EPA-907/9-77-007
August 1977
CONTROL
OF REENTRAINED DUST
FROM PAVED STREETS
U.S. ENVIRONMENTAL PROTECTION AGENCY
REGION VII
Kansas City, Missouri 64108
-------
EPA-907/9-77-007
CONTROL OF REENTRAINED DUST
FROM PAVED STREETS
Prepared by
PEDCo-Environmental, Inc.
2480 Pershing Road
Kansas City, Missouri 64108
Contract No. 68-02-1375
EPA Project Officer: Dewayne Durst
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Region VII
Kansas City, Missouri 64108
August 1977
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations in limited quantities from the
Library Services Office (MD35) , Research Triangle Park, North Carolina
27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
PEDCo-Environmental, Inc. , 2480 Pershing Road, Kansas City, Missouri
64108, in fulfillment of Contract No. 68-02-1375. The contents of this
report are reproduced herein as received from PEDCO-Environmental,
Inc. 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-907/9-77-007
11
-------
ACKNOWLEDGEMENTS
Several air pollution control agencies and public works departments
provided information and data during this study, and their assistance is
gratefully acknowledged. A special mention is due Mr. George Hinkle,
Senior Research Associate, American Public Works Association, who
provided PEDCo-Environmental with results of the recent APWA survey of
street cleaning practices and with consultation on questions concerning
street cleaning.
Street cleaning was done on special schedules in Kansas City and
Cincinnati by their city street departments. Coordination of cleaning
activities with PEDCo's field studies was handled by Mr. Don Ewbank of
the Kansas City Streets Division and Mr. Michael Donnelly of the Cincinnati
Highway Maintenance Division.
The EPA Project Officer was Mr. Dewayne Durst, Chief, Air Support
Branch, EPA Region VII. Messrs. Dallas Safriet, Edward Lillis and
Thompson Pace of the Monitoring and Data Analysis Division, and Mr. David
Dunbar of the Control Programs Development Division, Office of Air Quality
Planning and Standards, EPA, also provided project guidance and advice.
The principal authors of the report are Mr. Kenneth Axetell and
Ms. Joan Zell.
-------
CONTENTS
Page
EXECUTIVE SUMMARY 1
Scope and Purpose 1
Reentrained Dust 1
Material on Street Surfaces 2
Potential Control Measures 3
Field Studies 4
Implementation of Control Measures 6
Costs 8
1. INTRODUCTION 9
Scope of Work 9
Recognition of the Reentrained Dust Problem 10
Characteristics of Reentrained Dust 13
Potential Control Measures 15
2. MATERIAL ON STREET SURFACES 19
3. TESTING THE EFFECTIVENESS OF CONTROL MEASURES 24
Project Field Studies 25
Street Cleaning in Kansas City, Missouri 25
Street Cleaning in Cincinnati 28
Street Cleaning in Residential Areas 33
.Mud Carryout Control at a Construction Site 35
Assessment of Other Studies 38
New York-New Jersey 38
Kansas City, Kansas 41
Charlotte, North Carolina 43
Chicago, Illinois 46
Twin Falls, Idaho 48
Seattle, Washington 51
Untested Measures 54
Reductions in VMT 54
Modifications of Street Sanding Procedures 54
Adverse Environmental Effects from Control Measures 56
Summary of Control Measure Evaluations 58
4. COST DATA 62
Street Cleaning Costs 62
Capital Costs 62
Operating Costs 64
Modified Snow Control Costs 69
Mud Carryout Control Costs 71
-------
Cost-Effectiveness of Individual Measures 73
Flushing 73
Combined Flushing/Sweeping 74
Modification of Street Sanding Procedures 74
Control of Mud Carryout from Construction 75
Sites
Control of Mud Carryout from an Industrial 76
Area
5. OPTIMIZING REENTRAINED DUST CONTROL MEASURES 79
Street Cleaning 79
Current Street Cleaning Practices 79
Optimizing Street Cleaning 81
Mud Carryout Control 85
Current Mud Carryout Controls 85
Optimizing Mud Carryout Controls 85
6. LEGAL AND ADMINISTRATIVE ASPECTS OF REENTRAINED 87
DUST CONTROL
Programs for Prevention of Material Deposition 87
Existing State Regulations 87
Existing County and City Ordinances 90
Enforcement Procedures 94
Model Regulations 97
Intergovernmental Coordination 98
Programs for Improved Street Cleaning 98
The Concept of an Administrative Agreement 98
Development of an Administrative Agreement 101
Implementation 107
Public Acceptability 111
7. PROCEDURE FOR DEVELOPING A CONTROL STRATEGY FOR 114
REENTRAINED DUST
8. RECOMMENDATIONS FOR FURTHER STUDY 119
APPENDICES
A. CHARACTERISTICS OF REENTRAINED DUST
Concentration Gradients 123
Traffic Volume 124
Street Surface Loading 128
Wind Speed and Direction 132
Sampling Height 133
Emission Rates 137
Fallout 140
Particle Size 144
Lead 148
vi
-------
B. MATERIAL ON STREET SURFACES 152
Characteristics of Material on Streets 153
Deposition Processes 157
Removal Processes 162
Accumulation Rates 167
C. PROJECT FIELD STUDIES 171
Street Cleaning in Kansas City, Missouri 172
Study Design 172
Data from Study 174
Street Cleaning in Cincinnati 177
Study Design 177
Data from Study 177
Street Cleaning in Residential Areas 184
Construction Site 185
Study Design 185
Data from Study 187
Short-term Sampling Study 189
Study Design 189
Data from Study 191
D. METHOD FOR MEASURING STREET SURFACE LOADINGS 196
Objectives and Scope 196
Parameters to be Measured and Equipment 197
Design 197
REFERENCES 202
VII
-------
FIGURES
No. Page
2-1 Deposition and Removal Processes 20
3-1 Comparison of Particulate Concentrations on 40
Days with Flushing with Average Concentrations,
New York-New Jersey
3-2 Change in Average Particulate Concentrations 44
in Charlotte Area
3-3 Average Particulate Concentrations by Day of 49
Week in Twin Falls, Idaho
3-4 Change in Particulate Concentrations in the 52
Seattle Area
5-1 Effectiveness of Flushing and Rainfall in 83
Reducing Particulate Concentrations
5-2 Effectiveness of Street Cleaning in Reducing 83
Particulate Concentrations
6-1 Example Administrative Agreement 100
6-2 Example City Council Resolution 103
A-l Particulate Concentration Gradients Near Streets 125
A-2 Traffic Volume Versus Concentration for McGee 126
Street Sites, Kansas City
A-3 Reduction in Concentrations with Increase in 136
Height
A-4 Source Depletion Factors by Stability Class for 142
Ground Level Sources
B-l Street Loadings as a Function of Time 170
C-l Kansas City Street Cleaning Study Area 173
C-2 Daily Particulate Concentrations for McGee 175
Street Sites, Kansas City
C-3 Cincinnati Street Cleaning Study Area 178
C-4 Daily Particulate Concentrations for Hamilton 179
Avenue Site, Cincinnati
viii
-------
No. Page
C-5 Effectiveness of Rainfall in Reducing Particulate 181
Concentrations
C-6 Effectiveness of Flushing in Reducing Particulate 181
Concentrations
C-7 Effectiveness of Street Cleaning in Reducing 183
Particulate Concentrations
C-8 Location of High Volume Samplers at Construction 186
Site
C-9 Sketch of Typical Sampling Setup 190
D-l Machine Used for Collection of Street Loading 200
Samples
-------
TABLES
No. Page
2-1 Deposition Processes 22
2-2 Removal Processes 23
3-1 Comparison of Particulate Concentrations in 26
Kansas City during Three Cleaning Periods
3-2 Comparison of Particulate Concentrations in 30
Cincinnati for Four Cleaning Periods
3-3 Cleaning Efficiency of Street Cleaning Methods 32
3-4 Particulate Concentrations at Westwood and 34
Roeland Park Sites
3-5 Particulate Concentrations Associated with 36
Different Mud Carryout Control Measures
3-6 Summary of Control Measure Evaluations 59
4-1 Summary of Capital Costs for Selected Street 63
Cleaning Equipment
4-2 Summary of Annual Operating Costs for Street 65
Cleaning
4-3 Operating Costs by City Size 66
4-4 Operating Costs by Climate Zone 66
4-5 Operating Costs for Selected Cities 68
4-6 Costs of Mud Carryout Control Measures 72
4-7 Cost-Effectiveness of Control Measures 77
6-1 Summary of State Regulations Applicable to 91
Mud Carryout Control
6-2 Enforcement of Local Ordinances for Control 95
of Mud Carryout
6-3 Implementation of Control Measures 108
A-l Effect of Traffic Volume on Concentrations 127
-------
No. Page
A-2 Correlations Between Street Loadings and 130
Ambient Concentrations
A-3 Stepwise Linear Regression Analysis for 131
Short-Term Studies
A-4 Upwind-Downwind Relationships 134
A-5 Concentration Dependence on Sampler Height 135
A-6 Calculation of Vehicle-Related Emission Rates 139
from Short-Term Sampling Studies
A-7 Emission Rates in Areas with Different Land Uses 141
A-8 Particle Size Distribution of High Volume Samples 146
A-9 Suspected Origin of Particulate Matter in High 147
Volume Samples
B-l Distribution of Surface Material Across a Typical 154
Street
B-2 Summary of Street Loading Sample Data 156
B-3 Deposition Processes 158
B-4 Removal Processes 163
B-5 Accumulation Rate on Hamilton Avenue, Cincinnati 169
C-l Suspended Particulate Concentrations for Short- 192
Term Sampling Studies with Consistent Wind
Directions
C-2 Suspended Particulate Concentrations for Short- 193
Term Sampling Studies with Variable Wind
Directions
XI
-------
EXECUTIVE SUMMARY
SCOPE AND PURPOSE
This report is the first comprehensive review on the
topic of reentrained dust from paved streets. Its primary
purpose is to evaluate control measures for reducing emis-
sions of reentrained dust. Information for the review and
evaluation was obtained by several different methods: a
literature review, collection of unpublished data from
traffic-related air pollution studies, compilation of cost
data, survey of public works officials, and design and im-
plementation of five different field studies to evaluate the
effectiveness of specific reentrained dust control measures.
REENTRAINED DUST
The term "reentrained dust" refers to small particles
that are thrown from the street surface by contact with
vehicle tires or are induced to become airborne by the
vortexes from passing vehicles. In addition, motor vehicles
emit particulate matter directly in engine exhaust gases,
from wear of various parts such as bearings and brake linings,
and from abrasion of tires against the road surface. It
appears from studies of total traffic-related impact that
the reentrained portion is an order of magnitude greater
than the direct emissions accounted for by currently avail-
able emission factors for vehicle exhaust (0.34 g/veh-mi)
and tire wear (0.20 g/veh-mi).
All of the assembled data show agreement on the effect
of traffic-related particulate emissions and, in particular,
reentrained dust—this is one of the most important particulate
-------
source categories in every metropolitan area. In most urban
areas where particulate analyses have been performed recently,
some reduction in reentrained dust emissions has been shown
to be necessary in order to attain the primary ambient air
quality standards.
Particulate emissions from a street were demonstrated
to be directly proportional to the traffic volume along that
street within the ranges of traffic volumes tested. This
assumption has previously been made in basing emission factors
for reentrained dust on vehicle-miles of travel.
Suspended particulate samples taken near streets in the
present field studies had relatively large particle sizes—a
mass median diameter of 15 u and approximately 22 percent
by weight greater than 30 u. Calculated fallout rates
averaged 14 percent at 10 m, 26 percent at 20 m, and 34 per-
cent at 30 m from the street.
Mineral matter was identified microscopically (polarized
light microscope) to constitute an average of 59 percent by
weight of the material in the samples. Combustion products
were a surprisingly large 40 percent of the weight. Bio-
logical matter and tire tread together comprised a little
more than one percent of the particulate collected near
streets. Inorganic lead was determined chemically to account
for an average of one to two percent of the suspended par-
ticulate.
MATERIAL ON STREET SURFACES
Published information was compiled to identify the
important processes by which materials are deposited on
street surfaces and the processes, including reentrainment,
by which they are removed. Typical annualized deposition
and removal rates for each of the processes were estimated
for a four-lane street with a traffic volume of 10,000
vehicles per day. These deposition and removal rates, which
-------
should not be considered descriptive of any single day on a
particular street, are summarized below:
Deposition
process
Typical rate,
Ib/curb-mi/day
Mud and dirt 100
carryout
Litter 40
Biological 20
debris
Ice control 20
compounds
Dustfall 10
Pavement wear 10
& decomposition
Vehicle-related 17
(incl. tire wear)
Spills <2
Erosion from 20
adjacent areas
Removal
process
Reentrainment
Displacement
Wind erosion
Rainfall
runoff
Sweeping
Typical rate,
Ib/curb-mi/day
100
40
20
50
35
Deposition and removal rates are not equal in the above
tabulation because of approximations used for individual
processes. Based on this crude mass balance for materials
entering and leaving a typical street, it appears that
almost half the material may leave the street as particulate
air pollution by either reentrainment or wind erosion.
POTENTIAL CONTROL MEASURES
Several control measures were identified which could
possibly reduce the amount of suspended particulate caused
by reentrainment from paved streets. All but one of these
potential measures was concerned with reducing the amount of
material on street surfaces, under the initial assumption
that reducing street surface loadings would proportionately
reduce the amount of material that is available for reen-
trainment. The measures which were proposed and then
evaluated in subsequent project work are:
-------
Improved street cleaning
Broom sweeping
Vacuum sweeping
Regenerative air sweeping
Flushing
Modified snow and ice control procedures
Replace sand with salt or a salt/sand
mixture
Plow instead of sand or salt
Remove anti-skid material as soon as
possible
Apply less material
Control of dirt and mud carryout from major
sources (e.g., construction sites, truck
terminals)
Control of dirt and mud carryout from
ubiquitous sources (e.g., unpaved parking
areas, shoulders, and driveways)
Reduction in traffic volumes on a street
FIELD STUDIES
Five field studies were done in conjunction with this
project—three were comparisons of the air quality impacts
of alternative street cleaning techniques, one was a com-
parison of the air quality impacts of different methods of
controlling mud carryout at a construction site, and the
remaining one was intended to investigate the relationships
among traffic volumes, street surface loadings, and particu-
late concentrations near the street.
In addition, particulate air quality data were obtained
for six cities in which potential control measures had been
implemented—some type of street cleaning program in five
cities and control of trackout sources in the sixth. The
air quality data were reviewed to determine whether the pro-
grams had a discernible effect on particulate concentrations,
The findings of the project field studies and the six
other data analyses were inconclusive with regard to the
effectiveness of improved street cleaning as a control
-------
measure. Although some of the street cleaning methods on
occasion appeared to reduce particulate concentrations near
the streets that were cleaned, none of the methods proved to
be effective in all of the studies in which they were evalu-
ated, as shown below:
Studies in Studies in
which method which method
Cleaning method was effective was ineffective
Broom sweeping 1 2
Vacuum sweeping 0 2
Regenerative air sweeping 0 1
Flushing 2 2
Sweeping and flushing 0 1
In addition, it was concluded from data generated in
the field studies that there is no consistent relationship
between street surface loadings and nearby particulate
concentrations. Correlation between these two variables was
not improved when either the street loading in traffic lanes
or the loading of material less than 44 u diameter was sub-
stituted for total street loading. Since the rate of
reentrainment was not shown to be a function of the amount
of material on a street, reducing the amount of material by
street cleaning might not consistently reduce particulate
concentrations.
In contrast to the street cleaning studies, both of the
mud carryout control studies showed significant reductions
in particulate concentrations. At the construction site
study, intensive manual cleaning reduced nearby concentra-
tions by 10 to 20 ug/m . The area of impact of mud carryout
was estimated to be about a 1000 ft radius around the site.
Data from a subregional area of Seattle indicate that similar
reductions in concentrations can be obtained over a large
area through a comprehensive control program for mud carryout
-------
sources such as unpaved shoulders, unpaved parking areas,
and truck terminal lots.
IMPLEMENTATION OF CONTROL MEASURES
Implementation of control measures for reentrained dust
will require new approaches by air pollution control agencies.
Most importantly, authority and capabilities for reducing
reentrained dust from streets reside with city public works
departments. Air quality improvement may be of concern
within public works agencies, but it would not necessarily
be a consideration with high priority. Therefore, unless
these agencies receive a strong directive or increased
funding, only marginal changes in street cleaning or dirt
and mud carryout controls can be expected. Time require-
ments for programming and budgeting usually preclude rapid
implementation of program modifications.
Implementing a control measure for reentrained dust may
also require a well-developed educational presentation.
Reentrained dust, unlike a major point source, is not an
obvious, visible source of particulate emissions. It would
be difficult to get public or local government support for
the increased costs of an expanded street cleaning program
or a transportation control measure without strong documen-
tation of the air quality impact of reentrained dust and the
anticipated effect of the proposed measure.
The proposed method of implementation focuses on drafting
of an administrative agreement between the state air pollu-
tion control agency and the local government. This document
would formalize the approval of the city administration and
establish the framework for cooperative development of the
control measure, whether it is based primarily on modifica-
tions to the street cleaning program or control of dirt and
mud carryout sources.
-------
Street cleaning, at least for the level of cleaning
evaluated, is still unproven as a control measure and
should therefore be adopted cautiously. A pilot study
incorporating proposed changes is recommended prior to any
full-scale modifications of a city's street cleaning pro-
gram, at least until more data on the effectiveness of
improved street cleaning become available. The interest and
support of the public works department are required in order
for the measure to be successful. If the changes are not
supported or viewed as important, the expanded cleaning
program will probably not translate into an air quality
improvement.
Another reason for testing street cleaning modifica-
tions on a smaller scale is that study data from one city
may not be applicable in another due to great differences in
street systems (storm drainage, curbs and gutters, age and
type of surface, etc.) and productivities of street depart-
ments .
Mud carryout controls have a more consistent record of
success, and can be readily implemented by an air pollution
control regulation or ordinance for major sources such as
construction sites. An administrative agreement with local
governments may also be needed to control municipally-owned
mud carryout sources such as unpaved shoulders and streets.
The most difficult sources to encompass under a regulation
are the minor, privately-owned sources such as driveways,
unpaved parking areas, and storage lots.
The limited data available indicate that traffic re-
ductions could also be effective in reducing urban particu-
late concentrations. Traffic reductions would probably only
be considered in urban areas where a control plan requiring
such reductions would be needed for carbon monoxide or
oxidants.
-------
COSTS
In most cases, either mud carryout controls or improved
street cleaning (in areas where it can be demonstrated to be
effective) would be much more cost-effective in reducing
urban particulate concentrations than additional point
source or area source controls. The costs are estimated to
be on the order of $1,000 per ug/m reduction in annual
2
average concentrations over an area of one mi , and the
measures can be applied somewhat selectively in those areas
where particulate concentrations are highest (non-attainment
areas).
-------
1. INTRODUCTION
SCOPE OF WORK
*
According to several recent studies, reentrained dust
from paved streets and other traffic-related emissions are
major sources of suspended particulate in urban areas and
are a leading cause of concentrations above the ambient air
quality standards. Yet, these sources are not subject to
any controls.
This document was prepared to identify control measures
which can reduce the air quality impact of traffic-related
particulate emissions, to compile available information on
these control measures, and to report on several field
studies which were done specifically to evaluate the effec-
tiveness of control measures for reentrained dust. It also
proposes procedures for planning and implementing the con-
trol measures. Finally, the document provides detailed
information on the characteristics of reentrained dust and
of the material found on street surfaces.
Most of the effort in this assessment of controls for
reentrained dust was allocated to five field studies. Three
of the field studies investigated street cleaning methods,
one examined control of a mud carryout source, and the other
attempted to identify the variables which affect reentrained
dust emission rates. The field studies and their results
are discussed in Chapter 3.
*
These studies are described on pages 11 through 13,
-------
In addition, six cities were located in which potential
control measures had been implemented. Particulate air
quality data for time periods with and without the control
programs were obtained and analyzed to determine effects on
particulate concentrations. These data analyses are also
presented in Chapter 3.
During the project, many public works personnel were
contacted to obtain information. Cost data for street
cleaning and for mud carryout controls are summarized in
Chapter 4. Specific methods of reducing reentrained dust
impact, developed from results of the field studies and
through discussions with public works personnel, are pre-
sented in Chapter 5.
Potential regulatory approaches are discussed in Chapter
6. Existing public works and air pollution control ordinances
were reviewed to uncover innovative approaches.
An extensive literature review performed early in the
study was relied upon to provide much of the information on
characteristics of reentrained dust and material on street
surfaces. These two topics are covered briefly in Chapters
1 and 2, respectively, to provide the reader with appropriate
background information. More detailed presentations on
reentrained dust and material on street surfaces can be
found in Appendices A and B.
RECOGNITION OF THE REENTRAINED DUST PROBLEM
No references to reentrained dust from streets as an
emission source were found in the literature prior to 1972.
At that time, the Puget Sound Air Pollution Control Agency
performed some limited sampling of reentrainment by mounting
2
impaction plates on a trailer pulled behind the test vehicle.
Shortly thereafter, some microscopic analyses of high volume
samples for Chicago and Philadelphia were reported.3'4
10
-------
The samples had large quantities of mineral particles which
appeared to be road surface aggregate.
As the date for attainment of the National Ambient Air
Quality Standards (NAAQS) approached and many metropolitan
areas still had particulate concentrations well above the
primary annual standard for particulate, many studies were
initiated to better identify the sources contributing to the
high particulate concentrations. These studies were con-
ducted concurrently and were done independently of one
another using many different methodologies—microscopy,
emission inventory, statistical analysis of historical air
quality data, detailed survey of the areas surrounding the
samplers, additional sampling to determine concentration
gradients, etc.
All of these recent studies have arrived at similar
conclusions: reentrained dust from streets has an annual
impact of 10 to 30 ug/m at most urban sites and constitutes
10 to 50 percent of the particulate emissions in these
areas. The studies are described briefly below:
In a national assessment of the urban particu-
late problem which investigated 14 metropolitan
areas, the average impact from vehicle-related
emissions was found to be 15 to 23 ug/m3 in
residential areas and 22 to 30 ug/rn-^ in commer-
cial and industrial areas; at individual sites,
the impact was as much as twice these amounts.-*
At 10 temporary sampling locations on sidewalks
in downtown and suburban Philadelphia, micro-
scopic analysis indicated that 50 to 90 percent
of the particulate matter collected was from
vehicular traffic.
After analyzing eight years of air quality data
for Chicago, one researcher attributed 13 to 15
ug/m3 on an annual basis at all sites to reen-
trained dust from vehicular traffic.6 A separate
11
-------
study of Chicago's ambient monitoring network
which employed microscopy to identify particle
origin concluded that 10 to 80 percent by weight
of the collected material was reentrained, much
of it by auto traffic, and that 0»5 to 5 percent
was automobile exhaust emissions.
An analysis of the Kansas City area showed that 30
to 55 ug/m3 were contributed to annual average
concentrations in areas exceeding the primary
standard by reentrained dust from motor vehicles.
An emission factor of 10.7 g/veh-mi was used.
However, if the value of 3.7 g/veh-mi obtained
from the project field studies in Kansas City
were substituted, the impact from reentrained
dust would be 10 to 19 ug/m3.7
An emission inventory of Nashville and Davidson
County, Tennessee revealed that 50 percent of the
area source emissions in this urban area in 1975
were due to reentrained dust from motor vehicles.
Preliminary data from a study conducted in New
York City showed that average particulate concen-
trations at sites located near streets were 26
ug/m3 higher during the 12 hours of the day with
high traffic volumes than during the remaining
12 hours.^
Particulate diffusion modeling of Jacksonville
revealed that 10 to 20 percent of the ambient
concentrations in most areas of the city were
from vehicle-related emissions.10
In business and residential areas of Phoenix,
vehicle-related emissions were found to consti-
tute 35 to 50 percent of total particulate
emissions.H
Semiquantitative elemental analysis and micro-
scopic analysis of particulate samples from the
industrial area of Seattle showed an average of
39 percent of the collected material originated
from motor vehicles or reentrained road dust,
12
-------
with about 35 percent being attributed to reen-
trained dust.1^
In Colorado Springs, diffusion modeling indicated
that 29 percent, or about 25 ug/m3, of the annual
average concentrations in areas currently ex-
ceeding the primary standard result from reen-
trained dust from streets and that an additional
5 percent is contributed by motor vehicle ex-
haust and tire wear. 3
Particulate diffusion modeling of the Tampa Bay
area showed 10 to 25 percent of the ambient con-
centrations in areas exceeding the primary annual
standard were from vehicle-related emissions.14
If vehicle-related particulate emissions have not been
identified as one of the most important source categories in
any large city, this is probably because the emission inven-
tory for that city has not been updated to include reentrained
dust from traffic as a source.
CHARACTERISTICS OF REENTRAINED DUST
A brief summary of the important characteristics of
reentrained dust is included here to assist the reader in
interpreting data presented later in the report. Much of
the information for this section was obtained during the
field studies done for this project. The field study designs
are shown in Appendix C and the characteristics of reentrained
dust are discussed in much more detail in Appendix A.
According to precise definition, reentrained dust is
only that portion of vehicle-related emissions which is
thrown from the street surface by contact with vehicle tires
or is induced to become airborne by the vortexes from passing
vehicles. Other particulate matter is emitted directly by
the vehicles: engine exhaust, wear of various parts such as
bearings and brake and clutch linings, and from abrasion of
13
-------
tires against the road surface. However, many of these
direct emissions settle to the road surface, become part of
the street surface loading, and are subsequently reentrained.
It is very difficult to distinguish in sampling whether
the suspended material has been directly emitted or reen-
trained. It does appear from studies of total traffic-
related impact that the reentrained portion is an order of
magnitude greater than the direct emissions accounted for by
currently available emission factors for vehicle exhaust
(0.34 g/veh-mi) and tire wear (0.20 g/veh-mi).
The average vehicular emission rate calculated in the
field studies from short-term downwind ambient concentra-
tions using a line source dispersion equation was 3.7 g/veh-
mi, with a standard deviation of 3.3 g/veh-mi. This was
determined to be equivalent to an JJfuJMnJL emission rate of
4.9 g/veh-mi with the observed average fallout rates of 14
percent at 10 m, 26 percent at 20 m, and 34 percent at 30
m from the street. A recent EPA-sponsored emission factor
development study reported an average emission rate for
particles less than 30 u of 8.5 g/veh-mi, and a summary of
resuspension studies concluded that the resuspension rate
for particles less than 40 u is probably in the range of 1
to 5 g/veh-mi.15'16
Suspended particulate samples taken near streets in the
present field studies had relatively large particle sizes—
a mass median diameter of 15 u and approximately 22 percent
by weight greater than 30 u. The samples also displayed a
strong bimodal distribution, with a large percentage of the
material being less than 3 u in diameter and another large
percentage in the 15 to 30 u range.
The small particulate (less than 3 u) was identified
microscopically as primarily combustion product, while most
of the large material was of mineral origin. Combustion
14
-------
products (from both mobile and stationary sources) averaged
a surprisingly large 40 percent of the weight/ indicating
that exhaust emissions may constitute a larger portion of
total vehicle-related emissions than previously suspected.
Mineral matter constituted an average of 59 percent of the
particulate collected near streets, while biological matter
and tire tread together comprised a little more than one
percent. Inorganic lead was determined chemically to account
for an average of one to two percent of the suspended par-
ticulate.
Four variables were evaluated to determine their effects
on reentrainment—traffic volume/ traffic speed, street
surface loading, and wind speed. Particulate concentrations
near a street were found to be fairly closely related (r =
0.5 to 0.8) to traffic volume within the range of traffic
volumes tested. A weak inverse relationship (r = -0.3)
between particulate concentrations and traffic speed was
observed. No relationship was found between particulate
concentrations and street surface loadings. The data showed
that wind speed is directly (r = 0.3 to 0.4) rather than
inversely related to particulate concentrations, indicating
that the dust generating and transport effects of increased
wind speed more than offset the diluting effect of the
higher wind speed.
POTENTIAL CONTROL MEASURES
Prior to this study, no information was available on
methods of reducing the emissions or impact of reentrained
dust. Consequently, one of the initial tasks was to identify
measures which could possibly reduce the amount of suspended
particulate caused by reentrainment. The resulting list
includes five general approaches:
15
-------
Improved street cleaning
Broom sweeping
Vacuum sweeping
Regenerative air sweeping
Flushing
Modified snow and ice control procedures
Replace sand with salt or a salt/sand
mixture
Plow instead of sand or salt
Remove anti-skid material as soon as
possible
Apply less material
Control of dirt and mud carryout from major
sources (e.g., construction sites, truck
terminals)
Control of dirt and mud carryout from
ubiquitous sources (e.g., unpaved parking
areas, shoulders, and driveways)
Reduction in traffic volumes on a street
All but the last of these measures are concerned with
reducing the amount of material available for reentrainment
from street surfaces. Also, all of the measures are outside
the normal areas of authority of air pollution control
agencies.
Each of the four types of municipal street cleaning
equipment was proposed for consideration as a control mea-
sure. Broom sweepers utilize a rotating gutter broom to
move material from the curb area (where most of the debris
is located) into the main pickup broom and then onto a belt
and into the hopper. Most broom sweepers are equipped with
a water spray for dust suppression.
Vacuum sweepers use a broom to loosen and move the dirt
and a vacuum system to pick it up. All material picked up
by the vacuum nozzle is saturated with water on entry and
passes into a vacuum chamber where it drops out of the air
stream. Another type of unit similar to the vacuum sweeper,
the regenerative air sweeper, is designed to blast dirt from
16
-------
the road surface into the hopper, with part of the air
stream being recycled and part vented through the dust
separation system.
Street flushers consist of a water supply tank mounted
on a truck, a pump to provide pressure, and three or more
nozzles to spread the water in directional sprays. The
nozzles are individually controlled and are usually placed
so that one is directed across the path of the flusher and
the two others are pointed out toward the gutters. Flushing
is used primarily to displace dirt from the travel lanes of
the street into the gutter. The volume of water is not
sufficient to transport the accumulated material to the
nearest drain. Flushing is also presumed to reduce reen-
trainment for a short period of time by its wetting action.
None of the available street cleaning equipment was
designed for air pollution control purposes. Future equip-
ment could be more effective because of such design changes.
Specific measures which can be employed to control dirt
and mud carryout by trucks from locations such as construc-
tion sites and truck terminals include paving (for permanent-
traffic areas), use of coarse gravel or fragmented brick,
chemical stabilization, requiring trucks to have tire scrapers,
wheel washes at construction site exits, and required covering
or wetting of loaded trucks. Very little can be done to
prevent the trackout that occurs during excavation. Con-
tractors can be required to clean access streets manually or
with mechanical sweepers while excavation is in progress.
Another source of mud on streets near construction
sites is runoff. This source can be controlled by measures
such as temporary grading, on-site retention, compacting or
stabilization, revegetation, and limiting the acreage of
cleared area that can remain exposed.
A large number of unpaved areas adjacent to streets
(e.g., parking lots, road shoulders, driveways, industrial
17
-------
and warehouse lots) can increase the amount of material on
street surfaces. Control of these small sources can be
achieved by requiring paving, oiling, or seal coating to
limit trackout and to reduce directly-emitted dust.
18
-------
2. MATERIAL ON STREET SURFACES
Before control measures can be developed to reduce the
amount of material deposited on streets, it must first be
determined where most of the material found on street sur-
faces originates. Also, prior to analysis of the effect of
improved street cleaning on dust reentrainment rates, it
would be desirable to know the relative amounts of material
removed from streets by cleaning, reentrainment, and other
removal processes. Therefore, the purposes of this chapter
are to identify the various deposition and removal processes
which determine the amount of material on a street surface
and to attempt to quantify typical rates at which each of
these processes operate.
Most of the data used to estimate deposition and remov-
al rates was obtained from the literature rather than from
analysis of street loading samples taken during the field
studies to provide data for this document. References and
estimating procedures are presented in Appendix B.
The street loading measured at any time represents the
net balance of all the deposition and removal processes,
some acting almost continuously, some intermittently, and
some (e.g., street sanding) on only a few occasions per
year. These are shown schematically in Figure 2-1.
Two of the primary removal mechanisms, rainfall and
street sweeping, are intermittent. To the extent that
deposition exceeds removal rates between these intermittent
processes, material accumulates on streets.
19
-------
1 Pavement wear and decomposition
2 Vehicle-related deposition
3 Dustfall
4 Litter
5 Mud and dirt carryout
6 Erosion from adjacent areas
,7 Spills
18 Biological debris
\9 Ice control compounds
DEPOSITION
REMOVAL
1 Reentrainment
2 Wind erosion
3 Displacement
4 Rainfall runoff to catch basin
5 Street sweeping
Figure 2-1. Deposition and removal processes.
-------
"Typical" annualized deposition rates for nine pro-
cesses identified in the literature as significant contrib-
utors to street surface loadings are presented in Table 2-1.
It should be emphasized that these values are only order-of-
magnitude estimates and should not be considered representa-
tive of deposition on any single day on a particular street.
All of these processes except application of ice control
compounds occur at a fairly constant rate with time.
In contrast, most removal processes are intermittent
and their rates are primarily a function of the amount of
material on the street at the time they occur. Estimates of
typical rates for the five major processes for removal of
material from streets are shown in Table 2-2.
Some of the characteristics of the material found on
street surfaces influence removal rates:
Most of the material is too large to become
directly airborne, with the <44 u fraction
averaging 6 and 10 percent in two different
investigations.
Most of the material collects within a foot
of the curbs rather than being uniformly
distributed across the streets.
Normally, more than 95 percent by weight of
the material is inorganic, with a bulk den-
sity of about 1.6 g/cm3.
21
-------
Table 2-1. DEPOSITION PROCESSES
Source
Constituents
Typical depo-
sition rate, Range,
Ib/curb-mi/day Ib/curb-ni/day
1.
Mud and dirt
carryout
2. Litter
3.
4.
Biological
debris
Ice control
compounds
5. Dustfall
6. Pavement
wear and
decompo-
sition
7. Vehicle-
related
-Tire wear
-Brake and
engine com-
ponent wear
-Settleable
exhaust
8. Spills
9. • Erosion
(runoff and
blowing)
from adjacent
areas
Total
Soil from con-
struction sites,
unpaved parking
areas, etc.
Cans, bottles,
broken glass,
cigarette butts,
plastic, other
debrj s
Leaves, grass
clippings,
sticks, animal
droppings, in-
sect parts, etc.
Sand, salt, cin-
ders , calcium
chloride
Atmospheric
fallout
Asphalt, cement,
aggregate, ex-
pansion joint
compounds and
fillers
Rubber
Metals, lubri-
cants, brake and
clutch linings
Combustion pro-
ducts, fuel
additives
Sand, dirt,
chemicals
Soil
100
40
20
20
10
10
Extreme
Extreme
Extreme
0-60
2-25
5-150
10
5
6-50
2-25
1-10
No data;
est <2
20
240
Extreme
22
-------
Table 2-2. REMOVAL PROCESSES
Process
Typical removal
rate,
Ib/curb-mi/day Assumptions incorporated
Reentrainment 100
Displacement 40
Wind erosion 20
Rainfall runoff 50
Sweeping 35
For 10,000 ADT; net removal
rate =4.5 g/VMT
Estimated from dustfall rate
just beyond curb
Force of same magnitude as
reentrainment, but only
operative 20% of time
Removal efficiencies of 50%
for rain of 0.1-0.5. and
90% for rain of >0.5 in.
Average efficiency of
removal = 50%; weekly clning
23
-------
3. TESTING THE EFFECTIVENESS OF
CONTROL MEASURES
Several of the potential control measures identified in
Chapter 1 were investigated to determine their impacts on
particulate air quality as measured with high volume samplers.
There are three types of evaluations included in this chapter;
project field studies specifically designed to
quantify the effect of a control measure (per-
formed by PEDCo-Environmental in study areas in
Kansas City and Cincinnati) ;
evaluations in which a control measure in the
form of a municipal program or short-term test
was implemented in a city and available air
quality data were collected after-the-fact
to estimate the air quality impact of the pro-
gram; and
estimates based on theoretical approaches for
control measures which were not subjected to
field studies.
Each of the four project field studies and six program
evaluations is presented as a separate subsection. The
results of all these evaluations are then summarized and
collective estimates of control efficiencies made at the end
of the chapter.
More information on the four project field studies can
be found in Appendix C.
*
The fifth project field study investigated variables which
affect reentrained dust emission rates. Since it did not
evaluate a specific control measure, this field study is not
discussed in this chapter. Its study design is described in
Appendix C.
24
-------
PROJECT FIELD STUDIES
Street Cleaning in Kansas City, Missouri
Study Design - The primary purpose of this study was to de-
termine the effect of alternative street cleaning methods on
particulate air quality in the area being cleaned. Two
street cleaning methods were evaluated: broom sweeping and
flushing. Streets in the study area were cleaned once
weekly for a period of one month with each type of equipment.
Also, the streets were not cleaned for a one-month period.
Particulate concentrations were measured daily at eight
sampling sites in the study area for the entire three
months. Average concentrations associated with each cleaning
period were compared after being adjusted to account for
differences in regional particulate concentrations during
the three periods.
The four by seven block study area was located in a
commercial/warehousing district just south of Kansas City's
central business district. Almost all the streets in the
area have average daily traffic (ADT) volumes of 4,000 to
15,000 vehicles.
Results and Conclusions - Particulate concentrations were
relatively uniform at the eight sites on any one day in
comparison with the day-to-day variations at all the sites.
The plot of daily data also showed a consistent weekly
pattern of concentrations, with low concentrations on week-
ends, slightly higher concentrations on Mondays, and high
concentrations on other weekdays. The weekly pattern was
attributed to differences in daily traffic in the study
area.
Average concentrations during each of the cleaning
periods are shown in Table 3-1. The flushing period was
only three weeks due to the early onset of cold weather.
25
-------
Table 3-1. COMPARISON OF PARTICULATE CONCENTRATIONS IN
KANSAS CITY DURING THREE CLEANING PERIODS
Geometric mean concentration, ug/m
Site Broom sweep
no. 09/13-10/10
Sites in
cleaning
area
A-l
A- 2
A- 3
A- 4
A-5,6
A-7
A- 8
Avg.
Control
sites
C-13
C-14
C-15
Avg.
15 Regional
stations
"A" sites
- Reg. sites
"A" sites
- "C" sites
107.4
115.6
104.4
113.4
113.4
149. 8C
114.5
116.2
52.9
54.0
100.7
66.0
84.1
32.1
50.2
Flushing
10/17-11/07
98.0
107.6
90.0
98.6
97.3
105.4
94.4
98.6
59.0
59.6
84.8
66.8
74.6
24.0
31.8
No cleaning
11/08-12/12
,
125.6
140.8
111.7
126.9
130.8
134.2
114.5
126.0
71.1
72.4
89.2
77.1
92.5
33.5
48.9
Relative rank
of periods by
geom
High
N
N
N
N
N
B
N
N
N
N
B
N
N
.. mean"
Med
B
B
B
B
B
N
B
B
F
F
N
F
B
Low
F
F
F
F
F
F
F
F
B
B
F
B
F
Calculated using every other day sample values.
B = Broom sweeping
F = Flushing
N = No cleaning
Interference during sampling period from local street
repair work.
26
-------
This added an extra week to the no cleaning period. The
unadjusted data indicated that air quality was best with
flushing, next best with broom sweeping, and worst when
there was no cleaning. When the average concentrations for
the three periods were adjusted to account for weather-
related and seasonal variations between periods, the apparent
difference between concentrations during the broom sweeping
period and with no cleaning was eliminated (or broom sweeping
showed no reduction in particulate concentrations). However,
the flushing period still showed 8 to 18 ug/m (geometric
mean) lower concentrations after adjustment for external
differences in air quality.
Three samplers located in suburban areas five to six
miles from the study area were used to provide a comparison
of concentrations uninfluenced by the street cleaning
during the three periods. Also, concentrations measured
during the three periods at 15 high volume samplers in the
regional network were used for a secondary comparison.
Sampling was every sixth day at the regional network sites
versus daily in the study area.
Rainfall obviously reduced concentrations on days when
it occurred. In the Kansas City study area, the average
reduction in concentrations on days with 0.1 inch or more of
rain was 28 ug/m . However, on the first day after a rain
only a minimal residual effect was observed and on subsequent
days no residual effect was noted. A similar impact from
rainfall has been reported for other cities.
Days with flushing in Kansas City inexplicably showed
higher than average concentrations, although concentrations
on the first days following flushing were lower than average.
The characteristic change in particulate concentrations
following broom sweeper cleaning in Kansas City was a moder-
ate increase on the day of cleaning, a definite reduction on
the day after cleaning, and then sharply higher concentra-
tions for the next few days.
27
-------
Street Cleaning in Cincinnati
Study Design - The purpose and approach for this study were
the same as for the Kansas City street cleaning study—a
comparison of particulate concentrations in a five by seven
block area during periods with different street cleaning
methods. In Cincinnati, three cleaning methods (flushing,
broom sweeping, and vacuum sweeping) were evaluated for one
month each, plus there was no street cleaning for a non-
continuous one-month period. The cleaning cycles were ir-
regular, but consistent from one month to another: the
streets were cleaned early in the first week of each period,
not at all during the second week, twice during the third
week, and once in the fourth week. Coincidentally, there
were three days of rain between the first and second clean-
ing during eaqh period, so there were no times longer than a
week without cleaning or natural washing of the streets and
several periods with cleaning at three- or four-day intervals,
As in the Kansas City study, average concentrations for
each of the cleaning periods were adjusted to account for
differences in particulate concentrations between the
periods due to factors other than street cleaning (primarily
weather-related and seasonal variations).
The Cincinnati study area was located in a primarily
residential area with one commercial street, Hamilton Avenue.
This street has an ADT of 17,700 but most other streets in
the area have less than 1,000 vehicles per day. Seven
samplers were placed in the study area and operated daily.
Results and Conclusions - Concentrations in the Cincinnati
study area were much lower than in Kansas City, and the
impact from traffic-related emissions appeared to be much
less. No distinct weekly pattern of concentrations was
seen. Differences between weekday and weekend concentra-
tions were not expected, since traffic data indicated that
28
-------
traffic volumes were approximately the same for weekdays and
weekends.
Average concentrations for each of the sampling periods
are shown in Table 3-2. Without considering variations
between sampling periods due to weather-related factors, air
quality was best during the month of broom sweeping, nearly
the same during the months with flushing and vacuum sweep-
ing, and worst when there was no street cleaning.
There was no sampling site operated daily outside the
cleaning area that could be used to estimate differences in
air quality for the four periods. To provide some measure
of the variation in external air quality, concentrations at
seven regional control agency sites within a four mile
radius were determined for the four periods. The sampling
schedule at these sites was every sixth day. As indicated
by the data in Table 3-2 for the seven sites, the high
concentrations during the no cleaning period occurred out-
side the study area, too.
According to the air quality data, broom sweeping was
the most effective cleaning method. Concentrations averaged
6 to 20 ug/m less during this period than during the other
three periods regardless of whether data from sites outside
the study area were used to adjust for the differences in
concentrations between periods.
The average concentrations for the month of flushing
did not indicate a significant reduction by use of this
cleaning method, although a plot of concentrations in the
Cincinnati study area as a function of time since flushing
showed that concentrations were 16 ug/m lower on the days
when the flushing was done and 4 ug/m less than average on
the first day after flushing. The air quality impacts from
flushing and broom sweeping in the Cincinnati study were
reversed from what they were in the Kansas City study.
29
-------
Table 3-2.
COMPARISON OF PARTICULATE CONCENTRATIONS IN CINCINNATI
FOR FOUR CLEANING PERIODS
OJ
o
Geometric mean concentration, ug/m
Site Flushing Broom sweep Vacuum No clean
no. 09/20-10/17 10/18-11/14 11/22-12/17 4 weeks
Sites in
cleaning;
area
2-B
2-T
3-B
3-T
1
4
5
Avg.
Control
sites
7 Regional
sites
65.0
59.0
62.1
58.8
57.3
49.5
56.3
58.0
58.7
62.
53.
53.
48.
52.
44.
49.
51.
61.
0
3
4
3
7
2
4
7
4
72.
62.
60.
53.
60.
47.
53.
57.
54.
6
1
8
0
0
0
4
9
5
78
71
73
68
74
67
67
71
73
.2
.8
.4
.1
.9
.5
.4
.5
.7
Relative rank
of periods by
geom. meana
High 2nd 3rd Low
N
N
N
N
N
N
N
N
N
V
V
F
F
V
F
F
F
B
F
F
V
V
F
V
V
V
F
B
B
B
B
B
B
B
B
V
Study area -0.7
sites
- Reg. sites
-9.7
3.4
-2.2
F = Flushing
B = Broom sweeping
V = Vacuum cleaning
N = No cleaning
-------
Taken at face value, the air quality data from the
study area and from surrounding sites indicated that vacuum
sweeping of streets increased concentrations about 5 ug/m
compared to no cleaning. This was thought to be an anomaly
resulting from the noncorresponding sampling schedules for
the two data sets being compared—daily sampling in the
study area and every sixth day for the sites outside the
study area. Nevertheless, comparison of concentrations
during the vacuum cleaning period with those from other
cleaning periods could only lead to the conclusion that this
method was not effective in reducing particulate concentra-
tions in the study area.
This finding is unexpected considering the demonstrated
efficiency of the vacuum units in removing small size par-
ticles from street surfaces and their overall street clean-
17
ing efficiency in the study area. As shown in Table 3-3,
the vacuum sweepers had removal efficiencies consistently
higher than the broom sweepers. These cleaning efficiencies
were determined from street loading measurements taken
before and after each cleaning operation.
31
-------
Table 3-3. CLEANING EFFICIENCY OF STREET CLEANING METHODS
Type of
cleaning
Broom
Broom
Broom
Broom
Broom
Broom
Flush
Flush
Flush
Flush
Flush
Vacuum
Vacuum
Vacuum
Vacuum
Date/Location
Sep 22 K.C.
Sep 29 K.C.
Oct 20 CINC
Nov 03 CINC
Nov 06 CINC
Nov 10 CINC
Oct 20 K.C.
NOV 02 K.C.
Oct 06 CINC
Oct 09 CINC
Oct 13 CINC
Nov 24 CINC
Dec 07 CINC
Dec 11 CINC
Dec 15 CINC
Percent removal of material
Curbs and
traffic
lanes
41.8
10.9
40.4
57.5
24.2
60.9
-2.6
n.a.
1.3
5.9
18.4
62.1
70.4
60.0
53.2
Particle size range
<44 44-106 106-841
-77a -136
-11 -15
1 34
63 80
8 24
9 40
-38 -13
90 90
29 25
-1 16
21 25
34 62
79 86
-85 2
-17 48
62
11
52
62
23
52
3
-171
7
-3
69
59
75
60
63
, um
>841
65
45
30
1
28
78
-2
n.a.
-54
20
1
71
32
73
65
In traf-
fic lanes
only
52.1
33.7
-109.0
n.a.
n.a.
n.a.
43.6
58.4
34.9
-51.0
-55.6
n.a.
n.a.
n.a.
n.a.
to
Negative values indicate an increase in material following cleaning.
n.a. = data to calculate this value not available.
-------
Street Cleaning in Residential Areas
Study Design - This study was intended to show whether the
regular use of vacuum sweepers on low traffic density resi-
dential streets results in improved particulate air quality.
A high volume sampler was located in the center of the
City of Westwood, a residential suburb of Kansas City.
Streets and gutters in Westwood appear visibly clean at all
times. All the streets in the city are cleaned about once a
week with a vacuum sweeper, while surrounding communities
use broom sweepers. Particulate concentrations were also
measured on the same sampling schedule (every other day) in
the adjacent community of Roeland Park, where street clean-
ing is done less frequently with mechanical broom sweepers
and some of the streets are not curbed and guttered. An
effort was made to place the two samplers in comparable
exposures: both were located atop one story, flat roof
buildings at distances of about 150 feet from the nearest
streeto There are no particulate point sources in either
community, so any reduction in ambient concentration at the
Westwood site should provide a measure of the impact of
vacuum street sweepers on subregional particulate air quality,
Results and Conclusions - The difference in daily concentra-
tions for the sampling sites in Westwood and Roeland Park
for the 43 sampling days indicated that there was no signi-
ficant difference in air quality at these two locations.
The data are shown in Table 3-4. A t-test showed an average
difference of 0.1 ug/m and a corresponding t value of 0.08
compared to the calculated t value (at 95 percent confi-
dence) for rejecting the null hypothesis of 2.04.
Vacuum street cleaning appears to provide excellent
removal of material from residential street surfaces, but it
was concluded that optimal street cleaning in such a low
traffic density area does not improve air quality because
there is not a significant contribution from reentrained dust.
33
-------
Table 3-4. PARTICULATE CONCENTRATIONS AT
WESTWOOD AND ROELAND PARK SITES
Date
09-12
09-14
09-17
09-19
09-21
09-23
09-25
09-27
09-29
10-01
10-03
10-05
10-07
10-09
10-11
10-13
10-15
10-17
10-19
10-21
10-23
10-25
10-27
10-29
10-31
11-02
11-04
11-06
11-08
11-10
11-12
11-14
11-16
11-18
11-20
11-22
11-24
11-26
11-28
11-30
12-02
12-04
12-06
12-08
12-10
12-12
Average
Particulate
concentration ,
Westwood Roeland Park
60.5
69.6
85.4
39.5
57.1
59.6
56.8
43.0
51.3
77.9
47.6
28.3
43.4
56.4
45.5
81.9
207.1
65.5
43.8
52.5
36.4
58.3
n.d.
45.1
86.5
64.8
80.0
79.2
77.5
118.3
104.0
62.8
(45.0)
115.4
83.2
55.7
54.9
52.2
36.5
59.9
82.2
126.2
56.6
64.6
68.7
n.d.
68.4
60.1
77.3
95.2
37.7
53.0
65.5
62.4
44.3
58.6
61.6
50.3
29.6
48.9
47.0
48.4
75.6
200.6
72.7
39.0
50.9
36.7
72.3
n.d.
47.9
75.0
71.1
74.7
77.3
79.6
107.4
77.8
66.0
n.d.
115.2
84.6
45.7
50.2
60.5
38.8
64.7
93.2
113.9
60.5
67.1
77.6
(57.2)
68. 3
ug/m
Difference
+0.4
-7.7
-9.8
+1.8
+4.1
-5.9
-5.6
-1.3
-7.3
+16.3
-2.7
-1.3
-5.5
+9.4
-2.9
+6.3
+6.5
-7.2
+ 4.8
+1.6
-0.3
-14.0
-2.8
+11.5
-6.3
+5.3
+1.9
-2.1
+ 10.9
+26.2
-3.2
+ 0.2
-1.4
+ 10.0
+4.7
-8.3
-2.3
-4.8
-11.0
+ 12.3
-3.9
-2.5
-9.1
+ 0.1
34
-------
Mud Carryout Control At A Construction Site
Study Design - This study was designed to evaluate the
effect on air quality of potential control measures for
reducing the amount of material deposited on streets from a
major mud carryout source. The control methods evaluated
were: (1) frequent cleaning of the access area with municipal
street cleaners; (2) manual cleaning of the access area at
intervals with broom and shovel; and (3) immediate manual
cleaning of tracked material. A one-time tracer study was
also conducted at this site to determine how far material is
distributed from its initial carryout point by traffic on
the nearby streets.
Four high volume samplers were placed at three locations
near a building construction site in Kansas City. Samplers
were located in both directions along the access street from
the construction site entrance to ensure that the impact of
mud carryout would be monitored even if the mud tracking was
concentrated in one direction. The effectiveness of each
control measure was determined by comparing air qualities
during three cleaning periods with air quality during the no
control period.
Results and Conclusions - Particulate air quality in the
vicinity of the construction site was greatly affected by
excess reentrained dust caused by material tracked from the
site. Concentrations were 40 to 60 ug/m (geometric mean)
higher at the four sites in the study area than at the
closest sampler locations in the regional network. It did
not appear that much of this impact was from fugitive dust
emitted directly from the construction site.
Average concentrations at each of the sites during the
periods with different mud carryout controls are summarized
in Table 3-5. Comparison of particulate levels during the
35
-------
Table 3-5. PARTICULATE CONCENTRATIONS ASSOCIATED WITH
DIFFERENT MUD CARRYOUT CONTROL MEASURES
Control measure
Time period
No. of samples
Sampling site
B9b
BIO
Bll
B12
Ranking by air
quality
Average of 15
regional net-
work sites
Geometric mean concentration,
Minimal
cleaning
(no control)
09/12-10/14
15
108.7
132.3
127.5
124.4
3
84.1
Manual
cleaning at
intervals
10/15-10/24
4
112.0
118.4
119.9
106.8
2
78.5
Municipal
cleaning
10/25-10/31
4
118.8
135.7
191.8
142.9
4
74.9
ug/m
Continuous
manual
cleaning
11/01-12/10
21a
108.1
119.8
120.0
104.6
1
89.6
Only 7 samples for site BIO.
Directional sampler for southerly wind direction
(upwind of access street); concentrations may not
agree well with other data.
36
-------
different control periods revealed that concentrations were
lowest when the access area had intensive manual cleaning,
next lowest when manual cleaning was done intermittently,
third lowest during the no control period, and highest when
municipal street cleaning equipment was used along the
access street. Adjustment of these concentrations to account
for differences in regional air quality during the four
periods, as explained in the sections describing the Kansas
City and Cincinnati street cleaning studies, would not
change the order of the periods. In fact, it would increase
the differences in concentrations associated with the dif-
ferent control alternatives. It has been estimated from the
data in Table 3-5 that 10 to 20 ug/m reduction in concen-
trations at the samplers, or about 30 percent of the total
impact from mud carryout at the construction site, can be
achieved by immediate cleaning of the access street (s) after
material is deposited on it. It appears that about half
that reduction can be obtained by intermittent (i.e., daily)
manual cleaning of the access street. The results from the
municipal street cleaning period indicate that flushing or
watering in the access area may be counterproductive.
The tracer study showed that the distance of tracking
is very irregular and seems to depend on the relative traffic
volumes on different streets carrying traffic away from the
mud carryout source. The major impact area for reentrained
dust is probably confined to a distance of 1000 ft along the
primary access streets.
37
-------
ASSESSMENT OF OTHER STUDIES
New York-New Jersey
Study Design - The New Jersey State Bureau of Air Pollution
Control recently conducted a two year field study to char-
acterize the particulate material resuspended from road
surfaces within a large city. Sampling was conducted at two
different sites—nine months in Newark and fifteen months in
Brooklyn—both in commercial/business districts.
High volume samples were taken for sequential four-hour
periods (six samples per day) on most days during the two
years at locations four meters above street level and three
meters back from the curb. For the four-hour periods
concurrent with high volume samples, measurements of pre-
cipitation, traffic volume, average wind speed, resultant
wind direction, temperature, and average carbon monoxide
concentration were also taken.
On 15 days during the study, the streets were cleaned
by flushing at 5:00 a.m. The flushing was usually done on
two to three consecutive days, thus creating intensive
short-term cleaning periods with extensive noncleaning
periods. (Days of street cleaning in Newark were November
6, 7, and 19, 1974 and in Brooklyn were September 2, 3, 4,
5, 10, and 11, November 24, 25, 26, and 28, December 2 and
3, 1975).
Analysis of Air Quality Data - Average concentrations for
each of the four-hour periods on days with flushing were
compared with average concentrations for the same time
periods on days with no cleaning. No attempt was made to
comprehensively account for differences in meteorological
conditions on days with flushing compared with the other
days, but it was noted that no precipitation occurred on any
of the 15 days with flushing and that wind speeds were
38
-------
unusually high (18 mph) for only one of the four-hour
periods on these days. Therefore, it was concluded that
concentrations on flushing days were not biased downwind by
unusual meteorological conditions.
A similar comparison of average concentrations on the
first day after flushing (when this day was not another
flushing day) with average concentrations on noncleaning
days was attempted. Data were available for only three such
days, one of which was a Saturday. Therefore, this com-
parison was felt to be of limited value.
Results and Conclusions - As shown in Figure 3-1, particu-
late concentrations near the streets are consistently lower
on days with flushing or days after flushing than on non-
cleaning days. The average reduction on days with flushing
was 16.1 ug/m , or 15.7 percent.
The limited data from days after flushing would indicate
even greater reduction in particulate concentrations than on
the days of flushing. However, this anomaly was attributed
to inadequate sample size.
The diurnal variation in particulate concentrations is
closely related to traffic patterns on the nearest street (r
= 0.92), as shown in Figure 3-1. This relationship between
ambient concentrations and traffic volume has been found in
all field studies where the two variables have been measured
simultaneously for short-term periods.
This study, performed under relatively controlled con-
ditions and with detailed data for short-term time periods,
shows a definite reduction in ambient concentrations as a
result of flushing. Reductions on the day of flushing are
of the same magnitude as those observed in the Cincinnati
street cleaning study (16.1 ug/m in New York-New Jersey vs
15.8 ug/m reduction in Cincinnati).
39
-------
3000
2000
0)
3
Siooo
o
ISO
t»
3
.1
0
O
50
I I I T
I r
I I
FIRST DAY AFTER FLUSHING 13)
I I I
8
8
o
o
s
o
§
Hours
Figure 3-1. Comparison of particulate concentrations on
days with flushing with average concentrations, New
York-New Jersey.
40
-------
Kansas City, Kansas
Street Cleaning Program - Streets in Kansas City, Kansas
have been cleaned with broom sweepers in past years. Fre-
quency of cleaning was daily in the central business district,
once a week on major streets and in industrial areas, and
once a month in most residential areas. Beginning in early
1976, the city initiated a street flushing program in which
the streets were flushed on the same schedule on which they
were cleaned with broom sweepers, with the flushing opera-
tion following the sweeping. The new program was actively
implemented during the spring and summer months but was
curtailed in the fall as concern over the municipal budget
became a major election issue.
Analysis of Air Quality Data - The change in the street
cleaning program should be detectable on ambient concen-
trations in Kansas City, Kansas since it was implemented
throughout the city. There are three sampling stations in
the regional network that have been operating without
change in location in Kansas City, Kansas for the past four
years. The 1976 concentrations by quarter at each of these
sites were compared to the averages for the previous three
years (when no flushing was done) to determine the improve-
ment in air quality associated with the flushing. The
changes in air quality in 1976 in Kansas City, Kansas were
compared with corresponding changes in concentration at
other urban and suburban locations in the metropolitan area
(where no changes in street cleaning had occurred during
1976) .
Results and Conclusions - Air quality in Kansas City, Kansas
improved over previous years during the first and second
quarters of 1976. In the third quarter, concentrations at
41
-------
the three sites were the same as in previous years. Par-
ticulate concentrations were substantially higher during the
fourth quarter, as summarized below:
No. of concentration in 1976 from _
sampling previous 3 years' average, ug/m
Reduction in particulate
concentration in 1976 fro
/ious 3 years' average, u
Location sites 1 qtr 2 qtr 3 qtr 4 qtr
Kansas City, 3 5-3 0 -14
Kansas
Remainder of 12 7 9 -3 -20
metro area
Particulate concentrations in the remainder of the Kansas
City metropolitan area were also lower than normal during
the first two quarters and showed essentially the same de-
viation from previous years' readings as the Kansas City,
Kansas sites for all four quarters. Therefore, street
flushing in the subregional area where it was performed does
not appear to have produced an obvious change in air quality
relative to the rest of the region.
42
-------
Charlotte, North Carolina
Street Cleaning Program - Starting in 1973, Charlotte began
phasing out mechanical broom sweepers and began replacing
them with regenerative air units, which are similar in
operation to vacuum sweepers. By 1976, this replacement
program was complete and all downtown and arterial streets
are now cleaned with regenerative air sweepers and then
flushed. The frequency of cleaning is daily in the downtown
areas and twice per week on all arterials. The downtown
cleaning is also supplemented by manual sweeping during the
day.
Analysis of Air Quality Data - The improvement in the street
cleaning program should be detectable in the trend of air
quality at urban sites in Charlotte. Under this assumption,
air quality data were obtained for urban sites in Charlotte
(Mecklenburg County) for the years 1971 through 1976. For
purposes of comparison, data were also obtained for suburban
and rural sites in Mecklenburg County where the impact from
reentrained dust and street cleaning would be minimal. The
latter sites should not show a change as a result of the
improved street cleaning program.
Results and Conclusions - The trends in air quality present-
ed in Figure 3-2 show that air quality at the urban sites
has improved significantly during the period of the change
in the street cleaning program, whether all high volume
sampler sites or only selected sites considered to be more
representative of such regional air quality are used.
However, a comparable improvement in air quality has also
occurred at the suburban and rural sites in Mecklenberg
County during this period. Further investigation of the
observed trends revealed that point source emissions in
43
-------
0)
e
u
•H
4J
-------
Mecklenberg County were reduced from 35,286 ton/yr in 1973
to 5,416 ton/yr in 1975. This reduction in point source
emissions could certainly explain some of the uniform re-
duction in ambient particulate concentrations throughout the
county.
Due to the parallel decreases in particulate concen-
trations in Charlotte at urban and rural sites, the sub-
stantial improvement in air quality cannot be attributed to
the change in street cleaning techniques. It is not likely
that the effect on air quality of the street cleaning im-
provements would be uniform over the entire metropolitan
area.
45
-------
Chicago, Illinois
Study Design - The Departments of Environmental Control and
Streets and Sanitation in the City of Chicago conducted a
one-month study in May 1974 to evaluate the effect of street
sweeping frequency on particulate air quality. Four high
volume samplers were located in different blocks of State
Street in the downtown area. All were mounted on utility
poles just back from the curb at 14 ft height. A control
site less affected by vehicle-related emissions was operated
about four blocks away in Grant Park. The samples were
taken for 24 hours on five days a week (weekdays).
During the study period, State Street was swept daily
north of Monroe (in front of sites 1 and 2) and every other
day south of Monroe (in front of sites 3 and 4). All other
aspects of the street cleaning were the same for both sec-
tions of the street. Traffic volumes are fairly uniform
along State Street, so it was assumed that concentrations
would be equal at the four sites in the absence of the dif-
ferent street cleaning frequencies.
Analysis of Air Quality Data - Average concentrations at
sites 1 and 2 were compared with those for sites 3 and 4 to
determine whether daily sweeping reduced particulate concen-
trations. Also, the impact of vehicle-related emissions at
sites 1 through 4 was estimated by comparison of their
average concentrations with those from site 5. The sampling
data are summarized below:
Site
1
2
3
4
5
Particulate
Location Average
State
State
State
State
Grant
and Randolph
and Madison
and bet. Adams/
Monroe
and Jackson
Park
111
107
105
104
52
a 3
concentration, ug/m
High Low
192
163
173
170
90
17
62
56
64
15
24 days of data between 4-29-74 and 5-31-74.
46
-------
The relative concentrations on cleaning and noncleaning days
at sites 3 and 4 could not be determined because the daily
data were not available and there was no record of the exact
days that the sweeping was done on the southern portion of
State Street.
Results and Conclusions - The impact of vehicle-related
emissions on concentrations along State Street was estimated
to be at least 55 ug/m (arithmetic mean), this value assum-
ing no vehicular contribution to concentrations at the Grant
Park site. The great importance of vehicle-related emissions
on measured concentrations was confirmed by microscopic ex-
amination of several of the samples which all showed the
following percent compositions by weight:
calcite >25 auto exhaust 5-25
quartz .5-5 tire rubber .5-5
clay and <5 asphalt <.5
humus ,, ., , „
all other <14
The microscopists concluded that most of the calcite was
reentrained from street surfaces or from pavement wear.
At both sites with daily sweeping, concentrations were
slightly higher than at the sites with sweeping every second
day. Although concentrations at the four sites were so
close that statistically they might not be different than
zero (this could not be checked without the raw data), there
was certainly no air quality benefit from daily sweeping
versus sweeping every second day.
47
-------
Twin Falls, Idaho
Study Design - During the summer of 1973, GCA Corporation
conducted an intensive field study of suspended particulate
18
in Twin Palls, Idaho. High volume samplers were operated
for 91 consecutive days at nine different locations in the
Twin Falls area—three in the downtown area, four in sub-
urban locations, and two in rural areas. Meteorological
data on an hourly basis were collected for the entire three-
month period. There are no major particulate point sources
in Twin Falls, but average concentrations above the primary
standard were recorded at five of the nine sites.
Streets in the downtown area were flushed and swept
weekly on Mondays (and Tuesdays if not completed on Monday).
The broom sweeper was preceded by the flusher. Suburban
areas in Twin Falls were cleaned on an irregular schedule
and there was no street cleaning within a mile of either of
the rural sampling sites.
Analysis of Air Quality Data - Since streets were cleaned
weekly during the entire sampling period, the effect of
street cleaning on air quality could best be determined by
observing the variation in concentrations with time since
cleaning. Average concentrations by day of the week for the
three downtown sites (weekly cleaning) and the two rural
sites (no cleaning) are shown in Figure 3-3. Concentrations
at suburban sites (irregular cleaning) were not investigated
because they could not be equated with days since cleaning.
Results and Conclusions - Concentrations at the three down-
town sites on Mondays, the day the streets were cleaned,
were lower than on any other week day. Lower concentrations
on Saturdays and Sundays can be attributed to less traffic
on these days. The weekly variation in concentrations at
48
-------
Th
Sa
Su
Q)
4J
•H
W
-P
(0
C
o
•H
•P
(0
M
-P
c
0)
o
c
o
u
0)
Cn
(0
^
(U
O
4-1
O
•H
1.1
0.9
oa
12
1.1
10
0.9
08
Downtown sites
M
0
Rural sites
234
Days since cleaning
So
5
Su
6
Figure 3-3. Average particulate concentrations by day of
week in Twin Falls, Idaho.
49
-------
the two rural sites was essentially the same as at the
downtown sites, although the weekend effect was less pro-
nounced.
The shape of the curve for the downtown sites matches
the curves of days since flushing obtained in the Cincinnati
street cleaning study, but the curves for the rural sites
indicate that regional concentrations are low on Mondays
even if there is no street cleaning nearby. Therefore, one
of three conclusions can be drawn: the street cleaning was
not affecting particulate concentrations significantly; it
was affecting all sites near Twin Falls uniformly; or some
other factor was influencing weekly variation at the rural
sites in a manner similar to the street cleaning.
The GCA report on the Twin Falls particulate study con-
cluded that street cleaning was an ineffective control
measure for reducing street dust in Twin Falls.
50
-------
Seattle, Washington
Control Program - A 1972 study of Seattle's Duwamish Valley,
done by the Puget Sound Air Pollution Control Agency,
indicated that 5 to 18 ug/m were added to particulate con-
centrations by mud carryout from unpaved roads and parking
2
lots onto paved roads and its subsequent reentrainment.
The report recommended that mud carryout be reduced by
paving or oiling of roads and parking areas as the most
cost-effective control measure in this highly industrialized
area.
As a result of these recommendations, the city ini-
tiated a $2 million per year three-year program in 1974 to
strip pave or seal coat roads and traffic areas in all parts
of Seattle. A portion of this money was used on unpaved
roads in the Duwamish Valley. Streets which would have
required heavier paving because of truck traffic have been
oiled to reduce dust and mud carryout. Approximately $40,000
has been spent for oiling in the Duwamish Valley. By the
end of 1976, the initial program for control of mud carryout
in the Duwamish Valley was completed.
Analysis of Air Quality Data - The change in particulate
concentrations in the Duwamish Valley relative to the rest
of the Seattle (King County) area over the time period in
which the mud carryout control program was implemented
should provide a rough assessment of the effectiveness of
the program. Point source emission reductions in the
Duwamish Valley were nearly complete by the end of 1973.
There were seven high volume sampler sites located in the
Duwamish Valley during the period 1973 to 1976 and eleven
sites in other parts of King County, including one rural
location designated as a background site. Monthly average
concentrations for the four year period are plotted in
Figure 3-4.
51
-------
en
NJ
7 sites in
Duwamish Valley
10 sites in
King County
1 FMAMJ J ASONDJ
1973
0 J
0 J
1974
1975
1976
Figure 3-4. Change in particulate concentrations in the Seattle area.
-------
Results and Conclusions - The paving program in the Duwamish
Valley appears to have reduced ambient concentrations in
this area relative to the rest of the metropolitan area.
Most of the change in year-to-year averages at the other
sites in King County can be attributed to changes in back-
ground concentration (see annual averages in Figure 3-4).
In relation to either background concentrations or ambient
concentrations in the rest of King County, there has been an
improvement in the air quality of Duwamish Valley in the
range of 12 to 22 ug/m (arithmetic) since 1974, the first
year of the program. This is greater than the total impact
from traffic-related emissions originally estimated by the
Puget Sound agency but agrees quite well with a later esti-
mate made in 1974 that motor vehicles and reentrained dust
constituted about 39 percent of the ambient particulate in
3 12
the Duwamish Valley (or about 32 ug/m ). The improvement
due to reduced reentrainment cannot be isolated from that of
reduced unpaved road emissions. Both are thought to be
significant.
53
-------
UNTESTED MEASURES
Several potential measures were not evaluated in field
tests. Two of these—reductions in vehicle-miles traveled
(VMT) and modification of street sanding procedures—are
thought to have enough potential to warrant some estimate of
their effectiveness.
Reductions in VMT
According to data presented in Chapter 1 and Appendix
A, ambient concentrations are directly related to traffic
volumes on the nearest street and/or to traffic densities in
the surrounding area. Therefore, reductions in these traffic
volumes should result in proportional reductions in concen-
trations caused by reentrained dust. For example, a 20
percent reduction in VMT in an area where 45 ug/m is con-
tributed by traffic-related emissions should reduce average
concentrations by 9 ug/m .
Rerouting traffic to streets that are only a few blocks
distant from its present path would probably just shift the
location of the highest particulate concentrations slightly.
The transportation controls with the most potential for
achieving VMT reductions in a limited area with excessive
particulate concentrations are well-marked alternative
routes (several blocks away) which tend to disperse traffic
densities in the urban area and direct traffic restrictions
such as limited access zones or designated carpool lanes.
Modification of Street Sanding Procedures
In cities where sanding is used on streets for snow and
ice control, modifications can be made in the sanding opera-
tions to reduce air quality impact without increasing the
hazard of vehicle accidents. Some of these modifications
are:
54
-------
replace the sand with salt or a salt/sand mix;
plow streets instead of sanding;
clean the sanding material from streets as
soon as possible after each storm;
apply material only at intersections and on
hills and curves (reduce the amount applied);
use sand that has been washed and sized.
It is not possible to quantify the air quality impact of
each of these changes or their combinations.
In many cities which sand streets, seasonal particulate
concentrations are significantly higher during the first
quarter, or winter months. In Denver, particulate concen-
trations are at least 25 ug/m higher during the first
19
quarter than in the remaining three quarters. Much of
this increase during the winter is attributed to street
sanding. High 24-hour concentrations (in violation of the
primary standard) in Montana cities were attributed largely
to winter sanding operations. In contrast, a study in
Detroit, where streets are salted instead of sanded, showed
that suspended particulate in the winter averaged 6.1 per-
cent salt (sodium chloride) compared to a summer background
21
of 0.55 percent salt. Therefore, salting of streets
apparently increased average winter concentrations about 5
ug/m . In the Kansas City and Cincinnati street cleaning
studies, the effect of salting could not be identified on
days after snowfalls even though salt residue was visible on
streets in the study areas. Replacement of sand with salt
is therefore estimated to reduce the effect of this reen-
trained dust source by more than 50 percent.
55
-------
ADVERSE ENVIRONMENTAL EFFECTS FROM CONTROL MEASURES
Recently, numerous studies have called attention to a
relationship between dirt loadings and storm water pollu-
tion. This is a cause of concern to public officials who
have seen large investments in water treatment plants and
collection systems partially nullified by this type of
pollution. Sartor and Boyd found that runoff from street
surfaces is generally highly contaminated and contributes
considerably more pollutional load than sanitary sewage
during the same period of time. Shaheen concurred that
storm water runoff is frequently a significant portion of
22
the total annual water pollution. Thus, any reduction in
material on street surfaces has the dual advantage of im-
proving air quality and water quality.
Some of the control measures cited previously can serve
this dual purpose. For example, wherever a measure controls
soil erosion or trackout, the amount of material deposition
on streets is decreased and the impact of subsequent reen-
trainment is thereby reduced. Concurrently, the decrease of
material on streets reduces contamination of the urban water
runoff.
However, in some instances, measures that improve air
quality can have a detrimental effect on water quality.
Flushing, which was shown to reduce particulate concentra-
tions, does not remove material from streets but moves it
from traffic lanes to the curb area. In subsequent rains,
the suspended solids loadings in storm water runoff would
not be reduced at all by the streets having been flushed.
In fact, short-term concentrations might be increased be-
cause the material was already concentrated in the gutters.
A combination of flushing followed by a pickup of
material in the gutter, with either broom sweepers or vacuums,
would theoretically provide optimum cleaning for both air
and water pollution control. However, in the two studies
56
-------
reported in this chapter where a combination of flushing and
sweeping was performed (Kansas City, Kansas and Twin Falls),
no significant particulate concentration reductions were
observed.
Based on limited available data, it appears that salt-
ing presents less of an air pollution problem than sanding.
However, sanding (with clean sand) may be better from a
water quality standpoint because of reduced chlorides in
storm water runoff. In addition, the use of salt will
accelerate corrosion of vehicles, reduce the life of highway
structures, and cause damage to vegetation along areas
23
adjacent to the streets. Each of these adverse environ-
mental effects of salting must be balanced against the air
quality effects of sanding. Reducing the amount of either
antiskid material used is desirable from an environmental
standpoint. Use cannot be reduced to the point where acci-
dents might increase, but monitoring for overuse and main-
taining safe driving levels offers one feasible approach to
reducing reentrained dust emissions and suspended solids or
chlorides in runoff.
57
-------
SUMMARY OF CONTROL MEASURE EVALUATIONS
Results of all the field studies and control measure
evaluations are summarized in Table 3-6. None of the con-
ventional street cleaning methods were found to be effective
in every study in which they were investigated. Because of
the mixed results, definite conclusions cannot be drawn
relative to the effectiveness of street cleaning as a con-
trol measure for reentrained dust.
The relationship between cleaning and subsequent emis-
sion rates appears to be complex; emission rates are not
directly related to the percent of street surface loadings
removed from the traffic lanes (just as emission rates were
not found to be closely related to the street surface load-
ings) . Also, the street cleaning studies with streetside
samplers tended to show a positive effect from cleaning
operations, while those using regional network samplers in
general failed to show an impact from street cleaning.
Flushing apparently reduced particulate concentrations
significantly in two of the studies, showed no effect in two
studies, and did not reduce monthly average concentrations
but did reduce concentrations on the days with flushing in
a fifth study. Some of the data indicated that flushing has
an effect very similar to rainfall, with a substantial re-
duction in particulate concentrations on the day of flushing
but little or no residual effect on following days. Weekly
flushing in the two studies where no effect was noted may
not have been often enough to produce a detectable reduction
in concentrations. A reduction of 16 ug/m on the days of
flushing is recommended for estimating the impact of this
street cleaning method in high traffic density areas.
Broom sweeping reduced average concentrations in one
study but appeared to be ineffective in two others. This
cleaning method has two operational limitations that make
its use alone in reducing reentrained dust of doubtful
58
-------
Table 3-6. SUMMARY OF CONTROL MEASURE EVALUATIONS
Study
Control measure
Effectiveness
Comments
Kansas City,
Missouri
flushing
broom sweeping
Cincinnati flushing
Westwood
Construction
site
New York-
New Jersey
Kansas City,
Kansas
Charlotte
Chicago
Twin Falls
Seattle
broom sweeping
vacuum sweeping
vacuum sweeping
manual cleaning
(intermittent)
manual cleaning
(continuous)
municipal cleaning
flushing
flushing added to
existing broom
sweeping
regenerative air
cleaning
daily broom
sweeping
flushing and
broom sweeping
reduction of mud
carryout by paving
8-18 ug/m aver
negligible
negligible
6-20 ug/m aver
negligible
negligible
5-10 ug/nT
10-20 ug/m3
negligible
16 ug/m on day
of flushing
negligible
negligible
cone, on days of.,
flushing 16 ug/m
lower
compared to
broom sweeper
in area around
access streets
in area around
access streets
on subregional
scale
on subregional
scale
negligible com-
pared to every
other day cleaning
negligible
15-35 ug/m'
part of impact
due to reduced
emissions from
unpaved roads,
etc.
59
-------
value—it moves material from the gutter back into the
street for pickup and it is not efficient in removing fine
particles. There were some indications that the optimum
frequency for broom sweeping from an air quality standpoint
might be every two or three days rather than daily. The
application of this cleaning method in the Cincinnati study
was probably near its optimum—6 to 20 ug/m reduction in
average concentrations. It is not possible from available
data to estimate the air quality improvement that could
normally be expected.
None of the three studies which evaluated the air
quality impacts of vacuum or regenerative air sweepers
showed reductions in concentrations associated with their
use. In the Cincinnati study, street loading measurements
showed that vacuum sweeping removed a higher percentage of
material from the streets than broom sweeping, but the
vacuum sweepers always collected less total weight of material
in their hopper over the 13 curb-miles than the broom
sweepers. Dust was often observed escaping from the top of
the vacuum sweeper while it was operating* In the Westwood
study, there was apparently not enough contribution from
reentrained dust to ambient concentrations so that the
improved street cleaning was detectable in reduced concen-
trations.
Both of the mud carryout control studies showed signi-
ficant reductions in particulate concentrations. At the
construction site study, the impact from mud carryout was
reduced by 30 percent by intensive manual cleaning near the
access point. Even though control of such sources has an
effect only within the area of impact of the construction
site, about a 1000 ft radius around the site, that effect is
in the area where locally high concentrations occur.
Data from Seattle indicate that reductions similar to
those obtained at the construction site study can be achieved
60
-------
over a subregional area through a comprehensive mud carryout
control program for smaller sources such as unpaved parking
lots, roads, and shoulders. Annual average concentrations
3 2
were reduced 12 to 22 ug/m over a 16 mi industrial area by
required strip paving or oiling to eliminate trackout.
Although no field studies were conducted to evaluate
traffic volume reductions or modified street sanding pro-
cedures , some data are available which indirectly show that
both of these potential control measures can be effective if
they are applicable in a particular area. The air quality
improvement that can be obtained by modifying street sanding
procedures is, of course, highly dependent on the climate
and existing street sanding practices in the area.
Because of the large number of measures being evaluated
and the conflicting results on street cleaning as a control
measure, the data presented in this chapter provide only a
preliminary assessment of these measures rather than a •'"'
conclusive examination. There are still questions to be
answered with pilot studies before regional reentrained dust
control programs can be implemented.
61
-------
4. COST DATA
Cost data were compiled for all the control measures
except reduction in traffic volumes. Costs for street
cleaning activities were compiled separately for operating
and capital costs. The primary source of operating and
maintenance cost data was the American Public Works Associa-
tion (APWA) 1975 Survey of Street Cleaning Practices (unpub-
24
lished). Other operating data for specific cities were
compiled from the literature and by contact with public
works officials. Capital equipment costs for selected
equipment were obtained directly from vendors.
Costs relevant to enforcement or inspection activities
are not included.
STREET CLEANING COSTS
Capital Costs
Capital costs of street cleaning equipment are summar-
ized in Table 4-1. For the major manufacturers represented,
the cost of municipal sized vacuum units ranges from $43,000
3 3
(7 yd model) to $56,000 (16 yd model). A smaller model (3
yd ) for shopping centers and similar uses sells for about
$12,500. Broom sweepers range in cost from approximately
$20,000 (3-4 yd3 models) to $45,000 (5 yd3 models). The
regenerative air units cost an average of $30,000 (3.5-6 yd
models). Generally, vacuum sweepers are larger in capacity
than broom sweepers and cost approximately 24 to 72 percent
more. However, their cost per cubic yard capacity is less
than that of broom sweepers. Note that these cost data are
62
-------
Table 4-1. SUMMARY OF CAPITAL COSTS FOR
SELECTED STREET CLEANING EQUIPMENT
Manufacturer/
Model
Ecolotec
VACU-Sweep
Elgin
Whirlwind
Central Engr
Vac-all (E-5-16)
Central Engr
Vac-all (E-5-13)
Central Engr
Vac-all (E-10)
Schwarze Indus
Products, Inc.
Supervac
Tymco, Inc.
Model 600
Tymco, Inc.
Model 350
Eintyre
Elgin
Pelican
Elgin .
White Wing
Tennant Co.
92 Power Sweeper
Athey Products
TE-4
Athey Products
TE-3
FMC Corp.
Wayne Model 12
PMC Corp.
Wayne Model 984
FMC Corp.
Wayne Model 973
Type eauipment
Vacuum with 1 gutter
t 1 wide sweep broom
Vacuum with 1 gutter
& 1 wide sweep broom
Vacuum w/optional
gutter broom
Same
Vacuum-mounted on 3/4
or 1 ton truck, gutter
broom optional
Regenerative air with
1 gutter broom
Regenerative air with
1 gutter broom
Plusher
Broom sweeper with 3
wheels, 1 or 2 gutter
brooms, 1 wide sweep
broom
Broom sweeper with 3
wheels, 1 gutter broom,
1 wide sweep broom
Broom sweeper with 3
wheels, 1 gutter broom,
1 wide sweep broom
Broom sweeper with 4
wheels, 1 or 2 gutter
and 1 wide sweep broom
Broom sweeper with 4
wheels, 1 or 2 gutter
and 1 wide sweep broom
Broom sweeper with 4
wheels, 4 gutter and 1
wide sweep brooms
Broom sweeper with 3
wheels, 2 gutter and 1
wide sweep brooms
Sane
Sweep Sweep
speed, path,
cfm in.
6,000 84
10,000 93
12,000 57a
12,000 57a
12,000 57a.
108
108
96 to
120
96 to
120
66
90 to
120
90 to
120
120
96
96
Capacity,
Ib s, ydJ
13,000
7
10,000
7.4
68,000
16
13
50,000
10
3,000
3
6
3.5
n.a.
9,000
3
9,000
4
1,800
3
4
3
12,000
5
4
3
Cost,
1976 S
Base-43,000
Dual-49,000
45,000 to
55,000
56,000
53,000
50,000
12,500
32,000
28,000
27,300
25,000 to
38,000
20,000 to
28,000
15,000
35,000
34,000
45,000
42,500
40,000
Cost,
S/vd^ cpy
6,143
7,000
6,081
7,432
3,500
4,077
5,000
4,167
5,333
8,000
n.a.
8,333
12,667
5,000
7,000
5,000
8,750
11,333
9,000
10,625
13,333
to
to
to
estimated
Pelican model, without lifting hopper
63
-------
manufacturers' quoted prices for normally-equipped machines
and could vary considerably with the addition of optional
equipment. Also, prices requested by potential purchasers
through competitive bidding may be different than these
quotes.
Operating Costs
Annual operating costs for street cleaning summarized
from the 1975 APWA study are presented in Table 4-2. For
the items tabulated, reported costs were aggregated for all
equipment types in use by the survey respondents.
The costs for specific aspects of street cleaning pro-
grams, such as sign posting or water for flushing, were ex-
pected to vary over a wide range because they are total bud-
get costs for all sizes of cities. However, the cost data
normalized for such parameters as tons of material collected,
curb-miles cleaned, or population served should be much more
consistent. The enormous range of reported costs per curb-
mile cleaned has led to the conclusion that several survey
respondents misinterpreted the questionnaire and reported
total annual operating costs per curb-mile of streets in
their jurisdiction rather than the cost to clean one curb-
mile » Such errors would bias the median value upward slight-
ly and make the calculated mean value meaningless.
Table 4-3 presents the same Operating cost data dis-
aggregated by city size. The same problem with reported
costs.per curb-mile arises here, and the data indicate that
the misinterpretations were made by respondents from the
smaller cities. The costs per cubic yard of material col-
lected and per person do not reveal any definite relation-
ship between costs and city size.
The APWA operating cost data are tabulated by climate
zone in Table 4-4. With the exception of zone 2, the
Southwestern states, street cleaning costs per cubic yard
64
-------
Table 4-2. SUMMARY OF ANNUAL OPERATING COSTS FOR STREET CLEANING
en
en
Operating costs
$/ton
$/yd3
$/curb-mile
$/person
% possible
data
represented
33
58
69
81
Mean
20.79
17.55
86.61
1.33
Median
12.20
13.24
7.00
1.18
Standard
deviation
20.35
15.40
165.85
0.78
Data
range
10 percentiles 90
2.25
3.37
2.96
0.52
53.57
38.00
300.00
2.51
Specific
Sign posting
Street cleaning
Transport
Dump fee
Water cost (flushers)
Total
18
73
33
28
32
5,335
197,718
23,735
8,841
6,376
206,980
540
58,520
5,864
142
300
69,800
17,202
664,232
51,183
32,960
27,748
675,815
0
15,000
1,200
0
0
17,300
9,750
283,735
51,000
11,339
3,990
328,221
Generally includes these items: equipment operation and maintenance, labor, fuel,
depreciation, other related costs.
Source: APWA 1975 Survey of Street Cleaning Practices.
-------
Table 4-3. OPERATING COSTS BY CITY SIZE
Cost
parameter
$/yd3
mean
count
. s.d.
$/curb-mi
mean
count
s . d.
$/person
mean
count
s.d.
0-10
7.43
2
7.68
61.50
2
82.73
1.32
2
0.91
Population class,
10-25 25-50 50-100
16.58
36
14.82
103.07
43
175.09
1.21
57
0.78
19.57
28
17.34
99.84
34
192.92
1.45
35
0.71
17.50
14
16.90
66.87
16
129.05
1.31
18
0.74
x 103
100-250
14.94
4
10.92
7.08
5
4.30
1.74
5
1.11
250-500
28.22
2
1.41
9.02
3
5.43
1.51
3
0.87
>500
11.70
2
2.40
5.80
2
1.70
1.62
3
0.97
Table 4-4. OPERATING COSTS BY CLIMATE ZONE
Cost
parameter
$/yd3
mean
count
s.d.
$/curb-mi
mean
count
s.d.
$/person
mean
count
s.d.
1
California
dry summer ,
mild wet
winter
19.
16.
63.
136.
1.
0.
06
22
33
92
33
60
47
33
73
Climate zone
2 3
Southeast
hot wet sum-
Southwest mer, mild
arid, hot winter
7.
5.
5.
4.
0.
0.
38
7
28
23
7
45
64
7
38
17.
13.
36.
103.
1.
0.
79
17
58
85
22
18
59
25
86
• 4
Midcontinent
hot summer ,
short winter
19.
17.
195.
238.
1.
0.
52
23
54
14
20
44
11
32
77
5
Northern
warm sum-
mer, long
winter
16.
15.
97.
165.
1.
0.
94
19
40
15
23
63
37
26
67
Source: APWA 1975 Survey of Street Cleaning Practices.
66
-------
collected or per person do not appear to be closely related
to climate. Variations in cost within each climate zone,
as measured by the standard deviation, are much greater than
the variations in mean values for different zones. It is
surprising that street cleaning costs are not climate depen-
dent. Apparently, the extra costs of spring cleanups in
northern climates are offset by shorter cleaning seasons.
Table 4-5 presents operating costs for specific types
of street cleaning equipment. These data were obtained by
direct contact with street department officials and from the
literature. Equipment-related costs were most often reported
on the basis of hours of equipment use. To derive costs on
a $/curb-mile basis where this information was not reported
directly, it was assumed that all types of equipment cleaned
about 20 curb-miles per eight hour shift. Labor rates vary
somewhat in different parts of the country but not to the
extent of the variance in equipment-related costs. In order
to permit greater ease of comparison of total operating
costs for the three equipment types, the data are summarized
below:
Total operating cost, $/curb-mi
Vacuums Broom
City & R. air sweepers Flushers
^ a
Cincinnati
Kansas City
Charlotte
Milwaukee
Memphis
Beaumont
Love ' s Park
Sparks
Average
8.94
-
4.06
-
8.20
4.50
3.75
1.56
5.17
8.24
13.57
-
9.41
10.00
8.75
5.75
1.94
8.24
4.03
-
3.51
—
—
—
—
-
3.77
a 1976 $
67
-------
Tabe 4-5. OPERATING COSTS FOR SELECTED CITIES
City/equipment Equipment
type $/hr $/curb-mi
Cincinnati, OH
Vacuums 16.00 6.40?
Broom 14.40 5,76
sweepers
Flushers 3.85 1.56
Kansas City, MO
Broom - 11.34
sweepers
Charlotte, NC
Regen. air - 2.06
Flushers -" 1.51
Beaumont , TX
Vacuums
Broom - >-
sweepers
Love's Park, IL
Vacuums <- -
Broom -- *•
sweepers
Sparks, NV
Vacuums 0.64 0.13
Broom 1.98 0.55
sweepers
Milwaukee, WI
Broom *• -?
sweepers
Memphis , TN
Vacuums 15.09 6.16
Broom 13.40 7.28
sweepers
Operator Total3
$/hr $/curb-mi $/hr $/curb-mi
6.34 2.54b 22.34
6.23 2.49D 20.63
6.23 2.49b 10.08
2.23
5.00P 2.00b
5.QOC 2.00°
11.26
21.87
7.88
12.08
5.00 1.38 5.64
5.00 1.38 6.98
23.48
5.00° 2.04 20.09
5.00C 2.72 18.40
8.94b
8,25b
4.03b
13.57
4.06
3.51
4.50b
8.75b
3.75
5.75
1.56
1.94
9.41
8.20
10.00
3 All costs in 1976 dollars
Assumes an average cleaning rate of 2.5 curb-mi/hr
(8 hr shift)
c Assumed labor rate
68
-------
Several conclusions that may be drawn are:
broom sweepers are generally more expensive
to operate on a curb-mile basis than vacuums
or flushers;
flushers incur the lowest operating costs,
probably due to less required maintenance;
there are no major differences in street
cleaning costs based on either city size or
geographic location;
differences in accounting systems and methods
of cost reporting exist among jurisdictions;
these could explain some of the variation in
cost between cities (for example, the most
commonly accounted for cost items are equip-
ment operation, maintenance, parts, downtime,
fuel and labor, but capital depreciation may
or may not be included);
the average broom sweeping cost per curb-mile
determined by contacts with a few public works
departments, $8.24, agrees well with the median
cost of $7.00 reported in the 1975 APWA survey.
MODIFIED SNOW CONTROL COSTS
Basic cost data for snow control procedures such as
sanding, salting, plowing, and cleanup are presented in this
section to permit estimation of the extra costs of modifying
a city's snow control program. Information specific to the
city is also needed before the total costs of any program
modification can be calculated: miles of road to be deiced,
number of snowstorms per year, and current practices.
Some comprehensive cost-benefit analyses have been per-
formed to determine whether salt or sand is better antiskid
O O O C 1 C
material for use on public roads. ' ' These analyses
considered antiskid material, application, corrosion, water
pollution, and vegetation damage as cost items and traffic
safety, public health, time savings, and reduced fuel use as
benefits. Air quality was not considered. Results of these
studies have varied widely, depending on the input data used
69
-------
and assumptions made. Therefore, only the material and
application costs (which are minor compared to the indirect
costs and benefits cited) are presented herein.
27-32
Costs per ton for salt and sand are:
Average
Salt $17
Sand 5
Application costs, which include equipment, labor, gasoline,
supervision, and garage support, are the same per ton of
28 33
salt or sand applied—about $3.00. ' However, since sand
must be applied at higher rates than salt per mile of road,
it costs more per mile to apply. For example, with average
27
application rates of 500 Ib/mi and 800 Ib/mi for salt and
sand, respectively, the cost per mile would be $0.75 and
$1.20.
The limited data on plowing indicate that plowing a
29
mile of road costs approximately $24. If snow removal by
truck is required, the additional costs are about $315 per
29
mile cleared.
The costs for removing antiskid material on a regular
basis (after each application) would be the same as for
regular street cleaning. Therefore, data from the preceding
section can be used. Spring cleanup has been included as a
major cost item in several published articles on snow and
ice control. In cases where salt and sand are compared, it
is the concensus that cleanup is twice as expensive when
sand is used as when salt alone is used. '
The application of less material as a control measure
for reentrained dust would obviously have a negative cost
.and show to be the most cost-effective measure.
70
-------
MUD CARRYOUT CONTROL COSTS
Construction activity control measures for which cost
data were available include wheel washing at the site,
manual or machine street cleaning of the access street by
the contractor/city/private firm, truck covering, and
covering traffic routes on site with gravel. Costs were
also estimated for the paving and oiling of unpaved parking
lots. These data are summarized in Table 4-6.
Most large construction sites would probably require
implementation of more than one control measure. It is
estimated that one construction worker stationed near the
entrance could keep the area cleaned continuously with half
his available time, with the remainder for traffic control
and other assignments. One private street cleaning firm
estimates a labor savings to the contractor of $15-20 per
hour if they perform the cleanup job. They use flushers
exclusively and charge $26 per hour per truck with a $39
minimum charge. In Chicago, where mud carryout regulations
for construction activities are rigidly enforced, the clean-^
ing costs to the contractor are written into the total job
budget as a matter of routine and are not easily identified.
Truck covering is regarded as an ancillary control
technique and is probably the least costly of all the avail-
able alternatives. It should, however, be considered in
combination with one or more of the above measures. Cover-
ing traffic areas on the construction site with coarse
gravel may be a cost-effective measure depending on the size
of the areas.
Paving, oiling, and seal coating costs for parking lots
and other exposed areas have been estimated to cost from
2
$0.07-0.50 per ft of area covered or paved.
71
-------
Table 4-6. COSTS OF MUD CARRYOUT CONTROL MEASURES
Control Capital Operating
measure Cost, 1976 $ Cost basis Cost, 1976 $ Cost basis
For construction sites
Wheel
washers
Manual
cleaning
Cover w/
gravel
5,000-7,500 equipment $3/truck
type plus
installation
$5-12/hr£
Machine 12,500-56,000 cleaning $6-22/hr
cleaning machine
type
2
Truck 0.04-0.60/ft tarpaulin; $3/load
covering $/ft2 trk-
bed
0.12/ft'
area size
lost time
labor
labor/
equipment
lost time
For other carryout sources
2C
Parking lot 0.20-0.50/ft lot size
paving
2d
Oil/Seal Q.07-0.13/ft lot/area
coat size
Total cost to contractor; range of 10 small/large cities
(median = $7.68; mean = $8.37).
Table 4-1.
Cost varies with weight bearing capacity of paving laid.
Oil = least cost.
Table 4-5.
72
-------
COST-EFFECTIVENESS OF INDIVIDUAL MEASURES
Using the summarized data on control measure effective-
ness from Table 3-6, preliminary estimates of cost-effective-
ness for promising measures have been made to provide some
guidance in selection of measures. Control measures which
failed to demonstrate a significant reduction in particulate
concentrations have not been considered since their cal-
culated cost-effectiveness would be infinite. The five
measures reviewed were flushing, combined flushing/sweeping,
modification of street sanding procedures, control of mud
carryout from construction sites, and control of mud carryout
in a large industrial area.
Cost estimates have been developed for hypothetical
urban areas exceeding the annual particulate air quality
standards.
Flushing
The hypothetical area in which the annual particulate
air quality standards are being exceeded is three mi by five
mi in size, with approximately 100 mi of arterial streets
(200 curb-mi). The cost of flushing these streets once
would be $754, based on an average flushing cost of $3.77
per curb-mi. Flushing three days per week, eight months per
year would cost $78,400 and is estimated to lower concentra-
tions in the non-attainment area by 16 ug/m on the day of
flushing, based on data from the Cincinnati and New York-New
Jersey field studies. This would reduce the annual average
38 3
concentration by (16)(j)(j^) =4.6 ug/m . Similarly,
flushing five days per week for eight months per year would
cost $130,700 and is estimated to reduce particulate concen-
trations by 7.6 ug/m .
73
-------
Combined Flushing/Sweeping
For the same non-attainment area, the cost per curb-mi
of flushing followed by sweeping would be the sum of the
individual costs for the two cleaning operations, or about
$12.01 per curb-mi. The cost for cleaning the 200 curb-mi
once would be $2402, and the cost of a scheduled three day
per week cleaning program for eight months per year would be
$249,800.
No direct information was obtained in the field studies
on the air quality improvement associated with combined
flushing/sweeping. The Kansas City, Kansas data showed no
additional impact when flushing was combined with an existing
sweeping program. Twin Falls, Idaho data indicated that
weekly flushing/sweeping probably did not have a significant
effect on air quality in that city, although concentrations
were consistently lower on the days when the cleaning was
done. It has been assumed for the cost-effectiveness analysis
that flushing/sweeping would have the same effect on air
quality as flushing alone. This assumption precludes com-
bined cleaning from being as cost-effective as simply flush-
ing. One of the primary benefits of picking up the material
that is flushed to the gutter—that removal would reduce
storm water loadings on the sewage treatment system--does
not enter this cost-effectiveness analysis because effec-
tiveness is measured only in terms of air quality improvement.
Modification of Street Sanding Procedures
For the same non-attainment area, the direct costs of
modifying street sanding procedures by switching to salt and
cleaning as soon as possible after every application are
estimated from the extra cost of salt compared to sand and
the cost of the additional street cleaning. The costs do
not include the savings that should be realized by lower
spring cleanup costs if the streets are cleaned periodically
74
-------
during the winter and salt is used in place of sand, nor do
they include possible indirect costs associated with salting
such as decreased vehicle lives. Assuming a need for sand
or salt on 10 occasions per year, the amount of sand applied
on the arterials would be 400 tons per year (at an applica-
tion rate of 800 Ib/mi). This would cost $3200 if sand
costs $5.00 per ton and can be applied for an additional
$3.00 per ton. Substitution of salt for all 10 applications
25
at a rate of 500 Ib/mi and an average cost of $20 per ton
(including application at $3.00 per ton) would total $5000
or $1800 more than the sand.
Removal of salt residue and other material after each
application is estimated to cost $8.24 per curb-mi (average
broom sweeper costs), or a total of $16,500. Thus, the
additional cost of both .modifications of the snow control
program would be $18,300.
The improvement in air quality is approximated at 5
ug/m for the winter quarter (at least a 50 percent reduc-
tion in contribution and a residual impact with salting of
5 ug/m ), or a 1.2 ug/m reduction in the annual arithmetic
mean.
Control of Mud Carryout from Construction Sites ••
The most effective control measure at construction
sites was found to be immediate manual cleaning. The cost
of an employee on half-time assignment to clean the entrance
area to the construction site would be (8 hr)(1/2)($8.37/hr)
= $33.50 per day. For a year of five-day weeks at a con-
stuction site, the total cost would be $8400.
The improvement in air quality due to the cleaning
would be 15 ug/m over an area of about 1000 ft radius (0.1
mi ). Since this is a very limited area of impact, control
of mud carryout at construction sites cannot be compared
directly with other measures. Per unit area, the cost of
75
-------
construction site cleanup is more for each ug/m reduction
in particulate concentrations than the four measures with
regional impact. However, construction site control has its
effect entirely in an area with elevated concentrations.
Control of Mud Carryout from an Industrial Area
The control program for mud carryout sources in the
Duwamish Valley (see pages 51 to 53) provides the cost-
effectiveness data for this measure. An average reduction
3 2
of 17 ug/m over the 16 mi area of the valley was achieved
for a total capital cost of $1-2 million (assume $1.5
million). Costs for the other measures were annual costs
rather than capital costs. With an interest rate of six
percent (municipal) and a seven year assumed life for the
strip paving, the equivalent annual cost for the air quality
improvement is $268,700.
The cost-effectiveness data for the five measures are
summarized in Table 4-7. Comparison of the data for diff-
erent measures indicates that modification of street sanding
procedures is the most cost-effective, followed closely by
control of mud carryout from an industrial area and street
flushing. The costs per unit improvement in air quality for
these three measures with potential impact over an entire
subregional area fall within a relatively small range. Even
the more expensive construction site control measure can be
justified as the most feasible means of preventing concen-
trations above the ambient air quality standards in the
immediate vicinity of a construction site. There is no
information to indicate whether the combinations of these
measures have an additive effect in reducing particulate
concentrations.
Given the relative accuracy of the input data for the
cost-effectiveness analysis, it should be concluded that
76
-------
Table 4-7. COST-EFFECTIVENESS OF CONTROL MEASURES
$ per
Annual Av reduction- ug/m3 Area
Control measure cost,$ in cone, ug/m reduction affected
Flushing 78,400 4.6 17,000 3x5 mi
130,700 7.6 17,200 3x5 mi
Combined flush- 249,800 4.6 54,300 3x5 mi
ing/sweeping
Modification of 18,300 1.2 15,200 3x5 mi
street sanding
procedures
Control of mud 8,400 15.0 560 0.1 mi2
carryout from
construction
sites
Control of mud 268,700 17.0 15,800 16 mi2
carryout from
an industrial
area
77
-------
costs may not provide a sufficient criterion for selection
of a control strategy. The measure or measures that can be
adapted most readily to existing public works practices
would probably be preferable to the most cost-effective
measure.
Without data for a direct comparison, it appears that
all of these control measures for reentrained dust have
substantially lower costs than point source controls that
would have an equivalent impact on particulate air quality.
78
-------
5. OPTIMIZING REENTRAINED DUST CONTROL MEASURES
It is evident that the field studies did not provide
conclusive data on which recommendations for control actions
could be based. However, some specific questions were
answered by the field studies, and information from personal
contacts with public works personnel and from the literature
search revealed other methods of improving the efficiencies
of the control measures. This chapter summarizes the infor-
mation that has been assembled to date on optimizing reen-
trained dust control measures. There were no field studies
for street sanding or traffic volume reduction measures, so
the discussion is confined to improved street cleaning and
mud carryout controls.
STREET CLEANING
Current Street Cleaning Practices
Many cities have an intensive four to six week spring
cleaning to remove debris accumulated during the non-clean-
ing season and sand from the winter's snow and ice control
programs. They may operate two 8-hour shifts daily to cover
their entire street network. In some cities, any other
cleaning that is done is undertaken only as needed. More
commonly, a routine street cleaning program operates from
April through October to cover the entire city on a system-
atic basis.
In 1975, the American Public Works Association (APWA)
performed a survey of public works officials on various
24
aspects of street cleaning practices. From the 250 agencies
79
-------
included, a total of 152 responses were received. This
survey indicated that 42 percent of the cities clean their
central business district (CBD) daily; 10 percent of the
cities clean commercial areas daily and 40 percent clean
weekly; and 25 percent of the cities clean residential areas
weekly.
Most of these 152 agencies are presently using either
3-wheel or 4-wheel mechanical sweepers and almost half these
agencies also have flushers. A previous APWA survey con-
ducted in 1971 indicated that half the cities surveyed
included flushing as part of their street cleaning (20
percent flushed before, 45 percent flushed after, and 35
percent flushed both before and after sweeping). Thirty-five
agencies reported in the 1975 survey that they use vacuum
sweepers and 14 agencies use regenerative air sweepers.
Response to the questionnaire indicated further that fewer
agencies in the future will have the mechanical sweepers and
flushers but the number of agencies using vacuum sweepers
was projected to remain about the same.
In response to questions relative to their public works
budgets, most officials felt that street cleaning was gain-
ing in importance in their overall public works programs.
Responses to these questions are summarized below:
Response of agencies surveyed, %
Time Smaller Same Larger
span budget budget budget
Recent 13 56
past (1970-75)
Near 11 53
31
36
future (1976-80)
Distant 10 43 47
future (after 1980)
80
-------
It is interesting to note that the same officials who
thought that street cleaning would increase as an item of
budget also indicated a trend to fewer types of equipment in
their cities.
Optimizing Street Cleaning
Even cities with frequent, systematic street cleaning
have an air quality problem due to reentrained dust. Cer-
tain modifications of common practices offer a possible
means for improving air quality. Six of these modifications
are discussed below.
Equipment - Flushing showed the most promise for reducing
reentrained dust. It wets the streets, causing dust sup-
pression until the surface is completely dry, and moves
material out of the traffic lanes to the gutters. The only
limitation on the use of flushers is in areas where water
availability is restricted. Flushers use 3,000 to 4,000
gallons of water per mile of street, or up to 70,000 gal/day.
Therefore, street flushing could easily constitute 1 to 2
percent of a city's total water consumption.
Flushing can also be used in combination with subse-
quent material pickup by a broom, vacuum, or regenerative
air sweeper. This combination method results in very effec-
tive removal because the material is wet and is concentrated
in the gutter. However, data in Chapter 4 showed that
flushing followed by sweeping is much more expensive than
just flushing, with no demonstrated additional improvement
in air quality.
Broom sweeping has two operational characteristics that
make its use alone of doubtful value—it moves material from
the gutter back into the street for pickup and it is not
efficient in removing fine particles. It reduced particu-
late concentrations in only one of three field studies.
81
-------
Neither vacuum sweepers nor regenerative air units produced
reductions in particulate concentrations in field studies.
Frequency - Flushing is effective in reentrained dust con-
trol primarily on the day it occurs, with little residual
effect, as shown in Figure 5-1. In that respect, it is
similar to rainfall. Because daily flushing is as effective
pe.ft. t'ime. as infrequent flushing, from an air quality stand-
point it should be done as frequently as possible. In areas
where particulate concentrations are above the air quality
standard, flushing could be done daily or every other day.
This frequency of cleaning is not recommended for broom
or vacuum sweeping. Higher than average particulate concen-
trations were observed on days with sweeping, probably due
to the cleaning operations breaking loose material adhering
to the street surface, redistributing it, and temporarily
making more material available for reentrainment. The
curves of concentrations versus time since cleaning pre-
sented in Figure 5-2 show that the lowest concentration
usually occurs on the first day after cleaning, with a
continual increase thereafter. Assuming that broom or
vacuum sweeping does have some positive effect on air quality
by removing material from the street (even if this impact
was not detectable in the field studies), it appears that
sweeping should not be done more often than every two to
three days.
Location - The direct relationship observed between particu-
late concentrations and traffic volume indicates that there
are advantages to concentrating cleaning efforts on streets
with high traffic volumes. The relationship between traffic
and reentrained dust emissions suggests that frequent (daily
or every other day) cleaning of major arterials and less
frequent cleaning in the CBD and on residential streets
82
-------
50
ro
?s -
o
c
o
o
Q)
.4J
(0
O 0
•H
-M
tH
<0
0)
Cr>
OS
fi
t)
-25
-50
Note:
Data for Figure 5-1 and
5-2 from Cincinnati
street cleaning study.
4-
V I * •* ^ ^ v
Days since flushing or rainfall
Figure 5-1. Effectiveness of flushing and rainfall in reducing
particulate concentrations.
5
-------
might reduce total emissions. Many cities with comprehen-
sive street cleaning programs observe the following schedule:
daily cleaning in the CBD; weekly cleaning of major arterials
and industrial areas; and every other week or monthly clean-
ing of residential areas. A reallocation of the cleaning
effort to every other day in the CBD and on major arterials
with remaining equipment operating as often as possible in
industrial and residential areas appears to offer a poten-
tial for improving air quality.
Cleaning Seas.on - The removal of excessive materials as they
accumulate during the winter is preferable to an intensive
spring cleanup. The length of time between cleaning, the
application of antiskid material/ and the greater amount of
dirt that is deposited from vehicle undercarriages causes
higher street loadings in winter. Use of flushers would not
be practical, even during periods with temperatures above
freezing, because temperatures would usually be low at night
and could cause icing. However, sweeping the streets
whenever weather permits throughout the winter season would
be helpful in lessening the air quality impact from exces-
sively dirty streets. A similar situation occurs in the
fall when leaves should be removed from streets to reduce
this potential source of dust and subsequent reentrainment.
Operator Training - One reference recommended increased
training of equipment operators. This would be particu-
larly important with respect to the air quality benefits of
certain cleaning procedures. Training and performance are
thought to be directly related, but only 43 percent of the
cities surveyed by APWA in 1975 have a formal operator
24
training program. The average initial training period is
54 hours per operator with subsequent training of about 30
hours per operator per year. The authors of the above
84
-------
report on street cleaning recommend that operators be trained
not only in how their equipment can best be used and main-
tained but also in what material needs to be removed and
where this is commonly located.
Street Improvements - Two actions can be taken to reduce the
interference of parked vehicles with street cleaning.
First, signs can be posted along frequently cleaned streets
prohibiting parking during specified time periods. This is
more important for sweeping than flushing because sweepers
must have continuous access to the gutter area to be effec-
tive. The second action would improve flushing efficiency,
by marking parking lanes or changing curb/gutter configura-
tions to allow the free flow of water and flushed material
even in the presence of parked cars.
MUD CARRYOUT CONTROL
Current Mud Carryout Controls
There are currently a variety of means used to control
trackout or runoff of mud and dirt onto street surfaces.
These methods were discussed briefly in Chapter 1; regula-
tions requiring these methods are discussed in the next
chapter.
Optimizing Mud Carryout Controls
For permanent sources of mud carryout (e.g., operations
such as ready mix plants, truck terminals, warehouses), it
is preferable to select a permanent approach to control.
These controls can include paving, oiling, seal coating, or
use of nondusting gravel. This minimizes the carryout
problem so that it is not necessary to clean adjacent streets.
All of the above controls will produce approximately
the same air quality benefit. However, there are other
85
-------
considerations that would influence the owner or operator's
selection: capital cost, maintenance, ability to withstand
loads, and drainage and storm water runoff problems.
At temporary sources such as construction sites, it is
not possible to deal with the problems by paving, oiling,
etc. The data assembled for this report show that the best
control approach during the excavation period is immediate
cleanup of all mud and dirt carried onto streets to prevent
other traffic from spreading the trackout to other streets
in the area. A less desirable alternative is to require
cleaning of the adjacent streets at the end of each day's
work shift. Use of flushers on adjacent streets or any
watering control should be undertaken only where it will not
create a greater problem of mud trackout.
For most cleared areas, there is a significant problem
with runoff onto adjacent streets. At construction sites,
contractors use a number of control methods. Such measures
are usually implemented in a limited area. However, con-
tractors frequently clear a large area where construction is
proposed but may not be undertaken for a number of months or
even years. These areas contribute to runoff problems
because there is no paving or vegetation to control the soil
erosion. The optimum means for controlling these areas is
to limit the number of acres cleared and to require control
of erosion and runoff by some method of stabilization.
86
-------
6. LEGAL AND ADMINISTRATIVE ASPECTS OF
REENTRAINED DUST CONTROL
There are presently a large number of state regulations
and local ordinances that represent control measures for
material deposition. In addition, there is considerable
diversity in the street cleaning programs utilized by local
*
communities. The development of regulatory approaches and
administrative agreements as mechanisms for legally enforc-
ing control measures is one of the purposes of this docu-
ment. Some system of intergovernmental coordination would
obviously be desirable in view of the existing regulations
and street cleaning programs. It is important to find a
reasonable means of incorporating reentrained dust control
into State Implementation Plans (SIP).
PROGRAMS FOR PREVENTION OF MATERIAL DEPOSITION
Existing State Regulations
In 1971, EPA recommended certain reasonable precautions
that could be taken to prevent particulate matter from
becoming airborne and these were published in the Federal
*
Presentation of legal and administrative aspects of
material deposition will precede those on improved street
cleaning. Material deposition is more readily handled by
regulatory approaches, while street cleaning is usually the
responsibility of a local agency and may not be subject to
regulation. Although this reverses the order of presenta-
tion used in the other chapters, it allows for a more logi-
cal discussion in this case.
87
-------
Register (Appendix B, Section 2.2). Some of these recom-
mendations are actually mud carryout controls and include
the following:
"a. Use of water/chemicals for dust control at sites
of demolition, construction, road grading, or land
clearing.
b. Application of suppressants at stockpiles, on dirt
roads, or other dust producing surfaces.
c. Use of hoods, fans, filters, or other containment
methods at dust producing operations.
d. Covering, at all times when in motion, open bodied
trucks transporting materials likely to give rise
to airborne dust.
e. Conducting agricultural practices in such manner
as to prevent dust becoming airborne.
f. Paving roadways and maintaining roads in a clean
condition.
g. The prompt removal of earth or other materials
from paved streets onto which earth or other
matter has been transported by trucking or earth
moving equipment, erosion by water, or other
means."
A review of state air pollution control agency regula-
tions reveals 21 states that use portions of the Appendix B
recommendations for controlling air pollution from fugitive
dust sources. A total of 16 states make use of both para-
graphs d and g above, which expressly pertain to the problem
of mud carryout.
The Colorado Air Pollution Control Commission has an
abatement/prevention measure for unpaved roads and parking
38
areas (mud carryout sources). They recommend use of
88
-------
water, suppressants, shielding, paving, and speed restric-
tions. They also have control measures for preventing
deposits of mud and dirt on improved streets and roads.
They require that haulage equipment be washed, wetted down,
treated, or covered to minimize dust in transit and loading.
In New York, the air pollution control agency has
proposed use of fugitive dust control permits, to become
effective January 1, 1980, that would regulate the use of
39
paved or unpaved parking lots and terminal areas. The
regulation would also require persons responsible for the
use of any road not to permit its use unless it is cleaned,
treated, and traffic controlled in such manner that visible
particulates are not produced.
There are state statutes, implemented by agencies other
than air pollution control, that are applicable to mud
carryout control. In the states of Massachusetts, Michigan,
Minnesota, New Jersey, Texas, Utah, Virginia, West Virginia,
and Wisconsin, the administering agency is the highway
division. The regulations in these nine states require that
anyone operating a vehicle carrying a load on a public
highway must prevent that load or any part of it from spill-
40
ing. In six of these states, Massachusetts, Michigan, New
Jersey, Utah, West Virginia, and Wisconsin, it is also
required that loads be securely fastened to prevent leakage
40
and spillage. None of the state statutes specifically
authorize anyone to obtain an injunction to enforce the
statute; however, four states (Minnesota, New Jersey, Utah,
and Wisconsin) authorize seeking civil remedies and six
states (Michigan, Minnesota, New Jersey, Texas, Virginia,
and West Virginia) provide for criminal penalties with
maximum fines ranging from $100 to $1000 and maximum impris-
onment from 10 days to 12 months.
89
-------
Two states offer good illustrations of statewide regu-
latory programs with regard to covering trucks and control-
ling mud carryout. North Carolina Highway Commission has a
regulation which specifies that "no person operating a dual
wheel truck may track or cause mud to be deposited on the
paved portion of any state highway, creating a hazard to the
traveling public, and such deposits must be immediately
41
removed, under penalty of a misdemeanor." In addition,
there is a Highway Commission specification that requires
the state's contractor to comply with all state and local
air pollution regulations throughout the life of a project,
and suspension of work can follow within 24 hours of failure
42
to perform necessary measures of control.
The second state, Maryland, requires covering at all
times for vehicles moving material likely to create air
pollution, as well as prompt removal from paved streets of
earth or other materials that result from truck or earth-
43
moving equipment or erosion by water. Their highway main-
tenance supervisors and engineers operate a regular schedule
of inspection and all violations of the highway code are
reported and enforced by the highway patrol. Inspectors
require contractors at subdivision developments, shopping
centers under construction, and highway construction sites
to control such problems with calcium chloride and water
trucks and to clean up all spills and trackout.
These state level regulations described above are
summarized in Table 6-1.
Existing County and City Ordinances
A number of local ordinances further illustrate possi-
ble approaches to mud carryout control. The Polk County
(Des Moines), Iowa fugitive dust measure includes the Appen-
dix B wording of both paragraphs d and g relating to truck
covering and removal of mud from roadways.45 Two additional
90
-------
Table 6-1. SUMMARY OF STATE REGULATIONS
APPLICABLE TO MUD CARRYOUT CONTROL
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Air Pollution Control agency
Federal model provisiona Other
d. f. g. regulation
X X
X
X X
XX X
XX X
X ' X X
X X
X X
X X
XXX
XX X
X
X
XX X
X X ' X
X
xx x
X X
X X
XX X
X X
XXX
Other state
agency
regulation
X
X
X
X
X
X
X
X
X
X
X
see pages 87, 88.
91
-------
counties regulate against fugitive dust—Allegheny County,
Pennsylvania uses all the provisions of the Appendix B
recommendations with the exception of the paragraph on truck
covering; and Milwaukee County, Wisconsin uses all the
wording of the Appendix B recommendation with the exception
AC A **1
of the paragraph on mud removal from roads. '
Two cities use the Appendix B recommendations without
any modifications—Washington, D.C. and Cincinnati. '
Several cities have mud control ordinances that require
paving of all parking lots. Representative of these is the
Rapid City ordinance: "All areas devoted to permanent off-
street parking . . . shall be of a sealed-surface construc-
tion and maintained in such manner that no dust will result
from continuous use."
The city of Seattle has a street use ordinance, No.
90047 (1961), that prohibits deposit of materials on streets.
Regulation I, Section 9.15, of the Puget Sound Air Pollution
.Control Agency, a four-county regional agency in Washington,
requires use of water or chemical suppressants at construc-
52
tion sites, parking areas, and roads being cleaned or repaired.
At the end of each work shift, mud and dust on all public
roadways has to be removed. The agency also recommends that
log storage areas be equipped with truck wash-down facili-
ties so that trucks and log-hauling equipment can be cleaned
prior to entry on public roadways. In response to this
guideline, some sand and gravel operations in Seattle have
installed wheel washes consisting of two parallel spray
bars, approximately 25 feet long and 6 inches above ground,
activated by an electric eye.
Charlotte, North Carolina has a comprehensive plan for
community improvement and their Public Works Department is
responsible for making inspections as well as citing viola-
tions of mud carryout requirements. The maintenance of
construction sites is the responsibility of individual
92
-------
contractors and each employs his own manual labor to take
care of this cleanup. There is also a private local firm
that is employed to flush streets where the contractor
cannot adequately take care of the cleanup. Their construc-
tion site ordinance reads as follows: "In the event that
dirt, mud, construction materials or other debris shall be
deposited upon any street or sidewalk as a result of a
construction project in progress, the contractor in charge
of the project shall be required to remove said debris. Any
contractor failing to comply with this provision shall be
assessed a penalty of ten dollars ($10.00), and each and
every day during which such violation continues, shall be a
54
separate and distinct offense."
The Chicago Building Department controls problems of
mud carryout at construction sites with requirements con-
tained in the building permit issued to the contractor.
Initially, excavating contractors take responsibility for
cleanup of spills and trackout on surrounding streets and
use manual labor and brooms to control the problem. Follow-
ing excavation, some building contractors use street sweep-
ers and subcontract the rest of the work to a hauling company.
Others employ street sweepers, motor graders, and prepare
all construction sites and haul roads with fragmented brick
to control both dust and runoff problems.
Kansas City, Missouri prohibits mud trackout and
requires covers for trucks hauling materials to and from
construction sites. These two ordinances, Sections 34.145
and 34.148, are enforced by the Street Department, which is
concerned primarily with concrete or sand spills on city
streets. Local contractors use coarse gravel to prevent
mud buildup between truck tires and sometimes use broom
sweepers.and vacuums on streets adjacent to their sites.
Kansas City, Kansas prohibits overloading trucks,
spilling from trucks, or leaving construction materials on
93
-------
streets.57 Their ordinance is also under the jurisdiction
of the Public Works Department and it carries a maximum
penalty of $500 and/or 30 days imprisonment.
Enforcement Procedures
The extent of enforcement of control measures for
reducing material deposition on streets is difficult to
ascertain. The information gathered for this report indi-
cates that responsibility for inspection and enforcement is
usually delegated to a division within the city public works
department or within the state highway department. There is
uneven enforcement of the Appendix B type regulations by
state air pollution control agencies.
While all agencies respond to complaints if an ordi-
nance is violated, some rely on a team of inspectors to
watch for violations and to deal with those that occur. The
usual procedure is to notify the person in charge that an
ordinance is being violated and to give him a number of
hours to comply with the request. Where cooperation is not
obtained, there are a variety of steps taken to bring about
compliance.
Of the cities cited in the above section, several were
contacted to determine how they enforce their ordinances.
These findings are presented in Table 6-2. If these agen-
cies offer a fair representation of practices nationwide, it
appears that the most common and effective procedures are
systematic inspection, notice of request for compliance, and
enforcement through a fine or billing for cleanup done by
the city or state.
The most critical components of an enforcement program
are frequent inspection (perhaps in coordination with the
local air pollution control agency) and a penalty for lack
of compliance that is large enough to make it worthwhile to
comply with the law.
94
-------
Table 6-2. ENFORCEMENT OF LOCAL'ORDINANCES
FOR CONTROL OF MUD CARRYOUT
city/
Population
Charlotte/
241 ,000
Chicago/
3.367.000
Kansas City.
Kansas/
168.000
Milwaukee/
717,000
Rapid City/
48*000
Rapid City/
46,000
Seattle/
531.000
Agency
enforcing
Community
I improvement
Div
Dept of
Bldg; Bldg
Permits , Adm
Svcs Office
Street Dept
Bureau of
Street San-
itation
Streets and
Building
inspection
Engineers
Office
Typo of
ordinance
Oi'd requires
dirt and
debris from
conotj ord
requires
truck cover-
ing
Bldg per-
mits; anti-
Ord prohib-
its truck
•pillage and
prohibits
const ma-
terials being
left on
streets
Ord requires
truck
tires be
cleaned;
prohibits
all dirt
on streets
Ord prohib-
of materials
and requires
cleanup
Bldg permit
requires
scaled-sur-
face const
and mainten-
ance for all
off-street
parking
areas
Street use
ord
No. Of
citations,
annual
23 issued in
contractors
Depends on
weather con-
daily viola-
tions during
wet weather
2 to 3 court
actions in
1976
Daily vio-
whc never it
snows (30%
of the time
cleans; 70%
of the time
the city
cleans and
bills the
contractor)
Daily vio-
lations
Manpower
within
department
Division
plus nine
53 inspec-
tors (this
of many ord
included in
daily in-
there are
60.000
nits issued
annually as
well as
utility
work, which
•Iso would
create mud.
carry out
_
9 super-
visors (on
an 8 mo-
Apr thru
Bov — opera-
tion) who
handle vio-
lations
3 foremen
(to in-
spect) and
sweeper
operators
5 Building
inspectors
and 4 from
the Eng and
1 from tho
Fire Dept
13 inspec-
tors
Enforcement
procedures
Get compliance
at all sites.
cooperation;
ord for truck
tarping
enforced by
Police Dept
Bldg permit
requires ad-
bo kept clean;
a site is shut
down ( where the
not comply , un-
til he arranges
problem with a
wheel wash and
with sweepers.
This ord is en-
forced only if
complaints are
received by the
Police Dept or
the Street Div
Police Dept
closes down
const site un-
til compliance
is achieved;
all violations
const Is halted
aze fined S500/
day and, in addi-
tion, a traffic
violation is
Issued against
each truck that
continues to
operate from
the site
Have ord against
littering streets
spilling loads;
there is citizen
identifying vio-
lations and good
cooperation from
sand/gravel owners;
city also cleans
all mud runoff
from streets in
low lying areas
after rain/snow
Bldg permits for
control of all
parking lots; ord
requires surface
scaling; very few
unsurfaccd lots
(they use dust
preventives on
these during
tho summer)
Use a bond/do-
poait oystcm
Total manpower within tho dcpartmontj rcnponsiblo for inspection and enforce-
ment of other ordinances in addition to the mud carryout controls shown.
95
-------
It is necessary to have a team of inspectors (two or
more depending upon the size of the city) who can make daily
or every other day inspections of the entire community.
Some cities have a grid system and assign a single grid to
each inspector. The department responsible for inspection
and the manpower employed varies among cities. As shown in
Table 6-2, the provisions for mud carryout control are
usually contained within a building permit and thus are the
responsibility of the building department, or they are
street use ordinances and are enforced by the street department.
The three cities shown in Table 6-2 with the strictest
enforcement procedures are Milwaukee, Chicago, and Seattle.
In Milwaukee, the $500 daily fine is large enough that
payment of the fine is more costly than complying with the
58
ordinance. When the stop order is issued to halt con-
struction, every truck that continues to operate from the
site is also in violation of a traffic ordinance and is
fined. Compliance requires that the contractor use a desig-
nated route to and from his site and that he keep the route
clean with mechanical sweepers. At the large construction
sites, contractors must also build a coarse gravel road for
use by their trucks to shake debris from the undercarriages.
In Chicago, the building permit issued at the time con-
struction is begun requires the contractor to control mud
59
carryout. This usually involves a steel mesh decking at
the entrance to the site where truck wheels are hosed off,
and the "barricade permit" allows the contractor use of
adjacent streets for the period of construction as long as
he employs a street sweeper to keep the streets clean.
Chicago does not fine, since a stop order is issued where
contractors fail to comply with the provisions of their
building permit.
In Seattle, control is obtained with a bond/deposit
system where the contractor posts a bond with the city.51
96
-------
If the contractor fails to meet any city requirements for
maintenance of his site, the city performs the work and
bills the service against the deposit.
Model Regulations
Two control measures were discussed in Chapter 3 that
have potential for improving air quality—control of dirt
and mud carryout from major sources and control of dirt and
mud carryout from ubiquitous sources. A model regulation
for implementing these two measures would include the follow-
ing provisions:
Major mud carryout sources
0 Manual or mechanical sweeping of streets adjacent
to major sites, preferably with immediate removal
of material deposited on streets or a cleanup at
the end of each work day.
0 Use of a ground cover or chemical dust suppressant
for sites to control runoff and trackout.
0 Addition of a coarse gravel road on construc-
tion sites for trucks to drive over and shake
loose debris from the vehicle undercarriage.
0 Wheel wash mechanisms for vehicular construction
equipment and for tractor trailers at terminal
sites.
0 Truck tarping or required wetting of the load.
0 A limited number of cleared acres at any one
time in a given area unless soil erosion con-
trol measures are included.
0 Some method of soil control for cleared areas
where development of the area is delayed.
Ubiquitous mud carryout sources
0 A maximum number of vehicles using a parking area
unless it is paved.
0 Required application of seal coating or oiling for
parking areas used by fewer vehicles.
0 Oiling or seal coating for unpaved driveways and
road shoulders.
97
-------
Intergovernmental Coordination
if regulations such as those cited above are to be in-
cluded in SIP'S, theri some clearly defined mechafiism for the
cognizant air" pollution contr61 agencies to coordinate
enforcement of that existing regulation with the nOn^air
po'liution control agency must be established. The state has
to incorporate in its implementation plan the necessary
mechanisms for reducing reehtrained dust. They must then
delegate the implelrnehtation of these specific mechanisms to
io'cal Sgencie's; in most cases, this would mean the munici-
pal public works department.
If the state agency is not going to pass mud carryout
regulation^ and depends on another agency to enforce such
regulations, then establishment of some kind of administra-
tive agreement for inspection arid enforcement would probably
be necessary.
P&OGF&MS F0'6 IMPROVED STREET CLEANING
ThQ d.@hcept G~f. ah Administrative Agreement
Regulations requiring municipalities to adopt specific
street cleaning practices for air pollution control purposes
Would probably be unsuccessful, principally because present
knowledge of controls for reeritraihed dust are insufficient
t'6 define in general regulatory terms the optimum street
cleaning program for a city. As an alternative to regula-
tibri* the tide" of an administrative agreement between the
state air pollution control agency arid the local government
is proposed. Through this agreement, the local government
would be committed to institute changes in street cleaning
practices in order to improve particulate air quality.
The term "administrative agreement" as used herein
refers to a written agreement entered into by two or more
governmental units. Such an agreement would constitute a
98
-------
formal, but not necessarily legally enforceable, commitment
by one governmental unit to engage in a specific set of
activities. It would consist of a statement of local govern-
mental policy regarding street cleaning and air pollution
control and a plan by that body to implement programs,
including any necessary budgetary adjustments. A timetable
for implementing the actions would also be specified in the
written agreement. An example farthis type of agreement is
presented in Figure 6-1.
There has been no indication whether such a nonregu-
latory approach would be approvable by EPA if submitted as
part of an SIP revision. The Federal regulations (40 CFR
Sl.llf) do allow the state to delegate authority to c.a.fifiij
oat part of the plan:
The State may authorize a local agency to carry out
a plan, or portion thereof, within such local agency's
jurisdiction: Provided, that such plan demonstrates,
to the Administrator's satisfaction, that such local
agency has the legal authority necessary to implement
such plan, or portion thereof, and further: Provided,
that such authorization shall not relieve the State of
responsibility under the Act for carrying out such
plan, or portion thereof.37
The municipalities certainly have legal authority to
clean their streets. The uncertainty arises in whether
every measure identified in the control strategy must be
backed by a regulation. Also, the state has no basis for
seeking injunctive relief for nonperformance by the muni-
cipality and has no legal authority for cleaning the streets
if the municipality fails to do so.
No precedents for comparable situations to this could
be found in other EPA regulatory areas. The agreement shows
recognition of the state agency's objectives and an intent
to comply with program requirements that are not as yet
defined because they must be designed specifically for each
municipality. nn
-------
THIS AGREEMENT made and entered into on this the day of , 19 ,
by and between the STATE of / hereinafter referred to as the Party of the
First Part, and the CITY OF / a municipal corporation organized and
existing under and by virtue of the laws of the State of , hereinafter re-
ferred to os the Party of the Second Part.
HITNESSETH:
That, WHEREAS, the parties hereto recognize that there is a great need on the part
of their citizens for a program and/or system for controlling dust from paved streets
and parking facilities affecting air quality within the jurisdictions of the party of
the first part and the party of the second part; and
WHEREAS, the party of the first part has found that the institution of a program
and/or system for controlling air pollution from paved surfaces is necessary in order to
achieve the intents and purposes of the State Air Pollution Control Implementation
Plan; and
WHEREAS, the party of the second part is desirous and willing to define and implement
policies and execute programs which will produce effective control of dust from paved
surfaces owned and operated by the party of the second part.
NOW, THEREFORE, in consideration of the premises and in the further consideration of
the mutual agreements and covenants hereinafter contained, the parties hereby enter
into the following agreements:
1. The party of the second part is authorized by the party of the first part to
carry out those portions of the State Air Pollution Control Implementation Plan relating
to control of dust from paved streets within its jurisdiction.
2. It is agreed that the party of the first part, acting through its Department
of Environmental Protection, will provide technical assistance to the party of the second
part in defining the degree and extent of fugitive dust control required on paved sur-
faces within the boundaries and jurisdiction of the party of the second part.
3. It is agreed that the party of the second part will within two months oi the
date of this agreement adopt a resolution stating that it is the policy of the party of
the second part to modify street cleaning practices to meet air quality objectives;
within six months of the date of this agreement specify and submit to the party of the
first part a detailed program for implementing said policy, which program shall include
such elements as both parties agree to be necessary; and at the time of the next follow-
ing budget approval submit to the party of the first part an operating budget and capi-
tal expenditures budget which provides the fundings necessary to implement the programs
so developed.
4. It is agreed that the party of the first part and the party of the second part
.will equally share the expense of determining the degree and extent of dust control re-
quired, and that the party of the first part will assist the party of the second part
in. finding supplementary sources of funding to implement the programs developed as a part
of this agreement.
IN TESTIMONY WHEREOF, the parties hereto have caused this agreement to be executed,
in duplicate, by their duly authorized agents and/or officials, as of the day and year
first above written,
ATTEST: STATE OF
By
Recorder Executive Director,
Department of Environmental
Protection
ATTEST: CITY OF
Clerk Mayor
Figure 6-1. Example administrative agreement.
100
-------
Development of an Administrative Agreement
There are five basic elements to the process of de-
veloping the type of administrative agreement discussed
herein:
decision by the air pollution control agency
that improved street cleaning is necessary
in order to attain and maintain air quality
standards;
establishment of a joint institutional
mechanism with the public works department
to define the nature of such improvements;
provision of active technical assistance
on the part of the air pollution control
agency;
adoption of altered street cleaning prac-
tices as a matter of local governmental
policy;
development of program elements designed
to implement and assess those desired
improvements.
The first of these is amply discussed in other portions of
this report and need not be discussed further here.
Institutional Mechanisms - Modifications to the street
cleaning program determined independently by the air pollu-
tion control agency have no better chance of being imple-
mented than equivalent regulations adopted by the agency.
Rather, what is required is an institutional process for
agreeing upon changes that can be fitted into the existing
street cleaning operations with the optimum mix of minimized
change and maximized air pollution control effectiveness.
It is suggested that a committee comprised of repre-
sentatives of both organizations be appointed to determine
the technical means of achieving the desired degree of im-
provement in air quality. This committee process might be
rather difficult to implement in practice due to differing
organizational imperatives.
101
-------
Technical Assistance - A public works department cannot be
expected to know what must be done to reduce the air quality
impact of reentrained dust from paved streets. Nor can an
air pollution control agency be expected to know what must
be done to clean streets efficiently and effectively. Each
group must work with the other, as is implied by the admin-
istrative agreement concept, to determine what must be done
to produce mutually desirable results.
To this end, the air pollution control agency needs to
provide three services: guidelines as to what equipment and
procedures are most beneficial in terms of air quality;
directions as to what parts of the municipality are most in
need of improved air quality; and air quality sampling data
which can be used to monitor and assess the effect of modi-
fications in the existing street cleaning system.
Municipal Policy Development - In making policy decisions
that add, delete, or modify community objectives, municipal
administrations are usually responding to public pressure,
special interest groups, perceived need by the administra-
tion itself, or consideration of requirements or desires of
other governmental entities. The municipal administration
must determine the likely costs and benefits of the improve-
ments in air quality that can be expected from altered
street cleaning practices and where and how the necessary
modifications in activities and expenditures are to be made.
Once these questions have been satisfactorily answered the
municipal administration can then draft a resolution stating
the policy to be followed and submit it to the city council
for a vote under its prescribed rules of ratification.
Wording of the resolution in each municipality will vary,
but may approximate the model shown in Figure 6-2. Follow-
ing favorable adoption of such a resolution, administrative
and other legislative actions will be required to implement
102
-------
BY THE COUNCIL:
WHEREAS, this Council has, through the State Department of Environmental Pro-
tection, become aware of overall air quality plan objectives for the State and this
municipality; and
WHEREAS, this Council recognizes that national air quality objectives are embodied
in the State objectives; and
WHEREAS, this Council also recognizes the need to modify street cleaning practices
in order to meet air quality objectives and protect the health and welfare of the citi-
zens of this municipality; and
WHEREAS, this Council has been advised that by reordering public works expenditures
and procedures and modifying existing powers and programs that this municipality can
achieve a significant measure of air quality improvement through control of reentrained
dust;
NOW THEREFORE, BE IT RESOLVED by the Village/City Council of this municipality,
that program plans and implementation be directed toward a systematic modification of
street cleaning practices;
BE IT FURTHER RESOLVED that the Department of Public Works is directed to cooperate
with the State Department of Environmental Protection to define jointly the nature, de-
gree, and extent of those modifications which are necessary to implement this policy; and
BE IT FURTHER RESOLVED that the Clerk of this Council is hereby directed to certify
copies of this resolution to Village/City Directors of Public Works, Office of Budget,
and Community Planning for program planning and execution.
ADOPTED at a regular meeting of the Village/City Council of Any County, Every State
this day of , 19 .
Council-Person AYE NAY
Council-Person AYE NAY
Council-Person AYE NAY
Council-Person \ AYE NAY
Council-Person AYE NAY
CERTIFICATE OF CLERK
IT IS HEREBY CERTIFIED that the foregoing is a true and correct transcript of a
resolution adopted by this Council in session the day of , 19 .
IN WITNESS WHEREOF, I have hereunto set rr.y hand and affixed the official seal of
the Village/City Council of Any County, Every State this day of
Clerk
Village/City Council
Figure 6-2. Example city council resolution.
103
-------
the policy. The first administrative step involves program
planning and development.
.Program Planning and Development - Programs are action-
oriented media for carrying out policies. The increasing
complexity, number, and technical character of municipal
problems makes it necessary that program planning be accom-
plished in a systematic manner. Moreover, the multiplicity
of municipal needs that compete for funds makes it impera-
tive that any new program or major program modification be
well-conceived and well-presented.
Programming provides the framework for rational deci-
sions. This framework may be categorized into assessing
existing programs, determining where these programs can be
most effectively applied, identifying new areas of concern,
designing new programs, and scheduling new programs.
Programming options for controlling dust from muni-
cipally-owned paved surfaces are largely limited to public
works or street cleaning functions and activities. To
achieve improved reentrained dust control may require
improved maintenance of existing street cleaning equipment,
purchase of new or different equipment, improvement of
employee training programs, more frequent street sweeping,
and/or modifying the specific sequence of activities in a
given area. It may be desirable to program joint agreements
with neighboring jurisdictions. In addition, programs
related to Areawide Waste Treatment Management Planning
(Section 208, P.L. 92-500) requirements may need to be
integrated with those related to air quality.
According to the International City Manager's Associa-
tion (ICMA), there are four basic ingredients to the
actual process of developing street cleaning programs:
104
-------
0 adequate planning maps,
0 definition of desirable frequency and time
of cleaning for individual streets,
0 development of routes, and
0 scheduling control.
The air pollution control agency should obtain enough of
this type of information from the public works department so
that the agency can interactively determine with the depart-
ment what modifications need to be made as a part of the
effort to reduce reentrained dust from paved surfaces.
Budgeting Mechanisms - Implementing the changes agreed upon
during the process of program development will require mod-
ifications in the budget that would otherwise be prepared.
Such modifications may include the same budgeting mechanisms
as used for other municipal functions: the annual operating
budget and the capital improvements budget.
A reordering of existing budget priorities may be
necessary. For example, if more frequent cleaning is re-
quired in a specific part of the municipality, then either
services will be cut back in other parts of town, produc-
tivity will increase, funds will be shifted to the public
works department from other departments, or an increase in
overall financial resources will be obtained. These al-
ternatives clearly imply that a thorough analysis of how the
requisite changes can be implemented most efficiently is
required. Likewise, additional street cleaning equipment
may have to be purchased. The air pollution control agency
must make every effort to provide the public works depart-
ment with the best possible information concerning the air
pollution control effectiveness of different types of street
cleaning equipment.
Any effort to improve street cleaning practices will
have to be conceived with recognition that funding sources
105
-------
for local governments are limited and that the fundable im-
provements must first be incorporated in the next year's
approved budget, which would prevent very rapid implementation.
Evaluation of Program Effectiveness - The concept of "effec-
tiveness" refers to the degree to which a desired result is
achieved. Thus, there are two types of effectiveness mea-
surements that should be developed and assessed:
0 effectiveness in improving air quality, and
0 effectiveness in implementing the program
elements determined to be necessary to im-
prove air quality.
According to the ICMA and APWA, there are no generally
accepted ways of measuring changes in cleanliness of streets. '
Therefore, the air pollution control agency and the public
works department will have to agree upon some mutually
satisfactory indicators of effectiveness that can be used.
Some possibilities include:
0 volume of material collected,
0 miles of streets swept,
0 curb-miles of streets swept,
0 tons per curb-mile swept, or
0 any other mutually agreeable unit of measurement.
Given the fact that the desired changes will probably focus
upon frequency of cleaning in a given area, it is suggested
that the most useful indicator would be curb-miles swept by
area.
In the end, however, the true measure of the effec-
tiveness of the discussed changes will be air quality im-
pact. It is important that the agency design an adequate
monitoring network and reporting procedures to the public
works department so that changes in air quality can be
correlated with changes in street cleaning practices.
106
-------
IMPLEMENTATION
The preceding two sections have described the legal and
administrative elements of controls based on preventing
material deposition and improving street cleaning. Table 6-
3 summarizes these control measures, possible implementation
methods, and an estimate of the length of time for each to
become operational. Except for improved street cleaning,
implementation would require the adoption of a state regu-
lation or a local ordinance.
In the case of controls for major mud carryout sources,
implementation can be accomplished in one of two ways.
Either the state air pollution control agency can enforce
the measures or the responsibility can be delegated (within
the SIP) to a local agency, possibly a municipal building
inspection or public works department. Where the state is
responsible, they will need sufficient personnel to regularly
inspect all major sites. One or two years is probably
sufficient time to effect this type of regulation.
If a local agency is the responsible party, the muni-
cipality will need to pass ordinances to implement each
measure. For major but temporary sources, such as construc-
tion sites, control could probably be a provision of the
building permit or zoning permit issued to the contractor.
Because the control measures suggested can require purchase
of equipment (street sweepers) or arrangements for site
maintenance (wheel wash mechanisms for trucks), a time
period of one to two years may be required. This allows
contractors the lead-in time necessary for compliance.
For major permanent sources, such as ready-mix plants,
truck terminals, etc., the local air pollution agency could
require operators of problem sources to submit a compliance
plan on how they intend to control trackout and then issue
an enforcement order when the compliance plan is approved.
107
-------
Table 6-3. IMPLEMENTATION OF CONTROL MEASURES
Description of
control measure
Party
involved
Method of
implementation
Estimate of
time required
Control of major
mud carryout sources
Construction sites,
truck terminals,
etc.
ground cover
chemical sup-
pressant
building of
coarse gravel
road
tire wash mech-
anism
tire scrapers
truck tarping
Contractor,
private
owner
wetting of
aggregate
loads
number of acres
under const
soil erosion
for any cleared
area
Const
equip mfgr
Contractor,
state or
local
agency
Contractor
Contractor
Contractor
Local street
ordinances,
provision of
local building
permit, or air
pollution con-
trol agency
regulation
Federal legis-
lation
State highway
regulation and/
or local ordi-
nance
State highway
regulation and/
or local ordi-
nance
Local building
permit or zon-
ing ordinance
Local building
permit or zon-
ing ordinance;
air pollution
control agency
regulation
1-2 years,
possibly as
much as 5
years if some
sort of com-
pliance plan
were involved
for privately-
owned source
108
-------
Table 6-3 (continued). IMPLEMENTATION OF CONTROL MEASURES
Description of
control measure
Party
involved
Method of
implementation
Estimate of
time required
Control of
ubiquitous mud
carryout sources
Unpaved parking
lots
Private
owner
Local building
permit or zon-
ing ordinance;
air pollution
control agency
regulation
3-5 years
Unpaved driveways
Unpaved road
shoulders
Private
owner
State or
local
agency
Local street
ordinance; air
pollution con-
trol agency or
state highway
regulation
Phased control
over a period
of years
Control by
improved street cleaning
More frequent use
of flushers on
major arterials
Use of flushers
in combination
with pickup
equipment
Frequent street
cleaning following
sanding/salting
operations
Operator training
Local
street
depart-
ments
Require reallo-
cation or revi-
sion of street
cleaning priori-
ties; budget
approval, if
equipment has
to be purchased
or if it in-
creases the
total workload
1-3 years
On-street parking
restrictions
Modifications to
curb/gutter con-
figuration
Local
street
depart-
ments
Local
street
depart-
ments
Require a street
use ordinance and
budget approval
for signing
Require budget
approval for
curb and gutter
construction
1-3 years
1-3 years
109
-------
A provision for this type of enforcement action would also
require a local ordinance and a time period of one to two
years.
Controls for ubiquitous mud carryout sources involve
both municipal and private property. Unpaved road shoulders
are usually within the jurisdiction of the state or local
agency and control could reasonably be phased over a period
of years. Control of other sources, such as unpaved parking
lots and driveways, usually involves private ownership.
Passage of the necessary ordinances and allowance of time
for implementation would probably require three to five
years.
In the case of control by improved street cleaning,
there are apparent air quality advantages to be gained with
frequent use of flushers on major arterials, use of a com-
bination flusher/pickup method, and the prompt removal of
sand or salt throughout winter months. If the state air
pollution control agency provides the local public works
department with an understanding of the air quality benefits
that each measure offers, then the municipal department can
redesign its existing street cleaning program to incorporate
these measures. This could be undertaken on a test basis,
over a three to six month period, to give both agencies
involved an opportunity to make necessary modifications.
Where the recommended measures involve expanding the program
to include more frequent cleaning, to cover a greater area
of the city, to purchase new equipment, to increase the
department's staff, or to train operators in the optimum use
of their equipment, more time would be required for imple-
mentation. Budget approval, selection of equipment, requests
for bids, placement and receipt of orders is a lengthy
process but a time frame of one to three years is realistic.
Phasing in new equipment by replacing worn units over a
period of five years is a practical approach and is a
current practice in most municipalities.
110
-------
As outlined above, it is apparent that implementing
some of these control measures will involve not only con-
siderable time but possibly increases in personnel. There
is also the problem of establishing a measure of intergovern-
mental coordination between the cognizant agencies. The
coordination between a state air pollution control agency
and the municipal public works department for setting up a
mutually beneficial street cleaning program has already been
discussed in detail in this chapter. With coordination
between the two agencies, it is possible that the increases
in personnel required can be shared on an optimized basis.
Where trained employees already exist, as in the building
and zoning sections of public works departments, they can
perhaps include some of the inspections in their regular
assignments. For example, in Seattle the street department's
maintenance foremen cover the area daily and radio any
violations of the city's street use ordinance to their
inspectors. It is possible such an intradepartmental pro-
cedure can be followed on an interagency level. ___
One further phase of implementation concerns the tech-
nical difficulties that some of the recommended control
measures might cause. For example, use of tire wash mecha-
nisms that are improperly placed can cause mud carryout
instead of minimizing it. The use of tire scrapers that are
not properly adjusted can cause excessive tire wear. Where
operators of street cleaning equipment are not fully informed
of the optimum removal techniques, considerable material
critical from an air or water pollution aspect can be
missed. Coordination between the state and local agency can
minimize these difficulties.
PUBLIC ACCEPTABILITY
Because successful implementation of any of the controls
is contingent upon the support it would receive from members
111
-------
of the community, an effort was made to determine public
acceptability for such measures. With the cooperation of
the Kansas City, Missouri Action Center, citizen response to
the subject of improved street cleaning methods was examined.
The Action Center records each call received on a service
request form. A total of 170 street service requests were
reported in a three-year period (1974 to 1976). A review of
these requests showed that only a few citizens were con-
cerned with the general street cleaning program. About 90
percent of the calls were requests for litter removal from
the neighborhood street. One instance of citizen interest
was noted where an individual called to suggest that the
city post signs 24 hours prior to street cleaning so that
cars could be moved for more thorough sweeping.
In the spring and summer of 1976, a total of 626 persons
were contacted throughout the city in a citizen attitude
survey on the subject of city government and the services it
provides. On questions relative to condition of city
streets, respondents did not include need for additional
street cleaning as an existing problem. The list of prob^
lems with streets, in order of frequency mentioned, included
the following:
0 chuck holes
0 overhaul/widening needed
0 improper snow removal
0 rough streets
0 poor surfacing
0 problems with curbing
0 trash in streets
0 repairmen leaving holes
Trash on streets was low on the list of priorities and
general cleanliness of streets was not mentioned at all.
112
-------
In an interview with three city councilmen, citizen
attitude regarding the street cleaning program was explored
further. Their responses substantiated much of the input
received during the city's 1976 citizen survey. They did
not feel that there was a problem with the existing program.
They believed that it should serve an aesthetic purpose and
they were particularly concerned that it be vigorously
carried out in the central business district. They also
felt that the primary effort in cleaning should be directed
at removing litter, large spills of mud or sand, and any
accumulation of leaves around catch basins.
Based on the above three indications of public sensi-
tivity to street cleaning, it does not appear that the
general public is concerned about street cleanliness other
than in those instances where street litter is aesthetically
unpleasing or poses a safety hazard. It is not possible to
generalize from this single example, but it is obvious that
citizen concern would be helpful in promoting an expanded
street cleaning program.
113
-------
7. PROCEDURE FOR DEVELOPING A
CONTROL STRATEGY FOR REENTRAINED DUST
Five measures for control of reentrained dust from
paved streets have been identified and described in this
report. A procedure for developing a control strategy
containing one or more of these measures would have a number
of major steps. Some of these steps were discussed in
detail in preceding chapters. The steps are presented below
as a systematic procedure for implementing an effective
control strategy:
1. Define the Problem Area
The first step is to define the area exceeding particu-
late air quality standards and the approximate contribution
from vehicle-related emissions in this area. Many of the
studies which have already identified reentrained dust as an
important source category (references 3 through 14) can be
used for examples of how to inventory and estimate the air
quality impact of reentrained dust. The control strategy
should be designed to have its effect concentrated in the
non-attainment area, although individual control measures
may be implemented on a city-wide basis.
2. Determine Street Cleaning Practices and Locate Major
Mud Carryout Sources
It is important to know the existing level and type of
street cleaning to which modifications might be made. Also,
the overall efficiency of the existing program should be
evaluated, since an increase in budget or manpower in a
114
-------
poorly administered program would not translate into a
significant reduction in particulate concentrations. A
survey of sources contributing to street loadings in the
area is a prerequisite to selection of effective control
measures.
3. Design Modifications in the Street Cleaning Program
In conjunction with the public works department, several
alternative changes should be reviewed. At least two should
be designated for further investigation—one which is expected
to produce the maximum air quality impact and one which will
probably result in a substantial air quality impact with a
minimum expansion of the existing street cleaning program,
The proposed changes should describe types of equipment to
be used, cleaning frequencies by street type and area of
city, weekly equipment-hours and curb-miles by area of city,
projected personnel requirements, and comparable data for
the existing program. The air quality impact can only be
estimated approximately and would probably be based on the
expected percent reduction in emissions and the present
contribution of vehicle-related emissions to air quality in
the affected area. There are no regional-scale atmospheric
dispersion models for particulate currently available to
assess the effect of reentrained dust emission reductions on
ambient concentrations.
4. Consider Control Measures for Preventing Material from
Reaching Street Surfaces
If the survey of the non-attainment area reveals that
there are identifiable mud carryout sources, a regulatory
approach or municipal program for control of carryout should
be considered. At this point in control strategy develop-
ment, it would only be necessary to identify the specific
types of carryout sources to be covered by the proposed
115
-------
program (e.g., unpaved parking areas, unpaved shoulders,
construction sites), time period for its implementation, and
its effect on average street loadings and air quality.
5. Determine Estimated Costs, Public Acceptability, and
Adverse Environmental Effects
Fairly detailed descriptions of each control measure
would be needed in order to prepare a cost estimate. The
experience of the public works department should be utilized
extensively in compiling and selecting cost data. Public
acceptability may be based on the city council's agreement
to modify the street cleaning program. Adverse environment-
al effects should be identified by conferring with the 208
agency and others involved in environmental quality and
control.
6. Calculate the Cost-Effectiveness of Individual Control
Measures
From the data generated in the steps above, the cost-
effectiveness of each measure can be estimated directly.
Any measure with substantially higher costs per unit reduc-
tion in particulate concentrations should be reviewed for
possible replacement, elimination, or modification to make
it more cost-effective. The group of measures in the final
strategy should be capable of attaining the ambient standards
in the area defined in the first step or should contain all
reasonable measures expected to reduce concentrations in
that area.
7. Design a Demonstration Program
A further step that would be worthwhile in developing
a control strategy, but one that is not necessarily essen-
tial, is to design a demonstration program (pilot study) for
a smaller area. The proposed street cleaning or mud carry-
out controls could be tested and air quality impact observed
116
-------
within a limited area. Because of inconsistent results from
the field studies performed to date to evaluate control
measures for reentrained dust, this step would enable the
public works department to test any strategy to reduce
reentrained dust emissions prior to full-scale implementa-
tion. If possible, the test program should be conducted
with existing street cleaning equipment and manpower or, in
the case of mud carryout controls, with controls in the
demonstration area being implemented by the public works
department. This testing could be done while funding and
approvals for the full-scale program were being obtained.
As soon as more definitive data on the effectiveness of
reentrained dust controls become available, this demon-
stration program may not be required and the public works
agency could implement the full-scale program directly.
8. Obtain Funding and Approval for the Full-Scale Program
This step would include incorporation of the proposed
modifications into the budget approval process and the
passage of necessary ordinances or state regulations. In
certain cases, the funding could be contingent on the success-
ful demonstration of an air quality impact in the test area.
It is anticipated that this step would require approximately
one year.
9. Optimize the Control Program
If the demonstration program for a smaller area is
undertaken, there may be some instances where the results of
the study suggest adjustment in the design of the street
cleaning or the proposed mud carryout controls. Prior to
implementation of the control strategy, there should be an
opportunity for findings from the study to be analyzed and
used to adjust any elements of the strategy.
117
-------
10. Monitor the Full-Scale Program to Assess its Air
Quality Impact
It is important that the ambient sampling network in
the non-attainment area be able to adequately determine the
effect of the reentrained dust controls. This may require
the addition of some sampling stations to the regional
network for a period prior to and after implementation of
the strategy, until its impact has been established.
118
-------
8. RECOMMENDATIONS FOR FURTHER STUDY
This document represents the first comprehensive exam-
ination of control techniques for reentrained dust, so it is
anticipated that many more detailed studies will follow.
Some potentially productive areas of investigation which
were identified during preparation of the present document
are:
1. A field study which measures the air quality impact of
da
-------
2. A more detailed study of the potential applicability of
vacuum sweepers. Although vacuum sweepers had good street
cleaning capabilities and ability to remove fine particles
from street surfaces, their use failed to reduce particulate
concentrations. Manufacturers of vacuum street cleaning
equipment would probably cooperate in a demonstration pro-
ject to test operating or mechanical modifications that
would reduce air pollution levels. Possibly an industrial-
size vacuum cleaning unit with a cloth filter on the air
discharge could be tested along with a municipal-size
vacuum cleaning unit to determine whether discharge of fines
reduces the effectiveness of the larger units.
3. An evaluation of several non-conventional street clean-
ing techniques, such as regular winter street cleaning or
sweeping traffic lanes as well as curb areas.
4. Liaison and cooperative field work with construction
industry trade associations on reasonable methods to mini-
mize mud carryout from construction sites. There would be
no need for ambient air sampling in such a study; a good
method for measuring the amount of material tracked from the
sites would provide a better measurement of effectiveness.
5. Development of a street loading measurement procedure
that is accurate, representative of a long section of the
street, replicable, and that can be performed by technician-
level personnel. The American Public Works Association is
also very interested in developing such a procedure.
6. Further investigation of the relationship between street
surface loadings and vehicle-related emission rates. The
project field studies showed no correlation between these
two variables, but another EPA-sponsored study concluded
120
-------
that they were directly proportional. Some definitive
research on the effect of street loadings on reentrained
dust emission rates is needed.
7. Particulate emission tests of motor vehicle exhaust.
There were several indications that a significant portion of
the vehicle-related air quality impact was from direct
emissions rather than reentrained material: 40 percent of
the material was combustion products; the emission rate
decreased with increasing vehicle speed; and there was no
correlation between street loadings and downwind concentra-
tions. The current EPA-recommended particulate emission
r o
factor of 0.34 g/veh-mi for light-duty vehicles is based
on only a single testing study. Therefore, confirmation
of this value appears to be warranted.
8. Adaptation of a line source model such as APRAC or HIWAY
for modeling of suspended particulate. Since vehicle-
related emissions account for such a high percentage of
suspended particulate emissions in urban areas, a much more
accurate representation of the geographic distribution of
these emissions is needed than provided by currently avail-
able area source models. A fallout function should be
included in this line source model.
121
-------
APPENDIX A
CHARACTERISTICS OF
REENTRAINED DUST
-------
CHARACTERISTICS OF REENTRAINED DUST
The project field studies yielded much information on
the factors which affect reentrained dust emission rates.
That information is summarized in nine separate sections in
this appendix.
CONCENTRATION GRADIENTS
The profiles of short-term average particulate concen-
trations away from a street were demonstrated to be quite
similar to those for carbon monoxide, and can be approximated
*
by a Gaussian line source model. However, for the longer
averaging times of interest with particulate concentrations
(24-hour or annual average), changing wind directions and
relatively high "background" levels tend to reduce concen-
tration gradients in the vicinity of streets. There are two
contrasting, idealized traffic patterns which produce dis-
tinctive particulate concentration gradients near streets:
A single high traffic volume street or express-
way with almost no traffic on other streets in
the surrounding area.
A grid network of evenly spaced streets all of
which have approximately the same traffic volumes
*
Use of the line source dispersion equation to estimate
vehicular emission rates is discussed starting on page
137.
123
-------
These two situations and resulting concentration pro-
files are shown in Figure A-l. For the first traffic pattern,
the concentration measured at any point is highly dependent
on the distance of that point from the street. The con-
tribution added by emissions from the street can be roughly
estimated to be inversely proportional to distance from the
street (y - —§—r- where a and b are constants) . In an area
X » u f
with a dense network of streets, concentrations are relatively
independent of distance to the nearest street because proximity
to one street increases distances to other surrounding
streets which also contribute to the total concentration at
that location. Most urban areas have traffic patterns and
street networks that combine the characteristics of these
two idealized examples, so moderate concentration gradients
with increasing distance from the street would be expected.
TRAFFIC VOLUME
In field studies performed to provide data for this
document, a well-defined relationship between traffic volume
and measured concentrations was observed. At the location
where the most traffic counts and air quality measurements
were available for simultaneous time periods, a correlation
of 0.8 for these two variables was found by linear regression
analysis. Within the entire range of traffic volumes occur-
ring on this street, the relationship appeared to be linear,
as shown in Figure A-2. Therefore, the assumption made in
using emission factors for reentrained dust from motor
vehicles—that total emissions are directly proportional to
traffic volume—is borne out.
Although concentrations at a site near any one street
appear to vary quite closely with that street's traffic
volume, the added concentration per vehicle is not consis-
tent from one location to another. Comparison of the data
in Table A-l for two different streets demonstrates this
124
-------
Figure A-l. Particulate concentration gradients near streets,
125
-------
300
e
tn
c
o
•H
•i-l
M
4J
§200
0
c
o
o
tn
nJ
M
0)
100
• lite on east side of street
X site on west side of street
1000 2000 3000 400O 50OO 6000 7000
Traffic volume, 12- or 24-hr
8000
9000
10000
Figure A-2. Traffic volume versus concentration for McGee
Street sites, Kansas City.
126
-------
Table A-l. EFFECT OF TRAFFIC VOLUME ON CONCENTRATIONS
Location/
period
Av. traffic
volume for
period
Arith. av. particulate
concentration,3 ug/m^
Ground level Elevated
(5ft) sites (20ft) sites
McGee Street,
Kansas City
Weekday
Weekend
8 a.m. to 8 p.m.
(day)
8 p.m. to 8 a.m.
(night)
Hamilton Ave.,
Cincinnati
5360
1560
3190
863
140.4
92.2
196.4
100.9
129.6
81.4
177.3
80.6
Weekday
Weekend
8 a.m. to
(day)
8 p.m. to
(night)
8 p.m.
8 a.m.
17792
17325
12682
5110
70
66
72
59
.6
.6
.2
.8
63
60
63
56
.7
.8
.7
.7
All samplers are 35 ft back from edge of street.
127
-------
point; each vehicle on McGee Street has a larger incremental
effect on measured concentration than a vehicle on Hamilton
Avenue. There are two probable explanations for this finding:
0 Hamilton Avenue is the only arterial street in a
primarily residential area while McGee Street is
surrounded by other streets carrying comparable
or greater traffic volumes (similar to the two
examples in Figure A-l). The measured concen-
trations may be showing a relationship with
traffic density in the area surrounding the
sampling sites rather than with traffic volume
on the nearest street.
0 Average emission rates per vehicle may be con-
siderably different on different streets, vary-
ing with such factors as type and condition of
street pavement, traffic speed, or amount of
material on the street surface (see below).
STREET SURFACE LOADING
The recent EPA-sponsored study to develop emission
factors for reentrained dust from paved streets concluded
that the emission rate is directly proportional to the
weight of material on the street surface. This conclusion
was based primarily on samples taken on a street with ex-
tremely high artificial surface loadings. No other previous
investigations of the relationship between surface loadings
and reentrained dust emissions were found, although studies
of accumulation rates of material on streets indicate that
the rate of material removal between rains or cleaning
(assumed to be primarily by reentrainment) increases as the
street loading increases.
In the project field studies, no consistent relationship
was found between street surface loadings (either total or
just the material in the traffic lanes) and particulate
concentrations measured on the same day. Correlation
between these two variables was not improved when the weight
of material on the street subject to direct reentrainment
128
-------
(less than 44 u size) was substituted for total street load-
ing. Even when differences in traffic volumes during different
sampling periods were accounted for by stepwise linear re-
gression, reentrained dust emission rates and ambient con-
centrations near a street appeared to be independent of the
amount of material on the street. Results of the linear
regression analyses are summarized in Tables A-2 and A-3.
It should be noted that there is no standardized pro-
cedure for measuring the weight of material on street sur-
faces. In fact, one of the major conclusions of a recent
American Public Works Association study of street cleaning
was that a standard method is needed so that the efficiencies
of commercial street cleaning equipment can be evaluated.
The procedure used to measure street loadings in conjunction
with the reentrained dust study was to vacuum a 100 ft
length of the street from curb to curb with a large, gasoline-
powered vacuum cleaner, then to transfer the collected
sample with a household vacuum from the cloth bag to a
small, disposable paper container. The procedure is de-
scribed in detail in Appendix D. The collection efficiency
of the sampling and transfer operations combined were checked
and found to be approximately 85 percent, with much of this
loss probably being fines passing through the cloth bag or
released during sample transfer. However, surface loading
values obtained by this method were substantially lower than
those reported in a previous comprehensive study of street
loadings that has been widely used as an information source
on this subject.
The problem of a satisfactory sampling procedure is
compounded by the non-uniform occurrence of material along
the streets. The 100 ft section of street sampled may not
be representative of surface loading for a longer section of
the street being studied, even though surface loadings are
most commonly reported in units of Ib/curb mi.
129
-------
Table A-2. CORRELATIONS BETWEEN STREET LOADINGS AND
AMBIENT CONCENTRATIONS
Street
McGee
McGee
Hamilton
23rd Street
Brush Creek
Cleveland
31st Street
Street
loading No. of
samples samples
Traffic
lanes
Total3
Total
Traffic
lanes
Traffic
lanes
Traffic
lanes
Traffic
lanes
20
11
27
13
20
9
11
Correlation (r) with
particulate concentrations
total street Idg wgt <44u
-0.
-0.
0.
0.
0.
-0.
-0.
218
100
372
263
120
535
007
-0
-0
0
0
0
-0
-0
.237
.409
.176
.236
.192
.700
.155
Curbs and traffic lanes.
Sampling location in Cincinnati. All other locations
are in Kansas City.
130
-------
Table A-3. STEPWISE LINEAR REGRESSION ANALYSIS FOR
SHORT-TERM STUDIES
Dependent variable - downwind added concentration, ug/m
Correlation coefficient
Independent Cone. @ Cone. @ Cone. @
variable 10m 20m 30m
Single correlations
Traffic volume .474 .468 .298
Traffic speed -.262 -.343 -.325
Street loading -.117 .119 .140
Wind speed .405 .284 .289
Stepwise cumulative
Traffic volume
Wind speed
Traffic speed
Street loading
.474
.568
.595
no increase
.468
.561a
.511a
.573
b
b
b
b
Traffic speed is the second variable used.
Data not meaningful because traffic speed (r = -0.325)
was first variable used.
131
-------
There are some possible, but untested, reasons other
than the problems of measurement that could explain the lack
of correlation between street loadings and particulate con-
centrations :
Only a small portion of the material already on
the street surface may be available for reentrain-
ment due to its location in cracks, cementation
to the surface, etc., while a much larger portion
of freshly deposited material is subject to reen-
trainment. This would result in a short average
residence time for most of the material.
Reentrainment may be a function of surface loading
up to some threshold loading above which the rate
of reentrainment is sharply limited by another
unknown factor.
Whatever the explanation, these new data indicate that
the process of reentrainment from paved surfaces is complex.
For example, reducing surface loadings by street cleaning
may not proportionately reduce the air quality impact of
reentrained dust.
WIND SPEED AND DIRECTION
According to the data in Table A-3, wind speed is
directly rather than inversely related to the net (downwind
minus upwind) vehicle-related concentration from streets.
This means that the dust generating and transport effects of
increased wind speed more than offset the diluting effect of
the higher wind speed within the range of wind speeds
encountered.
For one-hour periods with consistent wind directions,
downwind samplers at 10 m from the edge of the road had
average concentrations 82.1 percent higher than concentra-
tions the same distance upwind. Over 24-hour periods in a
relatively open area (Hamilton Avenue sites), the pole
samplers which were downwind most of the day only averaged
132
-------
13.6 percent higher than the upwind samplers because of the
changing wind directions. Where samplers were located in a
street canyon (McGee Street sites), channeling and localized
wind flow during 24-hour periods tended to obscure the
distinct upwind-downwind relationship as determined from
wind directions measured at rooftop level. The pole samplers
on the east side of the street were 9.5 percent higher than
those on the west when they were downwind, but they also
measured 6.7 percent higher concentrations when they were
upwind according to the rooftop wind direction recorder.
Data on upwind-downwind relationships are summarized in
Table A-4.
SAMPLING HEIGHT
A well-defined reduction in particulate concentrations
with increase in height was found at sampling sites near
streets, as shown in Table A-5. At low levels (5 and 20
ft), the reduction in concentration with height was similar
to that determined from more extensive vertical sampling
studies performed in other cities at sites with general
urban exposures. ' The vertical profile of concentra-
tions up to at least 200 ft height from these previous
studies appears to show an inverse power relationship of the
form:
X = ah"b (eq.l)
where x = concentration at height h
a and b = constants
between height and concentration, as shown in Figure A-3.
An inverse power relationship has also been proposed for the
vertical profile of particulate concentrations in dust
66
storms.
133
-------
Table A-4. UPWIND-DOWNWIND RELATIONSHIPS
Sites
Short-term
Study
Days downwind 39
Av downwind 3 153.0
cone , ug/m
Av upwind ., 84.0
cone , ug/m
Av difference 69.0
Difference, per- 82.1
cent of upwind
cone
Std dev of 39.7
difference
Kansas City Cincinnati
East3 West East West
45 28 66 22
123.7 112.3 71.6 69.2
113.0 120.4 60.7 63.3
10.7 -8.1 10.9 5.9
9.5 -6.7 17.9 9.3
15.7 16.2 11.0 7.7
Side of street.
134
-------
Table A-5. CONCENTRATION DEPENDENCE ON SAMPLER HEIGHT
u>
Ul
Pollutant/Site
Av concentration, ug/m
At 5 ft At 20 ft Difference
Std dev of
difference
% reduction
from 5-20 ft
Suspended
particulate
KC east side
KG west side
Cine east side
Cine west side
Lead
KC east
KC west
side
side
127
117
73
66
1.
1.
.6
.4
.2
.4
90
73
116
109
65
60
1.
1.
.1
.0
.2
.9
63
49
11.5
8.4
8.0
5.5
0.27
0.24
15.3
11.2
9.7
7.1
0.15
0.19
9
7
10
8
14
13
.0
.2
.9
.3
.2
.9
-------
5' is assumed to be ground level
in plotting these values
0 5 10 15 20 25
Percent reduction from ground level concentration
Figure A-3. Reduction in concentrations with increase in
height.
136
30
-------
The data points from the reentrained dust field studies
also fit the "slant distance" curve proposed in the National
Assessment of the Urban Particulate Problem reasonably well.
This empirical curve was developed to describe the impact of
nearby major streets on measured concentrations:
'fa (ADT)
h2 + d2
where x = predicted concentration at sampling site
h = height of sampler
d = horizontal distance of sampler from
nearest street
ADT = average daily traffic on street
a and b = constants
The above equation was used with values of a = 45, b = 0.15,
ADT = 10,000, and d = 35 ft to plot the curve of concentration
versus height that is also included in Figure A-3.
The reduction in lead concentrations with height (again
at the 5 and 20 ft sampling heights) was greater than for
total suspended particulate. Comparison of lead and total
suspended particulate are discussed further in a following
section.
EMISSION RATES
Vehicle-related emission rates were estimated from
short-term (1 to 2 hour) high volume samples taken at dis-
tances of 10, 20, and 30 m from the downwind edge of a
street and at 10 m upwind of the street. Only sampling
periods during which the wind direction remained fairly
consistent with the original upwind-downwind sampling con-
figuration for more than 90 percent of the time were used in
calculations. Also, data from four sampling periods were
eliminated because of extreme wind speeds—average wind
137
-------
speeds for two were less than 2.0 mph and for two others
were greater than 16 mph.
The net downwind concentrations at each of the three
distances for each sampling period were substituted into the
Gaussian line source equation below with other appropriate
input data to calculate the apparent line source emission
•strength.
X = ^ (eq.3)
sin c()~v2iT a u
where x = plume centerline (ground level) con-
centration at a distance x downwind
from the highway, g/m^
q = line source strength, g/sec-m
= angle between wind direction and the
line source
a = the vertical standard deviation of
plume concentration distribution at
the downwind distance x, for the pre-
vailing atmospheric stability and
including an initial a of 1.5 m, m
z
u = mean wind speed, m/sec
The calculated values of q were normalized to an emission
rate per vehicle by dividing by the hourly traffic volume
for each sampling period.
The net downwind concentrations and calculated emission
rates per vehicle for 35 sampling periods are shown in Table
A-6. The average emission rate for all this data is 3.7
g/veh-mi and the standard deviation is 3.3, both of which
agree reasonably well with results of the EPA emission
factor development study (average rate for particles less
than 30 u of 8.5 g/veh-mi, range for different land uses of
1.6 to 16 g/veh-mi) and a summary of resuspension studies
which concluded that the resuspension rate for particles
138
-------
Table A-6. CALCULATION OF VEHICLE-RELATED EMISSION
RATES FROM SHORT-TERM SAMPLING STUDIES
Net downwind concentration,
Study
no.
3
4
7
10
12
13
16
18
20
22
23
25
26
27
28
29
30
31
34
37
38
39
40
41
42
43
45
47
49
50
51
52
54
56
57
av.
10m
76.9
46.7
85.3
90.5
34.1
40.2
35.9
16.8
9.6
62.9
42.9
48.7
72.3
73.2
132.1
36.5
80.3
62.1
30.1
46.1
85.5
44.8
166.1
54.7
30.7
73.3
61.2
64.6
71.0
76.4
90.4
90.3
148.6
99.1
98.5
67.9
ug/m-3
20m
40.5
40.6
59.3
54.4
7.8
5.7
20.9
5.5
1.6
37.8
37.0
48.4
40.7
59.9
119.7
23.7
71.3
53.4
2.7
27.6
62.5
28.9
132.4
20.9
15.9
44.0
39.8
29.5
29.6
58.0
57.7
58.4
95.7
51.5
81.0
44.7
Aver emission rate/vehicle,
g/veh-mi
30m
33.3
59.7
. 35.0
39.6
2.1
0
20.2
2.9
16.6
32.0
41.2
15.3
24.5
36.8
61.2
22.4
68.5
41.6
0
10.9
26.9
6.9
74.4
38.8
9.6
30.2
43.4
9.5
22.5
33.4
13.3
36.8
75.4
38.1
36.5
30.3
10m
2.98
2.83
1.65
7.99
2.13
3.38
1.99
1.32
0.42
1.62
6.66
1.10
5.36
2.87
16.67
5.86
6.44
1.46
1.54
5.15
4.31
3.15
3.38
7.84
2.38
5.68
5.01
1.84
6.88
1.23
4.45
1.89
8.12
7.69
4.34
4.21
20m
2.10
3.27
1.54
6.41
0.64
0.64
1.54
0.58
0.09
1.30
7.66
1.45
4.03
3.13
20.16
5.06
7.62
1.68
0.19
4.12
4.18
2.72
3.95
4.00
1.64
4.57
4.33
1.12
3.84
1.25
3.79
1.63
6.98
5.33
4.76
3.63
30m
2.16
5.98
1.13
5.83
0.21
-
1.87
0.38
1.20
1.37
10.66
0.57
3.04
2.42
12. -88
5.99
9.15
1.63
-
2.04
2.25
0.81
2.52
9.28
1.24
3.90
5.92
0.45
3.63
0.90
1.09
1.29
6.87
4.93
2.68
3.26
139
-------
less than 40 microns is in the range of 1 to 5 g/veh-mi.
EPA's final recommended emission factor for reentrained dust
will probably consider all three of these values.
It has been proposed, but not confirmed, that reen-
trained dust emission rates vary as a function of the land
use surrounding a street. Data for the 35 sampling periods
were obtained at four different sampling locations which are
in areas of contrasting land uses, so the average emission
rates for each location were compared to test the land use
hypothesis. As shown in Table A-7, there are variations in
average emission rates from different streets, but the
variations are not as large as the proposed correction
factors based on land use and do not indicate that commercial
streets have lower emission rates per vehicle than residen-
tial streets. Therefore, until more specific data on the
variables affecting reentrained dust emission rates become
available, it is recommended that a single average emission
rate for all streets be used rather than attempting to
adjust rates on the basis of land use or street surface
loadings.
FALLOUT
The data in Table A-6 show generally lower apparent
emission rates with increased distance from the street.
This reduction is attributed to deposition or fallout of
heavier particles.
The deposition of small airborne particles has been
investigated and shown to be a function of ground-level
particulate concentration and settling velocity. The rate
of deposition is best described numerically in terms of a
source depletion factor (q /q ), the ratio between the
J^ U
apparent emission rate (q ) at a distance x downwind and the
X
initial emission rate (qQ). Depletion factors for different
stability classes are presented graphically in Figure A-4.
140
-------
Table A-7. EMISSION RATES IN AREAS WITH DIFFERENT LAND USES
Land use
discription
Short-term study site
123
Undeveloped, Residential,
(no curbs some corn-
on streets) Park mercial
Extensive
commercial
and campus
No. of sampling
periods
Av. emission rate,
g/veh-mi
@ 10m
@ 20m
@ 30m
MRIa emission
factor (< 30u)
based on land
use, g/veh-mi
12
5.32
4.70
4.96
no factor
10
3.92
2.61
2.53
4.9
2.78
2.60
1.56
4.9
8
3.84
3.94
3.89
1.2
Source: Quantification of Dust Entrainment from Paved
Roadways. Midwest Research Institute, Kansas City,
Missouri. Prepared for U.S. Environmental Protection
Agency. March 1977.
-------
0.01
101 102 103 10*
Distance downwind (x), m
Vd2/U2
(eq.4)
Figure A-4. Source depletion factors by stability class for
ground level sources.
Source: Meteorology and Atomic Energy. U.S. Atomic Energy
Commission, Oak Ridge, Tennessee. 1968.
142
-------
For wind speeds (u) greater than 1.0 m/sec or deposition
velocities (V,) greater than 1.0 cm/sec, the source deple-
tion factor can be calculated as follows:
V,9/u
d2 2 (eq.4)
The deposition velocity (V,) may be greater than the
gravitational fall velocity (V ) for the same size particles
because it also includes such nongravitational removal
mechanisms as surface impaction, electrostatic attraction,
adsorption, and chemical interaction. For example, the
average V for the reentrained dust samples should be 1.5
cm/sec (equivalent to a mass median diameter of about 15
microns and an average density of 2 g/cm ). However, the
deposition velocity which matches the sampling data in
Table A-6 best is 5 cm/sec. If this is assumed to be a
representative deposition velocity for reentrained dust,
then depletion of impact would be fairly rapid with distance
away from the street. At one km downwind, the apparent
emission rates would be only 11, 22, or 34 percent of the
initial rates for B, C, or D stability.
The fallout of reentrained dust should be considered in
any regional scale dispersion modeling for suspended particu-
late. Also, for microscale analyses which include the
source depletion factor to account for fallout, the initial
emission rate (q ) corresponding to the average emission
rate of 3.7 g/veh-mi is 4.9 g/veh-mi. Short-term (up to 1
hr) impact downwind of a street can then be calculated with
the line source equation which has the depletion factor
incorporated:
143
-------
x -
siri *2fr a u
b
Where Sc - source depletion factor determined
q . from Figure A-4 and equation 4
V = traffic volume on street, veh/hr
and other variables are as defined in equation 3. Values of
68
d can be estimated from the equation
Z
cfz (x) = ax + aQ (eq.6)
where x = distance downwind/ m
a = 1.5 m
o
a = 0*1120 for C stability,
0.0856 for D stability
b = 0*910 for C stability,
0.865 for D stability
PARTICLE SIZE
Several of the high volume samples from the project
field studies were examined microscopically to determine the
particle size distribution of particulate matter collected
near streets and its percent composition by particle origin.
Of the 66 cellulose acetate filters exposed during the
studies, 38 had light enough loadings that the collected
material could be subjected to direct microscopic examina-
tion at 100 power. The microscope was equipped with a
polarizer and a calibrated Porton reticle. The four dis-
tinct particle origins which could be identified microscopi-
cally were: combustion product, mineral matter, biological
debris, and tire tread.
Percent distribution for each particle size range was
calculated by number of particles and by weight. The
144
-------
percent by number was determined directly from the counts
for each particle size range. The percent by weight was
determined by multiplying the number of particles by the
average particle volume and the weighted average density
(based on composition). Assumed densities were 1.0 g/cm
for combustion products, 2.6 for mineral matter, 1.1 for
tire tread, and 1.0 for biological debris.
The size distributions and compositions of the 38 sus-
pended particulate samples are presented in Tables A-8 and
A-9, respectively.
The particles were observed to be quite large—a mass
median diameter of 15 u and approximately 22 percent by
weight greater than 30 u diameter. The samples were rela-
tively uniform in their size distributions, even though they
were taken at several different locations at various distances
from major streets. Most studies of urban particulate have
indicated mass median particle sizes of less than 5 u. The
large particle size of the reentrained dust supports the
high observed fallout rates described in the previous section.
Most of the samples also displayed a strong bimodal
distribution, with a large percentage of the material being
less than 3 u in diameter and another large percentage in
the 15 to 30 u range. The small particulate (less than 3 u)
was identified as mainly combustion products, while most of
the large material was of mineral origin.
Combustion particles generally constituted more than 80
percent of the number of particles on the filters but, be-
cause they were much smaller than the other particles,
accounted for an average of 40 percent of the weight of sus-
pended particulate. Mineral matter averaged almost 59 per-
cent by weight of the material collected; biological debris
contributed 1.4 percent average; and tire tread particles
were usually a negligible component of the samples.
145
-------
Table A-8.
PARTICLE SIZE DISTRIBUTION OF
HIGH VOLUME SAMPLES
Location
McGee St
Locust St
Oak St
43rd St
50th St
53rd St
Al
A3
A3
A5
A5
A6
A6
A6
A8
A8
B9
B9
B9
BIO
BIO
Bll
Bll
Bll
B12
B12
C13
C13
C.I 3
C14
C14
C14
Stadium Dr D15
Hamilton
D15
Av 2B
2T
2T
3B
3B
3T
3T
Marlowe Av 4
Llanf air
Mean
Std dev
4
Av 5
Date
10-26
10-26
11-23
10-26
11-23
10-05
10-26
11-23
10-26
11-23
10-07
10-27
11-24
10-07
10-27
10-07
10-27
11-24
10-07
10-27
10-07
10-27
11-24
10-07
10-27
11-24
10-07
10-27
11-09
11-09
12-07
11-09
12-07
11-09
12-07
11-09
12-07
11-09
Percent by weight by particle size range
(microns)'
<3.8 3.8- 7.5- 15.0- 30.1- >42.6
7.5 15.0 30.1 42.6
15.3
21.0
14.1
36.8
22.2
46.3
34.6
28.3
20.6
13.4
34.0
21.0
22.4
45.8
20.2
28.4
12.6
21.2
38.0
15.5
37.4
35.2
19.7
31.4
31.5
22.7
24.2
29.8
39.3
34.7
17.9
29.9
38.9
58.7
26.6
46.7
79.0
43.9
30.5
13.5
4.1
5.7
2.7
4.5
5.5
4.3
11.6
5.1
4.5
2.5
5.4
2.6
4.7
6.1
3.5
4.3
1.8
3.4
5.9
3.4
3.7
3.5
3.7
3.4
4.2
5.6
4.2
12.2
9.7
10.5
2.5
5.4
6.9.
10.9
6.2
8.1
6.5
10.2
5.5
2.7
8.6
13.4
9.9
12.0
10.6
11. 5
18.2
12.8
11.5
9.7
19.5
11.8
11.9
19.5
13.3
17.1
7.4
12.3
16.1
13.5
11.8
14.8
11.2
12.9
15.7
12.2
11.6
22.2
11.8
13.2
6.9
14.1
13.0
16.'!
13.3
17.0
9.8
14.6
13.2
3.3
22.5
27.6
38.1
21.8
31.9
24.7
34.4
31.8
20.7
36.8
39.7
35.9
43.7
28.6
35.6
33.9
23.2
34.2
36.1
39.2
34.7
28.5
4'', .4
38.2
30.0
44.7
25.5
22.3
19.2
20.8
21.8
23.3
15.7
14.0
25.1
28.2
4.9
21.9
29.0
9.0
10.3
21.8
17.6
19.1
11.4
11.9
0.5
18.4
17.5
26.6
1.2
15.1
14.4
0
20.5
12.9
21.1
21.2
i:?
16.6
15.5
11.8
15.5
9.7
16.1
7.4
29.9
0.9
19.0
20.4
15.7
27.3
2.9
0
3.0
0
0.8
8.4
12.8
8.6
39.2
10.5
17.6
5.8
18.4
1.3
0.7
3.6
24.9
10.9
0.2
13.6
2.9
0
6.9
3.4
27.9
7.7
2.2
11.8
7.4
6.2
7.5
4.4
2.5
7.4
4.6
12.6
0
0.4
35.2
0
22.6
0
25.8
0
0
1.0
9.1
10.4
146
-------
Table A-9.
SUSPECTED ORIGIN OF PARTICULATE MATTER IN
HIGH VOLUME SAMPLES
Location
McGee St
Locust St
Oak St
43rd St
50th St
53rd St
Origin of particulate matter, % by wt
Combustion Biological Tire
Date products Mineral matter tread
Al
A3
A3
A5
A5
A6
A6
A6
A8
A8
B9
B9
B9
BIO
BIO
Bll
Bll
Bll
B12
B12
C13
C13
C13
C14
C14
C14
Stadium Dr D15
Hamilton
D15
Av 2B
2T
2T
3B
3B
3T
3T
Marlowe Av 4
Llanfair
Mean
Std dev
4
St 5
10-26
10-26
11-23
10-26
11-23
10-05
10-26
11-23
10-26
11-23
10-07
10-27
11-24
10-07
10-27
10-07
10-27
11-24
10-07
10-27
10-07
10-27
11-24
10-07
10-27
11-24
10-07
10-27
11-09
11-09
12-07
11-09
12-07
11-09
12-07
11-09
12-07
11-09
38.0
38.4
39.3
48.9
45.4
58.2
50.3
41.9
37.7
30.8
32.9
35.4
22.8
45.4
35.2
35.5
19.7
33.6
37.2
25.6
28.3
66.7
15.6
43.9
61.8
22.5
29.9
36.6
38.3
30.4
35.5
29.1
55.0
44.5
64.9
38.1
82.8
40.4
39.9
13.8
60.1
60.4
59.3
49.3
52.1
38.7
48.1
53.0
61.2
67.5
66.0
63.7
76.3
54.6
61.8
63.6
. 79.5
64.5
61.6
73.7
70.7
30.1
83.7
54.4
33.6
76.5
69.4
62.8
61.6
69.0
63.7
70.9
44.0
55.3
33.6
61.8
16.1
58.2
58.7
14.3
1.9
1.2
1.3
1.7
2.4
3.1
1.6
5.1
1.1
1.6
1.1
0.9
0.9
neg
3.0
0.9
0.8
1.7
1.2
0.7
0.9
3.2
0.6
li.7
4.6
1.0
0.7
0.5
0.1
0.6
0.8
neg
1.0
0.2
1.4
0.1
1.1
1.4
1.4
1.1
neg
neg
0.1
0.1
0.1
neg
neg
neg
neg
0.1
neg
neg
neg
neg
neg
neg
neg
0.2
neg
neg
0.1
neg
0.1
neg
neg
neg
neg
0.1
neg
neg
neg
neg
neg
neg
0.1
neg
neg
neg
<0.1
<0.1
147
-------
Combustion products were a surprisingly large per-
centage of the particulate at all sampling locations, in-
dicating that exhaust emissions may be higher than previously
suspected. The ratio of exhaust emissions (based on the
published emission rate of 0.34 g/veh-mi) to total vehicle-
related emissions (4.9 g/veh-mi) would indicate an average
contribution closer to 10 percent combustion products rather
/- *\
than the reported 40 percent in the ambient samples. Of
course, some of the measured concentrations are due to
sources other than vehicle-related emissions and a high
percentage of this particulate matter could be combustion
products.
LEAD
The project field studies provided several comprehen-
sive sets of data on ambient lead concentrations near streets,
Average lead concentrations at all sampling sites downwind
of streets were in the range of 0.7 to 2.0 ug/m . The
highest recorded 24-hour concentrations out of 269 samples
were 4.6 and 4.2 ug/m . The highest one-hour concentrations
out of 239 samples (short-term studies) were 5.8 and 3.9
ug/m .
Lead concentrations measured at 20 ft height were
almost always lower than concentrations during the same time
period at 5 ft, but the percent reduction was highly vari-
able. This reduction in lead concentration from 5 ft to 20
ft averaged 11.8 to 35.7 percent at four different loca-
tions. The reductions in lead concentration with height
were greater than for total suspended particulate; no expla-
nation was found for this.
Correlations between measured lead concentrations and
traffic volume during the sampling period were moderate at
sampling sites in all of the studies where traffic counts
were available—in the range of 0.3 to 0.5. Correlations
148
-------
between lead concentration and traffic volume were not quite
as high as correlations between suspended particulate con-
centration and traffic volume. However, traffic volume
appears to be the single variable which most strongly affected
ambient lead concentrations at the sampling sites near
streets.
An average lead emission rate of 0.069 g/veh-mi was
calculated using the same method as that for estimating the
vehicle-related particulate emission rate--a line source
dispersion equation, equation 3, and short-term downwind
ambient lead concentrations. A value of 0.076 g/veh-mi was
determined based on the weighted average lead content of
gasoline, assuming that 70 percent of the lead in the gaso-
line is exhausted from the tailpipe. The calculated
emission rate was quite uniform at the different locations
where the short-term sampling was conducted, but the standard
deviation of the calculated emission rate from all the
sampling periods was 0.047, almost as high as the average
emission rate.
The apparent emission rates calculated at distances of
10 m, 20 m, and 30 m were approximately the same, indicating
that lead was not falling out of the plume within this
distance from the street (in contrast with the deposition of
total suspended particulate described on page A-18).
Lead averaged one to two percent of the suspended
particulate at all sampler locations except the sites which
were always upwind of streets. This consistent lead frac-
tion was attributed to the similar exposures of the sam-
plers, 10 m to 30 m back from the street. Very few indi-
vidual samples had lead less than 0.2 percent of the total
particulate and none were as high as 5.0 percent. Corre-
lations between lead concentration and suspended particulate
concentration at individual sampling sites were in the range
149
-------
of 0.2 to 0.6, so the relationship between the two pollu-
tants was not nearly as strong as indicated by the narrow
range of average lead percentages.
The data for the above analyses of lead concentrations
and more detailed discussion of the analyses can be found in
a companion report to this document.
150
-------
APPENDIX B
MATERIAL ON STREET SURFACES
-------
MATERIAL ON STREET SURFACES
This Appendix describes the derivation of the values
reported in Chapter 2 for deposition and removal rates of
individual processes. The underlying premise for this
approach to the investigation of material on street surfaces
is that a mass balance of total deposition and total removal
rates can be used as a check on the estimated values.
Published information was relied upon to identify the
processes by which materials enter and leave the street
surface and to estimate the rate of flow associated with
each process. Extensive chemical and microscopic analysis
of street loading samples was not used to determine the
origin of the sample components because this procedure could
easily have consumed more resources than the primary effort,
the testing of measures to reduce reentrainment.
The random nature of deposition and removal mechanisms
and their great geographical variations limit the accuracy
that can be achieved in estimating average rates. Replicate
sampling on adjacent roadway sections reveals a standard
22
deviation of about 25 percent just for total weights.
Contributions by individual deposition processes would be
expected to vary by wider margins on adjacent roadway
sections and by orders of magnitude on streets in different
land use areas.
152
-------
.CHARACTERISTICS OF MATERIAL ON STREETS
Some of the pure substances that have been found on
22
street surfaces are:
Gasoline
Lubricating grease
Motor oil
Transmission fluid
Antifreeze
Undercoating
Asphalt pavement
Concrete
Rubber
Diesel fuel
Brake linings
Brake fluid
Cigarettes
Salt
Cinders
Area soil
The total weight of material is normally more than 95 per-
cent inorganic and has a bulk density of about 1.6 g/cm .
Most of the material on street surfaces is too large to be-
come directly airborne, according to published average
particle size data:
Particle size range, u Percent by weight
>2000a 24.4
840-2000 7.6
246- 840 24.6
104- 246 27.8
43- 104 9.7
< 43 5.9
The surface material is not distributed uniformly
across the streets, as shown by the data in Table B-l.
Particles larger than 1/4" are removed from samples before
weighing.
153
-------
Table B-l. DISTRIBUTION OF SURFACE MATERIAL
ACROSS A 'TYPICAL STREET
Street location,
distance from curb
0- 6
6-12
12-40
4.0-96
in.
in.
in.
in.
96 to center line
Normal weight
of material,
% of total
78
10
9
1
2
Weight of .material
just after sweeping,
% of total
45
35
13
6
1
Source;: Sartor, J.. D. and 'G. B. Boyd*. Water Pollution
Aspects of Street Surface Contaminants. U.S. Environ-
mental Protection Agency,, Washington,, B.C. Publication
Number :EPA-R2-72-'081. November 1972.
154
-------
Almost all the dirt and debris collects within a foot of the
curb, presumably due to air currents and direct displacement
by passing vehicles.
Several street loading measurements were made during
the reentrained dust field studies. These are summarized in
Table B-2. With the exception of samples taken on McGee
Street, at least 70 percent of the material on streets was
concentrated in the 2 ft wide curb sections (average 86.5
percent). However, the percent of material less than 44 u
in size was found to average almost twice that previously
reported in the literature.
The total street loadings for these recent measurements
were also significantly lower than those reported in two
other studies which involved extensive street sampling:
Street loading, Ib/curb-mi
No.
Study
samples Mean Median Std. dev.
Range
PEDCo
Shaheen
Sartor and
Boyd17
129
127
n.a.
(>72)
150
320
1500
80
180
n. a.
120
480
1200
9-
8-
500
3400
31-12000
The lower measured loadings could be due to differences
in the sampling methods used or in the types of streets
tested. PEDCo's sampling methodology was similar to those
for the other two studies except that streets were not
flushed after they were broomed and vacuumed. However, the
flushing fraction of samples in the other two studies con-
tained only about four percent of the solids by weight.
Some of the streets sampled to obtain data for this document
were considered to be fairly "dirty" and were located in
warehousing and industrial land use areas.
155
-------
Table 8-2. SUMMARY OF STREET LOADING SAMPLE DATA
Ln
Location
Kansas City
McGee Street'
-A location
-B location
Westwood
Roeland Park
School
Leeds
23rd Street
Brush Creek
Cleveland Ave.
Gillham
31st Street
Cincinnati
Hamilton Ave.
-A location
-B location
No. of
samples
11
11
9
6
6
14
19
11
5
10
14
13
Avg. particle
size dist. , %
<44u 44- 106-
106u 841u
8.3
6.8
9.8
15.4
11.0
12.6
10.5
12.8
13.4
9.0
5.7
5.0
18.5
18.7
25.8
14.4
12.6
11.7
15.5
12.4
15.0
20.9
15.1
15.8
64.5
64.0
49.8
49.5
63.3
51.7
50.2
55.4
59.0
57.4
51.6
52.2
>841u
8.7
10.4
14.5
20.7
13.2
24.0
20.8
lb.8
12.5
12.6
26.0
27.0
Avg. sample wt,
Ib/curb-mi
Curb Traffic
area lanes
139
137
40
69
196
34
32
111
268
322
126
208
6
33
49
14
6
13
10
13
81a
113a
Pet.
in
traf .
lanes
47.6
60.3
12.1
32.2
20.0
28.9
15.6
10.4
3.5
3.8
5.8
3.1
Combined curb and traffic sample.
-------
It should be pointed out that the PEDCo measurements
and Shaheen measurements are in closer agreement than the
Shaheen/Sartor and Boyd values, although the latter data set
has been most widely cited in related work. These differences
in average street loadings reported in the different studies
are important because much of the data on deposition and
removal rates compiled from the literature are based on
these values and are therefore higher than would be con-
sistent with the street loading data in Table B-2.
DEPOSITION PROCESSES _
Most deposition processes are linear with time, or
deposit material at a fairly constant rate. Many of the
processes are closely related to traffic volume, while
others are dependent on immediately adjacent land use.
"Typical" deposition rates for nine processes identified in
the literature as significant contributors to street surface
loadings are presented in Table B-3. The assumption has
been made that the typical street is four-lane, 50 ft wide,
and has an average daily traffic volume of 10,000 vehicles.
The relative importance of vehicle-related deposition pro-
cesses would increase on more highly traveled streets and
decrease on streets with less traffic.
The term "curb-mi" for street loadings indicates the
area bounded by the two curbs for a distance of one mile.
The same unit (curb-mi), when used in reference to street
cleaning, means a distance of one mile along one curb.
Since there are two curb-mi to be cleaned in the length of
street that constitutes one curb-mi for purposes of defining
street surface loadings, the term is obviously quite con-
fusing. In this chapter, it always has the first of the two
definitions.
The reader is cautioned again about accuracy of the
estimates for deposition and removal processes. The values
157
-------
.Table B-3. DEPOSITION PROCESSES
Source
Constituents
Typical depo-
sition rate,
Ib/curb-mi/day
Range,
Ib/curb-mi/day
1, Mud and dirt
carryout
2. Litter
3. Biological
debris
4. Ice control
compounds
5. Dustfall
6. Pavement
wear and
decompo^
sition
7. Vehicle-
related
->Tire wear
•=-Brake and
engine com-
ponent wear
^Settleable"
exhaust
8. Spills
9, Erosion
(runoff and
blowing)
from adjacent
areas
Total
Soil from con-
struction sites,
unpaved parking
areas, etc.
Cans, bottles,
broken .glass,
cigarette butts,
plastic, other
debris
Leaves, grass
clippings,
sticks, animal
droppings, in-
sect parts, etc.
Sand, salt, cin-
ders, calcium
chloride
Atmospheric
faUout.
Asphalt, cement,
aggregate, ex-
pansion joint
compounds and
fillers
Rubber
Metals, lubri-^
cants, brake and
clutch linings
Combustion pro-
ducts, fuel
additives
Sand, dirt,
chemicals
Soil
100
40
20
20
10
10
10
5
No data;
est <2
20
240
Extreme
.'
Extreme
Extreme
0-60
2-25
5-l§0 '
6-50
2-25
1-10
Extreme
158
-------
in Table B-3 are only order-of-magnitude estimates and
should not be considered representative of deposition on a
particular street. There are probably no streets that have
the exact mix of contributions as that calculated based on
typical rates for the individual processes.
Mud and dirt carryout, or tracking, appears to be the
largest source of material on streets. This finding agrees
with those from two other recent studies, Quantification of
Dust Entrainment' from Paved Roadways and Road Dust as
Related to Pavement Polishing (neither of these two studies
provided data to generate the estimate for the deposition
rate from mud and dirt carryout). ' The amount of material
deposited on a street by vehicular carryout on tires or
undercarriages is obviously subject to wide variations
depending on proximity to unpaved areas and the traffic
across these areas.
The estimate of 100 Ib/curb-mi/day was obtained by
averaging the results of two different approaches: (1)
calculating the difference between reported accumulation
22
rates on wet and dry days and (2) determining the amount
of material on vehicle tires and undercarriages which is
subject to easy removal (0.13 Ib/veh). The former approach
indicated an excess accumulation rate of 44 Ib/curb-mi/day,
which did not consider dirt carryout contributions on dry
days, and the latter approach yielded a rate of 163 Ib/curb-
mi/day fo a traffic volume of 10,000 veh/day and an average
trip length of eight miles.
Two additional large sources of material on streets are
litter and biological debris. However, litter and biologi-
cal debris are not major concerns in the dynamics of street
loadings from an air pollution standpoint because these
materials are not usually ground to the size where they
could become airborne and they are efficiently removed by
the other mechanisms. In fact, some analyses of street
159
-------
surface material have excluded full consideration of these
categories by limiting the size of materials included in the
surface loading to less than 1/4 in.
The weight of litter on streets and its accumulation
rate were measured separately from total street loadings in
22
one study. The average deposition rate was 45 lb/curb-mi/
day, but most of the streets sampled had very high traffic
volumes. Adjusted to traffic volumes of 10,000 veh/day, the
average rate was 26 Ib/curb-mi/day. No data were found on
the weight of biological debris in street loading samples,
although some items in this category might be considered
litter in segregating samples.
The application of ice control compounds to streets has
the greatest geographical variation in deposition rates of
all the processes. The estimate in Table B-3 is based on 10
applications per winter of 800 lb/curb-mi each, which is
representative of a moderate winter climate. Sand or salt
added to streets for ice control has more effect on street
loadings than indicated by its relative contribution to
typical deposition rates because it is all concentrated
within a few months during which it frequently causes a
doubling of average street loadings. Also, there is normally
no street cleaning to remove the sand or salt during the
winter. This deposition process is not a factor in southern
climates.
Dustfall contributions to street loadings are based on
the assumption that fallout rates are the same on streets
as they are at the locations where dustfall measurements are
taken. The deposition rate of 10 Ib/curb-mi/day is equiva-
lent to a dustfall rate of 16 ton/sq mi/month. The normal
range of dustfall rates in urban areas is 9 to 40 ton/sq
mi/month.J' Some dustfall measurements taken along road
cross sections have indicated much higher dustfall rates on
72
the road. However, this observation is attributed to the
160
-------
trapping of resuspended material in the dustfall container
and therefore does not represent new material entering the
street deposition/removal system.
Pavement wear and erosion of areas adjacent to the
street are two categories for which very little data on
deposition rates could be obtained. Most available data
indicate that neither process is consistently a major con-
tributor. Calculations based on pavement wear (or polishing)
73
rates show only about 10 Ib/curb-mi/day loss, but some
microscopic examinations of loose surface material have
shown pavement decomposition products to be almost half of
3 73
the sampled material. ' There are two possible reasons
why microscopic examination might yield higher than actual
contributions from pavement wear: the sampling procedure
may abrade the pavement and break loose aggregate particles
that are included in the samples (this has been documented
22
by resampling the same pavement area immediately); or
mineral matter (calcite and quartz) tracked"or blown onto
the road may be mistaken for aggregate from the pavement by
the microscopist.
The most information is available for material deposi-
tion processes directly related to vehicle operation—tire
wear, brake and engine component wear, and the settleable
portion of particulate exhaust emissions. Several studies
of tire wear have all shown the same magnitude of 0.2 to 0.8
g/veh-mi, and most agree that about 90 percent of the tread
loss is deposited on the surface or is initially in the
settleable size range. ' Estimates of deposition
from brake and engine wear were obtained from chemical
analysis of street loading samples and from material balance
22 7R
of expendables. ' They show that these sources contri-
bute an average of about half that of tire wear, or 5
Ib/curb-mi/day for a daily traffic volume of 10,000 vehicles.
Chemical analysis for engine exhaust products which settle
161
-------
to the street surface indicate that deposition accounts for
only about 25 percent of the emissions, with the rest remain-
ing airborne. In particular, most of the lead emitted is of
small particle size and remains suspended.
REMOVAL PROCESSES
Unlike the deposition processes, which occur at a
linear rate with time, most removal processes are inter-
mittent and primarily a function of the amount of material
on the street at the time they occur. In other words, they
operate with a fairly constant efficiency in removing material
from the streets. An apparent exception to this is removal
by reentrainment; according to data presented in Appendix A,
reentrainment rates are independent of street loadings.
The major processes which remove material from streets
are listed in Table B-4, along with estimates of typical
amounts of material removed by each and the assumptions used
in making these estimates.
Reentrainment is shown to be the largest removal pro-
cess. The rate of 100 Ib/curb-mi/day was estimated from a
net emission rate of 4.5 g/veh-mi (total rate of 4.9 g/veh-
mi minus direct emissions of about 0.4 g/veh-mi) and a
traffic volume of 10,000 vehicles per day. Since this
reentrainment rate agrees well with those obtained in other
studies and represents the mass of material that definitely
is emitted from the street, it is probably more accurate
than the estimates of the other removal rates. The supply
of material in the particle size range that can remain
airborne must be continually replenished by the mechanical
fracture of larger particles on the street surface.
A process closely associated with reentrainment is the
displacement of material from streets to just beyond the
curb, either by direct propulsion from the vehicles' tires,
by splashing (when the pavement is wet), or from air turbulence
162
-------
Table B-4. REMOVAL PROCESSES
Process
Typical removal
rate,
Ib/curb-mi/day Assumptions incorporated
Reentrainment 100
Displacement 40
Wind erosion 20
Rainfall runoff 50
Sweeping 35
For 10,000 ADT; net removal
rate =4.5 g/VMT
Estimated from dustfall rate
just beyond curb
Force of same magnitude as
reentrainment, but only
operative 20% of time
Removal efficiencies of 50%
for rain of 0.1-0.5. and
90% for rain of >0.5 in.
Average efficiency of
removal = 50%; weekly clning
163
-------
created by the vehicles. The displaced particles are too
large to remain airborne. One study showed that street
loadings were increased along sections of a street where
barriers at the curb prevented material from leaving the
22
street. Additional evidence of the magnitude of this
process is provided by the rapid soiling of snow beside
streets; a strip of material up to one meter wide is deposited
just off the curb and is highly visible on white snow. The
rate of removal due to displacement was estimated from
dustfall measurements taken right at curbside.
Removal of material by wind erosion is difficult to
distinguish from reentrainment because the material leaves
the street as particulate air pollution in both cases.
Also, the high wind speeds associated with wind erosion
augment reentrainment rates by keeping a greater percentage
of the reentrained material suspended. Wind erosion removes
material in the absence of vehicular traffic at instantaneous
wind speeds greater than about 12 mph, which occur less than
79
20 percent of the time in most parts of the country. This
removal process is important near curbs, where it lifts away
material that may be outside the range of the reentrainment
and displacement processes.
Available information on wind erosion rates are all for
surface areas which are entirely covered with material
subject to erosion, such as plowed fields. These values
cannot be applied directly to street surfaces because
erodible material is found on only a small portion of the
surface area. One report estimated that wind erosion pro-
duced 35 percent as much impact as vehicle-induced reen-
trainment on regional network samplers in Chicago, but this
included wind erosion over the entire urban area.
Rainfall removes material by flushing it from the crown
toward the gutter and then along the gutter to a catch
basin. Some of the material is dissolved, but most is
164
-------
carried in quasi-suspension. This removal mechanism has
been studied in great detail. It was found that the most
important factor in determining amount of removal was rain-
fall intensity—an intensity of more than 0.5 in./hr for a
period of 30 min will remove 90 percent of the surface
loading. A rain sufficient to wet the street completely
and produce flow in the gutter will generally remove about
50 percent of the loading.
If it is assumed that any rain of more than 0.5 in.
total removes 90 percent of the material present and any
rain of 0.1-0.5 in. removes 50 percent of the material,
climatological records for an area can be used to estimate
the number of cleaning occurrences per year. With the
additional assumption of a street loading at the time of the
rains, the total annual removal per curb-mi by rainfall can
be calculated directly and converted to an average daily
removal rate. The surface loading on days when rain occurs
would be greater than the average loading because these days
are at the end of accumulation periods. Depending on which
set of street loadings from page B-4 is used, the surface
loading at the time of rain would range from 300 to possibly
1500 Ib/curb-mi. If a loading of 500 Ib/curb-mi is assumed
for a hypothetical area with 18 days per year of greater
than 0.5 in. rain and 40 days with 0.1-0.5 in. rain, the
total annual removal would be:
(500 Ib/curb-mi)(0.9)(18) = 8100
(500 Ib/curb-mi)(0.5)(40) = 10000
18100 Ib/curb-mi/yr
(50 Ib/curb-mi/day)
In areas with much rainfall or where streets have
higher surface loadings than estimated above, removal from
rainfall flushing could be comparable to that from reen-
trainment.
165
-------
ACCUMULATION RATES
Material deposition on streets is fairly constant with
time and independent of the surface loading. However, the
accumulation rate is not linear, but approaches zero as the
rate of removal increases with the loading. It has been
found that the street loading reaches an equilibrium level,
at which the deposition and continuous removal processes are
about the same, within a fairly short period of time — three
to five days after a rain or street cleaning reduces the
22
loading. ~ If the net accumulation rate did not decrease as
the loading increased, uncleaned streets would eventually
become impassable due to excessive deposits.
The above description of material accumulation on
22
streets is presented in equation form below:
Deposition: — - = k, (eq.7)
dt X
Removal: _ r = k2L (eq.8)
dt
dL = k1-k2L (eq.9)
Net accumulation: dt
Loading (integration k k .
of eq.9): L = =-i - =-! e k2fc (eq.10)
where L = street surface loading at
time t
k.. = deposition rate
k_ = fractional removal
When the rate of removal equals the rate of deposition, the
loading remains constant at its maximum value:
167
-------
L = Lmax
As indicated by equation 12, the loading at any time is
probably best expressed in terms of the street's maximum
loading and the time since rain or cleaning. Although an
attempt was made to derive empirical values for k, and k2
and to incorporate the intermittent removal processes in the
series of equations, the large number of secondary variables
which affect the deposition and removal rates caused this
effort to be dropped. Some of the secondary variables are:
surrounding land use and surfaces, street surface type and
condition, presence of curbing, traffic volume and mix,
public works practices, season, etc.
From a time series of street loading measurements taken
on Hamilton Avenue in Cincinnati, a curve of loading versus
time was constructed, as shown in Figure B-2. The discrete
measurement points are connected by straight lines. There
appears to have been a fairly consistent maximum loading of
about 1000 g/100 ft (116 Ib/curb-mi) during the flushing
period and of 1500 g/100 ft (175 Ib/curb-mi) during the rest
of the study. The accumulation rates for periods which were
not interrupted by rain are presented in Table B-5. When
averaged, the accumulation rates produce the characteristic
curve such as described in equation 12.
168
-------
2500
2000- •
i ! ! I' I I !
A and B denote alternate street sampling
locations on Hamilton Avenue, with
length B usually sampled before cleaning
and length A after cleaning.
19
Figure B-l. Street loadings as a function of time.
-------
Table B-5. ACCUMULATION RATE ON HAMILTON AVENUE, CINCINNATI
Date
10/07-10/08
10/10-10/12
10/14-10/19
10/25-10/26
10/26-10/27
11/04-11/05
11/07-11/09
11/11-11/23
12/09-12/10
12/11-12/14
Days
since
cleaning
0-1
0-2
0-5
1-2
2-3
0-1
0-2
0-12
0-1
0-3
Increase
in loading,
gm/100 ft.
-205
170
355
-160
177
237
991
1032
890
739
Increase
Increase in traffic
per day lanes
-205 -10.1
85 -7.9
71 -0.4
-160
177
237
495
86
890
246
Increase
per day in
traffic lanes
-10.1
-4.0
-0.1
CTl
VD
Av. increase on first day =
Av. increase on second day =
Av. increase on third day =
Av. increase on >3rd day =
307 gm/100 ft
129
240
-------
APPENDIX C
PROJECT FIELD STUDIES
-------
STREET CLEANING IN KANSAS CITY, MISSOURI
Study Design
The primary purpose of this study was to determine the
effect of alternative street cleaning methods on particulate
air quality in the area being cleaned. Two street cleaning
methods were evaluated: broom sweeping and flushing.
Streets in the study area were cleaned once weekly for a
period of one month with each type of equipment. Also, the
streets were not cleaned for a one-month period. Particu-
late concentrations were measured daily at eight sampling
sites in the study area for the entire three months. Average
concentrations associated with each cleaning period were
compared after being adjusted to account for differences in
regional particulate concentrations during the three periods.
The four by seven block study area was located in a
commercial/warehousing district just south of Kansas City's
central business district (CBD). Almost all the streets in
the area have average daily traffic (ADT) volumes of 4,000
to 15,000 vehicles. A six-lane freeway separates this area
from the CBD. Many parking lots within the area are unpaved
or poorly paved. There are also several truck docks, salvage
yards, body shops, and other businesses which might contri-
bute to dirty streets.
Four of the eight samplers were located in one block of
McGee Street, at heights of 5 and 20 ft on two poles on
opposite sides of the street. Much of the auxiliary data
for this study, such as traffic counts and street surface
loadings, were monitored at this location. The other four
samplers were placed on rooftops on the periphery of the
study area, as shown in Figure C-l, for the dual purposes of
determining the impact of the street cleaning at different
points in the study area and of monitoring incoming air
172
-------
i
N
Cro«»town
Freewoy
Samplers at 5'and
20'ht .
A4.A3A
Samplers at 5'and
2°'hl
5200 to 17300 APT on different blocks
AS, A6
AA
2 directional
>amplers(N(S),
windspeed and
direction
2000ADT
j
7500 ADT
II II II || If
Figure C-l. Kansas City street cleaning study area.
Truman Rd.
16th st
17th st
Sampler at 20'ht
A
A8
60OOADT
A
Sampler at . _
20'hl
18th tt
0 200 400 ft
T—'—^P—i—I
SCAIE
-------
quality. Two of the peripheral samplers were directionally-
activated units at the same location, one sampling when wind
direction was from the north (CBD and freeway) and the other
when winds were from the south (commercial/warehousing study
area) .
The day before and the day of each cleaning, street
loading measurements were taken on McGee Street. This pro-
vided an estimate of the removal efficiency of each cleaning
operation. Particle size distributions of material collected
from the street were determined so that cleaning efficiency
by particle size could also be estimated.
The data collected in this study were used in several
other pertinent analyses, such as:
0 changes in air quality and street loadings as
a function of time since last cleaning,
0 effect of rainfall on street loadings and air
quality, and
0 variations in daily concentrations across the
study area.
Data from Study
Particulate concentrations at the four sites on McGee
Street for the three-month sampling period are plotted in
Figure C-2. The data indicate that concentrations were
relatively uniform at the four sites on any one day in
comparison with the day-to-day variations at all the sites.
The four peripheral sites (not shown in Figure C-2) also had
daily concentrations very close to those at the McGee Street
sites, so there was generally little variation in daily
concentrations across the study area. The plot of daily
data also showed a consistent weekly pattern of concentrations
with low concentrations on weekends, slightly higher concen-
trations on Mondays, and usually high concentrations on the
other weekdays. The weekly pattern was attributed to differences
174
-------
300
300
-.200
i I i i t i i i i i i i i
fl
WEST SIDE OF STREET
20" HT
___ 5'HT
O DOWN WIND
Fl JSHING PERIOD
BROOM SWEI PING PERIOD
- ' EAST SIDE OF STREET
tun i tun
I I I I I I I I i I I I I I I > I i I I i I I I I (Mill i i i I i I I i i i i i I I I i I i i
i i i I i I i i i i I i I I i i I I I I I i I i I I I I I I I i I I I
Figure C-2. Daily particulate concentrations for McGee Street sites, Kansas City.
-------
in daily traffic in the study area. Available traffic
counts supported this hypothesis, as previously discussed in
Appendix A (see .Figure A-^2) .
The differences in average concentrations during the
three cleaning periods were discussed in detail in Chapter
3 of the report. The effect of rainfall oh ambient con-
centrations was also discussed at that time.
176
-------
STREET CLEANING IN CINCINNATI
Study Design
The purpose and approach for this study were the same
as for the Kansas City street cleaning study—a comparison
of particulate concentrations in a five by seven block area
during periods with different street cleaning methods. In
Cincinnati, three cleaning methods (flushing, broom sweeping,
and vacuum sweeping) were evaluated for one month each, plus
there was no street cleaning for a noncontinuous one-month
period.
The Cincinnati study area, shown in Figure C-3, was
located in a primarily residential area with one commercial
street, Hamilton Avenue. This street has an ADT of 17,700
but most other streets in the area have less than 1,000
vehicles per day. All streets are curbed and clean in
appearance.
Seven samplers were placed in the study area and
operated daily—four on two poles on opposite sides of
Hamilton Avenue and three on rooftops at the locations shown
in Figure C-3. Street loading measurements were taken on
Hamilton Avenue.
Data from Study
Particulate concentrations at the four sites on Hamilton
Avenue for the four-month sampling period are plotted in
Figure C-4. Concentrations in the Cincinnati study area
were much lower than in Kansas City, and the impact from
traffic-related emissions appeared to be much less. Daily
particulate levels were fairly uniform at the seven sampling
sites, with the ground-level (5 ft) sampler on the east side
of Hamilton Avenue averaging only about 15 ug/m (geometric
mean) higher than the site farthest from that street. The
relatively low readings at the sites nearest Hamilton Avenue
177
-------
00
Figure C-3. Cincinnati street cleaning study area.
-------
•+ WeST SIDE Of STREET
i • • 3'HT
— — — 20'HT
5 27 2
SHT
-r —— — JO'HI
It) 1 I I I I ! I I 1 I I I I I I I I I I I I I I I I I I I 1 I I I I I I I I I 1 _L 1 1.1 I I 1 I I <_L I I I I I I 1 1 I I I i
1 ocl 4 od 3 o
-------
and moderate concentration gradient with distance away from
that major street were attributed to low reentrained dust
emission rates from this clean street compared to the streets
in the commercial/warehousing study area in Kansas City.
No distinct weekly pattern of concentrations was seen.
Differences between weekday and weekend concentrations were
not expected, since traffic data previously summarized in
Table A-l indicated that traffic volumes were approximately
the same for weekdays and weekends.
The differences in average concentrations during the
four cleaning periods were already discussed in Chapter 3 of
the report.
Rainfall obviously reduced concentrations on days when
it occurred. In the Cincinnati study area,.the average
reduction in concentrations on days with 0.1 inch or more of
rain was 18 ug/m ; in the Kansas City street cleaning study,
rain reduced concentrations by 28 ug/m . However, on the
first day after a rain only a minimal residual effect was
observed (e.g., from moist street surfaces) and on subsequent
days no residual effect was noted. A similar impact from
rainfall has been reported for other cities, as shown in
Figure C-5.
If part of the reduction in concentrations on days with
rain is from suppression of reentrained dust and from wash-
ing of street surfaces, then street flushing by trucks
should have a similar effect on particulate concentrations
after the flushing. In the Cincinnati study, the change in
concentrations as a function of days after flushing was
almost identical to that for rain--16 ug/m less than aver-
age on days with flushing and 4 ug/m less than average on
the following day. These data are presented in Figure C-6.
Days with flushing in Kansas City inexplicably showed higher
than average concentrations, although concentrations on the
first days following flushing were slightly lower than
180
-------
o
c
o
o
Q)
-p
(0
iH
3
O
-H
-M
H
(0
Cu
c
•H
QJ
(T
C
(0
50
Days since rain (>0.1 in)
Figure C-5 Effectiveness of rainfall in reducing particulate
concentrations.
50
ro
e
CP
o
o
CJ
o>
r-l
u
i-l
(ti
(U
171
c
-------
average. Based on the curves of concentration versus days
since rainfall and the Cincinnati flushing data, it appears
that daily or every other day flushing might provide a
substantial reduction in ambient concentrations.
The characteristic change in particulate concentrations
following broom sweeper cleaning in Cincinnati was a moder-
ate increase on the day of cleaning, a definite reduction
on the day after cleaning, and then a gradual increase
of concentrations for the next few days, as shown in Figure
C-7. The data from the Kansas City street cleaning study
demonstrated the same changes in concentrations after broom
sweeping except for a much sharper increase in concentra-
tions on the second and third days after sweeping. The
higher than average concentrations on days with broom sweep-
ing are probably caused by the cleaning operations breaking
material loose from the street surface, redistributing it,
and temporarily making more material available for reentrain-
ment.
On days following vacuum sweeping in the Cincinnati
study area, the change in particulate concentrations was
almost identical to that for broom sweeping—a moderate
increase on the day of cleaning, a definite decrease on the
next day, and then a gradual increase in concentrations on
successive days. The curve of concentration versus time
was also presented in Figure C-7. Although vacuum sweepers
do not have the mechanical action against the street surface
that would tend to break down aggregated particles and
redistribute material on the surface, they apparently cause
some temporary increase in the amount of material available
for reentrainment on the day of cleaning. The steady
increase in concentrations starting with the second day
after either broom or vacuum sweeping indicates that clean-
ing at two- or three-day intervals might be optimum from an
air quality standpoint.
182
-------
100
to
C
O
•H
•4-1
10
M
-P
C
a;
u
C
O
u
a;
-P
(0
rH
3
O
•H
-P
V<
10
a
c
-H
0)
(0
u
(U
tr>
(0
M
0)
75
25 -
50
234
Days since cleaning
Figure C-7 Effectiveness of street cleaning in reducing
particulate concentrations.
183
-------
STREET CLEANING IN RESIDENTIAL AREAS
The study design and data for comparison of air quality
in two suburban residential communities with different street
cleaning programs were presented in reasonable detail in
Chapter 3.
184
-------
CONSTRUCTION SITE
Study Design
This study was designed to evaluate the effect on air
quality of potential control measures for reducing the
amount of material deposited on streets from a major mud
carryout source. The control methods evaluated were fre-
quent cleaning of the access area with municipal street
cleaners, and cleaning of the access area manually with
broom and shovel. A one-time tracer study was also con-
ducted at this site to determine how far material is dis-
tributed from its initial carryout point by traffic on the
nearby streets.
Four high volume samplers were placed at three loca-
tions near a building construction site in Kansas City, as
shown in Figure C-8. The two westerly samplers were located
on the roofs of one-story buildings about one block west of
the single, well-defined access point to the construction
site. The third and fourth samplers were placed on the roof
of a building near the eastern edge of the construction site
on 43rd Street. These two samplers were directionally
activated when the wind was from the north (270° to 090°) or
south (090° to 270°) . Locating samplers in both directions
from the access point assured that the impact of mud carryout
would be monitored even if the mud tracking was concentrated
in one direction. The sampling sites to the west were far
enough away from the construction area that fugitive dust
from the site itself did not cause a bias. However, lack of
suitable sampling sites at a corresponding distance east of
the site prevented a similar sampler configuration in that
direction.
A log of construction activities and truck traffic at
the site was maintained during the study period. Truck
traffic was greatest during September and early October (the
185
-------
00
CTi
Oft
t
N
SCALE
JOOft
A
Bll
NEW MOTEL CONSTRUCTION
43rd St.
o
5900V«hitl»»/Doy
B12
B9
AV
610
2 Directional
Hi Voli
Figure C-8. Location of high volume samplers at construction site.
-------
no control period). However, the trackout during late
October, November, and December was comparable because con-
struction workers began using the construction area for
parking.
Data from Study
Particulate air quality in the vicinity of the con-
struction site was greatly affected by excess reentrained
dust caused by material tracked from the site. Concentra-
tions were 40 to 60 ug/m (geometric mean) higher at the
four sites in the study area than at the closest sampler
locations in the regional network.
Comparison of particulate levels during the different
control periods (presented in Table 3-5 of the report) reveals
that concentrations were lowest when the access area had
intensive manual cleaning, next lowest when manual cleaning
was done intermittently, third lowest during the no control
period, and highest when municipal street cleaning equipment
was used along the access street. The results of the muni-
cipal street cleaning period require some further explana-
tion. Three blocks of the access street, 43rd Street, were
flushed on Monday, October 25 and Friday, October 29. The
same area was swept with a broom sweeper on Thursday,
October 28. The air quality data by day indicate that the
street flushing may have increased mud carryout from the
site and the transport of street surface material away from
the immediate area of the construction site access road.
Because material was removed by the broom sweeper only once
during the week, the result appears to have been more material
than usual available for reentrainment.
This construction site study was not a controlled
experiment—the material being tracked from the site was not
constant over time and no method for measuring or quantify-
ing the amount of material carried out could be devised.
187
-------
Also, visual observations of street cleanliness reported in
the log did not always agree with particulate concentrations
at the sampling sites for the same time periods. Any quanti-
tative conclusions from this study must be qualified by
these limitations.
The area of impact around a major mud carryout site was
estimated by means of a tracer study. A mixture of sand and
zinc oxide powder containing 6 Ib of zinc oxide was spread
evenly across 43rd Street at the entrance to the construc-
tion site. Samples of street dust at 50 locations at
distances from 200 ft to 3000 ft from the entrance were
taken one day before, one day after, and eight days after
the tracer was released. Although the data contained scatter
due to nonuniform dispersal of the tracer, there were de-
finite elevated levels of zinc for about 1500 ft along 43rd
Street on the day after the tracer was released. None of
the material collected from surrounding streets showed
significant changes in zinc concentrations on the first day.
However, after eight days, elevated zinc levels were detected
for 400 ft in one direction (west) and 2000 ft in the other
direction on 43rd Street and for 0.5 mi north of 43rd Street
on Broadway and the next major north-south arterial to the
east, Main Street. There were also some high zinc levels on
Broadway and Main Street south of 43rd Street. The tracer
study showed that the distance of tracking is very irregular
and seems to depend on the relative traffic volumes on
different streets carrying traffic away from the mud carryout
source. The major impact area for reentrained dust is prob-
ably confined to a distance of 1000 ft in either direction
along the primary access street.
188
-------
SHORT-TERM SAMPLING STUDY
Study Design
The primary purpose of this study was to obtain con-
current data on partiqulate air quality, street surface
loadings, and traffic volumes at several different test
locations so that general interrelations among these vari-
ables could be determined. It was important to establish a
direct relationship between street loadings and particulate
concentrations because most of the proposed control measures
for reentrained dust are based on reducing average street
loadings by cleaning techniques or by preventing material
from reaching the streets.
Secondary objectives of this study were to investigate
the effect on air quality of:
0 distance from the street curb,
0 traffic speed, and
0 wind speed.
At each test location, the portable high volume samplers
were placed at distances of 10, 20, and 30 m from the street
curb in the direction parallel to the wind direction at the
beginning of the sampling period, as shown in Figure C-9.
Another portable sampler was placed upwind of the street.
The samplers and wind instrument were run for a continuous
sampling period of one to two hours.
Prior to or after the sampling period (so that traffic
flow was not interrupted), the street surface loading on a
cross-section downwind from the test site was measured by
collecting vacuuming samples. With the push type vacuum
device, it was necessary to close single lanes of the
street while the samples were being taken. The weight and
particle size distribution of material collected from the
street was determined in the laboratory.
189
-------
prevailing
wind
direction
upwind
beta
gauge
upwind
high
volume
area for street
loading sample
traffic counter
van
and
generator
beta gauge
indicating
wind
instrument
portable
high
volume
samplers
Figure C-9. Sketch of typical sampling setup.
190
-------
Traffic counts were tabulated by a mechanical counter
and recorded for the total sampling period. The average
traffic speed during the test period was determined by
driving a vehicle along the street with the traffic flow.
Wind direction and velocity was recorded at two minute
intervals during the sampling period.
The high volume samplers require electrical power.
Since the sites were in fairly open areas, a power source
was not usually readily available; therefore, a 5 kw mobile
generator was used as standard fixed equipment in the van.
Data from Study
A total of 60 sampling periods were conducted at five
test sites. For 35 of these sampling periods, wind direction
remained fairly consistent with the original upwind-downwind
configuration more than 90 percent of the time. Data from
these periods, shown in Table C-l, were used in subsequent
analyses. Data for the remaining 25 sampling periods are
presented in Table C-2. These values were not used in any
data analyses because of the difficulty in determining net
downwind concentrations.
The analyses of short-term downwind samples were already
described in Appendix A. The results of multiple linear
regression with traffic volume, street surface loading,
traffic speed, and wind speed as the independent variables
were presented starting on page A-6. Calculation of esti-
mated emission rates from the downwind ambient concentra-
tions, using a line source Gaussian dispersion equation, was
presented starting on page A-15. The important findings are
summarized below:
Concentrations measured 10 m downwind of the
streets were an average of 67.9 ug/m3 higher
than ambient concentrations in the same area.
191
-------
Table C-l. SUSPENDED PARTICULATE CONCENTRATIONS FOR
SHORT-TERM SAMPLING STUDIES WITH
CONSISTENT WIND DIRECTIONS
Study
no.
3
4
7
10
12
13
16
18
20
22
23
25
26
27
28
29
30
31
34
37
38
39
40
41
42
43
45
47
49
50
51
52
54
56
57
av
std dev
Suspended
10 m
upwind
90.0
35.8
54.6
53.4
143.7
351.7
29.9
76.2
90.0
30.7
8.3
71.4
26.3
69.3
56.1
111.4
64.1
120.2
12.2
53.0
43.4
10.2
16.8
53.5
33.3
41.6
66.8
191.8
67.3
26.1
76.0
105.0
179.3
78.3
•85.1
75.0
64.9
particulate
10 m
downwind
167.8
82.5
139.9
143.9
177.8
391.9
65.8
93.0
99.6
93.6
51.2
120.1
98.6
142.5
188.2
147.9
144.4
182.3
42.3
99.1
128.9
55.0
182.9
108.2
64.0
114.9
128.0
256.4
138.3
102.5
166.4
195.3
327.9
177.4
183.6
142.9
72.7
concentration, ug/m
20 m 30 m
downwind downwind
131.4
76.4
113.9
107.8
151.5
357.4
50.8
81.7
91.6
68.5
45.3
119.8
67.0
129.2
175.8
135.1
135.4
173.6
14.9
80.6
105.9
39.1
149.2
74.4
49.2
85.6
106.6
221.3
96.9
84.1
133.7
163.4
275.0
129.8
166.1
119.7
67.1
124.2
95.5
89.6
93.0
145.8
334.0
50.1
79.1
106.6
62.7
49.5
86.7
50.8
106.1
147.3
133.8
132.6
161.8
4.9
63.9
70.3
17.1
91.2
92.3
42.9
71.8
110.2
201.3
89.8
59.5
89.3
141.8
254.7
116.4
121.6
104.5
62.7
192
-------
Table C-2. SUSPENDED PARTICULATE CONCENTRATIONS FOR
SHORT-TERM SAMPLING STUDIES WITH
VARIABLE WIND DIRECTIONS
Study
no.
1
2
5
6
8
9
11
14
15
17
19
21
24
32
33
35
36
44
46
48
53
55
58
59
60
av
std dev
Suspended
10 m
upwind
73.0
77.8
130.0
98.6
31.4
44.8
52.6
78.6
114.8
103.0
54.8
66.6
32.3
24.1
78.5
60.4
95.8
69.6
74.4
136.7
105.3
388.6
324.8
256.3
254.0
113.1
93.3
particulate
10 m
downwind
62.7
81.6
105.1
120.5
37.0
67.1
47.6
205.0
129.8
76.7
104.9
87.7
42.5
31.0
72.2
77.0
68.1
73.2
136.1
269.4
199.5
550.1
330.7
259.6
255.5
139.6
119.0
concentration, ug/m
20 m 30 m
downwind downwind
61.9
71.1
89.9
95.4
30.2
57.2
49.8
189.1
103.7
58.4
86.3
58.7
31.0
35.0
52.0
58.6
56.7
84.8
94.8
220.2
147.4
471.9
295.6
196.5
188.6
115.0
100.7
56.5
68.5
90.0
85.6
29.3
52.6
48.2
185.0
105.7
47.5
93.8
65.5
48.1
36.7
54.1
38.9
43.3
82.4
94.7
184.6
115.5
437.0
264.4
193.9
203.3
109.0
92.6
193
-------
Concentrations were more closely related to
traffic volume along the street than to any
of the other variables evaluated.
No relationship was found between street
surface loadings and particulate concen-
trations near the street.
Wind speed had a direct rather than inverse
relationship with the net downwind concen-
trations. Although the correlation was low
(0.28 to 0.40), it indicated that the dust
generating and transport effects of increased
wind speed more than offset its diluting
effects.
There was a slight inverse relationship
between traffic speeds and concentrations.
The average emission rate calculated from
the downwind concentrations was 3.7 g/veh-
mi, with a standard deviation of 3.3 g/veh-
mi.
Variations in emission rates from four
different streets were not very great.
The deposition of material reentrained
from streets appears to follow traditional
particulate deposition curves in which the
rate of fallout with distance is a function
of atmospheric stability, wind speed, and
the settling velocity of the particles.
The settling velocity which best fit the
data was 5 cm/sec.
194
-------
APPENDIX D
METHOD FOR MEASURING
STREET SURFACE LOADINGS
-------
METHOD FOR MEASURING
STREET SURFACE LOADINGS
OBJECTIVES AND SCOPE
A procedure has been developed to enable the quantita-
tive measurement of material loadings (in Ib/sq ft and
Ib/curb-mi) and the particle size distribution of deposited
material on paved streets.
The methodology is similar to that of the measurement
procedure developed by EPA's Office of Research and Develop-
ment and several key design elements have been retained.
Most of the operational aspects are dissimilar, however,
since the EPA procedure was developed for analyses associated
with water quality applications, in contrast to the current
method, which will be utilized in conjunction with field
studies on the air quality impacts of reentrained dust from
streets.
In particular, the variables to be measured are:
the total street surface material loading
on a 100 ft length section of street over
a cross-section width measured from curb
to curb, to include all naturally* occur-
ing material found thereon;
the percent of material in the active
traffic lanes (defined as the total area
extending across the crown of the street
to within approximately 24 in. of each
*
Street dust, dirt, leaves, bottle caps, etc. Obviously,
a cement block which has fallen off a truck is not
naturally occurring.
196
-------
curb) versus the percent accumulated
within 24 in. of the two curbs, to be
determined by the acquisition of two
separate samples;
the particle size distribution of the
curb and traffic lane samples;
PARAMETERS TO BE MEASURED AND EQUIPMENT
1. Street surface particulate broom
loadings ROOF-push vacuum
vacuum with disposable
sample bags
portable generator
chalk/tape measure
traffic safety cones
2. Particle size distribution rotary sieves (to 37/um)
analytical balance
DESIGN
Samples of materials deposited on streets will be
collected using a combination of sweeping (broom) and
vacuuming techniques to determine the total particulate
street surface loading and the size destribution of said
material on any particular 100 ft test section. The pro-
cedure assumes that the street loading is uniform along any
given street which is otherwise free from biasing due to
localized mud trackout, etc. Further, it is assumed that a
100 ft test section will produce an accurate and reproduc-
ible sample.
The material removed will be collected in a stepwise
fashion and segregated into two samples: the fraction
deposited within 24 in. of both curbs and the fraction
deposited over the remaining surface of the street. The
procedure entails a preliminary brooming of the 2-24 in.
wide 100 ft curb areas, followed by a thorough vacuuming
with a push-vac. The collected material constitutes one
sample. The remaining road surface area over the active
traffic lanes will then be vacuumed (not initially broomed)
197
-------
with the push-vac to gather the second sample for the site.
It is emphasized that final flushing is not employed to
collect the residual matter remaining after vacuuming because
the primary purpose of flushing is to collect water soluble
components present in low mass quantities. The present
procedure was designed for the collection of particulate
matter only, so it was determined that flushing was unnecessary.
The designs of the fixed street cleaning studies in
Kansas City, Missouri and Cincinnati require that at least
two street loading samples be taken per week, one before the
streets are swept and one after. The locations for taking
the samples will be chosen such that they are not immediately
fronting any of the hi vol samplers in the study area. This
will minimize the introduction of bias into the measured air
quality at these locations due to the possibility of the
street loading sampling (actually a sweeping and cleaning
operation) producing artificially clean streets and low
particulate concentrations adjacent to the sampler. In
addition, there will be two different areas chosen for
street surface sampling, one for the "before sweeping"
sample, and one for the "after sweeping" sample. 'The two
100 ft sections should preferably be on the same street,
since a basic assumption with this procedure is that loadings
are uniform over a given street section subject to the same
traffic, environmental influences, and sweeping technique
and patterns. Using two areas in this manner will ensure
that a representative sample is taken both before and after
the streets are swept.
For the mobile studies conducted in Kansas City, six
street loading sampling areas will be chosen on streets with
medium traffic densities and bordered by expanses of open
areas where air sampling equipment may be extended and set
up. The medium traffic flow will ensure ease and rapidity
of sample collection, facilitate personnel safety, and
198
-------
result in minimum disruption of traffic flow. Of course,
the streets will be chosen so as to minimize biases to the
sample collection process due to the proximity of storm
drains, crosswalks, steep grades, heavy parking activity,
active mud carry out sites, or other activities which could
result in acquisition of a nonrepresentative sample.
The most dominant aspect of this design for the street
loading sampling procedure is the utilization of the multi-
purpose ROOF-Groundskeeper--a gasoline powered, push type,
vacuum machine. This machine incorporates a powerful 5 hp
high rpm (3600) engine to drive a high volume centrifugal
impeller which provides the necessary suction to pick up
virtually 100 percent of the material encountered on a
street surface, including rocks and litter. Most importantly,
it greatly reduces the time required for sample collection
over large areas and provides for operator safety by limit-
ing exposure time in the active traffic lanes of the street.
The machine is shown in Figure 1.
Material is collected in a large fabric dust bag behind
the machine, as shown. Fine particulates are trapped in the
interstitial fibers of the fabric. After collection, the
sampled material must be transferred to a suitable sample
storage container until weighing and analysis takes place.
This is accomplished by vacuuming the material from the
inside of the bag with a standard household type vacuum
cleaner. The material is collected in a disposable type
filter bag which can be removed from the vacuum, sealed, and
retained quite conveniently for analysis. It was estimated
that this two-step procedure is approximately 85 percent
efficient at collecting material from the street surface.
This efficiency was determined by collecting material
directly with the household vacuum over a 50 ft length of
street and comparing the weight with that obtained using the
standard collection procedure on an adjacent 50 ft section.
199
-------
D-l. Machine used for collection of street loading
samples.
200
-------
The samples will be shipped weekly to PEDCo's laboratory
in Cincinnati where they will be weighed to determine the
street surface loading and analyzed to determine the particle
size distribution. Initially, several sample bags will be
equilibrated and weighed so that an average tare bag weight
may be determined. In this fashion, the net sample weight
may be calculated. It is estimated that the weight of the
bags will be insignificant (1 oz) as compared to the weight
of the sample (1-2 Ib); therefore, variations in bag weight
should be relatively unimportant.
201
-------
REFERENCES
1. Selected and Annotated Bibliography for Resuspended
Particulate Material from Pavement. PEDCo-Environmen-
tal, Inc., Kansas City, Missouri. Prepared for U.S.
Environmental Protection Agency, Kansas City, Missouri.
December 1976.
2. Roberts, J. W., A. T. Rossano, Jr., P. B. Bosserman, G.
C. Hofer, and H. A. Watters. The Measurement, Cost,
and Control of Traffic Dust in Seattle's Duwamish
Valley. Puget Sound Air Pollution Control Agency.
(Presented at Annual Meeting of the Pacific Northwest
International Section of the Air Pollution Control
Association. Eugene, Oregon. November 15-17, 1972.)
28 p.
3. Harrison, P. R., R. G. Draftz, and W. H Murphy.
Identification and Impact of Chicago's Ambient Sus-
pended Dust. Unpublished. 15 p.
4. Draftz, R. G. and H. W. Blakeslee. Identification of
Ambient Suspended Particles from Philadelphia. IIT
Research Institute and Scott Research Laboratories,
Inc. (Presented at Annual Pacific Northwest Section of
Air Pollution Control Association. Seattle, Wash-
ington. November 28-30, 1973.) 10 p.
5. National Assessment of the Urban Particulate Problem.
Volume I: Summary of National Assessment. U.S. Envi-
ronmental Protection Agency, Research Triangle Park,
North Carolina. Publication Number EPA-450/3-76-024.
July 1976.
6. The Environmental Flow of Cadmium and Other Trace
Metals. Purdue University, West Lafayette, Indiana.
Progress Report July 1, 1974 to June 30, 1975.
7. Analysis of Probable Particulate Non-Attainment in the
Kansas City AQCR. PEDCo-Environmental, Inc., Cincin-
nati, Ohio.. Prepared for U.S. Environmental Protection
Agency, Kansas City, Missouri. Februray 1976.
8. Particulate Area Source Emission Inventory for Nash-
ville-Davidson County, Tennessee. PEDCo-Environmental,
Inc., Cincinnati, Ohio. Prepared for U.S. Environmen-
tal Protection Agency, Atlanta, Georgia. February
1975.
202
-------
9. Control of Suspended Particulates, Particulate Resus-
pension Study. New Jersey Department of Environmental
Protection, Trenton, New Jersey. Quarterly Progress
Report January 1976.
10. Air Quality Modeling in Jacksonville, Florida. Volume
I. PEDCo-Environmental, Inc., Cincinnati, Ohio. Pre-
pared for U.S. Environmental Protection Agency, Atlanta,
Georgia. May 1976.
11. An Implementation Plan for Suspended Particulate Matter
in the Phoenix Area. Interim Report: Emissions Inven-
tory. TRW Systems, McLean, Virginia. Prepared for
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina. May 1976.
12. A Study of the Suspended Particulate Problem in the
Duwamish Basin. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. Publication
Number EPA-68-02-1499. May 1975.
13. Colorado Air Quality Maintenance Area Analyis. [Draft
Chapter 3.] PEDCo-Environmental, Inc., Cincinnati,
Ohio. Prepared for U.S. Environmental Protection
Agency, Denver, Colorado. December 1976.
14. Air Quality Modeling in Hillsborough, Pinellas, and
Polk Counties, Florida. Volume I. PEDCo-Environmen-
tal, Inc., Cincinnati, Ohio. Prepared for U.S. Envi-
ronmental Protection Agency, Atlanta, Georgia. July
1976.
15. Quantification of Dust Entrainment from Paved Roadways.
Midwest Research Institute, Kansas City, Missouri.
Prepared for U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. March 1977.
16. Resuspended Street Dust. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. Unpub-
lished. 50 p. February 1976.
17. Sartor, J. D. and G. B. Boyd. Water Pollution Aspects
of Street Surface Contaminants. U.S. Environmental
Protection Agency, Washington, D.C. Publication Number
EPA-R2-72-081. November 1972.
18. Idaho Particulate Matter Study. Draft Final Report.
GCA/Technology Division, Bedford, Massachusetts. Pre-
pared for U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. October 1973.
203
-------
19. Characterization of Particulate Sources Influencing
Monitoring Sites in Region VIII Non-Attainment Areas.
U.S. Environmental Protection Agency, Denver, Colorado.
Publication Number EPA 908 1-76-007. June 1976.
20. Brant, L. A. A Preliminary Report on the Study of the
Abatement of Air Pollution from Winter Sanding Opera-
tions in Montana. March 1972. Unpublished. 17 p.
21. Warner, P. 0., J. 0. Jackson, S. Schaldenbrand, and B.
Raphael. Effects of Street Salting on Ambient Air
Monitoring of Particulate Pollutants in Detroit. (Pre-
sented at Sixth Central Regional Meeting, American
Chemical Society. Detroit, Michigan. April 21-24,
1974.) 15 p.
22. Shaheen, D. G. Contributions of Urban Roadway Usage to
Water Pollution. Biospherics, Incorporated, Rockville,
Maryland. Prepared for U.S. Environmental Protection
Agency, Washington, D.C. Publication Number PB245 854.
March 1975.
23. Brenner, R. and J. Moshman. Benefits and Costs in the
Use of Salt to Deice Highways. The Institute for
Safety Analysis, Washington, D.C. November 1976.
24. The 1975 Survey of Street Cleaning Practices. American
Public Works Association, Research Foundation, Chicago,
Illinois. Unpublished.
25. Madden, J. L. The Use of Salt in Ice and Snow Control
in Rochester, A Cost-Benefit Study. Systems Analysis
Program, University of Rochester, New York. Working
Paper Series Number 7124. May 1971.
26. Murray, D. M. and U. F. W. Ernst. An Economic Analysis
of the Environmental Impact of Highway Deicing. U.S.
Environmental Protection Agency. Publication Number
EPA 600/2-76-105. May 1976.
27. Personal communication with W. G. Dickinson, Salt
Institute, Alexandria, Virginia. March 30, 1977.
28. O'Brien, L. G. Winter Snow and Ice Control. Township
News, Pennsylvania State Association of Township Super-
visors. 28:9. September 1975.
29. O'Brien, R. G. Road Salt—One Town's Cost Analysis.
Better Roads. p. 21-23. July 1973.
204
-------
30. Personal communication with City of Boston, Public
Works Department, Boston, Massachusetts. March 22,
1977.
31. Personal communication with Mr. Sawyers, Department of
Public Information, Indianapolis, Indiana. March 29,
1977.
32. Personal communication with M. Enquist, Street Mainte-
nance Department, Minneapolis, Minnesota. March 22,
1977.
33. Hubbard, W. Sensible Salting Saves Dollars and Envi-
ronment. Rural and Urban Roads. July 1972.
34. Jermano, J. P. Cost Comparison of Alternate Ice Con-
trol Methods. Public Works. February 1976.
35. Personal communication with J. Gould, Pavement Cleaning
Service, Charlotte, North Carolina. February 1977.
36. Pitt, R. E. Control Measures for Reducing the Accumu-
lation of Street Surface Contaminants. Woodward-Clyde
Consultants, San Francisco, California. Prepared for
National Science Foundation and American Public Works
Association. December 1976.
37. Requirements for Preparation, Adoption, and Submittal
of Implementation Plans. 40 CFR Part 51, Subpart B,
Section 51.11{f); Subpart C, Appendix B, Section 2.2.
Federal Register, Saturday, August 14, 1971.
38. Colorado Air Quality Control Regulations and Ambient
Air Quality Standards. Colorado Air Pollution Control
Commission, Colorado Department of Health, Denver,
Colorado. 1974.
39. New York Proposed Air Pollution Fugitive Dust Regula-
tions. Part 210. July 28, 1976.
40. Compilation of Federal, State, and Local Laws Control-
ling Nonpoint Pollutants. U.S. Environmental Protec-
tion Agency, Washington, D.C. Publication Number EPA-
440/9-75-011. September 1975.
41. North Carolina General Ordinances of the State Highway
Commission. Raleigh, North Carolina. 1972.
42. North Carolina Highway Specifications Code. Section
107-13.
205
-------
43. Maryland Air Pollution Control Regulation. October 18,
1974.
44. Personal communications with L. McCarl, State of Mary-
land, Highway Division, Annapolis, Maryland. September
1976 and February 1977.
45. Polk County, Iowa Air Pollution Control Regulation.
Article IX, Section 5-23. November 3, 1972.
46. Allegheny County, Pennsylvania Air Pollution Control
Regulation. Section 1809. June 15, 1972.
47. Milwaukee County, Wisconsin Air Pollution Control Regu-
lation. Section 89.12. May 1974.
48. District of Columbia Air Pollution Control Regulation.
Section 8-2:712. March 12, 1972.
49. Cincinnati, Ohio Air Pollution Control Regulation.
Section 1001-3-F8 and 1007-11. June 11, 1976.
50. Code of the City of Rapid City, South Dakota. Article
V, Section 5, B.5.
51. Personal communication with J. J. Coover, City of
Seattle, Washington. January 1977.
52. Puget Sound Air Pollution Control Agency, Seattle,
Washington. Guidelines for Control of Air Pollution
from Parking Lots, Roadways, and Open Areas. Janurary
1976.
53. Personal communication with City of Charlotte Public
Works Department, Public Services and Information
Department, and Clean City Committee on several occa-
sions in late 1976 and early 1977.
54. City of Charlotte Community Improvement Ordinances,
Litter Control, Chapter 17, Article 1, Section 17-29.1.
55. Personal communication with City of Chicago Department
of Building, Department of Building Permits, and
Department of Records Administration. November 1976
and February-March, 1977.
56. Personal communication with City of Kansas City,
Missouri City Clerk's Office. January 14, 1977.
57. Personal communication with City of Kansas City,
Kansas City Attorney's Office. January 14, 1977.
206
-------
58. Personal communication with H. Thompson, Bureau of
Street Sanitation, City of Milwaukee, Wisconsin.
January-February, 1977.
59. Personal communication with Ron Pinkowski, City of
Chicago, Department of Building Permits. January 19,
1977.
60. Korbitz, W. E. (ed.) Urban Public Works Administra-
tion. Washington, D.C., International City Management
Association, 1976.
61. Hinkle, G. Performance Specifications for Street
Cleaning Equipment. American Public Works Association,
Research Foundation. (Presented at International
Public Works Congress and Equipment Show. Las Vegas,
Nevada. September 25-30, 1976.)
62. Compilation of Air Pollutant Emission Factors. U.S.
Environmental Protection Agency, Research Triangle
Park, North Carolina. Publication Number AP-42. March
1975.
63. Ter Haar, G. L., D. L. Lenare, J. N. Hu, and M. Brandt.
Composition, Size and Control of Automotive Exhaust
Particulates. J. Air Pol. Control Assn. 22:39-46,
January 1972.
64. Study of Suspended Particulate Measurements at Varying
Heights above Ground. Texas State Department of Health,
Air Control Section, Ambient Monitoring and Sampling
Group. Unpublished. 13 p.
65. Mainwaring, S. J. and S. Harsha. University of Mel-
bourne, Australia and University of Roorkee, India. A
Study of the Distribution of Particulates with Height.
p. 471-491.
66. Chepil, W. S. and N. P. Woodruff. Sedimentary Charac-
teristics of Dust Storms: II. Visibility and Dust
Concentration. Amer. J. Sci. 255:104-114, February
1957.
67. Meteorology and Atomic Energy. U.S. Atomic Energy
Commission, Oak Ridge, Tennessee. Publication Number
TID-24190. July 1968.
68. Sklarew, R. C., D. B. Turner, and J. R. Zimmerman.
Modeling Transportation Impact on Air Quality. U.S.
Environmental Protection Agency, Research Triangle
Park, North Carolina. January 1973.
207
-------
69. EPA's Position on the Health Implications of Airborne
Lead. U.S. Environmental Protection Agency, Washing-
ton, D.C. November 1973.
70. Lead Analysis for Kansas City and Cincinnati. PEDCo-
Environmental, Inc., Cincinnati, Ohio. Prepared for
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina. March 1977.
71. Shuler, L. M. and R. R. Hegmon. Road Dust as Related
to Pavement Polishing. The Pennsylvania State Univer-
sity, University Park, Pennsylvania. Prepared for
Pennsylvania Department of Transportation. Automotive
Safety Research Program Report S49. May 1972.
72. Sylvester, R. 0. and F. B. DeWalle. Character and
Significance of Highway Runoff Waters—A Preliminary
Appraisal. University of Washington, Department of
Civil Engineering, Seattle, Washington. Prepared for
Washington State Department of Highways and Federal
Highway Administration. December 1972.
73. Dahir, S. H. and W. E. Meyer. Bituminous Pavement
Polishing. The Pennsylvania State University, Univer-
sity Park, Pennsylvania. Prepared for Pennsylvania
Department of Transportation. Automotive Research
Program Report S66. November 1974.
74. Dannis, J. L. Rubber Dust from the Normal Wear of
Tires. Rubber Chem. Technol. 4_7:1011-1037, September
1974.
75. Pierson, W. R. and W. W. Brachaczek. Airborne Particu-
late Debris from Rubber Tires. Rubber Chem. Technol.
4_7 (5) :1275-1299, December 1974.
76. Subramani, J. P. Particulate Air Pollution from Auto-
mobile Tire Tread Wear. Doctoral dissertation. Uni-
versity of Cincinnati, Cincinnati, Ohio. 1971.
77. Cardina, J. A. Particle Size Determination of Tire-
Tread Rubber in Atmospheric Dusts. Rubber Chem.
Technol. £7:1005-1010, September 1974.
78. Cahill, T. A. and P. J. Feeney. Contribution of Free-
way Traffic To Airborne Particulate Matter. California
University, Crocker Nuclear Laboratory, Davis, Cali-
fornia. April 1974. 167 p.
208
-------
79. Skidmore, E. L. and N. P. Woodruff. Wind Erosion
Forces in the United States and Their Use in Predicting
Soil Loss. U.S. Department of Agriculture, Washington,
D.C. Agriculture Handbook Number 346. April 1968.
80. Horton, J. P. Protect Streams Economically with Vacuum
Street Cleaners. (Presented at Region II Spring
Meeting of the American Public Works Association. New
York, New York. April 26, 1976.)
209
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-907-9-77-007
4. TITLE AND SUBTITLE
Control of Re-entrained Dust from Paved Streets
7. AUTHOR(S)
Kenneth Axetell
Joan Zell
9. PERFORMING ORGANIZATION NAME AND ADDRESS
PEDCo Environmental, Inc.
2480 Pershing Road
Kansas City, Missouri 64108
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Air and Hazardous Materials Division - Region VI]
1735 Baltimore Avenue
Kansas Citv, Missouri 64108
15. SUPPLEMENTARY NOTES
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
August 10.77
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-1375, Task 35
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
16. ABSTRACT
This report is a comprehensive review on the topic of re-entrained dust from
paved streets. The purpose is to evaluate control measures for reducing emis-
sions of re-entrained dust.
17. . ' KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS b.lDENTI
Air Pollution
Re-Entrained Street Dust
Fugitive Dust
Emission Factor
Total Suspended Particulate
Sampling/Measurement
Monitors
13. DISTRIBUTION STATEMENT 19. SECUf
Unlimited Unc
20. SECU
Unc"
FIERS/OPEN ENDED TERMS C. COS AT I Field/Group
=MTY CLASS (This Report) 21. NO. OF PAGES
assified 221
=t IT Y CLASS (This page) 22. PRICE
assified
EPA Form 2220-1 (9-73)
------- |