EPA-R2-72-081
November 1972 Environmental Protection Technology Series
Water Pollution Aspects of
Street Surface Contaminants
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were"established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
H. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This v?ork provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards..
-------
EPA-R2-72-081
November 1972
WATER POLLUTION ASPECTS
OF STREET SURFACE CONTAMINANTS
By
James D. Sartor and Gail B. Boyd
Contract No. 1^-12-921
Project 1103^ FUJ
Project Officer
Francis J. Condon
Municipal Pollution Control Branch
Environmental Protection Agency
Washington, B.C. 20^60
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, B.C. 20^60
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $3.00
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommenda-
tion for use.
ii
-------
ABSTRACT
Materials which commonly reside on street surfaces have been found to
contribute substantially to urban pollution when washed into receiving
waters by storm runoff. In fact, runoff from street surfaces is
similar in many respects to sanitary sewage. Calculations based on a
hypothetical but typical U.S. city indicated that the runoff from the
first hour of a moderate-to-heavy storm would contribute considerably
more pollutional load than would the same city's sanitary sewage during
the same period of time.
This study provides a basis for evaluating the significance of this
source of water pollution relative to other pollution sources and pro-
vides information for communities having a broad range of sizes,
geographical locales, and public works practices. Information was
developed for major land-use areas within the cities (such as residen-
tial, commercial and industrial). Runoff was analyzed for the following
pollutants: BOD, COD, total and volatile solids, Kjeldahl nitrogen,
nitrates, phosphates, and a range of pesticides and heavy metals.
This report was submitted in fulfillment of Project No. 11034 FUJ,
Contract No. 14-12-921 under the sponsorship of the Water Quality
Office, Environmental Protection Agency.
-------
CONTENTS
Section Page
I CONCLUSIONS 1
II RECOMMENDATIONS 13
Operator Training 13
Effort 13
Data Collection 13
Street Maintenance 14
Auto Parking Controls 14
Equipment Adjustments 15
Gutter Brooms 15
Catch Basins 15
Separation of Storm and Sanitary Sewers 16
Freeway Runoff 16
Vacuum Wands 16
Special Curb System 17
Cost Effectiveness 17
Snow and Ice 18
III INTRODUCTION 19
Background 19
Objectives 20
Method of Approach 20
Project Overview and Scope 21
IV CHARACTERISTICS OF STREET SURFACE CONTAMINANTS 27
Common Sources and Constituents 27
Observed Loading Intensities 31
Pollutional Properties 43
Transport of Contaminants 81
-------
Contents (Contd.)
Section
VI
VII
VIII
IX
APPENDIXES
A
B
D
E
F
G
H
EFFECTIVENESS OF CURRENT PUBLIC WORKS PRACTICES
Existing Street Cleaning Practices
Street Sweeping Effectiveness
Discussion of Sweeping Effectiveness
Catch Basin Effectiveness
SIGNIFICANCE OF STREET SURFACE RUNOFF AS
A SOURCE, OF WATER POLLUTION
Summary of Contaminant Loads
The Effectiveness of Street Cleaning Practice
Significance to Street Cleaning Programs
ACKNOWLEDGMENTS
REFERENCES
BIBLIOGRAPHY
SAMPLE COLLECTION METHODS
SUMMARY OF CHARACTERISTICS OF TEST SITES
IN SELECTED CITIES
DATA SUMMARY AND INVESTIGATION OF
ACCUMULATION RATES
TYPICAL LAND-USE CATEGORIES
METHODS USED FOR ANALYSIS
QUESTIONNAIRE
STREET SURFACE CONTAMINANT SIMULANT
CATCH BASIN TEST PROCEDURES
Page
91
91
108
118
128
135
136
148
149
155
159
163
169
175
187
209
215
219
229
233
VI
-------
FIGURES
1 THE SYSTEMS NETWORK
2a ACCUMULATION OF CONTAMINANTS - HYPOTHETICAL CASE
(linear buildup, no sweeping, no rainfall)
2b ACCUMULATION OF CONTAMINANTS - HYPOTHETICAL CASE 32
(linear buildup with periodic sweeping but no rainfall)
3a ACCUMULATION OF CONTAMINANTS - HYPOTHETICAL CASE 33
(natural buildup, no sweeping, no rainfall)
3b ACCUMULATION OF CONTAMINANTS - HYPOTHETICAL CASE 33
(natural buildup with periodic sweeping but no rainfall)
4 ACCUMULATION OF CONTAMINANTS - TYPICAL CASE (natural 34
buildup with periodic sweeping and intermittent rainfall)
5 TOTAL SOLIDS LOADING ON STREET SURFACES - 39
Variation with Land Use
6 TOTAL SOLIDS LOADING ON STREET SURFACES - 41
Variations between Cities
7 DISTRIBUTION OF SOLIDS ACROSS STREETS 44
8 STREET SURFACE CONTAMINANTS AFTER DRY SIEVING 49
9 BOD LOADING INTENSITIES ON STREETS - 53
VARIATION BErWEEN CITIES
10 COD LOADING INTENSITIES ON STREETS - 53
VARIATION BETWEEN CITIES
11 VOLATILE SOLIDS LOADING INTENSITIES ON STREETS - 53
VARIATION BETWEEN CITIES
12 BOD LOADING INTENSITIES ON STREETS - 54
VARIATION WITH LAND USE
13 COD LOADING INTENSITIES ON STREETS - 55
VARIATION WITH LAND USE
14 INCREASE OF BOD AND COD CONCENTRATIONS IN 56
SOLIDS SAMPLES WITH INCREASED ELAPSED
TIME SINCE LAST RAINFALL
15 VOLATILE FRACTION OF STREET SURFACE CONTAMINANT 58
SOLIDS - DISTRIBUTION BETWEEN PARTICLE SIZE RANGE
16 NUTRIENT LOADING INTENSITIES AND WASTE 61
"STRENGTHS" - VARIATION WITH LAND USE
-------
Figures (Contd.)
Figure No.
17 VARIATION OF TOTAL PHOSPHATES WITH PARTICLE SIZE 63
18 VARIATION OF KJELDAHL NITROGEN WITH PARTICLE SIZE 64
19 VARIATION OF NITRATES WITH PARTICLE SIZE 65
20 HEAVY METALS LOADING INTENSITIES ON STREET
SURFACES - VARIATION WITH LAND USE , 70
21 HEAVY METALS CONCENTRATIONS - VARIATION
WITH LAND USE 71
22 HEAVY METALS CONCENTRATIONS - VARIATION
WITH PARTICLE SIZE 72
23 HEAVY METALS IN STREET SURFACE CONTAMINANTS -
VARIATION BY PARTICLE SIZE FOR BUCYRUS,
ATLANTA, TULSA, AND PHOENIX II 73
24 PESTICIDE CONCENTRATIONS - VARIATION WITH
PARTICLE SIZE 80
25 TRANSPORT OF STREET SURFACE CONTAMINANTS
BY RUNOFF 82
26 MOBILE RAIN SIMULATOR 83
27 RAIN SIMULATOR AND SAMPLE COLLECTION SYSTEM 84
28 PARTICLE TRANSPORT ACROSS STREET SURFACES -
VARIATION BY PARTICLE SIZE 86
29 PARTICLE TRANSPORT ACROSS STREET SURFACES -
VARIATION BY PARTICLE SIZE 86
30 PARTICLE TRANSPORT ACROSS STREET SURFACES -
VARIATION BY PARTICLE SIZE 86
31 PARTICLE TRANSPORT ACROSS STREET SURFACES -
VARIATION BY STREET CHARACTER AND RAINFALL INTENSITY 86
32 PARTICLE TRANSPORT ACROSS STREET SURFACES -
VARIATION BY STREET CHARACTER AND RAINFALL INTENSITY 87
33 RELATIONSHIP BETWEEN PARTICLE SIZE AND
PROPORTIONALITY CONSTANT 88
34 COMPARISON OF SWEEPER PERFORMANCE IN FOUR CITIES 107
35 EFFECTIVENESS OF CONVENTIONAL MOTORIZED STREET
SWEEPING ON PORTLAND CEMENT CONCRETE AT
THREE MASS LEVELS 111
viii
-------
Figures (Contd.)
Page
36 COMPARISON OF CLEANING PERFORMANCES OF
MOTORIZED STREET SWEEPING AND MOTORIZED
STREET FLUSHING 113
37 THE EFFECT OF PATTERN ON RESIDUAL DEBRIS 114
38 DEBRIS PICK-UP VS BRUSH SPEED 114
39 THE EFFECT OF SWEEPER SPEED ON THE RESIDUAL DEBRIS 114
40 COMPARISON OF RESULTS FROM SWEEPING EFFECTIVENESS
TESTS CONDUCTED UNDER VARIOUS CONDITIONS:
FOR DIRT/DUST FRACTION 120
41 PARTICLE SIZE DISTRIBUTION INITIALLY: FOR A 122
COMPOSITE SAMPLE
42 INITIAL AND FINAL LOADING ACROSS SWEPT 125
STREETS: COMPOSITE SAMPLE
43 REMOVAL EFFECTIVENESS WITH NUMBER OF PASSES 127
44 REMOVAL OF STREET SURFACE CONTAMINANT SOLIDS - 131
VARIATION WITH PARTICLE SIZE
45 COST EFFECTIVENESS PROGRAM FOR STREET CLEANING 152
-------
TABLES
No.
1 . STUDY TASKS - WATER POLLUTION EFFECTS OF STREET
SURFACE CONTAMINANTS 23
2 TOTAL SOLIDS, LOADING INTENSITIES (Ib/curb mi) 36
3 TOTAL SOLIDS, LOADING INTENSITIES (lb/1000 sq ft) 37
4 SOLIDS LOADING INTENSITIES DISTRIBUTION
ACROSS STREETS 45
5 PARTICLE SIZE DISTRIBUTION OF SOLIDS
SELECTED CITY COMPOSITES 48
6 OXYGEN DEMAND LOADING INTENSITIES ON STREETS 51
7 LOADING INTENSITIES ON STREETS - VARIATION
BY LAND USE 52
8 NUTRIENTS IN STREET SURFACE CONTAMINANTS -
VARIATION WITH LAND-USE CATEGORY 62
9 COLIFORM BACTERIA IN STREET SURFACE CONTAMINANT -
VARIATION WITH LAND-USE CATEGORY 67
10 HEAVY METALS LOADING INTENSITIES (Ib/curb mi) 69
11 DETECTION LIMITS FOR PESTICIDE ANALYSES 78
12 PESTICIDE LOADING INTENSITIES (10~6lb/curb mi) 79
13 PESTICIDE CONCENTRATIONS IN DRY SOLIDS (10~9 Ib
of pesticide/lb of dry solids) 79
14 PESTICIDE CONCENTRATIONS IN TOTAL SOLIDS (ppm) 80
15 SOME COMMONLY USED STREET SWEEPERS 93
16 COMMONLY USED STREET FLUSHERS AND EDUCTORS 96
17 CHARACTERISTICS OF CITIES SURVEYED 100,
18 STREET SWEEPING EQUIPMENT IN SELECTED CITIES 101
19 OPERATING SPECIFICATIONS FOR SWEEPERS IN
CITIES SURVEYED 101
20 FLUSHERS IN CITIES SURVEYED 102
21 CATCH BASIN CLEANING IN CITIES SURVEYED 102
22 SWEEPER DEBRIS COLLECTED, BY MONTH, FOR FOUR CITIES 103
23 CLEANING PRACTICES IN SELECTED CITIES 104
24 COMPARISON OF REMOVAL EFFECTIVENESS FOR MOTORIZED
SWEEPING AND VACUUMIZED SWEEPING 112
-------
Tables (Contd.)
No. Page
25 SUMMARY OF STREET CLEANING EFFECTIVENESS TESTS 116
26 SUMMARY OF STREET CLEANING EFFECTIVENESS 116
27 REMOVAL EFFECTIVENESS VERSUS PARTICLE SIZE
DISTRIBUTION 117
28 REMOVAL EFFECTIVENESS ACROSS STREET SURFACE 117
29 INITIAL LOADING DENSITY OF SIMULANT 119
30 SUMMARY OF RESULTS - CONTROLLED SWEEPER
EVALUATION TESTS 119
31 PARAMETERS WHICH AFFECT STREET SWEEPING PERFORMANCE 121
32 ESTIMATED STREET SWEEPER EFFICIENCY 123
33 SUMMARY OF CALCULATED REMOVAL EFFECTIVENESS VALUES 124
34 EFFORT REQUIRED TO ACHIEVE RESIDUAL MASS LEVELS 128
35 SUMMARY OF DATA ON CATCH BASIN CONTENT ANALYSIS 132
36 ANALYSIS OF CATCH BASIN CONTENTS 133
37 COMPARISON OF POLLUTIONAL LOADS FROM HYPOTHETICAL 136
CITY - STREET RUNOFF vs GOOD SECONDARY EFFLUENT
38 COMPARISON OF POLLUTIONAL LOADS FROM HYPOTHETICAL 137
CITY - STREET RUNOFF vs RAW SANITARY SEWAGE
39 COMPARISON OF STREET SURFACE CONTAMINANTS WITH 138
STORM SEWER DISCHARGES
40 POLLUTION LOADS BY SELECTED COMMUNITIES 139
(Ib/curb mi)
41 HEAVY METALS LOADS BY SELECTED COMMUNITIES 140
(Ib/curb mi)
42 PESTICIDE LOADS BY SELECTED COMMUNITIES 141
(Ib/curb mi)
43 TOTAL AND FECAL COLIFORM LOADING DISTRIBUTION 141
i BY LAND-USE CATEGORY
44 AVERAGE RATE OF ACCUMULATION OF POLLUTANTS 142
(Ib/curb mi/)
45 DISTRIBUTION OF CONTAMINANT LOAD BY 144
LAND-USE CATEGORY (Ib/curb mi)
46 DISTRIBUTION OF CONTAMINANT LOAD BY LAND-USE 145
CATEGORY (Ib/curb mi/day)
XI
-------
Tables (Contd.)
No. Page
47 FRACTION OF POLLUTANT ASSOCIATED WITH EACH 146
PARTICLE SIZE RANGE (% by Weight)
48 FRACTION OF HEAVY METALS ASSOCIATED WITH EACH 146
PARTICLE SIZE RANGE (% by Weight)
49 FRACTION OF PESTICIDES ASSOCIATED WITH EACH 147
PARTICLE SIZE RANGE (% by Weight)
50 DISTRIBUTION OF HEAVY METALS BY LAND-USE 147
CATEGORY (% by Weight)
51 SELECTED POLLUTANT REMOVAL PROJECTIONS - 150
BY STREET SWEEPERS
XII
-------
Section I
CONCLUSIONS
-------
Section I
CONCLUSIONS
Under the sponsorship of the Office of Research and Monitoring, U.S.
Environmental Protection Agency, research was conducted to investi-
gate and define the water pollution impact of urban storm water discharge
and to develop alternate approaches suitable for reducing pollution from
this source. At the start of this study, a comprehensive literature
search was conducted to collect existing data regarding the sources,
quantities, and pollutional properties of street surface contaminants
and refuse. It revealed the following:
a considerable amount of data and information exists
relating to pollutional loads associated with storm
water and combined storm and sewer systems
the data available on storm water pollutional loads are
not directly relatable to the materials contributed by
street surface contaminants
information was lacking on relationships between street
surface contaminants, their pollutional characteristics
and the manner in which they are transported during storm
runoff periods.
This study, therefore, focused on three principal areas:
determining the amounts and types of materials which
commonly collect on street surfaces
determining the effectiveness of conventional public
works practices in preventing these materials from
polluting receiving waters
evaluating the significance of this source of water
pollution relative to other sources.
The research led to the following conclusions:
JU Runoff from street surfaces is generally highly contaminated.
In fact, it is similar in many respects to sanitary sewage. Calcula-
tions based on a hypothetical but typical U.S. city indicate that the
runoff from the first hour of a moderate-to-heavy storm (brief peaks
-------
to at least 1/2 in./hr) would contribute considerably more pollutional
load than would the same city's sanitary sewage during the same period
of time , as indicated in the following table.
CALCULATED QUANTITIES OF POLLUTANTS WHICH
WOULD ENTER RECEIVING WATERS - HYPOTHETICAL CITY
STREET SURFACE
RUNOFF
(following
1 hr storm)
(Ib/hr)
RAW
SANITARY
SEWAGE
(Ib/hr)
SECONDARY
PLANT
EFFLUENT
(Ib/hr)
Settleable plus
Suspended Solids 560,000
BOD5 5, 600
COD 13,000
Kjeldahl nitrogen 880
Phosphates 440
Total coliform
bacteria (org/hr) 4000 x 10
10
1,300
1,100
1,200
210
50
460,000 x 10
10
130
110
120
20
2.5
4.6 x 10
10
Source: Tables 37 and 38.
The hypothetical city has the following characteristics:
Population - 100,000 persons
Total land area - 14,000 acres
Land-use distribution:
residential - 75%
commercial - 5%
industrial - 20%
Streets (tributary to receiving waters) - 400 curb miles
Sanitary sewage - 12 x 10 gal/day.
It should be noted that these calculations are for a situation in which
streets are cleaned (intentionally or by rainfall) on the average of
about once every five days. Thus, the above discharge of contaminated
runoff could conceivably occur many times in a year. On the basis of
this information, there is little question that street surface contami-
nants warrant serious consideration as a source of receiving water
pollution, particularly in cases when such discharges of contaminants
coincide with times of low stream flow or poor dispersion.
-------
2. The major constituent of street surface contaminants was consistently
found to be inorganic; mineral-like matter, similar to common sand and
silt.
This inorganic material, most of which is probably blown, washed, or
tracked in from surrounding land areas, does not constitute a serious
water pollutant by itself. However, along with this material is organic
matter, a small fraction of the total on the basis of mass. At a given
location, both fractions (organic and inorganic) increase in loading in-
tensity (Ib/curb mile) with increasing time since the last cleaning. Data
indicate, however, that the organic fraction tends to accumulate at a
faster rate than the inorganic fraction. However, within the time frame
of interest here (i.e. , a few days to a few weeks), the organic fraction
is still much smaller than the inorganic.
The quantity and character of contaminants found on street surfaces is
summarized in the following table. The tabulated values are for all
cities tested. They are weighted averages in which data for larger
cities are allowed to bias the reported loading intensities.
WEIGHTED MEANS
MEASURED FOR ALL SAMPLES
CONSTITUENTS (Ib/curb mile)
Total Solids
Oxygen Demand
BOD5
COD
Volatile Solids
Algal Nutrients
Phosphates
Nitrates
Kjeldahl Nitrogen
Bacteriological
Total Coliforms (org/curb mile)
Fecal Coliforms (org/curb mile)
Heavy Metals
Zinc
Copper
Lead
Nickel
Mercury
Chromium
Pesticides
p, p-DDD
p,p-DDT
Dieldrin
Polychlorinated Biphenyls
1400
13.5
95
100
1.1
.094
2.2
9
99 x 10
5.6 x 109
.65
.20
.57
.05
.073
.11
67 x 10~6
61 x ID'6
24 x 10~6
1100 x 10~6
Source: Tables 40, 41, 42 and 43
Note: The term "org" refers to "number of coliform organisms observed."
-------
Significant amounts of heavy metals were detected in the contaminant
materials collected from street surfaces; zinc and lead being the most
prevalent, as indicated in the previous table.
Heavy metal compounds have the potential of being highly detrimental to
biological systems, depending upon their specific chemical form. The
samples collected in this study have been analyzed only so far as to
indicate the total quantities of each metal present, not their specific
chemical form. The Office of Research and Monitoring of the U.S.
Environmental Protection Agency intends to develop more definitive
information from the samples collected in this study.
Substantial quantities of organic pesticides and related compounds were
found in the street surface contaminants. On the order of 0.001 Ib/curb
mile total was found for the cities tested, although the data showed con-
siderable variation from site to site. The chlorinated hydrocarbons p,p-DDD
and p,p-DDT were found rather consistently, as were polychlorinated biphenyl
(PCB) compounds (see Table 12 in Section IV). Although these have repeat-
edly been associated with adverse environmental effects in recent contro-
versies, the actual significance of these findings cannot yet be stated
since the environmental consequences of such materials have not yet been
established with any degree of certainty.
3. The quantity of contaminant material existing at a given test site was
found to depend upon the length of time which had elapsed since the site
was last cleaned; intentionally (by sweeping or flushing) or by rainfall.
The field sampling program focused on collecting materials from street
surfaces at single points in time (i.e. , no attempts were made to repeated-
ly sample a given site to develop information on how contaminants accumulate
with time). However, information was collected for each site to define
the elapsed time since the last substantial rain storm and/or cleaning.
Computer analyses of such data revealed correlations between antecedent
cleaning time and loading intensity. In general, industrial land-use
areas tend to accumulate contaminants faster than commercial or residen-
tial areas. Accumulation patterns as calculated here are shown in the
figure below (See Appendix I for details).
-------
uoo
n
5
Q
1200
800
ELAPSED TIMC SiNCf LAST CLEA.NiN
Source: Appendix I
4. The quantity of contaminant material existing on street surfaces was
found to vary widely, depending upon a range of factors. However, load-
ing intensities averaged on the order of 1400 Ib/curb mile of street for
the cities tested. The total solid loading intensities for the various
land-use areas tested are tabulated below.
NUMERICAL
MEAN
(Ib/curb mi)
Residential
low/old/single 850
low/old/multi 890
med/new/single 430
med/old/single i, 200
med/old/multi 1,400
Industrial
light 2,600
medium 890
heavy 3,500
Commercial
central
business district 290
shopping center 290
1,200
2,800
290
1,400
Source: Table 2
-------
The principal factors affecting the loading intensity at any given
site include the following: surrounding land-use, the elapsed time since
streets were last cleaned (either intentionally or by rainfall), local
traffic volume and character, street surface type and condition, public
works practices, season of the year, etc.
Contaminant loading intensities were found to vary with respect to
land-use patterns in the surrounding locale. In general, industrial
areas have substantially heavier than average loadings. All industrial
test sites (20 of them) taken together have an average loading of some
2800 Ib/curb mile; twice the mean for cities on the whole. This is
probably because industrial areas tend to be swept less often and because
generation rates of dust and dirt tend to be high (e.g., "fallout,"
spillage from vehicles, unpaved dirt areas, streets in poor condition,
etc.). Of these, heavy industrial areas showed the heaviest loadings,
medium industrial the lightest. The loadings varied so widely between
individual sites that it would be speculative to state why one type of
industrial area is dirtier than another.
Commercial areas have substantially lighter loading intensities than
the mean for cities on the whole (290 Ib/curb mile average vs 1400).
This is probably because they are swept so often; typically several times
weekly, daily in prime areas.
Residential areas were found to have an average loading intensity compa-
rable to the average for all land uses of all cities taken together:
1200 Ib/curb mile. Here again, the loadings varied widely from site to
site, and it would be speculative to state why one city is more heavily
loaded than another or why one type of residential neighborhood is cleaner
than another. The data in Table 2 (Section IV) implies, however, that
there is some tendency for newer, more affluent neighborhoods to be
cleaner; possibly because they are better maintained by residents and/or
are further from sources of contamination.
5. Perhaps one of the most important findings of this study is that
such a great portion of the overall pollutional potential is associated
with the fine solids fraction of the street surface contaminants.
Further, these fines account for only a minor portion of the total loading
on street surfaces. As shown in the following table, the very fine
silt-like material (< 43 microns) accounts for only 5.9 percent of the
total solids but about one-fourth of the oxygen demand and perhaps
one-third to one-half of the algal nutrients. It also accounts for over
one-half of the heavy metals and nearly three-fourths of the total
pesticides. This concentration of pollutants in a small amount of very
fine matter is of particular importance, considering that conventional
street sweeping operations are rather ineffective in removing fines
(sweepers were observed to leave behind 85 percent of the material finer
than 43 microns; 52 percent of the material finer than 246 microns)
-------
FRACTION OF TOTAL CONSTITUENT ASSOCIATED
WITH EACH PARTICLE SIZE RANGE (% by weight)
<43ju. 43/^ -~246Ju
TOTAL SOLIDS
BOD5
COD
Volatile Solids
Phosphates
Nitrates
Kjeldahl Nitrogen
Heavy Metals (all)
Pesticides (all)
Polychlorinated Biphenyls
5.9
24.3
22.7
25.6
56.2
31.9
18.7
37.5
32.5
57.4
34.0
36.0
45.1
39.8
i "s*,
51.2
73
34
>246Ju.
56.5
43.2
19.9
40.4
7.8
23.0
41.5
48.7
27
66
Source: Table 47, 48 and 49.
6. Chemical Oxygen Demand (COD) tests provide a better basis for
estimating the oxygen demand potential. It was found that due to
the presence of toxic materials in street surface contaminants
seriously interfered with BOD measurements. Such materials (particularly
heavy metals) were found to be present in many samples at levels far
in excess of those known to cause substantial interference.
7. Street surface contaminants are not distributed uniformly across
the streets. The solids loading intensity across a typical street is
gi ve n be 1 ow.
STREET LOCATION
(Distance from Curb)
0 - 6 in.
6 -12 in.
12 -40 in.
40 -96 in.
96 to center line
SOLIDS
LOADING INTENSITY
(% of Total)
78
10
9
1
2
Source: Table 4
-------
Typically, 78 percent of the material was found within 6 in. of the curb;
over 95 percent within the first 40 in. Presumably this is due to trans-
port by traffic (direct impact plus air currents) and because the curb
is a physical barrier, the gutter a protected zone.
The distribution of debris across a street after sweeping results in the
gutter being much cleaner; however, the sweeping operation moves much of
the material out of the gutter and redistributes it on areas which were
somewhat cleaner prior to sweeping. The redistribution is shown below.
Source: Fig. 42
The present design of gutter brooms is such that they tend to redistribute
the dust and dirt fraction (< 2000 fj.) over the surface of the street and
indeed are not particularly efficient in moving the dust and dirt fraction
out of the gutter.
-------
8. The rate at which rainfall washes loose particulate matter from
street surfaces depends upon three primary factors: rainfall intensity,
street surface characteristics, and particle size. Computer-assisted
analysis of data from a special series of field experiments revealed
that the wash-off phenomenon can be simulated by a simple, exponential
equation:
N = N (1 - e~krt)
c o
where N is the weight of material of a given particle size washed off
a street initially having a loading of N after t minutes of rainfall at
an intensity of r inches per hour. The proportionality constant k (units
of hr/in. min) depends upon street surface characteristics but was found
to be almost independent of particle size (at least within the range of
sizes of particular interest here; i.e., 10 to 1000 microns).
Street surface characteristics were found to have an effect on the con-
taminant loadings observed at a given site. For example, asphalt streets
had loadings about 80 percent heavier than all concrete streets. Streets
paved partially with asphalt and partially with concrete were inter-
mediate (their loadings were about 65 percent heavier than for all
concrete streets). The condition of street pavement is also important.
Streets in fair-to-poor condition had loadings about 2-1/2 times as
high as streets in good-to-excellent condition.
The design of future systems for controlling pollutional effects of
street runoff should take into account the fact that particulate con-
taminants arrive at point of entry to the sewer system in a manner
which is quite regular and predictable on the basis of a few, easily
measured parameters descriptive of the site and the design rainstorm.
Further studies should be conducted to develop design procedures which
can assure that the performance of such pollution control facilities is
consistent with their cost.
9. Current street cleaning practices are essentially for aesthetic
purposes and even under well-operated and highly efficient street
sweeping programs, their efficiency in the removal of the dust and
dirt fraction of street surface contaminants is low. The removal
efficiency of conventional street sweepers was found to be dependent
upon the particle size range of street surface contaminants as follows:
-------
PARTICLE SIZE
(Microns)
2000
8402000
246 840
104 246
43 104
< 43
Overall
SWEEPER
EFFICIENCY
(%)
79
66
60
48
20
15
50
Source: Table 32
The overall removal effectiveness for the dust and dirt fraction is 50
percent (that is, one-half of the dirt and dust fraction remains on the
street). The removal effectiveness for litter and debris, however,
(i.e., paper, wood, leaves, etc.) ranges from 95 to 100 percent.
10. Street cleaning effort, in terms of equipment minutes per 1000 sq ft
of area swept, required to achieve a greater removal effectiveness of the
dust and dirt fraction of street surface contaminants is several times
the effort normally expended in sweeping operations as indicated below.
EFFECTIVENESS
(%)
95
90
70
EFFORT
(equip rain/
1000 sq ft)
1.5
.85
.50
INCREASE OVER
NORMAL
(0.237)
6.3
3.6
2.1
Source: Table 34
Increased effort can be achieved by operating at a slower speed (normal
effort based on operating speed of 6 mph) or conducting multiple passes.
To achieve an overall effectiveness of 70%, two cleaning cycles would
be required. Effectiveness values greater than 90% are probably not
achievable with present state-of-the-art street sweepers.
10
-------
11. The technique of measurement of the effectiveness of street cleaning
practices, as related to the pollutions! properties of street surface
contaminants, was adequately demonstrated in this study. Additionally,
the following mathematical relationship was utilized to calculate the
removal effectiveness of the dust and dirt fractions by particle size
range:
* * -kE
M = M + (M - M )e
o
where M is the amount of street surface contaminants remaining after
sweeping, M is the initial amount and E is the amount of sweeping
effort involved in equipment minutes per 1000 sq ft (M and k are
dimensionless empirical constants dependent upon sweeper characteristics,
particle size of contaminants and street surface).
12. One of the most serious problems encountered in street sweeping
concerns vehicle parking. Increases in the use of vehicles and unavail-
ability of offstreet parking result in the occupancy of the gutters by
parked vehicles. In congested areas it is not unusual to find the
entire curb sides of streets occupied by parked vehicles. In some large
cities, "no parking" regulations have been instituted during scheduled
street sweeping hours.
13. The state-of-the-art regarding management information systems for
public works is not very far advanced. Existing cost accounting,
work reporting and equipment maintenance recording systems are fragmen-
tary and produce disparate comparative statistical data. There is need
for a system which will aid in providing public works with accurate
cost data associated with street cleaning practices.
14. Specially conducted field studies indicate that catch basins (as
they are normally employed) are reasonably effective in removing coarse
inorganic solids from storm runoff (coarse sand and gravel) but are
ineffective in removing fine solids and most organic matter. This is of
considerable importance because it is these latter materials which con-
tribute most heavily to water pollution effects. The material which
collects in catch basins comes from sources other than surface runoff.
Sample analyses indicated that much of the material found in urban and
suburban catch basins consisted of litter, leaves, used oil, etc. Upon
decomposition, the contents of catch basins become even more threatening
to receiving water quality.
11
-------
Section II
RECOMMENDATIONS
-------
Section II
RECOMMENDATIONS
OPERATOR TRAINING
« Street cleaning operations are generally focused on controlling those
types of contaminants and debris which are a nuisance from the standpoint
of aesthetics or public safety. The finer matter, shown here to be of
importance as a water pollutant, is seldom pursued. Although conventional
street sweeping equipment is not particularly effective in collecting
fines, with special attention on the part of operators a considerable
amount of the material normally "missed" could be collected.
Jt is recommended that street cleaning equipment operators be trained
not only in how their equipment can best be operated (i.e., vehicle
speed, broom speed, broom position, etc.) but also in what material
needs to be removed and where this is commonly located. Much of the
fine material which normally lays in the gutters could be picked up
if the operators had an appreciation for its importance relative to
water pollution effects.
EFFORT
This study has shown that the removal effectiveness of the dust and
dirt fraction of street surface contaminants is a function of the effort
expended in street cleaning operations, and to achieve a greater removal
effectiveness requires several times the effort normally expended in
sweeping operations. Effort is measured in equipment min/1000 sq ft
and effort can be increased by operating at a lower speed or sweeping
more often.
Jt is recommended that increased effort be expended on street cleaning
operations. Operating speeds should not exceed 5 miles per hour unless
operating on high-speed arterials. Additional cleaning cycles should
be scheduled on streets that are the principal vehicular arterials.
DATA COLLECTION
Acceptable methods of planning and evaluating the efficiency of street
cleaning programs are not available at the present time. The adequacy
and overall economy of street cleaning programs largely depend upon the
effective utilization of currently available street cleaning equipment.
An effective program planning technique requires accounts and detailed
13
-------
reporting on manpower and equipment utilization and equipment maintenance
and operating costs.
It is recommended that public works departments maintain accurate and
detailed records of street cleaning operations, including manpower
utilization, equipment utilization, and equipment maintenance. The
American Public Works Association, in their Special Reports No. 36
and No. 37, have created guidelines for developing standard procedures
to be used in collecting and reporting statistics and for measuring
and evaluating equipment performance. The procedures outlined in
these reports should be utilized in providing the necessary input data
to a cost-effective street cleaning program.
STREET MAINTENANCE
* Pavement type and condition were both found to have a substantial
effect on the total amount of loose particulate matter found on streets.
All-concrete streets were typically much cleaner than all-asphalt streets;
mixed concrete/asphalt streets were intermediate. Streets in good con-
dition were substantially cleaner than those in fair or poor condition.
These findings are as one might expect, although the specific reasons
(cause/effect relationships) have not been established, i.e., the
streets could be cleaner because they are easier to sweep or because they
themselves generate less material. Whatever the reason, it appears as
though there are distinct benefits to keeping streets in good condition.
It is recommended that public works departments pay increased attention
to maintaining pavements in good condition. When the material for
paving is being selected, it is recommended that this difference in
asphalt and concrete be taken into consideration, along with the fac-
tors normally included in such decisions.
AUTO PARKING CONTROLS
The field tests conducted in this study indicate that the bulk of the
material of primary concern (at least from the standpoint of water
pollutional effects) tends to accumulate near the curb. This is particu-
larly true where on-street parking is heavy. Discussions with public
works personnel revealed that no satisfactory means have been found
for effectively cleaning streets while cars are parked densely along
the curb.
Jt is recommended that cities give special consideration to ways of
restricting on-street parking on the days that sweepers or flushers
make their regular rounds. An effective approach (employed in Balti-
more) was to pass an ordinance restricting on-street parking, send
14
-------
public works crews out to educate local residents as to their street
cleaning program (a high degree of public support developed, with
neighbors reminding each other of the need to move their cars on
sweeping days), post signs along the streets, and enforce the ordinance
through citations and/or tow-away of vehicles. The program has allowed
the city to achieve substantially better control of all forms of street
contaminants and debris.
EQUIPMENT ADJUSTMENTS
A survey of equipment parameters (i.e., main broom speed, strike or
patterns, main broom pressure, gutter broom position, etc.) in various
cities showed a wide range of operational characteristics. The
effectiveness of sweeping can be improved by proper adjustments of main
broom, gutter broom, hydraulic systems, dust deflectors, elevator mech-
anisms, hopper operations , etc .
Jt is recommended that routine maintenance schedules include proper
adjustments to sweeper operating parameters as specified in Manufac-
turer's, Owner's, and Operating Manuals.
GUTTER BROOMS
» Gutter brooms were found to redistribute the dust and dirt frac-
tion (<2000ju) over the surface of the street, and in fact were not
particularly efficient in moving the dust and dirt fraction out of
the gutter.
It is recommended that the role of gutter brooms in street cleaning
be further evaluated and research directed towards the development of
new techniques for the efficient cleaning of gutters.
CATCH BASINS
Controlled field tests conducted on catch basins indicate that they
are relatively ineffective in preventing pollutant materials washed off
streets from entering the sewer system. Thus they serve little construc-
tive function. Further, they tend to accumulate large quantities of
organic matter (from a variety of sources) which subsequently decompose
and constitute a threat of massive slug pollution on being flushed out
during storms.
It is recommended that p'oblic works departments give serious consid-
eration to how necessary catch basins are in their particular systems.
Where a simple stormwater inlet structure would suffice, it is probably
desirable to get rid of the catch basin (either by replacing it or by
filling it in).
15
-------
An interim response which would be of considerable value in most com
munities would be to clean out dirty catch basins on a regular basis.
This would be particularly effective if they were cleaned just before
periods of major rainfall.
SEPARATION OF STORM AND SANITARY SEWERS
Modern sewer design practice has been influenced by the assumption
that storm runoff is quite clean, relative to sanitary sewage. Thus,
separate systems have been provided; one to convey the storm runoff to
direct discharge into receiving waters, the other to convey the sanitary
sewage to a treatment plant before it is discharged. This study along
with other recent information indicates that storm runoff is far from
being "clean" and probably warrants being treated in many instances.
Many older cities in this country were built with combined systems where
storm and sanitary sewage are not kept separate. There has been some
pressure for several years to encourage cities with combined systems to
separate their sewers (a very extensive operation, from the standpoint of
practicality and economics). The fact that both types of sewage have
been found to be important pollutant sources casts some doubt as to whether
sewer separation is warranted or treatment of all sewage is required.
It is recommended that further consideration be given to the desir-
ability of separating storm and sanitary sewers.
FREEWAY RUNOFF
This study focused on the contaminant materials which reside on urban
and suburban street surfaces. It intentionally omitted consideration
of freeways, even though these are a common element in most cities and
are surely heavily loaded. They were omitted because they are typically
subject to a somewhat different spectrum of contaminant sources and
because they cannot be cleaned by the same techniques as conventional
surface streets. Also, the techniques for studying them are necessarily
somewhat different.
It is recommended that special studies be conducted to determine the
amount and nature of materials which can wash off of urban freeways
during storms and identify means for controlling this source of water
pollutants. It would be important to conduct this study on well-defined
test areas having exclusive existing drainage systems. Since traffic
characteristics and aerial transport of fine solids both have a pro-
nounced effect on contaminant loadings, it is imperative that these be
studied concurrently with the freeway surface itself.
VACUUM WANDS
Recent developments in street maintenance equipment have provided
public works operations with a new type of equipment for collecting
loose leaves and litter via manually guided, truck-mounted, vacuum
"wands." Such devices or modifications thereof may be applicable to
16
-------
collecting street surface contaminants from areas normally rendered
inaccessible by parked cars.
Jt is recommended that special tests be conducted to evaluate the
feasibility (technical, operational, and economic) of vacuum wand
units in collecting street surface contaminants.
SPECIAL CURB SYSTEM
Field tests conducted in this study indicate conclusively that the
bulk of the material of primary concern is located near the curb. This
area is often swept only sporadically or missed completely. It is clear
that flusher units could be designed to wash materials over to the curb
with a high degree of effectiveness. If the water and contaminants
could then flow along the gutter to a pickup point, a great improvement
could be made in the control of subsequent street runoff. However, with
parked cars present, the flow of water down the gutter is seriously
impaired by curbed tires. Hence, with conventional curb/gutter con-
struction, the potential benefits of specially designed flushing systems
cannot be realized.
It is recommended that research and development be conducted to
explore special curb/gutter configurations which would allow free flow
of water and flushed debris to a pickup point, even in the presence of
parked cars. Other aspects of this study would be to develop special
mobile flushers (probably evolutionary extensions of the low volume/
high velocity units developed by the U.S. Naval Radiological Defense
Laboratories for removing radioactive fallout materials from street
surfaces), as well as special equipment and techniques for picking up
the water and contaminants after flushing.
COST EFFECTIVENESS
A cost-effectiveness program for street cleaning operations was
presented in this study which would assist public works directors in
evaluating the efficiency of street cleaning operations. However,
insufficient information was obtained during this study to adequately
prooftest the proposed model program.
It is recommended that a full-scale test program be conducted, in
cooperation with a municipal public works department, to examine the
overall effectiveness of street cleaning operations and the feasibility
of a cost-effectiveness model that could be utilized by municipalities
to upgrade current street cleaning practices. The full-scale test
program should include the evaluation of newly developed street clean-
ing equipment such as vacuumized sweepers, and broom sweepers, and
the general feasibility of adopting special public works practices
involving the use of special flushing units, modified gutter and inlet
designs, catch basins, and extra cleaning cycles for both catch basins
and urban streets.
17
-------
SNOW AND ICE
The sampling program in this study was conducted in several cities -
Seattle, Milwaukee, and Baltimore - that receive considerable amounts
of snow at various times of the year.
It is recognized that considerable quantities of water pollution are
associated with the enormous quantities of snow removed from urban streets
and dumped into nearby bodies of water or onto water supply watersheds.
However, no attempt was made in this study to conduct a sampling program
to measure such pollution.
In addition, large quantities of de-icing agents are applied to urban
streets during winter months for removal of ice and snow. There has been
growing concern over the environmental effects resulting from these
practices.
Jt is recommended that a study be conducted in several snow-belt cities
located near bodies of water to determine the extent and severity of
this problem. The results of such a study should serve to define possible
requirements for modifying current snow dumping practices and developing
safer means of ultimate snow disposal.
18
-------
Section III
INTRODUCTION
-------
Section III
INTRODUCTION
BACKGROUND
During the past few years, it has become increasingly obvious that
runoff from storms in urban areas is by no means "rainwater" in terms of
quality. Rather, storm runoff typically contains substantial quantities
of impurities; so much so that it is a more serious source of pollutants
than municipal sewage in many areas. Numerous studies have been and are
being conducted to help define this problem; to determine the amounts of
pollutant substances involved, their sources, their practical signifi-
cance, and possible means of control.
Urban runoff can contribute to a variety of problems, including direct
pollution of receiving waters, overloading of treatment facilities, and
impairment of sewer and catch basin functions. These problems are caused
in part by hydraulic overloading, but also by the various pollutants
contained within the runoff.
Previous studies by the American Public Works Association (Ref. 1) and
AVCO Corporation (Refs. 2 and 3) provide much valuable information on
the total problem of water pollution resulting from urban runoff. They
both point out the shock pollution loads which storm runoff from urban
areas can place on receiving waters. Among the sources of pollution in
urban runoff water are debris and contaminants from streets, contaminants
from open land areas, publicly used chemicals, air-deposited substances,
ice control chemicals, and dirt and contaminants washed from vehicles.
The APWA report (Ref. 1) suggested various means of reducing the pol-
lution problem created by urban runoff and emphasized the need for more
definitive investigations as to the source, cause, and extent of the pol-
lutants; the interrelationships and significance of the variables; and
the development of standard procedures, methods and/or techniques for
measuring the street surface contaminants. Among the concepts proposed
for limiting storm water pollution was the improvement of street clean-
ing methods and operations.
URS Research Company recently conducted a comprehensive literature search
(see Bibliography) to collect existing data regarding the sources,
quantities, and pollutional properties of street surface contaminants
and refuse, which has revealed the following:
0 a considerable amount of data and information exists relating
to pollutional loads associated with storm water and com-
bined storm and sewer systems
19
-------
« the data available on storm water pollutional loads are not
directly relatable to the materials contributed by street
surface contaminants
« information is lacking on relationships between street
surface contaminants, their pollutional characteristics
and the manner in which they are transported during storm
runoff periods.
OBJECTIVES
The broad objectives of this study were to investigate and define the
water pollution impact of urban storm water discharge and to develop
alternate approaches suitable for reducing pollution from this source.
The study focused on three principal areas:
determining the amounts and types of materials which
commonly collect on street surfaces
determining the effectiveness of conventional public
works practices in preventing these materials from
polluting receiving waters
evaluating the significance of this source of water pol-
lution, relative to other sources.
METHOD OF APPROACH
The above objectives were accomplished in nine major tasks, as follows
Task 1. Develop a planning and control technique for the
entire study (this Section)
Task 2. Establish a project review panel (this Section)
Task 3. Determine the current state of the art related to street
cleaning practices, specifically as they relate to water
, pollution control (Section V)
Task 4. Determine the characteristics of street surface contam-
inants and refuse (Section IV)
Task 5. Develop means of determining extent and significance
of pollutional materials not usually captured in
normal sampling techniques (Section IV)
Task 6. Develop standard techniques or procedures which can be
utilized for evaluating the performance of equipment
and street cleaning practices (Section V)
20
-------
Task 7. Identify the variables involved, their interrelation-
ships and their relative significance in water pollution
terms under most likely real-world conditions (Sections
IV, V and VI)
Task 8. Determine the feasibility of developing a mathematical
model and, if possible, the degree of sophistication
required (Section V)
Task 9. Prepare a final report.
Brief descriptions of the specific task units conducted to meet the
objectives of each major task are given in Table 1. Figure 1 presents
a systems network showing the interrelationships between the various
task units. The systems network also served as a scheduling tool which
allowed feedback and evaluation as the project proceeded.
PROJECT OVERVIEW AND SCOPE
The study, which required some eighteen months to conduct, involved a
broad range of research techniques, including:
field measurements and sample collection
sample analyses
experimental studies
literature reviews
surveys (by questionnaires and interviews).
The major efforts of the study centered around three elements :
collecting contaminant materials from street surfaces all
over the country
analyzing those materials to determine their physical,
chemical, and biological properties (insofar as these per-
tain to source identification, evaluating pollutional
potential, and/or possible means of control)
observing and evaluating various street cleaning practices
in several cities throughout the country.
In this study we have defined street surface contaminants as being
those materials found on street surfaces which are capable of being
washed off during common rain storms. Street surfaces are defined
as being the paved traffic lanes, any parking lanes, and the gutter;
i.e., the area typically bounded by curbs. In urban areas the total
contribution of contaminants comes from a much larger area than just
this "street surface." For instance, there are surely substantial
contributions from sidewalks, planter strips, yards, driveways,
21
-------
parking lots, roofs of buildings, etc. Tims, the quality of the
water entering the storm sewer inlet is only partly a function of the
contaminants washed from the street, per se.
The overall problem of controlling urban runoff pollution is complex
indeed. The general approach involves dividing the overall problem
into discrete segments which can be studied first separately and then
in relationship to each other. This project is but a part of that
overall approach; our "segment" is the "street surface." As the over-
all problem becomes understood, effective control measures can be
developed and implemented.
The rationale for selecting the "street surface" as the study area
and excluding various adjacent contributing areas is twofold:
« the street surface receives contaminants from sources
which do not contaminate surrounding areas (particularly
vehicular traffic)
several of the potential control measures can be applied
to street surfaces but not to surrounding areas (e.g.,
street sweeping).
22
-------
Table 1
STUDY TASKS
WATER POLLUTION EFFECTS OF STREET SURFACE CONTAMINANTS
MAJOR TASK
TASK UNIT OBJECTIVE
1 1.01 Describe the requirements for each major task and specific
task unit comprising this study
1.02 Design a systems network for accomplishing those tasks
1.03 Describe future efforts required to accomplish the
overall mission
2 2.01 Select review panel members
2.02 Conduct 1st review panel meeting
2.03 Conduct 2nd review panel meeting
2.04 Conduct 3rd review panel meeting
2.05 Conduct 4th review panel meeting
2.06 Prepare project critique
3 3.01 Describe current street cleaning practices including the
descriptions and specifications of equipment
3.02 Determine from available data the performance (effective-
ness and cost) of current street cleaning practices as
these relate to water pollution control
3.03 Conduct field evaluations of current street cleaning
practices to supplement the data found to be available
in task unit 3.02
3.04 Identify and analyze the deficiencies in existing street
cleaning practices as they relate to water pollution
control
4 4.01 Collect existing data regarding the sources, quantities and
pollutional properties of street surfaces contaminants
and refuse
4.02 Identify, on a preliminary basis, those physical, chemical
and biological properties of street surface contaminants
and refuse which are believed to be pertinent to street
cleaning operations, transport of the material by runoff,
and the action of the material as a pollutant
23
-------
Table 1 (Continued)
MAJOR TASK
TASK UNIT OBJECTIVE
4.03 Identify requirements for additional or more reliable
information
4.04 Design sampling and analysis program to obtain required data
4.05 Conduct sampling and analysis program
4.06 Determine the manner in which the properties of street
surface contaminants and refuse vary with factors such as
season and land use
5.01 Identify materials or objects which are generally excluded
from usual sampling techniques, including those materials
whose undesirable impact is related primarily to factors
such as aesthetics, hazards, or nuisances
6.01 Develop a program for establishing performance criteria for
street cleaning in a given area
6.02 Develop a program for evaluating the performance of street
cleaning equipment and/or practices
6.03 Develop a set of simulants (synthetic street refuse) for
evaluating performance of street cleaning equipment
and/or practices
6.04 Prepare results of task 6 in manual form for use in
evaluating performance of street cleaning equipment
and/or practices
7.01 Identify the variables involved, their interrelationships
and their relative significance ih water pollution terms
under most likely real world conditions
8,01 Develop empirical mathematical relationships to describe
the performance of selected street cleaning methods
8.02 Investigate the feasibility of developing a standardized
methodology for determining least-cost acceptable street
cleaning practice from alternatives using mathematical
modeling techniques
9.01 Prepare final report draft
9.02 Review final report draft
9.03 Submit final report
24
-------
1970
1971
Figure 1
THE SYSTEMS NETWORK
1972
41
o
U
JULY I AUG SEPT OCT NOV DEC JAN I FEB I MAR /
TASK UM IT WO
I. Of tOt
Task S.itrms^.
9:91 V
of Fane! Meeting rvieetirtg
Dt»r,b. Cerent Evaluate F^F^ar^ of ' ConH|[C( f.f d |nvM(iao
jiieel Ueamng if. eel Cleaning PracHE« W ^ Current Slreet Clean
t f '
*"
4o*
4.0Z 4.o« 4.0»
Id tf Pall t- Id If Da, D"i9"
f-roqrom
! !
I p 1,,,,^
hfiM Simulant
6.01 *** t
Develop o Program for Develop o Ftogrom For
Slreet Cleaning Piaclicci
7.01 ^
Ml
D,,.,ml,,
Sli eel Cl
PR
tM
ng Pra
[]
Somp
I
i
e Math
an.ng
MAY 1 JUNE JULY
*
i
f
ng Progra"! Ond Sample Anolyiis
I
Effectiveness
AUG
1
Anol/ze
~ Current
4.06
Deterr.
Refuse
J
Prept
£qui
brl'H/
SEPT OCT 1 NOV I
.o»
Descriptions of Future Efforts
f
1.0« Oh
Meeting ,
Defficiencies in __ __ l^^^
>treel Cleaning practices
with Seasan/Und Use
ire Manual
^
DEC JAN FEB 1 MA
I
2J36
Gitiqu, ^~
*»'
rjpcre Final Report Dtai
Completion Date
-------
Section IV
CHARACTERISTICS OF
STREET SURFACE CONTAMINANTS
-------
Section IV
CHARACTERISTICS OF STREET SURFACE CONTAMINANTS
This section deals with answering the basic question, "What are the
characteristics of street surface contaminants in terms of being potential
water pollutants?" This involves the discussion of:
common sources and constituents
observed loading intensities
transport mechanisms.
COMMON SOURCES AND CONSTITUENTS
In this study we have defined street surface contaminants as being those
materials found on street surfaces which are capable of being washed off
during common rain storms. Street surfaces are defined as being the
paved traffic lanes, any parking lanes, and the gutter; i.e. , the area
typically bounded by curbs. Excluded, therefore, are sidewalks, planter
strips, yards, driveways, roofs of buildings, etc. All of these surfaces
contribute both water and contaminants. However, as discussed in
Section III, they have been excluded here because they are not subject
to the same array of sources and because the means of controlling contami-
nants from street surfaces would generally not apply in these other areas.
Street surface contaminants are comprised primarily of particulate matter
but also include non-particulate soluble and suspendable matter which
are capable of being washed off the streets by rainfall (i.e., oils,
salts, saps, etc., are included even though they represent a minor loading
on a mass basis).
The sources of materials found on street surfaces are greatly varied.
However, the bulk of the contaminants commonly found comes from the
sources described in the following paragraphs. Obviously, the material
observed at any given location will be a composite of several sources;
the actual "mix" being a function of such factors as land use, geo-
graphical locale, season, weather, traffic volume and character, local
public works practices, etc.
Pavement
The street surface itself is a source of the materials we have defined
as being contaminants. Included here are asphaltic and Portland cement,
their various products of decomposition, and aggregate materials. In
addition, there are typically small amounts of road marking paints, crack
27
-------
fillers, and expansion joint compounds. On a weight basis, aggregate
materials account for the largest contribution of contaminants from this
source. Observation of photomicrographs was of particular value here.
The amounts found at any given location vary substantially and cannot be
predicted on the basis of information developed here. Three important
factors have been identified which appear to correlate with observed
generation rates of such materials:
the age and condition of street surfaces (old, worn,
cracked streets seem to generate more in the way of
contaminants)
the local climate (cold winters accelerate degradation
through freeze-thaw cycling, the use of studded tires ,
and the use of sand, ash, and chemicals for skid control)
leaks and spills of fuels and oils which hasten the
degradation of asphaltic pavements.
Motor Vehicles
This source of street surface contaminants contributes a broad range of
materials and in numerous ways. Although the contributions cannot be
quantified, they can be listed by general category:
leakage of fuel, lubricants, hydraulic fluids, and coolants
fine particles worn off of tires and clutch and brake linings
particulate exhaust emissions
dirt, rust, and decomposing coatings which drop off of
fender linings and undercarriages
vehicle components broken by vibration or impact (glass,
plastic, metals, etc.).
The generation rates of these materials have not been determined but
probably correlate with season, geographic locale, and local traffic
conditions.
Their importance as water pollutants varies substantially from material
to material. Fuel, lubricants, and hydraulic fluids degrade asphaltic
pavements, thereby increasing the amount of inorganic solids loadings
on receiving waters. Further, they float, causing films which are un-
sightly and hinder oxygenation. Like many other petrochemicals, they
are damaging to biological forms. The lead, nickel, and zinc compounds
used in their formulation are also harmful but to an undetermined extent.
The purpose of this discussion is not to define the extent to which
motor vehicles contribute to water pollution. That is a complex subject
in itself. Rather, we have listed a number of ways in which they affect
the amount and nature of the street surface contaminants investigated in
this study,
28
-------
Atmospheric Fallout"
This category has been included to establish a relationship between
street surface contaminants and air pollution. A large fraction of the
particulate matter contributing to the water pollution effects of street
surface contaminants are of a size fine enough that they could have been
transported by air currents prior to being deposited on the street sur-
face. The extent to which this actually occurs is not known, of course,
for the contaminants as a whole. However, certain contaminants surely
arrived on the street surface following air transport.
Major sources of such materials would be industrial stacks and vents,
construction and excavation projects, agricultural operations, and
exposed vacant land areas. Automotive traffic and heavy commercial air
traffic are also sources of fine airborne particles.
Many such forms of "fallout" are virtually inert and would add only tur-
bidity and suspended solids loadings to receiving waters. Others are
surely reactive and would impose loadings of oxygen demand, algal nutri-
ents, toxic metals, and pesticides.
Vegetation
This source includes leaves and other plant materials (pollen, bark, twigs,
seeds, fruiting bodies, grasses, etc.) which fall onto the street surface,
are blown there by wind, or are dumped or raked there. In any given lo-
cation, the generation rates are clearly a function of season, land use,
local landscaping, and public works practices. However, such materials
are distributed widely; substantial amounts of fragmented vegetative
matter were found in virtually all street surface contaminant samples,
irrespective of season or locale.
These materials are of interest in this study for several reasons. They
contribute to oxygen demand (immediate demand if they are allowed to
accumulate and decompose in catch basins, long-term demand after they
sink to form bottom deposits). Algal nutrients and pesticides are also
generally associated with vegetation.
Runoff From Adjacent Land Areas
A significant amount of both organic and inorganic matter found in street
surface contaminant samples originates in adjacent land areas and is
transported to the streets by runoff (some also blows there and is tracked
onto the streets by vehicles). The amount and nature of material so
imported varies widely as a function of topography, land use, season,
public works practices, etc. The major sources are, of course, areas
where soil is exposed rather than protected by vegetative cover, paving,
or other means (e.g., vacant lots and fields, unprotected cuts and fills,
ongoing construction and demolition projects).
29
-------
In addition to particulate matter and soil-like material, oils and
greases from parking areas, service stations, and commercial/industrial
operations are transported onto street surfaces by runoff.
Litter
This category of street surface contaminants is notable even though it
is probably not a major source of water pollution in the usual sense of
the term. Included here is the myriad of refuse items which are dis-
carded (intentionally or otherwise) by the public at large. Two major
components of this litter are packaging materials of all sorts (paper,
plastic, metal, glass, etc.) and printed matter (newspapers, magazines,
advertising flyers, etc.). Another source of litter is the intentional
disposal of waste material into the street when nearby occupants sweep
sidewalks and driveways or rake up plant debris and dispose of it in the
street.
As would be expected, litter exists on the street surface intact and in
various degrees of decomposition (photomicrographs of street surface
contaminant samples typically reveal the presence of dust-size fragments
of glass, clearly recognizable as ground-up soft drink and beer bottles).
Litter is of particular importance to this study because of its relation-
ship to conventional street cleaning operations. Most street sweepers
are employed for the primary purpose of cleaning up visible litter and
like-size materials, the intent being to maintain aesthetically pleasing
community streets. Section V of this report discusses the effective-
ness of such operations in controlling the pollutional aspects of street
surface contaminants.
The fact that many components of litter float in receiving waters makes
them a particular nuisance from the standpoint of visual aesthetics (e.g.
styrofoam cups, plastic bags, waxed paper cups, cigarette packages, etc.,
all float well and tend to be concentrated at the surface by eddy currents,
wind, and quiescent water). Because such conditions impair the receiving
water's suitability for certain uses, they are considered here to be a
pollutional effect of street surface contaminants.
Another effect of litter on water pollution is that litter tends to
collect in catch basins where its organic fractions gradually decompose,
causing increased oxygen demand, suspended solids, and turbidity in re-
ceiving waters, once there is sufficient storm flow to flush them out.
Further,- litter mechanically interferes with a catch basin's ability to
pass leaves, grass, and fine solids; hence, these also tend to be stored
and to decay, adding to the pollutional shock load on receiving waters.
Organics (food, animal droppings), another source of litter, are generally
present in substantially smaller quantities (on a weight basis) than the
dust and dirt fraction component of street litter and debris. Organics
could affect BOD readings; however, they are generally considered a
30
-------
source of nuisance rather than a serious water pollutant. Most of the
fecal coliform bacteria observed in samples of street surface contaminants
are probably associated with bird and animal droppings.
This category of street surface contaminants is well known but virtually
impossible to describe quantitatively, either in amount or character.
The major source of spills is vehicular transport. The types of materials
vary widely, but include primarily: dirt, sand, gravel, cement, various
bulk commercial and industrial raw materials and products, agricultural
products, and various types of wastes. Although a minor source in most
areas, discharges from backed-up, broken, or overloaded sewers also
contribute contaminants to street surfaces.
Anti-Skid Compounds
These include common salts (NaCl and CaCl2) plus a host of specially for-
mulated organic and inorganic compounds which are applied with the intent
of melting ice or inhibiting its formation during cold weather. Included
also are various types of relatively inert materials (e.g., sand and ash)
which are applied to act as abrasives in reducing skid hazards. The
total amount of anti-skid compounds applied during cold weather in north-
ern communities is considerable. This subject has been covered in detail
in two recent reports (Refs. 4,5) and therefore will be mentioned super-
ficially here and discussed further in Section V.
OBSERVED LOADING INTENSITIES
The major field sampling efforts conducted under this study were directed
toward determining the amount and nature of contaminants actually residing
on street surfaces at the time of sampling. The matter collected corres-
ponds quite closely with that matter which would wash off of a typical
street during a moderate-to-heavy rain of about an hour's duration (the
specifics of the field sampling techniques are presented and discussed
in Appendix A). It is important to note that the values reported herein
are observed loading densities (in weight per unit curb length or weight
per unit street surface area) rather than rates of accumulation, except
where specified. For each sample collected, the data concerning the
elapsed time since prior sweeping and the time since prior substantial
rainfall was recorded. This information, plus a description of each
test area, is given in Appendix B.
At the outset of the study it was recognized that the amount of contami-
nant material residing on the street surface would vary considerably
from place to place and from time to time, depending upon a number of
dominant factors. Much of this study has been devoted to identifying
those factors and establishing their relative importance.
31
-------
These dominant factors can be grouped into three major categories:
time since last cleaning or rainfall
season of the year
locale (actually, the activities that go on at the
particular location).
Before getting into the details as to how contaminant loadings vary in
response to these factors, it is important to discuss the issue of genera-
tion rate vs observed loading intensity. Consider a hypothetical area
of street surface which is (for the purposes of discussion) subjected
to a continual and uniform loading of contaminants (uniform with respect
to both time and spatial distribution). If there were no other activities
to disturb the contaminants, the loading intensity would increase linearly
with respect to time, as shown in Fig. 2a.
If the street were cleaned periodically but the cleaning operation were
unable to remove all of the deposit, the curve would be cyclic as shown
in Fig. 2b.
z
UJ
I
z
o
z
0
TIME
Fig. 2a. Accumulation of Contaminants - Hypothetical Case
(linear buildup, no sweeping, no rainfall)
Z
UJ
I
z
o
z
Q
o
TIME
Fig. 2b. Accumulation of Contaminants - Hypothetical Case
(linear buildup with periodic sweeping but no rainfall)
32
-------
In any actual situation it is clear that the plot cannot be linear but
rather would curve over and gradually approach a limit (see Fig. 3a).
If this were not the case, an unswept street would become impassable with
accumulated debris. While the mechanisms of removal cannot be described
quantitatively, they surely include wind, displacement by moving traffic,
and the like. Where periodic cleaning is practiced, the plot looks like
Fig. 3b. Note that this represents a case of uniform, continuous load-
ing and a regular cleaning (with the same degree of efficiency each
time and a uniform frequency).
LU
(
Z
o
z
O
o
TIME
Fig. 3a. Accumulation of Contaminants - Hypothetical Case
(natural buildup, no sweeping, no rainfall)
LLJ
t
O
z
0
TIME
Fig. 3b. Accumulation of Contaminants - Hypothetical Case
(natural buildup with periodic sweeping but no rainfall)
Figure 4 depicts the effect that intermittent rains would give . Large
storms would remove more than sweepers; small storms, less.
33
-------
z
UJ
h-
Z
O
z
Fig. 4.
Accumulation of Contaminants -
Typical Case (natural buildup
with periodic sweeping and
intermittent rainfall)
The final task here is to con-
sider how the plot would look
if all of these factors (i.e.,
sweeping efficiency, sweeping
frequency, rainfall frequency,
etc.) were allowed to vary'
randomly throughout their
normal range. Its shape would
be very complex indeed, looking
very little like the simplistic
curves of Figures 2b or 3b.
The purpose for discussing
this situation is to place into
context the meaning of the
"observed loading intensities"
reported herein. Streets were
sampled to determine their
contaminant loading intensi-
ties. At each sampling location, historical information was obtained as
to when the street had last been swept and when the last rain of consider-
able magnitude (one hour minimum, peaks to one-half in./hr). The thing
to note here is that, while these data are of some value, they are by no
means sufficient to describe the shape of the overall curve. As a matter
of fact, it is impossible to derive the rate of accumulation (the slope
at any point along the rising portion of the curve) on the basis of these
data. However, such was not the intent in this study; that point should
be made clear. What we are interested in is the answer to the question:
"How much material resides on a typical street which is subject to being
washed off by rainfall occurring at a random point in time?" To answer
that, it was necessary to look at streets in every stage representative
of the entire curve; streets which had not been cleaned recently, those
which had just been swept or had just been flushed by rain, etc.
A limited sub-study was conducted to determine if any consistent trends
could be found to relate the amount of contaminants found on streets
with the elapsed time since the last sweeping or substantial rainfall.
Since areas with widely differing overall characteristics were included
in the study, it was difficult to discern any dominant or repetitive
trends. The efforts involved in this sub-study are reported in Appendix C
General Observations
Total solids loading intensities were determined by collecting contaminant
materials from street surfaces by a combination of dry sweeping and
flushing with a jet of water. Sample areas of 800 to 1000 sq ft of
street were used (see Appendix A for a complete description of field
procedures and rationale for their selection). The total dry weight
of solids sample divided by the size of the area sampled is reported
34
-------
here as the "loading intensity." Two means of reporting such values have
been adopted:
the average loading intensity over the entire area
sampled (lb/1000 sq ft)
the average loading intensity along the length of the area
sampled (Ib/curb mile).
At the outset of the study, it was assumed that the loading intensities
would vary from location to location in response to such factors as
land-use category, season of the year, geographic locale, size of the
city, etc. For this reason, sampling sites were selected to represent
a broad range of all of these factors. In summary, we collected samples
in some ten land-use categories in twelve cities (large and small)
throughout the country. The overall solids loading intensities observed
are reported in Tables 2 and 3. (A summary of data on all observed
pollutant characteristics is presented in Appendix C along with an
analysis of accumulation rates.)
A review of the data reveals that solid loading intensities vary
significantly from city to city and from land use to land use. While it
is presumptive to report a single value representative of such a widely
varying population, the calculated means (weighted over both land uses
and cities) are: 16 lb/1000 sq ft and 1500 Ib/curb mile. Obviously,
they should be used only with some caution.
At the outset of the study , it was hoped that trends could be identified
to help relate the amount of street surface contaminants present with
certain local characteristics; in particular, parking, traffic, and
pavement.
The conclusions, after review of all the data collected here , are that
pavement composition and condition have a fairly consistent effect on
the amount of total solids present as street surface contaminants.
Specifically, streets payed entirely with asphalt have loadings about
80 percent heavier than all-concrete streets (streets paved partly with
concrete, partly with asphalt, are about 65 percent heavier than all
concrete). Streets whose pavement condition was rated "fair-to-poor"
were found to have total solids loading about 2-1/2 times as heavy as
those rated "good-to-excellent.
The other factors considered: traffic speed, traffic density, and park-
ing density, all surely have an influence, but no consistent trends
could be identified. It is probable that other, more dominant, factors
had greater effects (e.g., land use, season, etc.).
The following paragraphs deal with the factors which influence the loading
intensities observed at any given test site.
35
-------
Table 2
TOTAL SOLIDS, LOADING INTENSITIES
(Ib/curb mi)
San Jose I
Phoenix I
Milwaukee
Bucyrus
Baltimore
San Jose II
Atlanta
Tulsa
Phoenix II
Seattle
Decatur
Scottsdale
Mercer Island
Owasso
Nianevical
Mean
Weighted
Mean
RESIDENTIAL
LOW/OLD/ LOW/OLD/ MED/NEW/ MED/OLD/ MED/OLD/
SINGLE MULTI SINGLE SINGLE MULTI
840 1,100 290
770 1,900 180 310
720 560 280 6,900
1,900 410 1,900
1,300 1,200 1,400 500
620 470 200
590 31 330
120 620 150
1,600 1,100 380 500
470 540 260 140
1,200
1,200
93
250
850 890 430 1,400
1,200
INDUSTRIAL
LIGHT MEDIUM HEAVY
1,700 1,100
450 1,300
410 12,000
1 , 000 1 , 600
1,300 860 240
12,000 1,100
3,700 300
1,100 280
260 1,100
710
2,600 890 3,500
2,800
COMMERCIAL
CENTRAL SHOPPING
BUSINESS CENTER
DISTRICT
270 460
210 640
260 210
68 63
1,200 160
60 430
180
200 180
190 190
290 290
290
Weighted
Mean
910
650
2,700
1,400
1,000
6,000
430
330
910
460
1,500
Note: Tabulated values are the average loading intensities one would find if the contaminants were spread uniformly across the full width
of the street. The fact that they are distributed quite non-uniformly means that these figures must be used with caution.
-------
Table 3
TOTAL SOLIDS, LOADING INTENSITIES
(lb/1000 sq ft)
San Jose I
Phoenix I
Milwaukee
Bucyrus
Baltimore
San Jose II
Atlanta
Tulsa
Phoenix II
Seattle
Decatur
Scottsdale
Mercer Island
Owasso
Numerical
Mean
Weighted
Mean
RESIDENTIAL
LOW/OLD/ LOW/OLD/ MED/NEW/ MED/OLD/ MED/OLD/
SINGLE MULTI SINGLE SINGLE MULTI
6.3 8.6 2.2
5.8 14.0 1.4 2.4
12.0 9.2 3.5 66.0
27.0 6.5 31.0
14.0 14.0 24.0 4.5
69.0 6.3 2.5
8.6 4.5 0.6
1.8 12.0 2.9
22.0 19.0 6.0 6.3
8.8 6.8 4.1 2.9
20
13
1.3
3.9
12 11 5.6 15
INDUSTRIAL
LIGHT MEDIUM HEAVY
13.0 8.4
3.4 10.0
5.2 160.0
16.0 26.0
16.0 11.0 3.1
92.0 9.0
44.0 5.7
13.0 4.4
3.9 20.0
15.0
25 11 47
COMMERCIAL
CENTRAL SHOPPING
BUSINESS CENTER
DISTRICT
2.0 3.5
1.6 4.8
3.3 2.7
1.3 0.6
12.0 1.5
1.0 7.3
2.1 3.3
3.0 2.4
4.0 3.7
3.3 3.3
Weighted
Mean
8.6
8.5
32
18
11
56
4.7
4.6
12
6.4
16
Note: Tabulated values are the average loading intensities one would find if the contaminants were spread uniformly across the full width
of the street. The fact that they are distributed quite non-uniformly means that these figures must be used with caution.
-------
Land Us(3
At the outset of this study the assumption was that the amounts and
types of materials on street surfaces vary as a function of surrounding
land use. This was based on the knowledge that numerous related factors
(e.g., sweeping practices, traffic volume and type, parking patterns,
vegetation, etc.) all vary with respect to land use. Thus, the sampling
and analysis program employed here was designed to develop information
on a range of land uses. Ten categories were selected. While these ten
are not comprehensive (i.e. , there are areas typical of many communities
which are not included), they do account for a gross majority of the
land area comprising most non-rural communities. Furthermore, these
particular categories are easily recognized in most communities (in the
sense that they can be identified and readily distinguished from other
land-use categories).
The purpose for collecting samples in a variety of land-use categories
was two-fold: to provide a rational basis for characterizing an entire
city (samples from all land-use areas were combined to yield "city
composites") and to allow any significant trends between land-use cate-
gories to be identified. (Photographs depicting the general, overall
appearance of the areas comprising each of these land-use categories are
included in Appendix D to help orient the reader.)
Figure 5 depicts variations in total solids loadings from one land-use
category to another. The data plotted are from the eight cities listed
in Table 3.
Three major land-use "groupings" which are used throughout the remainder
of the report are introduced here: residential, industrial, and commer-
cial.
One conclusion that can be drawn from this information is that, although
there is considerable variation between land-use categories , when taken
all together, streets in industrial areas generally tend to be more
heavily loaded than residential streets; commercial streets, less heavily
loaded. These conclusions are consistent with expectations. Commercial
areas are commonly swept weekly (often daily); industrial areas often
rather infrequently and irregularly (details regarding common municipal
sweeping practices are given in Section V).
It was noted earlier that this study focuses primarily upon existent
solids loadings intensities (i.e., the solids one observes at a given
point in time rather than their accumulation rate over time). This
approach is essential if one is to predict the amount of contaminant
which will run off during a storm. For example, consider the trends plot-
ted on Fig. 5. A curb-mile of "residential" street will contribute
more contaminants than a curb-mile of "commercial" street even though
the commercial streets receive more contaminants per unit time. The
fact that commercial streets are cleaned so frequently (to maintain
38
-------
4000
3000
E
-e
1
to
z
LU
10 H-
- ?
O O
4/1 Z
O O
RESIDENTIAL INDUSTRIAL
2000
1000
I
COMMERCIAL
ro .i
-a
o
J!
s I
i> g
LAND-USE CATEGORIES
Fig. 5. Total Solids Loading on Street Surfaces
Variation with Land Use
39
-------
high aesthetic standards) tends to mask the fact that the accumulation
rate between sweepings is quite high.
Cities
Another factor expected to affect the amount and nature of contaminants
(observable at any location) was the particular city. There is little
doubt that the following factors affect contaminant loadings and that
they differ from city to city:
geographical locale - a factor with a substantial but
ill-defined effect on climatic conditions (seasonality of
snow, rainfall, wind, etc.), the community's proximity to
fixed area sources of airborne particulates (deserts,
plains, tilled fields, etc.), the amount and type of vegetation
(and associated leaf-fall), etc.
activities within the community - a generalized factor which
refers to the presence of point sources of airborne particulates
from residential, commercial, institutional, and industrial
activities (incinerators, power plants, industrial stacks,
construction projects, etc.)
public works practices and controls - a composite of factors
including street cleaning practices, street maintenance
practices, snow and ice control practices, and control over
such activities as refuse collection, litter abatement, use
of studded tires, etc.
non-specific characteristics of the community including land
area, population, land-use patterns, population density
and distribution, traffic density and patterns, general over-
all air pollution, and the important factor of public attitude
regarding public cleanliness and aesthetics (a factor which
is often reflected in terms of the size of the public works
department budget).
While it is plausible to assume that loadings would vary from city to
city, it had to be established that they do. Figure 6 shows graphically
how these loadings vary . However, the information developed in this
study does not provide a basis for predicting their value for a given
city (nor was that the intention of the study). The purpose for including
several cities of different types, sizes, and locations was to be sure
that a broad enough range of conditions was represented in the "sample
population" to assure the reliability and credibility of the research
findings. The eight principal cities studied were selected on the basis
of differences in their age, size, geographical locale, land-use
patterns, population and industrial growth patterns, topography, and
types of receiving waters.
40
-------
4000
3000
to
Z
o
Z
Q
<
o
O
2000
1000
-6000
111
~ a>
v
X _£
NOTE: The unusually high value for San Jose II
may be an anomaly caused by an unrepresentative
sample.
Fig. 6. Total Solids Loading on Street Surfaces
Variations between Cities
41
-------
Likewise, the purpose for repeating tests in some eities was not to
ascertain the exact variation in conditions from time to time during
the year. Rather, it was to determine if there are any seasonal effects
(obviously, there are marked effects) which should be given special
consideration in any subsequent studies of this type.
The dates of the samplings are listed below (further details of individual
sites are presented in Appendix B).
San Jose I Dec. 1970
San Jose II June 1971
Phoenix I Jan. 1971
Phoenix II June 1971
Milwaukee Apr. 1971
Bucyrus Apr. 1971
Baltimore May 1971
Atlanta June 1971
Tulsa June 1971
Seattle July 1971
Distribution on Street Surfaces
Contaminant materials are not distributed uniformly on street surfaces;
neither across them nor along them. Most of the material (especially
particulate solids) is concentrated toward the curb. This is as expected
considering the tendency of traffic to "blow" material out of the traffic
lanes. A cross section taken of the full width of a typical street would
show some accumulation down the very center, little in the traffic lanes,
quite a bit in the curb lane (especially if cars are allowed to park
there; not so much because they are the source but rather because their
presence arrests and fosters the accumulation of moving material).
The highest concentration of solids is in the gutter, as would be expected,
since the curb forms a barrier to any particles moving transversely. At
grade median strips are generally zones of accumulation for particulate
street surface contaminants. Raised medians are generally relatively
clean but, at points where breaks are provided, considerable accumulation
is common.
It is important to note that the distributions described above are for
particulate solids. Substances like oils, greases, and various liquids
which spill or leak onto the street surface are generally found in
heaviest concentration along the center of each traffic lane and down
the center of parking lanes.
It is important to recognize the distinct non-uniformity of distribution.
For one, this non-uniformity makes it somewhat risky to discuss loading
intensities on the basis of weight per unit area unless such values
are carefully explained^as to meaning (i.e., it is important to stress
that these are 'average" loadings over reasonably large areas which
42
-------
include most of the street's width). For this reason, most values report-
ed here are in terms of Ib/curb mile rather than lb/1000 sq ft.
Another important aspect of this distinct non-uniformity has to do
with the potential for cleaning streets. The fact that most of the
particulate matter is concentrated in a relatively narrow zone along the
curb (typically, on the order of 70 or 80 percent is located within 6
in. of the curb) means that cleaning efforts focused there could be
highly effective. On the other hand, since this is the very location
where cars park, it is virtually impossible to achieve desired cleanli-
ness when cars are present.
Figures 7a through 7d show the distributions measured on streets in
several of the cities tested. The distributions are not identical from
site to site, but the trends described above are generally substantiated.
The data are plotted in these figures and given in Table 4.
Contaminant materials are not distributed uniformly along the length
of the streets. Features which tend to cause major variations are inter-
sections, bus stops, special turning lanes, and even driveways. As
noted previously, any variations in parking patterns will also cause
variations in loadings. Field sampling results indicate that inter-
sections are loaded on the order of one-third as heavy as the normal run
of street (these tests focused on the particulate solids loading within
the first 7 ft of the curb - the path covered by a conventional street
sweeper). Further, driveways were found to be less heavily loaded than
the spaces between driveways although the variation was only about
30 percent.
POLLUTIONAL PROPERTIES
The preceding discussion presented data concerning the amounts and distri-
bution of contaminant materials found on street surfaces. The following
discussions are concerned with the nature of those materials, particularly
their potential to act as pollutants in receiving waters. Six principal
aspects are covered:
suspended and settleable solids
oxygen demand
algal nutrients
coliform bacteria
heavy metals
pesticides.
43
-------
CO
LU
oo
z
100
50
NOTE: S-l extends to centerline of street, and is therefore variable.
S-5 is the strip adjacent to curb.
PHOENIX II
low/old/single
n $$$t .._
(J Bf^ViWt'iyV'f-'-VRfflJiPTygffgB
tO ^T CO CN
oo in oo J>
100
50
<#$
BUCYRUS
low/old/single
IT) ->J CO W
J) UO CO CO 00
o
z
Q
co
Q
O
co
100
50
To 376
BALTIMORE
med/new/single
o alii
"O * CO CM
CO J, co J,
f To
201
100
50
IT) Tf CO CN
to i/i to to
BALTIMORE
med. industry
10'
scale
Fig. 7. Distribution of Solids Across Streets (Based on Table
4)
44
-------
Table 4
SOLIDS LOADING INTENSITIES
Distribution Across Streets
CITY AND
LAND USE
Baltimore
med /new/single
Baltimore
medium industry
Baltimore
heavy industry
Milwaukee
med/new/single
Bucyrus
low/old/single
Tulsa
light industry
Seattle
light industry
Phoenix-II
low/old/single
Atlanta
heavy industry
Percentage of
total loading
found in each
strip
(average values)
TEST STRIP NUMBER
S-l
lb/ lb/
103ft curb ml
.79 64.2
.73 59.4
.10 8.1
.11 8. a
6.7 545
18.4 1500
2.0 162
.02 i.o
2%
S-2
lb/ lb/
103ft2 curb mi
.37 30.1
1.7 136
.35 28.4
.35 28.4
6.4 520
2.0 162
.66 53.6
3.8 309
.06 4.»
1%
S-3
lb/ lb/
103ft2 curb ml
9.3 755
15.4 1250
2.2 179
0.4 28.4
51.4 4180
7.5 610
2.7 220
48.6 3950
0.6 50.5
9%
S-4
lb/ lb/
103ft2 curb mi
8.8 71 3
9.3 755
2.0 162
i.o 122
41.4 3360
19.8 1610
19.7 1600
47.4 3850
1.2 97.6
10%
S-5
lb/ lb/
103ft2 curb mi
376 30600
201 16300
54 4360
69 5610
112 9100
21 1690
33 2660
213 17300
105 8540
78%
Street <£
&&&
>. >»-&
Curb
-i 9A"
f ACl" »-
r* 12" -\
* H
6" | 1
S-5 S-4 S-3 S-2 i>-l
45
-------
These were selected for study because of their potential for impairing
receiving water quality and because they are commonly used in characteriz-
ing pollutants from other sources (this obviously facilitates evaluating
the importance of street surface contaminants, relative to other sources
of pollution).
Another important characteristic of street surface contaminants is their
particle size distribution. This is because the size of solids determines
their transport on the street (by wind, water, and traffic effects) and
the ease with which they are removed by various cleaning techniques (sweep-
ers, vacuum sweepers, flushers, even catch basins). Furthermore, parti-
cle size is important in terms of pollutional aspects (i.e., where the
particles end up and what types of effects they will have). In the
following pages information is reported as to the tendency for certain
pollutional aspects to be associated with certain particle size ranges.
Data to support the issues are presented along with the discussion in
most cases. However, for convenience, the bulk of the data for the
study has been reduced and summarized, with pertinent excerpts presented
in Appendix C.
Suspended and Settleable Solids
Microscopic examination has revealed that the bulk of materials, loose
contaminants found on street surfaces, consist of "inert" minerals
of various types (quartz, feldspar, etc.) which reflect components of
street paving compounds and local geology. This inert portion of the
total contaminant loading is similar in size, shape, and composition to
the materials geologists classify as "sediments" and will henceforth
be referred to simply as sediments.
Sediments entering receiving waters fall into two major categories on
the basis of size; both categories have environmental effects associated
with them. Depending upon local flow patterns and velocities, a portion
of the sediments will be suspended, while the remainder, by virtue of
size and weight, will enter receiving waters by saltation, traction or
decreased transport energy (reduced flow velocity). The mechanisms by
which these materials act as pollutants may be direct, indirect, or
both.
The indirect effects of high sediment loadings on biologic systems can
be very great. Such mechanisms include the physical burial of plants
and animals and changes in the nature of the substrata causing alteration
of fauna and flora. High suspended sediment concentrations reduce water
transparency, inhibiting the transmission of light required for photo-
synthesis. This also interferes with predator/prey hunting relationships.
High sediment loads increase the probability of transporting pesticides
nutrients, various organic pollutants and many microbiological forms
by acting as a mobile substrate on which they adsorb, absorb, or other-
wise adhere.
46
-------
The direct effect of sediments includes actual damage to biological
structures, burial of organisms, and the clogging of respiratory, feed-
ing and digestive organs. High sediment loadings can also contribute
to the possibility of clogging sewer lines, increased solids loadings at
treatment facilities, and the shoaling of waterways.
Table 5 presents data on the particle size distributions of composite
samples from representative cities. The data were determined by summing
values obtained by dry sieving, wet sieving, and sedimentation pipette
analyses. (This is a common method used for determining particle size on
the basis of settling velocity, via Stoke's Law relationships.) Analytical
procedures utilized are described in Appendix E.
The classifications "sand," "silt," and "clay" have been included here to
help communicate the general properties of street surface contaminants.
These classifications also roughly correspond to the behavior of the
materials in water; that is, sand will generally settle out at low
current velocities, clay will remain suspended, and silt will be inter-
mediate (some will settle, some will not).
It seems likely that the materials in suspension will have a long-range
environmental effect; the coarser materials, a short-range effect (they
will be removed locally by sedimentation). Although the percentages of
suspended material are small compared to the total loading, their actual
weight on a curb-mile basis may be of some significance in increasing
the suspended sediment concentration of the receiving waters. For
example, consider the case of Milwaukee. By estimating the amount of
runoff for a 0.5 in. rain of 1 hour duration over a distance of 1 mile,
the loading per size range data for Milwaukee (from Table 5) has been
employed to develop an estimate of the concentration of suspended
material in the runoff. Runoff velocities are assumed to be high enough
to suspend most silt and all clay-size material. The greatly reduced
flow rates that this material will subsequently be exposed to will still
probably be high enough to maintain a state of transport. Complete
suspension will result in an average runoff concentration of around 100
ppm. Several case studies are presented in Section VI to help put the
pollutional potential of street surface contaminants into perspective.
In reality, the initial sediment slug may be many times higher. This
concentration will be diluted substantially by the receiving water
through a factor governed by its volume and initial suspended sediment
concentration. Depending on circumstances, the concentration may be
elevated to levels which could interfere with various organisms.
In the interest of helping orient the reader, we can compare the
information in Table 5 with street sweeper performance data (details
of which are discussed in Section V). The thing to note here is that
given a conventional sweeper operating at maximum efficiency, on the
order of 70 percent of material sand-size and larger can be removed.
Smaller materials are not removed well at all, however. From the
47
-------
standpoint of normal public works objectives (i.e., keeping the street rela-
tively free of large aesthetically displeasing debris), conventional sweep-
ers do an effective job. However, from the standpoint of controlling
the fine particulate matter which contributes so heavily to water pollu-
tion, conventional sweeping is relatively ineffective. This is espe-
cially true, of course, when sweepers are poorly operated. Photographs
showing street surface contaminants after dry sieving into particle
size ranges are shown in Fig. 8.
Table 5
PARTICLE SIZE DISTRIBUTION OF SOLIDS
SELECTED CITY COMPOSITES
SIZE
RANGES
> 4,800 u
2,000-4,800 U
840-2, 000 u
246-840 u
104-246 U
43-104 u
30-43 u
14-30 p.
4-14 M-
< 4 |a
Sand %,
43-4,800 LI
Silt %,
4-43 u
Clay %,
< 4 u
Lb Sand/curb mi
Lb Silt/curb mi
Lb Clay/curb mi
MILWAUKEE
12.0%
12.1
40.8
20.4
5.5
1.3
4.2
2.0
1.2
0.5
92.1
7.4
0.5
2,480
200
13.5
BUCYRUS BALTIMORE
% 17
10.1 4
7.3 6
20.9 22
15.5 20
20.3 11
13.3 10
7.9 4
4.7 2
0
74.1 82
25.9 17
0
1,020 845
356 176
9
" .
.4%
.6
.0
.3
.3
.5
.1
.4
.6
.9
.1
.1
.9
.3
ATLANTA
14
6
30
29
10
5
1
0
0
91
7
0
394
33
1
c"
/o
.8
.6
.9
.5
.1
.1
.8
.9
.3
.9
.8
.3
.5
.3
TULSA
-
37
9
16
17
12
3
3
0
0
92
7
0
300
30
0
%
.1
.4
.7
.1
.0
.7
.0
.9
.1
.3
.6
.1
.3
Note :
u = microns.
48
-------
840-2000 microns (3X)
246-840 microns (3X)
104-246 microns (3X)
104 microns (3X)
Fig. 8. Street Surface Contaminants After Dry Sieving
-------
Oxygen Demand
One of the most significant detrimental effects a pollutant can have
on receiving waters is to depress the dissolved oxygen level. Minor
depressions can usually be tolerated by a free flowing, relatively un-
polluted stream without any serious effects, although some shift in the
aquatic ecological balance will result. However, in many situations
where receiving waters are already.subject to the physical^and chemical
effects imposed by urban areas, the ambient or "usual case" oxygen resource
is only marginal to begin with. Therefore, substantial loads of oxygen-
demanding substances often lead to undesirable conditions; perhaps the
most notorious being fish kills, foul odors, unsightly'discoloration,
and slime growths.
In short, then, materials which depress receiving water oxygen resources
are considered pollutants. For the most part, such pollutants are
organic substances which are consumed by the stream biota as food. Con-
current with their consumption, the biota (primarily aerobic bacteria)
"breathe in" oxygen that was initially dissolved in the water. This
has the effect of stabilizing the organic matter (which is a "plus") but
leaves the rest of the aquatic life with less oxygen to fill their needs
(this being the "minus").
The amount of such organic matter present in polluted water or in a
waste can be measured and is broadly described in terms of "oxygen demand."
Three indices of such demand or waste strength which are in relatively
common use are BOD, COD, and volatile solids. BOD stands for biochemical
oxygen demand and is a reasonably direct measure of what goes on in the
receiving water (actually the test employs conditions which are quite
similar to what happens in nature, i.e., the waste is "fed" to bacteria
and the oxygen "breathed" during a 5-day test period is measured).
COD stands for chemical oxygen demand, an index which is measured by
reacting the sample at high temperature with strong chemicals (a boiling
mixture of concentrated sulphuric acid and potassium dichromate) to
determine how much oxidizable matter is present. The COD test is rapid,
precise, and less subject to certain interferences than the BOD test,
and was therefore indicated here. A third index of waste strength is
the volatile solids test which involves merely "burning" a dried portion
of waste solids at very high temperature (600°C) under controlled con-
ditions. This test is rapid, even simpler and less subject to the fac-
tors which interfere with the BOD or COD tests. It has been included
here for that reason.
Oxygen demand loadings on street surfaces were found to vary over a
very wide range depending upon the city, the land use, time since last
rainfall or sweeping, etc. This', of course, was expected. Loading
intensities varied from a high of over 60 to less than 2 Ib/curb mile
for BOD and from 400 to 13 for COD (see Table 6) . While the importance
of this is difficult to judge directly, the case studies developed in
Section VI should help put the findings in perspective. in summary they
50
-------
indicate that the oxygen demand attributable to street runoff is quite
substantial indeed. This source contributes large quantities of oxygen-
demanding materials in "slugs" which unquestionably causes a short-term
impairment of receiving water quality in perhaps a majority of locations.
These conditions have been substantiated by actual observations (see
Refs. 6 and 7). Perhaps the most notable case would be where the Sandus-
ky River, loaded with street runoff, has turned a murky black and lost
most of its fish life as it flows past Bucyrus, Ohio.
Table 6
OXYGEN DEMAND LOADING INTENSITIES ON STREETS
SAN JOSE-I PHOENIX-I MILWAUKEE BUCYHUS BALTIMORE SAN JOSE-II ATLANTA TULSA PHOEHIX-II SEATTLE
BOD
(Ib/curb mi)
COD
(Ib/curb mi)
16 6.5 12 2.9 61 53 1.9 14 10 4.8
310 30 48 29 20 400 13 30 54 17
Note: Tabulated values are computed average extrapolated from observed loading intensities
in several land-use areas having different antecedent accumulation periods.
For this reason, the values should be used with caution.
It should be noted that while BOD tests were run for many samples collected
from street surfaces , the data should be viewed with some skepticism.
This is primarily due to the fact that the presence of toxic materials
can seriously interfere with measured BOD results. Such materials (par-
ticularly heavy metals) have been found to be present in many samples
at levels far in excess of those known to cause substantial interference.
Note that the interference is in the direction of yielding low results,
so that Our measurements should probably all be raised somewhat (by
how much we would not speculate).
The COD test provides a better basis for estimating the oxygen demand
potential, primarily because it is not subject to interference by
toxic materials. COD tests were run on the bulk of the samples collected.
Oxygen demand values were measured to evaluate the pollution potential
of street runoff but also to reflect suspected differences and trends with
respect to such factors as land use, geographic locale, particle size
range, etc. For the most part, the data are not very informative (i.e.,
differences are notable but rather inconsistent). Numerous cross-checks
were run to verify test data and point up any errors in procedure and/or
computation, but to little avail. Apparently, even though many large
samples were collected in many areas and multiple tests run on each, the
heterogeneous nature of the material collected inherently yields high
deviations.
51
-------
This lack of regularity is particularly evident in comparing oxygen de-
mand loading from city to city. Figures 9, 10, and 11 plot the loading
intensities for the same ten tests in terms of three indices of organic
matter (or oxygen demand); e.g., BOD, COD, and volatile solids.
Some trends are apparent where loading intensities are compared on the
basis of land-use category, as shown in Figs. 12 and 13. Specifically,
over the ten city samples included, light industry tends to be heavily
loaded with both BOD and COD and the commercial areas (suburban shopping
centers and central business districts) only lightly loaded. The reason
for this pattern is not understood. However, there is a tendency for
public works departments to concentrate sweeping efforts in commercial
areas. Spillage of loads from trucking operations could account for the
high values in light industrial areas (these are often dominated by
warehousing operations, bulk materials storage, and light manufacturing).
It is of interest to note that, on a total solids basis (organic plus in-
organic solids) , heavy industry is far dirtier than light industry (see
Fig. 5 and Table 7). Obviously, the dirt from heavy industry contains far
less organic matter (as would be expected).
Table 7
LOADING INTENSITIES ON STREETS - VARIATION BY LAND USE
MEDIUM/ MEDIUM/ CENTRAL SUBURBAN
LOW/OLD/ LOW/OLD/ NEW/ OLD/ LIGHT MEDIUM HEAVY BUSINESS SHOPPING
SINGLE MULTI SINGLE MULTI INDUSTRY INDUSTRY INDUSTRY DISTRICT CENTER
BOD
(Ib/curb mi)
5.1
13
2.5
COD
(Ib/curb mi) 27
23
34 190
6.0
Total Solids
(Ib/curb mi) 1000 1000 480 1300 2300
3900 280
As reported later in this section, it was found that rainfall washes
streets fairly clean, removing a substantial fraction of the contaminants.
Following a rain, contaminants build up, increasing with respect to time.
A subject of interest here was to determine the manner in which oxygen
demand loadings build up following a rain. BOD and COD loadings in lb/
curb mile were first plotted vs time since last rainfall. These increased
rather steadily, as would be expected. Of more interest however
are the plots of Fig. 14 which show how the oxygen demand strengths of
unit amounts of solids samples increase with time [i.e., percent by
52
-------
UJ
h-
1
1
il
Jin
.
,
S 3 £ £ S J!
1 5- 1 5 5 i
j £ I 5 Ji
40
20
;- o
Fig. 9. BOD Loading
Intensities on Streets -
Variation Between Cities
UJ
I
z
O
!l
8>
u ~-
400
300
200
100
Fig. 10. COD Loading
Intensities on Streets-
Variation Between Cities
200
O
z
Q
s
100
11
111
0
Fig. 11. Volatile Solids Loading Intensities on
Streets - Variation Between Cities
53
-------
INDUSTRIAL
40
30
RESIDENTIAL
20
o
z 5-
9 E
21
CO
10
COMMERCIAL
'5.
Q.
O
LAND-USE CATEGORIES
Fig. 12. BOD Loading Intensities on Streets
Variation with Land Use
54
-------
200
100
z
o
81
u ^
RESIDENTIAL
Wl
INDUSTRIAL
COMMERCIAL
I '= £
< < i- >
-S ?, ~° j>
a, Ji
S *'
LAND-USE CATEGORIES
Fig. 13. COD Loading Intensities on Streets
Variation with Land Use
55
-------
1.5
1.0
Q
_l
o
LO
U_
O
I 0.5
u
Q
Q
z
LLJ
O
>
X
O
©
A BOD
0 COD
©
COD'S
©
BOD'S
j_
_L
0 10 20
TIME SINCE LAST RAINFALL (days)
30
40
50
60
Fig. 14. Increase of BOD and COD Concentrations in Solids
Samples with Increased Elapsed Time Since Last
Rainfall
weight of total solids , not pounds per curb mile (percent by weight
equivalent to pounds of BOD per 100 Ib of total dry solids)]. The trend
is that the oxidizable fraction of the contaminants continually increases;
COD at a greater rate than BOD. Recognizing that the data are limited
and quite scattered, we refrain from speculating on the exact shape of
these curves (other than to say that they probably level out somewhat,
given sufficient time). The conclusion to be drawn here is that the
data support our initial assumption that organic materials (the oxidizable
fraction) tend to accumulate on the streets faster than inorganic materi-
als (otherwise the curves would have a negative slope). This conclusion
56
-------
leads back to the issue of the sources of street surface contaminants.
Whatever the sources, it is clear that they must be contributing organic
matter more rapidly than inorganic matter. Another way to say this is
that vehicular inputs, leaves, litter, etc. are dominant over sand
and dust-lilce material. Further, these data seem to indicate that fixed,
constant sources of material containing both organics and inorganics (the
street surface itself is the prime example) must be insignificant contrib-
utors to the total load since their curve (if considered alone) would
plot as a straight horizontal line on Fig. 14. Note also that there is
no evidence that the pollution strength of the solids decreases with
time of exposure (through weathering) even up through as much as 60 days.
As discussed previously, an important characteristic of street surface
contaminants is their particle size distribution. This is because the
size of solids determines their-transport on the street (by wind, water,
and traffic effects) and the ease with which they are removed by various
cleaning techniques (sweepers, vacuum sweepers, flushers, even catch
basins). Furthermore, particle size is important in terms of pollutional
aspects (i.e., where the particles end up and what types of effects they
will have).
Figure 15 shows how the organic matter in street surface contaminants
is distributed between particle size ranges (here, volatile solids is
being used as an indicator of organic matter). Composites representative
of various cities and groups of cities were analyzed. Note that in all
cases the finer sizes tend to contain more organic matter than the coarser
sizes. This is reasonable since organic matter is typically low in
structural strength and can easily be ground into fine particles. Fur-
thermore, since non-particulate organic matter often adheres to the surface
of particles, the finer the particles involved the more organic matter
will adhere (because fine particles have greater unit surface areas than
coarse particles). This association of higher volatile solids with fine
particle size ranges is quite consistent from composite to composite as
shown by the shaded zones for each plot in Fig. 15.
BOD and COD were also analyzed to identify any relationships between
oxygen demand and particle size. The resulting trends (the data are
tabulated in Appendix C) are similar to Fig. 15, although somewhat less
consistent (presumably due to interferences in the chemical interactions
in the analyses).
Differentiation into size ranges is important because it allows compari-
son with the efficiency of street cleaning devices , as determined in
Section V. The size ranges at which sweepers are essentially ineffective
( < 246 microns) are observed to contain 33.9 to 99.5 percent of the
total BOD and COD loading. In other words, the majority of the oxygen
demand observed will run off the street with rainfall.
57
-------
Ol
00
to
Q
Zj
O
to
o
>
20
15
10
Composite from:
Phoenix I
I
Atlanta
Tulsa
Phoenix I
San Jose
Seattle
r CM S
rj ^t -o
-* O Tp
(N
8
8
3 <)
o s
^- CN
T)- -O
O ₯
PARTICLE SIZE (microns)
Fig. 15. Volatile Fraction of Street Surface Contaminant Solids
Distribution between Particle Size Ranges
-------
Algal Nutrients
An important aspect of water quality is its aesthetic appeal. Any
visual evidence of pollution therefore limits a water's beneficial uses.
While algal nutrients generally do not in themselves affect the appear-
ance of water, the aquatic growths which they stimulate do by increasing
color, turbidity, objectionable floating matter, and slimes. The ultimate
effect of nutrient discharges to receiving waters is eutrophication.
This is the term used to describe waters in which there is a high level
of phytoplankton activity, the results often being highly turbid, colored
waters having objectionable tastes and odors. When significant amounts
of nutritents are present in the water system, it is virtually impossible
to reverse the process and lower the nutrient level. This is due in
part to the fact that plant activity results in the conversion of nutri-
ents into plant matter and proteins, but, upon decomposition, the
nutrients are released back into the water system in a closed cycle.
With phytoplankton activity, the surface layers become supersaturated
in dissolved oxygen during periods of photosynthesis. Then, during
periods of low illumination, the algae consume oxygen. Eutrophied bodies
of water can therefore exhibit marked fluctuations in oxygen content, a
situation which is unfavorable to most aquatic life. Possibly the most
notorious aspect of eutrophic waters is the occasional occurrence of an
algae "bloom," wherein the waters become loaded with tremendous amounts
of algae. Natural byproducts of algae metabolism often include substances
which can produce tastes and odors along with possible toxic substances.
Normally, such substances are too dilute to be of much concern. During
a bloom, however, they become a problem. Further, when the bloom dies
out, large quantities of decomposing algae can exert tremendous oxygen
demand, possibly leading to anaerobic conditions and stratification of
relatively quiescent waters.
Nitrogen and phosphorous compounds are generally considered to be the
most important common algal nutrient compounds in receiving waters. In
this study these were measured as total Kjeldahl nitrogen, soluble nitrates,
and total phosphates. Phosphate compounds exist in several chemical forms
in nature. The most available form is orthophosphate (organically bound),
while polyphosphate is of only minor consideration. Polyphosphates are
converted with orthophosphates in aqueous environments within several
days (usually within several hours). Therefore, total phosphates is a
valuable measure of phosphorous nutrient impact. Nitrogen also exists
in several forms in nature, but the forms of primary interest in terms
of availability are nitrates and ammonium nitrogen. Again, since the
nitrogen in an aqueous system can be converted in various ways to one
of these two forms, the total nitrogen test is indicative of nitrogen
nutrient availability.
59
-------
Limitations on nutrient levels of receiving waters are established to
prevent concentrations from building up which would
© lead to uncontrollable algal activity
e cause harmful physiological effects among consumers
e interfere with certain water treatment systems.
While there is much controversy as to how much nutrient is too much,
the following maximum levels have been recommended by the Committee on
Water Quality Criteria (Ref. 8) to prevent eutrophication:
Phosphates - 0.015 mg/f (ppm)
Nitrogen - 0.3 mg/f (ppm)
The U.S. Public Health Service (Ref. 9) recommends the following limit
on nitrogen in surface and drinking waters:
10 mg/f of nitrate nitrogen
It has been found that the consumption of water having nitrate levels
exceeding this limit by infants may lead to a serious blood disease:
methemoglobinemia ("blue babies"). The Committee on Water Quality
Criteria (Ref. 8) points out that difficulties with coagulation in
water treatment plants often result when concentrations of complex
phosphates exceed 0.1 mg/^ . (mg/f = milligrams of substance per liter
of water, on a dry weight basis.)
The eutrophication and methemoglobinemia problems are usually encountered
when pollutants enter the water system constantly (assuming the system
is not quiescent). Shock loadings (as might occur from street runoff)
would be of less consequence in well mixed, free-flowing rivers. In
the case where street runoff enters lakes or swamps, however, nutrients
could accumulate, eventually reaching and passing recommended levels.
Data on loading intensities of nutrients found on street surfaces are
summarized in Table 8. Percent-by-weight values can be thought of con-
veniently as the "strength" of the dry solids collected from the street
surface. These strength values vary somewhat from one land-use cate-
gory to another, but only over a moderate range. This is evident from
the plots in Fig. 16. These data, based on the analysis of samples from
numerous cities, imply that all street surface contaminants are similar
in composition from site to site (at least from the standpoint of phos-
phates, nitrates and Kjeldahl nitrogen). It would be pure speculation
to extend this conclusion very far, but it is interesting. It was assumed
at the outset of the study that algal nutrients would probably be found '
in greatest concentration in residential areas because of the use of
fertilizer in domestic gardening. The data here do not support this
hypothesis.
60
-------
PHOSPHATES
2000
1000
D)
III
X
**/> "
4
3
2
s .1
KJELDAHL NITROG
1
CL
' " innn
'ro
u
* 0
EN
I_ X 0
4- 1*
I ^
1 Hi
9- n
2 "= o
u
1
1 .
£ 1 §
NITRATES
a
200
0
.20
,10
n-fi
C D
w
J
1
Fig. 16. Nutrient Loading Intensities and Waste "Strengths'
Variation with Land Use
61
-------
Table 8
NUTRIENTS IN STREET SURFACE CONTAMINANTS -
VARIATION WITH LAND-USE CATEGORY
STRENGTH
(% by weight)
LOADING INTENSITY
(Ib/curta mi) (lb/1,000 sq ft)
Phosphates
Residential
Industrial
Commercial
0.113
0.142
0.103
1.07
3.43
0.29
12.3
39.4
3.41
Kjeldahl Nitrogen
Residential
Industrial
Commercial
0.218
0.163
0.157
2.04
3.94
0.45
23.8
67.1
5.17
Nitrates
Residential
Industrial
Commercial
0.0064
0.0072
0.0600
0.063
0.178
0.172
0.70
2.00
1.96
Note
: The term "strength," as used here, refers to the amount of
contaminant contained in the dry solids collected from the
street surface (on a weight basis), i.e., a phosphate value
of 0.1 percent would be equivalent to 1 Ib of phosphate per
1,000 Ib of sample.
Table 8 reports information on nutrient loading intensities as well as
strength. These values, expressed in terms of both pounds per curb mile
and pounds per 1000 sq ft, vary considerably with respect to land
use. However, the variations are due primarily to differences in total
solids loading intensities. Figure 16 indicates the range over which
values of loading intensity differ.
The distribution of nutrients by particle size is shown in Figs 17 18
and 19. Note that phosphates exhibit a distinct pattern, most being in'
the smaller size ranges. The values for total nitrogen vary widely,
exhibiting no definite pattern. Nitrates show a pattern similar to'
phosphates but less pronounced. The discrepancy between total nitrogen
and nitrates may be due to the presence of other nitrogen species which
were not measured.
62
-------
CTION of TOTAL
SOCIATED WITH EACH SIZE RANG
by wt)
2
F
ASS
(%
60
50
Composite from:
San Jose 1
20
10
Milwaukee
Bucyrus
Baltimore
1L
3
PARTICLE SIZE (microns)
" 3
Fig. 17. Variation of Total Phosphates with Particle Size
63
-------
01
O
N
oo
21
"So
30
20
10
Composite from:
San Jose
3
Milwaukee
Bucyrus
Baltimore
Atlanta
Tulsa
Phoenix II
i
3
n *3-
fl- o
Son Jose I
Seattle
fi
i
n ^r -o
" 2 S
PARTICLE SIZE (microns)
Fig. 18. Variation of Kjeldahl Nitrogen with Particle Size
-------
en
01
O
40
Composite from:
_J LU
11
o<
30
20
10
il
Milwaukee
Bucyrus
Baltimore
1
Atlanta
Tulsa
Phoenix
Illll
3
5
3
San Jose II
Seattle
PARTICLE SIZE (microns)
Fig. 19. Variation of Nitrates with Particle Size
-------
Coliform Bacteria
The presence and quantity of pathogenic bacteria in natural waters are
difficult to determine by routine analytical, methods. It has, there-
fore, become common practice to test for other bacteria which are known
to be associated with pathogens. The coliform group of organisms is
widely used for this purpose. These "indicator organisms" are found
naturally in the intestines of warm-blooded animals. Thus, when analysis
of water reveals the presence of these indicators, it is assumed that
contamination by feces has likely occurred. However, coliform organisms
also live naturally in common solids, although the particular type of
coliform is different. This difference can be determined readily through
routine bacteriologic analyses.
Two terms commonly used in describing the bacteriologic quality of water
are "total coliforms" and "fecal coliforms." These terms are actually
descriptive of test procedures rather than classes of organisms, but
are often used to describe both. "Fecal coliforms" are those types
which are found in warm-blooded animals and do not include soil bactexia.
Their presence has been found to correlate quite consistently with the
presence of various pathogenic organisms. "Total coliforms," on the other
hand, include both fecal coliforms and common soil bacteria. They are
not, therefore, considered to be as reliable an indicator of pathogenic
bacteria, given a water contaminated by an unknown source.
It is generally assumed that the presence of fecal coliform bacteria
in a water supply signifies contamination by sewage and, therefore, the
possible presence of pathogens. It has been shown (Ref. 10) that swim-
ming in water containing high total coliform counts increases the
probability of contracting paratyphoid, diarrhea-enteritis, minor gastro-
intestinal disturbance, and eye, ear, nose and throat infections.
Drinking from a water supply which has a high total coliform count
obviously increases the likelihood of contracting any of these illnesses.
The Public Health Service has established drinking water standards which
are accepted by many state and local regulatory agencies. The standards
for bacteriological quality are expressed as the maximum permissible
number of total coliform organisms measured per volume of water sample.
If the supply has less than 2.2 total coliforms/100 ml it is considered
generally acceptable. If greater than 4 per 100 ml, immediate remedial
action is required (Ref. 9).
In this study, both total and fecal coliforms were measured during the
field test series using standard membrane filter techniques almost
immediately after samples were collected. Table 9 summarizes the total
and fecal coliform counts observed on street surfaces, expressing them
by land-use categories.
66
-------
Table 9
COLIFORM BACTERIA IN STREET SURFACE CONTAMINANT -
VARIATION WITH LAND-USE CATEGORY
Fecal Conforms
Residential
Industrial
Commercial
Total Coliforms
Residential
Industrial
Commercial
STRENGTH (a>
(10s org/lb)
15.4
1.82
175
80.8
187
79.9
LOADING
(106 org/curb mi)
6,100
2,600
34,000
60,000
150,000
116,000
INTENSITY
(106 org/1,000 sq ft)
70
30
390
696
1,760
1,300
Note: The term '"strength," as used here, refers to the number of
coliforms observed in street surface samples, related to the
amount of sample collected (on a dry-weight basis). Standard
membrane filter techniques were used throughout for identify-
ing and enumerating coliform organisms. The abbreviation
"org" refers to "number of coliform organisms" observed in
the analysis.
The distribution of conforms by particle size was not determined
because the heterogeneous character of the material , the necessity of
performing the tests in the field to restrict any growth and reproduc-
tion of the coliforms , and the chance of physically disturbing the
clumps of fecal matter which would change the size distribution.
Comparing the coliform counts obtained from the dry swept samples and
from the flushed liquid sample indicates that the coliforms do not associ-
ate preferentially with either the liquid or the solids. This may imply
that the coliforms are distributed randomly throughout the size ranges.
The cleaning efficiency for the coliforms would therefore closely resemble
the cleaning efficiency of the solid material.
Note that the data for total coliforms are more consistent than those
for fecal (by land use as well as the spread of values found). This may
be because the fecal coliform test is more complex or because fecal
matter tends to be located in high concentrations in small areas (there-
by reducing:sampling reliability). The strengths of the street dirt may
be the most important issue here. The observed strengths vary with
land use for both total and fecal coliform counts. The total counts
show little variation by land use, with the residential and commercial
areas showing the lowest counts. The fecal coliform counts show a wider
67
-------
spread, with the industrial being the lowest and commercial the highest.
The loading intensities per unit area or length of curb reflect the
total amount of dirt collected.
It should be noted that these values cannot be used as a basis for
estimating the coliform levels in the receiving waters. In most instances,
coliforms die off rather rapidly in receiving waters (although notable
exceptional cases have been observed where rapid regrowth has occurred).
For this reason, given data for the amounts found on streets, it is un-
wise to speculate at all as to the coliform levels in receiving waters.
Heavy Metals
Heavy metals are of concern because of their high potential toxicity
to various biological forms. Samples collected from street surfaces
in many cities were analyzed for the following metals: zinc, copper, lead,
nickel, mercury, and chromium (samples were preconcentrated before analyz-
ing for mercury). Atomic absorption techniques were used. Early in
the sampling program tests were run for arsenic and cadmium but, since
only insignificant amounts were detected, these tests were discontinued.
The samples were composited in various ways and analyzed to reflect
trends of particular interest. Table 10 reports the heavy metals loading
intensities (in pounds per curb mile) found in each of the cities tested.
Figures 20 and 21 show how the heavy metals are distributed between major
land-use categories (considering all cities together). Figure 22 shows
distributions by particle size (for composites prepared from samples
collected in several cities).
The thing to note here is that, from the standpoint of concentration
alone, zinc and lead have the heaviest loadings, chromium and nickel the
lightest. These trends are borne out in all of the cities tested. It
should not be concluded, however, that these metals are necessarily the
worst polluters; they may be, but this cannot be stated at the present
time. The toxic effect of a given metal on an aquatic environment is
dependent upon a number of complex and rather poorly understood factors.
One of the most important factors is the form of the particular metal.
The data reported here are the total amounts of such metals present,
without regard to their chemical/physical states (i.e., their valence,
whether they are tied up into complex inorganic or organic compounds,
etc.). Analyses of such materials should be performed as part of a more
definitive future study. At this time it is possible only to consider
the significance of finding such metals in their most toxic form,
recognizing the dangers inherent in making such speculations. It is
strongly urged that the conclusions drawn below be adequately qualified
if ever quoted out of context.
68
-------
Table 10
HEAVY METALS LOADING INTENSITIES (Ib/curb mile)
San Jose-I
San Jose- I I
Phoenix- I I
Milwaukee
Baltimore
Atlanta
Tulsa
Seattle
Arithmetic
Means
ZINC
1.4
.28
.36
2.1
1.3
.11
.062
.37
.75
COPPER
.49
.020
.058
.59
.33
.066
.032
.075
.21
LEAD
1.85
.90
.12
1.51
.47
.077
.030
.50
.68
NICKEL MERCURY CHROMIUM
.19
.085
.038
.032
.077
.021
.011
.028
.060
.20
.085
.022
-
-
.023
.019
.034
.080
.10
.14
.029
.047
.45
.011
.0033
.081
.12
Before proceeding to a discussion of the potential significance of each
heavy metal, consider their distribution by land use and particle size.
Figure 20 shows how heavy metals are distributed by major land-use
category. For all metals except mercury, loading intensities (in Ib/curb
mi) are heaviest in industrial areas and lightest in commercial areas.
Before any conclusions are drawn regarding the implications of these
data, it is well to consider Fig. 21 which expresses the distribution in
terms of the "unit strength" of the street surface materials (percent
by weight; i.e., pounds of metals per 100 pounds of dry solids). Here
the distinct trends as to land use disappear. This is probably because
the dominant patterns in total solids loadings overshadow patterns in
concentration levels of metals.
Figures 22 and 23 show the distribution of heavy metals between particle
size ranges. The plots, which are based on data for samples collected
from five cities, show little trend except for lead. There seems to be
a distinct tendency for lead to be associated with fine particles. If it
is assumed that antiknock gasoline additives are the principal source of
lead found on street surfaces, then these results are as would be expected ,
since particulate exhaust emissions would be very fine indeed.
The following paragraphs provide specific information on each of the
heavy metals found here.
69
-------
E
_Q
O
u-i O
X _J
25
20
15
o 10
JL
1
' 8.
N
1
I
NOTE:
Mercury samples taken in SJ I only
. .-III.
2 z
ij
RESIDENTIAL
INDUSTRIAL
COMMERCIAL
Fig. 20. Heavy Metals Loading Intensities on Street Surfaces -
Variation with Land Use
-------
l/>
Q
_i
O
t/1
Z
>
Zinc
,10
.05
Copper
LAND-USE CATEGORIES
Lead
Nickel
Mercury Chromium
J,
Fig. 21. Heavy Metals Concentrations - Variation with Land Use
-------
to
At\
in
ou
*
r
9
20
S
N
X
_J LU
-S 10
21 10
"oQ
z £
§§
< 01
^ 01
(i ^ n
Zinc
Copper
1
Samples from SAN JOSE II and SEATTLE
Lead
Nickel
Mercury
1
Chromium
I
S-OOOo -^--OOoO -*-oOOO Ti-'OOOO ->**OOOo -^-OOoO
-^S88 2SS88 2SSg8 2SS88 2SS8S 2SS88
vs^S- "'s^S^ vs^i^ vs^2^ vs^S^ Vsi27<
^^jt CN^J- , CN-^- ^"^3; r~ ^3; CNI^
PARTICLE SIZE (microns)
Fig. 22. Heavy Metals Concentrations - Variation with Particle Size
-------
FRACTION OF TOTAL
ASSOCIATED WITH
EACH SIZE RANGE
(% by wt )
60
60
ZINC
50
40
30
20 -
10
COPPER LEAD
NICKEL
MERCURY
CHROMIUM
11
50
40
30
10
3>O O
Tj- -*
~- CSi 00
PARTICLE SIZE
Fig. 23. Heavy Metals in Street Surface Contaminants - Variation by
Particle Size for Bucyrus, Atlanta, Tulsa, and Phoenix II
-------
ZINC - Most common zinc compounds are not particularly
toxic in low-to-moderate concentrations; nor are they
particularly soluble in water. It is estimated that
people consume on the order of 10 to 15 mg of zinc
daily in their diets (Ref. 11). From the standpoint
of water supplies, 5 ppm is the USPHS drinking water
limit (Ref. 9) (Concentrations of 25 to 30 ppm have
an objectionable taste and appear milky.) Aquatic
organisms are more sensitive than humans to zinc.
Concentrations as low as 0.1 to 1.0 ppm have been found
lethal to fish and other aquatic animals (Ref. 10).
Copper is reported to have a synergistic effect with
zinc toxicity (i.e., a given concentration of zinc
becomes more toxic to certain species when copper is
present in the solution).
Analysis of street surface contaminant samples
indicates that zinc is present in higher loading
intensities than other heavy metals (see Table 10).
Observed values range from a low of 0.062 Ib/curb mile
(Tulsa) to a high of 2.1 (Milwaukee); the mean for all
cities tested was 0.75 Ib/curb mile. Zinc was not found
to associate with any particular size range of particles.
Sources of zinc in street surface contaminants have not
been identified specifically; however, substantial
quantities of zinc are used in formulating tire rubber
compounds.
COPPER - In humans and other higher organisms, copper
is not particularly toxic. It does not exhibit cumulative
effects, as do many other heavy metals. USPHS drinking
water standards limit copper to 1.0 ppm (Ref. 9). Recom-
mended limits for irrigation water are 0.1 ppm, 0.05 ppm
for salt water organisms, and only 0.02 ppm for fresh-
water organisms. These values recognize the fact that
copper is toxic to lower biological forms (indeed,
copper compounds are typically used in low concentrations
to control aquatic weeds and algae).
Loading intensities for copper on street surfaces
loadings of zinc and lead and the light loadings of
chromium and nickel (see Table 10). Observed loadings
range from a low of 0.02 Ib/curb mile (San Jose II) to a
high of 0.59 (Milwaukee); the mean for all cities tested
was 0.21 Ib/curb mile. Copper was not found to associate
with any particular range of particle sizes. The sources
of the copper in street surface contaminants have not
been identified.
74
-------
LEAD - The effects of lead on biological forms are also
quite varied. In vertebrate animals, lead is a cumulative
poison which typically concentrates in bone. It is esti-
mated that humans consume on the order of 0.33 mg daily
in their diets. USPHS drinking water standards limit
lead to 0.05 ppm. At somewhat higher concentrations, it
has been reported to be moderately toxic to fish and
other aquatic organisms (Ref. 10).
Loading intensities for lead in street surfaces were
quite high, second only to zinc (Table 10). They ranged
from a low of 0.030 Ib/curb mile (Tulsa) to a high of
2.0 (San Jose II); the mean for all cities tested was
0.68 Ib/curb mile. Figure 22 reflects the very strong
tendency lead has for being preferentially associated
with small particle size range solids (nearly 90 percent
of the total lead found was with particles smaller than
246 microns, the size of a fine, silty sand). The term
"associated" is used here because it is not known whether
the lead exists in a compound whose particles are this size
or if the lead is somehow adhering to particles of this
size. Probably both situations exist. If the primary
source of lead is gasoline antiknock compounds (a plausible
speculation, but one that should be investigated), then it
is consistent that the bulk of material found would be associ-
ated with very fine particles.
NICKEL - This heavy metal is not considered harmful to man
in normal concentrations; no USPHS limit for nickel in
drinking water has been established. It is, however, mod-
erately toxic to aquatic organisms and can be very toxic
to plant life, depending on the chemical form (Refs.
12,13).
Of all the heavy metals tested here, nickel was found
to have the lowest loading intensities, ranging from a low
of 0.011 Ib/curb mile (Tulsa) to a high of 0.19 (San
Jose I); the mean for all cities tested was only 0.060
Ib/curb mile. Nickel was not found to be concentrated
appreciably in any particular size range of particles.
The sources of nickel in street surface contaminants have
not been identified.
MERCURY - In both its free state and in many of its com-
bined forms, mercury can be highly toxic to a broad range
of biological forms. Indeed, this element has recently
been the subject of much controversy, in public as well
as technical circles. Many studies are presently under-
way to develop a better understanding of mercury's role
75
-------
In various environmental problems. Thus, a definitive
discussion at this point would probably be of little
long-term value. However, it is probably safe to state
that mercury can be expected to be detrimental to aquatic
ecosystems at concentrations as low as 0.005 pptn.
Mercury was found to have only moderate loading
metals (Table 10). Observed values ranged from a low of
0.019 Ib/curb mile (Tulsa) to a high of 0.30 (San Jose I);
the mean value for all cities tested was .081 Ib/curb
mile. Mercury was not found to associate with any par-
ticular size range of particles. The source of mercury
in street surface contaminants has not been identified.
CHROMIUM - The toxicity of chromium is distinctly dependent
upon its chemical form. The metal form Cr° is extremely
common but virtually inert, whereas the hexavalent ion Cr+
is extremely toxic. USPHS drinking water standards limit
hexavalent chromium to 0.05 ppm but state no limit for
trivalent forms (Ref. 9). While its physiological effects
are poorly understood, chromium is not known to be a cumu-
lative poison to humans (Ref. 10). Toxic effects on lower
biological forms are variable. Limits of 100 ppm for
fisheries and 5 ppm for irrigation water have been recom-
mended.
Chromium was not found in substantial quantities in
street surface contaminants (only nickel was lower).
Observed loading intensities ranged from 0.0033 Ib/curb
mile (Tulsa) to 0.45 (Baltimore), the mean value for all
cities tested being only 0.12 Ib/curb mile. These values
were for total chromium (i.e., all chemical forms taken
together). Considering the fact that vehicle bumpers and
trim are typically plated with chromium, these low values
would suggest that the amounts of trivalent and hexavalent
chromium present on street surfaces are probably very low
indeed.
Pesticides
The widespread presence of pesticides in the environment has recently
caused much public and private concern because of their poten-
tial for upsetting ecological balances. Gas chromatographic techniques
have revealed the presence of various combinations and concentrations
of organic pesticides in all the street surface samples tested.
Organic pesticides in particular were examined because of their
high persistence in the environment (Ref. 14). The specific
76
-------
substances analyzed for are listed in Table 11 along with their respective
detection limits (i.e., the lowest concentrations which can be measured
quantitatively by the gas chromatographic methods employed here). Not
all, however, were found to be present in street surface samples. Tables
12 and 13 report the loading intensities for all pesticides found in the
samples tested. In the interest of obtaining maximum benefit from
available funds, samples collected from land-use areas of several cities
were combined into composite samples prior to analysis (the analytical
costs for pesticide determinations are quite substantial) . Several con-
clusions can be drawn from the data in Table 12. First, the chlorinated
hydrocarbons: p,p-DDD, p,p-DDT, and dieldrin are typically found in
highest concentration. Also, PCB's (polychlorinated biphenyls) are pres-
ent in higher concentrations than pesticides per se. Finally, the amount
of these materials (taken all together, pesticides and PCB's) is really
rather high, being on the order of 0.00125 Ib/curb mile for the cities
tested.
It is well to discuss the role of PCB's here. While these industrial
chemical compounds are not used as pesticides, they do share many of
their properties. They are included here because they, like the chlorin-
ated hydrocarbons, are the subject of much controversy. They have
repeatedly been found to correlate with detrimental environmental effects
(primarily birth defects in wildlife). They have also been found to be
extremely long-lived and are believed to be widely distributed through-
out domestic and worldwide ecosystems (Ref. 15). Another important reason
for including PCB's here is that their presence can cause interference
with analyses for other pesticides. The magnitude or significance of
that interference cannot be estimated, however.
Samples from Milwaukee, Bucyrus, and Baltimore were analyzed to show
variation with respect to land use and particle size. Table 14, which
reports the concentration of each pesticide by major land-use category,
reveals no consistent patterns. Figure 24 indicates that ODD, DDT and
dieldrin all tend to associate with finer particles but that PCB's
associate with coarser particles. The association of pesticides with
fine particles supports the speculations made at the outset of the
study. No explanation is given for why the PCB's favor the larger
particles.
The interpretation of observed pesticide levels is difficult indeed.
It should be appreciated that, at the present state of the art, accept-
able levels of pesticides in the environment at large are very much a
matter of speculation (i.e., no one can say how much is too much).
Further, it should be understood that while we conducted many tests and
had many analyses run, this effort should be still considered as
"spot-checks" rather than an accurate representation of situations in
the country as a whole. The important factor, however, is that these
materials are present in rather significant quantities. Organic
pesticides are normally measured in parts per billion by weight (values
77
-------
Table 11
DETECTION LIMITS FOR PESTICIDE ANALYSES
Chlorinated Hydrocarbons
DDE
p,p-DDD
o,p-DDT
p,p-DDT
Chlordane
Dieldrin
Endrin
Lindane
0,-BHC
Heptachlor
Aldrin
Kelthane
Heptachlorepoxide
Me t hoxy chl or
Toxaphene
Thiodan
Polychorinated biphenyl
Liquid
Samples
(ppb)
0.1
0.1
0.1
0.1
0.5
0.1
0.2
0.1
0.1
0.1
0.1
0.2
0.1
1.0
2.0
0.1
1.0
Dry
Samples
(ppm)
0.01
0.01
0.01
0.01
0.05
0.01
0.02
0.01
0.01
0.01
0.01
0.02
0.01
0.10
0.20
0.01
0.10
Organic Phosphates (Methyl Parathion)
Liquid = 0.01 - 0.001 ppm
Dry = 0.05 - 0.005 ppm
78
-------
Table 12
PESTICIDE LOADING INTENSITIES
(1C-6 Ib/curb mi)
San Jose I
San Jose 11
and Seattle
Phoenix II,
Atlanta
and Tulsa
Milwaukee
Bucyrus
Baltimore
p.p-DDD
67
120
34
0.5
83
100
p,p-DDT
110
170
13
1.0
60
30
DIELDRIN
11
27
24
10
17
3.0
ENDRI N
2
0
0
0
0
0
LINDANE
17
0
0
3.1
0
0
METHOXY-
CHLOR
0
0
0
8500
1600
170
METHYL
PARATHION
20
0
0
0
0
0
PCB's
1,200
1,100
65
3400
650
1000
TOTAL OF ALL
PESTICIDES
AND PCB's
1,427
1,417
136
12,000
2,451
1,300
Table 13
PESTICIDE CONCENTRATIONS
Ib of pesticide/lb of dry solids)
San Jose I
San Jose II
Phoenix I
Phoenix II
Milwaukee
Bucyrus
Baltimore
Atlanta
Tulsa
Seattle
p , p- DDD
73
20
37
0.19
61
100
79
100
270
p,p-DDT
120
28
14
0.38
43
30
20
39
380
DIEiDRIN
12
4.4
26
3.8
12
3.0
55
74
59
ENDRI N
2.2
0
0
0
0
. o' -
0
0
0
LINDANE
19
0
0
1.2
0
0
0
0
0
METHOXY-
CHLOR
0
0
0 ,
3,100
1,200
170
0
0
0
METHYL
PARATHION
22
0
0
0
0
0
0
0
0
PCB's
1,300
180
71
1,300
470
1,000
150
200
2,300 '
79
-------
Table 14
PESTICIDE CONCENTRATIONS IN TOTAL SOLIDS (ppm)
p,p-DDD p,p-DDT DIELDRIN ENDRIN LINDANE METHOXY- METHYL PCB's
CHLOR PARATION
Residential
San Jose I
Milwaukee
Baltimore
Industrial
San Jose I
Milwaukee
Baltimore
Commercial
San Jose I
Milwaukee
Baltimore
0.082
0
0 .11
0 .060
0.
0 .020
0 .040
0 .020
0 .020
0.15
0
0.030
0 .091
0
0 .020
0 .030
0 .031
0 .031
0
0.009
0
0 .031
0
0 .018
0
0
0
0
0
0
0
0
0
0.058
0
0
0
0
0
0 .031
0 .001
0
0
0
0
0
2.5
0.19
0
3.6
0
0
1.8
0
0
0
0
0.037
0
0
0
0
0
0
2
0
1
2
1
0
0
0
.81
.0
.99
.5
.0
.0
.60
.99
.51
o
z
oo X
LLJ . ,
^<
_ I LU
Z
o<
58
< oo
50
40
30
20
10
Dieldrin
ODD
Pol /chlorinated
Biphenols (PCB)
ntt
p,p-DDT
PARTICLE SIZE (microns)
Fig. 24. Pesticide Concentrations - Variation with Particle Size
80
-------
of several ppb are not uncommon but are cause for some concern when found
in the environment). Street surface concentrations, when present, all
range in parts per million (on the order of a thousand times higher).
The fate and relative significance of the various pesticides must
be considered both in terms of residence on the street surfaces and
ultimately in the receiving waters. While pesticides' effects in soil
systems have received considerable study, aquatic mechanisms have not
been well documented to date.
While they reside on street surfaces (the site of net accumulation),
pesticides are subject to a number of degrading actions. Among these are
volatilization, decomposition by ultraviolet light and other radiation,
chemical degradation, microbial degradation, and sorption and desorption
by soil particles. Thus, depending on resident time and the above fac-
tors, a certain amount of in situ decomposition will occur on the street
surface. The importance of all this on the pollutional effects of street
runoff is questionable. When pesticides enter receiving waters, the
mechanisms listed above can apply to reduce their effect. However, at the
same time, biological "magnification" can occur. Degrading effects are
overshadowed by concentrating effects. Chlorinated hydrocarbons are
increasingly concentrated by many types of organisms with successive
steps up the food chain. This is especially true in the upper trophic
levels. Numerous cases of fish kills and damage to invertebrate popu-
lations have been reported (Refs. 16, 17, 18, 19). In addition, pesti-
cides tend to concentrate in sediments by adsorption, concentrating them
in regions containing additional biologic communities.
TRANSPORT OF CONTAMINANTS
Street surface contaminants are washed into receiving waters via
the route illustrated in Fig. 25. Contaminants are
freed from the street surface itself
carried transversely across the surface to the gutter
by the overland sheet-like flow
carried parallel to the curb line to the storm
sewer inlet by the gutter flow
dropped through a stormwater inlet and transported to
the receiving waters via storm or combined sewers
(where catch basins are present, some of the denser
particulates are caught by simple sedimentation).
The fact that contaminants move through this sequence is well known.
However, the relationships between the contaminants and the various
mechanisms involved are only poorly understood. Given this as a
81
-------
Fig. 25. Transport of Street Surface Contaminants by Runoff
starting point for this study, it was necessary to conduct a series of
substudies which would provide a basis for understanding (at least
empirically) what happens to elements of street surface contaminants on
their way to receiving waters.
The first such substudy was designed to experimentally determine
the manner in which contaminants are flushed from the street surface by
typical rainfall. A portable rain simulator was designed and built
(see Figs. 26 and 27). The simulator applies water uniformly over a fixed
section of street at various controlled flow rates. The water, supplied
from nearby hydrants, sprays vertically (4 to 8 ft) through hundreds of
small jets (0.018 in. dia), which break up into discrete droplets about
the size of common raindrops before they fall to the street. The device
produces a pattern on the street surface which has the appearance of a
moderate-to-heavy rainfall.
It was found that contaminant materials are removed via two mech-
anisms which operate simultaneously:
Soluble fractions go into solution; the impacting
raindrops and the horizontal sheet-flow provide good
mixing turbulence and a continuously replenished
clean "solvent."
82
-------
Fig. 26 . Mobile Rain Simulator
Particulate matter (from sand size to colloidal size)
is dislodged from its resting place by the impact of
falling drops. Once dislodged, even reasonably heavy
particles will be maintained in a state of "pseudo-
suspension" by the repeated impact of adjacent drops
creating a reasonably high general level of turbulence.
(A substantial amount of the contaminants was found to be
located down inside small pits, cracks, and other irregu-
larities in the street surface.)
The sheet-like flow of water across the surface carries the contaminant -
materials to the gutter. These mechanisms are easily discussed and
understood, but only on a qualitative basis.
Experimental studies were conducted in Bakersfield (California) to
determine the rate at which contaminants are washed off of streets, given
various levels of rainfall intensity. (Bakersfield was selected as a
site for the field tests because it was the nearest sizable city which
had not experienced any significant rainfall since the preceding summer
and therefore had a moderate-to-heavy loading of solids of all sizes
available to observe.) The influences of street surface characteristics
were also of interest here. Field tests were conducted wherein three
typical street areas (two asphalt and one concrete) were flushed by a
simulated rainfall for a period of 2-1/4 hours. Every 15 min during
that period, samples of liquids and particulates were taken for
subsequent analysis. At the end of the period, the streets were flushed
83
-------
Vacuum Box
Sidewalk
0' nmm' jo]
T '\
a
Gutter
Plan View
IHlsi; -'^11111?
Side View
Fig. 27. Rain Simulator and Sample Collection Syst
em
84
-------
thoroughly with a firehose to wash off any remaining loose or soluble
matter. Samples of this remaining material were also collected. Two
rainfall rates, 0.2 in./hr and 0.8 in./hr, were used. (The lower inten-
sity - 0.2 in./hr - is typical of a heavy rainfall. The high intensity -
0.8 in./hr - would be an unusually high sustained value for any area of
the country. However, since such values commonly occur for at least
short periods during ordinary storms, it was important to observe how
this very high rate removes contaminants.)
The preliminary flushing tests in Bakersfield provided much valuable
information. On the basis of that experience, we were able to make
several important modifications to our equipment and field testing
procedure. An important reason for conducting the tests was to determine
an appropriate sprinkling time and rate to be used as fixed parameters
in subsequent test series.
Samples were fractionated in Imhoff cones to separate settleable
and floatable matter from the water which contained dissolved, colloidal
and suspended matter. Each of these fractions was analyzed; the settle-
able solids were also separated into six particle size ranges by dry
sieving. The results of this test series are presented in Figs. 28
through 32.
The thing to note here is that, while the first runoff to reach the
curb was quite dirty, the subsequent runoff got clearer and clearer as
time went on. Admittedly this observation was predictable; yet it was
necessary to be established in terms of meaningful parameters. Stated
another way, it was observed that of the total amount of material which
could conceivably be flushed off by a given rainfall intensity, the
amount flushed off during each successive time period decreased in a
regular pattern. Likewise, the cumulative amount increased, approaching
the total loading as an asymptote. This pattern is shown clearly in
Figs. 28 through 32. The thing to note in these figures is that the
runoff patterns shown in the plots are remarkably uniform in shape and
vary little from test to test (even though a range of rainfall intensi-
ties and street surface types were used). It is also interesting to
note the similarity between curves of all particle sizes.
Mathematical analyses of the data plotted have revealed that the trans-
port of particles across street surfaces fits an exponential function
quite well, as shown in the following discussion.
It was assumed that the rate of removal of particles of a given size
from a unit area of street surface is proportional to the number of
particles of that size contained within the unit area, as well as to the
rate of water deposition on the area. These assumptions are expressed
85
-------
II
Si .0001
o i
FLUSHING TIME (hour,)
Fig. 28. Particle Transport
Across Street Surfaces -
Variation by Particle Size
1.00 ,=
0 I
FLUSHING TIME (hours)
Fig. 30. Particle Transport
Across Street Surfaces -
Variation by Particle Size
540-2,000k
5 i .0001
0 1
FLUSHING TIME (houn)
Fig. 29. Particle Transport
Across Street Surfaces -
Variation by Particle Sjze
r = in./hr
a Concrete 0.2
b Concrete 0.8
c Aged Atpholt 0.2
d NewAtpholt 0.2
e New Aipholt 0.8
10.00 c=
0 I
FLUSHING TIME (houn)
Fig. 31. Particle Transport
Across Street Surfaces -
Variation by Street Charac-
ter and Rainfall Intensity
86
-------
in mathematical terms as:
= krN
dt
where:
o i
FLUSHING TIME (hounj
Fig. 32. Particle Transport
Across Street Surfaces -
Variation by Street
Character and Rainfall
Intensity
N is the amount of particles of
the given size which remain on the
street surface at time t (expressed
in g/sq ft)
r is the rainfall intensity over
the area (expressed in in./hr)
t is time (in min)
k is a proportionality constant
(having the units of hr/in.min)
With the mobile rain simulator,
the intensity r is uniform with
respect to time and space (at
least within the bounds of the
test site). The above can then be
treated as an ordinary differential
equation whose solution is:
N = NQe
-krt
In the experiments, we measured the amount of matter removed with time,
rather than the amount remaining (indeed, the total amount, N0, was never
measured directly. The equation then becomes:
N =
where:
Nc is the amount of material
oi a given particle size
which has been removed during time interval t by a
rainfall of intensity r
N0 is the initial loading intensity of that material of
that particle size which could ever be washed from the
street by rain of intensity r (even as t approaches
infinity)
k is a proportionality constant dependent on street
surface characteristics.
The experiments carried out in Bakersfield were of sufficient duration to
establish the asymptotic values of NQ. These experiments established
the appropriateness of the developed relationship as shown in Figs.
28 through 32.
87
-------
Based on the exponential function that was derived from the prelimi-
nary flushing data, certain conclusions can be drawn on the rate and
amount of material that could be removed from a street by a given
rainfall.
The proportionality constant k for material removed from
a street surface by rainfall is dependent on street
surface properties but is not dependent upon rainfall intensity.
This means that the type of street (e.g., asphalt or con-
crete, coarse or fine surface, roughness, etc.), and the
condition of the street (i.e., old and cracked or new and
smooth) is a major controlling factor on how fast such
material would enter a storm drainage system.
« The amount of material (No) which is capable of being
removed from a street varies with the rainfall intensity r.
This means that for a given rainfall intensity on a street
surface for which the proportionality constant k is known,
the relative amount of material flushed into the sewer over
a given time period could be predicted.
It appears from the data that all particles of the size ranges examined
are removed from the street at approximately the same rate, given the
same rainfall conditions. Therefore, the street surface constant k is
virtually independent of particle size (i.e., dN/dt is not a function of
particle size). This is substantiated by the fact that the plot in Fig.
33 is essentially a horizontal line. Analyses of the liquid samples
showed that soluble, colloidal, and suspended materials removed from a
street surface show the same functional behavior as the settleable solids.
However, since no further separation of these fractions was made, it was
not determined if this is true for each of these fractions independently.
i .001
1,000
10,000
PARTICLE SIZE (micron.)
Fig. 33. Relationship between'Particle Size and Proportionality Constant
88
-------
Runoff carrying street surface contaminants flows across the street,
reaches the gutter, and mpves down the gutter toward the storm sewer
inlet, as shown in Fig. 25. While moving down the gutter, it mixes
with runoff from other sources (i.e., off sidewalks, driveways, surround-
ing land area, building drains, etc.). Thus, the runoff which arrives
at the storm sewer inlet is not street surface runoff, per se. This
study focuses on street surface contaminants only. The field techniques
employed here were carefully designed to include only those contaminants
which reside on street surfaces.
Since street surface contaminants are but a single source of all
contributions to storm runoff, we have attempted to determine how
important they are. Their "importance" can only be expressed relative
to all sewered storm runoff since the myriad of other contributions have
not yet been isolated for study (nor has unsewered storm runoff been
studied to any great extent). Table 37 in Section VI compares the
street surface runoff (calculated for a hypothetical city) with storm
sewer discharges observed in several U.S. cities.
89
-------
Section V
EFFECTIVENESS OF CURRENT
PUBLIC WORKS PRACTICES
-------
SECTION V
EFFECTIVENESS OF CURRENT PUBLIC WORKS PRACTICES
Current public works practices may: in some instances, result in a
reduction of pollution of receiving waters from storm water runoff.
Practices that may influence the pollution of receiving waters include:
street cleaning
catch basin cleaning
refuse and litter collection
street maintenance
sewer cleaning
snow and ice control
air pollution control
open area maintenance
construction
parking regulations.
Of the above mentioned practices, the role of street cleaning and the role
of catch basins in controlling or reducing the pollutional effects of
street surface contaminants were included within the scope of this study.
This section, therefore, deals with answering the question "HOW effective
are current public works practices in controlling pollution of receiving
waters from street surface contaminants?" More specifically, this in-
volves the discussion of:
existing street cleaning practices
street sweeping effectiveness
catch basin effectiveness.
EXISTING STREET CLEANING PRACTICES
Street cleaning practices throughout the nation were evaluated through
a review of the literature, and by conducting a detailed survey of current
practices in several sample cities. The effectiveness of current street
cleaning practices was also determined in each of the test cities and a
series of control tests was conducted utilizing a street surface contam-
inant simulant. (A description of the test areas utilized in each city
is given in Appendix B.) Effectiveness data from previously conducted
91
-------
street sweeping tests were evaluated and correlated with data obtained
in this study.
Present methods of cleaning streets fall into two categories: sweeping
and flushing. These methods , for the most part, are carried out by
machines specifically designed for that purpose - street sweepers and
street flushers. As an ancillary function, most municipal street clean-
ing departments are also responsible for catch basin cleaning and leaf
collection in the fall. In most northern cities a spring clean-up of
streets which have been snowbound all winter is common. Because the
bulk of accumulated trash and sand can be very great, this clean-up
often utilizes front end loaders, trucks, and hand crews which are
followed by sweepers and/or flushers. The following paragraphs will
describe the procedures and the equipment used for the above-mentioned
functions.
Street Sweeping
Machine sweeping accounts for the great majority of street cleaning
performed in most communities. This effort may be assisted by a limited
amount of manual sweeping in areas that machines cannot reach. Hand
cleaning is primarily used to clean those streets where the presence of
cars prevents the use of mechanical equipment. It is most often employed
in business districts where the emphasis is placed on keeping "visible"
pollution (such as papers, tin cans) under control. Manual methods are
also useful in supporting mechanical operations. For example, a hand
crew can follow a street sweeper and clean out catch basin inlets , sweep
up missed debris or assist in transferring debris from the sweeper to
trucks.
Motorized street sweepers are designed to loosen dirt and debris from the
street surface (this debris is normally most concentrated in the gutter
area), transport it onto a moving conveyor and deposit it temporarily
in a storage hopper; the sweeper also typically contains a dust control
system. Three basic types of sweepers are in use; as shown in Table 15,
the most common is a design which utilizes a rotating gutter broom to
move materials from the gutter area into the main pickup broom which
rotates to carry the material onto a belt and into the hopper. This
type of sweeper relies upon water spray to control the dust problem. A
wide variety of sweepers of this type is available. Included are those
which are self-dumping and those which have 3 wheels or 4 wheels. Three-
wheel sweepers are generally considered more maneuverable while 4-wheel
sweepers can generally travel at higher road speeds when not sweeping.
The second class of sweepers includes those which use a regenerative air
system. These sweepers are designed to "blast" the dirt and debris from
the road surface into the hopper with a portion of the air being recycled
A portion of the air is vented through the dust separation system. Such
sweepers may also use water spray for dust control.
92
-------
Table 15
SOME COMMONLY USED STREET SWEEPERS
TYPE
Pickup
Oroom
Air
NOTE:
MANUFACTURER
BUI Austin-
Western
Elgin
Elein
MB
Mobil
Wayne
Wayne
Wayne
Murphy
Murphy
Murphy
Tymco
Tymco
Tennant
Elgin
Ecolotec
MODEL
70
475
(White Wing)
Pelican
Cruiser
TE 4
984
945
933
4032
4042
4062
300
600
100
Whirlwind
Vacu-Sweep
Metro- Vac
This list is not comprehens
NO.
WHEELS
3
3
3
4
4
3
4
3
4
4
4
4
4
3
4
6
6
NO. MAIN BROOM
ENGINES WIDTH
(in.)
1
]
1
1
2
1
2
1
2
2
2
2
2
1
2
2
2
ive but does include
broom: the broom wid1
60
68
68
60
60.
58
58
58
58
58
58
80
96
42
60
60
the most
SIDE BROOM
DIAMETER
(in.)
36
36
36
47
42
45
45
45
45
45
45
Optional
Optional
32
28
20
20
commonly used
SWEEPING*"'
PATH
9 ft
8 ft
8 ft
n.a.
7 ft 6 in.
9 ft
8 ft
8 ft
10 ft
10 ft
10 ft
6 ft 8 in.
8 ft
6 ft
87 in.
72 "in
94 in.
sweepers. Sw<
SWEEPING
SPEED
(mph)
1-10
1-15
1-15
3/4 & up
1-12
2-8
1-15
1-15
1-12
1-12
1-12
1-8
1-10
0-12
n.a.
n.a
raping path
MAX.
TRAVEL
SPEED
(mph)
25
22
22
50
55
25
55
55
55
55
55
55
55
15
55
55
data are
head
WATER
SPRAY
yes
yes
yes
yes
yes
yea
yes
yes
yes
yes
yes
Optional
yes
yes
yes
HOPPER COMMENTS
CAPACITY
(3 cu yd)
4 j Model 40-B has 2- cu yd hopper
4j Model 375 has 3}- cu yd hopper
2j Self-dumping (8ft 6 in. lift)
4-1/3 Dumps from rear
4 Model TE-3 has 3-cu yd hopper
4 Model 973 has 3-cu yd hopper
4
3 Self-dumping (94 in lift)
4 Self-dumping (side dump)
"1
using 1 side broom.
-------
A third type, vacuum sweepers, has been in use in Europe for many years
and in limited use in this country for some time. Considerable interest
has recently been.generated by the introduction of new models. These
vacuum sweepers operate using both a broom for loosening and moving
street dirt debris and a vacuum system to pick up the debris. All
material picked up by the vacuum nozzle is saturated with water on entry
and passes into a vacuum chamber where the water-laden dust and dirt
drop out of the air stream.
Small, industrial-type sweepers may be considered as a subclass to the
vacuum sweepers since they generally utilize an enclosed vacuum system
for dust control. These small sweepers are most useful for cleaning
parking lots and parking garages. In industry they are used to sweep
factory floors and sidewalks. Since these machines are of very limited
use on city streets, no data are included here.
The basic procedure used when operating a street sweeper is for the
sweeper to travel next to the curb, cleaning one swath along the length
of the street and then returning on the other side. (Litter normally
accumulates in the gutter because of currents created by passing traffic
obviating the need to sweep the center portion of the street.) In some
cases, a second pass is made by the street sweeper along the curb to
increase the effectiveness of sweeping.
When the hopper of a street sweeper is filled, the material must be
dumped. It can be taken in the sweeper to a storage or disposal site or,
as is the more common practice, simply dropped in a convenient place
along the street sweeping route, preferably an inconspicuous side street.
In the latter case, the dirt and debris is later collected by truck crews
and usually a front-end loader. The majority of street sweepers dump
their hoppers from the bottom. However, several manufacturers make street
sweepers in which the hopper swings up on arms and can be dumped into a
truck directly, thus negating the necessity for a separate pickup crew.
The operating speed of most street sweepers falls in the range of 4 to
8 mph. This is an acceptable speed for performing sweeping operations
in residential and commercial areas where a sweeper has to maneuver
around cars which are blocking access to the curb. However, for cleaning
main arterial streets or freeways, an operating speed of 4 to 8 mph is not
only dangerous to the driver in the vehicle but can cause severe traffic
tieups. Therefore, several manufacturers offer a 4-wheel street sweeper
with an auxiliary engine to drive the brooms that can be used in sweeping
arterials (streets or freeways) at speeds up to 15 mph, thereby reducing
the danger somewhat.
Auxiliary engines provide constant speed and power to brooms and elevators
thus allowing the operator to vary sweeper speed as necessary for street
conditions (i.e., traffic, debris type and loading, etc.) and maintaining
broom speed. This is advantageous in minimizing debris left on streets
at intersections.
94
-------
One of the most serious problems encountered in street sweeping concerns
vehicle parking. Increases in the use of vehicles and unavailability of
off-street parking result in the occupancy of the gutters by parked
vehicles. In congested urban areas, it is not unusual to find virtually
the entire curb sides of streets occupied by parked vehicles. The City
of Baltimore has instituted a no-parking regulation during scheduled
street sweeping hours and has found that public acceptance (especially
residents of the street in question) has been encouraging.
Street Flushing
Street flushing as presently conducted serves only to displace dirt and
debris from the street surface to the gutter. The volume of water uti-
lized is insufficient to transport the accumulated litter to the nearest
drain. Most public works agencies use flushers for: (1) aesthetic pur-
poses or (2) moving material out of travel lanes quickly.
A street flusher consists of a water supply tank mounted on a truck or
trailer, a gasoline engine driven pump for supplying pressure, and three
or more nozzles for spreading the water as directional sprays. The large
nozzles on the flusher are individually controlled and are usually placed
so that one is directed across the path of the flusher and one on each
side is pointed out toward the gutters. This arrangement makes it possible
to flush an entire street and also provides flexibility in operation.
The capacity of the water tank on a street flusher varies from 800 to
3500 gallons. The nozzle pressure of the water usually is between 30 and
55 psi. The amount of water delivered must be proportional to the speed
of the vehicle and the pumps must be capable of supplying sufficient
water at suitable pressures. Specifications of street flushers are given
in Table 16.
During normal operation, a street flusher will travel to its assigned
route, fill its tank at a fire hydrant, and proceed along the length of a
street flushing material into the gutter. On narrow streets, the whole
street can be flushed in one pass. However, on wider streets (those
wider than about 22 ft) multiple passes are needed.
Catch Basin Cleaning
The major purpose of a catch basin is to intercept grit and other
materials which, if allowed to enter the sewer system, could form deposits
and clog the sewer. Catch basins, which are typically located under the
inlet structures, act as sedimentation basins and collect large objects
that enter the inlet structure. Over a period of time, these catch basins
become full and have to be cleaned. The material in a catch basin then
has to be periodically removed and hauled off to a selected dump site.
95
-------
Wayne
Central
Eng. Co.
Table 16
COMMONLY USED STREET FLUSHERS AND EDUCTORS
COMPANY
Etnyre
Etnyre
Etnyre
Rosco
M S
Wayne
MODEL
Leader
Clipper
Superliner
MTA
Vactor
Sanivac 1600
TANK CAPACITY
(gal)
800-3,000
800-3,000
800-3,000
1,200, 1,600
2,100
2,500
3,300
FLUSHING PUMP
WIDTH SIZE
(ft) (gpm)
Variable 750
Variable 750
Variable 750
Variable 750
42 650
45 600
COMMENTS
Flusher
Flusher
Flasher
Flusher
Used as
Used as
only.
only.
only.
only.
a vacuum truck.
a vacuum truck,
Sanivac 1300
VAC-ALL
2,600
1,700-2,200
45
42
capacity - 16 cu yd.
600 Used as a vacuum truck,
capacity - 13 cu yd.
650 Used as a. vacuum truck,
capacity - 10-16 cu yd.
There are three principal methods used to clean catch basins: manual,
eductor, and a clam-shell or orange-peel bucket. The most common manual
method is to bail out the water and then dip out the material deposited
in the catch basin, piling it on the pavement so it can be hauled away.
Long-handled dippers are generally used for lifting the material. The
catch basin material is then shoveled into trucks and hauled to dumps
and sanitary landfills.
The eductor method of cleaning catch basins consists of using a large
vacuum truck with a sewer jet hose which is lowered into the catch basin
from the inlet. The vacuum pump on the truck is utilized to suck the
catch basin material up into a large watertight tank on the truck. Most
of these trucks can serve a double purpose in that when they are not being
used for catch basin cleaning, their tanks can be filled with water so
they can be used as street flushers. The vacuum method is the most
sanitary of those in general use as there is little leakage and the catch
basin material does not run on the street; however, some material such
as very large rocks and boards, cannot be picked up by the vacuum'hose
and has to be removed manually. Specifications of eductors are given
in Table 16.
The second mechanical method employs a bucket machine or hoist. The
process consists of lifting the solid catch basin material with an
"orange-peel" or clam-shell bucket operated by a hydraulic crane which
96
-------
dumps the material into a truck. This operation is comparatively fast.
However, some catch basin inlets are too small to allow the passage of the
bucket through them and the loose, runny catch basin material tends to
run out of the bucket as it is transferred to a waiting truck, thereby
dirtying the street.
Special Problems
Keeping streets clean also involves various ancillary operations, in-
cluding leaf pickup, snow removal, and, in some cases, removal of aban-
doned automobiles and the disposal of dead animals. The following
paragraphs will present brief descriptions of these various operations
as they are generally practiced.
Leaf Collection. In parts of the country where deciduous trees abound,
street cleaning departments have the seasonal problem of collecting the
fallen leaves. The collection and removal of leaves is important for
several reasons: wet leaves on a pavement surface impair vehicular trac-
tion and create slippery conditions that may be almost as dangerous as
icy pavements. Also, leaves clog catch basins and inlet gratings and,
unless removed, impede the transport of runoff and can cause major
flood damage.
The collection and disposal of leaves is done by a variety of methods,
depending on the job conditions and equipment vailable to the municipal-
ity. Municipal collection schedules also vary widely. Some cities clean
frequently during the leaf season so that a relatively small amount of
leaves is picked up during each cleaning period. Other cities will pick
up leaves at regular intervals throughout the leaf season, while still
others will collect the leaves only once at the end of the season.
The methods of collecting leaves include manual and several types of
machine collection. In the manual method, street crews simply sweep the
leaves into a pile and load the pile on a truck. Machine methods for
leaf collection include using a street sweeper alone, a street sweeper
with a trash screen mounted on the front to push the leaves ahead of it,
and front-end loaders either to load collected leaves or to gather leaves
in their buckets. The method used by a front-end loader or a street
sweeper with a leaf blade in front to collect leaves is for the vehicle
to proceed down the street pushing the leaves in a pile ahead of it
until the leaves begin to overlap the front of the blade. At this point
the vehicle leaves the pile, proceeds around it, and starts pushing up a
second pile for collection. When the machine is operating as a sweeper,
it collects leaves (the same as it does other street debris) in a hopper
and dumps the Copper in a convenient location when full. Collection of
the piled-up leaves can be done using either a front-end loader or a
vacuum truck. In the case of the front-end loader, the leaves are
picked up by the front-end loader bucket and deposited into an accompany-
ing truck which hauls the leaves off for disposal. The vacuum truck
97
-------
uses a suction hose to suck the leaves into the body of the truck where
they are further compacted, then proceeds to a dumping place when the
turck is filled.
Snow Removal. The presence of snow on streets, of course, prevents normal
cleaning operations. This suspension of normal cleaning operations over
the winter months can in itself lead to the buildup of heavy deposits on
streets. However, snow and ice control procedures can also add to the
presence of pollutants on streets. Most highway authorities in the
United States have a policy of maintaining "bare pavement" to protect
lives and promote safety. Thus, ice and snow are removed as quickly as
possible from roads and highways. Deicing compounds (road salts) are
usually applied at rates of 400 to 1,200 Ib per mile of highway per
application. Over the winter season, many roads and streets commonly
receive on the order of 20 tons of deicers per lane mile. This is equiv-
alent to 100 tons of salt or more applied per mile of roadway for
multiple-lane highways.
The reported use of sodium chloride (Refs. 4 and 5), calcium chloride and
abrasives in the United States for the winter of 1966-1967 amounted to
6,320,000 tons sodium chloride, 247,000 tons calcium chloride and
8,400,000 tons of abrasives.
While most of the deicing salts applied in urban areas will eventually be
channeled into the sewer system with the runoff water in the spring
thaws, insoluble abrasive materials tend to remain on the streets and
gutters. Thus, the spring cleanup is a routine practice in many northern
cities and may involve the use of manual crews and front-end loaders to
help in digging out the heavy deposits.
Another concern, however, is the presence of dirt and debris and particu-
larly deicing compounds which are incorporated into the snow, slush and
ice which is picked up and removed. Ultimately, this finds its way into
local receiving waters. Usually this is done by carting the snow
directly to a body of water, dumping it in, and allowing it to melt.
Some cities, however, use snow melting machines which melt the snow as
it is collected. The melt water, including any salt, then flows directly
into the sewer system.
Abandoned Cars. Abandoned or "junk" cars are a problem in most cities,
but the problem is most obvious in those communities where parking regu-
lations are used in support of street cleaning operations. However, the
problem of parked cars, in general, is a major deterrent to street
cleaning so that the inclusion of the additional junk cars normally is
not of major concern. A survey conducted as part of this study indicated
that, in most cities, the police are responsible for removal of "junkers"
once notified by the street cleaning department.
98
-------
Disposal of Animals. Although many small animals and a few larger ones
are killed on streets and highways in the course of a year (thereby cre-
ating a concentrated source of pollution) they are generally not dealt
with as part of the street cleaning program. Rather, some organization,
either within the city government or under contract to the city, is
responsible for removing such bodies and ultimately disposing of them.
Hence, although such a function is found within some city governments
under the street cleaning program, it will not be considered further in
this report.
Survey of Street Cleaning Practices in Selected Cities
One of the subtasks in this study was to determine the type and extent of
various street cleaning practices across the nation. To this end a
9-page questionnaire (see Appendix F) based on one used by the American
Public Works Association was prepared and used as the basis for inter-
views in selected cities. The sample includes most of the cities where
the street surface sampling program was conducted (in a few instances
questionnaires were not returned) plus a few cities which were selected
for special characteristics (such as extreme winter conditions). The
cities for which data were obtained are listed in Table 17. The table
includes the miles of streets which are regularly swept, population data
and climatological data. A summary of the data obtained is given in
Tables 18 through 22. Although the cities selected generally fall in
the moderately large population category, they do represent a wide spec-
trum of climatic conditions and street cleaning programs.
This survey of street cleaning practices was not intended to be compre-
hensive, since other excellent data sources are available. These
include a recent survey undertaken by The American City magazine (Refs.
20 and 21) and the Western Pennsylvania Chapter of the APWA and Institute
for Urban Policy and Administration, University of Pittsburgh (Ref. 22).
In the following analysis of street cleaning practices, these sources
will be referenced where appropriate.
Table 23 lists cleaning practices in selected cities. All cities were
found to have a comprehensive sweeping program. About one-half of the
cities also had a flushing program; most of the remainder used flushers
to some small extent. Several cities relied heavily upon manual cleaning
programs (which includes both gang and "white wing" crews), and all but
one city used manual crews to some extent. Normally the use of manual
crews is restricted to downtown areas or small business areas and to
daytime hours on a daily basis. One interesting exception is San
Francisco where manual crews are used in many of the residential areas,
the reason being that parking is extremely limited in the city. In many
neighborhoods sweepers can never get to the curb (this also accounts
for the high use of flushers in San Francisco).
99
-------
Table 17
CHARACTERISTICS OF CITIES SURVEYED
San Jose
Phoenix
Milwaukee
Bucyrus
Baltimore
Atlanta
Tulsa
Seattle
Minneapolis
St. Paul
San Francisco
Lawrence, Ka.
NOTE:
Source
/IILES
r STREETS
n
s a
1,165
1,450
1,701
n.a.
2,000
1,750
n.a.
1,280
1,000
896
850
150
a
^ i
w
95
100
75
n.a.
75
85
n.a.
16
50
67
80
28
O
M
536, 965
580,275
709,537
13,200
895,222
487,553
328,219
524,263
167,685
107,848
704,217
45,143
0
to
CTJ
£
3 D1
K 3
138
187
90
n.a.
75
136
49
82
53
52
45
n.a.
of Population data: Statistical
J < ^
i M c
< Si 2
12
7
33
37
43
44
30
34
19
25
28
34
Abstract
3|~
<§i
T
0
30
29
23
2
12
45
74
44
0
20
of U.S.
Bureau of Census.
Source
of Weather
data:
National
Climatological
Summary
DAYS/YEAR
WITH
I PI TAT I ON
O
w
g £
62
51
124
135
102
107
82
150
127
133
69
98
, 1970,
of Data,
DAYS/ YEAR
W 33°F
2
i§
20
7
162
123
108
73
95
28
152
160
0
98
U.S. Weather Bureau, Vol. 20, 1969
In San Jose the area has increased from 56 sq mi in 1960 to 138
sq mi in 1970.
The information for Lawrence, Kansas, is from Ref. 23.
T = trace.
n.a. = not available.
100
-------
Table 18
STREET SWEEPING EQUIPMENT IN SELECTED CITIES
3-
WHEEL
San Jose
Phoenix
Milwaukee
Baltimore
Atlanta
Seattle
Minneapolis
St. Paul
San Francisco
Lawrence
1
-
13
-
-
-
1
2
11
3
SELF-
4- DUMPING
WHEEL (3 -Wheel)
14
21
9
26
24
18
17
5 7
3
-
AVERAGE
VACUUM LIFE
TYPE (yr)
9
7-8
12-15
5
7
8
5
5
10
4
AVERAGE
DOWNTIME
(%)
15
17-22
Varies
w/season
25
10
20
25
25
25
5
Note: Ref. 20 indicates that of some 250 cities surveyed, 68 percent
use 3-wheel sweepers, 27 percent use 4-wheel sweepers, while
5 percent use "others" (either 3- or 4-wheel) which include
air- or vacuum-type sweepers.
Table 19
OPERATING.J3PECIFICATIONS FOR SWEEPERS IN CITIES SURVEYED
San Jose
Phoenix
Milwaukee
Baltimore
Atlanta
Seattle
Minneapolis
St. Paul
San Francisco
Lawrence
OPERATING
SPEED
(mph)
6-10
5-7
5-6
5-6
6
7-8
2-15
2-15
4-8
4J-5
MAIN BROOM
MATERIAL
P-P
Wire
P-P
P-P
P-P
1 P-P
P-P
P-P
P-P
P-P
MAIN BROOM
LIFE
800-900 mi
927 mi
80-120 hr
1,200-1,500 mi
4-6 wk
1,000 mi
500-800 hr
4-6 wk
1,000-1,200 mi
1,300 mi
PATTERN
WIDTH
(in.)
6-8
6
4-6
6-8
7
5-7
5-6 in.
4-6
6
2-6
GUTTER
BROOM
MATERIAL
Wire
Wire
Wire
Wire
Wire
Wire
Wire
Wire
Wire
Wire
GUTTER
BROOM MAXIMUM
LIFE ROUTE
(curb mi/
shift)
300 mi
1-2 wk
140-160 hr
50-60 mi
2 wk
600 mi
2 wk
1-2 wk
350 mi
350 mi
34
32
22
32
30
33
28
30
35
30
NOTF ;
polypropylene or other plastics.
101
-------
Table 20
FLUSHERS IN CITIES SURVEYED
San Jose
Phoenix
Milwaukee
Baltimore
Atlanta
Seattle
Minneapolis
St . Paul
San Francisco
Lawrence
NOTE: n.a. =
AVERAGE
TOTAL LIFE
FLUSHERS (yr)
0
i 6-8
2
11 8
3 7
8 8
3 n. a.
7 10
10 9
2 10
not available.
Table
CATCH BASIN CLEANING
AVERAGE
DOWNTIME
ROUTE
(%) (curb mi/shift)
-
5
Phasing Out
25
3
3
n. a.
5
20
5
21
-
-
5-6
-
30
5
n. a.
45-night
35-day
IN CITIES SURVEYED
TOTAL NO. NUMBER NUMBER
CATCH FREQUENCY OF CLEANED/ OF
San Jose
Phoenix
Baltimore
Seattle
Minneapolis
St. Paul
San Francisco
BASINS CLEANING YEAR
n.a. as required 3,600
3,100 6/yr 18,600
32,200 1/yr or as about
required 68,000
20,000 1/yr 20,000
35,000 as required n.a.
10,000 as required n.a.
50,000 as required 12,000
METHOD EDUCTORS
Eductor 1
Hand
Eductor 1
Hand
Eductor 4
Eductor 8
Eductor 1
Hand
Eductor 2
Eductor 10
NUMBER
OF MEN
CREWS CREW
1 2
12 2
2 2
6 4
2 3
7 2
1 2
1 2
1 1
8 3
COST.
S/CATCH
BASIN
9.52
15.00
3.00
5.64
n .a .
15.00
n.a .
9.12
NOTE: n.a. not available.
102
-------
Table 22
SWEEPER DEBRIS COLLECTED, BY MONTH, FOR FOUR CITIES
(as reported by City Public Works Departments)
M
O
CO
Month
Jan
Feb
Mar
Apr
May
June
July
Aug
Sept
Oct
Nov
Dec
Baltimore
4826
4539
4982
5476
5536
7475
7880
6575
8129
5897
5270
5317
Curb Miles
San
Francisco*
3514
4251
3633
4444
4563
5572
4672
4182
3263
3903
4360
3953
Cleaned
San Jose
6263
5708
6857
6557
6168
6409
6584
6165
5241
7138
5443
5773
Debris Removed
Phoenix
17445
14533
14300
16490
14928
14462
14043
13625
13796
15201
12243
13986
Baltimore
3000
2884
3228
3068
3180
3792
4292
3716
4132
3008
2968
2936
San
Francisco
660
786
636
813
783
879
729
702
627
714
795
696
cu/yd
San Jose
1266
1292
1350
1187
1067
1238
1327
1242
1061
1064
878
1220
% Total Debris
Phoenix D
5282
3804
4553
3838
4120
3781
3841
3600
8497C
9708
6586
6683
Baltimore
7.5
7.2
8.0
7.6
7.9
9.4
10.7
9.2
10.3
7.5
7.4
7.3
San
Francisco
7.5
8.9
7.2
9.2
8.9
10.0
8.2
8.0
7.1
8.1
9.0
7.9
San Jose
8.9
9.1
9.5
8.4
7.5
8.7
9.4
8.8
7.4
7.5
6.2
8.6
Phoenix
8.2
5.9
7.1
6.0
6.4
5.9
6.0
5.6
13.2
15.1
10.2
10.4
71,902
50,310
74,296 175,052
40,204
8,820
14,192 64,293
3 Curb miles/mo for flushers follow a similar pattern with a yearly total of 67,511 curb-miles.
b In tons (weighed at landfill); a conversion factor of 1.0 tons = 1.0 cu yd can be used for comparative purposes.
C The sharp rise in September is due to the initiation of a street sealing program; residual chips are picked up by sweeping operations.
-------
Table 23 also lists the number of sweepers and flushers in the various
cities; from these data has been calculated the number of sweepers and
flushers per 1000 miles of street. In all cases sweepers are more nu-
merous than flushers. Since the results of the URS survey indicate that
route miles per shift covered are about the same for the two types of
equipment, then it certainly follows that sweepers provide the major
portion of street cleaning programs. Reference 21 indicates that in a
survey of 152 cities virtually all cities with a population in excess
of 500,000 use flushers extensively with the percentage dropping with
decreasing city size, ultimately reaching an average of 16 percent for
cities of under 25,000 population. This same study also shows that a
significant number of cities (estimated at about 20 percent of all cities)
use street flushing in direct support of sweeping operations. This operation
is usually performed only on selected streets (i.e., streets located in
central business districts) during selected times of the year.
Table 23
CLEANING PRACTICES IN SELECTED CITIES
MAJOR CLEANING PROGRAMS
SWEEPER FLUSHER MANUAL
San Jose x 0 0
Phoenix x M M
Milwaukee x M x
Baltimore x x x
Atlanta x M x
Seattle x x x
Minneapolis x MM
St. Paul x x M
San Francisco x x x
Lawrence x x M
EQUIPMENT
SWEEPER
15
21
22
26
24
18
18
14
14
3
FLUSHER
0
1
2
11
3
8
3
7
10
2
EQUIPMENT/ 1,000-mi STREET
SWEEPER
12
14
12
13
13
14
18
15
16
20
.9
.5
.9
.0
.7
.1
.0
.6
.5
.0
FLUSHER
0.
1.
5.
1.
6.
3.
7.
11.
13.
7
2
5
7
3
0
8
8
3
NOTE :
Manual cleaning normally used in business districts only.
x major use.
0 none.
M - minor use.
104
-------
Reference 21 also shows that smaller cities have, on the average, more
sweepers per thousand miles of streets than do larger cities; the median
value for cities under 25,000 population is 20 sweepers/I,000 miles of
streets whereas for the cities of over 500,000 population the equivalent
value is 15. This, of course, is not to imply that smaller cities
necessarily have cleaner streets since both the loading on streets and
the frequency of sweeping must be taken into account. In fact, an
interesting area for research would be to ascertain how effectively various
cities do utilize their sweepers; that is, what fraction of the time does
the sweeper actually engage in sweeping. For example, cities which uti-
lize their sweepers for both day and night operations show better utili-
zation than those that sweep only at night.
Another variable is downtime, which the URS survey indicated to average
about 25 percent; however, some cities reported downtime as low as 10
percent. This variation is partly attributable to the life expectancy
of the equipment which was found to range from 5 to 15 years. (Downtime
does increase with equipment age and presents an interesting problem in
optimization: that is, when does downtime rise to the point where replace-
ment becomes the desired mode?). Not surprisingly, flushers have a
considerably longer average life span than sweepers, and their downtime
is much lower, averaging about 5 percent.
In the cities that URS surveyed, 4-wheel sweepers were predominant;
however, in Ref. 21 the reverse is true. As will be discussed later,
no appreciable differences appear to exist between the cleaning effec-
tiveness of the two types of sweeper, although the supposedly better maneu-
verability of the 3-wheel sweeper might improve overall effectiveness
somewhat. The admitted advantage of the 4-wheel sweeper is higher travel-
ing speed which normally allows off-street dumping of collected debris.
Two interesting trends were also encountered. The first is that self-
dumping sweepers seem to be gaining wider acceptance, possibly because
they eliminate the need for street-side dumping and suosequent transfer.
The other even more recent trend is the development and acceptance of
vacuum-type sweepers. Several major manufacturers are now marketing such
sweepers and, based upon tests conducted previously (Ref. 24, which will
be discussed later), vacuum sweepers do indeed pick up more dirt and
debris than conventional sweepers. However, because such sweepers were
not available in the test cities, they were not evaluated during this
research study.
105
-------
Variables in sweeper characteristics and operations which are known to
affect sweeping effectiveness include main broom fiber, strike or pattern,
and sweeping speed. Main broom fiber as found in the URS survey was
predominantly polypropylene. However, a number of cities (25 percent as
reported by Ref. 21), do use steel bristle main brooms and about 16
percent of the cities use natural fiber brooms. Broom life, even for the
same type broom, is extremely variable and is difficult to compare since
it may be reported in different units. Steel wire bristles are generally
used in gutter brooms (all cities in the URS survey utilized steel
wire gutter brooms). Gutter brooms last somewhat longer than main brooms,
with cities reporting gutter broom life averaging from 300-600 curb miles
swept.
Strike (i.e., that fraction of broom circumference which touches the
pavement)averaged about 5 in. However the range was from 2 in. to 8 in.
and seemed to vary with both sweeping conditions and broom wear. Sweeping
speed was reported to range from 2 to 15 mph (the latter for sweeping
of main arterial streets during daylight hours). The average operating
speed was closer to 6 mph. Reference 21 also found that the median
operating speed was between 5 and 5-1/2 mph.
Sweeping costs have been long reported in dollars/curb mi swept. Reported
costs range widely. For example, The American City survey (Refs. 20 and
21) found average costs, by city class, to range from a low of $2.18 to
a high of $8.42. Variations for individual cities can be even greater
than this. This very wide range is partly attributable to labor rates
and labor utilization. For example, the URS survey revealed that equip-
ment operator's pay scales range from a minimum of $2.60 to a maximum
of $7.00 per hour. Another variable is equipment costs, with depreciation
and maintenance costs likely to differ considerably between cities.
Finally, cities typically use different overhead rates and accounting
procedures. The final result then is that attempts to compare
costs between cities is difficult and may lead to erroneous conclusions.
For this reason, we did not pursue the dollar costs per curb mile.
Rather, the URS survey focused on information relating to the number
of miles swept and the amount of debris picked up.
The most pertinent information found in the URS survey includes:
the average number of sweepers in use
the number of miles of city streets swept
the number of curb miles swept per unit time (usually
per month)
the quantity of debris collected per unit time.
The manner in which these four factors can be assessed is shown in Fig.
34 (data are shown in Table 22). Figure 34a shows the sweeper utilization,
based on the curb miles swept (per year) by each sweeper (that is, the
106
-------
H-
TO
SWEEPING FREQUENCY
(times swept/year)
SWEEPER UTILIZATION FACTOR
(curb miles swept/vehicle)
d
i»
4
H-
U)
o
o
H)
CD
(D
(D
d
(D
8
Baltimore
San Francisco
San Jose
Phoenix
CD
4
H.
O
O
(D
PICKUP RATE (cu yd/curb mile)
cr
PICKUP PER SWEEPER (cu yd/vehicle)
(S3
o
O
H-
H-
-------
total number of curb miles swept was divided by the number of sweepers in
active use). The variation is considerable and shows that Phoenix utilizes
its sweepers more effectively than Baltimore. Figure 34b illustrates
the unit pickup per sweeper, determined by dividing the total quantity of
debris collected (per year) by the number of sweepers. Again, Phoenix is
much higher than the other cities. The reason can now be deduced by
considering Fig. 34c which shows the average number of times streets in
the city are swept each year. This value is obtained by dividing the curb
miles swept per year by the total number of miles of swept street in the
city multiplied by 2 to approximate total curb miles. It must be pointed
out that this average includes streets that are swept daily along with
some that are swept only occasionally. Figure 34c then shows that Phoenix
sweeps its streets much more frequently than any of the other cities;
consequently, the number of curb miles swept per sweeper is higher. More
importantly, the amount of debris picked up per sweeper is higher. Data
presented in Table 21 show that street cleaning operations in Phoenix
remove the largest quantity of debris.
Another way in which the data may be expressed is shown in Fig. 34d
which shows the pickup rate for debris (determined by dividing the total
quantity of debris collected by the total curb miles swept). Again, there
exists considerable variation, but in this case Phoenix is in the middle
of the group. While the pickup rate does not constitute an absolute measure
of effectiveness(because it does not consider the debris loading on the
street surface) it might well provide a valuable comparative measure for
similar neighborhoods and land uses. For example,the output of two sweepers
and operators over an extended period of time could be recorded and
assessed to determine any relative differences. However^ if
pickup rate could be used as an absolute measure of effectiveness, it
would be simple to say that the sweeping operation that picked up the
most debris per mile of curb was the most effective.
At this point we are not attempting to impute any great significance to
any one of these particular forms of expressing sweeper utilization, but
we do believe that with a sufficiently broad data base from a number of
cities such a presentation would be most useful to an individual
city in determining the efficiency of its sweeper performance. As
suggested in a recent APWA publication (Ref. 25), frameworks for per-
formance evaluation need to be developed for street cleaning which will
allow the public works engineer to maximize performance and to minimize
cost.
STREET SWEEPING EFFECTIVENESS
The effectiveness of existing street cleaning practices , as related to
water pollution control, was researched in the following manner:
published data were reviewed and information obtained
from street cleaning equipment manufacturers
108
-------
in situ evaluation tests of street sweepers were
conducted in cities where the street surface sampling
program was conducted
controlled tests were done utilizing a simulant of
street surface contaminants.
The results of each investigation are described in the following paragraphs.
Review of Pertinent Literature
Information needed to establish the effectiveness of existing street
cleaning practices as related to water pollution control was obtained from
a review of published data and interviews with street cleaning equipment
manufacturers. The primary sources of data containing pertinent informa-
tion are in a series of reports (Refs. 26 through 29). These describe a
comprehensive series of tests conducted to determine the effectiveness of
street cleaning practices when utilized to remove dry particulate matter
from paved areas (synthetic fallout material). The particle size range of
the material utilized in these tests approximated the dust and dirt fraction
of street surface contaminants found to constitute the major portion of
the street pollution load.
Little or no data relating to cleaning effectiveness were obtained from
the major manufacturers of street cleaning equipment. References 30, 31
and 32 were provided by the Newark Brush Co., which has conducted tests
on the performance of various broom types. Reference 33 reports the
results of the sweeper efficiency study conducted by the Wayne Manufac-
turing Co. for the American Public Works Association in connection with
the APWA study (Ref. 1) on urban runoff.
A considerable amount of data relating to the cost of street cleaning
is available; however, the data are usually dependent upon the street
cleaning practices followed by the reporting city, the accounting prac-
tices followed by the city and the prevailing labor rates, fuel costs,
etc. As indicated previously, a recent APWA report (Ref. 25) describes
procedures for determining costs for street cleaning operations. A
summary of the pertinent findings obtained from the various sources
follows.
The usefulness of street sweepers and street flushers to decontaminate
paved areas was evaluated in a number of full-scale test programs conduc-
ted by the U. S. Naval Radiological Defense Laboratory (NRDL) during
the 1960's. The tests were designed to:
(a) determine the effectiveness of motorized and
vacuumized street sweepers and conventional
street flushers when removing dry particulate
matter of various particle size ranges and
initial mass levels
109
-------
(b) establish the limitations of existing street
cleaning equipment with respect to the removal
of dry particulate matter
(c) reveal equipment design or operational improvements
which would increase their effectiveness.
The various test parameters included:
Machine type
motorized sweeper (Wayne 450)
vacuumized sweeper (Tennant 100)
motorized flusher (Etnyre Nozzles)
Operational procedures
forward speed (1st, 2nd and 3rd gear)
Mass loading
20-600 g/sq ft (44-1300 lb/1000 sq ft)
Particle size
six particle size ranges (44 micron to 2000 micron)
Surface type
asphalt
concrete
The measurement techniques for determining mass loadings in these studies
utilized a radioactive tracer which allowed the direct measurement of
residual mass levels of less than 1 percent of the initial mass. This
technique is much preferred over a material weight-balance technique
which is subject to error when the residual mass levels are low.
The NRDL studies are perhaps of most interest because a theoretical
explanation of street sweeper performance has evolved from them. In
the studies undertaken at Camp Stoneman (Ref. 27), an equation was evolved
based upon results such as those shown in Fig. 35. The equation, which
is found to express well the variables under consideration, is:
M = M* + (M - M*) e~kE
o
where M = the mass remaining after sweeping (g/sq ft)
M = the initial mass before sweeping (g/sq ft)
M = an irreducible mass remaining after any amount
of sweeping (and dependent upon the type sweeper,
the surface, and particle size)
110
-------
k = a dimensionless empirical constant dependent upon the
sweeper characteristics
E = the amount of sweeping effort involved (equipment minutes/10UU
sq ft swept)
1,000 ^
100
EFFORT (man-min/10 sqft)
SOURCE: Ref. 27
Fig. 35. Effectiveness of Conventional Motorized Street
Sweeping on Portland Cement Concrete at Three
Mass Levels
This study showed that the amount of mass remaining on the street can be
effectively reduced by making repeated passes over the same area with the
sweeper. Also, certain operational changes can improve performance.
For example, it was found that a sweeper moving at 2-1/2 mph removed
almost as much dirt in one pass as a sweeper moving at 5 mph removed in
2 passes. This initial NRDL study gave rise to other similar studies
which, for economy, were of the strip test variety (Refs. 24 and 29).
Ill
-------
A series of strip tests conducted by NRDL to evaluate the effectiveness
of vacuuraized and motorized street sweepers revealed that, for a given
level of effort , vacuumized sweeping was more effective than motorized
sweeping. Table 24 compares the effectiveness for motorized sweeping and
vacuumized sweeping for a range of initial mass levels and particle
sizes and for similar levels of effort.
Table 24
COMPARISON OF REMOVAL EFFECTIVENESS FOR MOTORIZED
SWEEPING AND VACUUMIZED SWEEPING
MACHINE
TYPE
RELATIVE
EFFORT
20 g/ft
177-300u
GEAR (%)
100 g/ft2
74-177U
(%)
600 g/ft
74-177|a
(%)
Motorized 2.17 2 92.5 58.0 46.0
Vacuumized 2.88 2 95.0 94.5 89.5
Motorized
Vacuumized
4.32
5.83
1
1
94.5
98. 5
62.6
91.4
NOTE: Tests conducted on asphaltic concrete. Results are for 1 pass
in 2nd gear and 1 pass in 3rd gear. Reference 24.
o
g/ft = initial mass level.
IJL - Particle size range of simulant.
% = Removal effectiveness = Mo-M*/Mo x 100.
Effort as applied by a street sweeper is not a continuous function which
can be truly represented by curves or mathematical equations. This is
because sweepers are designed to operate at the governed engine speed
which produces the most effective broom operation. The series of dis-
crete forward speeds obtained with a set of transmission gears combine
with integral numbers of passes over the surface swept to produce dis-
tinct levels of effort that can be applied.
Effort is defined as a factor which is inversely proportional to the for-
ward speed and directly proportional to the time spent covering a given
area. Its units are equipment minutes per 1000 sq ft of area swept
112
-------
Relative Effort =
1200
Forward Speed (ft/min)
The constant factor 1200 was chosen arbitrarily so that none of the
sweepers had a RE < 1.0 at the fastest speed (3rd gear) tested. Unit
relative effort corresponds to a forward speed of 1200.
Area coverage rates for unit relative effort are dependent upon broom
widths and can be directly calculated from the relationship
Area Coverage Rate (sq ft/min)
broom width(in.)
12 (in./ft)
x 1200 (ft/min.)
The area coverage rates do not account for the overlap of sweeper passes
or the turn around or dump cycle time. As defined, RE values are additive.
Figure 36 compares the rela-
tive performance of street
flushing with street sweep-
ing methods. The results
were taken for similar test
conditions, i.e. , mass load-
ing, particle size and sur-
face type. For the test
conditions, it can readily
be seen that a flusher can
be a much superior method
for moving street contami-
nants in the dust and dirt
fraction. (Note that the
mobile flushing units em-
ployed in these tests were
specially designed with high
pressure pumps and highly
effective nozzles.) Con-
ventional street flushers
would not be nearly as ef-
fective. It should also be
noted that a street flusher
doe's not'pick up contami-
nants but transports ma-
terial to and along the curb.
In Refs. 30 and 31 the Newark
Brush Co. summarizes a series
of tests designed to measure
the cleaning efficiency of
various types of main brooms
on several types of street
surface contaminants. The
reports conclude that:
SURFACE: Asphalt
INITIAL MASS: 20 om/sq ft
0 2 . 4
RELATIVE EFFORT
12 14
Fig. 36. Comparison of Cleaning
Performances of Motorized
Street Sweeping and
Motorized Street Flushing
113
-------
It is more difficult to pick up fine road debris than coarse
material; sweeping pattern and broom speed are critical factors.
A worn broom sweeps all types of debris better than a new one.
Crimped wire and fiber brooms proved more efficient in these
tests than plastic or plastic-wire mixtures for all debris at
the broom patterns used. (Plastic-fiber brooms used substanti-
ally, smaller patterns as the subsequent text will explain.)
The sweeping pattern contributes greatly to cleaning efficiency;
small patterns leave uncleaned streaks in depressions on irregu-
lar road surfaces.
At faster road speeds, proportionally higher broom rotation speeds
should be employed.
Figure 37 shows the quantitative effect of sweeping patterns on the
efficiency of debris removal and Fig. 38 shows the effect of increasing
broom speeds on residual debris with the pattern and sweeper speed main-
tained constant. Figure 39 shows the effect of sweeper speed on the
residual debris.
0234
Pattern inch
Fig. 37 The Effect of Pattern
on Residual Debris
400
* 300
200
100
0
WIRE BROOM
J L
0 10 20 30 40 50 60 70 80
Percent Debris Remaining
Fig. 38 Debris Pick-up vs Brush
Speed
0123
Forward Speed mph
Fig. 39 The Effect of Sweeper Speed
on the Residual Debris
114
-------
A factor to be considered in the relationships shown in Figs. 38 and 39
is that these tests were conducted with a single engine sweeper. Thus,
within the limitations of several gear ratios, higher broom speeds result
from higher engine speeds. Higher forward speeds also result in a prob-
lem of maintaining contact between the broom and the pavement; this
reduces sweeping effectiveness. Thus, the desirability of an auxiliary
engine to maintain a constant broom speed at the most efficient broom
rpm is apparent. A conclusion one can reach when examining Figs. 38 and
39 is to sweep at a forward speed of approximately 4 mph with a broom
speed of 100 rpm.
These tests have indicated the importance of establishing performance
requirements for street cleaning practices and in particular, operational
guidelines, i.e., broom patterns, broom rpm, sweeper speeds, etc., to
achieve the desired effectiveness.
In Situ Sweeper Evaluation Tests
Field evaluations of current street sweeping practices were conducted
in a number of the cities included in the street surface sampling pro-
gram. A description of the test sites is given in Appendix B. Briefly,
the test procedure used was as follows:
A street was selected and two adjacent test areas were
cordoned off. The first area was used to determine
initial loading; the latter, sweeping effectiveness.
» Initial loading was determined by hand sweeping the test
area and picking up the accumulated debris, then flushing
with a water jet, collecting the runoff, and determining
the solid content. Material so collected was then analyzed
to determine the initial loading. (See Appendix A for
details on sampling procedure.)
The street sweeper (furnished and operated by the local
public works department) passed over the second test area.
The surface was again swept and flushed to ascertain the
remaining solids. All test areas were 40 ft in length and
8 ft wide.
The removal effectiveness, in percent, was determined by
the formula:
(initial loading)-(final loading)
Removal Effectiveness = (initial loading) ~ X 10°%
Table 25 summarizes the tests conducted and gives information on sweeper
type and other operating parameters. Table 26 summarizes the street
cleaning effectiveness obtained in each test. Table 27 summarizes the
particle size distribution of street surface loadings before and after
115
-------
sweeping and Table 28 summarizes the effectiveness of removal by loading
location across the street. The uniformity of the initial loading on
the adjacent test areas proved to be quite good. In a series of repli-
cate tests, 50 percent of the adjacent test areas had initial loadings
which did not vary more than + ;5 percent. In only one case (out of 8)
did the non-uniformity of the loading exceed + 20 percent.
Table 25
SUMMARY OF STREET CLEANING EFFECTIVENESS TESTS
CITY
Milwaukee
Milwaukee
Baltimore
Scottsdale
Atlanta
Tulsa
Phoenix
TEST
NO.
Mi-3
Mi-10
Ba-7
Sc-4
At-9
Tu-6
PII-2
LAND
USE
3
10
7
4
8
6
2
STREET
EQUIPMENT
TYPE CONDITION TYPE
Concrete
Concrete
Asphaltic
Asphaltic
Asphaltic
Concrete
Asphaltic
Good
Fair
Fair
Good
Good
(food
Pooi'
Wayne
Mobil
Wayne
Wayne
Elgin
Elgin
Mobil
945
TE-4
945
985
Pelican
Pelican
TE-3
PICKUP BROOM
CONDITION SPEED
(rpm)
Fair 2,000
Fair 1,200
New 2,000
Worn (50%)
Fair u.a.
Worn (50%) n.a.
Fair "1,"700
STRIKE
(in.)
8
7
5J
5
6
4
5
VEHICLE
SPEED
(Gear) (mph)
3rd
2nd
2nd
2nd
2nd
2nd
2nd
5
3
4
5
3
4
5
.5
.4
.0
.5
.4
.1
.5
NOTE :
See Table 2 for land-use identifiers.
See Appendix B for information on parking and traffic conditions.
All sweepers equipped with polypropylene main pickup brooms and steel gutter brooms. Gutter brooms left
.operating in all-tests. Spray bar used-on all tests. - ,.-.-;:-
n.a. not available.
Table 26
SUMMARY OF STREET CLEANING EFFECTIVENESS
TEST
NO.
Mi-3
Mi-10
Ba-7
At-9
Tu-6
PII-2
Sc-4
INITIAL LOADING
(g/sq ft)
1.69
1.07
- 4.93
2.58
6.01
10.03
3.36
(lb/1000 ft2)
3.72
2.36
10.86
5.68
13.24
22.09
7.40
RESIDUAL LOADING
(g/sq ft)
0.89
1.31
4 . 37 :
1.75
3.89
3.78
1.49
(lb/1000 ft2)
1.96
2.88
9.62
3.85
8.57
8.32
3.28
REMOVAL
EFFECTIVENESS
(%)
47
11
32
35
62
56
NOTE: Removal effectiveness is for dirt and dust fraction. Could not
determine residual mass on Mi-10 due to wet street conditions.
116
-------
Table 27
REMOVAL EFFECTIVENESS VERSUS PARTICLE SIZE DISTRIBUTION
PARTICLE
SIZE RANGE
(micron)
> 2,000
840-2,000
246-890
104-246
43-104
< 43
Total (g)
Overall Eff.
ATLANTA
INITIAL
LOADING
(g)
175
103
375
231
66
43
993
(%)
RESIDUAL
LOADING
76
14
56
29
136
187
498
50
TULSA
INITIAL
LOADING
(g)
1,438
418
690
544
415
324
3,829
RESIDUAL
LOADING
(g)
142
181 i
588
595
549
431
2,486
35
PHOENIX
INITIAL
LOADING
(g)
535
308
2,190
1,273
425
175
4,906
RESIDUAL
LOADING
(g)
240
107
224
381
614
498
2,064
62
SCOTTSDALE
INITIAL RESIDUAL
LOADING LOADING
(g) (K)
217
439
915
421
213
87
2,292
43
124
415
287
134
14
1,017
56
Table 28
REMOVAL EFFECTIVENESS ACROSS STREET. SURFACE
STREET
SECTION
S-l
S-2
S-3
S-4
S- 5
Total (g)
ATLANTA
INITIAL
LOADING
(g)
2
4
24
11
952
993
RESIDUAL
LOADING
(g)
2 .
24
18
8
446
498
TULSA
INITIAL
LOADING
(g)
191
271
391
279
2,697
3,829
RESIDUAL
LOADING
(g)
188
1,040
387
283
588
2,486
PHOENIX
INITIAL
LOADING
(g)
216
324
2,261
72
2,033
4,906
RESIDUAL
LOADING
(g)
62
56
944
469
533
2,064
SCOTTSDALE
INITIAL
LOADING
(g)
56
614
694
500
428
2,292
RESIDUAL
LOADING
(g)
88
286
180
256
207
1,017
NOTE: See Appendix B for parking and traffic conditions. In Scottsdale, street
sloped toward centerline to facilitate runoff.
Street layout as follows:
40'
J2
Curb S-5 S-4
96'
S-3
Variable
S-2
S-l
Street
117
-------
Controlled Sweeper Evaluation Tests^
A series of controlled sweeper evaluation tests was conducted to proof-
test procedures for establishing performance criteria for street sweepers
Test parameters included:
debris loading density
sweeper type
forward speed of sweeper
broom type
rotational speed of broom.
Other variable parameters, such as broom pattern (strike) and pressure,
were set at the manufacturer's recommended specifications. The tests
were conducted in the City of San Jose on a newly constructed asphalt
paved street with concrete gutters and curbing.
Test areas, 50 ft by 8 ft, were delineated on the test street and each
test area was cleaned thoroughly by vacuum cleaning and hose flushing.
After the surface was dry, a synthetic street surface contaminant (de-
scribed in Appendix G) was spread on the street surface utilizing a cali-
brated lawn fertilizer spreader. Several sampling pans (1 sq ft in area)
were placed on each test area during dispersal of the simulant for use
in determining the initial loading density. Table 29 summarizes the
initial loading density for each of the tests conducted.
Street sweepers, provided by the City of San Jose and a commercial
street sweeping organization (San Jose Commercial Sweeping Co., Inc.)
were utilized in the test areas. Table 30 summarizes the test results
and gives information on sweeper type and operating conditions utilized
during each test. A single pass was made over the test area, and the
residual synthetic street contaminant loading was determined by follow-
ing the procedures (see Appendix A) utilized in the street surface
sampling program.
The test procedures developed in this series of tests proved worthwhile
and effective for use in establishing performance criteria of street
sweepers.
DISCUSSION OF SWEEPING EFFECTIVENESS
Three general types of tests have been conducted: in situ street tests,
controlled tests in which paved areas are artifically given a variable
or uniform loading, and strip tests in which a narrow path of material
is laid down to be removed by the pickup broom (the gutter broom is
normally disengaged). Since the latter type of test is easily run and
readily reproducible, most of the data generally available on street
118
-------
Table 29
INITIAL LOADING DENSITY OF SIMULANT
TEST
NO.
1
2
3
4
5
6
ACROSS
S-l
0.76
1.89
3.50
1.40
3.10
1.20
STREET LOCATION
(g/sq ft)
S-2 S-3
1.
5.
12.
5.
14.
4.
58 13.5
00 23.7
20 40.4
50 12.5
30 49.6
80 15.7
TOTAL
LOADING
(g)
1,584
3,059
5 , 610
1,940
6,700
2.170
Street layout as follows:
"t of street - Test areas average 50 ft long
1
1
1 1
24"
24"
24"
curb
n
S-l
S-2
S-3
Table 30
SUMMARY OF RESULTS
.CONTROLLED SWEEPER EVALUATION TESTS
TEST
NO.
SWEEPER
TYPE
SWEEPER
SPEED
(mph )
1
2
3
4
5
6
Mobil-TE-3
Mobil-TE-3
Mobil-TE-3
Mobil-TE-3
Tymco-300
Tymco-300
4
5
6
6
1
1
.8
.6
.4
.4
.0
.0
PUB
ENGINE
(rpm)
1,750
1,200
1,750
1,750
(b)
(b)
SIMULANT
INITIAL
(g)
1,
3,
5,
1,
6,
2,
584
059
610
940
700
170
LOADI NG
RESIDUAL
(g)
629
682
2,334
1,425
3,735
1,388
REMOVAL
EFFECTIVENESS
('
60
77
58
26
44
36
7.)
.3
.7
.4
.5
.2
.0
NOTE: PUB - Main pickup broom.- Polypropelene utilized in Tests 1-4 , strike
measured 6-1/2 in. Tymco sweepers not equipped with pickup broom.
119
-------
100
z
LLJ
5
o
sweeping has been developed in this way. Since the strip test provides
nearly ideal operating conditions, it is not surprising that results
from such tests result in removal effectiveness of greater than 90 per-
cent, as shown in Fig. 40. Controlled tests prove somewhat less effec-
tive, and in situ street tests fall even lower in measured effectiveness.
However, street tests represent real-world conditions.
The sources of data for
-^ these various tests (Refs.
17? 24 through 33) include
o >- _S
5 J u the present study, some
_ ^ " _j work undertaken by the
APWA, evaluations by
manufacturers of street
cleaning equipment , and
finally the series con-
ducted by the Naval Radio-
logical Defense Laboratory
(NRDL).
Now let us turn to the
consideration of param-
50 4 ~ eters which have been
identified as most impor-
tant to street sweeping
effectiveness. Table 31
lists fixed, controllable
and untested parameters.
"Fixed" parameters refer
to those with which the
Public Works Engineer
is "stuck." For the pur-
poses of assessing removal
effectiveness of the dirt
and dust fractions, we
will consider the unit
Fig. 40. Comparison of Results from Sweep- mass level (expressed in
grams/sq ft), the parti-
cle size, and the uniform-
ity of material across
the street. Initial load-
ing, the first of the
fixed parameters, is dependent upon the land-use category, the season of
the year, and frequency of cleaning. The NRDL studies and the results
obtained in the controlled tests have shown that removal effectiveness
increases with increasing mass levels.
Q
a:
z
Street
Tests
Controlled
Tests
Strip
Tests
Comparison of Results from Sweep-
ing Effectiveness Tests Conducted
Under Various Conditions: For
Dirt/Dust Fraction
120
-------
Table 31
PARAMETERS WHICH AFFECT STREET SWEEPING PERFORMANCE
Fixed Loading : Mass level
Particle size
Uniformity
Surface: Type
Condition
Sweeper: Type
Controllable Sweeper
Operation: PUB type
PUB rpm
PUB diameter
PUB strike
Forward speed
Number of passes
Gutter broom
Debris deflector
Untested Operator skill
NOTE: Other fixed parameters which have been discussed elsewhere
include: land-use category, frequency of cleaning, and
seasonal variations.
PUB - Main Pickup Broom
The present URS study provides some pertinent data on the effect of
particle size on street sweeping effectiveness. Our results indicate
an overall removal effectiveness of 50 percent (that is, half the dirt
and dust fraction was picked up by the sweeper and half remained on the
street). However, when considering the removal effectiveness in terms of
particle size, the results are far different. Our analysis, done by
sieving the solid sample through standard Tyler sieves, indicated that
initially the middle-size fractions (i.e., the 104-840 micron fraction)
were predominant, as shown in Fig. 41. (For comparative purposes, the
less than 43 micron material has a consistency of flour, the 104 to
246 micron grouping is equivalent to fine sand, the 246 to 840 micron
fraction is equivalent to a coarse sand, while the greater than 2000
micron grouping consists of small pebbles, shards of glass, cigarette
butts, etc. (Large gravel and rocks were not included in the sample and
when present were rejected.) However, after sweeping, this proportional
distribution was found to change, with the smaller fractions (i.e.,the 43
104 micron range) showing an increase (see Table 27) while other size
121
-------
20
X
_Q
10
O
Z
O
o
o
O
O
(N
S*o o
XT "*t
C-il 00
I I I
n $ o
-* 2
PARTICLE SIZE
(microns)
Fig. 41. Particle Size Distribu-
tion Initially: For a
Composite Sample
ranges decreased somewhat. The
results indicate that removal
effectiveness is actually greater
than 70 percent for the larger
fractions (greater than 246 microns)
dropping somewhat for the middle-
size fraction and decreasing to an
insignificant amount for the small-
est fractions. This finding,
which was corroborated by the NRDL
studies, has serious implications;
namely, that the smallest fractions
are the most poorly removed by con-
ventional street sweeping procedures.
This is particularly significant
since the principal pollutant materi-
als have been found in highest con-
centrations in association with
the fine fractions.
Table 32 summarizes the removal
efficiency of the dust and dirt
fractions by particle size range as
determined by the iji situ tests and
as determined through utilization
of the NRDL equation:
* * -kE
M=M +(MQ-M)e
The following assumptions were
utilized in calculating the
removal efficiencies:
1.
2.
3.
Paved surface - asphaltic concrete, fair condition
Equipment - conventional motorized street sweeper with
three or four wheels, utilizing a polypropylene main
pickup broom and steel bristle gutter brooms
Effort - Based on average operating speed of 6 mph and
an 8 ft wide swath:
(8 ft) x
'6 miles\
1 hr
5280 ft
hr / I 60 min/ " V mile
or E = 0.237 equipment min/1,000 sq ft
\
/
= 4224 sq ft/min
122
-------
Table 32
ESTIMATED STREET SWEEPER EFFICIENCY
PARTICLE SIZE
(u)
> 2,000
840 - 2,000
246 - 840
104 - 246
43 - 104
< 43
NOTE: In-situ tests
REMOVAL
IN SITU TEST
78.8
66.4
69.5
47.7
< 0
< 0
are average removal
EFFICIENCY, (%)
EQUATION
-
-
49.2
48.7
22.2
15.8
efficiencies from
COMPOSITE
(estimate)
79
66
60
48
20
15
test resu]
"given in Table 25. The equation utilizes the relationship:
M = M* + (M - M*) e-KE
o
6.
Proportionality constant, k, is dependent upon sweeper
characteristics. For conventional sweepers, k = 0.330
(see Ref. 27)
M is the irreducible mass which would theoretically
remain after an infinite amount of sweeping (dependent
upon sweeper type, surface and particle size, of con-
taminant). For particles of 43 to 10/1^ , M = .75 g/sq ft
and for particles of 104 to 840^i , M = . 14 g/sq ft (estimated
from data in Refs. 26 through 29)
MO = the initial mass loading before sweeping in g/sq ft
Average loading from selected cities = 7.26 g/sq ft
Note that 1 g/sq ft = 2.20 lb/1000 sq ft
From Table 5, average particle size distribution of
solids in selected city composites:
123
-------
Particle
Size Range
>
840
246
104
43
<
2000
- 2000
- 840
- 246
- 104
43
Distribution
20
6
20
19
18
15
.1
.8
.2
.8
.0
.1
Loading
(g/sq ft)
1
1
1
1
1
.46
.49
.46
.44
.31
.06
Table 33 summarizes the calculated removal effectiveness values for
particle size ranges for which equation constants were available.
Table 33
SUMMARY OF CALCULATED REMOVAL EFFECTIVENESS VALUES
* * _kE
Based on: M = M + (M - M )e
o
PARTICLE SIZE (1) (2)
M M*
w
246-840 1.46 0.14
104-246 1.44 '- 0.14
43-104 1.31 0.75
< 43 1.06 0.75
M - M
-icncj M
0
(3) (4)
(Mo-M*) e-KE
1.32 0.457
1.30 0.457
0.56 0.457
0.31 0.457
100%
(5)
(3)x(4)
0.604
0.594
0.266
0.142
EFFEC-
M TIVENESS
0.744 49.2
0.734 48.7
1.016 22.2
0.892 15.8
The URS in situ tests also show the nonuniform distribution of debris
across streets. Figure 42 shows the variations in mass loadings
(g/sq ft) found across the street both before and after sweeping. As
might be predicted, the gutter was found to be the most heavily loaded
zone on unswept streets. Loading then drops rapidly moving out from
the curb. However, after sweeping, the gutter is much cleaner; this is
not so for some of the other areas. In short, it appears that the
124
-------
sweeping operation has moved much of the material out of the gutter but
has tended to redistribute it on areas which were somewhat cleaner prior
to sweeping. The function of gutter brooms on conventional motorized
sweepers is to move material out of the gutter into the path of the main
pickup broom. Dirt deflectors are utilized to assist in directing the
material for pickup. Since the distribution of material on streets is
such that a large portion (70-80 percent) of the material is located within
6 in. of the curb, a device such as a gutter broom is required for
efficient street debris and litter removal. However, the present design
of gutter brooms is such that they tend to redistribute the dust and
dirt fraction (< 2000 fj.) over the surface of the street, and indeed are
not particularly efficient in moving the dust and dirt fraction out of the
gutter.
14 JM. Initial loading , /f 2)
18 Final loading W '
Fig. 42. Initial and Final Loading Across Swept Streets: Composite
Sample
125
-------
Consider now surface type as a "fixed" parameter. Only the NRDL and
URS studies have attempted to differentiate between asphaltic concrete
and Portland cement surfaces. Relatively small differences were found.
However, surface condition, which was an important factor in the present
tests, does affect removal effectiveness. We have been unable to quanti-
tatively describe surface conditions although it is obvious that sur-
faces in poor condition, with large cracks, depressions, etc., are more
difficult to sweep and removal effectiveness suffers. Likewise, it is
difficult to assess the effect of curb type, street slope and street
contour. Further studies should probably be undertaken so that this
parameter can be considered in the design of new streets.
The final "fixed" parameter is sweeper type. A variety of tests con-
ducted on conventional 3-wheel and 4-wheel sweepers does not indicate
great differences between various models or types. Vacuum and air-
sweepers have been evaluated only on controlled tests and strip tests.
Sweepers of this type were not available in the test cities for in situ
tests.
Moving now to "controllable" parameters, we find a variety of machine
characteristics which can be varied by the operator to change the
cleaning characteristics of the machine. Historically, considerable
study has been done on the parameters listed in Table 30. Cited litera-
ture indicates that best cleaning performance is obtained with a steel
pickup broom, followed closely by. the natural fiber brooms and finally
by plastic pickup brooms. However, despite the apparent superiority of
the steel-type and natural fiber-type pickup broom in removal effective-
ness , the trend is definitely toward the adoption of plastic-type brooms,
since they do perform satisfactorily (at least on litter and larger
dust and dirt fractions).
The effect of broom diameter (that is, wear) has been studied by several
investigators, but results are ambiguous. One claim is that a new broom
sweeps cleanest whereas another claim is that shorter fibers are better.
In either case, the evidence suggests that the differences in performance
are rather small (perhaps the point is academic since brooms are generally
used until worn out anyway).
Pickup broom strike has been found by all investigators to be an
important consideration in cleaning efficiency. The greater the strike,
the better the cleaning efficiency. However, increasing the strike also
increases the wear on the broom.
The rotational speed of the pickup broom has not been demonstrated to
be a highly critical aspect of cleaning performance. However, for sweepers
with auxiliary engines, increasing the rpm of the pickup broom does seem
to improve pickup efficiency somewhat. Forward speed of the sweeper,
especially when it is geared directly to the rpm of the main broom, does
seem to have an important effect on cleaning performance. As previously
noted, the slower the machine moves, the better the cleaning performance
126
-------
appears to be. However, for machines in which the pickup broom rotational
speed is geared to forward speed, it appears possible that effectiveness
may actually drop since at the slower speed the main broom speed also
drops.
A final .variable which has not been assessed by any researchers is that
of operator performance and competence. However, the URS test team did
conclude from a subjective evaluation of operations in various cities
that the skill and competency of the operator is a very crucial factor.
While it is difficult to conceive of a test procedure that would identify
and quantify the skills required of a good operator, perhaps at some
point in the future such an index may be developed.
Turning again now to the equation developed by NRDL to explain removal
effectiveness, we see in Fig. 43 how such a curve would look for the
conditions that are normally encountered in street sweeping. The first
pass removed approximately 50 percent of material on the surface, the
second pass removed about 50 percent of that remaining for a total
removal effectiveness of 75 percent. Subsequent passes remove effectively
less of the remaining mass , although the overall effectiveness does
approach 100 percent with increasing number of passes, so we see that
one procedure for increasing removal effectiveness is to sweep the same
area two or more times.
100
75
50
I/I
LJJ
LU
u
£ 25
Based on:
M= M*+(M - M*)e'
o
M 10.0 a/ft2
M* = 0.5 g/ft2
I
-kP
0)23'
NUMBER OF PASSES (P)
Fig. 43 Removal Effectiveness with
Number of Passes
A comparison of the
effort required to achieve
lower residual mass levels
and increasing removal
effectiveness , as compared
to the effort expended
during typical street
cleaning operations, can
be illustrated by utiliz-
ing the NRDL equation
and the sweeper parame-
ters assumed in deriving
the calculations given
in Table 33.
Table 34 compares the
effort required, as de-
termined from the NRDL
equation, t;o achieve
different degrees of
effectiveness at several
initial mass levels for
a motorized sweeper on
asphaltic concrete. It
can readily be seen that
127
-------
the amount of effort required to obtain greater than 90 percent effective-
ness is several times the effort normally expended in sweeping operations.
Table 34
EFFORT REQUIRED TO ACHIEVE RESIDUAL MASS LEVELS
M^M EFF E INCREASE OVER
(g/sq ft) (g/sq ft) % (equip min/1,000 sq ft) NORMAL (.237)
20 1.0
2.0
5.0
50 2.0
5.0
120 2.0
5.0
95
90
75
96
90
98
96
1.50
.85
.50
1.35
.80
2.00
1.10
6.3
3.6
2.1
I
5.7
3.4
8.4
4.6
CATCH BASIN EFFECTIVENESS
Although most of this study is focused on street surface contaminants
and various means for removing them from street surfaces, a special
substudy was directed toward catch basins. The primary goal here was
to develop an understanding of how catch basins affect the quality of the
runoff water which passes through them. It should be noted that the
study was quite limited in scope and should not be viewed as being com-
prehensive in breadth or depth. Nonetheless, some of the information
developed should prove valuable in understanding the pros and cons of
catch basins as they relate to the pollutional aspects of street runoff
discharged to receiving waters.
To summarize our conclusions based on field testing and laboratory
analysis, we found that catch basins can be reasonably effective in pro-
tecting sewers from loadings of coarse granular material but have a defi-
nite potential for contributing to water pollution problems. The
following paragraphs describe the catch basins we studied, our test
procedures, and the rationale for these conclusions.
128
-------
In the way of background, it is of interest to note that catch basins
became quite popular appurtenances in sewerage systems during the years
before sound, well-engineered, paved streets were common and before
mechanical trenching made it practical to lay sewers on reasonably steep
grades. Catch basins were included in sewer design as a means of pre-
venting sewers from becoming clogged by the rocks and gravel-like material
which commonly entered the sewer system. Another function was to provide
a waterseal to control the escape of sewer odors and gases. In recent
years, public works and design engineers have questioned the merit of
routinely including catch basins in modern systems.
The subject of catch basins is included here not so much in the sense
of what their eventual fate should be, but rather to determine what effect
those currently in use might have on receiving water quality. Note that
the term "catch basin is rather nonspecific; devices of a very broad
variety of sizes and shapes are in general use. Their primary common
feature is that they act as miniature settling basins for removing dense
solids via simple sedimentation. The catch basins which were studied here
are located in a residential area of San Francisco. They are true catch
basins in that they were specifically designed and installed for this
purpose (many are not). They are of a standard design, minor variations
of which are found in urban and suburban areas throughout the country.
The experimental program can best be understood after considering that
two phenomena occur simultaneously in a typical catch basin during periods
of storm runoff:
Dissolved and particulate solids initially contained in
the catch basin (as a result of prior deposit) are stirred
up and swept out by the water flowing through
Particulate solids carried in the influent settle out and
are retained in the catch basin.
Clearly, the relative balance between these two phenomena determines
whether the unit is a benefit or a detriment. The fact that these
act simultaneously accounts for why so little quantitative information
on performance is available. Our study approach involved separating the
two phenomena so they could be examined independently. Two types of
tests were employed:
Clean water was run under controlled conditions into
several previously "dirty" catch basins (ones which had
not been cleaned for several months). The discharge
from these was sampled and analyzed. Several flow rates
were used to cbver the conditions of light, moderate,
and heavy storm intensities
129
-------
Dirty water was run under controlled conditions into a
"clean" catch basin and the discharge sampled and analyzed
to establish the unit's removal efficiency under different
flow rates. The "dirty" water was made up by carefully
introducing solids which had previously been collected
from street surfaces. (This "dirt" was mixed with water
from a fire hydrant supply with a time-varying concentra-
tion to simulate the variation in storm water solids con-
tent with time from the onset of a storm.)
During the course of the tests, numerous water samples were collected;
most of them shortly after the onset of the simulated storm to reflect
the "shock" loading effect. Subsequent analysis was directed primarily
toward determining the amounts and size distribution of solids.
Test results on the initially clean catch basins indicate that they are
reasonably effective treatment units for removing heavy solids from
storm runoff. The curves of Fig. 44 show that virtually all of the solids
larger than 246^x, diameter were removed. On the other hand, only a
small portion of the fine solids was removed. These curves were for a
test wherein a heavy rainfall intensity (1/2 in./hr) was applied over a
rather sizable catchment area (25,000 sq ft). (A word of caution is in
order here regarding the subsequent use of reported values. It must be
recognized that these catch basins are of a standard design and are
used routinely in a broad variety of situations. This means that virtu-
ally no consideration is given toward sizing them to be appropriate to
the expected flow [i.e., the same unit is used to receive runoff from
both large catchment areas and very small areas]. The net result of
this practice is that retention times vary tremendously from basin to
basin; likewise, turbulence levels vary and removal efficiencies vary.)
Other tests indicate that higher flow rates through the same basin result
in lower removal efficiencies, and lower rates give higher efficiencies
(as would be expected). Catch basins function as very simple sedimenta-
tion units and are, therefore, limited by the same factors that limit
any sedimentation process: turbulence and retention time. The catch
basins tested have no turbulence-controlling baffles at either their
inlets or outlets and operate under complete mixing during all but the
lowest flows. The retention time is extremely short, less than a minute
even for rather low flows. These facts explain why catch basins are
effective only in removing coarse materials. Since other phases of
this research project have identified the fine particle size ranges as
being most relevant to receiving water pollution, we conclude that even
the best conventional catch basins are ineffective in reducing pollution.
The curves of Fig. 44 show that removal efficiencies varied during
the test, generally decreasing with respect to time. Presumably this is
due to unstable conditions of hydraulic turbulence and resuspension
(although further tests would be required to support this speculation).
130
-------
i
z
a
UJ
I
40
20 -
The other issue examined
in this substudy has to
do with catch basin po-
tential for adding pol-
lutants to the flow
passing through them.
This was studied by
running clean water at
prescribed flow rates
through several catch
basins which had not
been cleaned for sever-
al months. The discharge
was sampled repeatedly
over a period of about
an hour (with most of
the samples taken short-
ly after the "storm11 on-
set). The catch basins
all had several thousand
pounds of solids in them,
with a layer of water
(and floating debris) up
to the outlet level.
The water first dis-
charged was very dirty;
composed primarily of
this supernatant water,
some of the floating
matter, plus particulate
matter suspended by the
turbulence flow. Within a few minutes, the water became reasonably clear
but still contained particulates. Even after nearly an hour's flushing,
the discharge contained much particulate matter. At the end of an hour,
inflow was stopped and the volume of basin contents was measured.. It was
found that only about 1 percent of the initial solids in the basin was
removed by the flushing action of any of the simulated storms (light,
moderate, or heavy). On the other hand, the material which was flushed
out as the initial slug would have a substantial pollutional impact on
the receiving waters. This is borne out by analyses of catch basin
contents.
The City of San Francisco Public Works staff provided URS with data devel-
oped as part of their studies of combined storm/sanitary sewers. URS
sampled and analyzed the content of several catch basins in Baltimore
and Milwaukee. Pertinent data on these catch basin contents are reported
in Tables 35 and 36. While these data reflect conditions during winter
and spring months, the "catch basin operation" can be considered essen-
tially uniform during all seasons of the year. In terms of operational
mode, the catch basin, acts as a short-term sedimentation basin and its
TIME SINCE FLUSHING BEGAN (min)
Fig. 44. Removal of Street Surface Contaminant
Solids - Variation with Particle Size
131
-------
efficiency, measured in terms of solids removal and retention, is generally
constant. A recent study (Ref. 34) indicates that the sedimentation
process does show improved efficiency when operated at elevated tempera-
ture, but, in the case of short term detention systems, such as catch
basins, this effect must be considered negligible. However, recognizing
that pollutant loads (in terms of specific constituents) do vary seasonally,
it would be expected that during summer months the pollutant load on
catch basins and the resultant effluents from them will be higher in
nitrates and phosphates due to increased use of fertilizers. It should
be stressed that this change in pollutant character and quantity is not
a function of catch basin efficiency but rather a function of increased
pollutant load to the environment.
Table 35
SUMMARY OF DATA ON CATCH BASIN CONTENT ANALYSIS
(from City of San Francisco)
CATCH BASIN
LOCATION
Plymouth and
Sadowa
7th and Hooper
Yosemite
40th and Moraga
Mason and
O'Farrell
32nd and Taraval
Haight and
Ashbury
Marina Area
Montgomery Street
Webster and Turk
Lower Selby
Upper Mission
FIRST SAMPLING SERIES
COD
(mgA)
3,860
15,000
739
9,060
8,100
153
37,700
701
6,440
1,440
288
5,590
BOD
(mg/-t)
190
430
11
40
130
5
1,500
100
390
44
6
50
TOTAL N
(mg/l)
10.9
33.2
1.8
16.1
29.7
0.5
1.4
7.0
18.8
14.0
1.4
12.0
TOTAL P
(mgA)
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
< 0.2
0.3
SECOND SAMPLING SERIES
COD BOD TOTAL N TOTAL P
(mgA) (mg/-t) (mg/-L) (mg/-£-)
8,610 122 2.8 0.3
2,570 170 2.0 < 0.2
21,400 120 4.6 < 0.2
51,000 130 12.0 < 0.2
7,720 85 16.5 < 0.2
708 15 1.4 < 0.2
143,000 420 14.6 < 0.2
8,600 40 < 0.5 < 0.2
8,160 300 3.9 < 0.2
NOTE: Both sampling series were conducted in winter 1970. All values based on analysis of
total basin contents after complete mixing.
132
-------
Table 36
ANALYSIS OF CATCH BASIN CONTENTS
TEST SITE
CODE
SOLID SAMPLES (Sediments)
LIQUID SAMPLES (Supernatant)
COD PHOSPHATES NITRATES COD PHOSPHATES NITRATES
(mg/-t) (mg/-t) (mg/t) (mg/g) (mg/g) (mg/g)
BA-6
BA-8
BA-2
150
175
1.10
2.2
Baltimore
4.0
5.5
31.0
12.0
0.60
0.17
0.50
0.90
Mi-5
Mi-8
8,250
1.5
MiIwaukee
9.0
7,750
11.75
3.0
0.09
16.0
0.70
NOTE: See Appendix B for site code key, giving cities and land-use categories. Both
sampling series were conducted in April/May 1971.
The successful operation of a catch basin, as a sedimentation process, is
a function of the solids retention capacity of the system. Basins which
are frequently cleaned have the capacity for operating at design efficiency
and retaining solids (with associated pollutants); however, effluents
from dirty catch basins (most basins in urban and suburban areas are
cleaned less than once per year and are categorized as "dirty") exert a
significant pollutional load on receiving waters and/or waste treatment
plants. A portion of the solids found in catch basins is not deposited
there by runoff. Rather, catch basins may act as convenient receptacles
for litter, leaves and garden cuttings, crankcase drainings, etc.
133
-------
Section VI
SIGNIFICANCE OF STREET SURFACE
RUNOFF AS A SOURCE OF
WATER POLLUTION
-------
Section VI
SIGNIFICANCE OF STREET SURFACE RUNOFF
AS A SOURCE OF WATER POLLUTION
The intent of this sectdon is to place the information obtained in this
study in perspective and to help answer the question of how important street
surface contaminants are, relative to other common sources of water pollu-
tion. To accomplish this, we have compared a city's street runoff with
both its treated sanitary sewage discharge and its storm water discharge.
In the interest of simplicity we have made these comparisons using a hypo-
thetical city, rather than a real one. This city's street surface contam-
inants have the properties determined in this study (means of the values
actually observed have been employed in the comparisons). The hypothetical
city has the following characteristics :
Population - 100,000 persons
Total land area - 14,000 acres
Land-use distribution:
residential - 75%
commercial - 5%
industrial - 20%
Streets (tributary to receiving waters) - 400 curb miles
Sanitary sewage - 12 x 106 gal/day.
The comparisons made here are for the first hour of a moderate-to-heavy
rainstorm; one which involves brief peak rates of at least 1/2 in./hr
during that first hour.
Table 37 compares the pollutants in street runoff (generated by that 1-hr
storm) with the pollutants which the city's municipal sewage treatment
plant would contribute during a typical hour's discharge (the plant is
assumed to be a well-operated secondary facility). Obviously, the street
runoff is a much greater source of short-term "slug" loadings. It should
be noted that, if this comparison were to be recomputed using the total
pollutant loading over an annual cycle, the street runoff would likely be
less than the treated sewage load. We have not made such comparisons
because of the difficulty of establishing meaningful weather patterns for
a hypothetical city.
135
-------
Table 37
COMPARISON OF POLLUTIONAL LOADS
FROM HYPOTHETICAL CITY -
STREET RUNOFF vs GOOD SECONDARY EFFLUENT
Settleable + ...
(d 1
Suspended Solids
BOD(C°
COD
-------
Table 38
COMPARISON OF POLLUTIONAL
LOADS FROM HYPOTHETICAL CITY -
Street Runoff vs Raw Sanitary Sewage
Settleable +
Suspended Solids
BOD
-------
chemical oxygen demand, Kjefdahl nitrogen, soluble nitrates and phosphates.
Also included are less common parameters; e.g., heavy metals (including
chromium, copper, zinc, nickel, mercury, lead and cadmium) and both
chlorinated hydrocarbon and organic phosphate pesticide compounds. However
only the following were found: dieldrin, DDD, DDT, methoxychlor, endrin,
methylparathion, and lindane. Polychlorinated biphenyls (PCB's) were also
sought and found in significant quantities in all cities studied. Addition-
ally, studies were conducted concerning the presence of both total and
fecal coliform bacteria on the streets.
Table 39
COMPARISON OF STREET SURFACE CONTAMINANTS
WITH STORM SEWER DISCHARGES
SOURCES OF POLLUTANTS
Storm Sewer
Samples
(reported by
others )
Street Surface
Contaminants
(collected in
this study)
Street Surface
Contaminants
(reported by
others )
East Bay
Sanit. Dist. (c)
Cincinnati (c)
Tulsa (d)
Bucyrus (e)
(a combined
system storm/
sanitary)
Tulsa
Bucyrus
Average over
ten cities
Chicago (c)
KuELDAHL TOTAL COLIFORM
BOD COD NITROGEN PHOSPHATES BACTERIA
(% by wt)(a) (% by wt) (% by wt) (% by wt) (10 org/lb)1
6.2 4
4.6 42.3
2.2 16 .16 .21 73
11 40 1 .93 4,400
4.2 8.8 .2 .17 200
.21 2.1 .087 .018
1.7 8.4 .18 .092 104
.5 4.0 . 048
(a) Concentrations of pollutant as a percent of total solids (dry weigh base).
(b) Concentrations of viable organisms associated with total solids (dry weight basis).
(c) Ref. 1.
(d) Ref. 2.
(e) Ref. 35.
Tables 40 through 43 summarize the quantity of pollutants identified in
the investigation in terms of the parameters cited above. Reviewing the
information in Table 40 reveals the rather extensive potential problem
posed by these pollutants. As an example, the weighted average of 5-day
BOD is shown in Table 40 to be 13.5 pounds per curb mile. In a typical
(hypothetical) community of 100,000 population having 400 curb miles of
streets, this would represent a potential load on the receiving water of
5,400 pounds of BOD estimated from an average 2- to 10-day buildup period
since last sweeping or rain. The same city with a well-run municipal
138
-------
sewage treatment plant would have a daily BOD discharge of only about 1/30
this figure. While this is, of course, a hypothetical calculation and does
not reflect the time span over which the street contaminant would be dis-
charged to the receiving water, it nevertheless reflects serious concern
over the non-point pollutant described.
Table 40
POLLUTION LOADS BY SELECTED COMMUNITIES
(Ib/curb mi)
CITY
San Jose I
San Jose II
Phoenix I
Phoenix II
Milwaukee
Baltimore
Seattle
Atlanta
Tulsa
Bucyrus
Weighted
Average
CURB
MILES
2,300
2,300
2,900
2,900
3,400
3,900
2,600
3,500
3,600
200
27,600
TOTAL
SOLIDS
910
6,000
650
910
2,700
1,000
460
430
330
1,400
1,400
VOLATILE
SOLIDS
66
460
40
92
180
96
29
18
19
150
100
BOD
5
16
53
7
10
12
-
5
2
14
3
13.
COD
(310)
(400)
30
54
48
17
13
30
29
.5 95
32a
KJELDAHL
NITROGEN
2.
11,
1.
2,
1.
1.
0,
0.
0.
1.
2.
,1
,0
.5
,9
,4
,9
,9
.5
,7
.2
.2
SOLUBLE
NI TRATE
0.
0.
0,
0,
0.
0.
0.
0.
0.
0.
.27
.29
.12
,052
.038
.027
.024
.012
.12
,094
PHOSPHATES
0.
4,
0
2,
0.
1.
0.
0,
0,
0
1.
.70
.5
.22
.8
.27
.0
.49
.26
.54
.25
.1
Excluding San Jose I and II
139
-------
Table 41
HEAVY METALS LOADS BY SELECTED COMMUNITIES
(Ib/curb mile)
CITY
San Jose I
San Jose II
Phoenix I
Phoenix II
Milwaukee
Baltimore
o Seattle
Atlanta
Tulsa
Bucyrus
Weighted
Average
CHROMIUM
0.
0.
-
0.
0.
0.
0.
0.
0.
~
0.
20
14
029
047
45
081
Oil
0033
11
COPPER
0.
0.
-
0.
0.
0.
0.
0.
0.
_
0.
50
02
058
59
33
075
066
032
20
ZINC
1.4
0.28
-
0.36
2.1
1.3
0.37
0.11
0.062
"~
0.65
NICKEL
0.13
0.085
-
0.038
0.032
0.077
0.028
0.021
0.011
0.05
MERCURY
0
0
-
0
(0
(0
0
0
0
"
0
.30
.085
.022
.082)8
.082)3
.034
.023
.019
.073
LEAD
1.9
0.90
-
0.12
1.5
0.47
0.50
0.077
0.030
~
0.57
CADMIUM
0
(0
-
(0
0
0
(0
(0
(0
.0033
.0031)
.0031)
.0032
.0026
.0031)
.0031)
.0031)
TOTAL
HEAVY METALS
4.5
1.5 j
-
0.63 "
4.5 ^
2.8 g
1.09
0.31
0.16
~
1.6
Note: Except for San Jose I, Phoenix I, Milwaukee, Baltimore, and Bucyrus, cadmium
estimates were based on other observations.
-------
Table 42
PESTICIDE LOADS BY SELECTED COMMUNITIES
(Ib/curb mile)
CITY
San Jose I
San Jose II
Phoenix I
Phoenix II
Milwaukee
Baltimore
Seattle
Atlanta
Tulsa
Bucyrus
Median
DIELDRIN
11
27
24
10
3
27
24
24
17
24
PCB
1,200
1,100
65
3,400
3,400
1,100
65
65
650
1,100
BP-DDD
67
120
34
0.5
100
120
34
34
83
67
METH-
OXYCHLOR
0
0
0
8,500
170
0
0
0
1,610
P ,P-DDT
110
170
13
1.0
30
170
13
13
61
61
ENDRIN
2.0
0
0
0
0
0
0
0
METHYL
PARATHION LINDANE
20 17
0
0
0 0
0
0
0
0
TOTAL
PESTICIDES
1,460
1,417
136
11 ,910
3,700
1,420
136
136
2,410
1,420
NOTE: All values by 10
Table 43
TOTAL AND FECAL COLIFORM LOADING DISTRIBUTION BY LAND-USE CATEGORY
TOTAL COLIFORMS FECAL COLIFORMS
(109 org/ (number/ (10^ org/ number/
AREA curb mile) gram solids) curb mile) gram solids)
Residential 60 160,000 5.8
Industrial 150 82,000 1.6
Commercial 120 110,000 18.
Combined
(Total) 99 130,000 5.6
16,000
4,000
5 ,900
14,000
Note: The per curb mile ratio is not equal to the per gram solids ratio
because extreme values in each matrix were eliminated prior to
determining the.weighted averages for e'ach land-use area. The
eliminated values from each matrix were not always representing
the same city and land use, hence the discrepancy.
141
-------
Based on information available concerning antecedent street cleaning and
rainfall patterns, an attempt was made to calculate accumulation rates in
terms of pounds of pollutant per curb mile per day. These summary^figures
are shown in Table 44. It should be appreciated that these "daily" values
are somewhat artificial in that there was no way to account for either the
effectiveness of street cleaning operations or the extent of the rainfall
and associated pollutant runoff prior to the street sampling procedures.
Nevertheless, it is apparent from Table 44 that the relative pollutant load
is significant, although the relationship between specific pollutants does
change. In the case of BOD, the mean value of 4.5 Ib/curb mi/day (or
1800 Ib for 400 curb miles for the hypothetical city) is roughly equivalent
to the amount of BOD discharged to the receiving waters daily by the sewage
treatment plant.
Table 44
AVERAGE RATE OF ACCUMULATION OF POLLUTANTS
(Ib/curb mi/day)
TOTAL VOLATILE KJEDAHL
CITY SOLIDS SOLIDS BODg COD NITROGEN NITRATES
San Jose I 70 5.3 1.2 24 0.16
San Jose II 860 65 7.6 56 1.5 0.038
Phoenix I 92 5.9 0.93 4.3 0.21 0.041
Phoenix II
Milwaukee 2,700 180 12 48 1.4 0.052
Baltimore 260 24 0.48 0.0095
Seattle
Atlanta 220 9.3 0.95 6.5 0.24 0.012
Tulsa
Bucyrus 690 74 1.4 25 0.60 0.060
Ivefagf 73° 51 4"5 26 °'66 °'029
TOTAL
HEAVY TOTAL
PHOSPHATE METALS PESTICIDES
0.054 0.34 110 v 10~6
0.64 0.22 200
0.031
0.27 4.3 12,000
0.25 0.68 940
0 . 13 0 . 21 68
-
0.12
0.37 1.3 200
Note:
1. No sweeping data available for Phoenix II, Seattle, and Tulsa.
2. Based on number of days since cleaned, either by rain or by sweeping,
whichever occurred closest to the test date.
3. Heavy metals include chromium, copper, zinc, nickel, mercury, lead, and
cadmium.
4. Pesticides include dieldrin, PCB, DDD, methoxychlor, DDT, endrin, methyl
parathion, and lindane.
In summary, all other identified characteristics exhibit similar relation-
ships in terms of wastes emanating from domestic treatment plants.
142
-------
The 'ihforirt'ation in Table 41 deals specifically with heavy metals. The
metals included here were found in sufficient concentration to be detect-
able in all cities of the study. This of itself is not a surprise due
to the sensitivity of the analytical procedures. However, the fact that
the weighted average of the total heavy metals is as high as 1.6 pounds
per curb mile is rather alarming.
Considering that the subject of individual heavy metal's effects on the
environment is only slightly understood at best, and considering further
that so little conclusive work has been done regarding the synergistic
effects of combinations of metals, there is every reason for concern over
the high quantities found in this investigation.
Information concerning pesticides is presented in Table 42. The concern
over pesticides has reached such proportion that many states have banned
their use and sale and the Federal government has taken an active role in
prohibiting extensive use of DDT and other long-lived synthetic pesticides.
The information in Table 42 is significant on two counts. First, the fact
that the median value for total pesticides found in the study is high
enough to be reported in terms of pounds (albeit a rather small number -
0.0014 Ib/curb mile). For the hypothetical community of 100,000 population,
the calculated pesticide loading was in excess of 1/2 Ib per precipitation
incident, which represents approximately 0.1 Ib/day. Secondly, about three-
fourths of the total weight of such materials were found to be polychlori-
nated biphenyls, a class of compounds over which there has been much recent
concern due to the high incidence of wildfowl deaths correlatable with high
PCB levels. It is premature at this point to speculate on the implications
of how PCB are being introduced to the environment or why there is such
a high incidence of PCB identified. However, recent studies have shown
PCB concentrations as high as 0.2 ppm in soft-shelled clams taken from
Chesapeake Bay and as high as 1.5 ppm in Atlantic Ocean zooplankton.
Based on an estimated 42,500 tons of PCB commercially produced in the U.S.
in 1970 and assuming a closed system use of approximately 60 percent, the
accumulation on urban and suburban streets, as reported in this study,
represents 0.15 percent of the total PCB production.
Table 43 summarizes the total and fecal coliform distribution found in
the study. For purposes of summary, values are given for the three major
land-use classifications within a community (i.e., residential, industrial,
and commercial), as well as a combined figure representative of the average
municipality. Two observations can be made here. The first concerns the
overall magnitudemore than 100 billion total coliforms and over 1 billion
fecal coliforms per curb mile. The second concerns the relative magnitude
of total to fecal counts.
The number of fecal coliforms found in the street samples is about a
thousand-fold less than densities commonly associated with the discharges
from domestic animals. The figures in Table 43 indicate that fecal coliforms
range from 4,000 to 16,000 per gram of solids. Comonly accepted figures
for animals are in the range of 8 to 23 million fecal coliforms per gram
of feces.
143
-------
Data reported in Table 43 indicates that the highest ratio occurs
in the industrial areas, the lowest in central business districts (which
are typically swept daily). On a comparative basis, reported ratios of
total to fecal coliform in raw sewage have been found to range from 2.1
to 12, while ratios for storm runoff (uninfluenced by domestic sewage
discharges) range from 10 to 300. Studies concerning the relationship
between total and fecal coliforms as a function of time indicate that the
fecal coliforms exhibit a more rapid die-off (they are much more sensitive).
Therefore the greater the ratio of total to fecal, the greater the time
interval since deposition. In other words, the high ratios exhibited in
industrial areas may well be attributed to a longer residence time on the
streets. In any case, all of the ratios seem to indicate that the bacteria
have resided in the test sites for some time and are probably not indicative
of fresh bacterial discharge.
However, the total impact of the non-point pollutants must be assessed
in terms of the product of the pollutant load per land-use category and the
actual amount of land area represented by the designated land-use category.
Thus , although Table 45 and 46 show the industrial category to have the
highest loading of pollutants per curb mile, there may be no problems in a
small community with minimal industry.
Table 45
DISTRIBUTION OF CONTAMINANT LOAD BY LAND-USE CATEGORY
(Ib/curb mile)
RESIDENTIAL INDUSTRIAL COMMERCIAL
Total Solids
Volatile Solids
BOD,.
3
COD
Kjeldahl Nitrogen
Nitrates
Phosphates
Total Heavy Metals
Total Pesticides
1,200
86
11
25
2.0
0.06
1.1
0.58
2,800
150
21
100
3.9
0.18
3.4
0.76
360
28
3
7
0
0
0
0
_.
.4
.18
.3
.18
144
-------
Table 46
DISTRIBUTION OF CONTAMINANT LOAD BY LAND-USE CATEGORY
(Ib/curb mile/day)
Total Solids
Volatile Solids
BOD
o
COD
Kjeldahl Nitrogen
Nitrates
Phosphates
Total Heavy Metals
Note: Based on number
RESIDENTIAL
590
44
3.6
20
0.60
0.019
0.37
1.2
of days since
INDUSTRIAL COMMERCIAL
1,400
77
7
81
1
0
1
1
cleaned
.2
.2
.055
.1
.6
, either b
180
14
0.99
5.7
0.12
0.055
0.10
0.34
y rain or by
sweeping, whichever occurred the closest to the test date.
The figure reported here as "Residential " was computed by
combining all of the observed data for the four residential
land-use categories sampled in each city. "industrial"
and "Commercial" figures were computed similarly.
Normally, pollutants are associated with liquid discharge and are, therefore
in terms of concentration (e.g. , mg/£ or ppm). This is both reasonable and
proper in that treatment facilities operate in physical, chemical, and/or
biological modes to remove pollutants from the liquid stream so as to
minimize their concentration and total impact on the receiving water.
However, the pollutants associated with street surface comtaminants are, by
and large, in the dry state until such time that they are hydraulically
conveyed to and through storm or combined sewer systems to the receiving
water. Treatment of this type of pollutant can take place either at the
source, at the point of discharge, or more typically, not at all. If
treatment (and in the broad context this means removal of constituents) is
attempted at the point of origin, then it is apparent that it is most
appropriate to characterize the pollutants in the dry state. For this
reason, extensive studies were conducted to establish relationships. A
review of Table 47, 48, and 49 summarizes these relationships. A review
of Table 47 clearly indicates that efforts to control or remove particles
larger than 2,000 microns from streets will generally remove no more than
10 percent of a broad spectrum of pollutants, even if the removal of
these large-size particles were 100 percent effective. Putting it in
somewhat different terms, approximately 75 to 100 percent (depending on the
specific pollutant) of the pollutants are associated with particles smaller
than 2,000 microns; perhaps of even more significance, between 40 and 90
145
-------
percent of the pollutants are associated with particles-,of less than 246
microns in diameter. Clearly then, the design of any treatment method for
controlling street surface pollutants must necessarily be effective at
removing a rather broad spectrum of particle sizes.
'Table 47
FRACTION OF POLLUTANT ASSOCIATED WITH EACH PARTICLE SIZE RANGE
(% by Weight)
PARTICLE SIZE (u)
> 2,000 840 -. 2,000 246 -. 840 104 -, 246 43 - 104 < 43
Total Solids
Volatile Solids
BODS
o
COD
Kjeldahl Nitrogen
Nitrates
Phosphates
Total Heavy Metals
Total Pesticides
24.4 7.6
11.0 17.4
7.4 20.1
2.4 ' 4.5
9.9 11.6
8.6 6.5
0 0.9
16.3 17.5
0 16.0
24.6 27.8
12.0 16.1
15.7 15.2
13.0 12.4
20.0 20.2
7.9 16.7
6.9 6.4
1
14.9 23.5 :
26.5 25.8
9.7 5.9
17.9 25.6
17.3 24.3
45.0 22.7
19.6 18.7
28.4 31.9
29.6 56.2
27.8
31.7
Table 48
FRACTION OF
Chromium
Copper
Zinc
Nickel
Mercury
Lead
Average
HEAVY METALS ASSOCIATED WITH EACH PARTICLE
(% by Weight)
> 2,000 840 -. 2
26.1 13.6
22.5 20.0
4.9 25.9
26.2 14.2
16.4 28.8
1.7 2.6
16.3 17.5
PARTICLE SIZE (u)
,000 246 - 840 104 - 246
16.3 16.3
16.5 19.0
16.0 26.6
15.3 17.2
16.4 19.2
8.7 42.5
" i
14.9 23.5
SIZE RANGE
< 104
27.7
22.0
26.6
27.1
19.2
44.5
27.8
146
-------
Table 49
FRACTION OF PESTICIDES ASSOCIATED WITH EACH PARTICLE SIZE RANGE
(% by Weight)
Dieldrin
PCB
ODD
Methoxychlor
DDT
Average
840^ 2,000
0.9
33.5
10.7
35.0
0
16.0
PARTICLE
246 - 840
21.3
32.4
29.3
27.0
22.3
26.5
SIZE (u)
104 246
36.0
15.7
30.3
11.0
36.1
25.8
< 104
41.8
18.4
29.7
27.0
42.1
31.7
Table 48 shows the distribution of heavy metals relative to particle size,
and Table 50 shows how these heavy metals are distributed by land-use
category. A review of Table 50 indicates that chromium, nickel, and cad-
mium are probably of less concern than the other heavy metals, at least
on the basis of total quantities. However, as stated earlier, the lack of
definitive information concerning the individual toxic effects of these
metals (and particularly their synergistic effects with each other or
other compounds) precludes the assumption that, although chromium, nickel,
and cadmium represent an insignificant amount of total heavy metals, they
have no serious impact on receiving waters.
Table 50
DISTRIBUTION OF HEAVY METALS BY LAND-USE CATEGORY
(% by Weight)
METAL
RESIDENTIAL
INDUSTRIAL
COMMERCIAL
TOTAL
Chromium
Copper
Zinc
Nickel
Mercury
Lead
Cadmium
5
10
38
1
10
36
100%
8
14
44
5
4
25
100%
5
20
24
3
20
28
__.
100%
7
11
40
3
4
35
100%
Note: The figure reported here as "Residential" was computed by com-
bining all of the observed data for the four residential land-
use catagories sampled in each city. "industrial and
Commercial" use were computed similarly.
147
-------
The results of the investigation to date emphasize several points. The
data presented here indicate rather clearly that a broad spectrum of
pollutants exists in significant quantities in all of the cities investi-
gated and in each of the land-use areas designated therein. Further,
these data provide a basis for estimating the anticipated uncontrolled
pollutant discharge to the receiving waters of other communities. Finally,
it is now possible to make realistic and meaningful comparisons concerning
the relative impact of this non-point pollutant source on a comparative
basis with discharges from municipal and industrial sources as well as
other non-point pollutant sources as they become quantified. The study
reveals, for the first time, the dominant relationships between pollutant
properties and the particle size distributions with which they are associ-
ated. This is an extremely important relationship because it allows a
sound engineering evaluation to be made concerning the value of various
means of street litter control in reducing this non-point pollution
source. It also allows quantification of the pollutant load to a receiving
water under a wide variety of water pollution control technologies. In
fact, the identified relationships give perspective to the impact of street
cleaning practices in terms of controlling street surface pollution runoff.
THE EFFECTIVENESS OF STREET CLEANING PRACTICE
It is apparent from the preceding discussion and supporting tables that
conclusive evidence now exists confirming not only the wide spectrum of
pollutants present on urban and suburban streets, but also the order of
magnitude of the loading intensities of these pollutants. This section
concerns means of controlling the quantity of pollutants which actually
reach the receiving water. This is clearly a function of the daily accu-
mulation of pollutants on the streets and their transport, by runoff, to
streams, lakes, bays, etc. The daily accumulation, in turn, is determined
in part by street cleaning operations. Obviously, if cleaning removed all
pollutants on the streets daily then this non-point pollution source would
be reduced to insignificance.
Street sweepers (brush or vacuum) are intended to remove those types of
materials which are of concern to the public, primarily because they are
aesthetically objectionable. Almost by coincidence, street sweepers also
remove particle-related pollutants. Studies were conducted in the 1950's
and early 1960's on sweeper efficiencies (using sweepers which still
represent the current state of technology). Those studies are
valuable in that they allow estimates to be made concerning sweeper effec-
tiveness in removing pollutants. Using information developed in those
prior studies, it was possible to establish relationships between sweeper
performance and particle removal. With these data it is possible to cal-
culate the hypothetical maximum removal of selected pollutants for any
given community. As an example, consider the hypothetical community
described earlier in this section for which it is possible to compute the
resultant pollution load after sweeping. The initial step involves cal-
culating sweeper effectiveness using data in Section V.
148
-------
The resulting calculations are shown in Table 51. It is significant to
note that the range of removal efficiencies is from a high of only 79
percent to a low of 15 percent. These figures are of particular interest
when it is realized that these efficiencies represent optimum operation of
carefully adjusted equipment and are probably seldom achieved with muni-
cipally operated sweepers. Even assuming these relatively high efficiencies,
there would still remain a residual street surface contaminant loading of
about 3200 Ib of BOD on the city's streets (this assumes that the contaminants
have had about 5 days to accumulate since the last sweeping or rain).
This is equivalent to about three times the daily output from a well-run
municipal treatment plant.
A limited sub-study was conducted to determine if any consistent trends
could be found to relate the amount of contaminants found on streets with
the elapsed time since the last sweeping or substantial rainfall. Since
areas with widely differing overall characteristics were included in the
study, it was difficult to discern any dominant or repetitive trends.
The efforts involved in this sub-study are reported in Appendix I.
SIGNIFICANCE TO STREET CLEANING PROGRAMS
The conclusion is inescapable: even under well-operated and highly
efficient street sweeping programs, the broad spectrum of pollutants
accumulated in urban and suburban streets represent a non-point pollution
potential well in excess of the presently allowable discharge from munici-
pal treatment plants. Either more efficient street cleaning equipment must
be developed and put into operation or storm water must be treated prior
to discharge to the receiving waters.
Attempts to treat storm water at the point of discharge have been made in
certain instances, such as in Chicago, New York City, Washington, D.C.,
etc,; generally by storing the storm water in ponds, lakes, or underwater
bladders, removing floating and suspended matter by screening and sedi-
mentation, then releasing the water at a controlled rate. Where the storm
water contains large quantities of suspended silt and sediment, this
approach is effective. The enormous volume of water which can originate
in an urban watershed in a single storm, however, requires extremely large
and expensive storage facilities.
The cleaning of urban streets has long been a routine function of munici-
pal government. The operation was developed to meet relatively subjective
cleanliness criteria, based on individual perceptions of satisfactorily
cleaned streets. Urban sociologists have observed that this perception is
subject to large variations, in part related to socio-economic status.
Even when the goal of an adequately clean street is defined and accepted,
municipal street cleaning operations differ in their ability to achieve
that goal.
149
-------
Table 51
SELECTED POLLUTANT REMOVAL PROJECTIONS - BY STREET SWEEPERS
TOTAL SOLIDS BOD5 COD KJELDAHL NITROGEN PHOSPHATES TOTAL HEAVY METALS TOTAL PESTICIDES
SWEEPER
PARTICLE SIZE EFFICIENCY
(U) (%)
> 2,000 79
840 - 2,000 66
246 -. 840 60
104 -. 246 48
43 - 104 20
< 43 15
Total Removal
Efficiency
Size Dis- Size Dis- Size Dis- Size Dis- Size Dis-
tribution Removal tribution Removal tribution Removal tribution Removal tribution
(%) (» (%) (%) (%)
24.4 19.3 7.4 5.8 2.4
7.6 5.0 20.1 13.3 4.5
24.6 14.8 15.7 9.4 13.0
27.8 13.3 15.2 7.3 12.4
9.7 1.9 17.3 3.5 45.0
5.9 0.9 24.3 3.6 22.7
55.2 42.9
(%) (%) (%) (%)
1.9 9.9 7.8 0
3.0 11.6 7.7 0.9
7.8 20.0 12.0 6.9
6.0 20.2 9.7 6.4
9.0 19.6 3.9 29.6
3.4 18.7 2.8 56.2
31.1 43.9
Size Dis- Size Dis-
Removal tribution Removal tribution Removal
(%) (%) (%) (%) (%)
0 16.3 12.9 0 0
0.6 17.5 11.6 16 10.1
4.1 14.9 8.9 26.5 15.9
3.1 23.5 11.3 25.8 12.4
5.9 27.8 5.6 31.7 6.3
8.4
22.1 50.3 44.7
Ol
O
-------
Effective methods of planning and evaluating the efficiency of street clean-
ing practices are not available at the present time to assist those public
works personnel responsible for street cleaning programs.
Figure 45 presents a cost effectiveness program which would assist public
works officials in evaluations and/or selecting the combination of equip-
ment and operational procedures which will provide the desired cleaning
effectiveness.
As shown in Fig. 45, cost-effectiveness indices should be derived for each
street cleaning practice and for the important particle size ranges of
street surface contaminants. For each combination of equipment and opera-
tional practice there is associated:
A total cost, including fixed and variable costs
A level of effectiveness represented by a particle
size removal efficiency for specific particle size
ranges
A relationship between the particle size range and the
the pollutional properties of street surface contam-
inants.
Operational practice is composed of two elements: the operator
and the equipment type being utilized. Operator skill and training
and crew size are important inputs to operational practice. Equipment
parameters include:
Equipment type
broom
vacuum
air
combination
Number of cleaning cycles
Speed of operation
Broom parameters
type of bristle
rotation speed
contact pattern (strike)
broom pressure
condition of broom
Pickup mechanisms
hopper size
gutter brooms
Auxiliary systems
vacuum
air spray
water spray
filtration system
151
-------
Operator
STREET CLEANING PROCEDURE P,
OPERATIONAL PRACTICE
Equipment
ENVIRONMENTAL PARAMETERS
I
CURB-MILES CLEANED/HOUR
I
Cost/Man- Hour Cost/Equipment Hour
Size Removal Efficiencies
E . . . EN
Cost/Operating Hour
CP
1
Pollutional Removal Efficiencies
p, - . . P.
Cost-Effectiveness Indices for
Water Pollution Control
C /P C /">
^-p/ r . v-p/. N
Fig. 45. Cost Effectiveness Program for Street Cleaning
152
-------
Environmental parameters include:
Quantity and amount of contaminants and
refuse on street surface
Pollution potential of the various com-
ponents (dust, dirt, litter, leaves, etc.)
of comtaminants and refuse
Particle size distribution of the dust and
dirt fraction
Street type, surface characteristics
Curb and gutter configuration
Pavement type and condition
Street repair practices
Catch basin design.
As discussed in the previous section, the state of the art regarding
management information systems for public works is not very far advanced.
Existing cost accounting, work reporting, and equipment maintanence
recording systems are fragmentory and produce disparate comparative
statistical data. There is a need for a system which will aid in providing
public works personnel with accurate cost data associated with street
cleaning practices.
The technique of measurement of street cleaning effectiveness as related
to the pollutional properties of street surface contaminants was adequately
demonstrated in this study. The techniques described in Appendix A for
collection of street surface samples could be utilized to determine the
size removal efficiencies and corresponding pollutional removal
required to determine the overall cost-effectiveness indices for each street
cleaning practice evaluated.
153
-------
Section VII
ACKNOWLEDGMENTS
-------
Section VII
ACKNOWLEDGMENTS
This report summarizes research conducted by URS Research Company for the
Water Quality Office, Environmental Protection Agency, under Contract No.
14-12 921 during the period July 17, 1970 through December 31, 1971. The
work was performed under the direction of Dr. Franklin J, Agardy, Executive
Vice President and Director of the Environmental Systems Division. Mr.
James D. Sartor served as Project Manager and Mr. Gail B. Boyd served as
Assistant Project Manager.
The URS Research Team was comprised of the following:
Senior Research Engineer
Research Engineer
Research Sanitary Engineer
Test Engineer
Staff Biologists
Laboratory Assistants
Graphic Communications
Editorial /Product ion
W.
c.
R.
R.
B.
S.
C.
M.
S.
P.
H. Van Horn
Foget
Pitt
Castle
Westree
Luoma
Brennan
Sartor
Hossom
Reitman
A Project Review Panel met quarterly during the conduct of the study to
suggest direction and evaluate work progress. The panel was comprised of
the following:
Mr. Francis J. Condon
Project Officer
EPA Water Quality Office
Storm and Combined Sewer
Pollution Control Branch
Washington, D. C.
Mr. Richard Sullivan
Assistant Executive Director
American Public Works
Association
1313 E. 60th Street
Chicago, Illinois 60637
Dr. F. Pierce Linaweaver
Director of Public Works
Department of Public Works
Baltimore, Maryland
155
-------
Mr. S. Myron Tatar!an
Director of Public Works
Department of Public Works
San Francisco, California
Mr. A. R. Turturici
Director of Public Works
Department of Public Works
San Jose, California
Dr. Ross McKinney
Dean of Engineering
University of Kansas
Lawrence, Kansas
Dr. John P. Horton
President
Newark Brush Company
260 Michigan Avenue
Kenilworth, New Jersey
Public Works organizations in the cities selected for the street surface
sampling program and the in situ street cleaning tests were most coopera-
tive in providing assistance in
selection of test areas
provision of street cleaning equipment and operators
provision of traffic control during testing.
Specifically, URS Research Company wishes to acknowledge the following
personnel for their invaluable help to our project staff:
City
San Jose, California
San Francisco, California
Seattle, Washington
Bakersfield, California
Milwaukee, Wisconsin
Tulsa, Oklahoma
Atlanta, Georgia
Personnel
Mr. Richard Blackburn
Mr. Gene Toschi
Mr. Ken Hall
Mr. Chester Spurgeon
Mr. Todd Cockburn
Mr. Mark Noonan
Mr. John F. Palmer
Mr. William Jing
Mr. Joe Albert!
Mr. Jasper Harwood
Mr. Paul Guhley
Mr. James Ralston
Mr. Richard Respress
Mr. 0. A. Powers
Mr. Ralph Hulsey
Mr. Jack Campron
156
-------
Bucyrus , Ohio Mr. Walter Lammerman
Baltimore, Maryland Mr. Gene Neff
Mr. Leonard Folio
Mr. Lou Votta
Scottsdale , Arizona Mr. Mark Stragier
The major street cleaning manufacturers were visited at the initiation of
the project and were very cooperative in providing specifications and
data relating to their equipment. We wish to acknowledge, in particular,
the following organizations:
Wayne Manufacturing Co.
1201 East Lexington Street
Pomona, Calif.
Tennant Company
701 N. Lilac Drive
Minneapolis, Minn.
Elgin Sweeper Company
1300 West Bartlett Road
Elgin, 111.
American Hoist & Derrick Co.
Mobil Sweeper Division
63 So. Robert St.
St. Paul, Minn.
URS Research Company also wishes to thank Mr. William A. Rosencranz, Chief,
and Mr. Francis Condon, Project Officer of the Storm and Combined Sewer
Pollution Control Branch of the Environmental Protection Agency, Washington
D. C. for their generous assistance and guidance.
157
-------
Section VIII
REFERENCES
-------
Section VIII
REFERENCES
1. Water Pollution Aspects of Urban Runoff - U.S. Department of the
Interior, Federal Water Pollution Control Administration WP-20-15,
January, 1969
2. Storm Water Pollution from Urban Land Activity. AVCO Economic Systems
Corporation, FWQA Contract No. 14-12-187, FWQA Publication No. 11034
FKL, April 1970
3. A Multi-Phasic Component Study to Predict Storm Water Pollution From
Urban Areas, AVCO Economic Systems Corporation, OWRR Contract
No. 14-31-0001-3164
4. Snow Removal and Ice Control In Urban Areas, Robert K. Lockwbod, ed.,
AWPA, Research Project No. 114, Volume 1, August 1965
5. Environmental Impact of Highway Deicing, Report No. 11040 GKK, Edison
Water Quality Laboratory Storm and Combined Sewer Overflows Section,
R&D Edison, New Jersey, June 1971
6. Barrett, Bruce R., "Monitors Solve Fish-Kill Mystery, Civil Engineering-
ASCE, January 1971 (p. 40)
7. Stream Pollution and Abatement from Combined Sewer Overflows, Burgess &
Niple, Limited, Columbus, Ohio, Contract No. 14-12-401, November 1969
8. Water Quality Criteria, Report of the National Technical Advisory Com-
mittee to the Secretary of the Interior, Federal Water Pollution
Control Administration, Washington, B.C., April 1, 1968
9. Public Health Service Drinking Water Standards - 1962, U.S. Public
Health Service, Publication No. 956, 1963
10. Water Quality Criteria, Second Edition, McKee and Wolf California State
Quality Control Board, Publication No. 3-A, 1963
11. Browning, E., Toxicity of Industrial Metals, Butterworths, London,
England, 1961
12. Vanselow, A. P., "Microelements in Citrus," Cal. Agr., 6, 1952
13. Jones, J. R. E., "The Relation between the Electrolytic Solution
Pressures of the Metals and Their Toxicity to the Stickleback,
Gaskerosteus aculeatus L., " Jour. Exp. Biol., 16, 1939
159
-------
14. Edwards, C. A., Persistent Pesticides in the Environment, CRC Press,
1970
15. "Controversy Continues Over PCB's," Anon, Chemical and Engineering News,
December 13, 1971
16. Hitchcock, S. W., Field and Laboratory Studies of DDT and Aquatic
Insects, Conn Agr Exp Stn Bull, 1965
17. Hunt, E. G. and Bischoff, A. I., Inimical Effects on Wildlife of Periodic
DDT Applications to Clear Lake, Calif. Fish & Game, 1960
18. Ide, F. P., "Effects of Pesticides on Stream Life," Develop. Ind.
Microbiol., 1968
19. Mount, D. I. and Putnicki, G. J., "Summary Report of the 1963
Mississippi Fish Kill," Trans. 31st N. Amer. Wildlife and Nat. Res.
Conf., 1966
20. Scott, John B., Director of Research, "The American City 1970 Survey of
Street Sweeping Equipment," The American City and The Municipal Index,
December 1970
21. Laird, Carlton W. and John Scott, "How Street Sweepers Perform Today ...
in 152 selected cities across the nation," The American City, March 1971
22. Street Cleaning Programs in Western Pennslyvania Jurisdictions - A Survey,
Western Pennsylvania Chapter, American Public Works Assn and Institute
for Urban Policy and Administration, University of Pittsburgh, December
1970.
23. McGhee, Mary F., Street Cleaning Practices in Lawrence, Kansas, research
paper, May 14, 1971
24. Clark, D. W. , and E. C. Cobbin, Removal Effectiveness of Simulated Dry
Fallout from Paved Areas by Motorized and Vacuumized Street Sweepers, U.S.
Na-wal Radiological Defense Laboratory, USNRDL-TR-746, August 8, 1963
25. Public Works Information Systems , APWA-SR-36, American Public Works
Association, October 1970
26. Sartor, J. D., H. B. Curtis, H. Lee, and W. L. Owen, Cost and Effec-
tiveness of Decontamination Procedures for Land Targets, STONEMAN I,
U. S. Naval Radiological Defense Laboratory. USNRDL-TR-196, December 27,
1957
160
-------
27. Lee, H., J. D. Sartor, and W. H. Van Horn, STONEMAN II Tests of Re-
clamation Performance, Volume III, Performance Characteristics of Dry
Decontamination Procedures, U.S. Naval Radiological Defense Laboratory,
USNRDL-TR-336, June 6, 1959
28. Owen, W. L., J. D. Sartor, and W. H. Van Horn, STONEMAN II Test of
Reclamation Performance, Volume II, Performance Characteristics of Wet
Decontamination Procedures, U.S. Naval Radiological Defense Laboratory,
USNRDL-TR-335, July 21, 1960
29. Clark, D. E., and W. C. Cobbin, Removal of Simulated Fallout from Pave-
ments by Conventional Street Flushers, U.S. Naval Radiological Defense
Laboratory, USNRDL-TR-797, June 18, 1964
30. Horton, J. P., "Broom Life isn't the Most Important Cost," American
City, July 1968
31. Horton, J. P., "The Street-Cleaning Revolution," American City,
March 1963, April 1963
32. American City, Tests Rate Cleaning Efficiencies of Sweeper Brooms,
July 1966
33. Test Report on Sweeper Efficiency, for the American Public Works
Association, Wayne Manufacturing Company, May 1968
34. Waste Heat Utilization in Wastewater Treatment, URS Research Company:
Environmental Systems Division, URS 7032, prepared for Water Quality
Office, September 1971
35. Waste-water Chlorination for Public Health Protection, California
Department of Public Health, 1970
36. Engineering Management of Water Quality by R.H. McGauhey, McGraw Hill
Book Company, 1968
37. Water and Wastes Engineering, January 1971
161
-------
Section IX
BIBLIOGRAPHY
-------
Section IX
BIBLIOGRAPHY
Sullivan, Richard H. , "Problems of Combined Sewer Facilities and Overflows,"
journal__WPCF, 41:113 (January 1969)
Weibel, S.R. , R.J. Anderson, and R.L. Woodward, "Urban Land Runoff as a
Factor in Stream Pollution," Journal WPCF, 36:914 (July 1964)
Bloom, Sandra C. and Stanley E. Degler, Pesticides and Pollution, BNA ' s
Environmental Management Series, Washington, D.C. ~19~69)
Potomac Ri^ver Water Quality - Washington, D.C. Metropolrtan Area, FWPCA,
U.S. Department of the Interior (1969)
Bacon, Vinton W. , Street Sweeping as a Means of ^Controlling Water Pollution
(The Effect of Urban Environment and Municipal Practices on the Quality of~
Storm Water Runoff) , The Metropolitan Sanitary District of Greater Chicago
Wischmeier, Walter H. , and D. Smith, Ra. i nfa 1 1_E ner gy_and _I t s^ Relationship
t£j3oil_Loss, Transactions American Geophysical~Union, "39727285 (April 1958)
Storm and Combined Sewer Demonstration Projects, FWPCA, U.S. Department of
the Interior (August 1969 and January 1970)
Treatment of Overflow from Combined Sewers, San Francisco (Progress Reports) ,
Engineering-Science, Inc. (July, August, October, December, 1966, and
March 1967)
Tucker, L.S., Availability of Rain fa, 11 -Runoff^ Data ^ for _PartlyJ3ewered
Urban Drainage Catchments
Combined Sewer Overflow Seminar Papers, FWPCA, U.S. Dept. of the Interior,
DAST^37 7l 1 02o7~03 77 0 ~(Nov embe^ r ~l 9 7 0 )~ ~~
Metcalf & Eddy, Inc. , Storm Problems and Control in Sanitary Sewers -
Oakland/Berkeley, Cali7ornia7~FWreA7~Contract No. 14-12 Ts"eptembef~1969)
Pravoshinsky, N.A. and P.D. Gatillo, "Calculation of Water Pollution by
Surface Runoff," International Association on Water Pollution Research,
Minsk, USSR (1969)
An Approach to Cost-Benefit Analyses of Water Pollution Hazard, Joint PAU/
~~~
Evans, F.L. , III, E.E. Geldreich, S.R. Weibel, and G.G.Robeck, Treatment
of_Urba.n_Stormwater Runoff , JWPCF 40:R162 (May 1968)
Poertner, H.C. , P.L. Anderson, and K.W. Wolf, Urban Drainage; Practices,
Procedures, and Needs, APWA Foundation Project 119 (December 1966)
"Water Pollution Control Financing Needs," Proceedings 1970 Legislative
Seminar, WPCF, Washington, D.C. (February 24, 1970)
163
-------
17. Burgess & Niple, Limited, Consulting Engineers, £tream_Pol.lu.tiojn_and_Abate-
ment from_Combined_Sewer_Overflows (Draft), Bucyrus, Ohio, FWPCA Contract"
~ (November 1969)
18. Aerojet General Corporation, Environmental Division, ^
Combined Sewage Pollution, Volume 1 (Draft), Sacramento, California,
FWPCA7~Department of the Interior, Contract No. 14-12-197 (October 1969)
19. Roy F. Weston, Environmental Scientists & Engineers, ^t-ej.imi^na.ry_Engineering
and Applied Research_Study_-- JVashJ^gtc^i^^._jl:^onibJ^e£^we£_S^tem
20. Huang, Chang Ju, and Liao Cheng-Sun, "Absorption of Pesticides by Clay
Minerals," J^_Sanitary Engineering, 96:1057,SA5 (October 1970)
21. Storm Water Po 1 lu_t i on _f r om Urban^Land _A^c tivity , FWPCA, U.S. Department
oITh "fnterlor "11034 ~FK L ~ ( 0 7/70 )
22. Combined s r0 ve r now_Aba t emejtil^ JTe chiiology , FWQA, Dept. of the Interior,
23. Claycomb, E.L. , "Urban Storm Drainage Criteria Manual from
Denver," Civil Engineering, ASCE (May and July 1970)
24. Tholin, A.L. , and Clint J, Keifer, "The Hydrology of Urban Runoff,"
J. Sanitary Engineering, SA2;47 (March 1959)
25. Johnson, R.E. , A.T. Rossano, Jr., and R.O. Sylvester, "Dustfall as a
Source of Water Quality Impairment," J_. _ Sanitary_Engineering , 245-267
(February 1966>
26. Mahan, R.D. , Flow Characteristics of a Catch Basin, Thesis for University
of Illinois (May~237~1949)
2 7 . Pollutional Effects of Stormwater and Overflows from Combined Sewer System
(Preliminary Appraisal), U.S. Dept. of HEW, PHS , Division of Water Supply
and Pollution Control (1964)
28. Palmer, C.L. , Feasibility of Combined Sewer System, 35:163, WPCF ,
(February 1963)
29. Palmer, C.L. , The Pollutional Effects of Storm-Water Overflows from Combined
Sewers Sewage and Industrial Waste (February 1950)
30. Inaba, K. , Extent of Pollution by Stormwater Overflows and Measures for_rts
Control, presented at the 5th International"₯p~Research Conference, HA-8
(July-August 1970)
31. Friedland, A.O., T.G. Shea, H. Ludwig, Quantity and Quality- Re la tionsjiipsjg.
Combined Sewer Flow, presented at 5th International WP ResearcVconference
1-1 (July-August 1970)
32. Soderlund, G. , H. Lehtinen, Friberg, Physiochemical and Microbiological
Properties of Urban Storm Runoff, presented at 5th International WP "
Research Conference, 1-2 (July-August 1970)
164
-------
33. AWPA Research Foundation, The Causes and Remedies of Water Pollution
From Surface Drainage of Urban Areas, WPCA contract WA 66-23 (June
1968) WP 20-15
34. Cleveland, J.G., G.W. Reid, and P.R. Walters, Storm Water Pollution
From Urban Land Activity, Reprint 1033, ASCE Meeting, Oct. 13-17, 1969
35. Problems of Combined Sewer Facilities and Overflows , FWPCA, U.S. Dept.
of Interior, WP 20-11, 1967
36. Hanes, R.E., L.W. Zelazmy, and R.E. Blaser, Effects of Deicing Salts
on Water Quality and Bloat (Lit. review recommended research
Virginia Polytechnic Institute, Blackburg, Va.)
37. Wischmeter, W.H., D.D. Smith, and R.E. Uhland, "Evaluation of Factors
in the Soil-Loss Equation," Am. Soc. of Agricultural Engineers (1957)
38. Laws, J.O., "Measurements of the Fall-Velocity of Water-Drops and
Raindrops," Papers Hydrology (1941)
39. Laws, J.O., and D.A. Parsons, "The Relation of Raindrop Size to
Intensity Transactions," American Geophysical Union
40. Gun, Ross, and D. Kinzer, The Terminal Velocity of Fall for Water
Droplets in Stagnant Air; U.S. Weather Bureau, Washington, B.C. 1948
41. Seah, H., and Hayes, Mattering & Mattern, Architects-Engineers,
Engineering Investigations of Sewer Overflow Problems (Roanoke,
Virginia), FWQA, U.S. Dept. of Interior 11024D M505/70 (1970)
42. Wilkinson, R., "The Quality of Rainfall Run-off Water from a
Housing Estate," Journal of the Institution Public Health
Engineers , April 1956
43. Smith, D.C., and D.M. Whitt, "Evaluating Soil Losses from Field
Areas," Architectural Engineering, 1948
44. Ramon, V., and M. Bandyopadhyza, "Frequency Analyses of Rainfall
Intensities for Calcutta," ASCE, SAG Dec 1969, p. 1013
45. Schranfrazel, F.H., and P.H. Engineer, Chlorides Committee on Water
Pollution, Madison, Wisconsin, August 1965
46. Hamitt, F.C., and J.F. Lafferty, Laboratory Scale Device for Rain
Erosion Simulation, Tech. Report 08153-3-TU of Michigan, August 1967
47. McDonald, J.C., "Rain Washout of Partially Wettable Insoluble
Particle," Journal Geophysical Research, Vol. 6 of No. 17,
August 1967
165
-------
48. Eutrophication (causes, consequences, correctives) National Academy of
Sciences, Washington, B.C., 1969, Antic, Urban Drainage as a Factor in
Eutrophication by Werbel, p. 383
49. Gambell, A. W. and O. W. Fisher, "Occurrence of Sulfate and Nitrate
in Rainfall," Jour. Geophysical Research, v. 69, No. 29, p. 4203, 1964
50. An Engineering and Ecological Study for the Rehabilitation of Green
Lake, by University of Washington, Dept. of Civil Engineering by
Robert G. Sylvester and George C. Anderson, Feb. 1960
51. Woman and Schick, Effects of Construction on Fluvial Sediment, Urban
and Suburban Areas of Maryland, Water Resources Research, Vol. 3, No.
2, 1967
52. Swanson, Dedrick, Dudeck, "Protecting Steep Slopes Against Water
Erosion," Highway Research Record, No. 206, National Research Council,
1967
53. Diseker, E.G. and E. C. Richardson, "Roadside Sediment Production and
Control," Am. Soc. Agr. Engrs., 4, 62-68, 1961
54. Diseker, E.G. and E. C. Richardson, "Erosion Rates and Control
Measures on Highway Cuts," Am. Soc. Agr. Engrs., 5, 153-155, 1962
55. Keller, F.J., "Effect of Urban Growth on Sediment Discharge, Northwest
Branch Anacostia River Basin, Maryland," U.S. Geol. Surv. Prof. Paper
450-C, pp. C129-131, 1962
56. American Public Works Association, Control of Infiltration and Inflow
Into Sewer Systems, for Water Quality Office, EPA, 11022EFF (12/70)
57. Department of Public Works, Portland, Oregon, Demonstration of Rotary
Screening for Combined Sewer Overflows, for Water Quality Office, EPA,
11023 FDD (07/71)
58. Edison Water Quality Laboratory, Environmental Impact of Highway
Deicing, for Water Quality Office, EPA, 11040 GKK (06/71)
59. Dodson, Kinney and Lindblom, Evaluation of Storm Standby Tanks,
Columbus, Ohio, for Water Quality Office, EPA, 11020 FAL (03/71)
60. Rosenthal, P., F. R. Haselton, K.D. Bird, and P.J. Joseph, Evaluation
of Studded Tires, National Cooperative Highway Research Program Report
No. 61, Highway Research Board, National Research Council, National
Academy of Sciences and National Academy of Engineering (1969)
61. The Oxygen Uptake Demand of Resuspended Bottom Sediments, by Seattle
University Seattle, Wash., for Water Quality Office, EPA, Contract
No. 14-12-481, 16070 DCD (09/70)
166
-------
62. AVCO Economic Systems Corporation, A Multi-Phasic Component Study to
Predict Storm Water Pollution from Urban Areas, Final Report for Office
of Water Resources Research, Dept. of Interior, Contract No. 14-31-001-
3164 (December 1970)
63. Lager, John A., Robert P. Shubinski, and Larry W. Russell, "Development
of a Simulation Model for Stormwater Management," Journal WPCF,
43:2424 (Dec 1971)
64. Division of Water Resources, Dept. of Civil Engineering, University of
Cincinnati, Urban Runoff Characteristics, for Water Quality Office, EPA,
11024 DQU (10/70)
65. Metcalf and Eddy, Inc., Palo Alto, California, University of Florida,
Gainsville, Florida, and Water Resources Engineers, Inc., Walnut Creek,
California, Storm Water Management Model, Vols. I-IV, for EPA, Contract
Nos. 14-12-501-3, 11024 DOC (July-October 1971)
66. Dow Chemical Company, Chemical Treatment of Combined Sewer Overflows,
for Water Quality Office, EPA, Contract No. 14-12-9,11023 FOB (09/70)
67. Melpar, Combined Sewer Temporary Underwater Storage Facility, for FWQA,
Dept. of Interior. Contract No. 14-12-133, 11022 DPP (10/70)
68. Franklin Institute, Selected Urban Storm Water Runoff Abstracts
(periodical) for Water Quality Office, EPA, 11024 EJS (date)
69. Southwest Research Institute, Impregnation of Concrete Pipe, for Water
Quality Office, EPA, 11024 EQE (06/71)
70. American Public Works Association, Prevention and Correction of
Excessive Infiltration and Inflow into Sewer Systems, for Water
Quality Office, EPA, 11022 EFF (01/71)
71. Black, Crow and Eiderness, Inc., Storm and Combined Sewer Pollution
Sources and Abatement, Atlanta, Georgia, for Water Quality Office,
EPA, 11024 ELB (01/71)
167
-------
Appendix A
SAMPLE COLLECTION METHODS
-------
APPENDIX A
SAMPLE COLLECTION METHODS
The goals of the sampling program were fourfold:
9 To determine the manner in which contaminants are
flushed from street surfaces by rainfall runoff
To determine the quantity as well as the physical/
chemical/biological characteristics of street surface
contaminants which are removed from street surfaces
by rainfall runoff and/or by sweeping
To determine how these quantities and characteristics
vary with respect to factors such as land use,
geographical locale, and season
To examine correlations between pollutants and the
physical fractions with which they are associated.
To fulfill these goals required the development of several sampling
and testing programs. The first conducted was the simulation of rain-
fall removal effects. From the results of these early tests, pro-
gressively simpler and more efficient procedures were evolved. The
following is a summary of the sample collection techniques and test
procedures utilized during the study:
Test Procedures
Initial field tests were conducted wherein three typical street
areas (two asphalt and one concrete) were flushed by a simulated rain-
fall. (Bakersfield, California, was selected as a site for the field
tests because it was the nearest sizeable city which had not yet
experienced any significant rainfall since the preceding summer.)
The system designed and built by URS to accomplish this is shown in
Figures 26 and 27. It sprinkles fine sprays of water which cover
an area approximately 40 x 25 ft (1000 sq ft) and is constructed
of four 16-pipe manifolds mounted on a small trailer. Each manifold
has four 8 ft sprinkler booms attached at 4-ft intervals plus a
4 ft boom at the outer end of the manifold.
The rainfall simulator is wheeled into position over the designated
sample area and connected by firehose to a nearby fire hydrant; a
flow-meter, pressure gage, and valve system controlling and measur-
ing water flow rate were also connected into the system. The rainfall
simulator was calibrated experimentally to relate rainfall intensity
to pressure. Simulated rainfall flushing was conducted for a period
of 2-1/4 hr at each test site. Every 15 min during that period,
169
-------
samples of liquids and partiiculates were taken for subsequent analy-
sis. At the end of the period, the streets were flushed thoroughly
with a firehose to wash off any remaining loose or soluble matter.
Samples of this remaining material were also collected. Two rain-
fall rates, 0.2 in./hr and 0.8 in./hr, were used.
The runoff collection system consists of several watertight vacuum
boxes of 160 gal capacity, a large industrial vacuum cleaner, two,
vacuum hoses and several sandbags. The sandbags are used to make
a small dam in the gutter a short distance downstream of the test
area. The vacuum cleaner is connected to one of the vacuum boxes,
drawing a vacuum on the box. A pickup hose from the box is placed
in the gutter in front of the dam and picks up all the water runoff
coming down the gutter. When one box is filled the vacuum cleaner
and pickup hoses are switched to another box and the runoff collection
is continued. The box is fitted with a cloth filter bag to collect
all but the finest particulate matter to be saved for subsequent
analysis; the water in the box was discarded after noting its volume.
A smaller (5 gal) vacuum can was used to collect liquid samples for
analysis; the inlet nozzle was withdrawn periodically from the gutter
to assure that the 5 gal were obtained throughout the 15 min
period.
In addition, dry samples were collected with an industrial-type vacuum
sweeper (Tennant Mfg. Co. Model HD-42, Figure A-l). The dust collec-
tion system on the sweeper has been modified somewhat to simplify
Fig. A-l. Motorized Vacuum Sweeper
170
-------
sample handling. This was done by removing the dust filter collector
and substituting a special replaceable dust filter bag. Note that
the unit is not intended to simulate the effect of cleaning strets
using conventional municipal sweepers; rather, it is intended to
simulate the results which could be attained by advanced state-of-the-
art equipment employing a combination of brush sweeping and vacuum.
Tests to date have shown the small unit to be quite effective in
removing all visible debris and all but a small amount of the very
fine particulate matter.
The preliminary flushing tests in Bakersfield provided much valuable
information. On the basis of that experience, we were able to make
several important modifications on our equipment and field testing
procedure. A primary reason for conducting the tests was to determine
an appropriate sprinkling time and rate to be used as fixed parameters
in subsequent test series.
i
Results of initial testing in Bakersfield allowed improvement of the
sampling technique. The following program was developed:
Test A: Sprinkling unswept street area with simulated
rainfall
Test B: Vacuum sweeping an unswept street area
Test C: Sprinkling a previously vacuum swept street
area by simulated rainfall
Test B': Hand sweeping an unswept street area
Test C': Flushing a previously hand swept street area
using a jet of water.
All tests were conducted on adjacent sections of street using the
following standardized procedure:
Test area: 25 ft x 40 ft of level street
'oriented parallel to curb
Rainfall intensity: 1/2 in./hr uniformly for one hour
Vacuum sweeping: Two passes with Tennant HD-42 (8 to
10 min)
Hand sweeping: Two passes with a stiff-bristle
street broom
Flushing: Water applied in a high-intensity jet
until street surface foaming ceases.
In San Jose only Tests A, B and C were conducted. These tests were
carried out on streets representing seven different preselected
land-use areas. In Phoenix all five tests were conducted on each
of the streets representing each of the eight land-use areas.
171
-------
Routine Sampling Procedures
Modifications of the preceding sampling program were developed since
it was found impractical to transport the rain-simulation and other
bulky equipment across country. The standard sampling routine was
essentially reduced to a combination of Tests B1 and C', that is,
hand-sweeping and hose flushing combined with a small-scale vacuum
recovery system. The procedure consisted of three phases, locality
data taking, hand-sweeping, and flushing-vacuum recovery of remain-
ing material.
Locality Data Collection
After setting up traffic control in the chosen test area, informa-
tion was gathered as to: location, date, land use, parking and
traffic characteristics; street, gutter, and curb composition, con-
dition and texture, test area and description of the adjoining area.
At this time photographs of the area were taken.
Hand Sweeping
Hand sweeping, or dry solids collection, utilized a standard stiff
bristled broom sweeping towards the curb while moving laterally
along the street. After concentration in the gutter, samples were
collected by whisk broom and dustpan and normally placed in clean
paint cans.
Hose Flushing
Flushing was conducted after sweeping to remove adherent soluble
films and otherwise nonsweepable material. The downslope gutter
was dammed with sandbags to create a collection area for flushing
water. A small vacuum collector was constructed using 5 gal paint
cans and connected to a rented industrial wet/dry vacuum cleaner.
The test area was first slightly wetted to facilitate removal of sol-
uble materials. Flushing was then commenced at the road crown using
a garden hose and spray nozzle connected to fire hydrants. All water
used was collected by vacuum box and measured. The samples were then
mixed by vigorous stirring and split to 1 gal volume. If pesticide
analysis was to be conducted, an additional sample was taken in quart
size glass containers. Plastic gallon bottles were used for all
other samples.
Across the street sampling and sweeper testing were conducted in con-
junction with routine sampling as required.
Distribution of Street Surface Contaminants Across a Street
The test procedure involved dividing a one hundred ft long section of
street into swaths from the center line to the curb. Widths of the
parallel swaths were as follows:
172
-------
S-5 6 in. (nearest to curb)
S-4 6 in.
S-3 28 in.
S-2 56 in.
S-l remaining distance to street centerline (variable
from site to site)
The width of swaths varied, decreasing from the center line to the
curb, in order to more closely define contaminant loading in the
areas where heavy accumulations of contaminants are known to exist.
Each swath was vacuum swept twice with the Tennant HD-42 Vacuum
Sweeper to collect the samples in San Jose I and Phoenix I, while
remaining collections were by hand sweeping. With the exception of
very large samples which were split, all of the material collected in
each swath was returned to the laboratory in plastic bags. Analyses
were for total dry solids.
Equipment Performance Tests
Two methods were employed to determine the performance of street
cleaning equipment. The first relies on measuring the amount of a
street refuse simulant left on a clean street after the passage of the
equipment to be tested. The second method compares before and after
sweeping loadings of adjacent dirty streets.
Simulant Test
Test areas 50 ft by 8 ft were laid out at the test site. Each test
area was vacuum swept twice and then hosed down until all foaming
ceased. These test areas were then allowed to dry.
A street refuse simulant (see Appendix G) of realistic specific grav-
ity, size distribution and shape was applied to test areas at a pre-
scribed loading density for each specific test. A pre-weighed amount
of simulant was placed in a calibrated lawn fertilizer cart and dis-
persed over the test area. Several shallow pans of 1 sq ft area were
placed on the test area, collected after distributing the simulant,
and weighed to check the initial loading density. The street sweeper
tested made one pass over the test area, operating under the conditions
designated for the specific test. Following each test, the test area
was again hose flushed and all the water collected. Solids were sepa-
rated by settling and subsequent decantation and weighed to determine
yield.
Routine Tests
Routine equipment evaluation tests were usually conducted in conjunc-
tion with across street sampling as previously described. To conduct
a sweeper test, two areas were used. The first was swept (usually
173
-------
using across the street methods) and hosed to determine the total
initial loadings. The second area, adjacent to the first, was swept
using a street sweeper. It was then hand swept, using across the
street procedures to determine any change in distribution of loadings,
and then hose flushed with vacuum collection.
Sample Handling and Preparation Procedures
All solids and liquid samples (except for San Jose) were shipped by
air from the test sites to the laboratory. Upon receiving the ship-
ment, the laboratory technicians placed the samples in a cold room
maintained at 5°C, and placed the dry samples in a room at ambient
temperature (about 20°C). The solids were stored in new unlined metal
paint cans, while the liquids were stored in plastic containers. All
samples designated for pesticide analysis were collected and stored in
glass containers.
All individual solid samples were dried under heat lamps (less than
100°F) and weighed.
A composite solid and liquid sample for each city was then prepared by
the following technique:
Solid Composite - Each individual land-use sample for a given
city was thoroughly mixed and an aliquot of a given weight
removed. The aliquot size was based on the land-use percent-
age of the city multiplied by the amount of material found on
the sample street in that land use.
Liquid Composite - Each individual land-use sample was thor-
oughly mixed and an aliquot taken based on the land-use
percentage of each city.
Size classification of solid sample was performed by standard sieve
analysis. Sieve analysis was run on all city, land-use and area com-
posites and on all street sweeper evaluation tests. The dried solid
sample to be analyzed was placed on top of the 2000 micron screen in
the nest of 5 screens (sizes 2000 microns, 840 microns, 246 microns,
104 microns, 43 microns and the pan). The screens were then plac-d in
a roto tap unit and agitated for 1/2 hr. The screens were then
removed and the material on each screen weighed.
Special sample preparation was used for the heavy metal analysis of
the liquid samples. All liquid samples were cotton filtered prior to
analysis to remove large settleable solids.
Solid Sample Preparation
Prior to chemical analysis aliquots of each solid composite sample
were taken and placed in a blender with a known amount of distilled
water (varied according to strength of pollutant) and homogenized.
174
-------
Appendix B
SUMMARY OF CHARACTERISTICS
OF TEST SITES IN SELECTED CITIES
-------
Appendix B
SUMMARY OF CHARACTERISTICS OF TEST SITES
IN SELECTED CITIES
GLOSSARY OF TERMS USED IN TABLES B-l through B-ll
(Self-explanatory terms omitted)
Street
Pavement:
Condition:
Volume of Water:
Parking Density:
Traffic:
Density:
Minimum distance from curb (ft)
Type of surfacing
Excellent - Very smooth surface, no cracks,
essentially new condition.
Good - Few cracks , near new condition.
Fair - Cracks, some pavement deterioration.
Poor - Many cracks , moderate to extensive
deterioration.
The amount of water utilized for collecting
street surface sample (in gallons).
Heavy - Parking mostly continuous.
Moderate - Around half of available areas
filled.
Light - Very few vehicles parked.
Predominantly automobile, trucks, or mixed.
Heavy - > 10,000 AADT (annual average daily
traffic.
Moderate - 500-10,000 AADT
Light - < 500 AADT
The distance between the curb and traffic
flow.
175
-------
Table B-l
DESCRIPTIONS OF TEST SITES IN SAN JOSE DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mph)
min. distance
from curb (ft )
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW
single
/3.2S
/2-/4 -7O
ASPHALT
GOOD
18
CONCRETE
GRASS
CONCRETE
LAWN
6,80
18
LIGHT
AUTO
LIGHT
/O
4
12
/OLD
multi
6J-I -2.
£. W/iL/AM
/3.2S
/P- 14-70
ASPHALT
fAIR
15
CONCRETE
CONCRETE
GRASS
CONCRETE:
LAWN
36,0
27
LIGHT
AUTO
LIGHT
10
6
/J
n.a.
SUSZfT
MED/ NEW
single
SJ-I-3
CAMUS f
tOMBARO
26.5
12-14-70
ASPHALT
GOOD
/<£
CONCRETE
CONCRETE
GRASS
CONCRETE
LAWN
600
27
MOD.
AUTO
LIGHT
IO - /5
4-
/Z
n.&
SWff'T
MED
single
/OLD
light
COMMERCIAL
JIO*
/q.o
/2-/3-70
ASPHALT
FAIR
25
CONCRETE
CONCR£TE.
ASPHALT
NONE
D/RT
/OOO
JO
L/GHT
MIXE.D
MOD.
25
IO
13
INDUSTRY
ST-I - 7
M/SS/ON
12 -/5-70
ASPHALT
GOOD
24
CONCRETE
CONCRETE
DIRT
NONE.
ffU/LD/NGS
880
25
MOD.
AA/X£D
HEAVY
JO -40
£ -a
/8
SWEPT
heavy
CENTRAL
BUSINESS
DISTRICT
SAN FERNANDO
4.5
/2 -15-70
ASPHAL T
FAIR
20
ASPHALT
CONCRETE
D/f?T
CONCRETE
PARK. LOT
80O
40
MOD.
AUTO
HEAVY
JO -35
6 - (<,
13
n.a
SUBURBAN
SHOPPING
CENTER
ffACE f
AUZER/AS
4.5
12 -15-70
ASPHALT
GOOD
20
CONCRETE
CONCRETE.
CONCRETE
CONCRETE
PARK. LOT
800
40
LIGHT
AUTO
MOD.
20
5
/8
/i. a
-------
Table B-2
DESCRIPTIONS OF TEST SITES IN PHOENIX DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (Ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mph)
min. distance
from curb (ft)
DAYS SINCE LAST RAIN
DAYS SI NCE LAST CLEANED
CLEANING METHOD
LOW/
single
PI -1
/8.5
/-fS-71
ASPHALT
FAIR
18
CEMENT
C£MENT
DIRT
££M£*JT
LAWN
fOOO
48
LIGHT
Jt/ro
LIGHT
/S-2O
(,-8
/2
8
StV£PT
'OLD
multi
PI -2
/93/e.ftxx
/-/4 -71
ASPHALT
f~AIR
CEMEN T
CfMEHT
CEMENT
CEMENT
LAWN
/OOO
/
-------
Table B~3
DESCRIPTIONS OF TEST SITES IN MILWAUKEE DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mph)
min. distance
from curb (ft)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW
single
Mi-l
6<*f£. LLOYD
4-28-71
ASPfJAL T
GOOP
IZ
ASPHALT
CONCRETE
D/RT
CONCRETE
GRASS
4-40
/o
L/6HT
AUTO
LI6HT
15-20
4
0
7
SWEPT
/OLD
multi
Mi -2
/& 3
4 -28-71
ASPHALT
POOR
to
CONCRETE
CONCRETE
Dl RT
CONCRETE
D/RT
460
8
NO PARK.
AUTO
LIGHT
IS ' -25
O
SWEPT
MED/ NEW
single
Mi -3
4-21-7I
CONCRETE
GOOD
IB
CONCRETE
CONCRETE
LAWN
CONCRETE
LAWH
0,00
/3
L/GUT
AUTO
LIGHT
20-25
6-8
0
7
SWEPT
MED,
single
'OLD
multi
Ml -5
LATHAM (SI 0T-
16.3
4- -28-71
ASPHAL T
fA/R
18
CONCRETE
CONCRETE
DIRT
CONCRETE
LAWN
80O
15
HO PARK.
AUTO
LI GHT
20-25
0
9
3U/EPT
light
INDUSTRY
medium
Mi -7
ffCCHCKfAiLIS
12.5
4-28-71
ASPHAL T
FA/R
l(e
ASPHALT
CONCRETE
CONCRETE:
CONCRETE:
BI//LP/NGS
&00
3
NO PARK.
MIXED
MOD.
15 -2O
4-6,
O
8
SWEPT
heavy
Mi-8
f BARCKY
12.5
4 -21 - 71
ASPHALT
FAIR
16,
ASPHALT
CONCRETE
DIRT
CONCRETE
O/RT
6,00
17
A/0 PARK.
rRucK
HEAVY
15 -20
O
8
CENTRAL
BUSINESS
DISTRICT
MI -9
MASON f
BROADWAY
4.7
4-27 -71
ASPHALT
fXCEL.
25
ASPHALT
CONCRETE"
CONCRETE
CONCRETE
BU/LDINCS
(oOO
3
NO PARK.
AUTO
HEAVY
JO -35
8
O
I
SUBURBAN
SHOPPING
CENTER
Mi- 10
27r*fPARN£LL
4.7
4 -2? - 71
CONCRETE:
25
CONCRETE
CONCRETE
DIRT
CONCRETE
PARK iOT
COO
25
LIGHT
AUTO
MOD
25 -JO
8
o
7
-------
Table B-4
DESCRIPTIONS OF TEST SITES IN BUCYRUS DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET povement
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mph)
min. distance
from curb (ft)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW/
single
Bu-l
SCHABERT
SMONNZTT
18
4 -JO -71
ASPHALT
POOR
15
CONCRETE.
CONCRETE
LAU/N
'CONCRETE
LAH/N
620
15
LIGHT
AUTO
LIGHT
15 -20
3-5
2
sZrr
OLD
multi
MED / NEW
single
VICTORIA.
j MARTHA
18
4 -30 -71
ASPHALT
£XCEL.
14
CONCRETE
CONCRETE
LAWN
NONC
LAWN
480
/4
LIGHT
AUTO
LIGHT
/5-20
<*
2
fi.a.
SH/fPT
MED/
single
Bu.-4
H/ALLACE j
EAST
4-30-71
ASPHALT
EXCEL .
14
ASPHALT
CONCRETE
LAWN
CSNCRETE
LAWN
480
^o
NO PARK.
AUTO
LIGHT
/S-20
S-7
2
n.a.
'OLD
multi
light
INDUSTRY
medium
Bu-T.
AVWf WAYNE
12
4-30-71
ASPHALT
EXCEL.
ASPHALT
CONCRETE
LAWN
HONC
LAWN
480
//
LIGHT
AUTO
L/GHT
20-25
4
2
na.
heavy
Bu.-8
sour HE UN
f HARRIS
a
4-30-71
ASPHALT
POOH
/4
CONCRETE
CONCRETE
GRASS
NONE
GRASS
480
//
NO PARK.
AUTO
MOD.
25-30
4
2
n.a.
SWEPT
CENTRAL
BUSINESS
DISTRICT
US. IVARRENr
fStNOVSKY
8
4 -JO -71
ASPHALT
FAIR
/7
ASPHALT
CONCRETE
CONCRETE
CONCRETE
BUILDINGS
60O
/2
MOO.
AUTO
MOO.
20-26
6-8
2
swr
SUBURBAN
SHOPPING
CENTER
-------
Table B-5
DESCRIPTIONS OF TEST SITES IN BALTIMORE DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mph)
min. distance
from curb (ft)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW
single
/OLD
multi
Ba. -2
MIL TON f
iANVALE
28.2
J-4 -71
ASPHALT
GOOD
tc,
ASPHALT
CONCRETE
CONCRETE
CONCRETE
BUILDINGS
&8O
S3
HEAVY
AUTO
MOD.
25
8.
SW f FLUSH
MED/ NEW
single
Ba-3
3EKO/S f
PICKWICK
14.1
3-4 -71
CONCRETE
GOOD
CONCRETE
CONCRETE.
LAWN
NONE
LAWt/
6,80
/y
LIGHT
AUTO
LIGHT
20-25
(,-8
^(,
/j
SlV.JFt.(/SH
MED /OLD
single multi
Ba-4-
H.I
6-4 -71
ASPHALT
EXCEL .
/O
COHCRZTC
CONCRETE
CONCRETE.
CONCRETE.
SHRUBS
44 O
/4
MOD.
AUTO
LIGHT
t-5-20
4-C,
sw.j FLUSH
Ba-5
BAUKfELWOOD
14.1
3-4 -71
ASPHAL T
EXCEL
18
ASPHALT
CONCRETE
CONCRETE.
CONCRETE
GRASS
840
/S
NO PARK.
AUTO
MOD.
25-30
6
26,
4
light
0&-C,
S. CAROLINE
f'FLEET
3-4 -71
ASPHALT
FAIR
18
GRANITE
CONCRETE
GRANITE
GRANITE
BUILDINGS
6OO
/8
A/0 PARK.
TRUCK
Hi A V/
25-30
6-8
26,
3
SW.f FLUSH
INDUSTRY
Bs-7
EASTERN t
EAST FALLS
3-5 -71
ASPHAL T
F~AIR
20
CONCRETE
CONCRETE
CONCRETE
PARK. LOT
&00
ir
NO. PARK.
MIXED
HEAVY
25-30
6-8
4
SlV.f FLUSH
heavy
Ba-8
KEY HI3HIMY
6.4
S -5 -71
CONCRETE
EXCEL .
JO
CONCRETE
CONCRETE
CdffC/fETE
CONCRETE
GRASS
C,oo
/9
LIGHT
MIXED
MOD.
40-45
12
SW.fFiUSH
CENTRAL
BUSINESS
DISTRICT
MAR/ON /
CATHEDRAL
6.8
3- 5 -71
ASPHALT
EXCEL.
25
ASPHALT
CONCRETE
CONCRETE
CONCK£TE
BUILDINGS
400
17
NO PARK.
AUTO
HEAVY
30-35
2-3
SWfFlUSH
SUBURBAN
SHOPPING
CENTER
Ba-'O
ATHOL f
fDMOHOSOfJ
4.0
-5-6 -71
ASPHALT
EXCEL-
20
ASPHALT
CONCRETE
DIRT
CONCRETE
SHRUBS
8OO
LIGHT
AUTO
'MOD.
25-30
(, -8
4
-------
Table B-6
DESCRIPTIONS OF TEST SITES IN SAN JOSE DURING SECOND TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET povemenl
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gol )
PARKING DENSITY
TRAFFIC main types
of vehicles
density
^ overage speed (mph)
fe*M.~- - from curb (ft)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW/
single
sj'jr-i
BERKLEY
fDOBERN
/3 25
ASPHALT
GOOD
18
CONCRETE
CONCRETE
MASS
CONCRETE
LAWN
&80
18
LIGHT
AUTO
LIGHT
/O
4
31
n.a
SlStPT
'OLD
mulli
sj-Tf-2.
WILLIAMS
/3.2S
6 -/5-7I
ASPHALT
FAIR
IS
CONCRETE
CONCRETE
GRASS
CONCRETE
LAWN
5&0
27
LIGHT
AUTO
UGHT
/O
6
51
n &
MED /NEW
single
SJjr-3
CAMOS r
LOMBARD
2C. 5
6-15-71
ASPHALT
GOOD
CONCRETE
CONCRETE
GRASS
CONCRETE
LAWN
600
27
MOD.
AUTO
LIGHT
/O -15
4
51
s^rr
MED,
single
'OLD
mull!
light
COMMERCIAL
flOV
6 -15-71
ASPHALT
FAIR
26
CONCRETE
CONCRETE
ASP HA i r
NONE
D/RT
/ooo
30
LIGHT
MIXED
MOD.
25
/O
51
n.i.
INDUSTRY
medium
JQ W f
MISSION
11.0
6- 15 -71
ASPHALT
GOOD
24
COMCRETE
CONCRETE.
£>l RT
NONE
ffU/LO/HGS
880
25
MOD
MIXED
HEAVY
30-40
6-8
61
n.a.
SWEPT
heavy
CENTRAL
BUSINESS
DISTRICT
SSJT-f
f.J&f'
SAN FERNANDO
4.5
(,-15-71
ASPHALT
FAIR
20
ASPHALT
CONCRETE
Dl RT
CONCRETE.
PARK. LOT
80O
40
MOD.
AUTO
HEAVY
JO -35
-5 -6
61
n a..
SUBURBAN
SHOPPING
CENTER
SJK-/0
AU2EUAIS
("RACE
4.5
6 -IS -71
ASPHALT
GOOD
20
CONCRETE
COMCftETf
CONCRETE
CONCRETE
PARK LOT
8OO
40
LIGHT
AUTO
MOD.
20
6
51
fl A
-SWEPT
-------
Table B-7
DESCRIPTIONS OF TEST SITES IN ATLANTA DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (Ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mphj
min. distance
from curb (ft)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW
single
At-l
WALNl/TJ
THURMOND
11.3
&-2Z-7I
ASPHALT
GOOD
16
ASPHALT
CONCRETE.
GRASS
NONE
GRASS
620
/<£
L/GHT
AUTO
LIGHT
/O
4-
2
/4
sw.f FLUSH
/OLD
At -2
DREWf
CLARRILLA
11. 3
6 -2 '2 '-71
CONCRETE
GOOD
ZO
CONCRETE
CONCRETE
GffASS
CONCRETE
LAWN
640
/J
LIGHT
AUTO
LIGHT
15
C,
2
/
Sw/rti/stt
MED/ NEW
single
At-3
FERNLEAF&
fFERNLEAFRd.
113
6-22-71
ASPHALT
GOOD
15
ASPHALT
GRANITE
LAWN
NON C
LAWN
6(,0
JO
LIGHT
AUTO
LIGHT
/O
i5
2
^l
sivf FLUSH
MED
single
/OLD
multi
At -5
BOLT ON Dr.
/7.J
d -22- 71
ASPHALT
POOR
IS
CONCRETE.
CONCRETE
GRASS
CONCRETE
LAWN
4OO
20
L/GHT
AUTO
LIGHT
20-25
8
2
28
Sw / FLUSH
light
At -6
n.a.
7.4-
6 -22-7\
ASPHALT
FAIR
Id,
ASPHALT
CONCRETE
GRASS
NON£
GRASS
64 O
24
//O PARK.
TRUCK
MOD.
40
8
2
JO
SIVf FLUSH
INDUSTRY
medium
At- 7
SEABOARD
INDUST RD.
74
(,-22-7\
ASPHALT
POOR
14
ASPHALT
GftAN/TE
6RASS
NONE:
GRASS
400
27
NO PARK
MIXED
MOD.
JO
4
2
7
SlV.fFLUSH
heavy
At -8
/&&( HOLLY
74
6 -22-71
ASPHALT
18
ASPHALT
GRANITE
GRASS
CONCRETE
GRASS
/4
NO PARK.
TRUCK
MOD.
JO
&
2
/O
S IV r FLUSH
CENTRAL
BUSINESS
DISTRICT
At -9
MARIETTA
fGRADY
.2
6-2Z-7I
ASPHALT
GOOD
/6
CONCRE TE
CONCRETE
CONCRETE:
CONCRETE
BUILDINGS
4-4O
9
NO PARK
MIXED
HEAVY
20
6
2
I
SW. {FLUSH
SUBURBAN
SHOPPING
CENTER
At -10
PIEDMONT
.2
ASPHALT
£XC£L
20
CONCRETE:
CONCf?£TE
CONCftETF.
CONCRETE
STONE WALL
440
2O
NO PARK
MIXED
HEAVY
2O -3O
4-
2
14
sw. f'n USH
-------
Table B-8
DESCRIPTIONS OF TEST SITES IN TULSA DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
* width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mphj
min. distance
from curb (ft)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW/
single
Tu-l
fATOM (
GREENWOOD
24 0
& -28 - 71
ASPHALT
POOR
/4
ASPHALT
CONCRETE
GRASS
CONCRETE.
31//LD/WGS
48 O
/&
MOD.
AUTO
MOD.
/S
3
1
tt.a
JIKSTLl/SH
'OLD
mulli
MED/ NEW
single
Tu-J
45 i
BRAD EN
35.0
& -25-71
CONCRETE
fA/fi
/4
COA/CR£T£
CONCRETE
GRASS
MONC
400
17
S./GHT
AUTO
/ IGHT
/5
S
t
na.
SIV.SrtUSH
MED/
single
'OLD
Tu-5
sr. LOUIS
(£ /4a
3S.O
t - 25 - 71
CONCRETE
FAIR
14-
ca/vcfffTE
COP/CRETE
GRASS
CONCRETE
STONEWALL
400
20
/V0. PARK.
AUTO
MOD.
20
J
?
71.3
SHtf FLUSH
light
Tu-d
W-XidB1*
2.0
6 -25-71
COMRSTE
GOOD
18
CONCRETE
CONCRETE
GRASS
NONE
GRASS
64 O
JO
it GUT
rfft/CK
S./GHT
20
6,
9
xa.
SWf'nuSH
INDUSTRY
medium
Tu-7
CATIM£R
rOWASSO
20
6-25-71
ASPHALT
f/i/fl
/&
CO/VCff£T£
CONCRETE
GRASS
NONE
0u/LD/f/6S
480
20
NO PARK
TRUCK
MOD.
2O
6
9
Md
SlV.ffLUSH
heavy
CENTRAL
8USINESS
DISTRICT
r*- 9
J«* f
BOSTON
.7
C, -25-71
ASPHALT
FAIR
20
ASPHALT
CONCRETE
coMCf?er£
COMC??£ T£~
PARK tor
C4O
/q
BUS STOP
M/X£D
HEAVY
JO
8
9
Ha
Sli/fTLUSH
SUBURBAN
SHOPPING
CENTER
TV -/O
CANTON (
£. 43-
.7
a -25 -71
CONCRETE
FA If!
/(,
CO/VCff£T£
COMCf>£T£
LAlA/fJ
NOME
£A WM
440
/7
A/0 PARK.
AUTO
MOD.
25
J
-------
Table B-9
DESCRIPTIONS OF TEST SITES IN PHOENIX DURING SECOND TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SJDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gol)
PARKING DENSITY
TRAFFIC « main types
of vehicles
density
overage speed (mph)
min. distance
from curb (^)
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW
single
pii-i
WWt-fCS/d*-
/8.S
6-24-71
ASPHALT
POOR
18
CONCRETE
CONCRETE
D/RT
CONCRETE
LAWN
6(>0
E2
MOD.
AUTO
LIGHT
15
&
60 +
n.a
JCAMB£U.
~5(,.7
6-28-71
ASPHALT
GOOD
14-
CONCRETE
CONCRETE
CONCRETE
CONCRETE
LAWN
48O
18
LIGHT
AUTO
LIGHT
10
4-
6O +
nA
SWEPT
MED/
single
'OLD
PZ-3
CULVCRtiB*
-5.8
6 -£S-7/
ASPHAL T
PAIR
14-
CONCRETE
CONCRETE:
GRASS
CONCRETE
LAWN
(,00
/8
HEAVY
AUTO
LIGHT
/5
6
&O'
71.3
SWEPT
light
rn-(,
MZ/VfriLLMOIIt
6.3
6-28-71
ASPHALT
GOOD
20
CONCRETE
CONCRETE.
DIRT
NONE
OIRT LOT
J2O
/7
MOD.
MIXED
MOD.
20
8
&O +
na
SWEPT
INDUSTRY
medium
PIT -7
37 *Sf.
2.5
6-Z8-71
ASPHALT
GOOD
25
CONCRETE
CONCRETE
ASPHALT
ASPHALT
PARK. LOT
44O
A5
NO PARKINS
MIXED
HEAVY
40 -50
8
GO*
n<3,
SWEPT
heavy
CENTRAL
BUSINESS
DISTRICT
PE -1
MONROEi '/*
3.8
6-21-71
A'SPHALT
FAIR
24-
ASPHALT
CONCRETE
CONCRETE
CONCRETE
BU/LDING
-52O
20
row A WAY
MIXED
HEAVY
20
8
60 +
n.a.
SWEPT
SUBURBAN
SHOPPING
CENTER
pjr-io
J3-J GRAND
J.8
6-28-71
ASPHALT
GOOD
15
CONCRETE
CONCRETE
CONCRETE
CONCRETE
PARK LOT
3(>O
24-
LIGHT
AUTO
LIGHT
20
<0
c,o+
n.s
SWEPT
-------
Table B-10
DESCRIPTIONS OF TEST SITES IN SEATTLE DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (Ft2)
VOLUME OF WATER (gal)
PARKING DENSITY
TRAFFIC main types
of vehicles
densrly
average speed (mphj
from curb (ft)
DAYS SINCF LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW/
single
Se-l
/K°V FIR
30.0
7-8-71
ASPHALT
POOR
/2.
ASPHALT
CCA/CRETE
GRASS
COMCflETE
iAU/tJ
400
f3
L/GHT
AUTO
L/GHT
/5
4-
/2
na
SW.jnt/SH
'OLD
multi
Se-2
2la f
YESLER
9.0
7-8-71
ASPHALT
GOOD
/&
ASPHAL T
CONCRETE
CONCRETE
CONCRETE.
BUILD/fJGS
600
/&
MO PARK.
Aaro
HEAVY
JO
8
/Z
7I.-3
SHfffll/Sff
MED /NEW
single
MED,
single
Sc-4
/2*-rf
£. THISTLE
3S.O
7- 7-71
CONCRETE
GOOD
/&
CONCRETE
COfJCR£T£
GRASS
CO fJ CRETE.'
/.AlVf/
3&O
?3
tlGHT
AUTO
LIGHT
/O
6,
/2
na
SIV.f FLUSH
'OLD
multi
Sc-5
St/MYS/PE<
GKEM LAM UH\
SO
7-8-71
ASPHAL T
FAIR
/O
ASPHALT
COfl/CRETE
COMCR£T£
CONCRETE.
PLANTS
3(eO
26
MOD.
AUTO
HCAVY
30
3
/2
n.s.
SWfTLUSH
light
Sc-C.
/oa^Ave.
2O.O
7-8-71
CONCRETE.
FAIR
/2
CONCRETE
CONCRETE
O/f>T
iVON£
DIRT
4OO
/7
MOD
MIXED
tfEAVr
JO
8
/2
n 3
Siv.f FLUSH
INDUSTRY
Se-6,-2
WALK£R f
6a
7 -8-71
cot/CRerz
FAIR
fO
co/vc/tere
CONCRETE
o/ftr
NONC
O/ffT
3ZO
Z3
MOD.
M/X£D
HfAVY
JO
8
/Z
n a
SIVj 'FLUSH
heavy
CENTRAL
BUSINESS
DISTRICT
JVr-9
3" f
V/ftG/N/A
.5
7-8-71
ASPHALT
FAIR
/O
ASP HA i. T
COMCKEri.
COWCR£T£
COMCF1ETE.
PARK LOT
3&0
/O
BUS STOP
AUTO
HEAVY
25 -JO
6
12
na
Sw.f FLUSH
SUBURBAN
SHOPPING
CENTER
Sc -10
//O r-* f
/V.3r-"
/.O
7-8-71
ASPHAL T
FAIR
/Z
ASPHALT
CONCRETE
C0HCft£T£
CONCRETE
BUILDINGS
4OO
15
NO PARK
AUTO
HEAVY
JO
8
/Z
ft. -3
S Its (FLUSH
-------
Table B-ll
DESCRIPTIONS OF TEST SITES IN MERCER ISLAND, WASH.; DECATAUR, GA.; OWASSO, OKLA. .
AND SCOTTSDALE, ARIZ. DURING FIRST TEST SERIES
CODE NUMBER
SITE LOCATION
PERCENT LAND USE
DATE
STREET pavement
condition
width (ft)
(crown to gutter)
GUTTER
CURB
PARKING STRIP
SIDEWALK
AREA BEYOND SIDEWALK
SIZE OF TEST AREA (ft2)
VOLUME OF WATER (gol)
PARKING DENSITY
TRAFFIC main types
of vehicles
density
average speed (mph)
min. distance
from curb (FO
DAYS SINCE LAST RAIN
DAYS SINCE LAST CLEANED
CLEANING METHOD
LOW /OLD
single
multi
MED/ NEW
single
Mrs -3
M£RCER IS.
n.a..
7-7-71
ASPHALT
GOOD
/<*
CONCRETE
CONCRETE
GRASS
GRASS
GRASS
56,0
21
MOD.
AUTO
LIGHT
/5
6>
/2
n.a.
n.a.
MED /OLD
single
DC -4
IV/VTER AVE.
f'iARK PL.
n.a.
(,-23-71
ASPHAL T
rA/R
/4
ASPHAL T
CO/VCRETE
GRASS
CONCRETE
LAWN
440
/8
MOO.
AUTO
LIGHT
10
5
2
n.a.
n a.
multi
OH/ ~4
W.3*-f
BE A MONT
n.a.
(,-2 H -71
ASPHALT
FAIR
/3
CONCRETE
CONCRETE
GRASS
CONCRETE
LAWN
480
23
L/6HT
AUTO
LIGHT
10
3
9
n.a
n a
light
Sc-4
£. 74TJ!S
/?OOSEV£LT
n.a..
(e-21-71
ASPHAL r
GOOD
20
CONCRETE
CONCRETE
CONCRETE
CONCRETE
LAWN
680
JO
LIGHT
AUTO
LIGHT
10
to
JO*
n.a
n.a..
INDUSTRY
medium
heavy
CENTRAL
BUSINESS
DISTRICT
SUBURBAN
SHOPPING
CENTER
i
i
-------
Appendix C
DATA SUMMARY AND INVESTIGATION
OF ACCUMULATION RATES
-------
APPENDIX C
DATA SUMMARY AND INVESTIGATION OF ACCUMULATION RATES
During the course of this study, it was suggested that an attempt be
made to determine the relationship between the amount of contaminant
material at a given site and the period of time which had elapsed since
that site had been cleaned by either rainfall or sweeping. This appen-
dix describes the attempts made to determine such accumulation rates.
In the way of background, it is important to reflect back upon the
discussion of Figs. 2, 3, and 4 in Section IV. The field sampling
programs carried out in this study were directed toward collecting
materials which exist on street surfaces at a given point in time. At
some prior point in time, these sites had been cleaned, by sweeping
and/or rainfall. Thus, we have two points on a curve but know virtually
nothing about the shape of the curve between those points. The dis-
cussion in Section IV suggests that the curve is surely not linear.
Clearly, a study could be conducted to develop basic data on accumu-
lation rates, but this would require making numerous repetitive field
measurements on the same sites.
Recognizing the above, it was still decided that a sub-study should be
conducted to investigate the possibility of establishing some type of
correlation between time and loading intensity.
The data employed were of three types:
the solids loadings intensities measured at each of the sampling
sites (in Ib/curb mile)
the elapsed time since the street was last swept (these data from
public works department records)
the elapsed time since rainfall had last cleaned the streets
(NCAA weather records were scanned to determine the antecedent
period for a 1/2 in. rain storm).
Only solids data were included in this sub-study. This grossly simpli-
fied data manipulation does not introduce any appreciable error
(primarily because the total contaminant loading correlates so well with
just the solids loading).
OVERVIEW
Various numerical analysis techniques were employed, using digital
computer solutions. Analyses were conducted in five parts:
187
-------
Table C-L
QUANTITIES OF POLLUTANTS FOUND ON STREETS (lb/1000 sq ft)
POLLUTANT
BOD5
COD
P01
NOg
N
Solids
Cd
Ni
Pb
00 Zn
00
Cu
Cr
HE
Endrin
Dieldrin
PCB
Methoxychlor
p,p-DDT
Lindane
Methyl
Pa rath ion
DDD
SAN
JOSE-I
0.15
2.9
0.0066
0.031
0.020
8.6
28xlO"6
0.0018
0.018
0.013
0.0046
0.0019
0.0028
.019xlO"3
.10x10-3
llxlO"3
0
1.0,10-3
0.16x10-3
0.19x10-3
0.63xlO-3
PHOENIX MIL-
I WAUKEE
0.085 0.14
0.39 0.58
0.0029 0.0033
0.0038 0.00062
0.020 0.017
8.5 32
38x1 O"6
0.00038
0.018
0.052
0.0071
0.00056
-
0
0.12xlO~3
41xlO"3
100x10-3
0.012x10-3
0.037x10-3
0
0.006x10-3
BALTI-
BUCYRUS MORE
0.038 0.67
0.38 0.22
0.0033 0.011
0.0016 0.00042
0.016 0.021
18 11
29xlO"6
0.00085
0.0052
0.014
0.0036
0.0050
-
0 0
0.22X10"3 0.033X10-3
8.5xlO-3 llxlO-3
21x10-3 1.9x10-3
0.78x10-3 0.33xlO-3
0 0
0 0
1.1x10-3 1.1x10-3
SAN
JOSE-I I
0.50
3.8
0.042
0.0025
0.10
56
-
0.00080
0 .0085
0.0026
0.00019
0.0013
0.00080
0
0.25xlO-3
lOxlO"3
0
1.6x10-3
0
0
LlxlO-3
ATLANTA
0.021
0.14
0.0029
0.00026
0.0053
4.7
-
0.00023
0.00085
0.0012
0.00073
0.00012
0.00025
0
0.26x10-3
0.72x10-3
0
0.14x10-3
0
0
0.37x10-3
TULSA
0.20
0.42
0.0076
0.00017
0 .0092
4.6
-
0.00015
0.00042
0.00087
0.00045
0.000046
0.00027
0
0.34x10-3
0.91X10-3
0
0.18x10-3
0
0
0.48x10-3
PHOENIX
II
0,13
0.70
0.036
0.0016
0.038
12
-
0.00049
0.0016
0.0047
0.00075
0.00038
0.00029
0
O.SlxlO-3
0.85xlO"3
0
0.17x10-3
0
0
0.44x10-3
SEATTLE
0.067
0.24
0.0069
0 .00038
0.013
6.4
-
0.00039
0.007
0.0052
0.0011
0.0011
0.00048
0
0.38X10-3
15x10-3
0
2.4x10-3
0
0
1.7x10-3
NUMERICAL
MEAN
0.20
0.98
0.012
0.0042
0.026
16
32x1 O"6
0.00064
0.0074
0.012
0.0023
0.0013
0.00082
-
-
-
.
-------
Table C-2
QUANTITIES OF POLLUTANTS FOUND ON STREETS (Ib/curb mi)
POLLUTANT
BODc
lb/curb mi
COD
Ib/curb mi
POf
lb/curb mi
N03
Ib/curb mi
N
lb/curb mi
Solids
lb/curb mi
l_j
00 cd
lb/curb mi
Ni
lb/curb mi
Pb
lb/curb mi
Zn
lb/curb mi
Cu
lb/curb mi
SAN
JOSE-I
16
310
0.70
3.3
2.1
910
0.0030
0.19
1.9
1.4
0.49
PHOENIX MIL-
I WAUKEE
6.5 12
30 48
0.22 0.27
0.29 0.052
1.5 1.4
650 2700
0.0032
0.032
1.5
2.1
0.59
BALTI - SAN
BUCYRUS MORE JOSE-I I ATLANTA
2.9 61 53 1.9
29 20 400 13
0.25 1.0 4.5 0.26
0.12 0.038 0.27 0.024
1.2 1,9 11 0.48
1400 1000 6000 430
0.0026
0.077 0.085 0.021
0.47 0.90 0.077
1.3 0.28 0.11
0.33 0.020 0.066
PHOENIX NUMERICAL
TULSA II SEATTLE MEAN
14 10 4.8 18
30 54 17 95
0.64 2.8 .49 1.1
0.012 0.12 0.027 0.043
0.66 2.9 0.90 2.4
330 910 460 1500
0.0029
0.011 0.038 0.028 0.060
0.030 0.12 0.50 0.68
0.062 0.36 0.37 0.75
0.032 0.058 0.075 0.21
AVERAGE
DEVI-
ATION
16
100
1.0
0.040
1 .8
1200
0.0002
0.043
0,60
0.63
0.20
WEIGHTED
AD/ AVER-
AGE
0 . 86 14
1.1 95
0.92 1.1
0.93 0.094
0.74 2.2
0.78 1400
0.069
0.72 0.05
0.88 0.57
0.85 0.65
0.95 0.20
-------
Table C-2 (Continued)
QUANTITIES OF POLLUTANTS FOUND ON STREETS (Ib/curb mi)
SAN JOSE- PHOENIX MIL- BALTI-
I I WAUKEE BUCYRUS MORE
POLLUTANT
Cr 0.20 - 0.047 - 0.45
Ib/curb mi
Hg 0.30 - -
Ib/curb mi
Total heavy Metals 4.4 - 4.3 - 2.6
Ib/curb mi
Total Coliforms - 49 - 46
Billion/curb mi
Pecal Coliforms - - 7.0 - 12000
Million/curb mi
Endrin 2.0 0 00
10"6 Ib/curb mi
Dieldrin 11 - 10 17 3.0
ID"6 Ib/curb mi
PCB 1200 - 3440 650 1000
10"6 Ib/curb mi
Methoxychlor 0 - 8500 1600 170
10~6 Ib/curb mi
p,p-DDT 110 - 1.0 60 30
10-6 Ib/curb mi
Lindane 17 3.1 00
10~6 Ib/curb mi
Methyl Parathion 20 - 0 00
10~6 Ib/curb mi
ODD 67 - 0.5 83 100
AVERAGE WEIGHTED
SAN PHOENIX NUMERICAL DEVI- AD/ AVER-
JOSE-II ATLANTA TULSA II SEATTLE MEAN ATION MM AGE
0.14 0.011 0.0033 0.029 0.081 0.12 0.11 0.92 0.11
0.085 0.023 0.019 0.022 0.034 0.081 0.075 0.93 0.073
1.5 0.31 0.16 .63 1.1 1.9 - - 1.7
72 32 170 48 160 82 46 0.57 99
590 2900 31,000 2400 1900 7200 8000 1.1 5600 median
value
0 0 000 -__-
27 24 24 24 27 - - 0 24
1000 65 65 65 1100 - 1100
0 0 000 ---_
170 13 13 13 170 - - - 61
0 0 000 -___
0 0 000 --__
120 34 34 34 120 - 67
Total Pesticides 1400
and PCB
12000
2400 1300
1400
140
140 1400
-------
Table C-3
ESTIMATED QUANTITIES OF CONTAMINANTS WHICH WOULD
WASH OFF STREETS IN A RAINSTORM
SAN PHOENIX-
POLLUTANT JOSE-I I
BOD (Ib) 38,
COD (Ib) 748,
P0» db) 1,
N05 (Ib) 7,
N (Ib) 5,
Solids fib) 2,200,
Cd (Ib)
Ni (Ib)
Pb (Ib) 4,
Zn (Ib) 3,
Cu (Ib) 1,
Cr (Ib)
Hg (Ib)
Number Total
Conforms x 10
Number Fecal
Coliforms x 10
Endrin (grams)
Dieldrin (grams)
PCB (grams) 2
Hethoxychlor (grams)
p,p-DDT (grams)
Lindane (grams)
Methyl Parathion
(grams )
DDD (grams)
land -use a
000 17,000
000 78,000
700 580
900 800
000 3,900
000 1,700,000
7.2
460
600
400
,200
480
720
-
-
3.6
20
,100
0
200
30
36
120
HI LWAUKEE BUCYRUS
43,000 610
180,000 6,100
'L , 000 53
200 23
5,000 250
10,000,000 290,000
12
110
5,400
7,600
2,100
170
_
1.8
25
0 0
16 1.6
5,600 62
14,000 150
1.6 5.7
5.1 0
0 0
0.82 7.9
BALTIMORE
300,000
98,000
4,900
190
9,300
4,900,000
13
380
2,300
6,400
1,600
2,200
-
2.
59
0
7
2,200
380
67
0
0
220
SAN
JOSE-I I ATLANTA
130,000
960,000
11,000
650
26,000
14,000,000
-
200
2,200
670
48
340
200
.2 1.7
1.4
0
29
1,200
0
190
0
0
130
3,200
22,000
440
36
820
730,000 1
-
33
130
190
110
19
39
.54
4.9
0
19
50
0
10
0
0
26
TULSA
50,000
110,000
1,900
43
2,400
,200,000
-
40
110
220
120
12
68
6.
110
0
39
110
0
21
0
0
S3
eral
would
PHOENIX-
II
26,000
140,000
7,300
710
7,500
2,400,000
-
99
310
940
150
75
57
1 1.2
6.2
0
28
77
0
15
0
0
40
SEATTLE
23,000
82,000
2,400
130
4.300
2,200,000
-
130
2,400
1,800
360
390
160
7.7
9.1
0
59
2,400
0
370
0
0
260
sh off these amounts of contaminants.
-------
Table C-4
QUANTITY OF POLLUTANTS FOUND ON STREETS
PER DAY SINCE LAST SWEEPING
POLLUTANT
BOD5
Ib/mi/day
COD
Ib/mi/day
P043
Ib/mi/day
N03-
Ib/mi/day
N
Ib/mi/day
Solids
Ib/mi/day
Cd
Ib/mi/day
Ni
Ib/mi/day
Pb
Ib/mi/day
Zn
Ib/mi/day
Cu
Ib/ml/day
Cr
Ib/mi/day
Hg
Ib/mi/day
Total Collforms
109/mi/day
Fecal Conforms
lOVmi/day
Endrin
10"6 Ib/mi/day
Dieldrin
10~6 Ib/mi/day
PCB
10~6 Ib/mi/day
Methoxychlor
10~6 Ib/mi/day
p , p-DDT
10"^ Ib/mi/day
Lindane
10~6 Ib/mi/day
Methyl Parathron
10~6 Ib/mi/day
DDD
10"6 Ib/mi/day
SAN
JOSE-I
Swept
13 Days
1.2
24
0.054
0.25
0.160
70
0.0002
0,015
0.15
0.11
0.038
0.015
0.023
0.15
0.85
92
0
8.5
1.3
1.5
5.2
PHOE- MIL-
NIX-I WAUKEE
Swept Swept
7 Days 6 Days
Pnor Prior
0.93 2.0
4.J 8.0
0.031 0.045
0.041 0.0087
0.21 0.23
93 450
0.0005
0.0053
0.25
0.35
0.098
0.0078
8.2
1.2
0
1.7
570
1400
0.17
0
0
0.083
BALTI -
MORE
Swept
BU- 4 Days
CYHUS Prior
15
5.0
0.25
0.0095
0.48
250
0.0007
0.019
0.12
iJ 0.33
3
5 0.083
o
0> 0.11
c
o
S"
01
a 12
3000
0
0.75
250
43
7.5
0
0
25
SAN
JOSE-I I
Swept
7 Days
7.6
57
0.64
0.038
1.5
860
0.012
0.13
0.04
0.003
0.020
0.012
10
84
0
3.9
160
0
24
0
0
17
ATLANTA
Swept
16 Days
Prior TULSA
0.12
U.81
0.016
0.0015
0.03
27
u.0013
0.0048
0.0069
0.0041 £
*
0.0007 E
o
0)
0.0014 *
g
2.0
180
0
1.5
4.1
0
0.81
0
0
2.1
AVER-
NUMERI - AGE
PHOE- CAL DEVI- AD/
NIX-I1 SEATTLE MEAN ATION NM
4.5 4.6 1.0
16 15 0.96
0.17 0.18 1.1
0.058 0.064 1.1
0.44 0.38 0.85,
300 240 0.82
0.0005 0.0002 0.40
0.01 0.0056 0.54
0.13 0.054 0.42
? U.17 0.14 0.83
§ 0.045 0.037 0.82
-M
"S 0.031 0.032 i.O
0)
% ** 0.012 0.007 0.58
0} o
z 8.0 3.0 0.38
800 1100 1.3
_
_
192
-------
Table C-5
QUANTITY OF POLLUTANTS FOUND ON STREETS PER DAY SINCE LAST MAJOR RAINFALL
POLLUTANT
BOD5
Ib/mi/day
COD
Ib/mi/day
P04=
Ib/mi/day
N03-
Ib/mi/day
N
Ib/mi/day
Solids
Ib/mi/day
Cd
Ib/mi/day
Hi
Ib/mi/day
Pb
Ib/mi/day
Zn
Ib/mi/day
Cu
Ib/mi/day
Cr
Ib/mi/day
Hg
Ib/mi/day
Total Coliforms
i09/mi/day
Fegal Coliforms
10b/mi/day
Endrin
10~6 Ib/mi/day
Dieldrin
10 Ib/mi/day
PCB
10"6 Ib/mi/day
Hethoxychlor
10~6 Ib/rai/day
p, p-DDT
10~6 Ib/mi/day
Lindane
10-6 Ib/mi/day
Methyl Parathion
10~6 Ib/mi/day
ODD
10"6 Ib/mi/day
SAN
SAN PHOENIX- MIL- BALTI- JOSE- AT-
JOSE-I I WAUKEE BUCYRUS MORE I I LANTA TULSA
Rain Rain Rain Rain Rain Rain Rain Rain
14 Days 12 Days 1 Day 2 Days 26 Days 59 Days 2 Days 9 Days
1.1 0.54 12 1.5 2.3
22 z.3 48 15 0.77
0.050 0.018 0.27 0.13 0.038
0 .24 0 .024 0 .052 0 .06 Q .002
0.15 0.13 1.4 0.60 0.073
65 54 2700 700 38
0.0002 0.0032 0.0001
0.014 0.032 0.003
0 .14 1 .5 0.018
0.10 2.1 0.05
0.035 0.59 0.013
0.014 0.047 0.017
0.021
49 !.
7 .0 460
0.14 000
0.79 10 s.o .12
86 3400 330 38
0 8500 800 6.5
7.9 3.1 30 i.ji
l.z 000
L.t 000
4.8 0.5 42 3.8
0.90 0.95 1 .6
b.8 6.5 3.3
0.76 0.13 0.060
0.0046 0.012 0.0013
0.19 0.24 0.073
100 220 37
0.0014 0.011 0.0012
0.015 0,039 0.0033
0.0047 0.055 0.0069
0,0003 0.033 0.0036
0.0024 0.055 0.0004
0.0014 0.012 0.0021
1.2 16 19
10 1500 3400
000
0.46 12 2.7
19 33 7.2
000
2.9 0.3 l.n
000
000
2.0 17 3.8
PHOENIX SE-
II ATTLE HU- AVER-
Rain 30+ Rain MERI- AGE
Days 12 Days CAL DEVI-
.33 0.40 2.1 L.y 0.90
i.O 1.4 11 10 0.91
0 .093 0 .041 0 .090 .29 3.2
0 .004 0.0023 0 ,040 0 .047 1 ,2
0.097 o .075 0.30 0 .31 1 .0
30 38 400 520 L . j
0 .0012 0 .0010 0 .83
0.0013 0.0023 0.0078 0.0074 0.95
0.004 0.012 0.22 0.31 1.5
0.012 0.031 0.30 0.46 1.5
0.019 0.0063 0.086 0.13 1.3
0.00097 0 .0068 0 .018 0.017 0 .94
0 .00073 0 ,0028 0 . 0065 0 .0063 0 . 97
1.6 13 15 12 0.79
80 160 800 940 I .2
0 0 0.016 0.028 1 .B
0.80 z.j 4.2 4.0 0.95
2.2 92 460 660 1 .4
0 0 1000 1700 1 . i
0.43 14 7.2 6.7 0.93
0 0 0.13 0.23 1 .a
0 0 0.16 0.28 l .B
1 .1 10 9.4 9,1 0.97
193
-------
Table C-6
QUANTITY OF POLLUTANTS FOUND
ON STREETS PER CURB MILE PER
DAY SINCE LAST SWEEPING OR
LAST MAJOR RAINFALL
SAN
JOSE-I
13 Days
BODg i.2
Ib/mi/day
COD 24
Ib/mi/day
PO4= 0.054
Ib/mi/day
NOs- 0.25
Ib/mi/day
N 0.16
Ib/mi/day
Solids 70
Ib/m i/day
Cd 0.0002
Ib/mi/day
Ni 0,015
Ib/mi/day
Pb 0.15
Ib/mi/day
Zn 0.11
Ib/mi/day
Cu 0.038
Ib/mi/day
Cr 0.015
Ib/mi/day
He 0.023
Ib/mi/day
Total Heavy 0.35
Metals
Ib/mi/day
Total Coliforms
109/mi/day
Fecal Coliforms
106/m i/day
Endrin 0.15
10~6 Ib/mi/day
Dieldrin 0.85
10~6 Ib/mi/day
PcB 92
10~6 Ib/mi/day
Methoxychlor 0
10~6 Ib/mi/day
p, p-DDT 8.5
10~6 Ib/mi/day
Lindane 1.3
10-6 Ib/mi/day
Methyl Parathion 1.5
10~6 Ib/mi/day
ODD 5 . 2
10~6 Ib/mi/day
Total Pesti- 110
aides and PCB
10~G Ib/mi/day
PHOE- MIL- BALTI-
NIX-I WAUKEE BUCYRUS MORE
13 Days 1 Day 2 Days 4 Days
0.93 12 1.3 15
4.3 48 15 S.O
0.031 0.2? 0.13 0.25
0 . 04 1 0 . 052 0 . 06 0 . 0095
0.21 1.4 0.60 0,48
93 2700 700 250
0.0032 0.0007
0.032 0.019
J..D 0.12
t.L 0.33
0.59 - 0.083
0.047 0.11
4.3 3 0.66
49 12
7 . 0 3000
0 00
10 8.5 0,75
3400 330 250
8500 800 43
3.1 30 7.5
000
000
0.5 42 25
12000 1200 330
SAN
JOSE-I I ATLANTA NU- AVER-
7 Days 2 Day« PHOENIX- CAL DEVI- AD/ AVER-
7,6 0.95 5.5 5.0 0.91 4.5
57 6.3 22 17 0.77 26
0.64 0.13 0.21 0.15 0.71 0.37
0.038 0,012 0.066 0.053 0.80 O.O29
1.0 0.24 C 3 0.67 0.47 0.70 0.66
860 220 700 620 °-89
0.0014 0.0012 0.86
0.012 0.011 0.018 0.006 0.33
0.13 0.039 0.38 0.44 1.2
0.04 0.055 0.53 0.63 1.2
0.003 0.033 a a 0.15 0.18 1.2
0.020 0.055 0.050 0.027 0.54
0.012 0.012 0,016 0.005 0.31
0.22 0.21 A>± 1<3
10 16 22 14 0.63
84 1500 1100 1100 1.0
0 0 e.O
3.9 12 820
160 33 710
0 0 =r 13
24 6.5
0 0
0 0
17 17 le
200 69 2400 200
194
-------
Table C-7
STRENGTH OF SOLID MATERIAL
SAN
POLLUTANT
BOD5 ppm 17
COD ppm 340
PO^ ppm
N0§ ppm 3
N ppm 2
Cd ppm
Ni ppm
Pb ppm 2
Zn ppm 1
Cu ppm
Cr ppm
Hg ppm
Endrin 10~3 ppm
Dieldrin 10"3 ppm
JOSE- PHOENIX-
I I
,000 10,000
,000 46,000
770 340
,600 450
,300 2,300
3.3
210
,100
,500
540
220
330
2.2
12
PCB 10~3 ppm 1,300
Methoxychlor 10 ppm
p.p-DDT 10"3 ppm
Lindane 10 ppm
Methyl Parathion
10~3 ppm
DDD 10~3 ppm
Number Total Coliforms/
gram solids
Number Fecal Coliforms/
gram solids
0
120
19
22
73
-
.
MIL-
WAUKEE BUCYHUS
4,400 2,100
18,000 21,000
100 180
20 89
530 890
1.2
12
560
1,600
220
18
-
0 0
3.8 12
1,300 470
3 , 100 1 , 200
0.38 43
1.2 0
0 0
0.19 61
40,000
5.7
BALTI-
MORE
61,000
20,000
1,000
38
1,900
2.6
77
470
1,300
330
450
-
0
3.0
1,000
170
30
0
0
100
100,000
SAN JOSE-
I I ATLANTA
8,900 4,500
68,000 30,000
750 620
45 55
1,800 1,100
-
14 49
150 180
46 260
3.4 160
23 25
14 53
0 0
4.4 55
100 150
0 0
28 30
0 0
0 0
20 79
26,000 160,000
200 15,000
TULSA
43,000
91,000
1,700
37
2,000
-
33
91
190
98
10
59
C
74
200
0
39
0
0
100
1,110,000
210,000
PHOENIX-
II
1 1 , 000
58,000
3,000
130
3,200
-
41
130
390
39
32
24
0
26
71
0
14
-
0
37
120,000
5,800
SEATTLE
10,000
38,000
1,100
59
2,000
-
61
1,100
810
172
172
75
0
59
2,300
0
380
-
0
270
770,000
9,100
NU- AVERAGE
MERICAL DEVI-
MEAN AT I ON;
17,000 13,000
73,000 68,000
980 610
460 630
1 , 900 570
-
54 34
530 510
760 450
200 130
120 130
93 120
-
28
780
500
75
-
-
82
330,000
38,000
AD/NM
0.76
0.93
0.62
1.4
0.32
-
0.63
0.96
0.59
0.65
1 .1
1.3
-
-
-
-
-
-
-
-
_
-------
Table C-8
OXYGEN DEMAND OF STREET SURFACE CONTAMINANTS - VARIATION BY PARTICLE SIZE
H
3
g
BOD *
5
*
COD
Volatile
portion
of solids
in
specific
size
range
SJI
Mi, Bu, Ba
At, Tu, PII
SJII, Se
SJI
At, Tu, PII
Mi, Bu, Ba
SJII and Se
PI
Mi, Bu, Ba
SJII
At, Tu, PII
Se
Particle Size, (|j.)
<43U 43-104JJ.
24 . 7%
30.3
19.8
22.7
1.4%
38.3
5.2
45.7
11.5%
13.3
16.7
21.4
14.1
14.4%
14.0
11.6
26.4
22.7%
43.1
2.8
26.5
10.0%
13.3
16.5
7.6
8.5
104-246U
18.2%
13.3
13.3
18.9
23 . 6%
9.0
1.5
15.4
9.5%
9.4
7.8
8.6
7.6
246-840|j,
26.4%
17.9
19.9
9.3
3 8. 4%
6.9
0.4
6.2
10.4%
7.9
4.2
5.3
3.9
840-2000u
13.4%
21.2
32.9
6.3
9.2%
2.7
0.1
6.2
5.1%
10.9
11.8
8.1
11.7
>2000u
2.9%
3.3
2.5
16.4
4.7%
-
0.0
-
2.8%
5.5
8.6
4.7
3.5
* Tabulated values are percents of total pollutant associated with each size range
(% by weight).
<£>
CD
-------
Part 1. The solids loading intensity data were grouped into resi-
dential, industrial, and commercial land uses, plus all
land uses combined. The following curve forms were fitted
by the least squares method to determine loading intensity/
unit time.
Y = A + BX Eq. (a)
Y = ABX Eq. (b)
Y = A exp(BX) Eq. (c)
Y = A + B/X Eq. (d)
Y = 1/(A + BX) Eq. (e)
Y = X/(A + BX) Eq. (f)
Part 2. Data on loading intensities for all land-use categories
were grouped by particle size categories (<246fji for small
and >246/i for large) and compared to days since last swept,
days since last rain, and days since last cleaning by
either sweeping or rain. The same curves used in Part 1
were then fitted.
Part 3. The solids loading data were again grouped by land-use
category (as in Part 1) and the following analysis was
performed:
Mean, variance, standard deviation, standard error,
coefficient of variation, minimum, 10th percentile,
1st quartile, median, 3rd quartile, 90th percentile,
maximum, quartile deveation, average deviation,
moment coefficient of skewness, and Pearson coeffi-
cient of skewness were computed
Histograms were plotted
Cumulative histograms were plotted.
Part 4. Using the data obtained from Part 3 with the loadings
grouped by land-use categories, values less than the 10th
percentile and greater than the 90th percentile were
rejected. The same curves used in Part 1 were then fitted.
Part 5. Additional computer runs were made to determine any dif-
ference between cleaning by sweeping or by rain without
dividing the data by land use. Only the mid-80 percent
of the rates were used. The same curves used in Part 1
were then fitted.
197
-------
RESULTS
The results of the five-part analysis of observed data are presented
in the following paragraphs, tables, and figures.
PART 1
This analysis (Table C-9) deals with the influence of land-use category on the
correlation between solids loading and elapsed time since streets were
last cleaned by either sweeping or by rainfall (whichever occurred first).
if
Table C-9
LOADING INTENSITIES/UNIT TIME FOR DIFFERENT LAND-USE AREAS
1. RESIDENTIAL LAND USE
EQUATION CURVE FORM
a. Y = 1306 - 62X
INDEX OF ,.
DETERMINATION
3.5 x 10
-2
-3
b.
c .
d.
e .
f .
Y = 654 \X """ A ^ )
Y = 603 exp (-2.08 x 10~3X)
Y = 435 + 1200/X
Y = l/(4.44 x 10~3 - 2.43 x 10~4X)
Y = X/(3.88 x 10~3 + 1.42 x 10~3X)
5.7 x 10
7.0 x 10~5
0.10
2.6 x 10~2
-2
5.1 x 10
2 . COMMERCIAL LAND USE
EQUATION CURVE FORM
a .
b.
c.
d.
e .
f .
Y = 306 + .54X
Y = 215 exp (7.03 x 10~ X)
Y = 354 - (59.7/X)
Y = I/ (6. 44 x 10~3 - 6.71 x 10~5X)
Y = X/ (4.93 x 10~4 + 5.83 x 10~3X)
INDEX OF
DETERMINATION
1.0 x 10~4
6.4 x 10~3
2.2 x 10~3
5.5 x 10~3
5.3 x 10~3
1.3 x 10~3
198
-------
3 . INDUSTRIAL LAND USE
EQUATION CURVE FORM
a. Y = 1450 + (-67.8X)
(x-5.24xlO-2)
b. Y = 957
c. Y = 998 exp (-2.50 x 10~ X)
d. Y = 1070 + 255/X
e. Y = I/(1.55 x 10~3 - 1.25 x 10~5X)
f. Y = X/(4.95 x 10~ + 1.29 x 10~3X)
INDEX OF
DETERMINATION
4.9 x 10
-2
2.3 x 10
9.2 x 10
4.9 x 10
9.8 x 10'
1.1 x 10
-3
i
-3
-3
-4
-2
4. ALL LAND USES COMBINED
9.3 x 10
EQUATION CURVE FORM
a. Y = 948 - 24.IX
b. Y=429(x°-144)
c. Y = 470 exp (1.20 x 10~2X)
d. Y = 756 + 169/X
e. Y = I/(4.40 x 10~3 - 1.70 x 10~4X)
3 ' 3
f. Y = X/(3.88 x 10 + 1.60 x 10 X)
X = elapsed time since last clean (days)
Y = solids loading on streets (Ib/curb mile)
INDEX OF
DETERMINATION
-3
1.6 x 10
2.4 x 10
3.3 x 10
1.9 x 10
7.1 x 10'
-2
-3
-3
i
-2
i
-2
PART 2
This analysis (Table C-10) deals with the influence of particle size on the
correlation between solids loading and elapsed time since street was last
cleaned (by rainfall, by sweeping, or by whichever occurred first).
Data for all land-use categories were combined here.
199
-------
Table C-10
LOADING INTENSITIES/UNIT TIME BY PARTICLE SIZE
a .
b.
t: .
d.
c .
f .
STREETS LAST CLEANED
LARGE PARTICLES
(>246U)
INDEX OF
DETERMIN-
CURVE FORM ATI ON
Y = 321 + 5.46X 0.24
Y = 314 \X8' 9 * 1 / 3.9 x 10"
Y = 316 exp (9.34 x 10~ X) 0.16
Y = 446 - 42.5/X 1.1 * 10"
V = l/(3.32 x 10~3 1.98 *. 10~ X) 0.11
Y = X/(-1.18 x 10"J + 3.07 x lo"3X) 3.0 x \0~*
BY RAINFALL
SMALL PARTICLES
a
b
c
d
e
f
CURVE FORM
Y = 167 + 13. 6X
Y = 210 \X ' )
Y = 202 exp (2.02 x 10~2X)
Y = 504 - 284/X
V = 17(5.51 x 10" - 5.55 *
Y = X/(-1.69 x 10"3 +4.83
INDEX OF
DETERMIN-
ATION
0.37
6.9 x 10"
0.28
3.4 x 10"2
10"5X) 0.13
x 10" X) 2.0 x W~*
«.
b.
c .
d.
c .
f .
STREETS LAST CLEANED
LARGE PARTICLES
(>246M)
INDEX OF
DETERMI N-
CURVE FORM ATI ON
Y = 586 - 18.2X 0.35
Y = 832 \X ' / 0.39
Y = 551 exp (-3.46 x 10 X) 0.26
Y = 221 + 1420/X 0.64
Y = 1/(2.00 + 6.50 x 10 X) 0.17
Y = X/(-6.03 x 10~ + J.43 x 10~ X) 0.44
a
b
c
d
e
f
BY SWEEPING
SMALL PARTICLES
(<246(i)
CURVE FORM
Y = 137 » 18. IX
Y = 113 lx°'411/
2
Y = 159 exp (5.77 x 10 X)
Y = 412 - 824/X
Y I/ (6. 28 x 10 2.24 x
Y X/(6.65 x 10"3 + 3.35 x
INDEX OF
DETERMIN-
ATION
0.25
0.15
0.20
0.15
10 X) 0.14
lO^X) 3.4 x l246f()
INDEX OF
DETERMI N-
CURVE FORM ATI ON
V = 528 - 13. 8X 0.29 a
Y = 564 (x~°'182) 0.28
2 b
V = 513 exp (-2.99 x 10 X) 0.27
Y = 356 + 301/X 0.21
-3 -5 d
Y = 1/(2.02 x 10 + 6.62 x 10 X) 0.25
-3 -3 e
Y = X/(-1.72 x 10 + 2.92 x 10 X) 0.25
SMALL PARTICLES
(<246n)
CURVE FORM
Y = 267 + 12. 3X
Y = 299 (X-175 * 10~2)
Y = 266 exp (1.94 X 10"2X)
Y = 338 28.2/X
Y = 17(3.85 x 10"3 + 1.20 x
Y = X/(-1.86 x 10"3 + «.68
INDEX OF
DETERMIN-
ATION
0.13
9.2 x 10~4
2.6 x 10"
3.9 x 10~3
10"5X) 4.8 x l
-------
Streets Last Cleaned by Rain
Large Particles
Curve Index of Determination
Y = 321 + 5.46X 0.24
Small Particles
Curve Index of Determination
Y = 167 + 13.6X 0.37
Streets Last Cleaned by Sweeping
Large Particles
Curve Index of Determination
Y = 221 + 1420/X 0.64
Small Particles
Curve Index of Determination
Y = 137 + 18.IX 0.25
By comparing the indexes of determination and, therefore, the regularity
of the data for specific groupings, it is possible to compare effective-
ness. It is seen that the effectiveness of removal of particles by the
rain is about the same for large particles as it is for small particles;
whereas for street sweeping large particles are removed better than small
particles.
PART 3
This analysis dealt with determining the nature of the statistical distri-
bution of all of the observed data (for use in Parts 4 and 5). Figure
C-l shows the computed histograms comparing frequency of rates of accumu-
lation. Table C-ll presents the results of the statistical analysis.
201
-------
O
of
z
UJ
o
z
Z
o
z
II
15
10
15
10
RESIDENTIAL
8
a
COMMERCIAL
INDUSTRIAL
I ALL LAND-USE CATEGORIES
to 26 COMBINED
M
8 8
Tt -O
RATE of ACCUMMULATION of SOLIDS ON STREET (Ib/curb mile/day)
Fig. C-l. Computed Histograms of Accumulation Rate of Solids on Street
202
-------
Table C-ll
STATISTICAL ANALYSIS OF ACCUMULATION DATA
STATISTICAL
PARAMETER
Mean
Variance
Standard Deviation
Standard Error
Minimum
10th Percentile
1st Quartile
Median
3rd Quartile
Maximum
Range
10-90 Percentile Range
Ouartile Deviation
Average Deviation
Moment Coefficient of
Skewness
Skewness
RESIDENTIAL
LAND-USE
(Ib/mi/day)
373
0.251 * 10S
501
107
1.34
24.0
31.2
74.5
125
491
966
1940
1920
934
208
378
1.69 '
1.49
INDUSTRIAL
LAND-USE
(Ib/ml/aay)
447
0.250 x 106
500
139
1.12
45.0
75. u
149
215
499
859
1850
1800
784
175
346
1.63
1.39
COMMERCIAL
LAND-USE
(Ib/mi/day)
226
0.835 x 105
289
74. B
1 . 28
8.00
30.8
64. u
204
223
263
1220
1210
233
79.3
143
2.62
0.23
ALL LAND-USE
CATEGORIES COMBINED
(Ib/mi/day)
348
0.200 x
447
63.
1.
8.
32.
74.
186
384
930
1940
1930
897
155
319
2.
1
io6
3
29
00
8
5
03
09
PART 4
This analysis deals with the influence of land-use category on the correla-
tion between solids loading and elapsed time since streets were last cleaned
by either sweeping or by rainfall (whichever occurred first). It is quite
similar to the analysis performed in Part 1 but does not include any data
lower than the 10th percentile or higher than the 90th percentile (see
Part 3). Table C-12 shows the best fitting equa ions for the standard
curve types.
203
-------
Table C-12
LOADING INTENSITIES/UNIT TIME SINCE LAST CLEANED
1. RESIDENTIAL LAND-USE
EQUATION CURVE FORM
a. Y = 495 + 34.8X
b.
c.
d.
e.
f.
Y = 446
Y = 426 exp (5.65 x 10~^X)
Y = 773 - 244/X
Y = I/(5.27 x 10~4 + 1.84 x 1Q3X)
Y = X/(5.27 x 10~4 + 1.84 x 10~3X)
INDEX OF
DETERMINATION
0.12
-2
9.1 x 10
0.17
3.6 x 10~
0.13
2.2 x 10~
2.
COMMERCIAL LAND-USE
EQUATION CURVE FORM
a. Y = 174 + 41.6X
b. Y = 170
c. Y = 162 exp (0.126X)
d. Y = 694 - 519/X
e. Y = 1/(7.19 x 10~ - 5.28 x 10~4X)
f. Y = X/(7.84 x 10~3 - 4.32 x 10~4X)
INDEX OF
DETERMINATION
0.61
0.46
0.32
0.88
0.12
0.25
3.
INDUSTRIAL LAND-USE
EQUATION
CURVE FORM
a. Y = 772 + 55.IX
b. Y = 556 (x°-397)
c. Y = 617 exp (8.91 x 10 ~X)
d. Y = 1300 - 777/X
e. Y = I/(1.96 x 10~3 - 1.55 x 10~4X)
~2
f. Y = X/(1.87 x 10
~3
6.01 x 10~4X)
INDEX OF
DETERMINATION
0.13
0.27
0.20
0.26
0.22
0.32
204
-------
4. ALL LAND USES COMBINED
INDEX OF
EQUATION CURVE FORM DETERMINATION
a. Y = 480 + 43.5X 0.12
b. Y = 296 (X°-511) 0.31
c. Y = 329 exp (9.85 x 10~2X) 0.21
d. Y = 994 - 652/X 0.25
e. Y = l/(4.29 x 10~3 - 3.25 x 10~4X) 0.14
f. Y = X/(5.09 x 10~ + 3.38 x 10~ X) 0.30
X = elapsed time since last clean (days)
Y = solids loading on streets (Ib/curb mile)
Figure C-2 illustrates the best fitting curves for each land-use category
when the extreme 10 percent values are disregarded. The appropriate
equations and indexes of determination are:
Residential land-use category:
Y=426 (e°-0565X); index of 0.17
Industrial land-use category:
Y = X/(0.00187 + 0.000601X); index of 0.32
Commercial land-use category:
Y = 694 - 519/X; index of 0.88
All land-use areas combined:
Y = 296 (x°'511); index of 0.31
205
-------
1400
1200
1000
,0
f-i
CJ
\
J2
iI
**^
O
}H
Q
§
Q
ii
O
Q
!=> 200
O
O
0
01 2345678
ELAPSED TIME SINCE LAST CLEANING BY SWEEPING OR RAIN (days)
Fig. C-2. Time Since Last Cleaning vs Solids Loading
It can be seen that the slopes of the straight line portions of the
commercial and residential land-use area curves are both smaller than
the slope of the industrial land-use regeneration curve. These slopes
were found to be as follows:
LAND-USE CATEGORY
SLOPE
(Ib/curb mile/day)
residential
industrial
commercial
all combined
28
60
10
57
206
-------
Part 5
This analysis (Table C-13) deals with the influence of cleaning method
(i.e., rainfall vs sweeping) on the correlation between solids loading
and elapsed time since last cleaning. Here too, only data between the
10th and the 90th percentile were included.
Table C-13
LOADING INTENSITIES/DAYS SINCE LAST CLEANED
BY RAINFALL OR SWEEPING
STREETS LAST CLEANED BY RAINFALL
INDEX OF
EQUATION CURVE FORM DETERMINATION
a. Y = 587 + 26.9X 0.057
b. Y = 435 (x0'317) 0.19
c. Y
d. Y
e. Y
f. Y
= 460e"' " A ""
= 1028 - 577/X
= l/(2.56 x 10~3
= X/(1.97 x 10~
0.13
0.17
- 1.18 x 10~ X) 0.17
+ 9.58 x 10~4X) 0.30
EQUATION
STREETS LAST
CURVE FORM
CLEANED BY SWEEPING
INDEX OF
DETERMINATION
a. Y = 324 + 68. OX 0.24
(x°-739)
b. Y = 191 Vx / °-51
c. Y = 214 (e°'162X) 0.37
d. Y = 986 - 768/X 0.37
e. Y = l/(6.32 x 10~3 - 6.13 x 10"4X) 0.26
3 5
f. Y = X/(7.56 x 10 + 6.27 x 10 X) 0.47
= elapsed time since last clean (days)
Y = solids loading on streets (Ib/curb mile)
207
-------
Figure C-3 illustrates the best fitting curves for the above
two groups of data (b and f, respectively), and the curve for all land-
use areas combined for days since cleaned, from the preceding study. It
can be seen that difference between cleaning by sweeping and by rain is
negligible on a total weight basis. See preceding computer study for
differences in cleaning effectiveness for small and large particle sizes
because of cleaning method.
o
K
l-H
Q
3
ra /-N
Q
-------
Appendix D
TYPICAL LAND-USE CATEGORIES
-------
RESIDENTIAL: low/old/single
RESIDENTIAL : med/new/single
209
-------
RESIDENTIAL : low/old/multi
RESIDENTIAL : med/old/multi
210
-------
RESIDENTIAL : med/old/single
INDUSTRIAL: light
211
-------
INDUSTRIAL : medium
INDUSTRIAL: heavy
212
-------
COMMERCIAL : shopping center
COMMERCIAL : central business district
213
-------
Appendix E
METHODS USED FOR ANALYSIS
-------
Appendix E
METHODS USED FOR ANALYSIS
The heavy metals were analyzed using standard atomic absorption (Ref. E-l)
methods by Metallurgical Laboratories, Inc. of San Francisco. The metals
analyzed included cadmium, nickel, lead, zinc, copper, chromium and mer-
cury. All valence states were measured combined.
The pesticide and PCB analyses were performed by Morse Laboratories of
Sacramento, using standard gas chromatograph (Ref. E-l) methods. All
chlorinated hydrocarbons were tested for in each sample, but only those
found were listed. All organic phosphates were tested for in selected
samples, and again, only those found were listed. PCB's were run separately
because of their interference on the pesticide analyses.
The soluble nitrate analyses were performed by Cook Research Laboratories,
Inc. of Menlo Park, using tentative methods (Ref. E-l). Total Kjeldahl
nitrogen includes ammonia and organic nitrogen, but not nitrite and nitrate
nitrogen. The total Kjeldahl nitrogen analyses were also performed by
Cook Laboratories, using standard methods (Ref. E-l).
The five day biochemical oxygen demand tests (BODs) were performed by
Cook Laboratories, using the standard (Ref. E) five day bottle technique.
The chemical oxygen demand (COD) tests were run by U.RSRC personnel. The
liquid samples were treated in the usual manner, and a slurry was made
from the solid materials. The selected samples were boiled with potassium
dichromate and sulfuric acid in reflex condensers, as per "Standard
Methods" (Ref. E-l), and the values were determined colorimetrically
using HACH (Ref. E-2) chemicals and a "spec-20" (Ref. E-3) colorimeter.
The samples were centrifuged before readings were taken to reduce back-
ground turbidity.
Total phosphates were determined by URSRC laboratory personnel, using the
standard (Ref. E-l) colorimetric procedure for ortho phosphates using
HACH (Ref. E-2) chemicals and a :'spec-20" (Ref. E-3) colorimeter. The
samples were boiled for 90 minutes in an acid mixture to convert the
polyphosphates to ortho phosphate, and centrifuged, before analysis.
Again, the liquid samples were handled normally, and a slurry was made
from the dry samples.
Total and fecal coliform counts were determined by URSRC personnel,
using the standard (Ref. E-l) (as of 1971) membrane technique. Millipore
apparatus was used along with their Endo and MFC broths for the total and
fecal coliform determinations, respectively. The liquid samples were
determined in the usual way, by passing a known amount of the sample through
the membrane filter. The solid samples were made into a liquid slurry of
215
-------
known concentration, and mixed in a high speed blender. Several different
"dilutions" were made of each sample in order to get the coliform density
reading within the desired range.
The total solids were determined by URSRC personnel. The liquid sample
from hosing each test area was analyzed for solids as per "standard Me-
thods" (Ref. E-l) and computed as amount of solids per unit area. The
solids swept were weighed and added to the amount obtained from the liquid
analysis to get the total solids for the test area. The solids were also
ashed in a muffle furnace according to "standard Methods" (Ref. E-l) to
determine the volatile fraction.
The time lapse between collection of the samples and running the tests
was kept to a minimum. The coliform determination of the liquid samples
was made the evening the samples were collected, while the "solid"
coliform counts were made immediately upon the return of the testing
personnel from the field trips. All other tests were also run as close to
the receiving date of the material as possible.
The results of the tests were all handled in a similar fashion: All
tests were run on liquid (hosed) and solid (swept) samples. Results were
reported as ppm of solids or as mg/f of the liquid. These values were
converted to amount of material per sq ft hosed and swept, which were then
combined to total amount of pollutant per sq ft. By knowing the street
and test site width, total amount of pollutant per curb mile was determined,
The strength of the solids in terms of each pollutant was determined by
dividing the total amount of pollutant per sq ft by the total solids
per sq ft.
Table E-l lists the pollutants, methods used for analysis, and test
samples analyzed for these pollutants.
216
-------
Table E-l
SAMPLES TESTED FOR EACH POLLUTANT
Annlytical
Method
Samples Tested
Lead, nlcl
chromium
Mercury
Cidnium
Soluble Nitrates
KJeldahl Nitrogen
Phosphates
BOD
5
Atomic Absorptioi
Atomic Absorption
Tentative Method
Standard Method
Standard Colorimeter Method
Standard 5-day Bottle
Method
R, I and C of SJ-1, Mil and Bolt. City composites of SJ-2, Atl, Tul, Pho-2,
840 2000 microns, and > 2000 microns.
same as lead, except R, I and C on SJ-1 only.
K, 1 and C of SJ-1, Mil and Bait.
Nine land-use composites. Ten city composites. SJ-1, Northern, Southern,
43 104 microns, 104 246 microns, 246 840 microns, 840 2000 microns,
and >2000 microns.
Same as Nitrates, except size analysis on SJ-1 and Northern composites only
Same as Nitrates.
COD Dichromate-Colorimetric Same as Nitrates.
Chlorinated Hydrocarbons Gas Chromatography R, I and C of SJ-1, Mil and Bait. Northern, Southern and Western composites
840 2000 microns.
Organic Phosphates Gas Chromatography SJ-1 city composite
Total Colifonns Membrane filter. Standard Every land use tested in each of the ten cities.
Fecal Coliforms Method (as of 1971)
Total Solids Standard Method Same as col i forms, plus the analysis of all ten cities ( < 43 microns,
43 104 microns , 104 246 microns , 246 840 microns , 840 2000 microns ,
> 2000 microns.
Volatile Solids Standard Method Nine land-use composites, ten city composites, plus on standard size ranges
of Pho-1 , Northern composite, SJ-2, Southern composite, and Seattle.
Kate: List of abbreviations used in table are as follows:
H: residential major land-use composite
I: Industrial major land-use composite
C: commercial major land-use composite
Southern composite: mode up of Atlanta, Tulsa and Phoenlx-2
Northern composite: made up of Milwaukee, Bucyrug and Baltimore
SJ-1: San Jose, winter test
SJ-2: San Jose, summer test
Pho-1: Phoenix, winter test
Pho-2: Phoenix, summer test
1111: Mil aukee
Bait: Ba timore
Sue: Buc rus
Atl: Atl nta
Tu: Tuls
Sea: Sea tie
Standard Methods for the Examination of Water amd Wastewater. American Public Health Association, 12 ed , 1965. (13 ed also, 1971).
217
-------
Appendix E
References
E-l. Standard Methods for the Examination of Water and Wastewater,
American Public Health Association, 12th ed. , 1965.
E-2. HACK Chemical Company, Ames, Iowa.
E-3. Product of Bausch and Lomb, New York, N.Y.
218
-------
Appendix F
QUESTIONNAIRE
This questionnaire was patterned after one
utilized by the American Public Works
Association in a similar survey.
-------
Appendix F
QUESTIONNAIRE
Street Cleaning Practices Utilized in City
GENERAL
Streets
Concrete
Asphalt
Other
Paved alleys
Condition (miles)
Good
Fair
Poor
>
Gutter (%)
Std.
Round
None
Swept
Other agencies responsible for street cleaning
Contractor
County/State
Name of responsible city department
Head URS Contact
PERSONNEL
No. supervisory personnel
Average salary
No. nonsupervisory personnel
Operators
, Nonoperators
Salary
Salary
219
-------
PROCEDURES
Frequency of cleaning (by neighborhood)
AREA
Downtown
Industrial
Arterial
Market
Residential
Lo income
Hi income
Frequency
Sweeper
Flusher
Hand
Time
Day
Night
Parking
Controls?
Extent of cleaning (by neighborhood)
Downtown
Industrial
Arterial
Market
Residential
Lo income
.Hi income
Route miles
Sweeper
Flusher
Hand
220
-------
EQUIPMENT
Number
Expected
Life
% Down
Time
Sweepers
# 1
Specify # 2
Type # 3
# 4
Specify
Type
Flushers
T, r*\r r*
J. M v>lXkJ
Specify
Type
Loaders
Specify
Type
Eductors
Specify
Type
Specify
Others
221
-------
SWEEPERS Operating Specs
Average operating speed , mph
Main broom fiber
Main broom strike
Main broom pressure
Main broom life
Main broom rotation speed
Gutter broom material
Gutter broom life
Hopper size
# 1
# 2
# 3
# 4
Sweeper performance by neighborhood
Downtown
Industrial
Arterial
Market
Residentia
Lo income
Hi income
Length
Shift
-
Curb
Mi per
Shift
Water
per
Shift
Gal
Cu yd
per
Shift
Crew
Size
per
Shift
Type debris (wt-%)
Wood/
paper
Glass/
metal
Dirt/
dust
Other
222
-------
Collection of sweeper debris
No. of trucks: Loaders
How covered:
Average size of pickup (yds)T
Ultimate dump site:
SPECIAL PROBLEMS
Leaves
Schedule
Method
Equipment
Quantity (per season)
rv-; ~« ~i
ui. ^ jj\j £y aj-
Dead Animals (if pertinent)
Abandoned cars (if pertinent)
Chemical Use (by city)
Quantity each season: Herbicides
Pesticides/insect icicles
Fertilizers
223
-------
Snow
Amount of chemicals used each season
Type chemical(s)
Amount of sand/cinders used each season
Snow disposal site
Estimated chemical load in snow disposed of
Is spring cleanup scheduled ?
Procedure:
Disposal site for spring cleanup debris;
Catch basins
Total number
Times oiled per gear
Times cleaned per year
Method of cleaning :
Hand
Bucket
Eductor
AVERAGE
No. of
Crew
Man/
Crew
Cost ,
$/ basin
Cost,
$/cu yd
Disposal
Site
^X^
224
-------
Sidewalks/parking lots
Frequency and procedure.
Extent of area involved:
Refuse Collection
Downtc-'-n
Industrial
Arterial
Market
Residential
Lo income
Hi income
Frequency
Colle
Curb
sction
Backyard
Contract
Litter
Baskets
Annual
Amount
Source of miscellaneous litter and debris:
Spillage from trucks
Litter from parades
Disintegration of
streets
Yard refuse
Animal droppings
Wind pockets/storms
Trash receptacles
Demolition
Streetside dumping
Street construction
Poor refuse collection
practices
Street trees
Lack of catch basin
inlets
Air pollution
Droppings from vehicle:
Downtown
Industrial
Arterial
Market
Residential
Hi
Lo
Rate Importance as: H (high), M (medium) or L (low)
225
-------
COST
Hand cleaning
Motor sweeping
Flushing
Disposal of
sweepings
Snow removal
Leaf removal
Dead animals
Catch basin clng.
Other
Total
Yearly Costs , $
Labor
Equipment
Other
Total
Cost per unit cleaned , $
Hand cleaning
Motor sweeping
Flushing
Catch basins
(curb mile)
(curb mile)
(mile)
(per catch basin)
226
-------
EFFECTIVENESS
In your judgement what is the effectiveness of your street
cleaning practices?
Good
Fair
Poor
Large
Visible
Dirt &.
Dust
Glass &
Metal
Organic
Sweeper Operations by Month
January
February
March .
April
May
June
July
August
Loads*
September
October
November
December
Curb-Miles
*Specify conversion factor to cubic yards
227
-------
Appendix G
STREET SURFACE CONTAMINANT SIMULANT
-------
Appendix G
STREET SURFACE CONTAMINANT SIMULANT
In developing standardized equipment and/or cleaning practice evaluation
methods, a street surface contaminant simulant was prepared for use in
test programs. This approach was used to enhance the reliability of
evaluating equipment parameters and various practices under standardized
street contamination conditions.
Requirements for such a simulant have been long recognized. In Operation
STREETSWEEP (Ref.G-l), ferromagnetic particles of two size ranges were
utilized to simulate weapons fallout particles; however, it became appar-
ent that the use of this simulant was inappropriate for ordinary street
dirt. In Operation SUPERSWEEP (Ref .G-2) three sizes of racliot ant alum-
tagged particles were utilized to simulate fallout particles. In both of
these experiments it was shown that small particles were the most diffi-
cult to remove. In operation STONEMAN (Ref.0-3) and STONEMAN II (Ref.Q-4)
a simulant covering a broad particle-size range was used. This simulant
consisted of a loam-type soil and several grades of sandblasting sand.
Results from the street surface sampling program in this URS study indi-
cate that the bulk of street surface contaminants is made up of particles
ranging from 43 to 2000 microns. Figure G-l shows the cumulative particle
size distribution derived from the sampling program in the cities of San
Jose and Phoenix. Since the distribution of particle sizes was found to
be fairly similar in each city, the particle size distribution selected
for the simulant to be utilized in the controlled street sweeping evalua-
tion tests was made an average of the two cities (shown by the dotted line
in Figure G-l). The composition by weight of the simulant for each size
range is given in Table G-l.
The synthetic simulant represents the dust, dirt and gravel fraction of the
street litter. An average of 92 percent by weight of the street litter
collected in the sampling program passed through a 200 micron screen
(10 mesh) and was mainly composed of dust, dirt, sand and gravel.
The material used for the simulant consisted of two grades of commercially
available Del Monte Sand, #60 mesh and #1 ground. These two grades of
river bottom sand contain a large percentage of the size fractions found
in the street surface samples. The sieve analysis for these two grades are
in Table G-2. The particles were separated into the required size
ranges on a commercial sieving machine (manufactured by Novo Corporation).
This machine (a vibratory type) feeds the raw material from a storage
hopper onto a screen where two fractions are obtained; one greater and
one smaller than the screen mesh opening. Selected screens are used to
produce the various size ranges required.
229
-------
5
x
UJ
z
PARTICLE SIZE (M)
Fig. G-l. Simulant Compared to Street Surface
Contaminant Samples, Tests B and C
Table G-l
SIMULANT PROPERTIES
Particle Size
2000
840-200
246-840
104-246
43-104
43
Composition
(% by weight)
8
20
30
20
16
6
Several hundred pounds of each
size range were produced for
use as the street surface con-
taminant simulant.
The simulant was formulated by
combining the desired weights
(Table G-l) of each particle
size group. If necessary, the
simulant combination can be
changed by varying the weight
composition of each particle size
range. Some important geometri-
cal properties of the various
size ranges are given in Table
G-3.
230
-------
Table G-2
ANALYSIS OF SANDS USED IN FORMULATING SIMULANT
No. 1
Ground
Del Monte
Sand
Nominal
60 mesh
Ground
Del Monte
Sand
Mesh
Opening
fo)
297
177
149
74
53
43
43
420
297
250
177
149
105
105
Percent
Re-
tained
3.8
20.8
13.4
32.2
12.2
7.2
10.5
8.2
48.4
21 .3
19.9
1.3
0.9
0
Cumulative Percent
Less Than
Stated Size
96.3
75.5
62.1
29.9
17.7
10.5
91.8
43.4
22.1
2.2
0.9
Table G-3
GEOMETRIC PROPERTIES OF PARTICLES
Size
Range
(M)
43- 88
88-175
175-350
350-750
Number of
Particles
(per gram)
6.98 x 106
6.20 x 105
7.54 x 104
7.09 x 102
Average
Surface Area
per Particle
(cm2)
6.69 x 10"5
3.21 x 10~4
1 .39 x 10"3
7.85 x 10~3
Average
Particle
Diameter
(At)
47
101
210
500
231
-------
Appendix G
References
G-l. R. A. Laughlin, J. Howell, et al. Operation STREETSWEEP.
Naval Radiological Defense Laboratory Report ADX-39,
2 December 1948.
G-2. F. R. Holden, R. A. Laughlin, et al. Operation SUPERSWEEP.
Naval Radiological Defense Laboratory Report ADZ-42,
4 October 1948.
G-3. J. D. Sartor, H. B. Curtis, et al. Cost and Effectiveness
of Decontamination Procedures for Land Targets. Naval
Radiological Defense Laboratory Report USNRDL-TR-196,
27 December 1957.
G-4. H. Lee, J. D. Sartor and W. H. Van Horn. Performance
Characteristics of Dry Decontamination Procedures. Naval
RadiolocaY De'fense Laboratory Report, USNRDL-TR-336, 6 June 1959.
232
-------
Appendix H
CATCH BASIN TEST PROCEDURES
-------
Appendix H
CATCH BASIN TEST PROCEDURES
Determination of the changes in street runoff resulting from passage
through a catch basin followed two approaches. The first of these in-
volved investigation of the hydraulic flushing effect of inlet water on
antecedent basin contents. This essentially consisted of discharging
fresh water into dirty catch basins under controlled conditions and samp-
ling the effluent. The second approach was the determination of the
solids removal effectiveness of catch basins. This was investigated by
discharging water and street litter simulant into a clean catch basin
and sampling the effluent.
The sampling site (in San Francisco at 40th and Moraga Avenues) is be-
lieved to be as typical an urban catch basin installation as might be
found. It contains three similar catch basins, all draining into a cen-
tral interceptor, which in turn drains into the sewer system (actually a
combined storm/sanitary sewer system). The installation is shown in
Fig. H-l. The central interceptor allowed a common sampling point for all
AVENUE
40* AVENUE
Fig. H-l. Map of Test Site for Catch Basin Tests (San Francisco)
233
-------
the tests. The intersection is on a slight grade, allowing rapid and
efficient drainage of water and sediment and an accurate determination of
individual drainage areas. The catch basins are of a standard, rather
conventional design: concrete with curb inlets and cast iron gratings
(a cross section is shown in Figure H-2).
Computation of a number of
factors was necessary prior to
the start of testing. The
amount of runoff for each
of the three drainage areas
was calculated at several
rainfall intensities. These
amounts were then converted to
flow rates (cubic feet per
second) to allow use of a
flow meter to regulate the
supplying fire hydrant. Using
previously collected street
loading data for the land
use encountered (i.e., resi-
dential) , the amount of ma-
terial that would be expected
to wash off the street was
determined. The amount of
water for each rainfall inten-
sity , and the drainage basin
loadings are shown in Table
H-l. Using this information
(along with information on
the rate at which rain removes
street surface loadings derived
from previous studies) , the
percentages of the total
loading washed off the street
introduction amounts and times
Fig. H-2.
Cross-Section Through
Catch Basin
were calculated on a time basis and sample
were established (as shown in Table H-2).
Table H-l
PARAMETERS USED IN CATCH BASIN STUDY
CATCH BASIN
DRAINAGE
AREA
(sq ft)
RAINFALL
INTENSITY
(in./hr)
FLOW RATE
INTO
CATCH BASIN
(gal/min)
SOLIDS INPUT RATE
at 2 g/sq ft
(lb/40 min)
A
B
C
41,700
25,000
11,050
0.07
0.49
0.72
32.1
126.1
82.2
144
88
38
234
-------
Table H-2
QUANTITIES OF SIMULANT USED
TIME SINCE AMOUNT OF SIMULANT MIXED WITH FLOW
BEGINNING INTO EACH CATCH BASIN (Ib)
OF TEST A B ' C
(min)
36 22 9-1/2
2
18 11 5
5
27 16-1/2 7
6
27 16-1/2 7
20
36 22 9-1/2
40
Total in 40-
min Intervals 144 88 38
The actual sampling procedure consisted of starting the flow of water
into a catch basin and then lowering a bucket into the central interceptor
to take a series of samples (one at each predetermined time interval). For
example, during the time interval 0 to 1 min, approximately five 1-gal
samples were collected and composited on the spot in the interceptor.
Other samples were taken in a similar manner for the intervals 1 to 2
min, 2 to 5 min, 5 to 10 min, and 10 to 20 min. In the laboratory, total
settleable solids and dissolved and suspended solids were determined for
all samples. In addition, samples from the "clean" catch basin tests and
the representative samples of the street simulant were analyzed by dry
sieving to determine size distribution.
235
-------
DIRTY CATCH BASINS
Clean water was introduced to catch basins A, B, and C (one at a time,
in three separate tests) at measured flow rates of 0.29, 0.51, and 0.81
in./hr, respectively. Sample compositing intervals were 2, 5, 10, and
20 min in Test A, with 1 min sample added in Test B and a 40 min sample
added in Test C. The measured output of each catch basin and the flow rate
are shown in Table H-3. The results indicate that most of the material
originally contained in a catch basin tends to remain there, regardless
of runoff flowing through it.
Table H-3
TEST OF "DIRTY" CATCH BASINS
CATCH
BASIN
CATCHMENT
AREA SIZE
(ft)
INFLOW RATE
IN TERMS OF
EQUIVALENT
RAINFALL
OVER CATCH-
MENT AREA
(in./hr)
WEIGHT OF
SOLIDS IN
CATCH BASIN
AT OUTSET
(Ib)
WEIGHT OF
SOLIDS
REMOVED BY
"STORM"
(Ib)
SOLIDS
REMOVAL
EFFICIENCY
(% by weight
of original)
A 41,700
B 25,000
C 11,050
0.29
0.51
0.81
2,047
2,559
3,481
26.6
30.0
21.6
1.2
1.1
0.6
Observed, but not measured in the sampling, was the general quality of
the water above and mixed with the sediments. Before testing, the non-
settleable contents were septic with an observable percentage of plant
material, oil and grease. Initial flows into the catch basin removed this
very quickly.
CLEAN CATCH BASINS
Catch Basin B was cleaned out by a combination of water-jet blasting and
wet vacuuming for this series of tests. The simulant material was intro-
duced by dumping a pre-weighed amount of material (pre-calculated for
each time interval) into the gutter several feet above the catch basin
grate and gradually eroding it with water supplied by a fire hydrant.
This proved to be an accurate method of introducing solids uniformly.
Sampling at the central interceptor was accomplished as described for the
dirty catch basin tests.
236
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
7 R..-,JOI-t No.
w
Title
Water Pollution Aspects of Street Surface Contaminants
EPA - R2 - 72 - 081
.'hnt(sj
Sartor. James P.. Boyd, Gail B.
i. Or?*r" atio"
URS Research Company
155 Bovet Road
San Mato, California 94402
12. Sfnsonv org»w ^tion Environmental Protection Agency OR&M
Environmental Protection Agency report
number EPA-R2-72-081, November 1972.
5. l\aport])ate
(j.
S. .f rivrmi ,£Org: matron
Report No.
10. Project Ho.
11. Contract/Grant No.
I.' Typ'. ' Rep(' . and
Period Coveted
16.
Materials which commonly reside on street surfaces have been found to contribute
substantially to urban pollution when washed into receiving waters by storm run-
off. In fact, runoff from street surfaces is similar in many respects to sanitary
sewage. Calculations based on a hypothetical but typical U.S. city indicated
that the runoff from the first hour of a moderate-to-heavy storm would contribute
considerably more pollutional load than would the same city's sanitary sewage
during the same period of time.
This study provides a basis for evaluating the significance of this source of
water pollution relative to other pollution sources and provides information
for communities having a broad range of sizes, geographical locales, and public
works practices. Information was developed for major land-use areas within the
cities (such as residential, commercial and industrial). Runoff was analyzed for
the following pollutants: BOD, COD, total and volatile solids, Kjeldahl nitrogen,
nitrates, phosphates, and a range of pesticides and heavy metals.
173, Descriptors
*Storm Runoff, Surface runoff, Urban runoff, *Pollution (water), BOD, COD,
solids, heavy metals
Ifb. Identifiers
*Street cleaning *Street surface contaminants
.'7c COWRR Field & Group
IS.
Abs
Availability
factor
19. Security Class.
10. Se
rity C< iss.
21. No. ot !i
Pages 1
-2. *,*'.,'
Send To :
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 2O24O
| Institution
------- |