion,
industrial process and solid waste operations. The second and third set
of recommendations are directed at all urban areas, both those that are
heavily industrialized and those that have only a moderate to light amount
of industry and area-wide combustion of oil or coal. Such a breakdown
presumes that most of the effective programs will already meet the first
set of recommendations for local control planning. However, additional
control may still be needed under large scale planning efforts as dis-
cussed later.
1. Control of Major Point Sources
a. Limitations on emissions from fuel-burning installations
of all sizes should be tightened considerably. There is
a factor of at least 4 between the typical regulation and
the most stringent, yet workable, regulation; while very
stringent requirements are likely not required in many
urban areas, in major industrial areas they are clearly
needed.
b. Emission limitations applicable to major oil-burning in-
stallations should be defined separately, and more strin-
gently, than those applicable to coal combustion. A single,
uniform combustion regulation that was designed to permit
some coal combustion will unavoidably permit relatively
poorly controlled residual oil combustion, simply because
of the inherently different emissions potential of the
two fuels. There is no need for this, nor for the atti-
tude that coal-to-oil conversion should be seen as a total
125
-------
abatement of the source; oil-fired sources can and
should be controlled.
c- All major combustion sources that have undergone coal-
to-oil conversion should be tested. There are significant
differences in effectiveness when control devices de-
signed for coal fly ash are utilized on oil-fired units,
yet it appears not uncommon for the original percent-
efficiency figures to be utilized in calculating emissions
levels.
d. Emissions from industrial process losses should be
regulated on the basis of industry-specific restrictions.
The present common practice of utilizing a general process-
weight curve necessarily means that relatively-easy-to-
control processes will only be required to meet such
standards as are appropriate for the more difficult to con-
trol ones. Major processes in any given area should be
regulated with a specific emission limitation economically
and technologically tailored to that industry.
e. Compliance determination procedures in general should be
significantly strengthened, primarily by increased use
of source testing and source surveillance. In general,
the degree of assurance of stated control efficiencies,
etc., appears to be less rigorous than needed to provide
good quantitative compliance knowledge. While it is not
recommended that all sources be physically tested, it is
believed that a much greater level of testing is needed
than is now practiced. All large sources and all unique
sources should be tested on compliance attainment and at
intervals thereafter, either by agency personnel or by an
approved independent testing organization. Reliance on
routine engineering calculations should be permitted only
for the simplest, most routine situations. Source tests
should take care to include not only stack emissions but
fugitive emissions from processes. Systematically and
randomly scheduled visits should be made with schedules
depending upon the source size, history of complaints and
problems, and compliance status.
2. Control of Fugitive Emissions
a. EPA should develop and make available an accurate me-
thodology for inventorying the emissions and estimating
the air quality impact in the vicinity of isolated sources
of fugitive emissions, such as quarries and rock-crushing
operations. In order to include such emissions in air
quality management planning, more adequate information must
be developed.
126
-------
b. In every heavily-industrialized area, the cognizant
state and local agencies should conduct, and EPA should
support, a major survey effort to identify and Inventory
fugitive emission sources. Such surveys should include
field monitoring, extensive inspections of industrial
premises, and the rough estimation of emission quantities.
While fugitive emissions from generally-isolated indus-
trial operations (such as quarries) are relatively easy
to identify, if not quantify, it is not easy to define
the degree to which fugitive emissions from dense heavy
industrial areas are a problem. Consequently, a serious
effort to define the nature and magnitude of the problem
is required prior to dealing with it.
c. Regulations applicable to fugitive emissions from indus-
trial property should be strengthened by the itemization
of specific control measures where possible and by the
institution of property-line air quality limitations
where required by the complexity of the area or the in-
dustrial operations involved. It is anticipated that in
cases of clear-cut, obvious sources it will be more ex-
peditious for the control agency to identify and specify
the required control measures. In contrast, in more com-
plex situations such as a major iron and steel facility, it
is anticipated that it will be more expeditious to require
the source to conduct property-line monitoring and be
responsible for the identification and implementation of
control measures on their own property so long as the
control measures required as a result of property line
monitoring are no less stringent than otherwise would be
required.
3. Control of Small Point Sources and Areawide Fuel Combustion
a. State and local agencies should reassess their point source
cutoff to ensure that a major percentage of their traditional
source inventory is receiving individual attention. Arbitrary
cutoff points of 25, 50, 100 tons/year are often used to
define point and area sources. Commonly, those cities that
have significantly reduced the emissions from major point
sources have used a high cutoff point. As the emissions have
been reduced, smaller sources are now of concern and should
be considered individually for modeling, compliance, and
enforcement purposes.
b. Regulations governing the allowable emissions from small
combustion units should be promulgated or, if already
promulgated, reviewed to reflect current fuel usage, fuel
availability, and control technology. Many jurisdictions
do not control small combustion units or allow emissions
higher than necessary for a well-maintained unit.
127
-------
c« Small incinerators should be banned in urban areas or
should at least have permits required to ensure proper
control and burning only under favorable meteorological
conditions. Many cities have already taken this route
with no major obstacles to its implementation.
Control of Nontraditional Sources
In contrast to the control of more conventional sources, which is pri-
marily a problem in industrialized areas, the management of particulates
arising from urban activity is likely to be required in essentially all
urban areas of any size. However, nontraditional sources are not such
a well-defined problem as to permit immediate and detailed control strat-
egy planning. Rather, no major national attack on such particulate
sources is considered appropriate for implementation until a variety of
preparatory steps have been taken. Consequently, the following recom-
mendations primarily concern preparation for, rather than actual imple-
mentation of, the control of urban activity sources.
1. Control of Urban Activity Sources
a. EPA should develop appropriate methodologies for inven-
torying emissions from nontraditonal sources in urban
areas and support their proper utilization by state and
local agencies. To the extent possible these inventories
will take into account particle size and spatial and tem-
poral emission rates.
b. EPA should develop and provide to the states a diffusion
model which adequately takes into account the deposition
characteristics and small scale diffusion of the in-
ventoried emissions. Both short-term (24-hour) and long-
term (annual) averaging models are needed which allow for
variations in parameters including precipitation, ground
cover, particulate loadings on roads (fluctuating with
street cleaning, precipitation, sanding operations), wind
speed, etc.
c. EPA should develop and implement a major effort aimed at
providing, in 1 to 2 years' time, information on the costs
and effectiveness of control measures potentially appli-
cable to urban reentrainment. The effectiveness, and to
some extent the cost, should be considered in light of
the potential for public acceptance and ease of implemen-
tation. Cross-media environmental impacts should also be
addressed; e.g., street flushing for reentrainment control.
128
-------
d. EPA should develop and implement a public education
effort aimed at developing recognition that nontra-
ditional sources are of significance with respect to
air quality. This effort should stress to the appro-
priate bodies of public opinion and governmental
agencies that such sources are a legitimate subject
of environmental concern.
e. State and local air pollution control agencies should
begin to attack the obvious visible sources of fugi-
tive dust such as construction, unpaved areas, etc.
Some agencies already have developed programs aimed
at these sources and most other agencies could control
many of the sources under current nuisance regulations.
Consideration should be given to requiring certain con-
trols on specified sources.
Control of Large Scale Influences
Additional reductions in TSP levels can be expected if precursors of
secondary particulates are controlled, not only within an urban area, but
also "upwind" of cities. Other contributions from "upwind" sources can
arrive directly via transport. Planning measures to control these con-
tributions require national direction for implementation, but will result
in a more equitable distribution of the stringency of control measures.
1. Control of Secondary Particulates
a. EPA should continue efforts to develop and document the
mechanisms of formation and models for the prediction of
secondary particulate levels, especially sulfates and
organics, on both the meso- and macroscale. The effects
of precipitation on scavenging of precursors, stagnating
air conditions, insolation, thermal radiation, etc.,
should be adequately considered.
b. State and local agencies should take into account the
formation of secondary particulates in their formulation
of control strategies for TSP. The urban excess of
secondary particulates must be either consciously in-
cluded in the TSP level that is considered uncontroll-
' able, thereby requiring further restrictions on tradi-
tional sources, or should be addressed through the con-
trol of the emission of precursor pollutants.
129
-------
c. EPA should not permit the use of supplementary control
strategies or tall stacks to satisfy immediately local
air quality needs since the result may be to increase
levels of secondary particulates at sites remote from
the source. Because of the continental scale and vari-
ability of meteorology, it is not likely that such con-
trols techniques would be permitted anywhere in the
country. In fact, additional SOX controls may be needed
at some sites to help solve the TSP problem elsewhere.
2. Transport and Primary Particulates
a. States should require air quality planning to be done
on as large a regional scale as is necessary to reach
into an area not impacted by transport. Where an auton-
omous local agency has control over only a small part of
a problem area they should have their authority extended,
through a contract with the state, a regional compact, or
some other arrangement.
b. EPA should support research efforts directed at the better
understanding of short- and long-range transported particu-
lates. Long-range transport under specific meteorological
conditions that contribute to excessive TSP levels and
cause violations of the 24-hour standards could become
predictable so that appropriate measures could be taken.
Short-range transport needs better documentation and anal-
ysis so that the intercity contributions to high TSP levels,
leading primarily to violations of the annual standard,
can be considered for air quality planning. Comprehensive
inter-EPA regional planning may be necessary for TSP
control.
RECOMMENDATIONS CONCERNING AIR QUALITY MANAGEMENT PLANNING
The second broad area of recommendations concerns improving the states1
ability to accurately and quantitatively develop plans for the attainment
and subsequent maintenance of the ambient standards. As has been noted,
the general failure of the SIPs to attain the standards despite signifi-
cant emission reductions is in part due to inadequate data bases and, to
a less extent, to inadequate planning methodologies. This area is thus
very fruitful for action, particularly short-term EPA action.
130
-------
1. Development of Improved Data Bases
a. EPA should support data gathering efforts and computer-
ized systems by state/locals which are compatible with
NEDS and SAROAD. A number of major changes and expan-
sions in NEDS and SAROAD are currently underway, and it
is essential that these receive continued support. The
fundamental concept of a joint federal/state/local pol-
lution control system depends on consistent, accurate,
up-to-date bases readily available to all.
b. EPA should provide a greatly expanded nonurban ambient
TSP data base, either through their own monitoring ef-
forts or by encouraging and utilizing such monitors run
by the states. There is a need for a clear understand-
ing of the variation in ambient levels as one moves from
remote nonurban areas through more proximate areas into
suburban areas, as well as a need for more detailed data
on the differences in levels between various portions of
the continent. Current networks need to be reviewed for
appropriateness, completeness, and possible local influ-
ences as discussed below.
c. EPA should conduct a study, based on existing nonurban
sites, concerning the effect of height and distance on
measured levels at such sites. Because of the long
history of many of these sites, dating back to the ear-
liest years of the NASN, they have apparently never been
studied carefully from a siting viewpoint, and in fact
many are apparently placed very close to the ground.
Since they must of necessity provide data to be extrap-
olated over scales of hundreds of miles, it should be well-
determined to what extent they reflect air masses on that
large scale, and to what extent very local impacts may be
important. The evaluation of the current network should
be conducted with the point of establishing consistently
sited monitors (height and neighborhood) and any changes
in the siting should be documented through concurrent
sampling for 'a period of time.
d. EPA should provide improved, more specific, network de-
sign and monitor siting guidelines for TSP monitoring
in urban areas. Although the study cities visited were
all major cities with active control programs and were
all meeting the minimum monitoring requirements, an un-
fortunate number of circumstances arose where the data
from these networks was inadequate to meet relatively
simple analysis needs. Emphases in the recommended ef-
fort should include the minimization of local effects
at sites, the development of inter-urban site standard-
ization, and the encouragement of increased monitoring
131
-------
frequency for problem diagnosis purposes. The number
of sites required for an area should not be based on an
arbitrary formula but should be determined by the need
to understand the full complexities of the TSP problem.
2. Development of Improved Planning Tools
a. EPA should promote increased use by state and local agen-
cies of special studies as planning tools. These studies
would be directed at providing analytical procedures and
control development tailored to the particular TSP situa-
tion and topographical, meteorological, industrial, and
social-economic characteristics of a region through coor-
dination with other environmental and regional planning
efforts.
b. State and local agencies responsible for emission inven-
tory maintenance should reevaluate previous years' emis-
sion inventories to make them compatible with those in-
ventories currently being used. A basic understanding
of how the emissions situation has changed in the past
and the resulting impact on air quality provides the
soundest basis for future planning.
c. EPA should support the development of the microscale dis-
persion models that are required to adequately cope with
the potential need to control vehicular traffic in urban
settings as a TSP control measure. At the present, the
quantitative consideration of such matters is limited to
the type of empirical data analysis performed herein.
While essential for many purposes, this type of treat-
ment does not allow for the consideration of hypothetical
alternatives, as is necessary for control strategy
formulation.
d. EPA should develop both the conceptual framework and the
requisite computer software needed to utilize long-distance
air mass trajectory modeling as an air quality management
tool available to the states. As increasing refinement of
control strategies is necessary, concern with background
transport, regional-scale secondary particulates, and
other large-scale considerations will become increasingly
important.
e. EPA should develop and promulgate at least informally guide-
lines for the use of particulate analysis by microscopy and
other analytical methods. The need to identify sources of
particulate matter will very likely result in an increase
in this work, and a mechanism is needed to assemble and make
available experience with the various approaches.
132
-------
f. EPA should assess and provide information to states on hard-
ware available for and design of special field studies.
OTHER ASPECTS - NAAQS REVIEW
The nature of the control measures mentioned above with respect especially
to many of the "nontraditional" sources of particulates and large-scale
problems will significantly modify the nature of many pollution control
programs. Implementation of some of the control measures will necessarily
extend the scope of control programs into areas of municipal services that
have not traditionally been involved in pollution control, and which will
necessarily be costly. Consequently, they will no doubt require extended
discussion and justification, not only in the eyes of public opinion,
elected officials, and municipal executives, but also in the eyes of many
personnel within the air pollution control community itself.
An important element in future planning concerns the actual standards them-
selves. The NAAQS are currently being reviewed by the National Academy of
Sciences for appropriateness and completeness. While the purpose of this
TSP attainment study did not involve addressing the need for standards re-
view or the ongoing review by the NAS, five specific issues became apparent
in the course of this study. These are listed below with the hope that they
will be considered in the review of the standards for TSP.
1. Particle Size - At present, the particulate standard is based
on the total mass of particulate matter suspended in the air;
there is no concern with particle size, save that the particles
be small enough to remain suspended. (Or actually, just to re-
main suspended long enough to reach the hi-vol, the proximity of
which thus becomes crucial.) Because the health effects of par-
ticles of various sizes differ significantly, and because the
size distribution of particles from various source types differs
significantly, any standard, whether an emission standard or es-
pecially an ambient air standard, that fails to recognize such
differences is unavoidably seen as oversimplified.
2. Particle Toxicity - Differential toxicity among particles of
various chemical nature is an issue very like that of particle
size. Known, obvious differences, such as the distinction be-
tween inorganic soil materials and organic coal tar derivatives,
133
-------
are not included in the basis for the present ambient standard,
except implicitly through the selection of epidemiological
evidence to support the standard. This unavoidably contributes
to the general impression that the standard is oversimplified
in letting important factors "average out." The composition of
the total particulates is already coming under review due to
recent concerns about toxic pollutants.
3. Monitoring Specifications - At present, neither the NAAQS them-
selves nor the specified Federal Reference Method address the
question of how hi-vol monitors should be placed relative to
the sources of the particulate matter. While "common sense"
has probably been adequate in distinguishing between ambient
and source-oriented situations in the case of major point
sources, it is clearly not so in the case of such low-level
dispe'rsed sources as street dust reentraimnent. In order to
be seen as appropriately precise in this area, the standards
should at least take note of the effects of height and distance
from a source such as a street by specifying a reference point
for monitoring and defining relationships by which data from
other points could be adjusted.
4. Time Scale of Standards - At present, the necessary recognition
of differing averaging times is handled by having standards for
both short-term and long-term (annual) averages, and this is
generally considered adequate. However, there are some phenom-
ena that operate on a longer-time period and tend to indicate
a desirability in considering longer-term values as well, such
as perhaps 5-year running averages. Such an approach would
offer one way of resolving the difficulties in handling such
features as meteorologically good and bad years in air quality
planning. It would also help in rationalizing the situation
caused by several-year temporary phenomena, such as major con-
struction, which presently are viewed as anomalies, causing
standards' violations that can be ignored because they are
temporary.
5. Spatial Averaging - Similar to the problem of monitoring speci-
fications is the concern over the Clean Air Act requirement
that each and every area of a city meet the air quality
standards. It can generally be assumed that every city has at
least one corner which can not meet the standards even though
the current monitoring network indicates no violations. In the
same manner, monitors may be placed so as to ignore the problem.
Average TSP levels within a certain area up to a set height may
be more appropriate.
134
-------
APPENDIX A
COMPARATIVE DATA ON THE 14 STUDY CITIES
135
-------
IO
E
220,
ZOO
180
160
140
I 20
too
80
60
40
20
BALTIMORE
NA BIRMINGHAM
+ CINCINNATI
CLEVELAND
A PHILADELPHIA
x ST. LOUIS
Heavily industrialized cities
1957
I960
240
220
2OO
180
160
140
120
100
80
60
40
20
1965
YEAR
1970
1974
o CHATTANOOGA
A DENVER
a PROVIDENCE
*. SEATTLE
Moderately industrialized cities
1957
I960
160
140
I 20
*> '00
E
4 80
a."
£ 60
40
20
1965
YEAR
1970
1974
O MIAMI
OKLAHOMA CITY
A SAN FRANCISCO
* WASHINGTON, D.C.
-A
Lightly industrialized cities
1957
I960
1965
YEAR
1970
1974
Figure A-l. Annual geometric mean TSP trends at NASN sites
136
-------
Table A-l. RANKINGS OF 14 CITIES BY METEOROLOGICAL/CLIMATOLOGICAL PARAMETERS'
Annual
precipitation
(in.)
Denver 15.5
San Fran. 19.5
Okla. 31.4
Clev. 35.0
St. Louis 35.9
Seattle 38.8
Wash.D.C. 38.9
Cin. 39.0
Phil. 39.9
Bait. 40.5
Prov. 42.8
Chat. 51.9
Birm. 53.2
Miami 59.8
No. of days with
precip > 0.01 in.
San Fran. 62
Okla. 82
Denver 88
St. Louis 109
Wash.D.C. Ill
Bait. 112
Phil. 116
Birm. 118
Chat. 121
Prov. 124
Cin. 129
Miami 129
Clev. 156
Seattle 161
Heating
degree days
(base 65°F)
Clev. 6154
Denver 6016
Prov. 5972
Seattle 5185
Cin. 5070
Phil. 4865
St. Louis 4750
Bait. 4729
Wash.D.C. 4211
Okla. 3695
Chat. 3505
San Fran. 3044
Birm. 2844
Miami 206
Surface
wind speed
(mph)
Chat. 6,3
Birm. 7.4
Denver 9.0
Cin. 9.1
Miami 9.1
Wash.D.C. 9.3
Seattle 9.3
St. Louis 9.5-
Bait. 9.5
Phil. 9.6
San Fran. 10.5
Prov. 10.8
Clev. 10.8
Okla. 12,9
Low- level
inversion
frequency^5
(% total hours)
Chat. 40
Denver 39
San Fran. 38
Birm. 37
St. Louis 36
Okla. 35
Cin. 31
Seattle 30
Clev. 25
Wash.D.C. 24
Bait. 24
Phil. 23
Prov. 22
Miami 10
No. of episode days
in 5-year period0
(Mixing ht. < 1000 m;
wind speed < 4.0 m/sec)
San Fran. d
Seattle 44
Denver 27
Birm. 22
Chat. 20
Wash.D.C. 20
Bait. 18
Phil. 13
St. Louis 12
Clev. 11
Cin. 10
Prov. 2
Miami 0
Okla. 0
u>
aExcept for episode days, data are mean annual values.
Obtained from Hosier, C. R., Low-Level Inversion Frequency in the Contiguous United States.
Monthly Weather Review, Vol. 89, September 1961, Fig. 1.
°0btained from Holzworth, G. C., Mixing Heights, Wind Speeds, and Potential for Urban Air Pollution
Throughout the Contiguous United States, EPA, Office of Air Programs Publication No. AP-101.
dlndex not quantitatively applicable to San Francisco.
-------
Table A-l (continued). RANKINGS OF 14 CITIES BY METEOROLOGICAL/CLIMATOLOGICAL PARAMETERS'
u>
oo
Mean wind speed through
surface layerc (m/sec)
AM
San Fran. 3.0
Denver 4.2
Chat. 4.5
Birm 4.6
Wash. D.C. 4.9
Bait. 5.1
Seattle 5.2
Clev. 5.2
Miami 5.3
St. Louis 5.5
Phil. 6.0
Cin. 6.0
Prov. 7.2
Okla. 7.8
PM
Seattle 5.4
San Fran. 5.5
Birm. 5.5
Chat. 5.5
Denver 6.3
Miami 6.5
Cin. 6.7
Wash. D.C. 6.8
Bait. 7.0
St. Louis 7.0
Clev. 7.4
Phil. 8.0
Prov. 8.2
Okla. 8.4
Mean mixing height0 (m)
AM
Denver 268
Okla. 377
Birm. 380
St. Louis 390
Chat. 440
Cin. 480
San Fran. 500
Wash. D.C. 528
Clev. 540
Bait. 620
Prov. 690
Seattle 705
Phil. 750
Miami 923
PM
San Fran. 1000
Prov. 1050
Seattle 1175
Clev. 1250
Phil. 1350
St. Louis 1350
Miami " 1351
Okla. 1382
Cin. 1390
Bait. 1450
Birm. 1500
Wash. D.C. 1570
Chat. 1600
Denver 2543
Median X/Q for 10-km cityc
(sec/m)
AM
Denver 17
Chat. 12.5
St. Louis 12
San Fran. 12
Birm. 12
Wash. D.C. 12
Cin. 11
Clev. 11
Phil. 10.5
Bait. 10.5
Okla. 10
Seattle 10
Miami 10
Prov. 9
PM
Denver 10
Birm. 10
Chat. 10
Seattle 10
San Fran. 10
Miami 10
St. Louis 9
Clev. 9
Cin. 9
Okla. 9
Wash. D.C. 9
Phil. 9
Bait. 9
Prov. 9
Except for episode days, data are mean annual values.
^Obtained from Holzworth, G. C., Mixing Heights, Wind Speeds, and Potential for Urban Air Pollution
Throughout the Contiguous United States, EPA, Office of Air Programs Publication No. AP-101.
-------
20
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Food
Textiles
Apparel
Lumber
Furn iturc
Paper
Printing
Chenicals
Petroleum
Ru!;ber ami Plastics
Lc at. her
St. one, Clay, Class
Primary Metal
Fabricated Metal
Machinery
Electrical Equip.
Transportation Equip.
1
I BALTIMORE
J
I
|
1
I
1
|
J
[
1
1
|
1
ill i i i i
BIRMINGHAM
CHATTANOOGA
10 20 30 40 50 60 70
1.0 20 30 40 50 60 70
10 20 30 40 50 60 70
20 Food
22 Textiles
23 Apparel
24 Lunber
25 Furniture
26 Paper
27 Printing
28 Chemicals
29 Pclroleum and Coal
30 Rubber and Plastics
31 Leather
32 Stone, Clay, Glass
33 Primary Metal
34 Fabricated Metal
35 Machinery
36 Electrical Equip.
37 Transportation Equip.
CINCINNATI
1
I i I
CLEVELAND
I
DENVER
10 20 30 40 50 60 70
10 20 30 40 50 60 70
10 20 30 40 50 60 70
Figure A-2. AQCR manufacturing employment (thousands of employees)
-------
20 food
22 Textiles
23 Appai-el
24 Lumber
25 Furniture
26 Paper
27 Printing
23 Chemicals
29 Petroleum and Coal
30 Rubber and Plastici
31 Leather
32 Stone, Clay, Clam
33 Primary Metal
3A Fabricated Metal
35 Machinery
36 Electrical Equip.
37 Transportation Equip,
MIAMI
20 30 40 50 60 70
OKLAHOMA, CITY
J
PHILADELPHIA
10 20 30 40 50 60 70
_L
J
10 20 30 40 50 60 70
20 Food
22 Textiles
23 Apparel
24 Lumber
25 Furniture
26 Paper
27 Printing
26 Chemicals
29 Petroleum and Coal
30 Rubber and Plastics
31 Leather
32 3ton», Clay, Glass
33 Primary Metal
34 Fabricated Metal
35 Machinery
36 Electrical Equip.
37 Transportation Equip,
PROVIDENCE
J
10 20 30 40 50 60 70
ST. LOUIS
I I II
SAN FRANCISCO
10 20 30 40 50 60 70 10 20 30 40 50 60 70
Figure A-2 (continued). AQCR manufacturing employment (thousands of employees)
-------
20
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Food
Textiles
Apparel
Lur.be r
Furniture
Pa pur
Printing
Chemicals
Petroleum and Coal
Rubber and Plastics
Leather
Stone, Clay, Class
Prir.iry Metal
Fabricated Metal
Machinery
Electrical Equip.
Transportation Equip.
|
-i SEATTLE
-J ,
Z_I
_-
_u
1
J
1
n
i
i
i
f
i i i i i h ,
10 20 30 40 50 60 70
20 Food
22 Textiles
23 Apparel
24 Lunber
25 Furniture
26 P.ipcr
27 Printing
28 Chcraicnls
29 Petroleum and Coal
30 Rubber and Plastic*
31 Leather
32 Stone, Clay, Glass
33 Primary Metal
34 Fabricated Metal
35 Machinery
36 Electric*! Equip.
37 Transportation Equip.
DISTRICT OF
COLUMBIA
10 20 30 40 50 60 70
Figure A-2 (continued). AQCR manufacturing employment (thousands of employees)
-------
6 a
Table A-2. REGULATIONS APPLICABLE TO FUEL-BURNING SOURCES' RATE OF EMISSIONS, Ib particulate/10 Btu
Baltimore, residual oil
Baltimore, solid fuel
Birminr.linm
Chattanooga
Cincinnat i
Cleveland
Denver
Miami
Oklahoma City
Philad,>lphiat
Provi done o
St . Louis
San Francisco
h
Seattle
Washington, D.C.
Arithmetic mean
Standard deviation
Fuel input, 10 Btu/hr
0.01
0.05
0.08
0.50
0.60
0.40
0.60
0.50
-
0.60
0.20
0.20
0.60
0.25
0.17
0.13
0.35
0.21
1.0
0.05
0.08
0.50
0.60
0.40
0.60
0.50
-
0.60
0.20
0.20
0.60
0.25
0.17
0.13
0.35
0.21
4.0
0.05
0.08
0.50
0.60
0.40
0.60
0.35
-
0.60
0.20
0.20
0.60
0.25
0.17
0.126
0.34
0.21
10
0.04
0.08
0.50
0.60
0.40
0.60
0.27S
-
0.60
0.20
0.20
0.60
0.25
0.17
0.102
0.33
0.21
20
0.04
0.08
0.37
0.50
0.33
0.52
0.23
-
0.55
0.20
0.20
0.53
0.25
0.17
0.086
0.29
0.18
40
0.04
0.08
0.27
0.42
0.27
0.45
0.19
0.10
0.4?
0.20
0.20
0.47
0.25
0.17
0.073
0.24
0. 14
70
0.03
0.08
0.21
0.36
0.22
0.42
0.16
0.10
0.39
0.20
0.20
0.42
0.25
0.17
0.064
0.22
0.13
100
0.03
0.08
0.18
0.33
0.20
0.39
0.15
0.10
0.35
0.20
0.20
0.40
0.25
0.17
0.059
0.21
0.12
200
0.03
0.08
0.13
0.28
0.17
0.33
0.13
0.10
0.29
0.20
0.20
0.35
0.25
0.17
0.050
0.18
0.10
400
0.03
0.05
0.12
0.23
0.13
0.29
0.11
0.10
0.25
0.20
0.10
0.32
0.25
0.17
0.043
0.16
0.09
700
0.03
0.05
0.12
0.21
0.11
0.26
O.JO
0.10
0.22
0.20
0.10
0.29
0.25
0.17
0.037
0.15
0.08
1,000
0.03
0.05
0.12
0.18
0.10
0.24
0.10
0.10
0.20
0.20
0.10
0.26
0.25
0.17
0.034
0.14
0.78
3,000
0.03
0.05
0.12
0.14
0.10
0.19
0.10
0.10
0.14
0.20
0.10
0.22
0.25
0.17
0.027
0.13
0.07
10,000
0.03
C.05
0.12
0.10
0.10
0.15
0.10
0.10
0.10
0.20
0.10
0.18
0.25
0.17
0.02
0.12
0.06
100,000
0.03
0..05
0.12
0.10
0.10
0.15
0.10
0.10
0.10
0.20
0.10
0.18
0.25
0.17
0.02
0.12
0.06
a 6
Appropriate conversions were made where allowable emissions were not directly expressed as lb/10 Btu.
Regulation for existing sources.
C0hio's regulation is the same as Cincinnati and was applicable to Cleveland sources as of July 1, 1975.
-------
Table A-3. REGULATIONS APPLICABLE TO GENERAL PROCESS SOURCES' RATE OF EMISSIONS IN 14
STUDY CITIES, Ib particulate/hra
City
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
St. Louis
San Francisco
Seattle
Vashlngton, D.C.
Arithmetic mean
Standard deviation
100
0.46
0.56
0.551
0.551
0.551
0.56
0.56
0.551
0.46
0.551
0.551
0.551
0.60
0.24
0.52
0.09
400
1.50
1.32
1.40
1.40
1.40
1.32
1.32
1.40
1.50
1.40
1.40
1.40
1.50
1.50
1.41
0.07
1,000
2.80
2.34
2.58
2.58
2.58
2.34
2.34
2.58
2.80
2.58
2.58
2.58
2.80
2.80
2.59
0.17
4,000
5.93
5.52
6.52
6.52
6.52
5.52
5.52
6.52
5.93
6.52
6.52
6.52
6.50
5.93
6.18
0.43
Process weight rate, Ib/hr
10,000
10.0
9.76
12.0
12.0
12.0
9.74
9.74
12.0
10.0
12.0
12.0
12.0
12.0
10.0
11.09
1.09
40,000
28.30
23.0
30.5
30.5
30.5
23.0
23.0
30.5
28.3
30.5
30.5
30.5
30.5
28.3
28.42
3.07
100,000
44.58
32.37
44.6
44.6
44.6
32.37
32.37
44.6
40.0
44.6
44.6
44.6
55.0
40.0
42.06
6.26
200,000
51.28
36.17
51.2
51.2
51.2
36.17
36.17
51.2
40.0
51.2
51.2
51.2
75.50
40.0
48.12
10.33
400,000
58.51
40.41
58.0
58.51
58.51
40.41
40.41
58.51
40.0
58.51
58.51
58.51
92.0
40.0
54.34
14.0
700,000
64.76
44.19
66.0
64.76
64.76
44.19
44.19
64.76
40.0
64.76
64.76
64.76
104.75
40.0
59.76
16.98
1 x 106
68.96
46.79
69.0
69.0
69.0
46.79
46.79
69.0
40.0
69.0
69.0
69.0
114.0
40.0
63.31
19.08
4 x 106
86.90
58.41
90.0
86.9
86.9
58.41
58.41
86.9
40.0
86.9
86. 9
86.9
138.0
40.0
77.97
25.33
3Numbers are approximate due to rounding and conversion of units.
-------
Table A-4. REGULATIONS APPLICABLE TO INCINERATORS'
RATE OF EMISSIONS IN 14 STUDY CITIES ,a>b
lbs/100 POUNDS OF REFUSE CHARGED
Cities
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
St. Louis
San Francisco
Seattle
Washington, D.C.
Mean
Tons /day capacity
1.0
0.0285C
0.20
0.40
0.20
0.20
0.1425
0.095
0.285
0.095°
0.076
0.285
0.1425
0.095
0.076
0.145
2.5
0,0285C
0.20
0.40
0.10
0.10
0.1425
0.095
0.19
0.095°
0.076
0.19
0.1425
0.095
0.076
0.145
> 50
0.0285
0.10
0.40
0.10
0.10
0.1425
0.095
0.19
0.095
0.076
0.19
0.1425
0.095
0.076
0.139
both state and local regulations were
applicable, the more stringent control is
assumed.
Numbers are approximate due to rounding
and conversion of units.
°New incinerators in this size range are
prohibited.
144
-------
Table A-5. TYPES OF FUGITIVE PARTICUIATE REGULATIONS IN STUDY CITIES
Cities
Kcco^ni t ion
Nuisance
Baltimore X
j
Birmingham
Chattanooga ) X
Cincint.ati . X
i
!
Cleveland X
Denver : X
Miami !
Oklahoma City
Philadelphia
Providence
St. Louis
San Francisco
Seattle
Washington, D.C.
X
X
Reasonable
precaution
X
X
X
X
a
X
a
X
X
X
X
X
Property line
Prohibition
X
X
X
a
a
Opacity
X
X
a
Concentration
X
&
Specific requirements
Type of source or control
Operating procedures, control equipment, and
maintenance requirements for coke ovens;
sealing of buildings and treatment of exhaust
Building emissions if a nuisance, building
cleaning and repair, spray application of
fiberateJ cementitious products
o
Building emissions if a nuisance
Fugitive dust from unpaved roads, construction/
demolition, mining, land development is regu-
lated; permits are required3
Hot mix axphalt plant; operating practices for
material handling
Building emissions
*State regulation; all other designations refer to local agency regulations
-------
APPENDIX B
MICROSCOPIC FILTER ANALYSIS
In all of the case study cities except Miami, where only NASN filters
were available, members of the GCA study team acquired hi-vol filters
from the 1974 filter banks of the cognizant state and/or local agencies.
In addition, several filter samples for 1974 and selected earlier years
were obtained from the federal filter bank. The filters were selected
to be as representative as possible of specific site and neighborhood
types - industrial, commercial, residential - and of periods of both
high and moderate TSP levels. The results of the analyses were used in
the process of analyzing the TSP problem in each city, within the
bounds of the reliability of the results, and were also aggregated into
composite results for each site, each city, and ultimately a national
composite. The primary presentation of the analytical results is in-
cluded in the separately-bound Volume II of the study report. The pur-
pose of this Appendix is to provide within Volume I a more detailed sum-
mary of the results than is appropriate in the main text; in particular,
the results of the quality control replications are included to provide
some perspective on the reliability of the results.
SUMMARY OF RESULTS
Occasional large differences in replicate results warn against relying
very heavily on the results of the analysis of any one filter. However,
a random match-up between analyst and filter sample minimizes any sys-
tematic bias in composited results, particularly with large numbers
of filters included. For summary purposes, only such composited
146
-------
results are presented hear although they should not be construed to be
more than semi-quantitative in nature. Table B-l presents citywide
summaries, and Table B-2 presents a composite of the results from all
14 cities. The results are given in percent by weight of the visible
material seen by the microscopist.
It is apparent that in nearly all of the study cities the predominant
component is mineral material. In Oklahoma City and Denver, two areas
with dry climate and acknowledged fugitive dust problems, minerals com-
prised more than 80 percent of the observed particulate. Even in areas
of high industrial fuel usage and substantial heating requirements, the
mineral components predominate. The products of combustion, though
significant, generally comprised only about one-quarter of the observed
particulate. In only two cities, Cincinnati and Cleveland, did the con-
tribution of combustion products exceed 40 percent. Biological material
generally was not a major contributor to the observed particulate matter,
except in Chattanooga where over 15 percent was pollen and plant tissue.
The average amount of rubber identified was highly variable from city to
city. San Francisco particulate was reported to average 16 percent
rubber, while two cities, Philadelphia and Washington, B.C. were repor-
ted to average only 2 percent rubber.
In Table B-2 all filters reported for the individual cities in Table B-l
have been combined. As stated previously, the predominance of minerals
(65 percent) is apparent. Of these minerals, two components - quartz
and calcite - account for 50 percent of the observed mass. The 25 per-
cent of the mass attributed to combustion products is comprised of about
equal quantities of oil soot, coal soot, miscellaneous soot, and glassy
fly ash. This 14-city composite also shows a 3 percent contribution
from biological materials and a 7 percent contribution from rubber.
In Table B-3 results from all the filters except the NASN filters have
been composited by site classification. These results show the highest
147
-------
percent component at the various site types to be: minerals at under-
developed (nonurban) sites, combustion products at industrial sites,
biological materials at residential sites, and miscellaneous (princi-
pally rubber) at the commercial (CBD) sites. The average percentages
in Table B-3 were applied to the average TSP concentration to provide
the rough estimates of mass contribution by component that were pre-
sented in Table II-6 of Section II.
QUALITY CONTROL RESULTS
The hi-vol filters obtained during the study visits were first returned
to a clean room at GCA/Technology Division offices for a visual inspec-
tion for artifacts or any evidence of sampler or filter malfunctions.
The filters selected for analysis were assigned a randomly-generated
five-digit number, which then subsequently served as the only identi-
fier for the filter. Hence the microscopist examining the particles
from the filter had no information concerning the city, site, TSP
loading, or possible local sources associated with the sample. In
addition, filters from at least two different cities were included in
each batch provided to the analysts, so that any potential carry-over
bias among filters in any one batch would be discouraged. The fact
that the analysts nevertheless found results that were consistent in
terms of similarities and differences among days, sites, and cities
indicates that advance knowledge of the circumstances of the sample is
not necessary for successful particle identification.
On the other hand, some significant concern about the quantitative es-
timation of various particulate components was caused by the results
of replicate analyses on some of the filters. Two different labora-
tories participated in the microscopic analyses, with Walter C.
McCrone Associates being responsible for the analysis of 400 filter
samples and Eastern Analytical Laboratories, Inc. being responsible
for the analysis of 45 filter samples. The use of randomly generated
148
-------
identifying numbers not only made the basic analyses as objective as
possible, but also permitted.submission of a large number of replicate
samples without the advance knowledge of the microscopists. Since
both laboratories utilized more than one analyst, these arrangements
resulted in as many as four microscopists observing samples from the
same filter and, in some cases, the same analyst examining replicate
samples from the same filter as many as 3 times.
Table B-4 summarizes the differences between results reported by dif-
ferent analysts and by the same analyst when examining samples from
the same filter. The two sections of the table, above and below the
diagonal, provide the results for minerals and combustion products,
the two classes of particles which together typically account for
about 90 percent of the filter loadings. Table B-4 shows some very
large average differences in the results reported, both by different
analysts and by the same analyst. It appears, however, that the re-
sults of replicate analyses by a single analyst are more consistent
that comparisons between analysts. Analysts A, B, and C were from
one laboratory and analysts D and E from the other laboratory.
Analyst D, who consistently differed substantially from the other anal-
ysts, was used very little during the program.
Volume II contains a more detailed presentation of the quality control
data. The balance of this appendix summarizes the data by analyst and
presents the statistical analysis of the replicate results. Table B-5
summarizes the interanalyst comparisons for two categories, minerals
and combustion products. It is apparent from the means that some very
sizable differences occur in the general run of two analysts' results.
To provide some quantitative perspective, a simple t-test was made on
the matched data from each pair of analysts. (The calculated Student's
t-vj'lae presented in Table B-5 was calculated in such a way as to take
account of the matched-pair nature of the data, so that the variability
of the various filters does not inflate the t-value.)
149
-------
Table B-6 presents the significance probabilities associated with the
t-tests, and the verbal description of the meaning that we would attach
to each probability. Although the assumptions underlying the t-test
are not strictly met by these data, the results of the comparisons are
sufficiently clear-cut that we have no important concern on that matter.
Only three of the eight pairs for minerals (AB, CD, CE) and two for
combustion products (AB, CD) exhibit sufficient consistency such that
one can reasonably attribute the difference to chance. Additionally,
two of the five pairs that do show agreement had only two comparisons
so the ability to illustrate disagreement is limited.
Table B-7 summarizes the intra-analyst comparisons for the three micros-
copists that examined replicate samples from the same filter. In gen-
eral the results were more consistent than were the comparisons between
two analysts, but there were some distressingly large differences.
Given the inconsistencies documented above, it was decided to look at
the more detailed components of the microscopy results for one pair of
analysts to see whether any ready explanation could be found for some
of the inconsistency. Table B-8 presents the results of this detailed
analysis, involving analysts B and C. Though this one case can by no
means be presented as being conclusive, it does bring up certain curiosi-
ties. Looking at the data in Table B-8, it is indicated that analyst B
had, on the average, higher mineral readings than analyst C. Analyst C
had higher readings for combustion products, and to a lesser extent
biologicals, than did analyst B. The major contribution to these dif-
ferences were calcite and quartz for minerals and incinerator fly ash
for combustion products.
150
-------
Table B-l. CITY-WIDE COMPOSITE SUMMARY OF FILTER ANALYSES IN 14 STUDY CITIES
Ln
City
No. of filters
Components
Minerals .
Quartz
Calcite
Feldspars
Hermite
Mica
Other3
Combustion
Products
Soot:
Oil
Coal
Misc. soot8
Classy
fly ash
Incinerator
fly ash
Burned wood
Burned paper
Magnetite
Carbon black
Other3
Biological
Material
Pollen
Spores
Paper
Starch
Misc. plant
tissue
Leaf
trie homer
Miscellaneous
Iron or steel
Rubber
Other*
Baltimore
27
Quantity,
percent
Average
(69)
31
18
3
15
<1
2
(25)
9
5
4
6
<1
<1
<1
I
( 3)
<1
<1
<1
1
2
( 3)
1
2
Range
52-88
10-50
2-41
0-6
2-46
0-46
11-61
0-60
0-52
0-50
0-42
0-12
<1-11
0-8
0-10
0-26
0-10
0-26
Birmingham
22
Quantity,
percent
Average
(66)
25
24
<1
17
<1
<1
(22)
4
2
12
3
<1
<1
1
( 2)
<1
<1
<1
<1
2
(10)
2
8
Range
14-90
8-67
3-52
0-4
3-65
0-2
0-3
2-86
0-86
0-10
0-71
0-20
0-15
0-8
0-8
0-50
0-25
0-50
Chattanooga
21
Quantity ,
percent
Average
(36)
10
16
8
2
<1
(35)
7
14
5
9
(16)
7
<1
<1
1
8
(13)
<1
13
Range
3-96
<1-30
2-93
0-20
0-12
0-3
8-78
0-40
0-40
0-76
0-26
0-90
0-25
0-2
0-2
0-5
0-80
0-45
0-4
0-45
Cincinnati
20
Quantity,
percent
A ve ra ge
(51)
19
22
3
6
<1
<1
(44)
9
6
<1
24
4
<1
<1
<1
( 1)
<1
<1
<1
<1
1
( <0
-------
Table B-l (continued). CITY-WIDE COMPOSITE SUMMARY OF FILTER ANALYSES IN 14 STUDY CITIES
Ui
City
No. of filters
Components
Minerals
Quartz
Calcite
Feldspars
Hematite
Mica
Other*
Combustion
Products
Soot :
Oil
Coal
Misc. eoot*
Classy
fly ash
Incinerator
fly ash
Burned wood
Burned paper
Magnetite
Carbon black
Other8
Biological
Material
Pollen
Spores
Paper
Starch
Misc. plant
tissue
Leaf
trichomer
Miscellaneous
Iron or steel
Rubber
Other3
Oklahoma City
27
Quantity,
percent
Average
(88)
29
45
3
11
<1
( 8)
5
2
1
<1
<1
<1
«D
<:
<1
<1
<1
<1
( 4)
<1
4
Range
63-99
10-65
15-75
-------
Table B-2. U.S. COMPOSITE SUMMARY OF FILTER ANALYSES3
No. of filters
300
Components
Minerals
Quartz
Calcite
Feldspars
Hematite
Mica
Otherb
Combustion Products
Soot:
Oil
Coal
Misc. soot
Glassy fly ash
Incinerator fly ash
Burned wood
Burned paper
Magnetite
Carbon black
Otherb
Biological Material
Pollen
Spores
Paper
Starch
Misc. plant tissue
Leaf trichomer
Miscellaneous
Iron or steel
Rubber
Otherb
Quantity,
percent
Average
(65)
29
21
5
10
<1
<1
(25)
7
5
5
6
2
<1
<1
<1
<1
<1
( 3)
1
<1
<1
1
1
<1
( 7)
<1
7
<1
Range
3-99
-------
Table B-3. ' COMPOSITE SUMMARY OF FILTER ANALYSES BY SITE CLASSIFICATION*
Site classification
No. of filters
No. of sites
Average concentration,
ug/m3
Components
Minerals
Quartz
Calcite
Feldspars
Hematite
Mica
Otherb
Combustion Products
Soot:
Oil
Coal
Misc. sootb
Glassy fly ash
Incinerator fly ash
Burned wood
Burned paper
Magnetite
Carbon black
Otherb
Biological Material
Pollen
Spores
Paper
Starch
Misc. plant tissue
Leaf trichomer
Miscellaneous
Iron or steel
Rubber
Otherb
Commercial
114
29
120
Quantity,
percent
Average
(63)
28
20
6
9
<1
<1
(26)
8
4
7
4
3
<1
<1
<1
( 2)
1
<1
<1
<1
1
( 9)
<1
9
<1
Range
6-99
2-80
1-75
0-35
0-45
0-6
0-5
2-86
0-86
0-34
0-84
0-55
0-45
0-47
0-45
0-2
0-2
0-10
0-17
0-50
0-4
0-50
0-10
Residential
90
21
92
Quantity ,
percent
Average
(65)
31
18
4
12
<1
<1
(25)
9
5
3
6
1
<1
<1
<1
<1
<1
( 4)
2
<1
<1
<1
2
( 6)
<1
6
<1
Range
3-99
-------
Table B-4. SUMMARY OF DISCREPANCIES IN DUPLICATE
ANALYSES
o
60
CD
o
l-H
O
T3
O
h
CX
C
o
H
4J
ra
3
i
o
Analyst
A
B
C
D
E
Minerals (above diagonal)
A
56
(8)
18
(8)
75
(14)
66
(10)
133
(6)
92
(15)
B
21
(14)
25
(7)
13
(7)
94
(8)
124
(18)
C
69
(10)
71
( 8)
8
(
5
(1
9
2)
0
3)
D
149
(6)
128
(2)
11
(
1
D
E
77
(15)
79
(18)
31
(13)
172
(D
48
(2)
49
(2)
Note: Each entry is the average difference between the
results reported for pairs of repeated analyses of
the same filter, expressed as the percentage of the
mean of the two reported results in each case. The
entries along the main diagonal summarize the rep-
licate analyses by the same analyst, while those off
the diagonal are averages of the comparisons between
two analysts. The number of paired analyses in each
comparison is given in parentheses.
155
-------
Table B-5. COMPARISONS OF MICROSCOPISTS: MINERALS AND COMBUSTION PRODUCTS
A B
A C
A D
A E
B C
B E
C D
C E
Minerals
Mean percentage by weight
Standard deviation
Differences between means
Degrees of freedom
t- value
69 78
22 11
9
13
1.6
56 31
26 14
25
9
2.9
70 6
31 2
64
5
5.1
62 37
28 16
25
14
3.0
75 36
9 10
39
7
12.4
77 35
16 14
42
17
11.5
30 5
23 0
25
1
1.6
38 39
10 12
1
12
0.1
Combustion products
Mean percentage by weight
Standard deviation
Difference between means
Degrees of freedom
t- value
22 16
25 10
6
13
1.1
29 43
26 13
14
9
2.3
27 86
32 19
59
5
2.9
31 61
29 18
30
14
3.5
16 40
10 13
24
7
6.3
15 59
12 15
44
17
11.3
38 95
16 0
57
1
5.0
38 59
15 14
21
12
4.0
-------
Table B-6. SIGNIFICANCE OF DIFFERENCES BETWEEN MEANS
Minerals
Analysts
A B
A C
A D
A E
B C
B E
C D
C E
Probability associated
with t
0.10 < P < 0.20
P « 0.02
0.001 < P < 0.005
0.005 < P < 0.01
P < 0.001
P < 0.001
0.30 < P < 0.40
P > 0.90
Difference between means
Not significant
Significant
Significant
Significant
Extremely significant
Extremely significant
Not significant
Not significant
Combustion Products
A B
A C
A D
A E
B C
B E
C D
C E
0.25 < P < 0.30
0.025 < P < 0.05
0.025 < P < 0.05
0.001 < P < 0.05
P < 0.001
P < 0.001
0.10 < P < 0.20
0.001 < P < 0.005
Not significant
Significant
Significant
Significant
Extremely significant
Extremely significant
Not significant
Significant
157
-------
Table B-7. CONSISTENCY OF INDIVIDUAL MICROSCOPISTS
Analyst
Average percent
difference between
observations
Number of
comparisons
Minerals
Analyst A
Analyst B
Analyst E
18
13
49
8
7
2
Combustion products
Analyst A
Analyst B
Analyst E
56
25
48
8
7
2
Biologicals
Analyst A
Analyst B
Analyst E
85
81
34
8
7
2
Rubber
Analyst A
Analyst B
Analyst E
118
117
0
8
7
2
158
-------
Table B-8. CONSTITUENT ANALYSIS: B C
Components
Minerals
Quartz
Calcite
Feldspars
Hematite
Mica
Coal
Combustion
products
Soot:
Oil
Coal
Fine soot
Glassy fly ash
Incinerator
fly ash
Burned wood
Burned paper
Magnetite
Biologicals
B
75
29
34
3
8
16
13
1
2
^~
2
C
36
16
6
12
2
40
6
7
1
27
10
Difference
(B -C)
39
13
28
-9
6
-24
7
-6
-1
2
-27
-8
All numbers represent percent contribu-
tion to the total TSP level and are mean
values
159
-------
APPENDIX C
A CASE STUDY OF LONG-RANGE TRANSPORT
INTRODUCTION
Long-range transport of participates can be significant at particle
sizes of less than about 10 ym in diameter. For such particles, the
normal turbulent motions of the atmosphere may be sufficient to keep
them suspended for considerable periods of time. Because of ever-
changing atmospheric conditions, there is, of course, no sharp cut-off
in particle size where the effects of gravitational forces are no
longer of consequence. However, for very small particles of diameters
of 1 ym or less, such as the secondary particulates formed by chemical
reactions within the atmosphere, transport over distances of many
miles is common before removal by deposition and washout. Such
long-range transport from very large sources of pollution, such as
major centers of industrial activity, large urban areas, and power
plant complexes, can be significant since dispersion in these cases
may well be inadequate to reduce concentrations to acceptable levels
within reasonable travel distances.
This long-range transport of particulates has often been blamed for
causing excessive levels of TSP. Specifically, the bulk transport of
both primary and secondary fine particulates into the northeastern part
of the United States is believed to be a significant contributor to the
particulate problem of that area. Conditions favorable for transport
into the region from outside sources occur during periods of sustained
southerly or southwesterly flow resulting from either a stagnating
160
-------
or slowly moving anticyclone over the eastern United States or from
the westward extension of the Bermuda High. Contributions are
greatest when the transport flow is restricted to a relatively shallow
layer by an upper level inversion.
The path of transported material is determined by the wind field over the
depth of the atmosphere through which the material has been mixed. By
making suitable assumptions, this path may be calculated in a step-by-
step fashion from wind speed and direction data obtained from synoptic
charts. Uncertainties are introduced by defining the appropriate air
space over which the vectorial mean velocity is to be calculated, es-
pecially if the calculations extend over several days and include periods
when large diurnal variations in mixing height take place. Nonetheless,
the use of computer-calculated trajectories is proving most helpful in
understanding the influence of long-range transport on air quality mea-
surements. This appendix presents an example of such a trajectory
analysis for 3 sampling days at Oklahoma City, ending with a day of ex-
ceptionally high TSP and sulfate levels.
CASE STUDY FINDINGS
The period described extends from the evening of 23 June 1974 through
29 June 1974, the date on which the high particulate concentrations were
observed. The surface weather observations show passage of a cold front
at Oklahoma City late in the evening of the 22nd. This front continued
southward, reaching the Gulf of Mexico by the morning of the 24th. On
the 24th a large elongated high pressure area with NE-SW axis was centered
over Lake Superior and covered the central United States. For the next
few days this high pressure area drifted slowly eastward, with the sur-
face ridge line remaining almost stationary during the 27th and 28th
*
The trajectories were calculated for a 300 to 1000 meter layer using
the trajectory model developed by the Air Resources Laboratories, NOAA,
and were provided by Dr. Lester Machta and Jerome Heffter of ARL.
161
-------
and extending from just west of Chicago to eastern Oklahoma. On the
29th, moderately strong southerly winds (averaging 16.5 miles per
hour) returned to Oklahoma City. Figures C-l through C-4 give the
sequence of surface weather maps at 48-hour intervals for this period,
and Table C-l gives detailed precipitation and wind data from the
Will Rogers International Airport in Oklahoma City.
TSP concentrations were measured at Oklahoma City on June 23rd, 26th
and 29th. Figure C-5 shows the trajectories of air parcels arriving
at Oklahoma City at noon the first 2 sampling days, and at the begin-
ning and at noon of the last sampling day. The average concentrations
observed at six sites on those days were, respectively, 68, 88, and
2
204 vg/m . Sulfate concentrations determined from seven hi-vol filters
3
on the 29th averaged 13.8 yg/m , an unusually high value for that part
of the country.
This analysis indicates that the meteorological event which led to the
high concentrations at Oklahoma City on the 29th is not the southerly
flow associated specifically with the immediate recent history of the
air mass reaching the city. Rather, the high concentrations are the
result of the sequence of weather patterns controlling the earlier
history of the air mass during which it moved south-southwestward from
Michigan, presumably accumulating particulates and sulfate precursors
from the industrial activities along its path.
162
-------
AI*THCR MAP C_^_j Z
lii*. WEATMEff K S
A M i s T r-*.-*
1008 ' - T^t*^
- . ^ f ,
X^rj--
Figure C-i. Surface weather map for 23 June 1974, 0600 CST
Figure C-2. Surface weather map for 25 June 1974, 0600 CST
163
-------
Figure C-3. Surface weather map for 27 June 1974, 0600 CST
-A *. X- - -. '^ ( J*3^°^-kv- *";; ^ / -.-.ahtuvr. ''Z' 'Xv./x'W'
A-vv-* :(~.;^.'37, / -1 ,i iV;^.i. s// "..'> -;^ n?" --^-. i
kl^"r -^-^ ^-'%;-
^V'.^V >^^.^y^ &^/*r ^ '^'&-&
3*-\ \.>»-(?*\ ^<^,^/ *-.( ^-l-^^>1- ;/[
^.-^\ j-\._--.-t ^v.-y.- /-I ..-.- **/#**/ .-/ I
" Xe \ \ / V >'"-'-"-- /.Oklolibma ^-11 / ^.T X-, - f" -W '
i^s. "if °*' " :'- t"/"'-:--»ti°/ *?*--,--"" ^^\
'^ij^'ijjp JQC ''^ V'JS^J
\ "r
r ^
"_ £_..;; L ,j
.. _ ^^
Figure C-4. Surface weather map for 29 June 1974, 0600 CST
164
-------
Table C-l. METEOROLOGICAL DATA FOR PERIOD 22 TO 29 JUNE 1974
Hour
(GST)
00
03
06
09
12
15
18
21
*
Wind
Direction
Speed
Direction
Speed
Direction
Speed
Direction
Speed
Direction
Speed
Direction
Speed
Direction
Speed
Direction
Speed
Average wind
speed
Resultant wind
direction
Precipitation,
inches
Day
6/22
160
11
170
11
200
6
220
14
240
6
10
9
20
8
40
7
9
190
0.02
6/23
30
13
40
12
50
8
30
11
30
11
40
16
30
12
50
5
11
40
0
6/24
40
5
60
6
330
6
20
7
40
10
10
7
40
10
50
8
7
30
0
6/25
70
7
C
120
4
50
9
50
6
50
7
80
7
100
5
6
70
0
6/26
80
7
C
C
170
5
60
10
90
7
110
8
110
6
5
100
0
6/27
110
5
120
5
140
3
180
10
140
5
180
10
170
7
120
6
6
150
0
6/28
120
7
*
150
8
160
6
160
15
200
15
170
10
160
14
140
9
11
160
0
6/29
150
11
150
11
170
11
200
15
200
21
180
21
150
14
180
11
14
180
0
Wind speed is in knots.
Wind direction is direction from which the wind blows in degrees.
165
-------
KEY :
SYMBOL SAMPLING DAY
6/23/74
6/26/74
A 6/29/74
A 6/39/74
6/27
Figure C-5. Trajectories of air parcels arriving at Oklahoma City at noon on three TSP
sampling days (filled symbols), and at midnight at the start of one of the
sampling day (open Triangle). Dates show positions at 24-hour intervals.
-------
APPENDIX D
ASSESSMENT OF TRADITIONAL SOURCE
IMPACT ON TSP STANDARDS ATTAINMENT
The most apparent of the many factors affecting attainment of the TSP stan-
dards are the particulate emissions from three obvious major categories of
pollution sources fuel combustion, industrial processes, and solid waste
disposal operations. These three categories have long been recognized as
significant pollution problems, and have traditionally been the primary
concern of air pollution control efforts. In most of the case study areas,
emissions from these traditional sources continue to be a major concern;
in some of the cities they constitute a major share of the problem, either
generally or only in some areas.
The aggregate traditional source emission estimates and air quality data
for each of the 14 study cities are listed in Table D-l; these emission
levels generally represent the central city or county in each area, with
appropriate modifications made to ensure the comparability of the numbers,
at least within the precision limitations typical of emission inventories.
A comparison of the first two columns of Table D-l indicates that particu-
late emissions from motor vehicles, the only other category routinely in-
cluded in the inventories, is roughly comparable among the cities. A
comparison of emissions with air quality levels in Table D-l suggests that
emissions from the traditional source categories fuel combustion, in-
dustrial processes, and solid waste are important factors in affecting
standards attainment. Unfortunately, the degree to which traditional
sources are- a factor in nonattainment cannot be judged from the emission
estimates, because emissions from other sources are typically not included.
167
-------
Table D-l.
COMPARISON OF EMISSIONS AND ANNUAL TSP
LEVELS IN THE 14 STUDY CITIES
Cities
Heavily
industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Average
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Average
Lightly
industrialized
Washington
Oklahoma City
Miami
San Francisco
Average
Inventoried emissions
estimates , tons/year
Traditional
sources
210,000
110,000
31,600
23,700
190,000
56,100
10,300
9,700
7,300
7,800
5,600
2,600
8,000
4,800
Otherb
6,400
3,400
6,700
5,700
4,200
1,800
5,400°
3,000
5,200
2,200
1,500
3,900
4,900
6,600
1974 ambient TSP
concentration ,
ug/nr
Highest
site
175
144
122
134
158
130
101
131
105
88
102
107
86
74
Citywide
average
116
99
87
95
88
80
94
71
109
63
61
76
67
66
59
52
62
See also Table 3.
^Generally transportation only.
Includes 3,900 tons per year fugitive dust emissions
168
-------
This appendix addresses the interrelationships between traditional source
emissions and the problem of attaining the TSP standards. Therefore, the
following discussion considers not only the actual emissions and air quality
data but also the stringency of controls being applied and the enforcement
of these controls.
TRADITIONAL SOURCE EMISSIONS
The relative contributions of emissions from the three major traditional
source categories are tabulated for each of the study cities in Table D-2.
Comparative judgments concerning either percentage contributions or actual
emission tonnages based on this table should not be interpreted too quanti-
tatively, because the data are subject to possible differences in defini-
tions among the cities as well as the normal difficulties with any emission
inventories. Nonetheless, fuel combustion is of some importance in most
of the cities except Birmingham, Miami, and San Francisco, all relatively
warm and/or primarily gas-burning areas. Similarly, industrial processes
are a significant proportion of traditional sou ce emissions in most of
the cities, including all the heavily industr .alized cities. Solid waste
disposal is currently'a minor share of the traditional source total; it
comprises a significant percentage only in those cities where some incin-
eration is still practiced, including Baltimore and some of the relatively
nonindustrialized areas.
Fuel Combustion Emissions
While the magnitude of fuel combustion emissions depends primarily on the
fuels burned, the degree to which they are or can be controlled also de-
pends on two further classifications: size and type of source. The dif-
ferent fuel combustion sources are typically identified in two size cate-
gories point sources and area sources. The major categories of fuel
combustion point sources are power plants and the industrial and other large
installations that use the energy from fuel combustion on their premises for
space heating or manufacturing products. Area source fuel combustion, on
169
-------
Table D-2. CONTRIBUTIONS OF MAJOR TRADITIONAL SOURCE CATEGORIES IN THE 14 STUDY CITIES
Cities
Heavily
industrial ized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miami
San Francisco
Estimated emissions, tons per year
Fuel
combustion
126,000
15,700
17,100
5,300
67,000
54,000
7,100
7,600
3,200
4,700
5,500
1,800
1,400
800
Industrial
process
losses
77,600
91,300
11,900
15,800
119,000
1,400
3,100
1,900
3,800
600
40
800
5,800
3,500
Solid waste
disposal
- 6,900
2,700
2,600
2,600
3,500
700
100
200
300
2,500
100
0
800
500
Total
210,500
109,700
31,600
23,700
189,500
56,100
10,300
9,700
7,300
7,800
5,600
2,600
8,000
4,800
Relative contribution, percent
Fuel
combustion
.
60
14
54
22
35
96
69
78
44
60
97
69
18
17
Industrial
process
losses
37
83
38
67
63
3
30
20
52
7
1
31
72
73
Solid waste
disposal
3
3
8
11
2
1
1
2
4
33
2
0
10
10
-------
the other hand, involves almost totally residences and small commercial and
apartment buildings, which burn fuel for space-heating in small furnaces
and boilers.
In any city, the amount of particulate emissions from fuel combustion de-
pends on the amount and type of fuel consumed, which is determined in turn
by the degree of urban and industrial development, the types of industries,
and the climate and location of the city. Table D-3 shows, for the 14
cities, the total energy use and the mix between fuels for the industrial
and utility sectors, which could be expected to be primarily point source
emissions. The total industrial energy consumption is highest in Phila-
delphia, San Francisco, St. Louis, and Cleveland, and lowest in Washington,
Oklahoma City, and Miami, as would be expected simply on the basis of city
size and degree of industrialization. Utility usage, which in this table
includes only those power plants within the city limits, is highest in
Baltimore and Philadelphia, older eastern cities where major, centrally
located plants cannot be, or have not been, replaced by newer facilities
in more remote locations.
Those heavily industrialized cities that have the largest emission problem
due to fuel combustion are the cities that use significant quantities of
coal for industrial purposes (Cleveland, Cincinnati) and/or power generation
(Cleveland, St. Louis). Baltimore and Philadelphia, the two cities where
the primary fuel for major point sources is residual fuel oil, are seen to
have a lesser problem from fuel combustion, at least as judged by emission
inventories.
Emissions from fuel combustion are tabulated separately for point and area
sources in Table D-4, to the extent possible considering the different emis-
sion inventory formats available. Table D-5 presents the distribution of
fuel use in residential housing units, the major portion of area sources.
The absolute contribution of area source fuel combustion is determined by
the fuel used and the climate, while the relative contribution is a function
of the magnitude of other sources.
171
-------
Table D-3. POINT SOURCE FUEL CONSUMPTION PATTERNS IN THE 14 CITIES
N>
Cities
Heavily
industrialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miami
San Francisco
Q
Industrial fuel usage
% of total Btu
Gas
63
77
46
20
75
54
69
84
4
35
66
99
59
97
Oil
4
9
52
55
4
6
8
12
95
65
34
< 1
41
3
Coal
33
14
2
25
21
40
23
4
0
0
0
0
0
0
Total 1?
Btu x 10
101.1
56.9
207.9
67.0
106.3
48.0
28.0
28.5
31.8
19.2
3.0
5.4
6.8
126.4
Utility fuel usage
% of total Btu
Gas
0
0
0
0
< 1
100
0
42
0
0
0
99
34
96
Oil
1
0
100
81
1
0
0
5
100
100
84
< 1
66
4
Coal
99
0
0
19
99
0
0
53
0
0
16
< 1
0
0
Total ._
Btu x 101
31.8
0
65.2
98.4
39.5
5.8
0
30.2
0.7
13.1
35.5
35.7
1.2
32.7
Data for SMSA from reference 1.
Data for plants within city limits from reference 2.
-------
Table D-4. COMPARISON OF AREA SOURCE AND
POINT SOURCE FUEL COMBUSTION
EMISSIONS
Cities
Heavily
1 ndus trialized
Cleveland
Birmingham
Philadelphia
Baltimore
St. Louis
Cincinnati
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miami
San Francisco
Emissions, tons per year
Point
source
114 ,500
15,100
9,200
3,200
66,100
54,000
6,900
500
1,100
1,100
3,900
1,500
1,400
200
Area
source
11,500
600
7,900
2,100
900
a
200
7,100
2,100
3,600
1,600
300
a
600
Total
126,000
15,700
17,100
5,300
67,000
54,000
7,100
7,600
3,200
4,700
5,500
1,800
1,400
800
Not separately Inventoried
Table D-5. RESIDENTIAL FUEL USAGE IN 1970
(PERCENT OF TOTAL DWELLING UNITS)3
Cities
Heavily
industrialized
Cleveland
Birmingham
Philadelphia8
Baltimore
St. Louis
Cincinnati
Moderately
industrialized
Chattanooga
Denver
Seattle*
Providence
Lightly
industrialized
Washington, D.C.
Oklahoma City
Miami a
San Francisco
Gas
95
88
46
48
83
87
22
95
28
31
55
93
35
92
Oil
2
1
49
47
12
8
6
1
44
66
35
<1
9
1
Coal
<1
5
2
1
2
2
9
<1
<1
<1
4
<1
<1
<1
Electricity
2
6
3
4
3
3
63
4
28
2
6
7
56
7
SMSA Data are for central county unless indicated
as being for the SMSA
173
-------
Process Emissions
Stack Emissions In any city the magnitude of industrial emissions of
particulates will depend on the degree of industrialization and the types
of industry. Among the 14 cities studied, as shown in Table D-2, inven-
toried industrial process emissions are very significant in the cities
with a high degree of intense industrial activity St. Louis, Birmingham,
and Cleveland, and to a lesser extent in Baltimore, Philadelphia, and Cin-
cinnati. This breakdown of cities suggests that process emissions are
higher in those cities' where a substantial proportion of the heavy industry
is iron and steel, with the associated coking and foundry operations; the
detailed inventories confirm that these are the major contributors.
With the exception of Birmingham, where gas is burned, the correlation be-
tween the estimated emissions from fuel combustion and those from industrial
processes is strong. This strong relationship is expected since heavy in-
dustrial processes and industrial fuel use would be related, and other fac-
tors such as the stringency of control programs could be expected to apply
comparably to both categories. Thus these source categories are jointly
responsible for that portion of the nonattainment problem attributable to
traditional sources; that portion was judged to be of major importance in
Cleveland, St. Louis and Birmingham.
Even in the less industrialized, generally cleaner cities, the industrial
processes need to be assigned a share of the blame for nonattainment prob-
lems. In cities such as Miami, Seattle, and Oklahoma City, process emis-
sions, while obviously much smaller in magnitude than in the industrial
cities, are still a proportionally larger problem than fuel combustion
sources. These emissions are generally from the type of small mineral
industries cement plants, asphalt plants, sand and gravel operations,
and stone quarries that necessarily are located in or near most urban
areas, independent of the degree of heavier industrial activity.
174
-------
Attempts to document reductions in inventoried industrial process emissions
due to control strategies adopted under State Implementation Plans were
successful in only a few cities. Most areas either did not maintain updated
compilations of emissions or had inventories which were not comparable
between years. This latter situation arose especially where control efforts
were fairly recent as a result of the constantly developing program which
was still identifying sources to be included in the inventories; therefore,
previous data bases could not be considered complete.
In those areas where inventories were felt to be sufficient for trend de-
termination, the stated reductions in industrial process (stack) emissions
ranged between 90 percent in Oklahoma City to 1 percent in the Cleveland
AQCR, with some areas reporting small increases in emissions due to growth.
Generally, the reductions were around 20 to 40 percent. The types of industry
that had the greatest reductions in total emissions were usually the same
industries that were originally producing the most emissions (e.g., steel,
coke, foundries).
Fugitive Emissions One type of industrial emission which has been tradi-
tionally recognized but for which only minor control has been pursued to
date is those emissions that are not vented out through stacks. In many
operations, especially where materials handling is prominent, particulates
are generated both inside and outside of plant facilities (but on plant
property) in processes that are either ignored or insufficiently controlled
so that particulate matter (and gaseous pollutants) are emitted into the
working environment. For industrial processes that operate outdoors, such
as coke ovens and rock crushing operations at quarries-, these pollutants
are directly emitted to the ambient air. When such processes are enclosed,
the pollutants in the working environment may escape to the atmosphere through
windows, doors, roof ventilators, or even unsealed cracks in walls. In
any case, these pollutants, when they enter the outside ambient air, are
defined as. fugitive emissions.
175
-------
Fugitive emissions may result from poor maintenance of process equipment,
environmentally careless operations or difficult to control processes.
For example, fugitive emissions can be the result of leakage from warped
doors on coke ovens as well as the charging operation itself. Failure to
observe proper occupational safety procedures in the design and placement
of hoods over process areas causes excessive fugitive emissions as well as
unnecessary exposures for the workers. Storage piles for sand and gravel,
coal, grain, and other materials that are kept in the open become sources
whenever a strong wind blows over them. Similarly., dirt and gravel parking
lots and roads on industrial property can become sources of particulates
due to either wind erosion or traffic.
Fugitive emissions are generally assumed to be small when compared with
stack emissions; however, stack emissions have been coming under control,
with as little as 1.0 to 0.1 percent of the previously uncontrolled emis-
sions being allowed to escape to the ambient air. Therefore, the relative
importance of fugitive emissions has been growing and they may now comprise
a significant portion of nationwide emissions. For example, EPA has es-
timated that total fugitive emissions of particulate from electric arc
furnace charging can be 5 to 50 times the total emissions of the furnace
4
which are emitted from the stack downstream of control devices.
In the course of this study, no attempt was made to survey the industry in
each city to determine the amount of fugitive emissions or the extent of
the impact. However, in the review of monitoring sites, local influences,
including fugitive emissions, were noted. Of the 154 monitoring sites
visited, 20 (13 percent) were felt to be subject to fugitive emissions.
The type of source and exposure varied from monitor to monitor, but most
sources could be classified as either rock and stone crushing, sorting
and storage operations (including cement plants), or primary metals pro-
cessing. A couple of sites were exposed to grain handling operations, and
other sources included a junkyard, lumberyard, and a fabric filter bypass
at a foundry. The impact that such fugitive emissions had on air quality
levels are discussed later in this appendix.
176
-------
Solid Waste Disposal
Of the several methods of solid waste disposal, incineration and open burn-
ing have traditionally been the most popular in urban areas and, therefore,
significant sources of particulate emissions. The larger point sources of
solid waste disposal emissions are the municipal incinerators and occasion-
ally a large industrial installation, while the smaller residential/commercial
incinerators and open burning in dumps are typically viewed as area sources.
The magnitude of particulate emissions from solid waste disposal in a city
depends on the population of the city and the disposal methods commonly
used. The estimates of emissions presented previously in Table D-2 show
that the highest emissions occur in the larger cities and those cities
where older municipal incinerators and/or numerous smaller incinerators
were in operation Cleveland, Birmingham, Philadelphia, Baltimore, St. Louis,
and Providence. In several of the cleaner cities, municipal incinerators
are among the largest point sources of particulate emissions.
While incineration emissions are not a major problem in many areas, they
remain a source category where some improvements can be made since more
attractive options exist. These can be either the development of modern
recycling centers with the use of combustibles for fuel or, in smaller
less urban areas, the use of landfills. The apparent trend in dealing with
municipal waste is clearly in these directions, so that solid waste dis-
posal emissions should continue to diminish as a significant concern in
attaining clean air.
TRADITIONAL SOURCE CONTROL
An integral part of the impact that traditional sources may have on stan-
dards attainment is involved with the nature of the pollution control effort
applied to them. Since fuel combustion, industrial processes and solid
waste disposal have long been perceived as important sources of particulate
emissions, the promulgation and enforcement of regulations and surveillance
177
-------
of these traditional sources have been the major activities of most air
pollution control programs since their inception. While it was beyond the
scope of this study to gather detailed information concerning these control
efforts, a comparison of the stringency of regulations being applied and the
types of surveillance/enforcement programs being pursued provided an impor-
tant understanding of the role of control approaches in TSP standards attain-
ment.
Regulations
Fue1 Combustion Regulations for particulate emissions from fuel combustion
generally specify the allowable emissions in pounds per million Btu heat
input or in grains per standard cubic foot of effluent air (gr/scf). Emis-
sions from small area fuel combustion sources usually have more lenient emis-
sion standards than those of large facilities; often residential and small
commercial furnaces and boilers with less than 1 million Btu per hour heat
input are excluded from the regulations. Some agencies have different emis-
sion standards for the various types of fuel used, thereby allowing more
stringent emission standards for combustion of cleaner fuels and less
stringent standards for dirtier fuels whose emissions are more difficult to
control. Design standards and operating practices may be specified, and
permits to build and operate large sources may be required.
Among the 14 study cities, the stringency of fuel-burning regulations varies
over the entire range of input rates (see Table A-2). The two cities with
the most stringent regulations are Washington, B.C., and Baltimore; Baltimore's
regulations are more stringent for installations up to 20 million Btu per
hour input, and Washington's regulations are somewhat more stringent for
higher input rates. The least stringent regulations of the 14 cities are
those of St. Louis, although Cleveland's were essentially equivalent until
July 1, 1975, when Ohio's regulations became effective. Regulations for
Chattanooga and Oklahoma City are equivalent to those of St. Louis at low
fuel input rates, but tighten to above-average stringency at high fuel
input rates.
178
-------
To understand how these regulations compared with those throughout the
country, the regulations listed in the World's Air Quality Standards,
Volume II, were reviewed to find maximum and minimum stringency regula-
tions. This review took the most and least stringent controls for each
size category and, as such, the maximum and minimum controls do not neces-
sarily reflect the controls in any one jurisdiction. The result of this
9
review indicated that Baltimore, below 3 x 10 Btu/hr, and Washington, D.C.,
above that level, had the most stringent controls in the nation. The max-
imum and minimum control regulations are presented graphically in Figure D-l,
along with the average of the 14 cities studied plus and minus one standard
deviation, represented by the shaded area. For the purpose of comparison,
the allowable emission level for the burning of solid fuels under reason-
ably available control technology (RACT) as defined in Appendix B of the
EPA's Requirements for Preparation, Adoption, and Submittal of Implementa-
tion Plans (40 CFR 51) is also plotted on this figure.
The impact of a change in the stringency of a fuel-burning regulation can
best be seen in Table D-6 where the maximum allowable emissions are given
for various fuel inputs under the complete range of controls. While it
may be impossible to enforce maximum stringent regulations in some areas
due to problems of fuel availability, the tightening of regulations from
minimum to average stringency can be expected to provide a 50 percent re-
duction in total emissions due to fuel combustion.
Table D-6. REIATIVE STRINGENCY OF PARTICIPATE EMISSION
STANDARDS FOR FUEL BURNING SOURCES - ALLOW-
ABLE PARTICULATE EMISSIONS-IN lb/106Btu
Relative
stringency
Minimum
Average
Maximum
Fuel input, 10 Btu/hr
0.01
0.60
0.35
0.05
1
0.60
0.35
0.05
10
0.60
0.33
0.04
100
0.43
0.21
0.03
1,000
0^29
0.14
0.03
10,000
0.27
0.12
0.02
100,000
0.27
0.12
0.02
179
-------
00
o
1.0
CD
"o
o
g 0.10
ui
u.
o
UJ
0.01
I I I I
i 11 i niit i i 11 nj
MINIMUM STRINGENCY
tv.
-------
Industrial Processes Industrial processes result in either stack emissions,
fugitive emissions, or both. Because of the different characteristics of
these emission types and the degree of attention paid to each, they gener-
ally come under base regulations that are quite different. Regulations
covering each of these industrial emission types are discussed in turn below.
Stack emissions Stack emissions from industry are normally regulated by
process weight rate regulations. In this type of regulation, a maximum
emission rate, in pounds per hour, is established for a complete range of
process feeds, also in pounds or tons per hour. The process feed is usually
defined as the total weight of all materials entering the process excluding
any fuels that are consumed (these would be covered under the fuel-burning
regulations) or normal process air. The stringency of the regulation depends
upon the maximum emission rate which can vary with the level of process
feed. Different jurisdictions may define the process to be the complete
transformation of the raw materials to the finished product or simply one
step along the way. Similarly, the emissions may be measured at each stack
or be the total for all stacks. Given the same emission rate for a certain
feed, greater stringency results from more complete aggregation of steps and
stacks.
All of the cities reviewed in this study have a general process weight
rate regulation which is applicable to all industrial processes unless
specifically exempted. Seven of the 14 cities (Chattanooga, Cincinnati,
Cleveland, Oklahoma City, Providence, St. Louis, and San Francisco) have
essentially the same standards, falling between the extremes of strin-
gency exemplified on the low end by Seattle and on the high end (low
emission rate) by Denver and Birmingham up to a process weight of 400,000
pounds per hour and, above that, by Philadelphia and the District of
Columbia (see Table A-3).
As was done with the fuel combustion regulations, the regulations in the
14 cities were compared with those throughout the country using the
World's Air Quality Standards, Volume II, to find maximum and minimum
181
-------
stringency regulations. This review also combined the most and least
stringent controls for each size category and does not reflect the con-
trols in any one jurisdiction. The maximum and minimum are plotted in
Figure D-2, along with the average of the 14 cities studied plus and minus
one standard deviation, represented by the shaded area. Again, the allow-
able emissions from general processes as defined by Appendix B (RACT) is
included on this figure for comparison purposes.
The impact of a change in the stringency of a regulation can best be seen
in Table D-7 where the maximum allowable emissions are given for various
process weight rates under different degrees of control. From this table,
it is apparent that a significant percentage reduction in emissions can
occur if large sources that are currently under minimum control are re-
quired to have average control. Changes in regulations from minimum to
maximum stringency would reduce emissions from sources with process rates
greater than 100,000 pounds/hour 70 to 90 percent. The same order of mag-
nitude, on a percentage basis, would be seen for similar changes in the
regulations for small sources (less than 1000 Ibs/hr). The selection of
such a control measure would be appropriate if there are large contribu-
tions to the emission inventory from these size categories.
Table D-7. RELATIVE STRINGENCY OF PARTICUIATE EMISSION
STANDARDS FOR GENERAL PROCESS SOURCES -
ALLOWABLE PARTICUIATE EMISSIONS (Ib/hr)
Relative
stringency
Minimum
Average
Maximum
Process weight rate, Ib/hr
100
0.60
0.52
0.16
1,000
2.80
2.59
1.40
10,000
15.20
11.09
5.00
100,000
78.1
42.1
22.3
1,000,000
263.7
63.3
34.5
In addition to a general process weight rate regulation, many areas have
process weight regulations for specific sources. For example, in Pennsyl-
vania over 30 different types of industry are regulated for particulates
182
-------
100
oo
u>
10
z
o
UJ
u.
o
1.0
O.I
MINIMUM STRINGENCY
-MAXIMUM STRINGENCY
i i i
i iii
10'
10*
I04 I05
PROCESS WEIGHT RATE, Ib/hr
I06
Figure D-2. Range of stringency of process weight rate regulations (shaded
area represents the average of the regulations in the 14 study
cities plus and minus one standard deviation)
-------
with several processes within each source type requiring different levels
of control. As is usually the case, these source specific regulations
supersede the general regulation. The purpose of these specific regula-
tions is to require different allowable emission rates for the particular
sources usually a lower emission rate for sources which are major con-
tributors to total particulate emissions and for which more stringent con-
trols are possible, but sometimes a higher emission rate for sources which
find controls difficult to apply and are not major problems.
Regulations for control of industrial emissions may also specify maximum
permissible particulate concentrations in the exhaust gas or they may
specify certain operating procedures and/or design requirements which
serve to reduce particulate emissions. The scope of these latter regula-
tions may range from banning certain types of furnaces used in a process
to the designation of the control equipment or control efficiency that may
be employed. To some extent, this type of regulation is easiest to en-
force since the equipment itself is visible proof of compliance; however,
as with any regulation, it cannot be patrolled continuously to ensure that
it is not being bypassed. Similarly; the requirement of control equipment
may provide the needed reduction in total emissions only when operated at
maximum efficiency. Where this regulation is most useful is in control of
fugitive emissions as discussed below.
Fugitive emissions With respect to regulations, fugitive emissions are
aptly named since they have generally escaped the imposition of specific
control measures. Currently, regulations for control of fugitive emissions
may be divided into three basic categories: (a) nonspecific regulations
which recognize that emissions do occur and can be a nuisance; (b) regula-
tions which quantitatively control the amount or type of particulate cross-
ing the property line; (c) regulations that require particular control mea-
sures on different types of sources. Considerable variation exists within
each of these categories among the various jurisdictions due to differ-
ences in wording, and many jurisdictions may use more than one type of
regulation.
184
-------
Under the first category of regulations, in which the problem of fugitive
emissions is recognized, there are two major approaches. Many agencies
will include in their regulations a general provision that bans any emis-
sion that is considered a nuisance or that endangers human health or wel-
fare. Fugitive emissions can be controlled under this type of regulation,
but complaints and the agency's judgment are needed to determine whether a
source is a nuisance. The other form of "recognition" regulation is the
requirement that reasonable precautions be taken to prevent particulate
matter from becoming airborne. Most such regulations are modeled after
the example regulations given in Appendix B of the EPA Requirements for
Preparation, Adoption, and Submittal of Implementation Plans (40 CFR 51)
which also lists measures that are considered to be reasonable. While a
reasonable precaution regulation provides a better stance from which the
air pollution control agency may operate, it still requires judgment in
the determination of what is reasonable and often means that the problem
must first be brought to the attention of the agency. Though this type of
regulation may not be considered the most stringent, many agencies feel
that it is sufficient for enforcement and yet allows them the flexibility
to deal with each situation as it arises.
The second major type of regulation specifies the quality of air crossing
the property line. The property line requirement appropriately separates
consideration of the air quality within the plant property, which is under
the control of the state or federal occupational health code, from the am-
bient air outside the plant property and under the jurisdiction of the air
pollution control agency. The type and degree of property line restriction
varies widely from opacity parallel to the property line (generally equal
to Ringelmann 0 or 1) or simple prohibition on particles crossing the
boundary, to specific quantitative limits on the amount, size, increase
in concentration, or fallout resulting from particulates crossing property
lines. The selection of appropriate limits on these parameters can ef-
fectively limit the TSP levels to prevent violations of the standard. En-
forcement of this type of regulation is relatively easy due to its quanti-
tativeness; however, with the exception of an opacity or total prohibition
185
-------
regulation, the determination of actual compliance involves the set-up and
operation of a monitoring network that could be expensive. Restrictions
on particle size may involve sizing instrumentation, and a regulation which
sets a maximum increase in TSP concentration in the air crossing a facility
necessitates having a monitoring network both upwind and downwind of the
property.
The third category of regulations is sometimes considered a variation or
extension of the reasonable precautions type of regulations. In this case,
instead of stating that reasonable precautions are required of all sources,
certain sources are singled out for control and specific control measures
may be required. Sources often controlled this way are hot mix asphalt
plants; sand, gravel, and stone crushing operations; cement plants; con-
crete batching; mica and feldspar processing; quarrying and mining; coal
handling; and general earth-moving activities such as road or building
construction. In addition to requiring the application of certain control
measures including the use of waters, chemical stabilizers, oil, or covers
and the installation of hoods, fans, fabric filters, and other control
equipment, these regulations may specify standard operating procedures,
as in charging of coke ovens, or maintenance requirements for process
equipment and buildings so that they are properly sealed and vented through
exhaust systems. Of the three major categories discussed, this one is
felt to be most effective and the easiest to enforce. Not only can the
regulatory agency quickly determine compliance with the regulations but
the source is aware of exactly what control is required of it. Should
the source feel that another control could be as effective, the burden
of proof would lie with the source.
The type of regulations for control of fugitive particulate matter than
were applied in the 14 study cities are summarized in Table A-5. Almost
all cities had some form of reasonable precaution requirement, and many
of them also could control fugitive emissions under a nuisance regulation.
Five areas prohibited particulate matter from crossing property lines and
186
-------
three restricted the opacity of the plume. Cleveland set a maximum con-
3
centration of 500 ug/m (1-hour average) for particulates crossing the
property line, and the Commonwealth of Pennsylvania limits the concentra-
tion to 150 particles per cubic centimeter above the background concentra-
tion. Specific sources are controlled in only a few areas and a limited
number of sources come under control.
The regulations generally do not make a distinction between fugitive dust
(see Appendix E) and fugitive emissions. However, reasonable precaution
type regulations primarily deal with fugitive dust, while regulations for
specific sources are usually directed toward fugitive emissions. A pro-
perty line standard may be made applicable to both types of emissions.
As part of this study, regulations for 76 jurisdictions were reviewed,
including those for all 50 states, the 14 TSP cities, and several other
cities and counties. Table D-8 is a summary of the regulatory approaches
to fugitive particulate emissions used by the various jurisdictions; as
might be expected, there is considerable variation within the categories
due to differences in wording and varying degrees of detail. However, the
approaches remain essentially the same. Figure D-3 shows the geographic
distribution of the principal categories.
About two-thirds of the jurisdictions have a regulation requiring that
reasonable precautions be taken to prevent particulate matter from be-
coming airborne. In addition, possible sources and appropriate measures
that would consistitute reasonable precautions are sometimes enumerated.
It is in this respect that the greatest variation among the regulations
appears, since many of the jurisdictions only list possible controls or
nothing at all. About 8 percent of the jurisdictions regulate fugitive
particulate emissions on a source-by-source basis without a reasonable
precautions section.
187
-------
Table D-8. SUMMARY OF REGULATIONS APPLICABLE TO FUGITIVE PARTICULATES
Regulation type
1.
2.
3.
4.
5.
Reasonable precautions required
(a) reasonable precautions regulation
alone
- including restrictions on fugitive
emissions from buildings
- including a property line restriction
(b) reasonable precaution regulation with
additional restrictions on specific
sources
- including a property line restriction
Regulation by specific source
Nuisance regulation alone
No regulation
Other
(a) restriction on particulate concentra-
tion at property line
(b) restriction on the size of particulates
at property line
(c) fallout limitation
TOTAL
Number of jurisdictions
48
8
4
5
11
76
39
9
5
2
3
6
6
2
188
-------
CO
STATE REGULATIONS
| 1 NO REGULATION OR NUISANCE
I 1 REGULATION ONLY
Y/A. REASONABLE PRECAUTION REGULATION
REGULATION BY SPECIFIC SOURCE
mrm RESTRICTION ON PARTICULATE CONCENTRATION
[LIU II AT PROPERTY LINE
mjmil RESTRICTION ON PARTICLE SIZE AT PROPERTY LINE
W//////A FALL OUT LIMITATION
Figure D-3. Geographic distribution of categories of regulations governing fugitive particulates
-------
Restrictions on the particulate concentration at the property line and on
the size of particles allowed to cross the property line, and fallout
limitations are summarized in Table D-9 for ten jurisdictions. Three
basic methods are employed in instituting the particulate concentration
standards given in Table D-9. Hawaii and Kansas specify an allowable in-
crease over the background or upwind concentration; Texas, Missouri, Omaha
and Cincinnati specify a maximum allowable concentration; and three states
(Hawaii, Mississippi and Nevada) have a fallout standard, with the Hawaii
and Mississippi standards being expressed as an allowable increase over
the background. Use of the high-volume air sampler is specified by Hawaii,
Kansas and Missouri.- Missouri also makes use of a soiling index and, like
Illinois, restricts the size of particles allowed to cross the property
line to 40 microns and smaller.
Indiana has three standards for fugitive particulate matter. Tne first
is that the downwind particulate concentration cannot exceed the upwind
or background concentration by more than 67 percent. The second standard
concerns the respirable particle content of the fugitive dust. If more
than 50 percent of the emission is respirable particles, defined as 0.5
to 6.0 microns in size, the allowable increase over the upwind concentra-
tion is reduced. The third standard is that the ground-level ambient
concentration cannot exceed the background concentration by more than
50 micrograms per cubic meter for a 60-minute period. In addition, there
is a visible standard which states that the fugitive dust cannot be visible
crossing the property line unless it can be proved that the three stan-
dards above are being met.
In summary, it should be emphasized that the fugitive emission problem is
mostly a local problem, affected in part by the local meteorology and
terrain. The effectiveness of a regulation should be judged by how much
it reduces emissions and helps meet the primary and secondary ambient
standards. All of the techniques discussed above can be effective with
suitable enforcement. The reasonable precaution approach gives considerable
latitude in interpretation to the local enforcement agency. It is beneficial
190
-------
Table D-9. PROPERTY LINE STANDARDS
Jurisdiction
Standards
Hawaii
Kansas
Texas
Omaha
Missouri
Mississippi
Nevada
Illinois
Cleveland
Indiana
150 (ag/m , above upwind concentration, 12-hour average
3.0 g/m2, fallout above upwind concentration, 14 day
period
3
2.0 mg/m , above background concentration, 60 minute
average
o
100 M-g/rn , 5-hour average
200 |ag/m , 3-hour average
400 ug/m , 1-hour average
500 i-ig/tn , 60 minute average
o
80 M.g/m , 6 month geometric mean
200 M-g/m , 2-hour arithmetic mean
0.4 coh/1000 linear feet soiling index, 6 month
geometric mean
1.0 coh/1000 linear feet soiling index, 8-hour
arithmetic mean
> 40 |-im
2
5.25 g/m , fallout above background
2
2 tons/mi , 24-hour period
> 40 urn
2
500 M-g/m , 60 minute average
See text
191
-------
if the reasonable precautions are clearly defined. Regulation by specific
source is effective near major sources and where the principal contribu-
tors can be identified. This approach would seem to have the least amount
of difficulty in enforcement. Standards for particulate concentration or
fallout at the property line are of particular value in controlling emis-
sions from specific industrial or commercial sources, but require sampling
to determine compliance.
Solid Waste Disposal (Incineration) Regulations on particulate emissions
from solid waste disposal by incinerators specify the allowable emissions
in pounds per 100 pounds of refuse charged for different size incinerators,
in pounds per hour of operation, or in grains per standard cubic foot of
effluent gas per minute. Some jurisdictions also have a design standard
which requires incinerators to be of the multiple-chamber, indirect feed
type with gas washers, and some have operating standards. Chattanooga's
operation standard requires incinerator operators to be knowledgeable on
correct incineration practices by testing and licensing them. Small in-
cinerators are banned entirely in some jurisdictions Baltimore and
Philadelphia have banned incinerators with charging capacities of less than
10,000 pounds per hour. Regulations have banned open burning almost every-
where except for certain purposes such as campfires or fire-fighting
training which will not substantially affect particulate emissions be-
cause of their small size and low frequency. Typically, cities will allow
open burning only when a permit has been issued by the air pollution control
agency, only if certain allowed types of materials will be burned, and only
when pollution potential is low.
The range of variation in the stringency of the incinerator emission regu-
lations for the 14 study cities is shown in Table A-4. Baltimore has the
most stringent regulation of the 14 cities by far, while Chattanooga's is
the least stringent. The stringency of the regulations for open burning
with permits can be varied according to the needs of the jurisdiction by
reducing the number of permits issued and restricting the areas where open
burning with a permit can take place.
192
-------
Surveillance, Compliance and Enforcement Programs
Although the stringency of the regulations that are being applied and en-
forced is obviously important in determining the reductions in particulate
emissions actually obtained, equally important are the enforceability of
the regulations, the strictness of the enforcement, and the manner in which
compliance is determined. As discussed above, the enforceability of regu-
lations affects the ease and speed with which emissions are controlled and
can vary widely with the types of regulations in effect. For example, in
dealing with numerous small incinerators, agencies find a standard requiring
specific equipment (e.g., multichamber, indirectly fed incinerators) much
easier to enforce than an emission standard; in dealing with coke ovens,
Birmingham has found most efficient an operation standard that specifies
maintenance and operating conditions; in dealing with fugitive emissions,
a source-specific regulation is considered much more effective than a nui-
sance regulation.
Similarly, the achievements of control programs depend on the matching of
enforcement activities to the nature of the particulate emissions problem.
Among the 14 cities, the largest actual emission reductions from tradi-
tional sources have been achieved in those cities where surveillance and
enforcement programs are comprehensive and vigorous. The activities of
such programs generally include the following: constant surveillance and
patrols, frequent inspections of problem sources, a general knowledge of
all of the traditional sources, rigorous compliance determination, prompt
action when a violation or upset occurs, issuance of compliance orders that
are strict yet reasonably attainable in the opinion of an appeals board,
and strict enforcement of compliance schedules.
Almost all pollution control agencies conduct visual surveillance of the
traditional sources some on a routine basis and some only in response
to complaints. In addition, almost all agencies respond to complaints
and investigate their causes. More detailed source inspections are per-
formed by most agencies, but the frequency and detail of inspection varies,
193
-------
as does the size of the sources normally inspected. The detailed source
inspections generally consist of walking through a plant, looking over the
processes and air pollution control devices, and checking records. Some
agencies inspect sources only when a permit to operate or alter is issued
or renewed. Other agencies inspect the larger sources and the sources
with problems or variances more frequently, as much as two or three times
a month, and the smaller sources less frequently, depending on size and
problems. Only a few agencies (of those in the 14 cities) require stack
tests of sources at regular intervals; the stack test is either performed
by the agency or, more typically, by the source or its consultants under
the supervision of the agency.
The primary purpose of the surveillance and enforcement programs is the
determination of compliance. Determination of the compliance of sources
with the regulations varies greatly in rigor, ranging from simple calcula-
tions of emissions, based on emission factors and reported control ef-
ficiencies, to more elaborate walk-through inspections of facilities and
equipment and stack tests of emissions. In the State of Illinois, a per-
formance bond may be required of sources which are installing control
equipment the bond to be forfeited if a stack test shows that the source
is not in compliance by its scheduled date of compliance.
Since the compliance date toward which most control programs were working
May 31, 1975 had recently passed at the time of the city visits, the
agencies could provide information on the status of compliance and would
also indicate how the determination of compliance was made. While some
agencies had extensions or initially later compliance dates so that sub-
stantial reductions in emissions were expected in the future, most agencies
in the study cities reported that compliance of particulate sources was
nearly complete with some sources still under specific compliance plans.
However, reports of full or near compliance may not actually mean that all
traditional sources are in strict compliance with the regulations. While
not a major purpose of the present study, some understanding of the methods
194
-------
used for compliance determination was obtained during discussions on over-
all compliance status. Typically, compliance of a source is determined
when a permit to operate is issued or renewed, when new equipment is installed,
or at other points. However, all too often it appears that agencies will
assume, without checking, that the source continues to be in compliance even
though control equipment may be inoperative or deteriorating, process loads
may be changed, or other difficulties may be encountered. Beyond this
concern with follow-up, there was also some apparent difficulty with the
rigor of the compliance determination techniques themselves. Only a few
of the agencies visited conduct or require actual stack tests and then not
on a routine basis. More commonly, compliance is determined on the basis
or walk-through inspections and theoretical calculations based on process
loads, emission factors, control efficiency specifications, and similar
data. While this is obviously appropriate for some sources and control
measures when done by well-trained agency personnel, it is equally in-
appropriate for other, more complex sources with untried control tech-
nology, particularly when done by inexperienced agency personnel. The
overall impression was that the level of effort in compliance determina-
tion was such that, frequently, little confidence could be placed in many
agencies1 determinations of strict compliance with their regulations.
AIR QUALITY IMPACT
While the above discussions have focused upon particulate emissions from
traditional sources fuel combustion, industrial processes, solid waste
and the control programs affecting them, a primary objective is to develop
an understanding of the impact that traditional sources are having on TSP
levels. Specifically, it is important to place traditional sources in a
proper perspective with respect to the problem of standards attainment;
i.e., how do changes in emission levels affect air quality; what is the
current contribution of traditional sources to TSP levels above the stan-
dards; what further reductions in TSP may be expected under additional
control?
195
-------
Because of the wide variety of information available and the complexities
of the situation, several different types of analyses were conducted on the
data collected in the 14 study cities. Each of these analyses is presented
separately below, but together they serve to provide the assessment of
the traditional source impact on TSP standards attainment given in the con-
clusion and summarized in Section III.
Current Emissions
In general, the impact of traditional sources on nonattainment depends very
heavily on the nature of the urban area in question. The 14 case study
areas included both cities where traditional sources totally dominated the
picture and cities where they are not and probably never have been a major
share of the problem. Based on analysis of aggregate emission inventories
and compliance trends, the status of traditional sources in the urban
areas studied have been categorized as follows:
I Traditional sources are still a problem in (1) Birmingham
because many major sources are still on compliance sched-
ules and another 80 percent reduction in traditional emis-
sions is expected; (2) Cleveland because little enforce-
ment has been pursued and most sources are neither in com-
pliance or on schedules; and (3) St. Louis because, al-
though most sources (on the Missouri side) are determined
to be in compliance, the regulations are not very stringent.
II Major traditional emissions have been significantly re-
duced to the point where they are no longer totally domi-
nant in Philadelphia, Baltimore and Cincinnati due to their
long-standing control programs. However, some further re-
duction of traditional source emissions is possible along
with control of nontraditional sources.
Ill Moderate emissions are reasonably well controlled in Chatta-
nooga, Seattle, Providence, and Denver.
IV Traditional sources were likely never a very serious prob-
lem in Miami, Oklahoma City, and San Francisco. Washing-
ton, D.C., no longer has much of a problem with traditional
sources.
196
-------
For the purpose of assessing the TSP attainment problem this type of
breakdown provides an indication of the further reductions that may
reasonably be expected through traditional source control; i.e., those
cities in the first category listed above (major uncontrolled traditional
emissions) should be able to reduce traditional emissions (and TSP levels)
at least to those currently in the second category (major controlled
traditional emissions). If the cities in the first two categories are
expected to be no dirtier than the other cities, almost all of which are
violating the standards, then all cities in the first two categories would
have to reduce their traditional source emissions to the levels in the
bottom two categories.
While the most direct relationship between traditional source emissions
and TSP concentration is one of increasing TSP concentration with increasing
emissions, another simple parameter that addresses the distribution of
sources in an area and the dispersion of emissions is emission density
(tons per year per square mile). Table D-10 provides a listing of the
total tons per year and emission density for the cities under the above
classification and also gives the citywide TSP concentration with the non-
urban particulate levels and sulfate/nitrate excess subtracted out. These
TSP contributions were removed to account for the changes that occur simply
due to different locations around the country (nonurban levels) and those
sulfate and nitrate TSP levels in the city (above those in nonurban areas).
(See Section III for a detailed explanation of the numbers used in this
calculation.) The selection of areas for the calculation of emission
densities was made on a judgmental basis considering the distribution of
major sources, the ambient monitoring network, and the air basin.
As may be seen in Table D-10, considerable scatter and overlap exists
between the total emissions, the emission densities, and the air quality
levels given, due in part to differences in dispersion characteristics
of the areas. However, on an average basis, there is apparently no real
difference between those areas with no major industry (Category IV) and
197
-------
Table D-10. COMPARISON OF TRADITIONAL SOURCE EMISSIONS, EMISSION DENSI-
TIES, AND TSP LEVELS IN THE 14 STUDY AREAS
Cities
Category I
Birmingham
Cleveland
St. Louis
Average
Category II
Baltimore
Cincinnati
Philadelphia
Average
Category III
Chattanooga
Denver
Providence
Seattle
Average
Category IV
Miami
Oklahoma City
San Francisco
Washington, D.C.
Average
Tons per
year
110,000
210,000
190,000
167,000
23,700a
56,100
31,600
37,100
10,300
9,700
7,800
7,300
8,500
8,000
2,600
4,800
5,600
5,250
Tons per year
per square mile
488
335
411
411
240
135
245
207
88
60
30
37
52
30
2
60
92
46
Citywide TSP
minus nonurban and
secondary urban
excess
60
84
55
66
56
42
45
48
31
83
28
42
34
34
40
32
31
34
alncludes both Baltimore City and Baltimore County.
Denver is not included in these averages because of its
adverse, complicating meteorology.
198
-------
those cities with a moderate amount of industry (Category III) but primar-
ily under control. Cities that are fairly heavily industrialized but have
most sources under compliance with moderately stringent regulations average
10 to 15 yg/m3 higher than the lower two categories; those cities that
still need to bring major sources under compliance (Category I) average
30 yg/m3 higher. When a regression analysis was applied to these sets of
numbers (excluding Denver), the following equation was derived (r = 0.769;
standard error = 10.5 yg/m3):
TSP = 7.6 x ED + 31.9
cw
where TSP = the citywide TSP level above the nonurban and secondary
urban excess levels, in |ig/m3
ED = the emission density, in 100 tons/yr/sq mile.
The average values of emission density and TSP levels for each category of
cities are plotted in Figure D-4, along with the ranges, which reflect
the differences that arise because of topography and meteorology. This
figure indicates that even the cleanest cities (with respect to tra-
ditional sources) are still 30 to 40 yg/m3 above the nonurban levels.
3
Therefore, with nonurban levels ranging between 20 to 35 ng/m , only a
very few cities, with favorable meteorology, can be expected to meet the
3
secondary annual standard (guide of 60 p.g/m ) on a citywide basis through
control of traditional sources. Since these levels are citywide averages,
all cities of the size considered here can be expected to have at least
one site exceeding the secondary annual standard. At the same time, this
figure indicates that control of traditional sources is warranted and
3
necessary to meet the primary annual standard (75 ug/m ) when traditional
source emission densities are greater than 100 to 200 tons/year/square mile,
Modeling Studies
Aside from considering the impact of the total traditional sources on TSP
air quality, the relative contributions of each of the major categories
199
-------
90
K)
E
^
o>
CO
to
IU
o
x
UJ
03
o:
O
U
UJ
v>
CO
O.
CO
0.
to
h-
UJ
o
80
70
60
50
40
CD
CC
O 30
z
20
^
| > CATEGORY I
1
CATEGORY II
CATEGORY
_L
_L
0 100 200 300 400 500 600
TRADITIONAL SOURCE EMISSION DENSITY, tons/year/sq. mile
Figure D-4.
Traditional source emission density versus
above nonurban TSP levels
200
-------
need to be understood. While the relative contributions of each of the
categories to the total impact from traditional sources may be expected
to be in direct proportion to the emission inventories, the different
types of emissions (process versus combustion), the spatial distributions,
and the stack heights can have an effect on the eventual contributions.
As part of this effort to assess the attainment of the TSP standards,
modeling studies were performed by EPA for three of the 14 cities -
ft 7 8
Cincinnati, Cleveland, and St. Louis. The studies used past and cur-
rent emission data to determine the emission sources which were most
affecting the TSP concentrations recorded at certain selected TSP sampling
sites. The use of modeling, as opposed to the rollback technique, allows
the position (i.e., distance and height) of sources with respect to the
sampling site to be taken into account. The model predicted the air quality
impact separately for point sources (divided into industrial process, fuel
combustion, and solid waste incineration) and for area sources. Because
all of these cities had over 97 percent of their inventories in traditional
sources, the assumption was made that the area source contribution came
primarily from area source fuel combustion.
Table D-ll presents the results for the three cities, showing the range of
percentage contributions of the different source categories to the above-
background concentrations for all modeled sampling sites in a city and
also to the total traditional source inventory for each area. The closest
correlation on the basis of source category is in solid waste, which is a
small part in any of the cities; on a city-by-city basis, the highest
overall correlation was in St. Louis. In Cincinnati and Cleveland, the
industrial contribution is apparently larger than would be predicted
solely from the emission inventories. Two explanations for this are
evident: (a) the monitoring sites for these cities were more often
located in industrial areas so the sites evaluated would have a greater
impact from industry; (b) the contribution from industrial emissions to
the calculated TSP levels is greater because of the effect of parameters
such as stack height and location throughout the city.
201
-------
to
O
Table D-ll.
MODELED TSP CONTRIBUTIONS VERSUS CONTRIBUTIONS FROM THE
INVENTORY FOR TRADITIONAL SOURCES IN THREE CITIES
Source category
Fuel combustion
Industrial
Solid waste
Cincinnati
Percent con-
tribution above
background TSPa
65 - 85
15 - 34
< 1
Percent of
traditional
sources
96
3
1
Cleveland
Percent con-
tribution above
background TSPa
21 - 42
57 - 78
1-4
Percent of
traditional
sources
60
37
3
St. Louis
Percent con-
tribution above
background TSP3
23 - 65
35 - 76
< 1
Percent of
traditional
sources
35
63
2
Background TSP is the value calculated in the model for the TSP level that cannot be explained due to the
inventoried sources. The values for each of the cities are: Cincinnati 58 ug/iiP , Cleveland 44 ^g/nP,
St. Louis - 32 |.ig/m3.
-------
The impact that fugitive emissions had on individual monitors could
also be determined from these modeling studies. Where the model was
found to provide large underpredictions for certain sites, fugitive
emissions (and other local sources) were prominent influences. These
3
underpredictions were usually in the range of 15 to 40 (j.g/m , center-
ing around 25 ug/m . This range agrees with the analysis presented
later under the discussion of site analysis.
Microscopic Analysis
In addition to the modeling studies, a number of hi-vol filters were
analyzed microscopically for each of the 14 cities to determine the
components of the collected particulates. The component analysis
could then be used in an attempt to identify the sources of the
particulates. One category used during the analyses was combustion
products, including oil soot, coal soot, and glassy fly ash. Table D-12
shows the citywide composite summary of oil and coal combustion products
(average percent by weight) determined during the filter analysis which
gives a rough idea of the relative impact of fuel combustion on the weight
of particles (larger than 1 urn) collected during sampling. The results
show that coal combustion products (coal soot and glassy fly ash) are a
relatively important portion of TSP samples in Chattanooga, Cincinnati and
Cleveland where coal usage is substantial. Oil soot is relatively impor-
tant in Providence and Washington, B.C., which are primarily residential
cities where oil is an important fuel. Comparing the composite combustion
products data with fuel usage data for the 14 cities showed a correlation
between percent oil soot and percent of homes using oil for space heating
(r = 0.699) and a stronger correlation between percent coal combustion
products and percent of industrial fuel usage being coal (r = 0.826).
A comparison of residential coal usage with coal combustion products showed
no correlation.
203
-------
Table D-12. PERCENT WEIGHT OF SELECTED FUEL COMBUSTION
PRODUCTS IN 14 CITIESa
City
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
.Oklahoma City
Philadelphia
Providence
St. Louis
San Francisco
Seattle
Washington, D.C.
Oil
soot
9
4
7
9
9
0
4
5
8
13
1
3
10
13
Coal combustion
products
Coal
soot
5
2
14
6
12
1
3
2
2
2
< 1
9
6
3
Glassy
fly ash
6
3
5
24
15
1
2
1
5
2
2
1
< 1
7
Total
11
5
19
30
27
2
5
3
7
4
< 3
10
< 7
10
Data from composite of microscopic analysis of
selected filters
Monitoring Site Analysis
A major part of this study was the visiting and reviewing of monitoring
sites in the study cities to determine the possible impacts on the TSP
levels being measured. Over 150 such hi-vol sites were reviewed and
the data compiled provided important information on the relative contri-
butions of the various source sectors in different neighborhoods and
different types of cities.
Fugitive Emissions - One of the principal objectives and results of the
monitoring site reviews was to determine the impacts of fugitive emis-
sions and fugitive dust emission (see Appendix E). In the review of
monitoring sites, local influences, including fugitive emissions, were
noted. Of the 154 monitoring sites visited, 20 (13 percent) were felt
204
-------
to be subject to fugitive emissions. The type of source and exposure
varied from monitor to monitor but most sources could be classified as
either rock and stone crushing, sorting, storage operations (including
cement plants), or primary metals processing. A couple of sites were
exposed to grain handling operations and other sources included a junk
yard, lumber yard, and a fabric filter bypass at a foundry.
Though the total level of fugitive emissions from a process may be
small in comparison with the stack emissions, the low height at which
they emanate means that very little dilution occurs before they reach
the breathing zone of those residing or working in areas near the faci-
lity. Even if the emission is only the air inside of a building es-
caping to the outside, the levels can be quite high. The maximum
allowable 8-hour average concentration for nuisance dusts in an indus-
3 3
trial environment is 15 mg/m (5 mg/m respirable), as set by the
Occupational Safety and Health Standards. This is almost 60 times the
primary 24-hour air quality standard for particulates.
Although the degree of impact on ambient levels is dependent on the
total mass of emissions, the extent of the impact depends upon the size
of the particles. Many operations which are primarily materials hand-
ling, crushing, or sorting processes generate particles that are fairly
large. These particles may be expected to fall out or be deposited
close to the source, probably within a radius of a half mile. Other
fugitive emissions, however, may contain fumes or combustion products
which are much smaller, can remain suspended for a long period of time,
and therefore have a much wider range of impact.
Since fugitive emissions are intermittent and local in nature, they
have their greatest impact over short time periods. The data collected
in this study indicated that in Philadelphia in 1974, the highest
3
24-hour TSP level found was 624 (ig/m at a site near a gravel storage
pile and a grain loading operation; the filter was reported to contain
205
-------
high levels of grain dust. This one site, sampling once every 6 days,
was responsible for half of the violations of the primary 24-hour
standard reported for all of the monitors sampling in the City of
Philadelphia. In Oklahoma City, a special field study of fugitive
emissions was conducted by the Division of Environmental Health.
Short-term concentrations on the premises of a concrete and asphalt
3
plant ranged from 121 to 1140 ug/m while the TSP levels at a monitor
3
outside the plant ranged from 43 to 743 ug/m . When the outside mon-
3
itor was upwind of the plant, it averaged 78 ug/m (11 percent of
3
in-plant concentrations); downwind of the plant, it averaged 200 ug/m
(42 percent of in-plant concentrations).
Since fugitive emissions are associated with industry, most of the ex-
posures were found in industrial areas; in fact, 32 percent of all
sites visited in industrial areas were subject to fugitive emissions.
Because major differences exist in the placement and exposure of mon-
itors, a comparison of the average value at sites near fugitive emissions
and at those which were not so obviously exposed in each city gave a
3 3
range of values from over 50 ug/m higher to 10 ug/m lower at the fugi-
tive emission sites on an annual average basis. However, based on the
visits to the monitoring sites it was possible to make judgemental
groupings of monitors to reflect comparable industrial exposures with
and without fugitive emissions. Then the sets of monitors exposed to
3
fugitive emissions averaged 25 ug/m higher than comparable monitors in
the same city not noted for fugitive emissions. The effect of this
3
average difference of 25 ug/m can be seen when the distribution of
values is studied. As shown in Table D-13, when the TSP levels at the
3
industrial monitors are distributed in 25 ug/tn intervals above the
primary standard, those sites in industrial areas that are influenced
by fugitive emissions are the only ones that had annual means more than
150 ug/m , and all of these monitors were above the primary annual
3
standard of 75 ug/m .
206
-------
Table D-13.
DISTRIBUTION OF ANNUAL MEANS AT INDUSTRIAL
SITES WITH AND WITHOUT FUGITIVE EMISSION
INFLUENCES
TSP concentration,
|ig/m3
151 - 175
126 - 150
101 - 125
76 - 100
< 75
Total
Number of industrial sites
With
fugitive
emission
influence
3
4
3
3
0
13
Without
fugitive
emission
influence
0
7
9
9
3
28
Total Impact The review of the monitoring sites also allowed an estimation
of the total impact of emissions from traditional sources, especially major
point sources, by comparing the TSP concentrations at different sites in
the highly urbanized, industrial cities to those in areas with low emission
levels. For example, in Miami's residential areas, where industrial emis-
sions are low and space heating emissions are almost nonexistent, the
3
average level is 50 |ig/m . This type of difference in TSP levels has
already been demonstrated in Table D-10 in which the heavily industrialized
3
cities were shown to have citywide TSP levels 20 to 50 ng/m higher than
those in lightly industrialized areas.
Table D-14 is used to summarize and compare the TSP levels in different
neighborhoods of the study cities among the .ndustrial groupings. As
was done in Table D-10, the nonurban levels were subtracted from the neighbor-
hood values for each city and the resulting levels were averaged for each
of the categories defined above. Again, the TSP levels are highest in the
heavily industrialized areas and lowest in the lightly industrialized
areas when compared on a neighborhood basis. However, while the percent
207
-------
Table D-14. TSP LEVELS AND STANDARDS VIOLATIONS BY CITY CATEGORY
City
category
I
II
IIIs
IV
3
TSP concentration above nonurban levels, yg/m
Residential
Range
29
to
45
26
to
30
18
to
21
17
to
26
Average
36
27
20
23
Commercial
Range
41
to
86
36
to
48
29
to
42
31
to
49
Average
61
41
35
37
Industrial
Range
74
to
113
58
to
71
54
to
62
NA
Average
92
65
58
NA
Percent of sites above standard
Annual
Primary
55%
38%
19%
16%
Secondary
84%
69%
33%
35%
24-hour
Primary
26%
16%
7%
33%
Secondary
90%
45%
30%
70%
O
00
Denver is not included in these numbers because of its adverse, complicating meteorology.
-------
of monitors with violations of the annual standards decreases as one goes
from Category I to Category IV cities, this pattern is not totally consis-
tent for the violations of the 24-hour standards. Therefore, 24-hour stan-
dard violations can logically be assumed to be influenced to a large
extent by other than total emissions; e.g., meteorology, monitor siting.
These data indicate that the traditional sources were still adding about
3
5 u.g/m to levels in residential neighborhoods in those cities where major
traditional emissions had been significantly reduced; the lack of control
of major traditional emissions (Category I) made residential levels another
3
10 fig/m higher. Similarly, levels in commercial neighborhoods of cities
3
in Category II were about 5 ng/m higher than in Categories III or IV but
Category I cities had TSP levels in commercial neighborhoods averaging
3
20 |ig/m higher than Category II cities. The same pattern is followed in
3
the industrial neighborhoods with Category II cities averaging 7 (ig/m
3
more than Category III cities and 27 |j.g/m less than Category I cities.
The consistent increase in TSP levels, regardless of neighborhoods,
between Category II and Category III and IV cities suggests that the problem
in Category II cities is more likely due to areawide sources that have had
insufficient attention while the major point sources were being controlled.
Such sources may be fuel combustion, small industries scattered around the
city, or nontraditional sources that may be related to city size and in-
dustrial activity. The increasing difference in TSP levels between
Category I and Category II cities as one goes from residential neighbor-
hoods to industrial neighborhoods indicates that industrial processes are
still the major concern; this is in agreement with the findings in each
city. Commercial sites have a larger increase than residential sites
because they are usually closer to the industrial areas.
209
-------
CONCLUSIONS
Traditional sources have received the greatest attention from control pro-
grams to date and significant reductions in emissions from these sources
have resulted due to regulations both before and after the requirements for
State Implementation Plans. Much of the reduction in the areawide problem
has come about because of switching to cleaner fuels such as gas and dis-
tillate oil over the past couple of decades. While more could still be
done with respect to fuel combustion where cleaner fuels are available,
control programs under SIP's have primarily directed attention at indus-
tries, power generation, and solid waste disposal.
With respect to the three major categories of traditional sources - fuel
combustion, industrial processes, and solid waste disposal - the relative
contributions very significantly with the nature of the different urban
areas. The fuel combustion contribution to inventoried emissions ranges
from less than 20 percent in clean-fuel, industrialized areas to well over
90 percent in totally nonindustrial areas, with industrial processes
accounting for most of the balance. Solid waste disposal emissions are
generally less than 5 percent.
Traditional sources are still a major concern in many areas. Significant
reductions in TSP levels can occur in these areas through traditional
source control. This potential for reduction has been well demonstrated
in several of the study cities which have undertaken to control traditional
sources and equally well demonstrated by the lack of success in those
cities which have not undertaken such programs.
Not only are more stringent controls warranted in many cities, but signi-
ficant reductions in emissions can be expected simply from the improvement
of surveillance, compliance, and enforcement programs. Generally, further
control of traditional source emissions can be expected to be appropriate
if the traditional source emission density is over 200 tons/year/square
210
-------
mile. When the density falls much below this value, control programs for
other sources may be necessary. As was shown in the review of the 14 study
cities, even cities with very few traditional sources in their inventories
3
cannot be expected to bring citywide TSP levels down below 60 jig/m , the
secondary annual standard, through traditional source control. In addition,
most cities have at least one site violating the primary annual standard
of 75 ug/m .
Though the reductions in TSP concentrations that can be expected under
traditional source control vary widely with the meteorology, monitor siting,
and industry type, if one starts with a poorly controlled, heavily indus-
trialized area, the concentrations in residential, commercial, and indus-
trial neighborhoods can be expected to decrease about 10, 20, and 25 to
3
30 ug/m , respectively, under moderately stringent controls. However, to
3
provide an additional 5 to 10 ug/m reduction in all neighborhoods would
require very stringent controls. Control of fugitive emissions can
3
reduce TSP levels 25 ug/m at individual monitors.
211
-------
REFERENCES
1. U.S. Census of Manufacturers, 1972.
2. Steam Electric Plant Factors. National Coal Association,
Washington, B.C. 1974.
3. U.S. Census of Housing, 1970.
4. Lillis, E.J. and D. Young. EPA Looks at "Fugitive Emissions."
J Air Pollut Contr Assoc. 25_: 1015-1018, October 1975.
5. Martin, W. and A.C. Stern. The Worlds Air Quality Management
Standards, Volume II: The Air Quality Management Standards of
the United States. Office of Research and Development, U.S.
Environmental Protection Agency. Washington, D.C. EPA-650/9-75-
001-b, October 1974.
6. A Modeling Study to Assess Attainment of the S02 and TSP NAAQS
in the Cincinnati AQCR. U.S. Environmental Protection Agency,
December 1975.
7. A Modeling Study to Assess Attainment of the S02 and TSP NAAQS
in the Cleveland AQCR. U.S. Environmental Protection Agency,
January 1976.
8. A Modeling Study to Assess Attainment of the S02 and TSP NAAQS
in the St. Louis AQCR. U.S. Environmental Protection Agency,
March 1976.
212
-------
APPENDIX E
ASSESSMENT OF PARTICIPATES FROM NONTRADITIONAL SOURCES
Even in cities where TSP emissions from traditional sources are relatively
3
minor, citywide averages are 30 to 40 |o.g/m above the nonurban levels, and
the secondary annual standard is being violated. Monitors in apparently
clean areas of a city or in smaller, nonindustrial cities have measured
high TSP concentrations which cannot be explained by complex modeling or
which fail to decrease as expected under controls of the State Implementa-
tion Plans. These findings indicate that a certain level of particulate
is being caused by those sources not traditionally considered in air pol-
lution control strategies. These activities are here collectively desig-
nated "nontraditional" sources.
Nontraditional sources of particulates may be divided into two categories.
One category consists of obvious, distinct sources of emissions; these in-
clude construction and demolition activities, particulate emissions from
tailpipes, and tire wear. The other category refers to the more general
problem of natural and man-induced disruption of particles on the ground
or surfaces in an urban area so that particulates can become entrained or
reentrained. These two categories, specific sources and reentrainment
particulates, form the framework for this Appendix.
Before the discussion of these categories, two terms merit differentiation:
fugitive emissions and fugitive dust emissions. Fugitive emissions include
those particulate (and gaseous) emissions that result from industrial re-
lated operations and which escape to the atmosphere through windows, doors,
vents, etc., but not through a primary exhaust system, such as a stack,
213
-------
flue, or control system. Fugitive emissions may result from metallurgical
furnace operations, materials handling, transfer and storage operations,
and other industrial processes where emissions escape to the atmosphere.
As they are part of the total emissions from industrial operations, they
are discussed under traditional sources in Appendix D. Fugitive dust
emissions, on the other hand, are generally related to natural or man-
associated dusts (particulate only) that become airborne due to the forces
of wind, man's activity, or both. Fugitive dust emissions may include
windblown particulate matter from paved and unpaved roads, tilled farm
lands, exposed surface areas at construction sites, etc. Natural dusts
that become airborne during dust storms are also included as fugitive
dusts. In all, over a dozen sources of fugitive dust emissions due to
man's activities have been studied. These include paved roads; sand on
paved roads; unpaved road; unpaved parking lots; residential, industrial,
and commercial construction; highway construction; agriculture (tilling);
land development; cattle feedlots; off-road recreational vehicles; un-
paved airstrips; quarrying, mining, and tailings; and aggregate storage.
Since the problem of attainment of standards is most prominent in urban
areas, this appendix focuses upon those fugitive dust emissions that occur
in the urban environment. Fugitive dust in rural areas that results from
man's activities or natural causes is discussed primarily under Large
Scale Considerations in Section III.
URBAN REENTSAINMENT
The category of nontraditional sources that has received the most atten-
tion recently is that of reentrainment, especially in urban areas. These
The commonly used term 'resuspended1 is considered inappropriate because
of possible confusion with the standards referring to total suspended par-
ticipates and the difficulty of distinguishing a resuspended from a sus-
pended particulate. In addition, all particles which become entrained may
not remain suspended. While the term 'reentrainment1 may pose similar prob-
lems in distinguishing the entrained from reentrained particulates, it may
be assumed that most particulates on the surfaces of the city were once
airborne for a short time as in sanding operations.
214
-------
sources have generally been ignored both in emission inventories and regu-
latory programs; controls for such sources are harder to implement and
often outside the purview of the air pollution control agency. Because
the actual reentrainment of particles is difficult to control, one approach
to controlling the impact of reentrained particulate matter is to reduce
the particles available. Therefore, it is important to understand where
and how these particles accumulate on the surfaces in the city. The
sources of particulates are discussed briefly below prior to the analyses
of natural and man-induced reentrainment.
Sources of Particulates for Reentrainment
The most widespread source of particulates for reentrainment in an urban
area is the fallout and deposition of particles that occur continually in
both the natural and urban environment. While there is no exact relation-
ship between ambient TSP concentrations and dustfall rates, there is a
direct correlation between these two parameters such that higher dustfall
rates are observed in areas with higher TSP concentrations. In a study
of 77 midwestern U.S. cities, dustfall was measured over a 4-month period
in different types of neighborhoods. The results indicated increasing
dustfall rates as the level of activity in the area increased:
Neighborhood
Residential
Commercial
Industrial
Rate of
total dustfall
g/m2/month
3.16
5.45
7.07
As the majority of the cities in this survey were small, with no appre-
ciable industry, these dustfall rates may be considered low for larger,
industrial areas. For instance, in a report on reentrained dust in
2 2
Chicago, the dustfall rate was determined to be 7 to 14 g/m /month.
215
-------
Two additional comments need to be made regarding dustfall measurements.
First, dustfall levels are generally measured by capturing all particu-
lates which fall into a collection apparatus whether by natural sedimen-
tation or induced by rainout and washout. Since rainfall in any large
amounts tends to help wash particulates off of surfaces and, in an urban
setting, down sewers, the actual accumulation of particulates on a sur-
face will not be continuously rising. Rather, given a constant sedimen-
tation rate, the loading intensity as a function of time would be similar
to that in Figure E-l.
ot
z
QZ
-------
Emissions from vehicles Emissions of particulates from
motor vehicles are usually calculated using AP-42^ emis-
sion factors. Depending upon the type of vehicle, the
emission rate varies from 0.33 g/km (0.21 from exhaust,
0.12 from tire wear) for light-duty, gasoline-powered
vehicles to 0.75 g/km for heavy-duty, diesel-powered ve-
hicles. As discussed later under Specific Urban Sources,
about three-fourths of the tailpipe emissions and less
than 10 percent of tire wear stay suspended; the remainder
presumably settles on the street or bordering.
Dirt and mud carryout from unpaved parking lots and roads -
This is extremely dependent upon the number and use of such
areas in a city, precipitation, and other factors. While
it is not possible to provide average citywide factors for
deposition from carryout, an understanding of the possible
magnitude in an area with numerous unpaved areas may be
obtained from a study done in the industrial Duwamish Valley
of Seattle.** This study stated that a car driven 7.5 miles
at 10 mph on a wet gravel road would gain approximately 80
pounds of mud and that an average of 0.74 pound per vehicle
was carried out from dirt parking lots, based upon brushings
from the tires.
Spillage from trucks - This is similarly difficult to esti-
mate on an average citywide basis. Spillage will be most
prominent in industrial areas and around construction sites
where materials are likely to be transported. Spillage was
shown to be the predominant reason for nonattainment of the
primary annual standard in Miami at one site.
Sand and salt for snow control Levels of sand and salt
added to the road will depend upon meteorological condi-
tions and approach to snow control (plowing versus salting)
among cities and, within each city, the importance of keep-
ing certain roads open. In Denver a little less than 16,000
tons of sand were added to the roads in the county (area =
77 mi2) in 1974.
The resulting accumulation of particles on streets will be affected by
removal mechanisms such as rainfall and street sweeping. A long-range
loading intensity graph could be represented by Figure E-2. It is
interesting ,to note the similarity between the buildup of particulates
on roads and the increase in TSP levels alter rainfall (Figures F-5,
F-7, and F-10); however, insufficient data was available to indicate
any cause-effect relationship.
217
-------
TIME
Figure E-2.
Accumulation of contaminants -
typical case (natural buildup
with periodic sweeping and
intermittent rainfall)
As may be expected, the particulate loadings on the streets due to these
mechanisms vary widely with the type of neighborhood. The loadings should
be much lower in a residential neighborhood than in an industrial area
where materials handling is prominent, fallout from the ambient air is
heavy, and unpaved parking lots and roads may be prevalent. In a light
commercial area, especially in the center city where street sweeping is
frequent, the total loading may be small. Figure E-3 presents these find-
ings from a water pollution study concerning urban runoff.
Natural Reentrainment
Erosion and entrainment of particulates from the soil have been exten-
sively investigated by agricultural scientists because wind erosion
has an obvious major impact on crop fields. One of the best known ex-
amples of this impact is the "dust bowl" situation in the late 1930's
when the Great Plains were severely affected by a combination of drought,
high temperatures, and periodic high wind velocities. According to a
study of Great Plains' duststorms in the 1950's, the average duststorm,
3 10
lasting 6.6 hours, had a median TSP concentration of 4.85 mg/m .
218
-------
4000
£ 3000 h
JQ
Z
UJ
I-
2 2000 h
o
z
§
o
CO
O
O
CO
1000 -
RESIDENTIAL INDUSTRIAL
-
CT
'1 1'
> X
/ ' x
/ __ x /
' / ' /
/ / X /
/ / / X
J ' s '
' / s /
7
\
/
/
/
/
/.
y
/
s
/
/
/t
/
/>
/
U UJ _ _ |_
o o t] j o
>v ^ \ ^
0*00
J w _J _l
O Z o O
X. V. >^ V»
* Q * °
O UJ o w
1 ^ 1 «
7
s
s
f
f
/
f
/
A
2
3
^
^
^
/
^
y
f
s
s
4
y
^
^
COMMERCIAL
-
-
__
KJ M^
I/J I/I
1^^ i yi
ryl 1^1
>. K u
> H «
i 3 \\
(9 5
Z (0
s -
I ^
,n Z
LAND-USE CATEGORIES
E-3. Total solids loading on street surfaces
variation with land
219
-------
While this discussion is primarily limited to the reentrainment of par-
ticulates in an urban environment, the mechanics and physics of soil
erosion are similar to those of reentrainment from a smooth surface.
Basically, wind forces can initiate three types of soil movement - surface
creep, saltation, and suspension. Surface creep involves particles of the
size 500 to 1000 jam that are rolled along the ground by strong winds and
by the exchange of momentum after impact with smaller particles in salta-
tion. Saltation consists of individual particles jumping and bouncing
usually within a few centimeters of the surface; particles that saltate
vary in size from 100 to 500 (am depending on shape and density, and are
therefore quickly brought back to the earth by the pull of gravity. The
third type of soil movement, suspension, consists of particles below 100 jam
being lifted off the ground and completely borne up by the wind. The dif-
ferences in these mechanisms and the particle sizes involved are especially
important in an air pollution context. The wind will support particles
as long as their falling (terminal) velocities are overcome by the upward
vertical velocities of the wind. Particles under 10 ^im can be suspended
almost indefinitely; particles with a mass median diameter of less than
5 (jm are generally assumed to be deposited in the lungs.
Fugitive dust emissions from wind erosion may be estimated from the wind
9
erosion equation developed by Woodruff and Siddoway. This equation ex-
presses a functional relationship between soil "loss" (tons/acre/year)
and the following variables:
Soil erodibility index This index is a measure of the
potential soil loss from a bare, smooth, noncrusted surface
and is related to the erodible fraction of the particular
soil (fraction of the soil below 0.8 mm particle size).
Soil ridge roughness This parameter is an estimate of
the amplitude of soil ridges or undulations.
« Local climatic factor This factor is a function of av-
erage wind speed and soil moisture conditions. The emis-
sion rate's dependence on the cube of the wind speed is an
important consideration in the analysis of worst day par-
ticulate loading.
220
-------
Length of the area subject to erosion This length, mea-
sured along the wind direction, is a necessary parameter
since some distance is required before the soil flow can
reach its theoretical maximum value.
Equivalent quantity of vegetative cover This parameter
is really a function of three other indices which measure
the quantity, kind, and orientation of the vegetative cover,
Although the soil erosion equation was developed basically for agricultural
areas, reasonable estimates of dust emissions from any exposed ground sur-
face (dirt roads, desert areas) could be made with an appropriate choice
of source parameters. One approximation which can be made is that the
relative mass of soil loss through saltation and suspension processes is
equal to the ratio by weight of the suspendible size fraction of particles
to the fraction greater than 10 \jm. A calculation of the wind necessary
for suspension of particulates in an urban setting (Chicago), taking into
account the surface roughness and the log-wind profile near the ground,
indicated that a velocity of about 16 mph, measured at a height of 22.5
2
feet, would be the critical point. Above this speed, suspension is very
likely if the dust is sufficiently dry and includes particles both in the
Q
size range 100 to 500 |im and below 100 |im. A recent study has revealed
that aerosol derived from the soil has a size distribution which follows
l
the size distribution of the soil itself for a range of particle radii
between 1 and 10 urn, so that ample particles would be available for
resuspension.
Similar analyses were conducted during this study and are presented in
Appendix F of this report. The results indicate that winds above 10 to
12 miles per hour are likely to be contributing to the TSP levels. Above
this speed, the reentrained dust maintains the TSP level above what
would be projected based upon the dilution effect of the ventilation
accompanying the wind. Therefore, the overall impact of wind neither in-
creases nor decreases the daily TSP levels. However, cleaner surfaces
(determined by comparisons after rainfall) did allow the dilution effect
221
-------
to reduce levels further than when the wind was blowing over dirtier
surfaces (measured on days before which there was no rainfall).
In an urban area, highest winds and turbulence are often experienced in
street canyons where the greatest among of particulates and larger par-
ticles are also found. Therefore, the problem of natural reentrainment
must consider the vortex circulation developed in these areas. At the
same time, however, the surface roughness of the urban area itself can
greatly reduce the wind speeds from those measured at an outlying station.
A discussion of the various conditions of wind flow perpendicular or par-
allel to the street canyons, under light or strong wind regimes, and their
impacts, are beyond the scope of this study. Further analysis and model
development are expected in a major study to be performed for EPA in
Philadelphia.11
Vehicle-induced. Reentrainment
As a measure of man's general activity in an urban area, probably no ac-
tivity is so important to reentrainment as that of motor vehicles. Not
only does the level of traffic reflect the total activity in an area, but
it also represents the greatest general disruption of the surface dust.
Much as the natural saltation process of particle movement may induce
smaller dusts to become entrained in the passing air, man's activities
on settled dust or dirt and sand deposited on roadways can disperse par-
ticulates into the ambient air. The wheels of the vehicles grind up the
larger, nonsuspendible particles into smaller ones, break up the cohesive
bonds of the dust, and impart kinetic energy to particles.
Vehicular Traffic on Paved Roads In several studies which have used
microscopy to identify the particles collected at monitors near street
levels, the contribution of vehicular traffic to the TSP concentrations
has been estimated to be as high as 90 percent.12'13 However, the con-
clusions of these studies have several problems in their derivation.
First, the microscopic analysis indicated that many of the particles
222
-------
being analyzed were greater than 50 \w in diameter and as high as 160 |jm.
As discussed earlier, such particles are not really suspendible and their
measurement is most likely a result of the height of the monitor being
used (~ 10 feet). Second, the analyses did not attempt to compensate for
any incoming particulate from nonurban areas or any industrial contribu-
tions. Rather, all mineral products (quartz, limestone, olivine, feld-
spars) were directly attributed to the grinding and entrainment of the
asphalt pavement. In addition to these problems in the interpretation of
results, microscopic analysis is subject to the interpretation of the
microscopists and cannot be considered an exact science, as discussed in
Volume II.
The calculation of total emissions from paved roads, as was done in the
Colorado report for developing particulate attainment strategies,^ is
based primarily upon a study of emissions from dusty paved roads in the
4
Duwamish Valley of Seattle, Washington. The emission rate developed in
this study was determined by isokinetically sampling the air 8 feet be-
hind a moving vehicle at a height of approximately 4 feet. A total of
four tests were run, three on a dusty paved road with no curb and one on
a paved road that was cleaned fairly frequently (see Table E-l). The re-
sults of these few tests indicated that significant differences can exist
between dusty and clean streets. In both cases, the majority of the
emissions were particles greater than 10 |im in size.
4
Table E-l. EMISSION FACTORS FOR TRAFFIC DUST FROM PAVED ROADS
Speed
mph
20
20
Type of road and test site
Dusty paved roads -
no curbs
Paved roads with curbs -
flushed weekly - swept
biweekly
Particulate, lb/VME
Total
0.83
0.14
Below
10 |om
0.17
0.0055
Below
2 urn
0.022
Percent
below
10 pjn
20.3
3.82
Number
of
tests
3
1
VM = vehicle mile
223
-------
In the Colorado study, the emission factor of 0.17 Ib/VM was used to
calculate the total tonnage of particulates that may be expected to be
emitted due to traffic on paved roads that had been sanded. An analysis
of the results of that study are given in the individual report on Denver
(Volume III in this report series); the major conclusion is that the
total tonnage determined to be emissions was much greater than the data
would actually support if such levels were considered to be suspended
for any length of time. The major reason for this conclusion was that
over 35 percent of the sand placed on the roads would have had to be sus-
pended to have such a high level of particulate emissions (5,800 tons)
directly attributed to sanding operations.
An analysis of the impact of both natural and vehicle reentrainment that
attempted to integrate the physics of reentrainment with observed values
2
was conducted in Chicago by Abel. Using over 8 years of data, this study
concluded that the total amount of particulates which become reentrained
3
and suspended is approximately 18 to 20 |ig/m on an average annual basis
in Chicago. Vehicular traffic was responsible for three-fourths of this
3
level and winds reentrained the other 5 [ig/m . This value was determined
to be reasonably constant from year to year, not a given percentage of
the total TSP value, because the reentrainment is a function of the traf-
fic volume, precipitation, and frequency of high wind velocities. Anal-
yses similar to the one in Abel's thesis were conducted in the course of
this study and are discussed in Appendix F.
The actual transport of particles from an asphalt road caused by car and
truck traffic has been investigated by using zinc sulfide (ZnS) tracer
particles which were initially less than 25 [im i-n diameter. General
findings from that experiment indicated that the fraction of tracer reen-
-5 -2
trained per vehicle ranged from 10 at 5 mph to 10 at 50 mph, and 20
to 30 percent of the material reentrained was found to be redeposited
within 30 feet of the road. The results of an analysis of the exposure to
reentrained ZnS at different heights are of particular interest to this
study because of their implications for monitoring of ambient air quality.
224
-------
Exposure, in units of mass/area, is the integrated product of particle
concentration, air speed, and sampling time. Typical exposures are shown
in Figure E-4 as a function of height and distance from the road. Expo-
sure is seen to decrease with both height and distance, decreasing rapidly
with increasing height over 8 feet. The results tend to support the find-
ings of this study, which also found decreasing concentration with increas-
ing distance from the road.
£ 10-5
10
DISTANCE FROM ROAO EDGE ,fl
A '0
D - 20
_ X -.--30
6 6
HEIGHT, ft
10
Figure E-4.
Airborne exposure profiles (truck driven
through at 15 miles h"1, 4-day weathering,
6,9 miles h"1 average windspeed at 4.9 ft
height)15
225
-------
The data gathered in the course of the city case studies provided several
opportunities for estimating the impact of vehicular-induced reentrain-
ment by comparing comparable monitoring sites. The best data were the
result of special studies that had been or were being conducted by the
local agencies to determine for themselves the impact of traffic on the
measured TSP levels. Generally this was done by monitoring in one loca-
tion but at either different heights or different distances from the
road, or both. The analyses of these data are given in the individual
city reports; the findings are listed below.
In Seattle two monitors, both at a height of 15 feet, were
operated 25 feet and 300 feet from a heavily traveled road
(average daily traffic, ADT = 13,000) in an industrial/
commercial area. Forty sets of paired measurements were
made between October 1974 and June 1975 with geometric mean
concentrations of 89 and 55 |j.g/m , respectively.
In San Francisco two monitors, both about 40 feet back
from the road, were operated at heights of 15 feet and
93 feet in the central business district (CBD). One
hundred eight paired observations in the first 8 months
of 1971 indicated geometric means of 47 and 42 ug/m ,
respectively. (The lower monitor had poor ventilation.)
In Philadelphia two monitors, in the same quadrant of the
block, were operated at heights of 13 and 60 feet in a heavi-
ly traveled area in the CBD (areawide ADT = 105,000; however,
nearest roads = 2,900 and 7,700 ADT). Forty paired obser-
vations were made between November 1974 and June 1975 with
geometric means of 110 and 83 yg/m , respectively.
In Philadelphia two monitors, within two blocks on a
heavily traveled road (ADT = 30,000), were operated at
heights of 13 and 35 feet and distances back from the
road of 35 and 100 feet, respectively. Twelve pairs of
concurrent observations in September through December
1974 indicated geometric means of 115 and 65 p.g/m ,
respectively.
While these data are in agreement with the experimental results discussed
previously in terms of showing a decreasing impact from traffic with
distance, they are not sufficiently detailed in any instance to provide
more than a linear relationship between two points. Similarly, the
226
-------
differences between the various monitor settings, especially the degree
of impact of traditional sources, varied considerably so that compari-
sons between different sets of paired observations were difficult.
In an attempt to bypass some of these problems and develop a quantita-
tive relationship between the traffic activity and measured TSP levels,
available traffic and TSP data were analyzed for Miami and Providence,
two areas which were believed to be not influenced by industry. Where
possible in each city, average daily traffic (ADT) levels were obtained
for monitors that were potentially influenced by traffic. Initial
comparisons between ADT and TSP levels without considering the monitor-
ing site did not produce any good relationships. However, when the
distance of the monitor from the traffic was taken into account by cal-
culating the slant distance(sD = J(height)2 + (distance)2 ) , a compar-
ison of the ADT/SD to the TSP concentration implied that a linear rela-
tionship could be assumed to exist with good correlation. The result
of this analysis is plotted in Figure E-5 along with the relationship
derived and correlation coefficients. The constant value of 46 in the
Providence relationship and 44 in the one for Miami can be interpreted
as representing the nontraffic-related contribution to the citywide
average TSP; i.e., if ADT = 0 or SD -» ».
Because of the lack of a constant, citywide value for a nontraffic-
related contribution in cities with a larger industrial base, compar-
able analyses in such cities failed to provide similar results. In
other nonindustrial cities (Washington and Oklahoma City), construction
activity was prominent and hindered the isolation of the effect of
traffic activity on measured TSP levels.
The form of the relationships derived for Miami and Providence was con-
sidered to be in accord with the experimental results presented in the
literature; i.e., TSP concentration decreasing with slant distance.
Several attempts were made to find better correlation using the data
from these two studies. Various weighting factors for the height of
227
-------
ro
e
O.
90
80
NOT
- INCLUDED IN
REGRESSION-)
70
60
50
40
42
0 MIAMI
A PROVIDENCE
100
200
300
400
500
ADT Per SLANT DISTANCE
Figure E-5. Providence/Miami comparison of ADT per slant
distance versus TSP (nearest major street)
228
-------
the monitor and its distance back from the road failed to provide any
increase in the correlation; attempts to derive exponential relation-
ships were similarly unsuccessful. The form of the relationships pre-
sented was thus assumed to be the most appropriate one for the data
being used.
The difference in the two slopes is believed to reflect the difference
in monitor placement in the two cities. In Providence, where the slant
distances varied more because of changes in height, the relationship
indicates that the TSP level due to motor vehicles changes faster with
an equivalent change in ADT/SD than in Miami; Miami's monitors were
almost consistently placed at the same height but the distance back
from the road varied greatly. Such an effect is expected based on the
literature discussed previously.
Under the assumption that the variation in TSP with ADT/SD was a valid
relationship, similar relationships were investigated for the two-point
data sets found in the city studies. In Seattle, where the terrain was
fairly open and the only difference in the monitors was th'eir distances
back from the road, the slope of the line was found to be 0.07 (ATSP =
ADT
0.07 x A rr)- In Philadelphia, for the two sites on the heavily
OLJ
traveled road and where the difference in slant distance had a major
component from both height and distance back from the road variations,
the slope was approximately 0.10. These findings appear to confirm
that the change in height has a slightly greater effect than the change
in distance. A specific relationship that addresses these variations
is expected to result from a current study of reentrained dust in
Philadelphia.
The overall impact of vehicular activity on citywide TSP levels and
the attainment of the standards can only be estimated from these find-
ings. As suggested earlier, the constant values in the Miami and
Providence relationships can be considered representative of the city-
wide TSP values without the influence of traffic. Therefore, the
229
-------
difference between those constants and the actual citywide values would
be the citywide TSP contribution from reentraintnent dust. In Miami,
3
the citywide TSP level in 1974 was 62 (jtg/m so the reentrained contri-
3
bution would have been 18 ug/m ; in Providence, the citywide TSP level
3
of 61 indicates a 15 ng/m contribution from reentrainment on the city-
wide level.
In conclusion, reentrained particulates from traffic on paved roads can
have a major impact on the citywide TSP levels. At individual monitors
the actual impact varies with the horizontal and vertical distance from
the road as well as with the amount of traffic on the road. From the
data analyzed in two cities and two special cases, the vehicle-related
3
TSP may be as high as 50 ug/m on an average basis for monitors close
to heavily traveled roads. For shorter time periods with different
meteorological and traffic conditions, this value may vary widely. If
many of the monitors in a city are located close to streets, there
would be a significant impact on the measured citywide TSP levels.
The above analysis basically addresses the question of impact of traf-
fic on measured particulate levels with no distinction between reen-
trained particulates and emissions from tailpipes or tire wear. How-
ever, the later discussion of these sources indicates that the majority
of the vehicular contribution at monitors close to the road is from re-
entrained particulates. For instance, even if the emission factor of
4
particulates under 10 p.m from a clean road (as given by Roberts ) is
used, each car would be reentraining 2.5 grams/mile as opposed to
emitting 0.31 grams/mile from tailpipes. Therefore, at least 85 percent
of the particulate matter originally suspended due to vehicles results
from reentrained particles.
230
-------
Vehicular Activity on Unpaved Areas - In many urban areas dirt or gravel
roads and parking lots are used by individual establishments or in
industrial areas because of the expense of adequate paving. As dis-
cussed previously, these areas can be sources of dirt for carryout to
paved areas and may also serve as areas for naturally reentrained dust.
In addition, vehicular activity on these areas can cause particle re-
entrainment.
4
One of the major sources of concern in Roberts' thesis was gravel
roads. Isokinetic sampling with the same equipment as used on paved
roads provided estimates of the emissions from dry gravel roads at
different speeds. These tests signified that emissions increased ex-
ponentially with speed and the percentage of particulates below 10 urn
also increased with speed so that almost 16 times as much particulate
below 10 um was determined to be emitted at 30 miles per hour as at
10 miles per hour.
The emission factors developed from Robert's work are given in Table
E-2. Almost 10 times as much particulate was found to be emitted from
gravel roads than from a dusty paved road on a per vehicle mile basis
at the same speed. This would imply a greater loading from gravel
roads than from paved roads; however, the ADT on gravel roads would
probably be more than on order of magnitude less than on paved roads.
Therefore, the contributions to local TSP levels from gravel roads may
be expected to be comparable to those of paved roads. Since gravel
roads are much less widespread in an urban setting, the impact on city-
wide TSP concentrations may be considered minimal.
231
-------
Table E-2. EMISSION FACTORS FOR TRAFFIC DUST FROM GRAVEL ROADS
Speed,
mph Test site
10 Gravel road
No. 1
20 Gravel road
No. 1
30 Gravel road
No. 1
20 Gravel road
No. 2
Ib/VM
total
particulate
3.5
7.0
22.2
7.3
Ib/VM
below 10
microns
0.58
1.9
9.0
2.0
Ib/VM
below 2
microns
0.10
0.24
0.77
Percent
below
10 microns
16.7
27.4
40.4
27.1
Number
of tests
1
17
1
1
Other estimates of emissions from unpaved roads have been made by mea-
suring ambient and deposited particulates. In one study, the emission
factor was determined to be 1.94 Ib/vehicle-mile at 30 mph for dirt
roads. Taking into account the silt content for roads (gravel road silt
content ~ 12 percent), another sampling study developed the following
equation to express the emission factor for particles smaller than 100 fim:
e = 0.81(s)(S/30)
where e = emission factor (pounds per vehicle-mile)
s = silt content of road surface (percent)
S = average vehicle speed (miles per hour)
18
Taking into account the number of dry days per year, another study de-
termined the following emission factor for particles smaller than 30 (am
(the assumed cutoff point for a hi-vol sampler) :
where d = the number of dry days per year
232
-------
This latter factor was used in developing a national emission inventory
for fugitive dust which included unpaved roads, unpaved airstrips, con-
struction and agricultural tilling. (Due to the widespread nature of
the study, many assumptions were made to allocate these emissions in
each county.) A comparison of emissions from unpaved roads with those
from traditional sources is given in Table E-3 for those central
counties analyzed in the course of this study. The data in this table
imply that the unpaved road emissions in counties that are not totally
urbanized can be 10 to over 30 times the emissions from traditional
sources. Even in urbanized areas, where unpaved roads are not common,
the fugitive dust from unpaved roads may be over 10 percent of the
traditional emissions in the county.
Based on the analyses conducted above, which indicate that the impact
of fugitive dust from vehicular activity decreases quickly with dis-
tance, these numbers are believed to be inappropriate for direct use
in air quality planning. If these fugitive dust emissions were treated
the same as traditional emissions and used in rollback calculations,
excessive, undeserved emphasis would be placed on these sources at the
expense of control of traditional sources. While these amounts of par-
ticulate may be temporarily reentrained"due to vehicular activity,
they are not suspended for any length of time. If they were, the rural
areas of counties would be expected to have TSP levels as high as those
found in the cities; such is obviously not the case. Therefore, the use,
if any, of these numbers would have to be limited to inputs to models
that adequately reflect the deposition and other removal of the par-
ticulates.
Control Approaches
The above discussions illustrate that a significant contribution to the
measured TSP levels in a city can result from reentrained particulates,
but that the degree to which the impact is measured is highly dependent
on the siting of the monitors. The reentrained contribution is mostly
233
-------
NO
U)
Table E-3. COMPARISON OF FUGITIVE DUST EMISSIONS FROM UNPAVED ROADS
WITH TRADITIONAL SOURCE EMISSIONS IN MAJOR COUNTIES OF
14 AQCR'S
AQCR
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
San Francisco
Seattle
St. ' Louis
Washington, B.C.
Major county
Baltimore City3
Jefferson
Hamilton
Hamilton
o
Cuyahoga
o
Denver
Dade
Oklahoma
Philadelphia3
Providence
o
San Francisco
King
St. Louis City3
a
District of Columbia
Emissions from
unpaved roads ,
tons per year
510
193,000
N.A.
76,990
18,800
1,270
70,940
85,920
2,660
139,830
660
197,770
l,220b
100
Emissions from
traditional
sources ,
tons per year
7,000
110,000
10,300
56,100
210,000
9,700
8,000
2,600
31,600
7,800
4,800
7,300
15,600
5,600
Ratio
0.07
1.75
1.37
0.09
0.13
8.88
33.55
0.08
17.88
0.14
27.17
0.08
0.02
These counties are almost totally urbanized so the travel on unpaved roads is ex-
pected to be minimal.
This represents the emissions from St. Louis County, a much larger, less urbanized
area; actual reported emissions for St. Louis City were 159,900 tons/year suggesting
a coding error confusion with the county.
-------
man-induced, and the majority of this in an urban area is from ve-
hicular traffic on paved roads. It is, therefore, this latter area
that should receive greatest attention in any formulation of a con-
trol strategy.
Under current EPA regulations and guidance, two basic approaches are
possible. The great diversity of monitor siting allowed in various
cities has already resulted in either highlighting or concealing the
problem of reentrainment. In the 14 cities studied under this contract,
the mean monitoring height varied from 10 feet (3 meters) in Birmingham
to 56 feet (17 meters) in Providence; distances back from the road
19
varied even more. Since current EPA guidelines allow hi-vols to be
placed from 2 to 15 meters above ground and no guidelines are set for
3
the distance back from a road, reductions of 50 ug/m or more may be
reasonable in areas of high traffic flow simply by relocating samplers
as witnessed at the two sites on South Broad in Philadelphia. There-
fore, the problem of monitor siting guidance is the first one that
needs to be resolved; this issue is discussed further under Monitoring
Considerations in Section III.
Once the monitor location is set and the TSP concentration due to traf-
fic on paved roads still provides an important contribution to the level
of TSP measured, control methods applicable to vehicular activity must
be considered. Since reduction of ADT is an extremely difficult task
and has seen only small success as a control method for CO, hydrocarbons,
and other traditional automobile-related pollutants, such a measure is
not discussed below. Rather, those measures which act to prevent the
deposition of particles for reentrainment or serve to remove such par-
ticles are given primary consideration.
Deposition -Control A viable approach to controlling reentrained par-
ticulates is to reduce the amount of particulates available for reen-
trainment; such measures are also beneficial for water pollution con-
trol. The feasibility of thio type of control varies with the source of
the particulate:
235
-------
Fallout This contribution to the total particulate loading
on the streets is small and no direct control is appropriate.
The fallout level will decrease automatically with control of
traditional source emissions and the reduction of lead in
gasoline.
Carryout In areas where dirt and mud carryout from unpaved
roads and parking lots is found to contribute significantly
to the loading of dust on paved roads, requirements for
paving or stabilization of the unpaved areas would be warranted.
Spillage The additional loading on streets due to spillage
from trucks may be easily regulated against; however, enforce-
ment may be a problem. Further study on this matter is needed
but it is probably possible to require specific equipment on
trucks to prevent spillage.
Sanding Because the short-term hazard of slippery roads is
more prominent than the hazard due to reentrained dust, sand-
ing and salting operations will obviously continue. Analysis
of the efficiency of sanding may result in procedures that
apply less sand more effectively; however, on-the-road imple-
mentation of such procedures is difficult. (Systematic road
cleaning after sanding operations would be more appropriate.)
Street Cleaning Since preventing all deposition of particles on a
paved surface is not possible, an alternative or auxiliary approach to
control of particle reentrainment is to remove the particles from the
surface. As part of a water pollution study, street cleaning practices
throughout the nation were evaluated through a review of the literature,
a detailed survey of current practices in several sample cities, and
several controlled tests. This study indicated that present methods of
cleaning streets fall into two categories of machines street sweepers
and street flushers.
Machine sweeping accounts for the great majority of street cleaning per-
formed in most communities. Motorized street sweepers are designed to
loosen dirt and debris from the street surface (this debris is normally
most concentrated in the gutter area), transport it onto a moving con-
veyor and deposit it temporarily in a storage hopper; the sweeper also
typically contains a dust control system. Three basic types of sweepers
are in use. The most common is a rotating gutter broom that moves
236
-------
materials from the gutter area into the main pickup brook, which ro-
tates to carry the material onto a belt and into the hopper. A second
class of sweepers uses a regenerative air system. These sweepers are
designed to "blast" the dirt and debris from the road surface into the
hopper; a portion of the air is recycled and the rest of the air is
vented through the dust separation system. A third type, vacuum
sweepers, has been in use in Europe for many years and in limited use
in this country. These vacuum sweepers use both a broom for loosen-
ing and moving street dirt debris and a vacuum system to pick up the
debris. All material picked up by the vacuum nozzle is saturated with
water on entry and passes into a vacuum chamber where the water-laden
dust and dirt drop out of the air stream.
In the controlled tests to determine street sweeping efficiency (using
only the first type of sweeper above), an overall effectiveness of 50
percent was found. However, the experiments indicated that this re-
moval was almost entirely in the larger size particle range (greater
than 100 um) and that, below 100 urn, more particles were found after
sweeping than before. In addition, the street sweeping operation,
which is principally directed at removing particles along the curb,
tended to redistribute the loading on the street so that more particles
were located on previously cleaner areas; i.e., toward the middle of
the street. These findings raise serious doubts about the use of
sweepers with only brooms for air pollution purposes. Those sweepers
that use air to blast the dust off the street or use a vacuum system
are expected to show better efficiency at lower particle sizes, but data
for a careful evaluation were not available.
Instead of, or in addition to, street sweeping, many cities will flush
the street with water. A street flusher consists of a water supply tank
mounted on a truck or trailer, a gasoline engine-driven pump for supply-
ing pressure, and three or more nozzles for spreading the water as
directional sprays. The large nozzles on the flusher are individually
controlled and are usually placed so that one is directed across the
237
-------
path, of the flusher and one on each side is pointed out toward the
gutters. Street flushing, as presently conducted, serves only to dis-
place dirt and debris from the street surface to the gutter. The volume
of water used is insufficient to transport the accumulated litter to
the nearest drain, and therefore the dust is not actually removed from
the streets. Should flushing result in transport into the drain (past
the catch basins), additional loading is put on the sewer system.
SPECIFIC UR5AN SOURCES
In addition to reentrained dust, other sources in urban areas have not
been traditionally considered major contributors to the TSP levels and
therefore have received little attention in the formulation of control
strategies for particulates. Yet these sources may be considered true
emission sources as the particulates arise directly as a result of the
individual activity rather than as a by-product. As with reentrained
dust, several recent studies have suggested that these sources may be
having more of an impact than previously thought. Of particular interest
in a crowded urban area are the tailpipe emissions from automobiles
and the emission of rubber due to tire wear. In addition, construction/
demolition activities are constantly occurring in cities and may add to
the total TSP levels measured. Each of these sources is discussed below.
Transportation Sources
Although particulates from the transportation sector have been inventoried,
they have seldom been regulated except through ordinances prohibiting
smoking vehicles and Federal restrictions on aircraft. Controls on motor
vehicles have centered around emissions of carbon monoxide, nitrogen
oxides, and hydrocarbons which are one to two orders of magnitude greater
than emissions of particulates. In addition, particulates from the
transportation sector have generally been assumed to be insignificant
when compared with emissions from traditional sources. However, as
238
-------
emissions from traditional sources have been reduced under implemen-
tation planning, particulates from the transportation sector have be-
come increasingly more important. To provide an idea of the relative
importance of the transportation sector, Table E-4 lists the emissions
from all sources categorized as transportation (including tailpipe
emissions, tire wear, aircraft, off-road vehicles, vessels) in inven-
tories for each of the major counties in the 14 AQCR's studied. When
these emissions are compared with the total emissions from traditional
sources, transportation emissions are shown to add a significant con-
tribution to the total inventoried emissions. This discussion focuses
on the contribution of tailpipe emissions and tire wear to the ambient
TSP levels.
Tailpipe Emissions - The emission of particulates from automobile and
truck tailpipes is not a source that has been ignored in the past in
terms of being included in emission inventories. Emission factors
published in AP-42 are most often used in determining emission levels;
these are listed in Table E-5.
Table E-5. PARTICULATE EMISSION FACTORS FOR
VEHICLES FROM AP-423
Vehicle type
Light-duty
Heavy-duty
Fuel
Gasoline
Diesel
Gasoline
Diesel
Exhaust emission
rate, g/mi
0.34
0.73
0.65
1.2
Attempts to separate out the contribution of motor vehicles to the total
TSP measured have centered around the use of lead as a tracer element.
On a national basis, automotive emissions of lead were calculated to be
20
75 percent of the total lead emissions in 1970. In most urban environ-
ments where lead, copper, and zinc smelters, grey iron foundries, or
239
-------
Table E-4. COMPARISON OF PARTICULATE EMISSIONS FROM TRANSPORTATION WITH TRADITIONAL
SOURCE EMISSIONS IN MAJOR COUNTIES OF 14 AQCR'S
AQCR
Major county
Emissions from
transportation,
tons per year
Emissions from
traditional
sources,
tons per year
Ratio
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
San Francisco
Seattle
St. Louis
Washington, D.C.
Baltimore City
Jefferson
Hamilton
Hamilton
Cuyahoga
Denver
Dade
Oklahoma
Philadelphia
Providence
San Francisco
King
St. Louis City
District of Columbia
2,960
3,050
1,420
3,490
6,410
3,060
4,870
6,650
2,180
1,095
5,190
1,360
1,500
7,000
110,000
10,300
56,100
210,000
9,700
8,000
2,600
31,600
7,800
4,800
7,300
15,600
5,600
0.42
0.03
0.14
0.06
0.03
0.32
0.61
0.43
0.21
0.28
0.23
0.71
0.10
0.27
a
Does not include 2,820 tons/year from aircraft.
Includes particulate emissions from tailpipes, tire wear, aircraft, off-road
vehicles, vessels.
-------
or other major point sources of lead are not prevalent, ambient lead
levels are assumed to be due almost entirely to vehicular activity.
Therefore, if the ratio of the total suspendible particulate emissions
from tailpipes to the suspended lead emissions is known, the ambient
levels of lead can be multiplied by this ratio to provide the ambient
TSP contribution due to total tailpipe emissions.
Such analyses of lead and total emissions have been conducted in several
studies with values for the ratio ranging from 3 to 7.5 depending upon
the model year of the car, gasoline used, and definition of suspendible
particulates. Basing their calculations on these previous studies and
taking into account the vehicle age distribution in the Los Angeles
21
area, another study developed the following estimates of weighted
average emission factors for light-duty motor vehicles:
Total tailpipe particulates = 0.43 g/mi
Suspended tailpipe particulates = 0.31 g/mi
Total lead emissions =0.19 g/mi
Suspended lead emissions =0.08 g/mi
Using these data, the ratio between suspended tailpipe particulates and
suspended lead would be 3.9. However, this ratio uses a vehicle age
distribution specific to Los Angeles in 1972, was calculated based upon
interpretation of a range of experimental data, and ignores differences
between light- and heavy-duty vehicles and between gasoline versus
diesel fuel which contains no lead. For different areas of the country,
this ratio may thus be assumed to range from 3 to 5.
Ambient lead levels are routinely measured in major cities through analy-
sis of NASN filters and many state and local agencies also perform their
own studies for lead. (California has an air quality standard for lead of
3 22
1.50 yg/m for a monthly average). One report reviewed ambient lead
levels contained in the National Aerometric Data Bank (NADB) and found
o
average annual concentrations ranging from 0.5 to 2 yg/m with a few
241
-------
cities measuring 3 to 4 yg/m . These lead data suggest that tailpipe
3
emissions are contributing from 1 to 20 yg/m to the total particulate
levels measured.
City study findings - Similar data from the NADB were provided for the
14 cities studied under this effort. To the extent that the data were
available, Table E-6 lists the average lead concentrations at the NASN
site in each city for the years 1972 to 1974. These values reflect the
3
middle range of those reported above; i.e., around 1 yg/m . Therefore,
it may be assumed that tailpipe emissions are generally contributing
3
3 to 5 yg/m to the ambient levels.
Table E-6. LEAD CONCENTRATIONS (yg/m ) AT
NASN SITES, 1972 TO 1974
City
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
San Francisco
Seattle
St. Louis
Washington, D. C.
1972
1.6432
1.0599
1.2176
1.5155
1.29763
1.9525a
1.3148
1.1518
2.0144
1.2391
1.1787
1.1574
0.9407
1.3292a
1973
0.9244
1.1581
(0.6475)b
NA
NA
(3.0351)C
1.5827a
(0.6343)b
(1.0772)c
0.8929a
1.1607
1.5176
(0.6079)°
NA.
1974
1.1533
1.1346
(0.6960)b
(0.8341)b
NA.
NA
1.0840
1.0255a
1.6071
1 . Ol64a
0.9340
(1.0l62)b
1.1845
(0.4889)C
a
Average of 3 quarters' data
Average of 2 quarters' data
Q
1 quarter's data
242
-------
In most of the cities where data are complete enough for comparison, the
lead levels have apparently decreased 10 to 20 percent between 1972 and
1974. This decrease may be a result of increased use of lead-free gas-
oline and also reduced vehicular activity in response to the energy
crisis. The percent decrease in TSP due to total tailpipe emissions
would not necessarily be equivalent to the percent reductions in lead
concentrations since the ratio of total to lead emissions will be in-
creasing over time as new cars replace older ones. In any event, the
decrease is negligible compared to the range of uncertainty in the ratio.
In addition to the NASN data, several local agencies provided to GCA
results of their metals analyses at particular monitoring sites. Selected
filters from these cities were analyzed for metals by EPA. The results
and additional supporting data are discussed in the appropriate city
reports and are summarized below; the analytical techniques used by the
various agencies are presented in Volume II.
Baltimore - The Maryland Bureau of Air Quality Control runs
chemical analyses on monthly composite samples to provide
monthly and yearly average levels for six stations. The an-
nual average concentration of lead ranged from 0.82 to 1.73
yg/nr* and monthly averages ranged from 0.36 to 3.08 yg/rn-^..
In all cases the lead contribution was less than 2 percent
of the total suspended particulate concentration.
Miami - Annual arithmetic averages of lead concentrations at
12 Miami monitors in 1974 were provided by the Metropolitan
Bade County Pollution Control Department. These values ranged
from 0.6 to 2.3 yg/m At the seven sites for which average
daily traffic (ADT) and slant distance (SD) were known, a re-
gression analysis indicated a positive correlation (r^ = 0.664)
between lead concentration and ADT/SD; however, the regression
equation implied a very small change in lead concentration with
large changes in ADT/SD, The lead contribution was generally
less than 2 percent of TSP.
In a special study conducted by EPA for this effort, 2-hour
elemental concentrations were compared with the hourly traffic
counts for 1 week. These data illustrated the expected rela-
tionship of increasing lead concentration (as high as 4.6 yg/m^
for a 2-hour average) with increasing traffic, especially in the
early morning when rush-hour traffic started before mixing
height and wind speed increases caused a drop in lead concen-
tration. The average lead concentration for the week was about
0.4 yg/nH.
243
-------
Oklahoma City - Analyses of hi-vol filters by the city-county
agency indicated annual average lead concentrations ranging
from 0.15 yg/m3 in a remote, rural area to 1.54 yg/m3 in the
center city-commercial area. At one site, in a location con-
sidered fairly rural, a nearby highway elevated levels to
0.66 yg/m3. The highest monthly average reported in 1974
2.38 yg/m3 was measured at a residential site. EPA analysis
of individual filters selected by GCA provided similar results
but with a wider range - 0.11 to 3.60 yg/m3.
Philadelphia - The Philadelphia Air Management Services labora-
tory provided the results of metals analyses done to date on
their filters. Two sites had a year of lead data. One in a
dense residential area, had an annual mean of 1.19 yg/m3 while
the other site, located in an outlying, light-industrial set-
ting, had an annual mean of 0.85 yg/nP. The highest daily lead
concentration measured at each site was 6.19 and 3.18 yg/m3,
respectively. A special 6-month study^-3 conducted at four
sites for the Air Management Services found average lead
values of 1.30 to 2.08 yg/m3. The lead data at the residential
site in Philadelphia were reported on a daily basis with 95
percent of the days in 1974 accounted for. An analysis of the
day of the week fluctuations in lead indicated that levels of
lead were 40 percent lower on Sunday and 25 percent lower on
Saturday than during the rest of the week. This weekly trend
was consistent with the estimated changes in daily traffic.
San Francisco - Four years of complete lead data were presented
in a report by the Bay Area Air Pollution Control District.
This report illustrated a continuing decline in the District-
wide average from 1.30 yg/m3 in 1970 to 0.78 yg/m3 in 1973.
The highest annual level in 1973 of 1.67 yg/m3 was recorded at
a site downwind from a major freeway; the lowest value of 0.35
yg/m3 was monitored in an area subject to some industry but
little traffic.
Six sets of paired filters from two sites, one in San Francisco
County and the other at a site with the highest annual average
due to fugitive dust, were analyzed by EPA. The results in-
dicated that the monitor exposed to traffic was consistently
measuring levels of lead more than twice those of the more
rural site. In addition, while the San Francisco site averaged
2.5 percent lead, the site influenced by fugitive dust had less
than 0.7 percent lead on the hi-vol filters.
Seattle - An analysis of the constituents of total suspended
particulates in the industrial valley.of Seattle had been con-
ducted previously under EPA contract.2^ This study found av-
erage lead values in the valley of 1.50 to 3.94 yg/m3 over a 5-
month period. The highest level was at a site influenced not
244
-------
only by traffic but also by activity at a steel mill, a
junkyard, and lead battery reclaiming facility. The lowest
value was at the site discussed under reentrainment which
was subject primarily to traffic.
St. Louis - EPA also provided 2-hour elemental concentration
at two sites in St. Louis for about 1 week. Unfortunately,
simultaneous traffic counts were not available so a direct
correlation between traffic and lead could not be shown.
Although no diurnal pattern was obvious, the lead concentrations
showed the expected weekly pattern of lower values on Saturday
and even lower values on Sunday.
Washington, B.C. - Daily values of lead in 1974 were reported
for the NASN site by the District's Department of Environmental
Services. These values ranged from 0.37 to 2.9 yg/m^ with an
annual average of 1.03 yg/m .
The special sampling studies conducted by EPA in Miami and St. Louis
also provided data on particle sizes by elemental composition. Results
in both cities indicated that the lead particles being sampled are
extremely small. Only a small percentage are greater than 4 ym in
diameter, and the largest percentage was collected in the last impactor
stage, implying the particles had an effective aerodynamic diameter of
less than 0.25 ym. This small diameter means that the lead would be
dispersed and transported much as a gas with very little fallout with
distance. Therefore, it may be expected to be measured at rooftop le-
vels or even in more remote areas.
The particle size of lead can be contrasted with that of tracer elements
often associated with reentrained dust - aluminum, silicon, and calcium.
These elements were usually found in higher quantities at the larger
size ranges (greater than 2 ym). While this size of particles is also
suspendible, the transport can be expected to be lessened to some extent,
A compilation of all the sites for which some lead data were available
provided 49 monitors from six cities (Baltimore, Miami, Oklahoma City,
Philadelphia, San Francisco, Washington) with annual average lead con-
centrations as well as individual filter analyses from several sites in
245
-------
the other cities discussed above. Those monitors with annual data were
grouped according to their site classifications and then averaged to
provide a mean concentration. Because the monitoring sites had a wide
range of local influences affecting the measured lead levels, Figure
E-6 presents not only the mean values for each of the site classifications
but also the range of values found. Since there were only four monitor-
ing sites under the classification of rural and industrial, these av-
erages and ranges may not be as representative of situations found in
other cities.
By using the ratios developed earlier, these lead data can be extra-
polated to a determination of the total concentration contributed by
tailpipe emissions. While the actual impact will depend upon variables
such as monitor siting and local traffic, the contribution from tail-
pipes may be assumed to be in the range of 2 to 10 yg/m3 in industrial
3
and commercial areas, 5 yg/m or less at residential sites, and less
than 2 yg/m3 at monitors remote from any heavy traffic activity.
Control approach - In urban areas where tailpipe emissions are found to
be contributing 5 to 10 yg/m3 to the TSP loading, control of tailpipe
emissions may be worthwhile. The most common suggestion for lower tail-
pipe emissions is to reduce or remove the lead in gasoline. Although
the lead particulates were estimated above to be 26 percent of the tail-
pipe emissions, the removal of lead from gasoline has a synergistic
relationship as other metal constituents are also reduced. It has been
estimated that the combustion of lead-free gasoline produces 85 per-
25
cent less particulate. Since diesel fuel, accounting for 10 percent
of the tailpipe emissions on a national average, has no lead in it,
the requirement for all cars to burn lead-free gasoline may reduce
O
center-city particulate concentrations around 5 yg/m . The current
use of lead-free gas by cars with catalytic mufflers and other post-
1971 cars is already accomplishing much of this strategy and should be
considered in future air quality planning.
246
-------
10
E
ZA
2.2
2.0
1.8
? 1.6
2 1.4
S
a:
UJ
o
o
o
o
<
UJ
1.2
1.0
0.8
0.6
OA
0.2
0
INDUSTRIAL
(4 SITES)
- \
COMMERCIAL
(28 SITES)
RESIDENTIAL
( 13 SITES)
RURAL
(4 SITES)
Figure E-6.
The range and average of annual lead concentrations
found at monitoring sites
247
-------
Tire Wear - Aside from direct tailpipe emissions and reentrained dust,
automobiles are known to generate particulates simply from the deteriora-
tion of the body and parts. Rust, corrosion, and friction of one part
on another are all known sources. However, their magnitude is small
compared with the wear that is seen on tires. In the U.S., an esti-
26
mated 660,000 tons of tire-tread are worn away each year. Since over
half of all vehicle-miles traveled (VMT) is in urban areas, approximate-
ly 350,000 tons of rubber are added to the urban environment each year.
Considering the size and widespread nature of this source, its impact
on air quality warrants study.
For the calculation of TSP emissions due to tire wear, EPA has used a rate
of 0.20 g/mi or about 60 percent of the tailpipe emission rate from light-
o
duty gasoline-powered vehicles. Tire wear's contribution to the total
TSP concentration would thus be expected to be about half that of tail-
pipe emissions. Concentrations would then be on the order of 1 to 5
3
yg/m in commercial areas and only a very minor contribution elsewhere.
An even lower estimate of the impact from tire wear was derived in a
fairly detailed analytical study comparing the elemental and chemical
26
composition of measured TSP. That study determined that 5 to 10 per-
cent of the tread material that disappears from tires in use becomes
airborne and the relative contribution to total TSP is on the order of
20 percent of the contribution from vehicle exhausts. This provided
an average concentration of airborne tire particulate matter in urban
3
areas of about 1 yg/m .
Two separate studies by the same group of analysts have provided con-
flicting data on the levels of rubber tire fragments in samples col-
12
lected in Philadelphia. In the earlier study, the analysis of hi-
vol filters implied an estimated weight percentage due to rubber tire
3
fragments of 10 to 40 percent, amounting to 10 to 50 yg/m . In a
1 o
follow-on study, rubber tire fragments were found to represent only
a minor (0.5 to 5 percent) or trace (<0.5 percent) amount of the
248
-------
material collected on a weight percentage basis. No explanation of
these differences was given.
City study findings - Although filters were selected from each city for
microscopic analysis, neither the selection of a few filters nor the
accuracy of the microscopy were believed to be sufficient to characterize
the cities. However, the numerous filters from among the cities were
considered to be adequate for averaging contributions according to the
various classifications for the monitoring sites.
By multiplying the average percentage contribution of rubber tire frag-
ments to total visible loading (diameter >1 urn) on the hi-vol filters
at each sampling site selected for analysis by the annual mean TSP for
that site for 1974, and under the assumption presented in Section II
that 85 percent of the loading was visible, average annual rubber load-
ings could be estimated for each site. These estimated loadings were
then grouped according to site type and averaged; these average values,
along with the range of values observed for each site type, are plotted
in Figure E-7. It must be cautioned that because of the nonrandom process
of filter selection, the discrepancies noted in replicate microscopic
analysis, and the use of only one to six filters to characterize the site
3
for the whole year, the values given could easily be off by several ug/m .
However, with these cautions in mind, it is interesting to note that the
results of the analysis show that commercial sites, generally most exposed
to traffic, have the highest contribution of rubber while underdeveloped
or rural sites barely measure any rubber.
In the course of the careful evaluation of the monitoring network in
each city, monitoring sites were also rated on the basis of local in-
fluences, including paved roads. Sites with an expected paved road
influence (10 in all) had rubber concentrations twice as high as sites
3
for which no such influence had been noted - 9.9 and 4.9
respectively.
249
-------
O\J
fO
E
^ 28
4.
to" 26
z
^ ZA
<
£ 22
Ul
2 20
Si 18
m
00
* 16
u.
o
14
z
o
H I 2
< ' c
cc
2 10
LU
O
Z
o 8
o
6
4
2
O
-
-
-
-
~
^
INDUSTRIAL
(15 SITES)
_
_
-
-
-
_
i^
,
COMMERCIAL
(35 SITES)
«
.
.
RESIDENTIAL
(15 SITES)
.
T RURAL
1_(5 SITES)
Figure E-7.
Average and range of TSP loadings due to tire wear
at different monitoring site classifications
250
-------
The values found for rubber in the course of this study are several
times what were expected based on the literature reviewed. The reason
for this discrepancy is not apparent. This experiment may have been
better formulated to give a good cross-section of values or the high
levels may be an artifact of the monitoring site and filter selection.
Currently, however, there is no reason to doubt the validity of these
data.
Of some interest in planning for control is the particle size distri-
bution of the rubber. The. average size range of the rubber tire frag-
ments (13 to 135 ym) is much larger than that of any other particulate
identified in the microscopic analysis (See Appendix B). Some par-
ticles were found to be 200 ym in length. Normally, such large particles
are not considered suspendible for any length of time and are too large
to be of concern for inhalation. Despite their size, however, no dif-
ference was discernible in average concentrations of rubber by monitor
height. Several monitors, 50 to 100 feet above ground level, measured
3
concentrations of rubber in the 5 to 15 yg/m range while other, lower
monitors recorded no rubber.
Construction/Demolition
The movement of materials associated with construction and demolition
activities usually results in the emission of particulates into the am-
bient air. Major demolition programs, whether using a ball and crane
or blasting (low-yield), will emit particulates up to a height equal to
that of the building being removed. On the other hand, construction
involves much more movement of materials continuously for periods of
several months to over a year. Emissions are generated by a wide
variety of operations over the duration of the construction, including
land clearing, blasting, ground excavation, and on-site traffic as well
as the construction of the facility itself.
251
-------
Two studies have attempted to estimate the total emissions from construc-
tion activity based on the size of projects underway. This is a very
difficult task as most construction jobs, except for road construction,
are different. In addition, particulate emissions vary substantially
from day to day depending on the level of activity, the specific operations,
and the prevailing weather. Original estimates for an emission factor
for particulates from construction were between 1.0 and 1.4 tons/acre/
month. * A later revision to develop state-specific emission factors
considered the average duration of construction projects in each
18
state. The resulting emission factors and total emissions due to
construction in major counties for each of the AQCR's reviewed are
given in Table E-7 along with the traditional source emissions.
The data in this table indicate that emissions from construction activity
in most of the major counties reviewed in this study would be over-
whelming in comparison to those sources traditionally considered. In all
but one case the emissions from construction are comparable to or
greater than traditional source emissions. When combined with the emis-
sions from unpaved roads, the fugitive dust emissions in King County and
Oklahoma County amount to over 50 times the traditional emissions.
To develop the emission factors used for deriving the emission estimates
presented above, several ambient monitoring experiments were conducted.
These sampling studies were not conducted in urban settings but rather
in more open terrain near construction sites where more surface preparation
and travel on the construction site itself would occur. This may explain
why the emission factors are so large.
Of greater importance to the use of these numbers in air pollution control
planning are the findings of the above sampling studies with respect to the
extent of the impact of the construction activity on ambient TSP levels.
These studies showed that any effect that construction activity had on
252
-------
Table E-7. COMPARISON OF FUGITIVE DUST EMISSIONS FROM CONSTRUCTION WITH TRADITIONAL
SOURCE EMISSIONS IN MAJOR COUNTIES OF 14 AQCR'S
Ln
AQCR
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
San Francisco
Seattle
St . Louis
Washington, D.C.
Major county
Baltimore City
Jefferson
Hamilton
Hamilton
Cuyahoga
Denver
Dade
Oklahoma
Philadelphia
Providence
San Francisco
King
St. Louis City
District of Columbia
Emission
factors,
tons
per acre
13.9
16.5
16.2
17.0
17.0
15.7
15.7
17.2
17.4
11.2
17.5
17.2
16.5
13.4
Emissions from
construction,
tons per year
19,320
23,000
N.A.
109,650
202,300
46,000
39,410
85,140
292,230
7,200
147,180
204,680
68,800
12,060
Emissions from
traditional
sources ,
tons per year
7,000
110,000
10,300
56,100
210,000
9,700
8,000
2,600
31,600
7,800
4,800
7,300
15,600
5,600
Ratio
2.76
0.21
1.95
0.96
4.75
4.93
33.26
9.24
0.92
30.66
28.12
4.40
2.15
-------
TSP concentrations was limited to within less than a mile from the con-
struction site. In one experiment, the contribution, if any, of the con-
struction activity to levels measured at monitors more than a mile from
the construction site could not be distinguished from other, more local
influences. In another experiment, a monitor approximately 4000 feet from
a construction site did not reflect any measurable impact from the con-
struction. On days of heavy construction activity, other sites closer to
3
the activity recorded concentrations 100 yg/m or more above levels
measured when there was no activity.
City Study Findings - To help provide an understanding of the impact of
construction activity on local and city-wide TSP levels, data on
current and recent construction were gathered in several of the cities
studied under this effort. Initial efforts to try to relate particular
parameters of construction activity, such as building permits and
dollars of new construction, to city-wide trends in TSP concentration
failed in each of the cities for which it was tried (Providence, Phila-
delphia, Denver). As these data were collected on an annual basis, the
Impact of construction could not be isolated from that of meteorology
and traditional source emissions. In addition, the type of data avail-
able for construction activity was not applicable to the size or the
type of activity underway. Building permits are required regardless
of the size of the project, and cost may refer more to the types of
materials being used in construction than to size.
Therefore, the remaining effort was primarily devoted to estimating the
impact of known major construction on nearby monitors. The findings
are discussed at length in the respective city reports and are sum-
marized below.
Baltimore - Baltimore City has been conducting a large urban
renewal program over the past 10 years, resulting in a con-
siderable amount of demolition, construction, and rehabili-
tation; highway construction has also been underway for the
254
-------
past 3 years. Four monitoring sites within 1 mile of construc-
tion showed an increase in TSP levels while the average values
at other sites were declining. If similar declines had
occurred at those sites near construction, their TSP levels
would have been an average of 11 yg/m lower in 1974. Road
construction progressed to within one block of one monitor in
1974; its TSP level was 19 yg/m3 higher in 1974 than in 1973.
Cincinnati - Urban renewal and highway construction have been
prominent over the past 15 years, predating much of the
monitoring network. No substantial impact on any one monitor
could be established and, on an average, those sites potentially
influenced by construction activity showed no difference in
trends when compared with other monitors.
Denver - A comparison of total construction activity, through
the use of permits and dollars of construction, with city-
wide TSP levels failed to produce any correlation. However,
fluctuations of 10 to 15 yg/m3 in annual geometric means at
several monitors over the past 5 years were interpreted by
the state agency to be caused by nearby construction and urban
renewal activity.
Miami - The 1974 construction activity associated with sewer
installation for a residential neighborhood showed no mea-
surable impact on the TSP levels at a monitor several blocks
away.
Oklahoma City - Monitors in the downtown area where urban
renewal has been underway for over 5 years have some of the
highest values reported in the city, averaging about 10 yg/m3
higher than in other parts of the city. However, a comparison
of the trends at these sites versus those at other sites showed
no apparent difference. The monitors started operation after
the urban renewal effort.
Philadelphia - Extensive analyses of construction activities
and TSP levels were conducted in Philadelphia on several
levels of data - city-wide, administrative area, and site-
specific. The city-wide and administrative areawide analyses
showed no noticeable relationship between construction activity
and measured TSP concentrations. A review of individual construc-
tion activity and nearby monitoring sites did not provide any
definitive correlation between TSP levels and distance from the
construction activity. The results indicated some probability
that construction activity would affect measured air quality
levels within 4000 feet of the activity, but wide-scale impacts
were precluded.
255
-------
Providence - Similar analyses were conducted in Providence;
however, no impact on city-wide TSP levels could be detected
over time. While a couple of sites were within 400 feet of
major construction activity, the influence of the construction
on measured TSP levels could not be separated from changes
observed at other sites due to meteorological conditions.
St. Louis - Urban renewal projects have been in operation in
St. Louis for the past 20 years, predating the monitoring net-
work. Particulate levels have been decreasing over the years
at all sites; however, at sites which are closest to the areas
of construction, levels have been decreasing at a slightly
slower rate than in other parts of the city. This suggests that
the continuing emissions due to construction may be helping to
maintain elevated levels at these sites; the projected average
difference is 3 ug/m3-
Washington, D.C. - A significant amount of construction has
taken place in the Washington area in the last 2 years; in
particular, the METRO transit system development is a focus
of construction activities of various types around the city.
According to the District agency's 1974 "Annual Report on
the Quality of the Air in Washington, D.C.," subway construc-
tion influenced measured particulate concentrations at three
sites, and a fourth site (CAMP) was felt to be subject to
other construction activity. Trends at these four stations
were compared to trends at other sites in the District of
1972-1974; the comparison suggested that the 1974 levels were
elevated 10 to 15 yg/m , on an average, at the sites near
construction. One monitor 600 yards from construction in 1974
showed levels 19 yg/m3 higher than in 1973. Those monitors
exposed to the cut-and-cover or surface type of construction
were the ones which showed the increases, while a monitor near
tunneling activity did not have higher levels in 1973.
Two other activities were available to indicate the impact
of construction. At the CAMP monitor, a vigorous enforcement
program was begun in May 1974 to eliminate fugitive dust from
the construction site closest to the monitor. This program
produced a substantial decrease in the number of violations
of the 24-hour secondary standard (150 yg/m3) from 16 during
the first 4 months to four for the rest of the year. At
another station, a month of construction for the foundation
of a greenhouse only 100 yards from the monitor provided a
monthly geometric mean 80 yg/m3 higher than would have normally
been expected.
256
-------
The above findings illustrate that construction activity does have an
impact on very local TSP levels but that the effect is not readily pre-
dictable. Construction will generally elevate concentrations downwind
from the site for distances up to a mile. The amount of increase is re-
lated to the level of activity, type of activity, distance from the
activity, and control measures employed. Monitors within half a mile
o
of construction may have annual geometric means 10 to 15 yg/m higher
than normal. Therefore, if 10 percent of the monitors in an urban area
are near construction activity, the calculated city-wide average TSP
level would be 1 to 2 yg/m^ higher than otherwise expected.
These measured impacts are much less than would be expected based on a
simple interpretation of the emission levels for each county given in
Table E-7. Obviously, the use of those emission levels must be
restricted to input into modeling programs which adequately account
for the fallout and deposition of particles. In addition, new emis-
sion factors to reflect the type and degree of activity and any con-
trol measures would be more appropriate. Future development should
emphasize that construction's impact is felt most strongly on the
short-term where conditions change from day to day.
Control Approaches - As with fugitive emissions (discussed in Appendix
D), the control of dust from construction and demolition activities is
most often approached through the use of nuisance or reasonable pre-
caution regulations. Under these regulations, the air pollution control
agency usually responds to complaints about fugitive dust and then re-
quires whatever controls are felt necessary to restrict the emission of
particulates. Such regulations are as stringent as the agency wishes
to enforce them. In some cases when the agency has reviewed authority
over construction permits, the construction firm is notified of the need
to control fugitive dust and specific control measures may be suggested.
257
-------
The property line regulation, which allows a certain increase in con-
centration as the air passes over the property and/or restricts the TSP
concentration or opacity of the air crossing the property line, is also
applicable to fugitive dust emissions from construction. This type of
regulation is enforceable to the degree that quantitative limits are
set; however, it requires a special monitoring network set up around
the construction activity to provide the necessary data. An appropriate
network may be expensive to set up and operate, and controls under this
type of system are unlikely to be pursued to any greater degree than
reasonable precautions.
A final type of regulation, used in only a few states, requires specific
controls on particular operations. For instance, in Colorado the
state agency has called out construction and demolition activities for
control. Permits are needed for these activities and the types of
control measures to be used must be described in the permit application.
Fugitive dust control measures include, but are not limited to:
wetting down, including prewatering;
landscaping and replanting with native vegetation;
covering, shielding or enclosing the area;
paving, temporary or permanent;
treating, the use of dust palliatives and chemical
stabilization;
detouring;
restriction of the speed of vehicles on sites;
prevention of the deposit of dirt and mud on improved
streets and roads;
e disturbing less topsoil and reclaiming as soon as
possible.
258
-------
Specific requirements imposed by Colorado include the following;
Sequential blasting shall be employed whenever or wherever
feasible to reduce the amounts of unconfined particulate
matter;
Such dust control strategies as revegetation, delay of sur-
face opening until demanded, or surface compacting and
sealing, shall be applied;
Haulage equipment shall be washed or wetted down, treated,
or covered when necessary to minimize the amount of dust
emitted in transit and in loading;
These measures shall also be enforced during periods when
actual construction work is not being conducted, such as
on weekends and holidays.
Control methods such as watering and chemical stabilization have been
evaluated and control efficiencies of 50 to 80 percent have been
suggested. Such control efficiencies are directly related to the
extent of the control measure applied; e.g., amount and frequency of
watering. Therefore, an agency should stipulate not only the type
of control measure but also the degree of application. The construc-
tion firm then knows exactly what is expected of it and enforcement
may be easier. The problem with this approach lies in the selection
of the appropriate control measure for each type of activity.
CONCLUSIONS
While the above discussions have not covered all topics possible under
the heading of nontraditional sources, they did center on those
sources that have been identified as probable major influences. From
these sources alone it is evident that there are contributions to the
total TSP levels simply from man's activity and, in an urban area
where man's activity is greatest, the contributions are the highest.
Similarly, the closer to the activity, the larger the impact. These
variations have been addressed above for each of the sources, but an
259
-------
overall combined assessment is needed to indicate the extent of the
total impact of nontraditional sources on TSP levels, and thereby on
the problem of attaining standards.
Because of the range of TSP levels that may be contributed by the
various nontraditional sources, it is not possible to identify at this
stage either the exact impact at any monitoring site or even the aver-
age impact in any one city. Such a determination would require exten-
sive data and modeling, most of which are not available. Rather, the
intent is to provide a measure of the range of impacts that may rea-
sonably be expected in most situations and an understanding of the re-
lative importance of nontraditional sources for standards attainment
on a national basis. Therefore, this conclusion should not be taken
as sufficient to preclude detailed analysis in each city but as guidance
to the development of national priorities for further planning measures.
The average and range of TSP levels attributed to tailpipe emissions and
tire wear were given in the above analyses by site type. Recognizing
that there were wide ranges of values found at sites in different cities,
average contributions can still be calculated. Tailpipe emissions pro-
vided an average level of TSP in industrial and commercial areas of 4 to
o 3
5 yg/m and approximately 3 yg/m in residential areas. Tire wear added
~ 3
rubber concentrations of 5 yg/nr at industrial sites, 8 yg/m at commer-
2
cial sites, and 3 yg/m at residential sites.
Construction activity is more difficult to present on an average basis
because of the wide range of possibilities that may occur. Some cities
have construction currently underway at individual, widely dispersed
locations which are not close to monitoring sites, while others may have
similar activity but within one or more monitoring sites. Levels of TSP
o
due to construction are expected to range between 0 and 15 yg/m with
the closer the monitor, the greater the impacts. If only one or two
monitors out of a network of 20 are near construction activity, a city-
3
wide average will only be affected by 1 to 2 yg/m annual geometric mean.
260
-------
However, in some cities major construction programs such as urban renewal
and subways are going on in concentrated areas of the city. These acti-
vities are apparently causing higher-than-normal values at a large number
of nearby monitors and will provide elevated average values in the com-
mercial section of the city.
While construction is obviously a localized source, the reentrainment
problem exists wherever there are roads and traffic. Since the level of
reentrained matter could not be exactly determined through microscopy or
elemental analyses, as with rubber and tailpipe emissions respectively,
other measures were necessary. Based upon the analyses of average daily
traffic and slant distances in Miami and Providence, the city-wide TSP
contribution from reentrained dust can be assumed to be in the range of
o
15 to 20 yg/m . Also, by calculating the excess levels at residential,
commercial, and industrial sites that could not be explained after
accounting for nonurban levels and traditional sources (see Section III),
the total nontraditional impact on the annual geometric mean TSP levels
3 3
averages around 20 to 25 ug/m at residential sites and 30 to 35 ug/m
at commercial and industrial sites. By subtracting the above levels
estimated to be due to tire wear, tailpipe emissions, and construction,
the average contribution due to reentrainment and any other nontraditional
3 3
sources would be approximately 20 ug/m at industrial monitors, 18 ug/m
3
at commercial monitors, and 14 ug/m at residential monitors. (The higher
levels at industrial monitors are likely the result of dirtier roads in
the area.) These neighborhood levels of TSP are comparable to those
calculated on a city-wide basis.
261
-------
REFERENCES
1. Hunt, W. F., C. Pinkerton, 0. McNulty, and J. Creason. A Study in
Trace Element Pollution of Air in 77 Midwestern Cities. Presented
at the 4th Annual Conference on Trace Substances in Environmental
Health, University of Missouri, Columbia, Missouri, June 23-24, 1970.
2. Abel, M. D. The Impact of Refloatation on Chicago's Total Suspended
Particulate Levels. A Thesis Submitted to the Faculty of Purdue
University in Partial Fulfillment of the Requirements for the Degree
of Master of Science, August 1974.
3. Compilation of Air Pollutant Emission Factors. U.S. Environmental
Protection Agency, Office of Air and Water Programs, Office of Air
Quality Planning and Standards. Publication No. AP-42, April 1973.
4. Roberts, John W. The Measurement, Cost and Control of Air Pollu-
tion From Unpaved Roads and Parking Lots in Seattle's Duwamish
Valley. A thesis submitted in partial fulfillment of the require-
ments for the degree of Master of Science in Engineering, University
of Washington, 1973.
5. Sartor, J. D. and G. B. Boyd. Water Pollution Aspects of Street
Surface Contaminants. Prepared for U.S. Environmental Protection
Agency. Washington, B.C. EPA-R2-72-081, November 1972.
6. Bagnold, R. A. The Physics of Blown Sand and Desert Dunes.
Methuen and Co., Ltd., London, 1959.
7. Chepil, W. S. and N. P. Woodruff. Estimation of Wind Erodibility of
Field Surfaces. J Soil Water Conser. 9:257-265, 1954.
8. Gillette, D. A., I. H. Blifford, Jr., and C. R. Fenster, Jr. Measure-
ments of Aerosol Size Distributions and Vertical Fluxes of Aerosols
on Land Subject to Wind Erosion. J Appl Meteorol. 11:977, 1972.
9. Woodruff, N. P. and F. H. Siddoway. A Wind Erosion Equation.
Soil Sci Soc Am Proc. 1965.
10. Hagen, L. J. and N. P. Woodruff. Particulate Loads Caused by Wind
Erosion in the Great Plains. Presentation at the 66th Annual
Meeting of Air Pollution Control Association, June 1973.
11. Philadelphia Particulate Study being conducted by GCA/Technology
Division for U.S. Environmental Protection Agency. Contract No.
68-02-2345.
262
-------
12. Scott Research Laboratories, Inc. Measurement of Street Level
Pollutants. Prepared for City of Philadelphia, Air Management
Services, May 1973.
13. Scott Environmental Technology, Inc. A Study of the Nature and
Origin of Airborne Particulate Matter in Philadelphia. Prepared
for City of Philadelphia, Department of Public Health, January
1975.
14. PEDCo-Environmental Specialists, Inc. - Investigation of Fugitive
Dust - Sources, Emissions, and Control - for Attainment of Sec-
ondary Ambient Air Quality Standards, Colorado. Prepared for
the U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards.
15. Sehmel, G. A. Particle Resuspension From an Asphalt Road Caused
by Car and Truck Traffic. Atmos Environ. 7:291-309, July 1973.
16. PEDCo-Environmental Specialists, Inc., Investigation of Fugitive
Dust-Sources, Emissions and Control. Environmental Protection
Agency. Contract No. 68-02-0044, Task Order No. 9, May 1973.
17. Cowherd, C. Jr., K. Axetell, Jr., C. M. Guenther, and G. A. Jutze.
Development of Emission Factors for Fugitive Dust Sources. Pre-
pared for the U.S. Environmental Protection Agency, Office of Air
and Waste Management, Office of Air Quality Planning and Stand-
ards. Publication No. EPA-450/3-74-037, June 1974.
18. Cowherd, C. C., et al. Emissions Inventory of Agricultural Tilling,
Unpaved Roads and Airstrips and Construction Sites. Prepared for
the U.S. Environmental Protection Agency, Office of Air and Waste
Management, Office of Air Quality Planning and Standards. EPA-450/
3-74-085, November 1974.
19. Guidance for Air Quality Monitoring Network Design and Instrument
Siting (Revised). Environmental Protection Agency, Research
Triangle Park, North Carolina. OAQPS No. 1.2-012 (Revised),
July 1975.
20. Hamilton, W. F. and P. A. Boys. Control of Airborne Lead Emissions
What's the Cost? Presentation at the 67th Annual Meeting of
the Air Pollution Control Association. J Air Pollut Contr Assoc.
No. 74-97, June 1974.
21. An Implementation Plan for Suspended Particulate Matter in the Los
Angeles Region. Technical Support Documents 1, 2, 3, 4. Prepared
for U.S. Environmental Operations. EPA Contract No. 68-02-1384.
263
-------
22. Lillis, E. J. and D. R. Dunbar. Impact of Automotive Particle
Exhaust Emissions on Air Quality. U.S. Environmental Protection
Agency, November 13, 1975. Unpublished.
23. Lead Concentrations in the Bay Area, 1970-1973. Bay Area Air
Pollution Control District. Information Bulletin, August 29, 1974.
24. Olsen, R. H., M.Y. Almassy, and A. LeWingert. A Study of the Sus-
pended Particulate Problem in the Duwamish Basin. Prepared for
the U.S. Environmental Protection Agency, Region X, Seattle.
EPA-68-02-1499, May 1975.
25. Supplement No. 5 for Compilation of Air Pollution Emission Factors,
Second Edition. EPA AP-42, April.
26. Pierson, W. R. and W. W. Brachaczck. Airborne Particulate Debris
From Rubber Tires. Ford Motor Company, Dearborn, Michigan,
January 13, 1975. Unpublished.
264
-------
APPENDIX F
INFLUENCE OF METEOROLOGY AND CLIMATOLOGY ON TSP LEVELS
GENERAL
The primary purpose of this appendix is to put into perspective the role
of meteorology and climatology in determining which urban areas or regions
of the country have met or are likely to meet the national particulate
standards by summarizing the impact of certain meteorological variables
and meteorological conditions on TSP levels in as quantitative a fashion
as appears reasonable at the present time. Meteorological factors affect
the rates at which locally generated particulates are introduced into the
atmosphere, and control local dispersion and transport from the urban
area. Additionally, the concentration of particulates in air as it enters
an urban area is determined in large part by the air's recent trajectory
and past meteorological conditions. And finally, depletion processes are
also influenced by meteorological events.
This appendix provides a brief general discussion of the ways individual
meteorological factors affect TSP concentrations, followed by a detailed
presentation of attempts to quantify the affects of precipitation, wind
speed, and degree days. The appendix ends with a summary of regional
meteorological and geographical influences.
Precipitation
The effect of precipitation is twofold: (1) it cleanses the atmosphere
by capturing particles within the cloud and by the washout of particles
below the clouds; and (2) it suppresses fugitive dust.
265
-------
Good summaries of the effectiveness of rainfall in removing particles have
1 2
been presented by Engelmann and Pasquill and will not be repeated here.
3
A basic reference is Chamberlain in which the washout coefficient is
shown as a function of the rate of rainfall and the terminal velocity of
the particle. From Chamberlain's curves it is evident that the fraction
of particles removed per unit time increases not only with increasing
rainfall rate but also with increasing particle size. Calculations made
from these curves show that sustained, light-to-moderate precipitation is
extremely effective in removing the larger particulates, a conclusion that
checks well with observational experience. For example, the use of
£n(X/X ) = -W t to express the fractional depletion of particle concentra-
tion X where W is the washout coefficient and t is time, as done by Stern,
4 P
et al., indicates that 50 percent of particles 8 ym in diameter and of
unit density would be removed in about 35 minutes in a rain of 0.10 inch
per hour. In addition to the inertia effects of the raindrops and particles
which were considered in developing these curves, electrical effects, which
are less well understood, are believed to play an important role in the
removal of very small particles. Such small particles, of the order of a
micron or less in diameter, have residence times within the atmosphere of
several days.
Snow is also an effective cleansing agent. Although the feathery nature
of the flakes and probable electrostatic charge suggest that snow is more
efficient in cleansing the atmosphere than an equal amount of precipita-
tion in the form of rain, this has not been firmly established.
The effectiveness of rain in suppressing fugitive dust emissions is a
result of the strong cohesive forces of absorbed water films surrounding
soil particles. As a result of the relationship between soil moisture
and fugitive dust emissions, soil must be very dry to be easily suspended
by the wind. Although rainfall tends to create a surface crust on soil
which also inhibits suspension, this crust is readily destroyed by any
266
-------
mechanical action, such as the passage of a vehicle. Heavy or sustained
rainfall also is effective in washing paved streets. On the other hand,
the amount of mud transferred after rainfall from unpaved parking lots
and roads to paved roads by vehicles can result in a significant increase
in street dirt in some areas.
Wind Speed
The effect of wind speed is also complex. As the speed of the wind
increases, the effective volume of air available for dilution increases
and, for constant source strengths, downwind concentrations tend to be
inversely proportional to wind speed. However, in the case of par-
ticulates, total emissions are not invariant with windspeed since the
wind is the natural agent by which soil and dust particles are resuspended.
The amount of fugitive dust resuspended depends on the moisture content
and nature of the soil (or dust) and the wind speed. At speeds below
10 to 15 miles per hour, however, the amount is probably neglibible even
under dry conditions. At greater average speeds, and particularly under
gusty conditions, fugitive dust contributions can be substantial. Wind
speed also indirectly contributes to fugitive dust emissions by increasing
the rate of evaporation, and hence speeds the drying of the soil and dust
particles.
Wind Direction
On a local scale, the wind direction determines the polar distribution of
pollutants around their sources, and hence is requisite to the understand-
ing of source-receptor relationships and the design of source-specific
sampling networks. On a regional scale, the background concentrations in
the air masses entering an urban area have been determined by the past
history of the air mass which is best estimated by trajectory calculations,
although wind direction roses can also be helpful.
267
-------
Temperature
Temperature, and seasonal temperature patterns, are the best single
indicators of space heating requirements, and hence correlate with
particulate emissions from these sources. Emissions from city to city
for the same number of degree days will vary with the type of fuel burned.
Temperature plays a part in fugitive dust emissions by affecting the evap-
oration rate of water and by its influence on the growth of vegetation.
Solar Radiation
In addition to being the driving energy source for weather systems,
solar radiation relates to particulate emissions through temperature,
evaporation, plant growth, and the formation of photochemical aerosols.
Stability
The stability of the air, as measured by the change of temperature with
height, controls the rate of vertical turbulent diffusion and the mixing
depth. It therefore is a measure of the dilution power of the atmosphere
in the vertical. The diurnal variation of wind speed caused by the ver-
tical transfer of momentum is closely related to the diurnal variation
in stability.
Stagnations
Stagnating air masses permit the accumulation of pollutants and the
development of stable transport patterns.
The geographic distribution of these and other meteorological and clima-
tological factors influencing TSP levels will be discussed later in this
section.
268
-------
MAGNITUDE OF METEOROLOGICAL EFFECTS
Although the ways in which meteorological parameters affect TSP levels
are quite well understood, the magnitudes of the various impacts have not
yet been well defined. Among the reasons for this lack of definition are
the complexity of the interrelationships among the meteorological param-
eters, the confounding effects of local particulate sources, and a general
lack of daily TSP observations. This study attempted (1) to investigate
quantitatively the effects of precipitation and wind speed (the most
obvious meteorological influences) on day-to-day changes in concentra-
tions, and (2) to make rough estimates of the impact of annual variations
in meteorology when feasible. These results must be considered prelim-
inary, but they do provide some guidance in judging meteorological impacts.
3hort-Term Effects of Wind Speed and Precipitation
The most detailed day-to-day analyses were carried out using precipitation
(expressed as its water equivalent in the case of snow), average wind
speed, and TSP observations in Birmingham and Denver, with supplemental
analyses in several other cities. Birmingham was selected because of its
large data base with daily observations and the presence of major industrial
sources; Denver was selected because of its recognized fugitive dust prob-
lem. Of the 14 cities studied, Denver has the lowest average annaul
precipitation (15.5 inches) while Birmingham, with 53.2 inches, has next
to the highest, being exceeded only by Miami, with 59.8 inches. In these
studies of short-term variations, the effects of amount and frequency of
precipitation and of average daily wind speed were investigated using
local airport meteorological data. TSP levels were handled as concentra-
tions in some cases and as ratios of 24-hour concentration to a smoothed
average concentration in other cases to minimize seasonal effects and
normalize the results. Weekend periods were omitted from the analyses
because of the existence of strong weekly cycles in average TSP levels.
269
-------
Results in Birmingham TSP levels at three sites were analyzed in
Birmingham. Of these, one is industrial (North Birmingham), one is
within the Central Business District (Downtown Birmingham), and the
third (Mountain Brook) is in a residential area well removed from in-
dustrial sources and the principal center of urban activity. Data for
1974 only were analyzed at the two Birmingham sites where observations
were made daily. At Mountain Brook, where observations were made every
6th day, 1973 data were also included to broaden the data base.
Daily TSP levels were plotted against wind speed at each site for three
48-hour precipitation categories: (1) £0.02 inch, (2) 0.03 - 0.24 inch,
*
and (3) >0.25 inch. Figure F-l shows the results of little or no rain-
fall (£ 0.02 inch) using Downtown Birmingham as an example. Two features
of Figure F-l are apparent. First, the range of concentrations is very
large at the lower wind speeds but decreases greatly at higher wind
speeds. Second, the average concentration decreases with increasing
wind speed for speeds up to about 8 knots and then remains essentially
constant. Furthermore, the decrease of average concentration with wind
speed, as shown by the dashed line, is approximately linear from 2 to
8 knots, as would be expected if dilution were the controlling influence.
The change in slope shown at 2.5 knots can be attributed to stagnation
conditions at very low wind speeds. The plot of average concentration
versus wind speed, shown in Figure F-l for the Downtown Birmingham site,
was transferred to Figure F-2 and similar curves added for the North
Birmingham and Mountain Brook sites. It can be seen that the curve for
North Birmingham reflects the same relationships as that for Downtown
Birmingham, but that the average concentration at Mountain Brook appears
to be invariant with wind speed. Thus one might postulate that the
The 48-hour precipitation period was selected initially on the basis
of work done by Lazenka and Weir of the Philadelphia Air Management
Services Laboratory. This selection is supported by results of this
present study, reported later.
270
-------
*>
6
^
4.
^
|
oc
H
CONCEN
&.
ouu
280
260
240
220
200
180
160
140
120
100
80
60
40
20
O
/
-
- X *
\
\ . t .
1
\
\
I
\ *
- v..
x
N.
. :.:.^: '
* * ^
0 > *
« »x
*»** ^ »» ** *
i *^ _ A
«. * * N
**;*»***
*
-
i t t i i i i i i i i i i
56789
WIND SPEED .knots
10 II 12 13 14
Figure F-l. Relationship between TSP le ols and wind speed at the
Downtown Birmingham site on days with 48-hour precip-
itation amounts
-------
450
400
350
10
v
300
250
200
a.
{2 iso
100
50
-NORTH BIRMINGHAM
(INDUSTRIAL)
\
V
DOWNTOWN BIRMINGHAM
- (COMMERCIAL)
_? _.__._
MT. BROOK (RESIDENTIAL)
x* NORTH BIRMINGHAM
o = DOWNTOWN BIRMINGHAM
=MT. BROOK
J I
5 6 7 8 9 10 II 12 13 14
WIND SPEED, knot*
Figure F-2.
Relationship betwee'p TSP concentrations and wind speed
at three selected sites in Birmingham on days with 48-
hour precipitation amounts _< 0.02 inch
272
-------
Mountain Brook mean concentration of 52 ug/m represents a concentration
level below which the Birmingham mean concentrations will not drop by
dilution. The tendency for particulate concentrations to be invariant
with speed for speeds about 10 to 15 miles per hour, or perhaps to in-
crease at high wind speeds has also been reported by Abel in an inves-
tigation of particulates in Chicago. In this study Abel postulated that
the departure from the dilution phase where concentration is inversely
proportional to wind speed represents the contribution from fugitive
dust emissions.
To avoid possible influences from seasonal patterns, Figure F-3 shows
the relationship between relative TSP concentrations and average wind
speed under dry conditions at the Downtown Birmingham site using the
ratio of the 24-hour concentration to the running 5-week mean weekday
concentration as the measure of relative concentration. Again, as in
Figure F-l, the dilution effect appears to dominate the relationship
below wind speeds of about 9 or 10 knots. Above this speed, concentra-
tions average about 70 percent of the 5-week mean, as was also the case
when the procedure was carried out using Chattanooga data.
A number of other possible relationships among TSP levels, wind speed,
and precipitation at the three sites were also investigated using the
Birmingham data. Table F-l shows the geometric mean concentration for
five precipitation classes at each of the sites and the mean value of
the ratios of 24-hour to 5-week average concentration for the precipita-
tion classes at the North and Downtown Birmingham sites. As expected,
concentration decreases with increasing precipitation amounts at North
Birmingham and Downtown Birmingham; however, no systematic change is
shown by the more limited Mountain Brook data. At North Birmingham,
the average decrease in concentration which occurred between periods when
the precipitation on the observation and previous day totaled less than
0.03 inches and when the total exceeded 0.24 inches was 46 percent, while
at Downtown Birmingham the average concentration decreased 32 percent.
273
-------
9.0
*
Z.O
0.7
0.6
O.4
O.3
0.2
« :
.
*:-
I 1 I I
6 7 t
WIND SPEED, kMtt
10 II 12 IS
Figure F-3.
Comparison of TSP levels to average concentration
as a function of wind speed at the Downtown Bir-
mingham on days with 48-hour precipitation amounts
£ 0.02 inch
Table F-l. RELATIONSHIP BETWEEN 48-HOUR PRECIPITATION
AMOUNTS AND TSP CONCENTRATIONS AT SELECTED
BIRMINGHAM SITES, MONDAY THROUGH FRIDAY
48 -hour
precipitation
la.
0
0.01-0.02
0.03-0124
0.25-0.99
>1.00
TSP levels
North Birmingham
n
99
27
38
56
28
Hg/m
195
141
128
106
103
Ratio
1.14
0.88
0.75
0.66
0.65
Downtown Birmingham
n
100
27
39
57
26
jig/m
117
96
85
80
75
Ratio
1.08
0.88
0.81
0.77
0.72
Mountain BrooW
n
43
3
12
17
4
Hg/tn
53
41
56
41
46
Note: Concentration* and ratios are geometric mean values.
274
-------
Since both wind speed and precipitation clearly affect particulate
concentrations, concentrations were examined for various wind speed-
precipitation regimes. Figure F-4 shows the effect of 48-hour precip-
itation amounts on TSP levels as a function of wind speed at the North
Birmingham site. Mean values were plotted for selected wind speed
classes and three precipitation categories. It is apparent from Fig-
ure F-4 that substantial reductions in concentration were brought about
by precipitation at wind speeds below about 9 knots, but that at higher
wind speeds concentrations show little variation with either precip-
itation or wind speed. Results obtained at the Downtown Birmingham
site were similar. Thus suggests that an average minimum value for a
given site can be reached either by increasing the wind speed to greater
than about 9 knots, or, except at quite low wind speeds, by increasing
the amount of precipitation.
40O
300
i
«.
100
< 0.03 in
NORTH BIRMINGHAM
X«<0.03 In
o r 0.03 -0.24 in
00.24 in
o
I
I
5678
WIND SPEED, knots
10
13 14
Figure F-4. Effect of 48-hour precipitation amounts on TSP concentra-
tions at various wind speeds at the North Birmingham site
275
-------
If the leveling off of the average concentration curve at higher wind
speeds is due to fugitive dust emissions, as postulated, concentrations
at these speeds would also be expected to respond strongly to precipita-
tion. The rather minimal reduction actually observed perhaps is the
result of the small sampling size available for these meteorological
conditions. A large part of the reductions observed at speeds below
9 knots may reasonably be attributed to the combined effects of washout
and reduced contributions by vehicular-induced reentrainment.
In a further attempt to isolate the contributions of fugitive dust to
Birmingham TSP levels, geometric mean concentrations were calculated at
the downtown and industrial sites for three wind speed classes under four
different precipitation regimes. The first regime represented dry con-
ditions in which the 48-hour precipitation amounts did not exceed 0.02
inch, and hence washout would be negligible but vehicular-induced reentrain-
ment would be easily accomplished and fugitive dust emissions would be
possible at the higher wind speeds. The three other regimes all begin with
a day having at least 0.10 inch of precipitation on the day prior to the
observation day but with different amounts of precipitation on the observa-
tion day itself (0 - 0.02 inch, 0.03 - 0.024 inch, and >0.25 inch).
Table F-2 presents the results. Unfortunately, the number of observations
on days with high average wind speeds following days with 0.10 inch of
precipitation are very few. A number of general comments can be made
based on the table, however. First, the dilution effect of wind speed is
obvious under all regimes except heavy precipitation. Second, the effect
of precipitation on concentrations the following day, if dry, appears to
be very strong at the North Birmingham site but negligible at the downtown
site (compare columns b and c). And third, the effect of moderate or
heavy precipitation on the day of observation again is most obvious when
winds are light, a condition which is most faborable for the washout of
local particulates. In interpreting Table F-2, it must be remembered
that certain wind speed and precipitation regimes are associated with
weather situations which may have additional effects on TSP levels. For
276
-------
Table F-2. AVERAGE TSP LEVELS FOR VARIOUS WIND
SPEED AND PRECIPITATION CATEGORIES
AT TWO BIRMINGHAM SITES
a. North Birmingham
Wind speed,
knots ,
a
2.0 - 4.9
5.0 - 8.9
9.0 - 13.9
n
Geometric
mean
n
Geometric
mean
n
Geometric
mean
48 -hour
precipi-
tation,
in.
0 - 0.02,
b
74
210
39
145
9
106
Precipitation on observation day
following days with at least
0.10 in. of precipitation, in.
0 - 0.02,
c
16
167
14
95
2
74
0.03 - 0.24,
d
7
140
4
108
1
37
£0.25,
e
9
94
3
87
3
78
b. Downtown Birmingham
T T J J
Wind speed
knots,
a
2.0 - 4.9
5.0 - 8.9
9.0 - 13.9
n
Geometric
mean
n
Geometric
mean
n
Geometric
mean
4 8 -hour
precipi-
tation,
in.
0 - 0.02,
b
74
132
40
85
10
71
Precipitation on observation day
following days with at least
0.10 in. of precipitation, in.
0 - 0.02,
c
16
129
14
87
2
76
0.03 - 0.24,
d
7
111
4
72
1
52
£0.25,
e
9
59
3
84
3
62
277
-------
example, the highest concentrations of all tend to occur under stagnating
conditions typified not only by light winds and no precipitation but also
by subsidence and little vertical exchange in the lower atmosphere. Addi-
tionally, some of the differences between the various columns of Table F-2
may be caused by systematic differences in wind directions which can affect
not only the impact of local sources but control entering background con-
centrations as well. One should therefore be cautious about using the
differences in concentrations shown in Table F-2 to speculate in detail on
the relative contributions of various source categories. Note, however,
that approximately 30 percent of the loading measured at North Birmingham
during the last half of a dry 48-hour period appears to have been eliminated
by a moderate amount of precipitation prior to the measurement period,
whereas no change was noticed at Downtown Birmingham. This indicates a sub-
stantial difference in the source characteristics of particulates reaching
the two sites. Further, the fact that the North Birmingham reduction oc-
curred at low as well as moderate wind speeds indicates that this part of
the loading was not entirely wind induced.
The data at the two sites were examined in more detail to determine the
average length of time precipitation is effective in reducing TSP levels.
For this investigation, the ratio of observed 24-hour concentration to the
5-week running average concentration was used as the measure of concentra-
tion to permit easy comparison with results in other areas. The ratios
were divided into six precipitation regimes and averaged, giving the results
shown in Figure F-5. The calculations were carried out twice, using dif-
ferent levels of 24-hour precipitation to classify a day as one with pre-
cipitation. In one set of calculations, any day with measurable precipita-
tion (0.01 inch) was considered to have had precipitation; in the second
set of calculations, only days with at least 0.25 inch of precipitation were
considered as days with precipitation. Figure F-5 shows that, on the aver-
age, TSP levels remain depressed the day after rain at North Birmingham.
At Downtown Birmingham the levels return to near normal more quickly, as was
pointed out earlier in the discussion of Table F-2. The effect of rainfall
278
-------
1.50
1.40
130
o 1.20
i-
K 1.10
Ul
3 i.oo
ac.
Ill
> 0.90
<
0.80,
0.70
0.60
>O.OI IN.
(29)
(10
146)
I 2 3 4 5 OR MORE
NUMBER OF DAYS AFTER PRECIPITATION
a) NORTH BIRMINGHAM
0.01 IN.
I 2 3 4 5 OR MORE
NUMBER OF DAYS AFTER PRECIPITATION
b) DOWNTOWN BIRMINGHAM
Figure F-5.
Duration of rainfall effectiveness in reducing TSP
levels at two Birmingham sites. Number of observa-
tions is shown in parentheses. Ratio is ratio of
24-hour concentration to 5-week running mean
concentration
279
-------
of different intensities is shown by the difference between the two curves
for each site. The occurrence of the peak average ratio at 4 days after
precipitation at North Brimingham reflects a few very high ratios apparently
associated with periods of dry, light-wind, poor-dispersion conditions
lasting for several days.
Results in Denver - An analysis somewhat comparable to that carried, out
with Birmingham data was also carried out for Denver; the principal dif-
ference was in the preparation of the data bases. The Denver monitoring
network is operated on a 4-day schedule and a 4-year time period was chosen
(1971-1974), instead of the single 1974 base year used in Birmingham, to
increase the reliability of the results. The second major difference in
the data sets was that a city-wide TSP index was used in Denver rather
that attempting to differentiate effects among site types. This TSP
index was obtained by averaging concentrations within the valley at five
locations extending from Englewood in the south to Adams City in the north.
These sites were: Englewood, Gates Building, School Administration Building,
Sewer Plant, and Adams City. Again, weekend data were not used in the
analysis. Data for the 4 years were pooled to obtain the reference con-
centration used in calculating the concentration ratios. The procedure
followed was to calculate 4-year average monthly means and then to smooth
these monthly averages slightly by means of a weighted running average
(1-2-1).
The results of the Denver analysis again show that precipitation has a
measurable effect on TSP levels, but in this case any relationship between
concentration and wind speed is largely obscured. Table F-3 gives the
average concentration for five 48-hour precipitation classes. The reduc-
tion shown by these average values in going from completely dry periods
3
to periods with £0.25 inch of precipitation is 74 |ag/m or 48 percent.
It should be borne in mind, however, that precipitation amount, in addi-
tion to being a measure of washout and the suppression of fugitive dust,
is to some extent indicative of the more general meteorological conditions
280
-------
affecting diffusion rates as well. The average ratio between 24-hour con-
centration and the smoothed monthly mean concentration also shows a sys-
tematic decrease with increasing precipitation. A breakout of those
observation days with precipitation that follow days with at least 0.10
inch of precipitation, as was done in Birmingham (Table F-2), yielded too
few cases for meaningful conclusions.
Table F-3. RELATIONSHIP BETWEEN 48-HOUR PRECIPITATION
AMOUNTS AND TSP CONCENTRATIONS IN DENVER,
MONDAY THROUGH FRIDAY
48 hour
precipitation,
in.
0
Trace
0.01 - 0.02
0.03 - 0.24
> 0.25
5-station average
n
96
43
27
44
22
ug/m
154
132
116
102
80
Ratio
1.10
1.02
0.92
0.80
0.61
Note: Concentrations and ratios are
geometric mean values
Figure F-6 shows plots of the 5-station average TSP concentration and of
the 24-hour to smoothed-monthly-concentration ratios against average air-
port wind speed for days with 48-hour precipitation amounts £ 0.02 inch.
The lack of any useful relationship between wind speed and these measures
of TSP level is obvious, with the possible exception that no above-average
concentrations were observed when winds were in excess of 15 miles per hour.
To estimate the period of time over which precipitation is effective in
reducing TSP levels in Denver, the 24-hour to monthly-mean-concentration
ratios were grouped according to the number of days after precipitation
and averaged. The following three precipitation levels were used to define
a day with precipitation: (a) £0.01 inch, (b) £0.10 inch, and (c) £0.25
inch. The results are shown by the curves in Figure F-7, all of which
281
-------
400
10
300
200
iii
o
o
u
100
;v ..'- . * .
**"« *"t v \ '
vJ ***'::';*;.'**.*".
* . ** 'i '
* ' ..
4 6 ' 8 10 12 14 16
WIND SPEED, mph
18 20
? K
2.0
1.5
a:
0.9
0.8
0.7
0.6
0.5
0.4
*
»
L.
1
8 10 12 14
WIND SPEED, mph
16
18 20
Figure F-6.
Relationship between areawide TSP levels
in Denver and wind speed on days with 48-
hour precipitation £.0.02 inch. Ratio
is the ratio of 24-hour concentration to
smoothed monthly concentration
282
-------
o
<
0.01 inches
Ul
O
a:
UJ
1.20
MO
1.00
0.90
0.80
0.70
0.60
I Z 3 4 5 OR MORE
NUMBER OF DAYS AFTER PRECIPITATION >0.10 inches
o
£
Ul
O
<
IE
Ul
l.30i
1.20
1.10
1.00
0.90
0.80
0.70
0.60
0.50
(c)
(184)
1 2 3 4 5 OR MORE
NUMBER OF DAYS AFTER PRECIPITATION > 0.25 inches
Figure F-7. Duration of precipitation effectiveness in re-
ducing TSP levels in Denver. Number of ob-
servations is shown in parentheses. Ratio is
the ratio of 24-hour concentration to smoothed
monthly concentration
283
-------
indicate a recovery time of between 2 and 3 days. It is of interest to
examine the seven values that make up the high average ratio found the
fourth day after precipitation >0.25 inch in Figure F-7- To investigate
the possible contribution of sanding operations to this high average
ratio-, precipitation amounts and type have been listed for the 5-day
periods ending with each observation day in Table F-4. Of the 4 days
with ratios greater than 1.0, 3 had some snowfall on the third and fourth
days prior to the day that the higher-than-average concentrations were
observed, and the fourth had snow 2 days before the observation day. Of
the 3 days with ratios lower than 1.0, 2 had precipitation in the form
of rain only during the 5-day period, and the third had snowfall on the
day of observation which accumulated to 2.3 inches over a 13-hour period.
Although the number of cases reported here are few, the results support
the view that following sanding operations a widespread increase in par-
ticulate levels occurs as a result of vehicular induced reentrainment
once the streets have become dry.
Table F-4. PRECIPITATION AMOUNTS AND TYPE
FOR SELECTED 5-DAY PERIODS IN
DENVER
Character of
precipitation
Water equiv.
Snow, ice
Water equiv.
Snow, ice
Water equiv.
Snow, ice
Water equiv.
Snow, ice
Water equiv.
Snow, ice
Water equiv.
Snow, ice
Water equiv.
Snow, ice
Days prior to observation
precipitation, inches
4
0.41
4.9
0.79
2.1
0.56
0
0.46
4.6
0.40
3.4
0.31
0
0.64
0
3
0.17
3.4
0.09
0.9
0.01
0
T
T
0.20
2.4
0
0
T
0
2
0
0
0.1
0.1
0.10
0
0
0
0.06
0.7
0
0
0.21
1.0
1
0
0
0
0
0.19
0
0
0
0
0
T
0
0
0
Obs.
day
0.09
0.7
0
0
0.02
0
0
0
0.13
2.3
0
0
0
0
TSP
ratio
1.56
1.58
0.91
2.07
0.71
0.76
1.25
Date of
TSP obs.
4/06/73
5/04/73
7/23/73
1/25/74
3/22/74
9/06/74
10/16/74
284
-------
Results in Other Cities - Similar, but more limited, studies of precipita-
tion anji wind speed effects on TSP levels were carried out using Chatta-
nooga and Oklahoma City data. For the Chattanooga study, 1971 TSP levels
measured at the City Hall were used. In Oklahoma City, 1974 data for
three sites were used. Two of these were in the center city: Site No. 17
was at 800 N.E. 13th and Phillips Street, and Site No. 18 was a traffic-
oriented site at 2045 N.W. 10th Street. The third site (No. 5) was located
in the small community of Jones in argicultural land northeast of the city.
Even though the number of observations at each of these three sites was
quite limited, the sites were examined separately to see if any effect of
site differences would appear. The results, presented in the same format
as those of Birmingham and Denver, are given in Figures F-8, F-9, and
F-10 and Table F-5. Briefly, the relative TSP level as indicated by the
ratio of the 24-hour concentration to an average concentration over an
approximate 5-week period is nearly invariant with wind speed up to about
8 knots in Chattanooga. At Chattanooga, the time needed to return to
average TSP levels after precipitation is 2 days when all precipitation
days are considered and increases to about 3 days when only days with
>0.25 inch of precipitation are considered. The sketchy Oklahoma City
data suggest a more rapid return to normal levels. Table F-5 shows a
3
decrease in TSP levels of 22 (ig/m or 22 percent between days with neg-
ligible 48-hour precipitation amounts and days with >.0.25 inch of
precipitation at Site No. 17, but a decrease of only 9 ug/m or 9 per-
cent at Site No. 18 between corresponding periods, and no change in
average level at Jones, where the mean concentration is low under both
precipitation regimes.
285
-------
9.U
2.5
2.0
1.5
0.8
0.6
0.4
r . SITE NO. 18
-
_
*
"". .* '. * ""
_ ^ , t - » -
r './
i i i i i i i i i i i
8 10 12 14 16 18 20 22 24
WIND SPEED,mph
2.5
2.0
1.5
0.8
0.6
0.4 -
SITE NO. 17
6 8 IO 12 14 16 18 20 22 24
WIND SPEED,mph
2.5
2.0
1.5
0.8
0.6 -
0.4 -
SITE NO. 5
16
18 20 22 24
Figure F-8,
4 6 8 10 12 14
WIND SPEED, mph
Relationship between wind speed and relative TSP
level at three Oklahoma City sites on days with
negligible precipitation. Ratio is ratio of
24-hour concentration to 5-week average
concentration
286
-------
3.0
2.5
1.5
1.0
°-9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
*
468
WIND SPEED,knots
10
12
Figure F-9.
Relationship between wind speed and relative TSP
levels at the Chattanooga City Hall on days with
negligible precipitation. Ratio is ratio of 24-
hour concentration to 5-week average concentration
287
-------
1234
DAYS AFTER PRECIPITATION
a)CHATTANOOGA
5 OR MORE
I 2 3 4 5 OR MORE
DAYS AFTER PRECIPITATION fcO.25 inches
b)OKLAHOMA CITY
Figure F-10. Duration of precipitation effectiveness in reducing
TSP levels in Chattanooga (a) and Oklahoma City (b).
Number of observations is shown in parentheses.
Ratio is ratio of 24-hour concentration to 5-week
average concentration
288
-------
Table F-5. REIATIONSHIP BETWEEN 48-HOUR PRECIPITATION
AMOUNTS AND TSP CONCENTRATIONS AT THREE
OKLAHOMA CITY SITES, MONDAY THROUGH FRIDAY
48-hour
precipitation,
inches
0 - Trace
0.01 - 0.24
> 0.25
TSP levels
No. 17
800 N.E. 13th
and Phillips
n
42
7
13
, 3
ug/m
101
88
79
No. 18
2045 N.W. 10th
n
43
6
12
/ 3
M-g/m
103
105
94
No. 5
Jones
n
34
8
5
ug/m
39
44
39
A different analytical approach was used by Lazenka and Weir to investi-
gate the effect of heavy rainfall on TSP levels in Philadelphia. In this
study TSP levels in 1972 and 1973 from three stations (500 South Broad
Street, AMS Laboratory, and the Southeast water treatment plant) measured
on the day after rain, the second day after rain, and the third day after
rain were compared with those measured the seventh day after rain. Days
of rain had to have at least 0.25 inch of rain during the day or a rate
of 0.1 inch or more in 1 hour. Other parameters such as wind direction
and speed were matched. The study concluded that, after periods of heavy
rain, particulate levels were significantly reduced on the first day (95
percent confidence) and the second day (90 percent confidence) when com-
pared with the seventh day after rain but that, by the third day after
rain, no significant difference was found. Over the 2-year period, the
concentrations on the first day averaged about 62 percent of the seventh
day values.
Conclusions From the above analyses, some of which are admittedly sketchy,
we draw the following conclusions for interim guidance until more defini-
tive analyses have been carried out. In carrying out these analyses, the
data were sometimes stratified by precipitation only and sometimes by pre-
cipitation and wind speed. No other attempt was made to separate the ef-
fects of interrelated meteorological parameters. As a result, estimates of
289
-------
the reductions attributed to precipitation may be biased due to typical
sequences of weather and air mass changes associated with precipitation
events.
Precipitation is very effective in reducing TSP levels
in highly polluted areas. In these areas, concentrations
measured during the last half of a 48-hour period with
precipitation j> 0.25 in. averaged approximately half
of concentrations measured during 48-hour periods with
negligible precipitation. In low pollution areas where
concentrations are generally below 60 pg/m3, the reduc-
tion was too slight to be detected.
At locations with high TSP levels, average concentrations
decrease steadily with increasing 48-hour precipitation
amounts.
Precipitation is most effective in reducing TSP levels on
the day it occurs. Concentrations at typical urban sites
(excluding clean residential areas) on days with measurable
precipitation are about 75 to 85 percent of average values
for the site and time of year.
The effect of precipitation lasts, on the average, about
2 days, with the duration varying somewhat with location
and the amount of precipitation.
In nonindustrialized urban areas, average TSP levels do not
appear to be related to airport wind speeds.
In industrial areas where large contributions are made by
point sources, average TSP levels decrease with increasing
wind speed up to speeds of about 10 miles per hour and
then remain invariant with wind speed.
The above conclusions have been drawn from average results. An inter-
pretation of individual observations would require detailed examination
of meteorological and emission data.
Effects of Annual Variations in Precipitation and Temperature
Multiple regression analysis was used to estimate the effects of annual
variations'in meteorological parameters over the 5-year period from
1970 to 1974 on the annual mean TSP level in the study cities. The
three meteorological parameters initially considered as independent
290
-------
variables were precipitation, temperature (heating degree days), and
wind speed. After initial calculations, plus reflection on the apparent
lack of short-term correlation between wind speed and TSP level except
in industrial areas, it was decided to exclude wind speed from the
analysis. Allowance for a linear trend was made by designating 1970 as
year 1, 1971 as year 2, and so on.
The data available for the 5-year analysis consisted of 11 sets of five
*
annual averages for 10 of the TSP study cities. Each set contained TSP
annual means for each of three site types (industrial, center city, and
suburban/residential) plus the meteorological data. Precipitation was
entered as inches per year and heating demand as degree days in thousands.
In the analysis for which the results are given here, each site type aver-
age in each city was treated as a separate observation. Dummy variables
were used to permit different intercepts for the several cities and
the three site types, while the meteorological effects were estimated
based on data from all cities. This approach in effect assumes that the
meteorological parameters operate in roughly the same manner throughout
the country, which is clearly neither an obvious nor a trivial assumption.
Previous analyses that permitted the meteorological effects to differ from
city to city did indicate that the effects found in the various cities
were quite similar in magnitude, and it is on this basis that the assump-
tion was made.
The resulting equation was:
TSP = C - 2.9Y - 0.43P + 2.5T
where TSP = annual geometric mean concentration in
3
C = constant in ug/m
Y - year, 1 to 5
Cincinnati, Cleveland, Denver, Miami, Philadelphia, Providence, St. Lcuis
(both Mo. and 111.), San Francisco, Seattle, and Washington, D.C.
291
-------
P = annual precipitation in inches
T = heating degree days in thousands.
Table F-6 gives values of the constant C for the 11 urban areas and three
site types.
Table F-6. CONSTANTS FOR REGRESSION EQUATION RELATING
TSP LEVELS TO METEOROLOGICAL PARAMETERS
Urban area
Cincinnati
Cleveland
Denver
Miami
Philadelphia
Providence
St. Louis (Mo.)
St. Louis (111.)
San Francisco
Seattle
Washington, D.C.
Site type
Suburban/
residential
84
122
93
83
92
66
73
98
53
58
59
Center city
100
138
109
99
108
82
89
114
69
74
75
Industrial
124
162
133
123
132
106
113
138
93
98
99
The implication of the general regression equation is that, within the
2
study cities, concentrations have been lowering at the rate of 2.9 yg/m
per year over the last 5 years, that an increase of 1 inch in annual pre-
3
cipitation decreases the mean concentration by 0.43 yg/m , and that an
increase in heating degree days of 1000 increases the mean concentration
o
by 2.5 yg/m . This result can be used to get some feel for the magnitude
of the effects of annual changes in precipitation and temperature on TSP
levels within a city, and for the differences to be expected among cities
due to these two meteorological variables. Examination of the variations
in precipitation and temperature over the 5-year period in each of the
14 study cities showed that the smallest range in precipitation (6 inches)
292
-------
occurred in St. Louis while the greatest range (27 inches) occurred in
both Chattanooga and Providence.
The minimum range for heating degree days was 188 in Miami and the maxi-
mun range was 1189 in Cleveland. The implied differences in TSP levels
resulting from these annual variations in precipitation range from 2.6
3 3
yg/m in St. Louis to 11. 6 yg/m in Chattanooga and Providence, and the
differences resulting from variations in heating requirements range
3 3
from 0.5 yg/m in Miami to 3.0 yg/m in Cleveland.
It should be recognized that this equation was developed using both var-
iations within the 10 individual cities and among these cities. The
equation is an attempt to generalize the relationships between TSP le-
vels and the three variables and should therefore not be expected to
have great precision at any one location. Furthermore, no attempt was
made to account for differences among the various cities in the mix of
fuels used for space heating.
Seasonal Variations
Just as changes in the annual levels of precipitation and heating degree
days can affect the annual TSP levels, so too may the seasonal variations.
Whereas changes in climatological factors can be 20 percent or more on
a yearly basis, they can often vary by an order of magnitude between
seasons and months. Therefore, the impacts that meteorology and clima-
tology have on TSP levels, and thereby the implications for control
planning, are much more important on a seasonal basis.
In most of the city reports (Volumes III to XVI), various meteorological
influences were investigated for each city by comparing the 1974 monthly
mean TSP levels with the monthly meteorological data. In this discussion
seasonal variations of several of the readily available meteorological
parameters influencing TSP levels are illustrated by presenting normal
monthly values (based on a 30-year period) for the 14 study cities.
293
-------
Because of the wide geographical distribution of these cities, these
variations provide a fair sample of the magnitude of seasonal changes
likely to be encountered nationwide. In addition, 1974 monthly preci-
pitation and TSP concentrations are presented graphically so that major
relationships or lack of relationships between the two variables during
a specific year may be seen.
The meteorological parameters presented by monthly averages are: wind
speed, total precipitation, number of days with measurable precipitation,
and degree days. The mean morning and afternoon mixing heights, taken
Q
from Holzworth, are also given, but by season only. The monthly data
are provided in bar graph format in Figures F-ll to F-16 and the mixing
height data are listed in Table F-7. The mixing height data in Table F-7
for Denver, Seattle, Miami, Oklahoma City and Washington are given as
tabulated by Holzworth; data for the remaining cities were interpolated
from his isopleth maps.
Figure F-ll shows mean monthly wind speeds for the 14 cities. Signifi-
cant features of this display are: (1) the annual average speed among
the cities varies by a factor of two and ranges from 6.3 miles per hour
in Chattanooga to 12.9 miles per hour in Oklahoma City, and (2) the
seasonal variation pattern differs considerably from city to city.
However, with the notable exception of San Francisco, wind speed is at a
maximum during the spring or winter months, and at a minimum during the
summer. In the case of Chattanooga, Birmingham, and Seattle, the minimum
extends into the fall season.
Normal monthly precipitation amounts for the 14 cities are shown in
Figure F-12. Annual precipitation among the cities varies by a factor
of 4, ranging from a low of 15.5 inches in Denver to a high of 59.8
inches in Miami. The seasonal variation is highly dependent upon the
location of the city. Extreme examples are the west coast cities with
summertime minimums and Miami where large amounts of rain occur from
May through October. Figure F-13 shows the average number of days with
precipitation amounts equal to, or more than, 0.01 inches, and the
294
-------
VD
JAN
FES
VAR
APR
1 BALTIMORE
I
1
1
! BIRMINGHAM
MAY l_i
JUN
JUL
AL;G t
SEP jl
OCT "I
NOV j_
DEC
[EEEH
AVG. 9.3
1
ZZZD
AVG. 7.4
"1
1 CMATTANOOei
I
I
AVG. 6.3
1 1 ' 1
1 CINCINNATI
i
U
H AVG. 9.1
1
CLEVELAND
AVe
0 4 " A n 16 i.O 0 4 8 12 16 10 0 4 8 12 16 20 0 4 8 12 16 20 0 4
12 16 20
JAN I
FESL.
MJS L
APP! j
MAY r j
JUN I 1
! DENVER
1 MIAMI
r-
AVG 90
AUG
SEP
OCT
NOV
DEC
JAN j_
FE8L
MASu
AP« j-
MAY L
J'JNl
FLT-
AVG. 9.1
I_J
J OKLAHOMA CITY
AVG 12.9
.1
J PHILADELPHIA
1
AVG. 9.8
o
i
12 16 20 C> 4 a 12 16 20 0 4 6 12 16 20 0
.
J PROVIDENCE
AVO. io.e
6 12 16 20 0 4
, I . i
12 It 20
) ST. LOUIS
_ 1
-------
JAM
ro
MAR
APR
KAY
JVNB
JULY
AUC
SEW
ocr
NOV
DEC
JAN
FEE
MAK
AFR
MAY
JUNE
JULY
AUC
SEPT
ocr
NOV
DEC
JAN
mrm
.^
MAY
JUSE
JVLY
AUC
SEPT
OCT
NOV
DIG
J BALTIMORE
1
1 (40.5)
1
1
]
1
i
1
1
1
_l_ 1 _l_ 1 1
0 2 4 6 8 10
D DENVER
Z)
, 1 (1>.J)
ZZD
1
]
1
)
ZD
Z]
I]
1 ! . 1
0 2 4 6 8 10
J ST. LOUIS
]
T ,_. -*
V 7. )
I
1
J
1
1
^
. «>^^
^
_^
I
12468:
0
1 BIRMINGHAM
J
1 (53.2*
1
1
1
1
1
~\
1
J
i ill i
D 2 4 6 8 10
1 MIAMI
' ]
1 (59.8)
T
1
J
1
1
J
'I
1
It r l i
0 2 4 6 8 10
JSAN FRANCISCO
\
J
D
1
"^
"~
i 11
i' i- 1 is i
i
0
1 CHATTANOOGA
1
1 «!.»)
1
1
~|
1
J
1
1
1
1 . 1 . i
> 2 4 6 8 10
1 OKLAHOMA CITt
' 1
1 (11.4)
" ' ' 1
I
1
1
I
1
^J
1
1 i l l i
D 2 4 6 8 10
I SEATTLE
1
I /
-------
NJ
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUC
SEP
OCT
NOV
DEC
JAN
FE8
MAR
APR
MAY
JUN
JUL
AUG
CCT
NOV
DEC
i:
JAN
FEB
MAP-
APR
MAY
JUN
JUL
AUO
SEP
OCT
NOV
DEC
0
.A 1 BALTIMORE
::>. \
:.* \
1*
1
1
1 1
"l«
/f 1 ~1 1 1 \ '
'* BIRMINGHAM
1*
1*
1
1
}
1
|
ii13
i *
, i* .
H _ J CHATTANOOGA
3 1
1*
1 *
1
1
TOTAL '121/2
1
J
]*
. I* . .
'/A J CINNCINNATTI
VA \
4 1
J*
1
1
| TOTAL'129/7
i=!
i*
A 1
'A , 1 .
'<-. '/A \ ctFVr: AM
^'.^ !
X<- 1
r* -i
j*
i
| TOTAL I5b/l8
1
1
1*
''.\ \
_
V///A i i ili
D
i 10 1) 20 25 05 10 IS 20 25 0 3 10 15 20 25 0 J 10 IS 20 25 0 5 10 IS 20 2S
221 D DENVER
HED
'//.'t \
l»
]
] 10TAL-UB/IU
1
]
1 1
1 MIAMI
TOTAL '179/10
1
i
1
1
1
^=1 .
-, -.0 15 20 29 0 5 10 IS 2O 2
Ji_ i ST. LOUIS
"'- 1
1
= u1'
1
nnn
1
j
S '0 15 20 2
5 (
. 1 * SAN FRANCl'
J
1 1
in
3 TOTAL' 62 /*
5 (
CO
) 5 10 IS 20 25 (
a . 1 OKLAHOMA CITY
1
1
J
j TOTAL'82/3
3 1 i
> t 10 iS 20 2
M ' SEAT
'I \
'4 \
1»
' 1
1 TOTAL'ISI/S
1
1*
1*
n ; ; ; ri
) S 10 IS 20 2
i
TLE
7?, 1 pm(.AOfi PHI*
% 1
'A 1
_j *
i
i
1 TOTAL1 116/6
1
=?
1*
fl , 1 .
1
T!TT 1 pqnutnf»'rF
':x- I
l/'v'il t
J*
1
1
J IOIAL'I<24/IO
|
^^.
1
^ , . 1 .
) 5 10 IS 20 25 O S IO IS 20 tt
'I 1 WASHINGTON D
:-j I
1
~1
* TOTAL HI/9
1
1 1
c.
KEY
1 .TOTAL PRECIPITATION
L__J -DAYS
V777A DAYS WHERE PRECIP
K'r''-J iTATION OCCURS AS
SNOW OH ICE
SO S 10 IS 20 25
Q TOTAL SHOWS'TOTAL DAYS WITH PRECIPITATION O.OI'lnchtt OP MORE/TOTAL DAYS WITH PRECIPITATION AS SNOW OR ICE 1.0 Ineht* OR MORE.
* INDICATES PRECIPITATION AS SNOW OR ICE OCCURS. BUT LESS THAN ONE-HALF A DAY.
Figure F-13. Average number of days with measurable precipitation
in 14 study cities
-------
VO
00
COOLING HEATING
JAN
rn
MAR
Ant
MAY
JUNE
JULY
ADC
SEPT
oct
NOV
DEC
e
BALTIMORE
C
C
1
LZ
C
,_
1
1
1
D
1108/4729
]
ZD
1
i i 1
00 400 0 400 800 UO
COOLING
BIRMINGHAM
1
C
LZ
LZZ:
c
i
00 400
HEATING
1
1
D
1928/2844
D
1
i i
o 400 800 12
io B
COOLINC
C1IATTANOCK1A
d
LZZ
cz:
c
i
)0 400
IIEATINU
1
1
)
^
1
1636/3505
D
]
r< 1' ,
0 400 800 12C
» 8(
COOLINC
CINCINNATI
C
CZ
LZZ
C
[
1
0 400
HEATING
1
1
J
ZZD
D
1080/5070
]
ZD
J
i i i
400 800 IZ(
o e
COOLINC
CLEVELAND
[
C
LZ
C
(
00 400 0
HEATING
^
1
^
1
ZD
3
613/6154
3
ZZD
1
i 1
400 800 12<
rn
MAR
APR
MAY
JUNE
JULY
AVC
SEPT
OCT
NOV
ore
8
JAM
rn
HAt
APR
HAY
JUNE
JULY
AUC
sen
OCT
NOV
DEC
nirg
DENVER
C
n
c.
[
)0 400 C
ST. LOUIS
I
C
LZZ
LZZZ
LZZ:
c
t
i
00 400
i
1
J
1
ZD
D
630/6016
D
, 1
400 800 12
J
1
1
ZD
H
1475/4750
1
ZD
. , i
3 400 800 1,
00 8
too ac
MIAMI I
C
d
1ZZ
CZ
, r~
00 400 (
SAN
FRANCISCO
[
1
1
10 400 (
3
4038/206
1
400 800 12
1
3
ZD
ZJ-
H
J 108/3042
D
]
D
ZD
Tl i
) «00 tOO 12
oo t
00 8
CITY
[
C
cz:
CZ
c
1
JO 400
SEATTLE
C
[
t
00 400
1
1
~~1
Z)
1
1876/3695
D
) 'iob BUU \'i
.. .1
1
ZZJ
D
] 129/5185
]
D
ZZD
i i
0 400 800 12
'0 g
00 8
C
CZ
LZZ:
LZZ
c
1
10 400
WASHINGTON
D.C.
C
LZZ:
LZ
,
00 400 0
1
J
1
D
1104/4865
]
ZD
1 ,1
i 400 800 12'
.. _ ,]
1
Z3
]
1415/4211
Z]
, 1
400 BOO 12
iO fl'
DO
C
LZ:
LZ
[
6 400 (
i
J
i
13
}
512/5972
D
400 600 12!
Figure F-14. Normal monthly degree days in 14 study cities (Base 65 F).
Annual degree days are given as: cooling/heating.
-------
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
JAN
FEB
MAR
APR
MAY
JUN
JUL
SEP
OCT
NOV
DEC
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
. 1 BAITIU
1
1
1
1
) (76)
1
1
1
1
1
, ,| , (
ORE
1 RIRUIK)
1
1
1
1
1 (107)
1
1
1
1
J
. 1 .
SHAM
,..) CHATTANOOI
1
1
,J
1
1 (64)
i
1
1
1
1
n i i
IA
| CINCINNATI
1
1
1
I
1 (72)
J
I
'1
l
1
, 1 , , .
3 30 60 90 I2O ISO O 30 60 90 120 ISO 0 3O 60 90 120 ISO 0 30 60 90 120 1!
In
]
1
]
" '
1 (05)
1
]
1
1
1
1
i ill
:NVER
1 UIAUI
.1
1
T
1 (go)
1
1
1
1
1
1
i 1 , , ,
1 AKLAHOUA Cn
1
, . 1
,1
1 (63)
l
.1
J
1
1
1
i 1 ,
Y
. J PHIl AOF
J
1
1
l (go)
1
1
I
1
I
1
, 1,
JO 60 SO 120 10 0 30 60 90 120 ISO 0 30 $0 90 120 ISO 0 30 60 90 120 19
1 ST. LOUIS
1
1
1
1 (79)
1
1
i
J
' 1 1 1
JO tO M IZO li
« <
! 1 SAN FRANCISCO
1 (50)
1
1
l
-"
-1
i »o to 90 iao ii
k) i
1 SEATTLE
1 (47)
1
i
\ ao co to 120 u
to <
1 WASHINGTON D
I
1
-1
1 (61)
1
1
1
. .1
> 10 «0 tO 120 1
>0 (
.PHIA
O
c.
10
1 r^ft/FL
1
1
J
1 (102)
i
i
J
1
i
l
i
> 30 60 90 120 H
I pflQviorut^
i
J
' (48)
1
1
1
1
I
, 1 , , ,
3 JO CO 9O 120 19
NUMBER OF
CITY SITES USED
BALTIMORE 12
BIRMINGHAM 13
CHATTANOOGA 5
CINCINNATI U
CLEVELAND 16
DENVER IS
MIAMI 13
OKLAHOMA CITY 14
PHILADELPHIA 10
PROVIDENCE 16
ST. LOUIS 17
SAN FRANCISCO 17
SEATTLE 14
WASHIN6TON D.C.
Figure F-15. 1974 monthly geometric mean TSP concentrations in 14-study areas
annual means are given in parentheses
-------
JAN
FES
MAR
APR
MAY
JUL
AUO
SEP
OCT
NOV
DEC
JAN
FE5
APR
MAY
JUS
0 JUL
O AUG
SEP
OCT
MOV
DEC
JAN
FES
MAP
APR
MAY
JUN
JUL
A'^3
SEP
NOV
DEC
1 BALTIMORE
Z]
f
1
1 ( 37.8)
1
13
i
ZZI
j. 1
1
3 2 4 6 8 1C
[ZJ DENVER
Z!
77D
]
) ( 14.0)
1
^
.. J
i.
D
1 I
r "j
i
: 1
n
02 4681
I
! 1 ST. LOUIS
~|
! (
I 1 (36.8)
i ..I
I J
| j
zz^
1 1
1 1
1
1
) C
3
1
1
) BIRMINGHAM
1
1 (56.0)
")
1
1
1
ZZI
1
1
> 2 4 e si
1 MIAMI
"1
1
'1 (49.0)
^ 1
1
f !
1
i
1
1
32 46 81
..I SAN FRANCISCO
' 1
~I
| (15.6)
]
]
Z!
J
1
3 (
3
1
j
1 CHATTANOOGA
1
1 (51.3)
j
1
" 1
1
1
> 2 4 6 8 l(
OKLAHOMA CITY
1
1
1
I (59.4)
J
D
i
i
i
j
]
02 4681
1
J
-] (37.9)
1
ZU
)
1
1
) (
0 (
| CINCINNATI
1
1
1
J (46.8)
1
1
b- _, '
Z3
J
1
) 2 4 6. 8 !<
) PHILADELPHIA
~-~ 1
J
]
1 (37.8)
J
^
i
\
j
j
T
) 2 4 6 8 1
I WASHINGTON .D.C.
]
1
1 (36.0)
_ 1
1
1
ZZZ3
1
l
> c
3
..... .,1 CLEVELAND
)
1
!
1 (39.91
1
1
j'
ZD
5
'1
) 2 4 6 8 1C
1 PROVIDENCE
1
!
1
1 <40.T>
1
1
J
1
"|
1
1
32 4 6 6 10
02 468 10 02 498 10 02 468100246610
Figure F-16. 1974 monthly precipitation in 14-study cities (inches).
Annual averages are given in parentheses.
-------
Table F-7. MORNING AND AFTERNOON MEAN MIXING HEIGHTS
City
Baltimore
Birmingham
Chattanooga
Cincinnati
Cleveland
Denver
Miami
Oklahoma City
Philadelphia
Providence
St. Louis
San Francisco
Seattle
Washington, D.C.
Time
of day
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a .m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
a.m.
p.m.
Mixing height, meters
Winter
750
1,050
500
1,050
510
1,100
550
900
720
900
219
1,482
707
1,221
342
859
900
1,000
870
900
430
800
450
700
824
718
672
1,054
Spring
680
1,850
470
1,700
520
1,870
600
1,730
550
1,580
423
3,070
980
1,459
457
1,506
800
1,600
780
1,200
490
1,550
700
1,200
838
1,577
585
1,890
Summer
500
1,900
420
1,850
420
1,900
380
1,700
380
1,550
255
3,458
1,071
1,383
367
1,862
650
1,700
490
1,200
330
1,650
460
800
576
1,419
421
1,924
Autumn
590
1,350
350
1,450
350
1,500
400
1,350
540
1,200
174
2,161
933
1,341
343
1,302
750
1,200
700
950
370
1,300
450
950
585
987
436
1,412
301
-------
average number of days with snow or ice equal to, or more than 1 inch.
Comparison of Figures F-12 and F-13 show roughly similar seasonal
patterns with the principal exception being Cleveland where the number
of days per month with precipitation is greatest during the winter and
early spring, but the average amount of precipitation per month is low
during the winter.
Figure F-14 shows the normal monthly heating and cooling degree days
for the study cities and is of use principally in comparing space
heating requirements among the cities. Rankings of the 14 cities by
degree days and other meteorological parameters are given in
Appendix A.
Figures F-15 and F-16 permit comparison between the monthly changes in
TSP levels in each city and the monthly changes in the amount of pre-
cipitation. The TSP levels represent areawide concentration, with the
number and types of sites included varying from one area to another,
while the precipitation amounts are airport data. An overall impression
gained from these two figures is one of greater regularity in the TSP
profiles. Presumably this is due in part to smoothing resulting from
spatial averaging, but also is likely to be a reflection of a number of
seasonally-integrating factors. It should also be borne in mind that not
only is showery type precipitation frequently spotty, but also that TSP
sampling may not be carried out on days directly affected by the
precipitation.
Although there appears to be no clear-cut seasonal trends in TSP levels
in many of the 14 cities, three types of seasonal patterns can be noticed.
The first type has a strong wintertime maximum and is represented solely
by Denver. The second has a broad summertime maximum as shown by Cleve-
land, and to some extent by Cincinnati. And the third has a late summer
and fall maximum, and is represented by Seattle and San Francisco.
302
-------
The seasonal TSP pattern in Denver follows the same general pattern as
other conservative pollutants, such as CO, and results principally from
seasonal changes in stability and the height of the mixing layer. Inspec-
tion of Table F-7 shows that Denver's average morning mixing heights are
the lowest of any Of the cities and that its afternoon mixing heights are
the highest, reflecting its dry climate and relatively intense solar radi-
ation. Associated with this large diurnal variation in mixing height is
a very pronounced seasonal change. The average of the morning and after-
noon mixing heights shows the greatest seasonal change among the 14
cities, varying from 850 meters in the winter to 1856 meters in the sum-
mer. Also, it is generally agreed that the contribution from vehicular
suspended street dust increases during the winter months due to sanding,
but the magnitude of this contribution is still ill-defined.
The summertime maximum in Cincinnati and Cleveland are associated with
the season of lowest wind speeds (see Figure F-ll), and lowest morning
mixing heights (see Table F-7). In Cleveland, low-level stability is re-
inforced in the summer months by undercutting cool air from Lake Erie.
In Seattle, the 3 months with relatively high average concentrations
(August, September and October) occur at the end of the dry season and
during the season with maximum inversion frequency and lowest morning
mixing heights. In San Francisco, the highest average concentrations
also occur at the end of the dry season as the strong summertime sea
breeze regime comes to an end.
It is also interesting to note months with unusually high concentrations,
shown in Figure F-15, compared to adjacent months and to check relative
precipitation amounts during these periods. In general, months with
high relative concentration had relatively little precipitation, but
not always. Months showing the expected correspondence are: July in
Baltimore; October in Birmingham and Chattanooga; July in Cincinnati;
and July in Oklahoma City. On the other hand, January in Chattanooga
had substantially, higher concentrations than February, but also had more
precipitation.
303
-------
REGIONAL METEOROLOGY AND GEOGRAPHICAL FACTORS
The purpose of this discussion is to provide a brief overview of several
large scale features of the meteorology and geography of the country which
play an important role in determining TSP levels. These features include
the general circulation pattern; the major storm tracks and resulting dis-
tribution of precipitation; wind speed with its multiple influences
through dilution, transport, and contribution to fugitive dust emissions;
frequency of low level inversions and stagnating air masses, with their
associated poor dispersion conditions; major topographical features,
which serve as barriers to air transport; and evaporation rate and an
index used in estimating the potential for fugitive dust emissions.
Mean Flow Patterns
Except for southern Florida and the Gulf Coast, the contiguous United
States lies in the belt of the prevailing westerlies. In very general
terms, therefore, unpolluted air enters the continent from the west
and accumulates particulates while traversing to the east coast. In
actuality, however, this flow is impeded by mountain ranges, has super-
imposed upon it day-to-day weather patterns, and undergoes significant
seasonal change. A more realistic view of average flow is given by
Figures F-17 and F-18 which show the mean resulting surface winds at
principal reporting stations during January and July. Large open
arrows have been added to these maps to depict average flow conditions
during these two months of maximum seasonal difference. During both
seasons air enters the west coast from the Pacific, and flow through
the mountain states, though rather ill-defined due to topographical
influences, shows an average movement from the south and west. Very
marked seasonal differences occur east of the Rockies, however. During
January, major inflow from Canada dominates the pattern over the north
central and northeastern states. In the far south, flow from the Gulf
moves north over Texas, and air from the Atlantic crosses southern
304
-------
'->.._MEAN RES'ULXANT SURFACE WIND" DIRECTION AND SPEED"
' /... MIDWINTER MONTH JANUARY
^1 T --iT"f
s I 0 .
,.JJ. Xi.
J / too^ ' ^*"r / v*^>
k^'^
NOTE'
RESULTANT WIND IS-THE
VECTORiAL AVERAGE OF ALL
WIND DIRECTIONS AND SPEED
DURING THE MONTH.
9 ?
SCALE. mph
Figure 7-17. Mean resultant surface winds for January. (Resultant wind map
taken from Climatological Atlas of the United States )
ffiAN RESyLTANT SURFACE WIND DIECTION
MIDSUMMER MONTH\-
NOTE =
RESULTANT WIND -IS THE
VECTOR IAL AVERAGE OF ALL N
WIND DIRECTIONS AND SPEED
DURING THE MONTH.
05 10 15 20
ill i )
SCALE,mph
Figure F-18.
Mean resultant surface winds for July. (Resultant wind map
taken from Climatological Atlas of the United States7)
305
-------
Florida. Northerly and southerly winds occur with nearly equal fre-
quency at Oklahoma City, with a net resultant flow from the west.
In July, in the mean, all of the states east of the Rockies are do-
minated by southerly flow which is particularly steady and strong
over the central states and veers to the southwest over the north-
eastern states. This predominant south to southwest current provides
the vehicle for long-range transport of particulates from many of
the industrialized regions of the country. Incursions of Canadian
air are less frequent and penetrate less deeply into the United
States than in winter.
Storm Tracks and Precipitation Patterns
Some of the most frequent routes of cyclonic weather systems are
sketched in Figure F-19. These paths have been drawn to give a gen-
eral picture of the movement of low pressure centers and do not refer
to the exact movement of individual storms. This figure shows that,
regardless of their origin, many storms pass over the eastern states.
In the development of storms moving from the west, moisture is sup-
plied principally from the Gulf of Mexico or Atlantic Ocean, and the
combination of this moisture supply and preferred storm paths lead to
high annual precipitation amounts in the southern and eastern states,
as shown in Figure F-20. The second area with high annual amounts of
precipitation lies along the coastal region of Washington and Oregon
where storm activity from the Pacific is frequent, particularly during
the winter and early spring. Figure F-20 was prepared from mean pre-
cipitation amounts given for state climatic divisions, and precipi-
tation amounts at specific locations may differ substantially from
those indicated by the figure. The drying effect of the western
mountain ranges on air masses moving in from the Pacific is reflected
in the extensive areas in the west with less than 20 inches of annual
precipitation.
306
-------
Figure F-19. Frequent routes of cyclone centers
\ \NIO- 19'
,_JJ
Figure F-20.
Mean annual precipitaition in inches. (Prepared
from data in Climatic Atlas of the United States7)
307
-------
As reviewed in earlier discussions, the effectiveness of precipitation
in cleansing the atmosphere, suppressing fugitive emissions, or en-
couraging plant growth depends not only upon the total amount of pre-
cipitation that falls during a year but also upon its distribution.
The most readily available climatological statistic bearing upon this
distribution is the number of days in a year with measurable precipi-
tation (i.e., _> 0.01 in.). Figure F-21 gives the geographic distri-
bution of this information. It can be seen from Figures F-20 and F-21
that the number of.days with precipitation and the annual precipitation
have similar geographic patterns; the principal difference is in the
eastern part of the country where total precipitation amounts are
greatest in the south and the number of days with precipitation is
greatest in the north.
Wind Speed
Wind speed is commonly observed at well exposed airport locations.
While these observations are generally acceptable for estimates of
dilution rates above neighboring urban areas, they are less likely to
be appropriate for use near ground or street level in estimating fu-
gitive dust emission rates within built-up areas. Relative comparisons
for various sections of the country are of value, however, and can be
made by examining the isopleths of wind speed averaged through the
Q
mixing layer which have been prepared by Holzworth. Figure F-22
shows the pattern of annual wind speed averaged through the morning
mixing layer, the main features of which are characteristic of other
seasons and times of day. The highest average wind speeds occur over
Oklahoma; the lowest occur inland in California and Oregon, and in a
band extending northward from the Mexican border to Wyoming. Moderate-
ly low wind speeds also occur extensively in the southeast. Maps
q
prepared by Hosier shows that the percent frequency of nighttime wind
speed equal to or less than 7 miles per hour has a somewhat similar
pattern; the fewest number of light nighttime winds are observed in
308
-------
Figure F-21.
Mean number of days with 0.01 inch, or more of precipita-
tion in 1 year. (Prepared from data in Climatic Atlas of
the United States')
Figure F-22.
Isopleths (m sec" ) of mean annual wind speed averaged
through the morning mixing layer (from Holzworth^)
309
-------
Oklahoma and Kansas and the greatest in a belt extending from lower
California and Arizona northward through Nevada to southern Oregon.
The second principal region with a larger number of light nighttime
winds (greater than 70 percent) covers much of the southeastern
part of the country.
Inversions and Stagnating Air Masses
Maps showing the frequency of low-level inversions (defined as an
inversion with base within 150 meters of the surface) throughout the
contiguous United States by season and year have been published by
9
Hosier. These maps show isopleths of inversion frequency in percent
of total hours and thus make general comparisons from one region of
the country to another possible. Figure F-23, taken from Hosier,
gives annual isopleths and shows two centers of maximum frequency,
one centered over western North Carolina and the other, larger and
more elongated, centered in the west. Low frequencies occur along
coastlines, particularly along the Washington-Oregon coast and much
of the eastern seaboard. The lowest frequency of all is at the
southern tip of Florida.
For detailed statistics on the distribution and frequency of stagnating
anticyclones east of the Rockies, reference should be made to Korshover10
who has summarized data on all cases of stagnation lasting for 4 or more
days which occurred during a 30-year period. This study showed that the
number of cases ranged from four to 16 per year, and that the maximum
frequency in both number of cases and number of days stagnation occurred
in an area covering parts of Georgia, South Carolina, and North Carolina.
Seasonally, the frequency of occurrence peaks in October and has a sec-
ondary maximum in May. Low frequencies occur during the winter and a
secondary minimum occurs in July. Figure F-24, sketched from data in a
figure presented by Korshover, shows the geographic distribution of the
average annual number of stagnation days for the 1936 to 1965 period.
310
-------
40
Figure F-23,
Annual inversion frequency in percent total hours
(from Hosier,9 p. 322)
Figure F-24.
Average annual number of stagnation days, 1936 to
1965 (see Korshover10)
311
-------
In a study of air pollution potential for the western United States,
Holzworth carried out a somewhat similar investigation of the
occurrence of stagnating anticyclones, although the criteria and study
techniques had to be different because of the iregularity of the
terrain. In this investigation an occurrence was counted if it per-
sisted for 2 days or longer. Holzworth found that these conditions
were extremely rare except during winter and late fall. The average
annual number of occurrences over the total study area during an
8-year period was 20, and the area having the highest frequency
(about one event per year) was centered over Idaho and Wyoming.
Evaporation Rates and a Meteorological Index of Soil Moisture
Fugitive dust emissions are a function of the dryness of the material
and, for wind-induced emissions, the speed of the wind. Because of
the many complicated relationships involved, no satisfactory index
has yet been developed by which the potential for fugitive dust emis-
sions across the country can be well defined. Current practice in the
calculation of emission factors for fugitive dust emission sources has
incorporated the use of Thornthwaite's precipitation effectiveness
(PE) index as an indicator of soil moisture. In concept, precipitation
effectiveness was to have been calculated by dividing the monthly pre-
cipitation by the monthly evaporation to find the P/E ratio. Because
of the general lack of appropriate data on evaporation, Thornthwaite
developed a formula based on precipitation and temperature for the
calculation of his PE index. Figure F-25, based on a more detailed
map prepared for EPA by Midwest Research Institute12 from annual pre-
cipitation and temperature data, illustrates the geographic distri-
bution of this index, referred to as the precipitation-evaporation
index in this report.
312
-------
Figure F-25. Precipitation-Evaporation (PE) values (after MRI Report12)
Mean annual class A pan evaporation in inches (adapted from
Climatic Atlas of the United States)
313
-------
One additional observation of some relevance to soil moisture is ob-
tained directly by exposing a pan of water and measuring the rate of
evaporation. Figure F-26 shows the geographic distribution of the
annual evaporation measured in this way.
Heating Degree Days
Figure F-27 is a map of the average annual heating degree days over the
contiguous United States, prepared for purposes of overview only. In
the western part of the country large differences in temperature (degree
days) occur due to differences in elevation. No attempt has been made
to show these details in Figure F-27, and reference should be made to
the climatological records of the city in question for quantitative data.
LOCAL METEOROLOGY
The previous discussion has reviewed the principal meteorological factors
which affect the concentrations of typical pollutants within an urban
area, plus additional factors which contribute to the generation of par-
ticulate emissions from fugitive dust sources; maps have been presented
showing how these factors vary from one region of the country to another.
While it is true that each of these factors may have a significant impact
in determining the TSP level which results from the particulate burden
introduced into an urban atmosphere, the extent of that impact is con-
trolled by local conditions and cannot be estimated without an under-
standing of the local meteorology. In many instances the favorable or
unfavorable aspects of the city's topography, because of their influence
on meteorology, determine whether or not air quality standards will be
achieved.
In particular, topographical barriers restrict horizontal dispersion,
and deep valleys and basins serve to collect nocturnally cooled surface
air from higher elevations. Under light gradient wind conditions and
314
-------
u>
I-1
Ul
LEGEND
(rvl > 10,000
=| 8,000-10,000
(Tfl 6,000-8,000
4,000-6,000
2,000-4,000
<2,000
Figure F-27. Average annual heating degree days, base 65 F (sketched
from more detailed map in Climatic Atlas of the United States)
-------
radiational cooling, a strong inversion develops and the resulting pool
of stagnant air continues to collect pollutants throughout the night and
early morning. Under unfavorable conditions, the inversion may not lift
sufficiently during the following day to permit the flushing of the area
with fresh air. A modification of this effect is found in Denver, where
the Platte River valley which passes through the city is quite shallow,
but heating and cooling of the slopes of the Rocky Mountain foothills to
the west of the city result, under light wind conditions, in a diurnal
reversal in air flow. Under this meteorological regime, part of an al-
ready polluted air mass may be resubjected to Denver's emission during
its return trip. In other areas, such as western Pennsylvania, narrow
valleys serve to channel air flow so that polluted air repeatedly follows
the same path.
A different, generally favorable, local meteorological effect results
from the close proximity of a city to a large body of water. In
addition to providing an area with minimal emissions, the smooth water
surface offers little frictional resistance and consequently contributes
to higher wind speeds. The development of land and sea breezes, which
result from temperature contrasts between land and water surfaces, also
contributes to increased ventilation under initially light wind con-
ditions, but may also aggravate local problems when the inland drift of
cool air from the water undercuts warmer onshore air to form a shallow
stable layer, as is the case in Cleveland.
Another noteworthy example of the importance of local terrain and meteor-
ology is Los Angeles, where level terrain and the proximity of the sea
coast are favorable for dispersion, but where the presence of a semi-
permanent subsidence inversion serves as a. lid and the surrounding
hills and mountains slow the flow of air in and out of the Los Angeles
Basin.
316
-------
The preceding examples are given simply to indicate some of the factors
controlling local meteorology. Understanding the details of how these
factors impact on TSP levels in a particular area requires an intimate
knowledge of the emissions and physical nature of the area in question.
SUMMARY AND CONCLUSIONS
We have found no simple meteorological index during these exploratory
studies that adequately describes the relative potential for TSP pollution
throughout the United States. This is not surprising in view of the com-
plex interrelationships involved. Nonetheless, since the way various
meteorological parameters affect TSP levels is known qualitatively, pre-
liminary estimates can be made as to which areas can probably achieve the
national standards because of good fortune in their location and which
areas will undoubtedly require more extreme control measures if the stan-
dards are to be achieved. The use of maps showing the geographical dis-
tribution of relevant meteorological parameters, such as those presented
in this section, are helpful in gaining general perspective, but final
judgment must be based on an intimate knowledge of local meteorology and
topography, plus an understanding of the contribution of each emission
category to existing TSP levels.
Special studies carried out during the program have provided estimates
of the effectiveness of precipitation in reducing TSP concentrations,
and indirectly provide an upper bound to the effectiveness of sprinkling or
watering programs for widespread TSP control. These studies also pro-
vide general guidance for estimating the annual variation in TSP levels to
be expected as a result of year-to-year changes in precipitation and
degree days.
Some of the principal conclusions drawn during this study about meteor-
ological influences on TSP levels are summarized by topic in the following
paragraphs.
317
-------
Transport and Background Levels
The concentration of particulates in air entering an urban area is the
base line value upon which urban emissions build. It varies according
to the history of the air mass, and is lowest when the air has had a
long water trajectory and highest when it has had a recent trajectory
over major industrial areas. Concentrations in coastal regions are
affected by the relative frequencies of onshore and offshore winds.
Thus levels are low along the west coast and along the southeast coast
of Florida where the prevailing winds are onshore, but higher along the
eastern seaboard farther north where prevailing winds are from the south-
west and west during much of the year. Concentrations are also low in
remote, undeveloped areas, -and are generally low in rural areas except
during periods of soil preparation under dry conditions. An under-
standing of land use and mean atmospheric circulation patterns, roughly
indicated by Figures F-17 and F-18, are useful in obtaining a perspec-
tive of regional nonurban levels. Fine particulates, such as secondary
particulates formed during atmospheric transport, may remain suspended
for several days and thus contribute to TSP levels at distances remote
from their primary source.
Effects of Precipitation and Heating Demand
Precipitation is very effective in reducing TSP levels in areas with
high concentrations which have resulted from either industrial or fugi-
tive dust sources, and average concentrations decrease steadily with
increasing 48-hour precipitation amounts in these areas. The effect of
precipitation is greatest on the day it occurs and lasts on the average
about 2 days. In our case studies of high-concentration areas, concen-
trations measured during the last half of a 48-hour period with precipi-
tation > 0.25 in. averaged approximately half of concentrations mea-
sured during 48-hour periods with negligible precipitation. Concentrations
318
-------
at typical urban sites (excluding clean residential areas) on days with
measurable precipitation were about 75 to 85 percent of average values
for the site and time of year.
A multiple regression equation, based on year-to-year variations in TSP
levels, annual precipitation, and heating degree days in 10 of the study
cities indicated that an increase of 1 inch in annual precipitation de-
3
creases the mean annual concentration by 0.43 yg/m , and an increase in
heating degree days of 1000 increases the mean annual concentration by
3
2.5 yg/m . The TSP data used in developing the regression equation were
drawn from a mix of site types, and from cities where the annual preci-
pitation ranged from a low of 14.0 inches in Denver to a high of 51.3
inches in Chattanooga, and annual heating degree days ranged from 208
in Miami to 6154 in Cleveland. Although conclusions based on this
equation must be considered tentative, the relationship provides a ready
means for comparing precipitation and heating demand effects throughout
the country, For example, the results of applying the equation to the
climatological precipitation pattern (Figure F-20) can be displayed as
the relative effect of differences in total annual precipitation on the
annual TSP level, as has been done in Figure F-28. In this figure a
precipitation rate of about 35 inches a year corresponds to the "0"
relative effect isopleth. The maximum geographical difference in the
3
annual mean shown by the figure is approximately 35 yg/m (from -25 to
+10 yg/m3).
The use of the relationship between heating degree days and TSP level in
conjunction with the geographical distribution of degree days shown in
Figure F-27 suggests a maximum contribution to the annual mean from space
3
heating of 25 yg/m . The spatial variation in degree days found in the
western United States has been drastically smoothed out in Figure F-27,
so attention should really be focused on the area east of the Rockies.
Here the typical contribution ranges from about 5 yg/m in the southern
3
tier of states to 22 yg/m in the most northern states. Again, this is
319
-------
to
N>
O
-HO
Figure F-28. Relative effect of annual precipitation on annual TSP level
in a hypothetical urban area. (Estimated from regression
equation, p. 27
-------
an attempt to generalize the effect of space heating using data from a
mix of cities with widely different fuel usage characteristics, and the
results therefore are not necessarily appropriate for any specific
city.
Effects of Stagnation Periods
We have made no special attempt during these studies to isolate the
effects of inversions or stagnation periods on TSP levels. However,
several general statements can be made. First, the highest 24-hour
concentrations are observed during stagnant conditions since by defini-
tion these are periods with very low wind speeds and inversion condi-
tions over at least a 24-hour period. Under these conditions, concen-
trations typically are about two or two-and-a-half times the average
values for the site and time of year. It is under these conditions
that the 24-hour standards are most likely to be exceeded. It is not
3
unusual for this increase to be 100 yg/m or more in the more polluted
areas. With this in mind one can use the average annual number of
stagnation days from Figure F-24 to approximate the effect of stag-
nation days on the annual mean. Thus for the area with the maximum
number of stagnations as defined and determined by Korshover, stagnations
occur on an average of 3.8 percent of the days in a year. If on these days
the average concentration is 2.3 times the annual mean, the increase in the
annual mean due to the stagnation days is approximately 5 percent.
Effects of Wind Speed
The following conclusions are based principally on 24-hour average TSP
concentrations and daily average airport wind speeds in four cities.
One of these (Birmingham) is heavily industrialized, two (Chattanooga
and Denver) are moderately industrialized, and one (Oklahoma City) is
lightly industrialized. Birmingham and Chattanooga have above-average
321
-------
precipitation and low average wind speeds, Denver has little precipi-
tation and below-average wind speeds, and Oklahoma City has nearly
average amounts of precipitation and high average wind speeds.
The results of these studies showed that the dilution effect of wind
speed was noticeable below speeds of about 10 miles per hour in indus-
trial areas where major contributions were made from point sources. At
higher wind speeds and in nonindustrial urban areas, average TSP levels
did not appear to be related to wind speed. We were unable to discover
to what extent this invariance with wind speed resulted from an inter-
play between dilution and wind-induced fugitive dust contributions.
In this connection, it should be pointed out that the collection effi-
ciency of the hi-vol for the larger particles is believed to decrease
with increasing wind speed.
322
-------
REFERENCES
1. Engelmann, R. J. The Calculation of Precipitation Scavenging in
Meteorology and Atomic Energy. U.S. Atomic Energy Commission,
Division Technical Information, 1968.
2. Pasquill, F. Atmospheric Diffusion. Ellis Horwood, Ltd.,
2nd Edition, 1974.
3. Chamberlain, A. C. Aspects of Travel and Deposition of Aerosol
and Vapour Clouds. AERE, HP/R 1261, HMSO, 1953.
4. Stern, A. C., H. C. Wohlers, R. W. Boubel, and W. P. Lowry.
Fundamentals of Air Pollution. Academic Press, New York.
5. Abel, M. D. The Impact of Refloatation on Chicago's Total Sus-
pended Particulate Levels. A Thesis Submitted to the Faculty of
Purdue University in Partial Fulfillment of the Requirements for
the Degree of Master of Science, August 1974.
6. Lazenka, C. and T. Weir. Rain Particulate Study. City of Phila-
delphia Air Management Services. Unpublished Preliminary
Document, January 4, 1974.
7. Climatic Atlas of the United States. U.S. Department of Commerce,
Environmental Science Services Administration, Environmental Data
Services, June 1968.
8. Holzworth, G. C. Mixing Heights, Wind Speeds, and Potential for
Urban Air Pollution Throughout the Contiguous United States. U.S.
Environmental Protection Agency. AP-101, January 1972.
9. Hosier, C. R. Low-Level Inversion Frequency in the Contiguous
United States. Monthly Weather Review. Vol 89, September 1961.
10. Korshover, J. Climatology of Stagnating Anticyclones East of the
Rocky Mountains, 1936-1965. U.S. Department of Health, Education,
and Welfare, National Center for Air Pollution Control, Cincinnati,
Ohio, P.H.S. Publication No. 999-AP-34, 1967.
11. Holzworth, G. C. A Study of Air Pollution Potential for the
Western United States. J Appl Meteorol. 1:366, September 1962.
12. Cowherd, C., Jr., K. Axetell, Jr., C. M. Guenther, and G. A. Jutze.
Development of Emission Factors for Fugitive Dust Sources. Pre-
pared for the U.S. Environmental Protection Agency, Office of Air
and Waste Management, Office of Air Quality Planning and Stand-
ards. Publication No. EPA-450/3-74-037, June 1974.
323
-------
APPENDIX G
PRELIMINARY ASSESSMENT OF SOURCE CONTRIBUTIONS
ASSESSING SOURCE CONTRIBUTIONS
The development of an appropriate control strategy for improving air
quality in a given area involves developing a quantitative knowledge of
the contributions of various sources and source categories to the TSP
levels in the particular area. Various approaches to assessing these
contributions are available; they range from data analysis efforts to
extensive particle analytical studies and special field monitoring.
Some of the study approaches that are typically applied include:
Air Quality Data Analysis Day-of-the-week patterns,
seasonal patterns, and time trends and similar paper
studies can often be interpreted to provide some
information. Wind direction analyses are particularly
good in cases of single, directionally clear sources.
These simple analyses can permit a good quantitative
assessment of nearby source impact if the meteorological
circumstances are appropriate.
Modeling and Other Emission Data Analysis Emissions
data analysis may also be helpful, depending on the
data available; in appropriate situations, primarily
those dominated by traditional source emissions from
well-defined stacks, meteorological dispersion model-
ing may be particularly useful. However, the lack
of reliable inventories of emissions from nontraditional
sources and the lack of adequate information on in-
coming concentrations make most TSP modeling of ques-
tionable usefulness.
Special Analytical Investigations The potential for
identifying particles by various analytical techniques,
such as microscopy, wet chemistry, or more elaborate
elemental analysis, is great. However, the extrapolation
324
-------
from particle identification to source identification
can be intricate and generally requires significant
knowledge about the nature of candidate sources. Con-
sequently, specific, well-planned efforts based on
some preliminary knowledge are more apt to be produc-
tive than general screening studies.
Special Field Studies Some short-term field monitor-
ing efforts may more reasonably be the first special
studies attempted. The simplest is merely to increase
the frequency of hi-vol sampling, to provide more ex-
tensive data, or to use additional sites, specifically
placed to provide meaningful comparisons or contrasts.
More elaborate efforts would involve sampling with
special equipment to provide data with a shorter time
resolution, directional sampling, particle size anal-
ysis, or sampling media more adaptable to chemical
elemental or morphological analysis.
PRELIMINARY ASSESSMENT METHOD
In designing a study effort aimed at assessing source contributions, it
is helpful to initially develop a preliminary hypothesis concerning the
source contributions, to be iteratively modified as further information
is developed by whatever study approach is taken. The following is a
suggested approach to forming such a preliminary judgment using the
results of this study. It is based on systematic comparisons of the
TSP levels in the area of concern, suitably adjusted, with the typical
values found in similar neighborhoods in the 14 study cities. Following
generally the contributions to typical urban TSP as discussed herein,
the suggested approach considers in turn the following additive portions
of the particulate problem:
Nonurban particulate levels
Urban secondary particulate
Areawide urban activity component
Industrial or commercial neighborhood influence
Nearby source influence
325
-------
Determination of Nonurban Particulate Levels
The first step in identifying the source of urban particulate levels is
to establish just how much of the measured levels in fact originate
within the urban areas of the AQCR, as opposed to being transported in
with incoming air masses. Generally, the careful definition of the
urban particulate contribution requires a much more precise definition
of the nonurban particulate level than is commonly available or under-
taken. Preferably, data on nonurban levels should be available from
hi-vol sites established at enough appropriate locations to provide a
meaningful average annual value, even weighted by wind prevalence if
such sophisitcation is warranted. Unfortunately, such data is rarely
available, and it is necessary to use data from existing NA.SN sites,
such as is presented in Figure 4 in Section III. When such data from
existing nonurban sites is used, the nature of the site must first
be investigated. Many nonurban stations are influenced by nearby low-
level fugitive dust sources, as are some urban sites. The purpose of
measuring TSP concentrations in nonurban areas, however, is to measure
the concentration in the incoming air mass; consequently, such monitor-
ing is one situation in which elevated hi-vol monitoring sites are
appropriate.
Determination of Urban Secondary Particulates
The chemical components of the TSP must also be considered in determining
the constituents added by the urban area, particularly the major secondary
constituents such as sulfates and nitrates. These are known to be both
elevated over very large geographical regions, and then elevated addi-
tionally within an urban area. The analytical determination of such
chemical species in urban samples, such as might be undertaken for
strategy development, will require a similar determination concerning the
incoming air in order to be most useful. The difference between urban
secondary levels and carefully determined nonurban secondary levels can
be interpreted as the portion of the urban TSP that will not be susceptible
to control techniques directed at particulates per se.
326
-------
Separation of Urban Activity and Neighborhood Components
The study found that a certain portion of particulate in any urban area
is dependent upon the areawide influences from nontraditional sources
such as construction, automotive emissions, re-entrained dust, unpaved
areas and general urban activity. This component must be separated from
neighborhood influences by comparing air quality levels at sites in
different neighborhoods. Consequently, it is necessary to classify
monitoring sites so that the typical influences associated with each of
these sites can be accounted for properly.
Classifying Sites The neighborhoods in which the existing monitoring
sites are located should be carefully studied, and then classified as
residential, commercial or industrial to permit comparisons with the
study city data. Although most sites are already classified, at least
in the SAROAD site description forms, the sampling of sites provided
by the cities in this study suggests that such classifications are not
commonly done with much precision and should be reviewed.
A variety of definitions of neighborhoods were considered, and the follow-
ing general distinctions proved to be the most useful categorization for
TSP analysis:
Residential Those sites that are in completely residential
areas, except for a single fire station, school, or other
similar building on which the monitor is placed.
Commercial The following varieties of neighborhoods are
all considered commercial:
~ Light commercial retail activity, including dense
apartment residential, with hotels and motels and
other urban activity mixed in.
-" ' Strip commercial major traffic artery lined on both
sides with commercial activity. Land use may be
totally residential behind the commercial establishments.
327
-------
Heavy commercial this includes both the central busi-
ness district and light industrial establishments with
space heating emissions.
Industrial includes only those sites which are in very close
proximity (several blocks) to, or actually within, an area of
heavy industrial land use.
Classification of Urban Area The second step of problem identification
is classifying the urban area under study with respect to the degree of
heavy industrialization-, in order to indicate in general the likely
dominance of emissions from traditional sources on areawide levels.
Since reliable emission inventory data on nontraditional sources do not
exist, it is difficult to develop a good overall perspective on the rel-
ative balance between traditional and nontraditional sources. Consequently,
classification must generally be based only on data concerning traditional
sources. The overview of the 14 case study cities does provide some gen-
eral guidance that is applicable to this judgment. Table G-l lists the
total emissions attributed to traditional sources in the 14 case study
cities, not only as total emissions but also on per capita, manufacturing
employment, and emission density bases. These figures are not perfectly
comparable because of differences in the emission inventories and the
differences in the geographic distribution of sources from area to area,
but they have been based on comparable areas as best possible.
The primary reason for classifying cities is to direct the general nature
of the strategy development effort by obtaining a rough estimate of the
traditional source contribution. This determination involves the degree
to which an a-rea should approach the development of control strategies for
nontraditional sources at present. In those heavily industrialized cities
where traditional source emissions still dominate (generally cities with
emission densities over 200 to 250 tons per square mile per year), it does
not appear possible to use air quality data to quantitatively study or
design control strategies for nontraditional sources. It is also likely
that any such strategies proposed would be very difficult to support in
public. It is therefore recommended that appropriate action in these
328
-------
Table G-l. TRADITIONAL SOURCE EMISSIONS IN CENTRAL COUNTIES
OF 14 STUDY AREAS
u>
N5
vD
Cities
Heavily
industrialized
Uncontrolled
Cleveland
Birmingham
St. Louis
Heavily
industrialized
Controlled
Cincinnati
Philadelphia
Baltimore3
Moderately
industrialized
Chattanooga
Denver
Seattle
Providence
Lightly
industrialized
Washington
Oklahoma City
Miami
San Francisco
Tons per year
210,000
110,000
190,000
56,100
31,600
23,700
10,300
9,700
7,300
7,800
5,600
2,600
8,000
4,800
Tons per year
per capita
x 1,000
122
171
164
60.7
16.2
15.5
40.4
18.8
6.3
13.4
7.4
4.9
6.3
6.7
Tons per year
per square mile
335
488
411
135
245
240
88
60
37
30
92
2
30
60
Percent
manufacturing
employment
36.1
29.3
35.1
37.7
34.1
29.3
52.6
20.8
26.7
44.2
7.2
19.0
10.5
24.4
Includes both Baltimore City and Baltimore County
An area other than central county boundaries was used to calculate emission
density in some cases (see Section III).
-------
cases is to concentrate on the control of traditional sources and to
gather background information to support nontraditional source control
when and if it becomes necessary.
At the other end of the spectrum are those lightly industrialized cities
where traditional sources are totally absent or completely controlled
(generally those with less than 50 tons per square mile per year) . These
areas must necessarily plan for or commence nontraditional source control
if citywide TSP levels are to be further lowered. It is recommended that
such cities minimize the portion of their resources expended on traditional
source control, completing the control of existing sources, maintaining
compliance, and assuring control of new sources.
Estimation of Undue Influences from Nearby Sources
In order to consider sites as being representative of overall conditions
in the particular neighborhoods they represent, the impact of any nearby
sources must be accounted for. Typical neighborhood levels may then be
compared with the results of the present study. Information and descrip-
tions accumulated during the site visits suggested at first glance that
about half of those sites had an effect from a nearby source that appeared
to be an undue impact. Upon subsequent comparative analysis of all these
sites, however, it became apparent that it was necessary to redefine the
subjective definition of "undue local impact" with which we were operating.
The aggregate result is that fewer of the sites (about one-tenth) are seen
as having "undue" local influences, and more of the observed local in-
fluences are incorporated into the neighborhood definitions.
Local impacts from nearby sources must be considered differently in dif-
ferent land use neighborhoods. The following guidelines are recommended
for identifying nearby impacts:
Residential sites: Any noticeable particulate source within about
100 meters, including any small fugitive dust sources within 50
meters, is considered a nearby source.
330
-------
Commercial sites: Traffic is considered an undue effect only if
the traffic source is a major expressway or if the site is very
near the street (less than 6 feet high au-i within 10 feet of a
street carrying more than approximately 5,000 cars/day). Building
and highway construction are considered local sources if significant
projects are located within 300 to 400 meters, or if even very small
projects are located within 50 meters.
Industrial sites; Nearby sources at industrial sites, especially
low-level fugitive dust and reentrainment sources, were so preva-
lent that no sites in industrial areas are considered to have un-
due influences, but rather that such influences are routine in
industrial neighborhoods.
In planning, the potential range of impact of these nearby sources should
be considered so that the areawide influence at the site can be better
estimated. Developing carefully defined relationships for making these
adjustments was beyond the scope of the present study, but the overall
perspective on a large number of sites does permit some rough estimations.
Table G-2 lists a variety of situations with undue impact from nearby
sources of one type or another, with estimates of the potential range of
influence on the annual geometric mean at nearby monitors. Note that
the traffic effect is identified only when major expressways are a fac-
tor or when the site is very near the street. Other sites, less extremely
affected by traffic, are judged not to have a truly undue effect from the
streets. The relationship between TSP levels and traffic volumes per-
mitted the development of a rough quantitative relationship (see Appendix E)
which could be used to adjust typical commercial sites to a common
denominator.
The estimated effects in Table G-2 should be used with considerable
judgment and care; they are based on too small a data base to be authori-
tatively used as firm corrections or adjustments. The best way to use
them is iteratively with the development of average neighborhood levels
across the city (next step). A site with a suspected undue nearby source
influence should first be compared with other sites in similar neighbor-
hoods (e.g., all residential) to establish the excess of the site in ques-
tion over the typical level at such sites. This comparison can only be
331
-------
done roughly, but if the undue influence is truly significant, it should
be apparent. If the excess of the site is roughly consistent with
Table G-2, it could be used as supportive evidence for the nature of the
influence on that site and a rough approximation of what the air quality
would be without that influence.
Table G-2. UNDUE EFFECTS OF NEARBY SOURCES
Effect
on annual mean,a
Nature of source
25-30
15-20
15-20
25
20-25
30-35
35-40
25
20-25
30-40
Residential neighborhood small apartment house con-
struction on adjacent property (about 40 m)
University setting subway construction along right-
of-way 500 to 700 m away
Three-story high CBD site with intermittent express-
way construction within 100 to 200 m
Mixed commercial-light industrial with extensive array
of local fugitive dust reentrainment sources
Elevated site within several-block area of ongoing
building construction
Low-level site on traffic island at intersection of
major arterial streets
Low-level site adjacent to elevated expressway and
access ramp; in slightly dry city
Similar to previous entry in wet city
Extensive unpaved areas and nearby quarry
Low-level CBD site subject to extensive ongoing con-
struction activity
ese concentrations are estimates based upon averages of levels found
at the monitoring sites visited in the 14 study cities; the actual effect
at any one monitor may vary widely outside the ranges given.
332
-------
Comparison of Typical Neighborhood Values to Study Examples
After assessing the various sites used to characterize a type of neighbor-
hood and considering any undue nearby source effects, the next logical
step is to group the sites being studied into neighborhood types. This
study provides some relatively well-based estimates of typical norms for
these neighborhood types, against which the levels at each site in a city
can be roughly compared. The final step in this analytical scenario is
to make this comparison and interpret the results in the light of knowledge
of local sources. This final section presents the TSP levels suggested
for comparison and possible reasons why another given city might be higher
or lower.
Nummary of Typical Source Contributions The following is a brief descrip-
tion of the methodology used in developing the typical values for each of
the components of the urban air at residential, commercial and industrial
sites.
Residential Sites TSP concentrations at residential sites can be thought
of as consisting of the following components nonurban particulate levels,
urban area sulfates and nitrates, a general urban activity component, and
possibly an industrial influence. The average residential levels in the
3
14 study cities ranged from 41 to 77 /ig/m after sites with undue influ-
ence were removed from the data base. After subtracting nonurban partic-
ulate levels and urban sulfate-nitrate excess (as calculated in Section III)
3
the average residential values ranged from 18 to 48 fig/m . After consider-
ing this range in conjunction with the nature of the cities, this urban
3
activity component was typically 20 /ig/tn in the lightly industrialized
cities with the remainder added by industrialization. The variation
3
(18 to 26 /jg/m ) was roughly correlated with "wet" and "dry" cities re-
spectively, so that precipitation was presumed to account for the range.
3
The industrial influence varied from 0 to 30 fig/m with the degree of in-
3
dustrialization; typically 5 to 15 fig/m in moderately industrialized and
333
-------
3
well controlled heavy industrialized cities and 20 to 30 jzg/m in heavy
industrial areas. While other analyses (Section III) suggest that ex-
tensive use of fuel oil for domestic space heating can reasonably account
3
for up to 5 jig/m , this could not be confirmed from the data base
utilized.
Commercial Sites - A similar analysis of commercial sites yields a com-
3 3
mercial urban activity range of 25 to 40 jig/m (typically about 30 jug/m )
It was not apparent that any factor other than increased congestion and
3
traffic accounted for this increase over residential of around 10 jzg/m .
Industrial contributions at commercial sites were generally 5 to 10
3
pig/m higher than at the residential sites, due to closer proximity to
the industrial areas.
Industrial Sites Industrial sites were subjected to a similar analysis
3
They ranged from 30 to 70 jig/m above residential values. Those at the
lower end of the scale are the cities where hi-vols reflect scattered,
isolated industrial activity. Those at the upper end were cities with
good representation of sites near major industrial (steel mill) valleys.
334 .u.s. UOVtRNMENT PRINTING OFFICE: 1077-740-110/306 REGION NO. 4
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 450/3-76-024
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
National Assessment Of The Urban Particulate Problem;
Volume I - Summary Of National Assessment
5. REPORT DATE
July 1976
6. PERFORMING ORGANIZATION CODE
GCA-TR-76-25-G(l)
7. AUTHOR(S)
Robert M. Bradway, Gordon L. Deane, Rebecca C. Galkiewic
David A. Lynn; Frank A. Record, Project Director
8. PERFORMING ORGANIZATION REPORT NO.
z
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
GCA Technology Division
Burlington Road
Bedford, MA 01730
11. CONTRACT/GRANT NO.
68-02-1376
Task Order 18
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
is. SUPPLEMENTARY NOTES Vol ume I, National Assessment - EPA 450/3-76-024; Volume II,
Particle Characterization - EPA 450/3-76-025; Volumes III-XVI, Urban Area Reports -
EPA 450/3-76-026a thru 026n.
16. ABSTRACT
This report is the summary report of a study of 14 urban areas to determine the rea-
sons for attainment or nonattainment of the National Ambient Air Quality Standards.
This summary report combines the data from the particle characterization and urban
area reports to provide a national assessment of the problem.
The report evaluates three major factors affecting attainment - (1) large scale
effects from transported primary and secondary particulates, (2) traditional sources
such as fuel combustion, industrial process and solid waste and (3) nontraditional
sources such as mobile source exhaust, tire rubber, construction, paved and unpaved
roads, etc. Two factors which modify measured levels, meteorology and monitor siting,
are also evaluated.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Particulate Matter
Total Suspended Particulate
Emission Sources
Control Methods
Air Quality Measurements
Optical Microscopy, Trans-
port, Secondary Particulate
Background, Fuel Combustion,
Process Emissions, Fugitive
Emissions, Fugitive Dust,|
Mobile Sources, Meteorology,
Monitor Siting
18. DISTRIBUTION STATEMENT
Release Unlimited.
Available for a fee, Thru the National
Technical Information Service, 5285 Port
Royal Road. Springfield, VA 22151
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
355 pages
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
335
------- |