DEMONSTRATION OF
NON-POINT POLLUTION MANAGEMENT
ON CASTRO VALLEY CREEK
First Annual Report
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
WATER PLANNING DIVISION
WASHINGTON, D.C, 20460
NOVEMBER 1979
BY
ALAMEDA COUNTY FLOOD CONTROL AND WATER CONSERVATION DISTRICT
HAYWARD, CALIFORNIA 94544
-------
27942
DEMONSTRATION OF
NON-POINT POLLUTION MANAGEMENT
ON CASTRO VALLEY CREEK
First Annual Report
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
WATER PLANNING DIVISION
WASHINGTON, D.C. 20460
NOVEMBER 197 9
WRITTEN BY
GARY SHAWLEY AND ROBERT PITT
H.A. FLERTZHEIM, JR., DIRECTOR OF PUBLIC WORKS, ALAMEDA COUNTY
PAUL E. LANFERMAN, ENGINEER-MANAGER
ALAMEDA COUNTY FLOOD CONTROL AND WATER CONSERVATION DISTRICT
HAYWARD, CA 94 544
GARY SHAWLEY, PROJECT MANAGER
ALAMEDA COUNTY FLOOD CONTROL
AND WATER CONSERVATION DISTRICT
ROBERT PITT
PRINCIPAL INVESTIGATOR
PRIVATE CONSULTANT
-------
DISCLAIMER
This document has not been formally released by the Environmental
Protection Agency (EPA) and does not represent agency policy. It is being
circulated for comment on its technical accuracy and policy implications.
-------
ABSTRACT
This annual report presents the results and preliminary conclusions
from the first year of study in the EPA-sponsored non-point pollution manage-
ment project on Castro Valley Creek. As part of the San Francisco Bay Area's
208 Continuing Planning Program, this project is the first prototype project
that is part of the EPA's National Urban Runoff Program. This is the first
research project designed to correlate street cleaning and receiving water
quality.
The study area is a 1,542-acre watershed that was divided into three
test areas for street cleaning and sampling convenience and to develop
information about topographical affects on accumulation rates of the pollu-
tants associated with street dirt. To demonstrate the relationship between
street cleaning and receiving water quality, the project is designed to
measure the following: (1) street cleaning effectiveness to identify the
quantity of pollutants removed and the initial and residual loadings before
and after cleaning; (2) weekly street surface pollutant loadings to identify
the loading on the streets at any time, and (3) receiving water quality to
know the quantity of pollutants washed off the watershed for various types of
rainstorms. Curve fitting analysis was used to correlate street surface
pollutant loadings before rain events with changes in receiving water pollu-
tant mass yields.
Although only a small quantity of receiving water data was obtained
during the first year, it was determined that street cleaning could signifi-
cantly affect the quality of receiving waters. Forty percent of the total
solids, lead, and zinc in urban runoff could be removed by changing from
monthly to weekly street cleaning schedules. The conclusions drawn from the
first year of study are based on limited data and conclusions in the final
report may be significantly different based on a larger sample population.
The final results will identify the role that street cleaning should pursue
to meet the goals of the 1972 Clean Water Act.
-------
CONTENTS
Page
Disclaimer
Abstract
List of Figures
List of Tables
Section 1. Conclusions 1
Section 2. Introduction 4
Section 3. Description of Study Area 6
Land Use 6
Section 4. Tests of Castro Valley Street Cleaning Equipment 13
Results of Year One Experimental Design 13
Street Cleaning Tests 14
Sources of Urban Runoff Pollutants 16
Street Surface Contaminants Accumulation Rates 25
Street Cleaning Demonstration Results 26
Section 5. Castro Valley Creek Monitoring 37
Analytical Program 41
Pollutant Removal Capabilities of Monitored Storms 44
Effectiveness of Street Cleaning in Improving
Receiving Water Quality 46
Section 6. References 55
Section 7. Appendix 56
-------
LIST OF FIGURES
Number Page
1. Improvement in Receiving Water Lead Yield as a Function
of Street Cleaning Effort 2
2. San Francisco Bay Area Showing the General Location of
the Castro Valley Watershed 7
3. Aerial View of Castro Valley 8
4. Castro Valley Typical Land Use Photographs 9
5. Study Area Divisions 10
6. Sawtooth Pattern Associated with Deposition and Removal
of Particulates 14
7. Deposition and Accumulation Curves - Lower Test Area 27
8. Deposition and Accumulation Curves - Middle Test Area 28
9. Deposition and Accumulation Curves - Upper Test Area 29
10. Street Dirt Removed as a Function of the Street
Cleaning Effort 33
11. Initial Street Loading as a Function of the
Street Cleaning Effort 34
12. Residual Street Loading as Function of the
Street Cleaning Effort 35
13. Runoff Flows at Monitoring Station Sites 38
14. Total Solids and COD Creek Yields for Storms of
Small Rainfall Quantities and Long Accumulation Periods 47
15. Total Solids and COD Creek Yields for Storms of
Large Rainfall Quantities and Short Accumulation Periods 48
16. Lead, Zinc, Chromium, Orthophosphate and Total Phosphorus Creek
Yields for Storms of Small Rainfall Quantities and Long
Accumulation Periods 50
17. Lead, Zinc, Chromium, Orthophosphate and Total Phosphorus Creek
Yields for Storms of Large Rainfall Quantities and Short
Accumulation Periods 51
18. Arsenic, Copper and Kjeldahl Nitrogen Creek Flows for
Various Street Surface Loading Values 52
-------
LIST OF TABLES
Number Page
1. Test Area Characteristics 11
2. Street Cleaning Schedule 15
3. Urban Runoff Pollutant Concentrations from Major Areas 17
4. Potential Significant Urban Runoff Pollutant Sources 19
5. Major Urban Areas and Delivery Yields to Outfall 20
6. Contribution from Various Areas to Outfall Runoff Yields 22
7. Relative Source Annual Depositions 23
8. Relative Runoff Yield Contributions 24
9. Castro Valley Total Solids Accumulation Rates 30
10. Street Cleaning Effectiveness 31
11. Castro Valley Rain Events 39
12. Significant Rains During First Year Field Activities 40
13. Summary Castro Valley 1978-1979 Monitored Rain Year 42
14. Parameter List for Receiving Water Analysis 43
15. Observed Street and Creek Yields 45
16. Estimated Effectiveness of Street Cleaning in
Improving Receiving Water Quality 53
-------
SECTION 1
CONCLUSIONS
Research funded by the Environmental Protection Agency (EPA) is being
conducted to determine whether street cleaning activities can significantly
affect non-point source water pollution.
With the limited receiving water data obtained in the first year of this
study, the following conclusions have been reached.
1. Street cleaning can significantly affect the quality of receiving
water. The amounts of lead, total solids, and zinc in storm mass yields can
be reduced 40 percent by changing from monthly to weekly street cleaning
schedules (see Figure 1). If street cleaning was increased from a "no
effort" to a "maximum effort" of three times per week, these pollutants could
be reduced by as much as 65 percent.
2. Street cleaning could be an important factor in affecting the
quality of receiving water for rain events that normally occur in the fall.
These rain events are characterized by long accumulation intervals between
rains and small rainfall quantities.
3. If an area is cleaned weekly it would be effective to maintain
that schedule and start weekly cleaning in other areas that are cleaned
monthly or less frequently (e.g., industrial areas). Similarly, if an area
is cleaned more than three times per week then the cleaning effort should be
reduced to either twice or once per week (See Figure 1) and other areas
should be cleaned using the new resources. In these cases, street cleaning
would be better applied to other areas because after a certain effort
additional cleaning is wasted. It is much more efficient in terms of
reduced annual loadings to receiving waters to apply additional street
cleaning effort to areas that have high street surface loading values rather,
than areas with low loading values.
1
-------
FIGURE: I
CHANGES IN KCCfclV/ING WATER LEAP YlfrLP
Ai A FUNCTION Of 6TRWT CLCAMJW6 EWORT
+50 - -
9
tu
V
Ck
2
o-a-® o-e-o
Q J " " * (VfrO O •
o a o -»-c o-» o-®-o o-o-a-
I
i
a
i
z
a
i
5
a
a!
-50
- ioo -.
-150"
CURRENTLY PO MOT CLEAN STREETS
CURRENTLY CLEAN STREETS QUARTERLY
o o o o o o o CURRENTLY CLCAKl STREETS MONTHLY
CURRENTLY CLEAN STREETS WCEKLY
CURRENTLY CLEAN STREETS 5 TIMES OR MORE PER WEEK
-200 '
+
-+-
-H
50
-4-
"1 ' ifc ' do ' ilo
20
I- MONTHLY
>— QUARTERLY
40
60
'-WEEKLY 3 TIMES/WEEK
PROPOSE? NUM&ER OF STREET CLEANING PASSES PER YEAS
-------
4. In the California coastal climate category, the first significant
rain of the year usually occurs in September or October and can account for
almost one-half of the total annual pollutant runoff from street surfaces.
That is, one-half of the total annual street surface pollutant load. In
contrast, January and February runoff contains large amounts of erosion
from areas adjacent to street surfaces and are not as sensitive to street
surface contaminant loading. This conclusion illustrates the need for resched-
uling of street cleaning programs. Concentrated street cleaning before the
first significant rain and less during the winter could reduce the amount of
annual urban runoff loading. This relationship between street surface
loading and street cleaning schedules could be optimized to protect receiving
waters.
5. Under certain storm conditions, pollutant deposition in the storm
sewerage system (i.e., inlets and laterals) may be an important component in
the total urban runoff mass balance equation. Another important component
which remains to be quantified for this mass balance equation is the loading
from adjacent land areas.
3
-------
SECTION 2
INTRODUCTION
In order for the reader to better understand the relationships between
street surface pollutant loading, street cleaning and receiving water quality,
a simplified conceptual model is offered. This urban runoff mass balance
model consists of the following two equations:
1. Receiving water quality is defined as the pollution load in the
receiving water. Therefore, receiving water quality when it is not
influenced by "point sources" is a function of the pollution load
from stream baseflow plus surface runoff plus aerial fallout.
2. Surface runoff quality is a function of the pollution load from
street surfaces plus adjacent land areas plus storm sewerage.
The purpose of this project is to determine whether removing the pollution
load from the street by street cleaning has an affect on surface runoff
quality from street surfaces and consequently an affect on receiving water
quality. Throughout the project, data will be collected to compare the
monitored mass pollutant flows of the storms with the total pollutant removal
of street cleaning programs. A valid data set for this analysis consists of
a data point for a monitored runoff event that occurs between adjacent street
surface monitoring data points. During the first year of this study,
eight valid data sets were collected.
Data will also be collected during the second year to develop adequate
street cleaning programs, determine street loading values, and monitor rain
events. A large number of these data points could demonstrate that street
cleaning significantly changes receiving water concentrations as well as
storm mass yields.
Demonstrating that a reduction in the quantity of street surface con-
taminants improves the quality of receiving waters may have strong policy
implications. Most cities have street cleaning programs and the financial
resources devoted to those programs are significant portions of local public
works departments operating expenses. As local decision-makers balance
priorities with revenues, street cleaning for aesthetic purposes is usually
relegated a low priority. Street cleaning was a candidate 208 control
measure in the San Francisco Bay Area Environmental Management Plan, and
local decision-makers are looking forward to the results of this project. If
it can be demonstrated that street cleaning has a positive affect on water
quality, then decision-makers may have more reason to give their street
cleaning budgets higher priority. A rescheduling of cleaning effort within
the existing budgets in order to maximize water quality benefits may be in
order. The water quality benefits of street cleaning must be proven in these
times of serious fiscal constraints.
4
-------
In 1971, the Castro Valley Watershed was selected by the U.S. Geological
Survey (USGS) and the Corps of Engineers (CORPS) (Hydrologic Engineering
Center) for urban runoff monitoring. This watershed was selected because it
was considered to be typical of residential basins in the San Francisco
Bay-Delta region (Sylvester and Brown, 1978). The main purposes of this early
federal effort were to (1) determine relationships between land use and land
surface characteristics and the quantity and quality of urban runoff, and (2)
determine whether control of urban runoff is necessary to meet regional water
quality objectives.
Castro Valley Creek has few beneficial uses (as defined in the San
Francisco Bay Basin Plan) and none have been officially established by the
state. However, the Castro Valley Basin probably has more data about urban
runoff than any other basin in California.
Many agencies have continued to monitor urban runoff since 1971. To
date, portions of about 45 storms have been monitored. Two continuous
recording stream gages and two automatic ISCO water samplers have been
installed by the USGS and the Alameda County Flood Control and Water
Conservation District (ACFC&WCD) in Castro Valley Creek. The upper station
(USGS No. 11181004) is located at Seaview Street and measures the contrib-
utions from the relatively undeveloped portions of the study area. The lower
station (USGS No. 11181006) is located at Knox Street and in addition to the
upper undeveloped portions of the study area, measures the contributions from
the urban areas. The CORPS has been monitoring another station at the
confluence of Chabot and Castro Valley Creek.
San Lorenzo Creek, downstream from the confluence of Castro Valley and
Chabot creeks, is a large watercourse with contiguous urban development.
This creek carries the flow to its discharge point in San Francisco Bay.
The ACFC&WCD is the lead agency in this project and the Association
of Bay Area Governments (ABAG) is the project sponsor. The USGS , the CORPS,
and the State Regional Water Quality Control Board (RWQCB) are responsible
for technical review assistance. The Project Manager, Gary Shawley, is a
research scientist and private consultant under contract to the ACFC&WCD, and
the Principal Investigator, Robert Pitt, is an environmental engineer and
private consultant to Woodward-Clyde Consultants.
5
-------
SECTION 3
DESCRIPTION OF STUDY AREA
The Castro Valley Watershed covers approximately 5.6 square miles
and is located within the San Francisco Bay Area (Figure 2). An aerial view
of the watershed is presented in Figure 3. The Castro Valley Creek branch of
the Castro Valley Watershed was selected as the study area to reduce the
study area to a more manageable size (from 5.6 square miles to 2.4 square
mi les).
The study area is predominantly residential. The majority of the
residential land use consists of single family housing with lot sizes varying
from 5,000 square feet to 10,000 square feet. The estimated residential
population density is about 20 people per acre. Residential land use occupies
2,464 acres (70 percent), commercial land use occupies about 246 acres (7
percent), and the remaining land is open space (809 acres or 23 percent).
Development along the stream banks in Castro Valley is intense and houses are
frequently constructed directly over the existing streambed. Some light
commercial areas, more than a dozen schools, and a short portion of Interstate
Highway 580 are also in the area. Photographs of typical Castro Valley land
uses are shown in Figure 4 A, B, and C.
Topography within the drainage basin is highly variable, and the land
slopes range from 10 percent to 70 percent in the upper end of the basin to
slopes as low as 1 percent in the valley portion near San Lorenzo Creek. The
Castro Valley Creek streambed in the lower portions of the drainage basin
range from 20 feet to 50 feet in width and 8 feet to 10 feet in depth. The
streambed is often strewn with litter and debris.
The study area was divided into four sub-areas (Figure 5). These
horizontal divisions across the watershed (based on topography and street
patterns) increased the amount of useful data obtained from the street
surface monitoring activities.
There are many similarities in the three lower urban sub-areas (see
Table 1). The most important of these similarities are the types of gutters
and the condition of the street surfaces. Seventy-five percent of the
gutters are concrete and 25 percent are asphalt. The shapes of the curbs
(straight or rolled) influence the amounts of street surface contaminants
that remain on the streets (and are available to street cleaners), and the
amounts that are transported to the shoulder of the road (and are not
available for pickup in normal street cleaning operations). The condition of
street surfaces determines the nature of contaminants and the performance of
street cleaning equipment. In the study area, 91 percent of street surfaces
in the lower three urban test areas are in fair condition, with little
variability in condition or width (95 percent of the streets in these sub-
areas are 20 feet to 40 feet wide).
The variable that may significantly influence the quantity of nutrients
removed by street cleaning operations is the amount of leaf material on the
streets. In the study area, the largest accumulation of leaves on streets is
in the upper urban sub-area, but the difference among the three areas is not
large.
6
-------
OS 10 15
Milei
FIGURE 2. SAN FRANCISCO BAY AREA SHOWING THE GENERAL
LOCATION OF THE CASTRO VALLEY WATERSHED
7
-------
Figure 3
Aerial View of Castro Va ley
3
-------
-------
FIGURE 5. STUDY AREA DIVISIONS (CASTRO VALLEY CREEK-
DRAINAGE AREA - 2.41 SQ. MI.)
RURAl
AREA
6EAVIEW AVE. STREAM
GAGE STATION
.SCALE-'I*<2003
middle URBA)
SAMPLING ARE,
LOWER
J^RBAN
WPLINS
AREA
KNOX ST. STRTEAM-
6AGE STATION
fIC CENTER STRE
GAGE STATION
RAIN GAGE STATION
I
/ CASTRO VALLEY WATERSHED
y OUTSIDE STUDY AREA
RAIN GA6E STATION
(C.V. FIRE STATion}
RAIN GAGE STATION
CASTRO VALLEY WATERSHED
(CASTRO VALLEY £CHA&OT CREEKS)
COM&INCP PKAINAGC AR£A-5.59 SS-Ml-
*»°
•toOlS-
10
-------
TABLE 1. TEST AREA CHARACTERISTICS (PERCENT, UNLESS OTHERWISE NOTED)
DRAINAGE AREA =2.4 SQUARE MILES
LOWER
MIDDLE
UPPER
TOTAL
Number of Gutters
0
0
6
0
2
1
0
0
11
4
2
100
89
89
92
4
9
5
0
2
Gutter Type
Concrete
93
62
78
76
Asphalt
7
38
17
22
Mixed
0
0
5
2
Gutter Shape
Straight
0
0
5
2
Rolled
41
52
69
55
Mixed
59
48
26
43
Median Strip
Yes
0
5
0
2
No
100
95
100
98
Street Condition
Poor
0
0
3
1
Fair
74
97
97
91
Good
26
3
0
8
Street Width
20 to 40 feet
93
95
100
96
40 feet
7
5
0
4
Landscaping Type
Deciduous
78
74
46
65
Evergreen
11
5
0
5
Mixed
11
21
54
30
Landscaping Quantity
None
11
0
0
3
Some
85
89
91
89
Much
4
11
9
8
Leaves on Street
Few
81
74
43
65
Some
15
21
45
28
Much
4
5
12
7
Parking Density
None
4
2
0
2
Light
55
50
89
65
Moderate
37
42
11
30
Heavy
4
6
0
3
11
-------
TABLE 1. TEST AREA CHARACTERISTICS (PERCENT, UNLESS OTHERWISE NOTED)
DRAINAGE AREA =2.4 SQUARE MILES (CONCLUDED)
LOWER
MIDDLE
UPPER
TOTAL
Traffic Density
Light
79
63
77
72
Moderate
7
26
23
20
Heavy
14
11
0
8
Traffic Speed
25 MPH
52
39
40
43
25 - 40 MPH
48
58
60
56
40 MPH
0
3
0
1
Topography
Flat
66
39
3
34
SI ight
30
34
3
22
Moderate
4
11
37
18
Moderate/Steep
0
3
11
5
Steep
0
13
46
21
Adjacent Land Use
Low Income, Old, Single Family
28
13
0
12
Medium Income, Old,
Single Family
57
66
98
75
Commerci al
7
13
0
7
Multiple Family
4
2
0
2
Vacant
0
2
0
1
School
4
4
2
3
Street Surface Particulate Loadings
Mean (lbs/curb-mile)
369
675
563
552
Median (lbs/curb-mile)
321
536
467
432
Standard Deviation
(lbs/curb-mile)
219
540
363
425
Standard Deviation/Mean
(ratio)
0.59
0.80
0.64
0.77
Minimum(lbs/curb-mile)
48
100
82
48
Maximum(Ibs/curb-mile)
821
2970
1260
2970
Number of Valid Data Points
27
37
34
98
12
-------
SECTION 4
TESTS OF CASTRO VALLEY STREET CLEANING EQUIPMENT
The objectives of the street cleaning equipment performance
in Castro Valley were:
• To determine the accumulation rates of street surface contaminants.
• To determine the characteristics of street contaminants in relation
to particle sizes.
• To investigate various street cleaning practices under actual
field conditions (including various street surface conditions,
residual particulate loadings, traffic densities, parked cars,
and climatic conditions) to determine the range of possible
cleaning performances.
RESULTS OF YEAR ONE EXPERIMENTAL DESIGN
Street surface samples were collected from narrow strips that were
the width of the street. The analytical procedure used to determine the
number of sub-samples needed involved weighing individual sub-samples in the
study area to calculate the standard deviations (c) and the means (x) of
street surface loading values.
From these two values, the number of sub-samples necessary (N),
depending on the allowable error (L), were determined. An allowable error
value of about 25 percent, or less, was used.
The formula used (Cochran 1963) was:
With a 95 percent confidence limit, this formula determines the number of
samples needed to determine the true value for loading within a range of
+ L. Initially, individual samples were taken at 100 locations in the
three study areas to determine the loading variability. Loading varied
within the study area but the median values in the three test areas were
fairly close. The overall minimum loading was about 50 lbs/curb-mile, the
overall maximum value was about 3,000 lbs/curb-mile, and the overall median
value was about 400 lbs/curb-mile. The median values in the three areas were
about 320, 540, and 470 lbs/curb-mile.
Preliminary statistical analyses (using Student "t" tests) have shown
little loading variations between candidate sub-division groups. The topo-
graphical grouping was made for sampling convenience, as it was difficult to
sample the complete study area twice in a single day (as required for the
street cleaning tests). The relative variations of particulate loadings
within the three test areas did vary.
13
-------
STREET CLEANING TESTS
Several street cleaning programs using various levels-of-effort were
evaluated during the first year of study. The street cleaning program for
the first year is presented in Table 2. The first year of the testing
program was divided into five segments and different street cleaning programs
were conducted in each of the three urban test areas. The values in Table 2
for each week refer to the number of street cleaning equipment passes
conducted in each area during that week. An initial cleaning was conducted
during the first project week. The next week of the study involved monitor-
ing a typical leaf removal operation which consisted of continuous front-end-
loader and dump truck activity, 5 days a week. The street cleaning program
studies in following weeks varied from one to five passes every week.
From 6 to 15 weeks passed before the regular street cleaning tests began
so that street surface contaminants could accumulate. This street cleaning
schedule permitted street surface loading conditions to be evaluated under a
wide range of cleaning practices.
This information was used to evaluate the street surface particulate
accumulation rates and street cleaner performance. This information is
compared to Castro Valley Creek pollutant flows during storm events in the
next section of this report.
Street cleaning equipment performance is very sensitive to operator
skill and equipment maintenance. The equipment must be adjusted adequately,
and maintained and operated in a manner to optimize debris removal and
minimize cost. The operators and maintenance personnel used during these
tests were supplied by the Public Works Department of Alameda County.
They were all well trained, skilled and operated the test equipment in an
optimum and recommended manner.
Long-term and frequent street surface sampling in the test areas made
it possible to directly measure accumulation rates of street surface contam-
inants. Street surface samples were collected within a few hours before and
after street cleaning by the procedures described in the First Year Work
Plan. The idealized loading pattern resulting from sampling at these inter-
vals, a sawtooth pattern depicting the deposition and removal of street
surface particulates is illustrated in Figure 6. The pattern observed during
year one generlly agreed with Figure 6 (First Technical Progress Report on
Castro Valley). Accumulation rates for the various contaminants can be
determined by calculating the angle of the slope between adjacent sampling
periods.
FIGUtt £
sawtooth panted associate with
PfrPOSITlON AMP REMOVAL Of PARTICULATES
II
1
%
*3
J • Anarf W jtrcrt
* tarfac* lamftmq
wm ^ ^
MtaMUatm OT» '"^SUndual ft
a titan
TIMB-
14
-------
TABLE 2. STREET CLEANING SCHEDULE 1
5-Day
Upper
Mi d d 1 e
Lower
Work-
Weeks
Urban Area
Urban Area
Urban Area
11/20
11/24/78
1(2)
1(2)
1(2)
f "i \
11/27
12/1
4L<3)
4L<3)
4L'3'
12/4
12/8
0
0
0
12/11
12/15
0
0
0
12/18
12/22
0
0
0
12/25
12/29
0
0
0
1/1
1/5/79
0
0
0
1/8
1/12
0
0
0
1/15
1/19
0
1
0
1/2 2
1/26
0
1
0
1/29
2/2
0
1
0
2/5
2/9
0
1
0
2/12
2/16
0
0
0
2/19
2/23
0
0
0
2/26
3/2
3
0
0
3/5
3/9
0
0
0
3/12
3/16
. 0
5
0
3/19
3/23
0
0
1
3/26
3/30
0
0
1
4/2
4/6
0
0
1
4/9
4/13
0
0
1
4/16
4/20
0
0
1
4/23
4/27
0
0
1
4/30
5/4
1
0
0
5/7
5/11
1
0
0
5/14
5/18
1
0
0
5/21
• 5/25
1
0
0
5/28
6/1
1
0
0
1 Number of street cleaning tests per week are shown - a)? performed w/Mob?l
(brand name of equipment)
2 Not monitored; a starting date
3 Leaf removal tests.
15
-------
The data collected in these test areas were used to identify the
range of performance that can be expected from the available street cleaning
equipment. Differences in removal values (lbs/curb-mile removed) instead
of percentage removals (percentage of the initial loadings removed) for the
test conditions are used as a more meaningful measure of equipment perform-
ance. The lbs/curb-mile measure is more realistic because it provides the
specific amount of material removed.
The following elements summarize this program:
• All samples ("accumulation" and "before and after street cleaning
test") were sieved for particle size analyses with a 0.25 wire screen
(6370 microns), Tyler Screens No. 10 (2000 microns), No. 20 (850
microns), No. 30 (600 microns), No. 40 (250 microns), No. 60 (106
microns) and No. 325 (45 microns); and a pan. The pan collected the
material which passed through the finest screens.
• The loading values expressed in lbs/curb-mile for each particle size
and the percentage of sample in each particle size was calculated for
each sample.
• The samples for each size, each test area and test phase were combin-
ed to analyze the amounts of lead, zinc, volatile solids, chemical
oxygen demand, phosphorous, orthophosphates, arsenic, copper, chro-
mium, sulphur and Kjeldahl nitrogen.
SOURCES OF URBAN RUNOFF POLLUTANTS
One of the major problem areas yet to be sufficiently addressed concern-
ing urban runoff is the quantification of the relative contributions from
pollutant sources in the adjacent land areas of the watershed to the outfall
yield. This adjacent land area contribution is one of the factors discussed
in the urban runoff mass balance conceptual model described in the introduc-
tion section. The contribution from these non-street adjacent land areas
will be quantified in the final report. That is why this discussion of
sources of urban runoff pollutants is included here. The following discusses
the specific pollutants deposited in the major urban source areas and the
quantity that actually reaches the storm outfalls.
Usually sources that are farther from storm drainage systems require
overland flow and have a very low yield when compared with parking lots or
street surfaces which are impervious and located adjacent to the drainage
system. Table 3 presents the preliminary results from the second phase of
the Coyote Creek project (Pitt and Bozeman 1979), which examined potential
sources of urban runoff pollutants. In the second phase, runoff samples were
collected during rainstorms from different areas within the South San Francisco
Bay Area. These areas were all small and included runoff from different
types of building roofs, parking lots, and gutter flows. Rainfall and
outfall samples were also collected for chemical analyses. In most cases,
rain had the lowest pollutant concentrations. Puddles in cities had the
highest concentrations of total solids, specific conductance, and nitrates.
16
-------
TABLE 3 URBAN RUNOFF POLLUTAWT CONCENTRATIONS FROM MAJOR AREAS
(rag/1, imless otherwise noted)
Commercial
Residential
Parking
Park
Tar and
Composition
Parameter
Outfall
Gutter Flow
Lot
Puddles
Gravel Roof
Shingle Roof
Rain
pH, pH Units
7.8
7.5
7.0
7.3
. 7.5
6.5
6.1
Specific Conductance,
(uiti os/cm)
185
130
15
2)00
155
11.2
10.1
Turbidity, NTU
29
100
26
21
1
<1
<1
Total Solids
162
235
310
2110
186
18
B0D5
8
13
22
3
7
3
3
COD
97
172
176
69
131
19
12
P-POl
0.23
0.12
0.17
0.32
0.02
0.08
0.03
Total POl
0.3"
0.31
0.19
0.12
0.07
0.10
0.03
KJeldahl N
1.52
2.11
1.17
1.32
1-37
0.71
0.61
NH3
0.25
0.12
0.35
1.223
1.06
0.50
0.36
N03
0.71
0.12
0.13
285
0.22
0.09
0.09
S
M
2
<1
15
5
<1
<1
SOU
13
7
<1
38
21
<1
<1
As
<0.01
<0.01
0.02
0.10
<0.01
0.01
<0.01
Zn
0.06
0.11
0.23
0.01
0.08
0.18
0.01
Pb
0.08
0.67
1.09
0.035
0.019
0.017
<0.01
Cr
0.009
0.019
0.071
0.010
<0.005
<0.005
<0.005
Cu
0.013
0.029
0.016
0.031
0.11
<0.005
0.010
Total Coliform Bacteria
(MPN/100ml)
>2100
>2100
510 '
19
170
<2
8
Fecal Conform Bacteria
(MPN/100ml)
>2100
920
350
19
9
<2
2
Fecal Strep. Bacteria
(HPN/100ml)
>2100
>2100
>2100
920
17
920
<2
Fecal Coliform/Fecal
Strep. Ratio
-
<0.1
<0.2
0.5
0.5
<0.002
>1
Source: Pitt and Bozeman 1979
17
-------
Table 4 is a generalization of urban runoff pollutant sources for
common pollutants in the South Bay Area. No one source area is expected to
contribute'significant quantities of the pollutants, but some of the areas
are expected to be quite important. Street surfaces are expected to con-
tribute significant amounts of heavy metals. Oxygen demanding materials and
nutrients are thought to originate mostly from landscaped and vacant areas.
Table 5 is also a generalization of urban runoff pollutant sources and
attempts to show the major contributors affecting the major areas and the
approximate delivery yields from each of the source areas to the outfall
yield. Vacant lots and landscaped areas are the most pervious surfaces in
urban areas and are located farthest from the urban drainage system; there-
fore, they contribute little flow. Landscaped and vacant areas sometimes
constitute almost half of the total area in residential communities.
However, only 5 percent of the rainfall in these areas is expected to contrib-
ute to the outfall flow. Similarly, very little of the potential pollutant
yield from these areas is expected to affect the outfall. Rooftops, which
can constitute about 15-20 percent of residential communities, are also
located a relatively long distance away from the storm sewerage system.
Rooftops are impervious; however, a large part of the roof drainage systems
in Castro Valley are not directly connected to the storm sewerage system and
require considerable overland flow. Therefore, the outfall runoff yield from
rooftops is expected to be about 30 percent. Sidewalks, generally constitute
about 5 percent of residential communities, and are located closer to the
storm drainage systems. However, some of their drainage flow is directed
towards adjacent landscaped or other pervious areas. Consequently, only
about half of the runoff yield from sidewalks enters the receiving water
flow. Parking lots can make up about 7 percent of an area and in the Castro
Valley area are usually paved and impervious. Again, some of the runoff from
the parking lots (especially near homes and apartments) is directed towards
adjacent impervious areas, and only about half of the parking lot runoff is
expected to reach the receiving waters. Street surfaces, however, are
located close to the storm drainage systems and are mostly impervious. In
Castro Valley, street surfaces and parking lots constitute about 26 percent
of the watershed (Ellefsen and Raburn 1978). Most of the runoff originating
from impervious areas is expected to reach the outfall. However, some of the
street surface flow does not reach the outfall because of infiltration and
evaporation or streets in poor condition. Dust fall and precipitation affect
all of the components. However, dust fall, is not a major pollutant source,
instead it mostly transports pollutants. Most of the dust fall monitored in
urban areas is resuspended particulate matter from street surfaces or wind
erosion products from vacant areas. Some point source air pollutant emis-
sions also contribute todust fall pollution. The bulk of the dust fall,
however, is contributed by other major pollutant sources.
Automobile tire wear is a substantial source of zinc in urban runoff
and is deposited mostly on street surfaces and adjacent areas. About half of
the particulates from tire wear settle on the street and the remaining
material settles within about 20 feet of the roadway. Auto exhaust particu-
lates are also significant pollutant contributors of heavy metals, particu-
larly lead, and mostly affect street surfaces and adjacent areas. Other
18
-------
TABLE 4. POTENTIAL SIGNIFICANT URBAN RUNOFF POLLUTANT SOURCES
Common Urban
Runoff Pollutants
Rooftops
Potential Significant Polluant Sources
Street Parking Landscaped Vacant
Surfaces Lots Areas Land
Construction
Sites
Other
(Industrial
and Solid
Waste
Runoff)
Sediment
X
X
X
Oxygen Demanding
Matter
X
Nutrients
X
X
Salts
Bacteria
X
Heavy Metals
X X
Pesticides/
Herbicides
X
Oils and Grease
X
X X
X
Floatinq Matter
X
Other Toxic
Materials
. X
X
X
Source: Pitt and Bozeman 1979
19
-------
TABLE 5 MAJOR URBAN AREAS AND DELIVERY YIELDS TO OUTFALL (Percent)
Pollutant
Contributors
Lawn and
Landscaped Areas
5%
Vacant Lots
5%
Rooftops
30%
Sidewalks
45%
Parking Lots
50%
Street Surface
75%
Dustfal1
X
X
X
X
X
X
Precipitation
X
X
X
X
X
X
Tire Wear
X
(Adjacent)
X
X
X
Auto Exhaust Particulates
X
(Adjacent)
X
X
X
Other Auto Use
(Fluid Drips, Wear Prod.)
X
X
X
Vegetation titter
X
X
X
X
X
Construction Erosion
X
Other Litter
X
X
X
X
Bird Feces
X
X
X
X
Dog Feces
X
X
X
Cat Feces
X
X
Fertilizer Use
X
Pesticide Use
X
Source: Pitt and Bozeman 1979
-------
heavy metals and asbestos are important pollutants associated with these
other automobile losses. Most of these pollutants directly affect parking
lots and street surfaces and some material is transported to adjacent areas
by wind.
Vegetation is usually a significant source of pollutants in most
source areas. In Castro Valley the fall months are an important source of
leaf litter. While animal feces contributes quantities of important nutri-
ents and bacteria to urban areas, only vacant and landscaped areas are
usually affected. Fertilizer and pesticide use is mostly associated with
landscaped areas, but large amounts of pesticides can be used to control
plant growth on impervious surfaces and fertilizers can be used in road
maintenance operations.
Table 6, based on preliminary results from the current Coyote Creek
project (Pitt and Bozeman 1979), estimates the percentage contribution of the
various pollutants from the different source areas studied. In most cases,
rooftops contribute the least amount of pollutants, while pervious areas are
the major source of solids, oxygen demanding materials, and some nutrients.
Parking lots, street surfaces, and sidewalks are expected to be a major
source of heavy metals, bacteria, and some nutrients of the total outfall
runoff yield.
Most of the street surface dust and dirt materials (by weight) are
local soil erosion products, some of which are deposited by motor vehicle
emissions and tire wear. Minor contributions are made-by wear of those
smooth street surfaces in good repair. The specific makeup of street surface
contaminants is a function of many site conditions and varies widely. Many
pollutant sources are specific to particular areas and activities. For
example, iron oxides are associated with welding operations and strontium,
used in the production of flares, would probably be found on streets in
greater quantities during holiday times or at the scenes of traffic accidents.
Relative deposition values for pollutants from source areas are summa-
rized in Table 7. These deposition values are percentages of the total
pollutants deposited in urban areas and are much larger than the pollutant
yields to the outfall. In comparison, Table 8 shows the relative yields from
these source areas to the total outfall runoff yield. The deposition rates
for some pollutants are relatively high for some of the impervious areas, but
these source yields are substantially reduced when infiltration is considered.
Automobile activity is responsible for most of the heavy metal yield in the
runoff and up to half of the total solids yield. Vegetation sources con-
tribute most of the oxygen demand materials, and domesticated animal feces
and fertilizers contribute most of the nitrogen in urban runoff.
In conclusion, if the total solids pollutant depositions in an urban
area were summed, only about one-third would reach the outfall. Only
about 10 percent of the nutrients and oxygen demanding materials deposited
would affect the quality of receiving waters, but most of the heavy metals
deposited in the area would affect water quality. Those pollutants that
21
-------
Table 6. ESTIMATED CONTRIBUTION FROM VARIOUS AREAS TO OUTFALL RUNOFF YIELDS (Percent)
Sources
Total
Solids
COD
B0D5
Ortho
P°4
Total
P04
Total
Kjeldahl
N
nh3 no3
Total
S
Rooftops
2
5
2
10
5
15
20 2
5
Landscaped Areas
and Vacant Lots
50
50
50
20
15
25
20 5
50
Parking Lots
10
15
20
30
20
10
10 1
1
Street Surfaces
and Sidewalks
40
25
25
45
65
50
50 20
45
S04
As
Zn
Pb
Cr
Cu
Total
Coliform
Bacteria
Fecal
Col iform
Bacteri a
Fecal
Strep
Bacteri a
Rooftops
5
10
20
1
1
10
1
1
30
Landscaped Areas
and Vacant Lots
40
70
1
1
2
10
1
1
3
Parking Lots
1
15
20
20
20
20
5
3
10
Street Surfaces
and Sidewalks
50
1
60
80
80
60
95
95
60
Source: Pitt and Bozeman 1979
22
-------
TABLE 7. RELATIVE SOURCE ANNUAL DEPOSITIONS (Percent)
Source
Total
Sol ids
bod5
TKN
Pb
Zn
Cr
Cu
Precipitation
10
3
5
1
15
1
10
Tire Wear
5
1
10
80
10
5
Auto and
Street Use
20
1
1
90
1
70
50
Vegetation
40
95
15
Construction
Erosion
20
1
1
2
10
2
Bird Feces
2
1
1
1
1
Dog Feces
40
Cat Feces
1
Fertilizer Use
40
Other
3
1
10
30
Source: Pitt and Bozeman 1979
23
-------
TABLE 8. RELATIVE RUNOFF YIELD CONTRIBUTIONS (Percent)
Source
Total
Sol ids
bod5
TKN
Pb
Zn
Cr
Cu
Precipitation
10
10
20
1
20
1
5
Tire Wear
5
1
1
80
3
2
Auto* and
Street Use
50
10
10
100
90
80
Vegetation
30
70
10
Construction
i Erosion
5
1
1
1
1
1
Bird Feces
5
1
1
1
1
Dog Feces
30
Cat Feces
1
Fertilizer Use
30
Other
3
5
5
15
Source: Pitt and Bozeman 1979
24
-------
are washed off the source areas and do not reach the outfall would be
accumulated in other areas of the urban environment. The most significant
pollutant "sinks" in the urban area are expected to be soils and plants. For
example, many studies have shown significant concentrations of heavy metals
in roadside soil and vegetation (Farmer and Lyon 1977; McMullen and Faoro
1977; Olson and Skogerboe 1975; Pitt and Amy 1973). As noted earlier, much
of this material (about 15 percent of the total deposition) can be associated
with dust fall. However, most of this dust fall is resuspended particulates
from streets and vacant areas and is not an actual source of urban runoff
pollutants, excepting point source air pollution emissions that settle out.
STREET SURFACE CONTAMINANT ACCUMULATION RATES
The objective of this portion of the study was to determine specific
accumulation rates in the test areas. This information must be obtained
before an effective street cleaning program can be designed. In addition,
this information was necessary to evaluate the Castro Valley Creek pollutant
flows. The rainfall patterns during this study were examined and those
periods in which rains removed significant amounts of street surface contam-
inants were eliminated from the accumulation rate calculation. To determine
accumulation rates of different pollutants, the samples were analyzed by
particle size (as previously described). This procedure is essential
because different particle sizes have different concentrations of pollutants.
Equipment performance also varies with particle size and this affects the
overall amount of pollutants that can be removed by street cleaning.
Approximately 58 days had measurable rain during the 1978-1979 portion
of this project. Twenty-three of these days had what was considered to be
significant rainfall. A significant rain is one that removes as much of the
street surface contaminants as possible. However, these rains can also add
material to the street surface through erosion of adjacent areas. A signif-
icant rain is defined as having a total rainfall of about 0.2 inches, or
more, within about 1-day (irrespective of traffic conditions); rain with a
peak instantaneous intensity (5-minute duration) of 0.5 inches per hour
(irrespective of traffic conditions); a rain with an average intensity of 0.1
inches per hour, or greater, with moderate to heavy traffic. Rains and
traffic conditions which meet one of these criteria are capable of imparting
enough energy to the street surface to loosen contaminants and suppling
enough water to flush them along the street surface and gutters to the storm
sewerage inlets. If sufficient amounts of water are not available to carry
the particulates through the storm sewerage to the outfall, material will be
deposited in the sewerage system. Rainfall intensities and removal effec-
tiveness relationships were studied by Sartor and Boyd (1972) and discussed
by others (including Pitt and Field 1977).
Periodic and unexplained decreases in loadings were observed. These
data resemble the sawtooth pattern similar to that shown in Figure 6. These
decreases in loadings may be caused by high winds. Winds greater than
about 13 miles per hour have been shown to be capable of removing particulates
from the street surface in the absence of traffic (McMullen 1977). In some
25
-------
cases, significant rains cause decreases in street surface loadings while
they cause an increase in others. Increases are thought to be caused by
erosion. As described previously, all of the test and accumulation samples
were separated by particle size. Samples of each particle size category for
each test area were then analyzed for the various pollutants.
Figures 7, 8, and 9 show estimated curves for total solids accumulation
and deposition rates. The loading values and associated time periods since
last cleaned were grouped by test area and season to identify the best
fitting curves. Loading values that were affected by rains were eliminated
from this analysis. These figures are highly influenced by residual loading
values. In general, loading values are the cleanest the streets can be and
are the values immediately after street cleaning or certain rains.
The resulting loading, deposition, and accumulation rates were quite
different for each test area. Loadings did not level off although the
accumulation rates decreased with time. Table 9 presents calculated average
accumulation rates for the three test areas for various times after street
cleaning or a significant rain. The lower test area had appreciably greater
accumulation rates and loading values than the other two test areas. The
middle area had intermediate rates and values, while the upper area had the
lowest loading values and accumulation rates. The deposition rates are
assumed to be equal to first day accumulation rates (70, 60 and 30 lbs/curb
mile-day for the lower, middle, and upper areas, respectively). The two
factors which affect the accumulation rate are the deposition rate and
the removal rate. The accumulation rate equals the deposition rate minus the
removal rate. The deposition rate is a function of the characteristics of
the area, such as climate, land use, traffic, and street surface conditions.
Removal of pollutants can be accomplished either by street cleaning or
naturally by winds or rains. The difference between the accumulation and
deposition rates (at any time) is assumed to be loss from the street surface
by wind or traffic-induced turbulence. This material can become suspended in
the air, but most of the material settles about 10 feet off the roadway (Pitt
1979).
The average street surface total solids loadings immediately after
street cleaning are also shown in Table 10. These values are also greatest
for the lower test area and the least for the upper test area. Significant
rains can produce cleaner street surfaces in the lower and middle test areas
when compared to street cleaning, but in the upper test area, street cleaning
can result in cleaner streets more than significant rains.
CASTRO VALLEY STREET CLEANING DEMONSTRATION STUDY RESULTS
The design of an effective street cleaning program requires not only
a determination of accumulation rates but also an assessment of the specific
street cleaning equipment performance for the actual conditions encountered.
Service goals, which consider effects on water quality, air quality, public
safety, aesthetics, and public relations are the driving force in establish-
ing a street cleaning program. The major objective of this part of the study
was to determine the effectivenesses of street cleaning equipment for various
levels-of-effort for comparison with pollutant flow information.
26
-------
Figure 7
PfcPOSlTlON ^ ACCUMULATE RATES
FOR LOWtR. Tt6T AWtA
PAY£> SINCE LAST CLEANE:P Ol? SIGNIFICANT RAlM
-------
FIGURE 8
PCPOSITIOM <£/ACCUMULATION RATftS
FOR MIPPLt TfrST AREA
PAYS 3INCE- LAST CLE-AfJtP OR SIGNIFICANT" KAIM
-------
FIGURE 9
P&P06ITI0N ^ACCUMULATION RATtS
FOR UPPER TfcST AKtA
ro
A
W
3
u
-------
TABLE 9. CASTRO VALLEY TOTAL SOLIDS ACCUMULATION RATES
Lbs./Curb Mile
Middle Area
Loading Accum.
Value Rate
Days After
Street Cleaning or
Significant Rain
Lower Area
Loading Accum.
Value Rate
Upper Area
Loading Accum.
Value Rate
0
1
3
5
7
10
15
20
25
30
35
530
600
725
840
930
1,050
1,180
1,240
70
63
58
45
40
26
12
500
560
630
700
760
850
960
1,050
1,100
1,160
1,200
60
35
35
30
30
22
18
12
10
8
285
315
365
405
440
490
560
605
30
25
20
18
17
11
9
Average first day
loading values after:
Street Cleaning:
Significant Rains:
620
490
590
420
280
350
30
-------
TABLE 10. STREET CLEANING EFFECTIVENESS
Test Area
Date
lbs./Curb-Mile
Before After Removal
Loading Loading lbs.
Removal
X
Days since
Last cleaninq
Days since
Last rain
Significant
Rain
Interference
Lower
3/23
1,385
1,182
203
15
121
8
Yes
3/30
564
557
7
1
7
3
Yes
4/5
955
727
228
24
6
9
4/12
535
365
170
32
7
16
4/18
1,131
1,371
-240
-21
6
22
5/14
1,012
664
348
34
26
18
Yes
5/15
807
732
75
9
0.8
19
5/16
758
470
288
38
1.0
20
5/17
679
668
11
2
1.0
21
5/18
568
535
33
6
1.0
22
Middle
1/19
454
381
73
16
58
1.4
Yes
1/25
703
473
230
33
6
8
'2/1
462
455
7
2
7
15
2/7
674
429
245
36
6
21
3/12
1,070
729
341
32
33
12
Yes
3/13
629
623
6
1
1
13
3/14
756
824
-68
-9
1
14
Upper
2/27
488
199
289
59
97
3
Yes
2/28
332
202
130
39
0.9
4
3/2
236
249
-13
-6
1.9
1.5
Yes
3/5
298
193
105
35
3
4
4/25
420
344
76
18
51
29
Yes
5/3
400
424
-24
-6
9
8
5/9
209
182
27
13
6
14
5/23
551
340
211
38
14
28
5/31
575
357
218
38
8
36
Average Removal * 115 lbs./curb-mile
Average Removal percentage ¦ 18.4X
31
-------
Street cleaning performance depends on many conditions including
the character of the street surface, street surface initial loading charac-
teristics (total loading value and particle size distribution), and other
environmental factors. Street cleaning variables that most affect street
cleaning performance include the number of passes the equipment makes and the
intervals in street cleaning. The most important measure of street cleaning
effectiveness is Ibs/curb-mile removed for a specific program condition.
This removal value, in conjunction with unit curb-mile costs, allows the cost
for removing a pound of pollutant for a specific street cleaning program to
be calculated. The "percentage-of-the-before-loading-removed" has often
been used as a measure of street cleaning equipment performance. It is very
misleading, however, because it is not a measure of the magnitude of the
amount of material removed. A street cleaning program may have a very low
percentage removal value but a high total amount removed if the initial
loading value is high.
The loading and removal rates for all pollutants combined in each test
area for all street cleaning programs combined are summarized in Table 10.
The percentage removal values for the total solids are similar to other
pollutants monitored; however, the removal rates expressed on a lbs/curb-mile
removed basis vary greatly. The lbs/curb-mile removed values can be used to
estimate the quantity of pollutants that are removed over both a large
area and long time period.
Figures 10, 11, and 12 are production functions developed from the
first year information for the three test areas. Figure 10 presents the
average for all three test areas for the amount of street dirt removed as a
function of the street cleaning effort. For example, monthly street cleaning
may result in removals of about 2,500 lbs/curb-mile each year. Weekly street
cleaning, at four times the cost, may remove about 8,000 lbs/curb-mile each
year, and daily street cleaning may remove a total of about 25,000 lbs/curb-
mile each year. Unit costs increase as the street cleaning effort increases.
Therefore, it is important that a minimum level-of-effort be used to meet the
program objectives. Additional street cleaning operations conducted in an
area are usually less efficient than if part of the street cleaning effort is
redirected to other areas.
Figure 11 is a graph showing the initial total solids loading values.
These values are the total solids particulates values on the street before
cleaning with different levels-of-effort. These values are the dirtiest that
the streets are likely to be for the various street cleaning programs and
test areas. Again, there is a decrease in unit effectivness as the number of
cleaning passes increased in a year. After about two passes per week, very
little additional cleanliness can be obtained in the test areas. However,
the lower area may be about four times as dirty as the upper area at this
mimimum value.
Figure 12 presents the residual street surface total solids loading
values. These residual values are the cleanest that the streets are likely to
be with these various street cleaning programs. Again, the upper area can be
made noticably cleaner than the other two areas. There is not a significant
improvement in these residual loadings after about two or three passes per
week. There can be major changes in going from monthly to weekly street
cleaning.
32
-------
FIGURE1 10
5TREBT P/RT RCMOVCP AS A FUNCTION
OP TH6 STREET CLEANING EFFORT
NUMUK OF PASSES PtK YEAR.
-------
GJ
-ps»
FIGURE 11
INITIAL dTREeT L0APIN6 AS A
FUNCTION OF THE CLEANING EFFORT
y 1500 -f-
&
V
r>
u
v3
Z
^ 1000 - •
OJ
u
o
u.
flj
(0
>3
2
5
«<
o
s
500 --
LOWte AREA
MlPPLe AREA
o
-------
FIGURE 12
RESIPUAL STREET L0APIN6 AS A
FUNCTION OP THE CLEANING EFFORT
co
c_n
1500 4-
a
3
Vi
*--v
a
-------
These figures are based on preliminary information and will change
after data from the complete study is analyzed for the final, second year
report. However, the figures do point out important aspects of street
cleaning programs. The most important of these aspects is that additional
cleaning after a certain level-of-effort is not productive and that street
cleaning effort would be better applied to other areas. The information
relating to street cleaning program effectiveness and accumulation rates is
used to address the major objectives of this demonstration project. This
major objective is relating different magnitudes of street cleaning effort to
Castro Valley Creek pollutant flow information and is presented in Section 5
of this report.
36
-------
SECTION 5
CASTRO VALLEY CREEK MONITORING
The objectives of the study of pollutant mass flow characteristics
of Castro Valley Creek were:
o To estimate the beneficial effects of street cleaning on the
quality of receiving waters.
o To investigate the quantity of pollutants removed from the watershed
by the various types of rainstorms.
Runoff event monitoring began on December 5, 1978, after two receiving
water monitoring stations were installed by the USGS at Seaview Avenue (USGS
#11181004) and Knox Street (USGS #11181096). Some of the monitored flows at
these two locations are shown in Figure 13.
Nineteen individual storm-days have been monitored. However, for
most data analysis purposes, a monitored storm cannot be used unless a street
surface sample is collected from the entire study area before and after a
storm. If an unmonitored storm occurs in a series of monitored storms
between adjacent street surface sampling, the complete runoff yield for that
storm series cannot be calculated or compared to the in street surface
loadings. Consequently, from the 19 storm-days, eight valid data sets
have been obtained for analysis.
Since the onset of the 1978-1979 wet season, measurable rain has
occurred on 58 days (with 1-day of rain prior to initiation of field activ-
ities). These events are summarized on Table 11. In comparison to an
annual average of 21.7 inches, the total amount of rain from September 1978
to April 17, 1979 was almost 17 inches. Individual rains have lasted from
15 minutes to 23 hours per day and the storm periods have ranged to a
maximum of 7 days.
Table 12 summarizes the significant rains that occurred during the
field activities of the first year of this study and compares rainfall
conditions observed at the Proctor and Fire Station raingage locations. The
rain total and intensity values observed at these two stations agree within
reasonable limits. The times of the rains observed at the Proctor Station
are also indicated on this table. The total rainfall during any one of these
significant rain days ranged from 0.18 inches to 1.8 inches. Average inten-
sity ranged from 0.01 inches to 0.21 inches per hour. Peak 5 minute intensi-
ties during these rains were substantially greater. Peak 1 hour intensities
ranged from 0.4 inch per hour to 0.67 inches per hour.
Table 13 summarizes the 1978-1979 monitored rain year. Almost seventeen
inches of precipitation occurred during this rain year, which is about 77
percent of the normal annual rain. Rain occurred on a total of about 58
days, while 19 of these days were included in the Castro Valley Creek mon-
itoring effort. Significant rains occurred on a total of about 23 days
during this period of time. January and February accounted for most of
the rain that occurred in this year, while June, July, August, and October
37
-------
FIGURE 13. RUNOFF FLOWS AT MONITORING STATION SITES
Discnarge at Knox Street Station
B. Discharge at Seaview Avenue Station
C. Discharge and Staff Gage
at Seaview Avenue Station
38
-------
TABLE 1 1 CASTRO VALLEY RAIN EVENTS DURING FIELD ACTIVITIES1
Total
Duration
Average Intensity
Peak Intensity
Date
(inches)
(hours)
(inches/hour)
(L-Khes/hour)
Dec. 17, 1978*
.39
12.5
.03
.14
Dec. 18
.05
15.75
.003
.03
Dec. 19
.01
0.25
.01
.01
Jan. 3, 1979
.10
3.25
.03
.03
Jan. 1
.03
9.25
.003
.02
Jan. 5
.01
0.25
.01
.01
Jan. 7*
• 31
14.75
.02
.05
Jan. 8*
1.21
6.0
.21
.10
Jan. 9
.18
8.25
.02
.04
Jan. 10#
.78
1.25
.18
.39
Jan. 11
1.80
20.75
.09
.27
Jan. 11*
1.43
20.75
.07
• 33
Jan. 15
.28
12.75
.02
.09
Jan. 17
.2"
5.75
.01
.11
Jan. 30
.01
.25
.01
.01
Feb. 3
.01
.25
.01
.01
Feb. 13*
1.11
13-25
.08
.25
Feb. It
.09
9.25
.01
.01
Feb. 15
.01
.25
.01
.01
Feb. 16*
¦ «9
11.75
.04
.21
Feb. 17
.01
.25
.01
.01
Feb. 18
.15
8.75
.05
.20
Feb. 19*
.06
1.25
.05
.05
Feb. 20*
.71
17.75
.04
.25
Feb. 21*
.41
11.0
.04
.17
Feb. 22*
.61
13.25
.05
• 30
Feb. 23
.18
23.25
.01
.07
Feb. 25
.07
2.75
.03
.04
Feb. 26
.09
9.50
.01
.06
Feb. 28»
.58
6.25
.09
.17
Mar. 1
.01
'.25
.01
.04
Mar. 3
.03
1.5
.01
.01
Mar. 15
.05
2.7
.02
.04
Mar. 16*
.69
17.3
.04
.15
Mar. 17
.01
.25
.04
.01
Mar. 26
.140
13.15
.03
.13
Mar. 27*
1.15
23.0
.06
• 38
Mar. 28
.13
20.85
.01
.06
Apr. 6
.02
.25
.08
.08
Apr. 9
.01
.25
.04
.04
Apr. 16
.05
9.15
.01
.04
Apr. 17
• CW
6.15
.01
.01
Apr. 22
.01
1.85
.02
.03
Apr. 23
.11
20.0
.01
.09
Apr. 25
.01
.25
.04
.04
Apr. 26*
¦ 37
12.15
.03
.12
May 5
.02
.5
.04
.02
May 6
.05
23-5
.002
.03
May 7
.01
12.75
.003
.02
May 8
.03
2.0
.015
.02
May 15
.01
.25
.(W
.01
Station discontinued for remainder of water year
1 Proctor School Rain Gage, USGS #71-1810.08
• Monitored Events
39
-------
TABLE 12. SIGNIFICANT RAINS DURING FIELD ACTIVITIES OF FIRST YEAR
Time of
Total
Rain (in.)
Average Intensity (in/hr)
Peak Intensity (in/hr)
Rain
Date
Proctor
Fire Sta.
Proctor
Fire Sta.
Proctor
Fire Sta.
Proctor
Dec. 17,1978
0.39
0.44
0.03
0.02
0.14
0.1
0530
1800+
Jan. 7, 1979
0.34
0.38
0.02
0.02
0.05
0.05
0830
2315+
8
1.24
1.35
0.21
0.21
0.40
0.67
0945
1545+
9
0.18
0.12
0.02
-
0.04
-
0445
1300+
10
0.78
0.64
0.18
0.13
0.39
0.33
1925
2400+
11
1.80
1.57
0.09
0.01
0.27
0.23
0045
2130
14
1.43
0.58
0.07
0.09
0.33
0.15
0245
2330+
15
0.28
0.57
0.02
-
0.09
-
0130
1415
17
0.24
0.23
0.04
-
0.11
1630
2215
Feb. 13,1979
1.11
1.18
0.08
0.05
0.25
0.34
1020
2345+
16
0.49
0.48
0.04
0.04
0.21
0.28
0115
1300+
18
0.45
0.31
0.05
0.03
0.20
0.07
0830
1715+
20
0.74
0.84
0.04
0.04
0.25
0.21
0530
2315
21
0.41
0.57
0.04
0.05
0.17
0.27
0045
1145
22
0.64
0.90
0.05
0.08
0.30
0.43
0645
2000+
23
0.18
0.18
0.01
-
0.07
-
0045
2400+
28
0.58
0.58
0.09
0.08
0.17
0.19
1500
2115+
Mar. 16,1979
0.69
0.65
0.04
0.03
0.15
0.12
0230
1945+
26
0.40
0.45
0.03
0.03
0.13
0.13
0930
2245+
27
1.45
1.56
0.06
0.07
0.38
0.45
0100
2400+
Apr. 26,1979
0.37
0.35
0.03
0.08
0.12
0.19
0500
1715
40
-------
of 1978 did not have any rain at all. Also shown on Table 13 is the average
number of days between these rains. For the 23 significant storm days,the
longest dry accumulation interval was about 150 days (this occurred before
the September rain). In contrast, only about 4 dry accumulation days occurred
between significant rains during January and February. Therefore, it can be
expected that the first rain of the year, typically occurring in September or
October, can account for almost one-half of the annual pollutant runoff from
street surfaces. January and February rains are mostly influenced by erosion
(non-street surface runoff) and are not significantly affected by street
cleaning.
The collected runoff samples were analyzed individually and in selected
composites. The more important parameters were investigated at different
times during several initial rains to see how the concentrations changed as
the rain progressed. Other parameters were analyzed only once during each
monitored rain. These individual composite samples were based on a flow-
weighted basis and represent the total quantity of pollutants flowing in
Castro Valley Creek at the two stations.
Nineteen rain days were monitored during the first year of study (Table
13). The pollutant flow information from the Knox and Seaview monitoring
stations was used to calculate the amounts of pollutant mass discharge that
entered the creek.
ANALYTICAL PROGRAM
This project is concerned with total mass flows of pollutants and
pollutant removals associated with various types of street cleaning programs
and meteorological conditions. In order to maximize the number of monitored
events, composite sampling and restricted analyses are necessary. Automatic
water samplers directly connected to flow measuring devices are used at
the two monitoring stations to enable flow-weighted composite samples to be
obtained. Samples for any storm can be directely combined into a single
flow-weighted composite sample representative of the total mass flows during
the storm. Analyzing this single sample will enable calculation of the total
mass flow that occurred during the runoff event.
The runoff measuring devices were not calibrated until several storms
had occurred, so the first few storms necessitated collecting discrete
samples throughout the period of the storm for individual analyses. These
concentrations, represent!'ative of different time periods in the storm, were
later used to calculate mass flows. The first storm samples involved analyz-
ing 10 discrete samples spread over the period of the runoff flow period.
The next two runoff events were sampled utilizing an estimated stage-flow
relationship, enabling a significant reduction in the amount of samples
necessary and the associated costs. Additional storms were analyzed on a
total storm flow-weighted composite basis during the first project year.
Table 15 lists the parameters included under three different sampling
and analytical priorities. The first priority samples are the most important
ones and are expected to be most sensitive to street surface loading condi-
tions and associated street cleaning programs. The heavy metal storm yields--
specifically lead and possibly zinc--are expected to be most sensitive to
41
-------
TABLE 13. SUMMARY OF CASTRO VALLEY 1978-1979 MONITORED RAIN YEAR
Number of Number of Number of Avg. No. of Avg. No. of
1978-79 Total Rain-days
"Rain Year" Rain (in) Monitored
Days with Insignif.
Rain >0.01 in. Rain Days
Days between
Any Rain
Rain between
Signif. Rain
30 June,1978
0
0
0
0
30*
30*
31 July
0
0
0
0
31*
31*
31 August
0
0
0
0
31*
31*
30 September
0.35
0
1
1
30
30
31 October
0
0
0
0
31*
31*
30 November
0.84
0
5
1
6
30
31 December
0.81
1
4
I
8
31
31 January,1979
6.44
4
12
8
3
4
28 February
4.94
7
15
8
2
4
31 March
2.77
6
8
3
4
10
30 Apri1
0.65
1
8
1
4
30
31 May
0.15
0
5
0
6
31*
TOTAL
16.95
19
58
23
~Whole month; add to preceeding months value for total accunulation between storms:
Significant
Storm Number
for Year
1
2
3
4-11
12-19
20-22
23
Approximate
Month
Sept.
Nov.
Dec.
Jan.
Feb.
March
Apri 1
Approximate
Accumulation
Interval (days)
149
61
31
4
4
10
30
42
-------
TABLE 14. PARAMETER LIST FOR RECEIVING WATER ANALYSES
FIRST PRIORITY
Total Solids
Lead
Zinc
Specific Conductance
Turbidity
PH
SECOND PRIORITY
Total dissolved solids Iron
Chemical oxygen demand Chromium
Phosporous Copper
Orthophosphate Nickel
Kjeldahl nitrogen Cadmium
Sulfate Mercury
Arsenic Ammonia
Suspended Solids Nitrites
Volatile suspended solids Potassium
Bicarbonate Calcium
Carbonate Magnesium
Chloride Sodium
THIRD PRIORITY
Organic/pesticide-herbicide screen (by GC/MS)
Elemental screen (by SSMS)
43
-------
street cleaning programs. The total solids analyses are important to relate
relative concentrations (milligrams of pollutant per kilogram of total
solids) of the runoff with the street surface samples. Specific conduc-
tance, turbidity, and pH measurements were taken on each individual sample
throughout the project period as indicators of concentration variations
during the storm period. The first priority list was analyzed on about five
discrete samples during the two storms utilizing an assumed stage-flow
relationship program in the sampling equipment. The second priority list is
a comprehensive list of parameters generally analyzed for urban runoff
samples reflecting the important parameter groups of solids, organics,
nutrients, heavy metals, and major ions. A single time composited sample was
analyzed from the next three storms for this list as a general indication of
the runoff and receiving water characterization. The third priority list
includes two types of screening procedures basically designed to identify other
parameters not included in the normal monitoring program that may be present
in important concentrations.
The total watershed yeilds were measured at the lower monitoring
station (located at Knox Avenue) and the non-urban portion of the watershed
was measured at the upper monitoring station. Street cleaning activities
were only conducted in the lower urbanized sections of the watershed, so the
mass yields were corrected and most of the analyses reflects the urbanized
area only.
POLLUTANT REMOVAL CAPABILITIES OF MONITORED STORMS
Table 16 presents the total solids and various pollutant loading
conditions observed on the street surfaces immediately before each of eight
monitored rain periods. Table 16 also shows the urban runoff entering Castro
Valley Creek between the two monitoring stations. Calculations were made to
determine changes in the street surface contaminant loadings and the runoff
yields. These calculations include factors to reflect expected street
surface loading conditions immediately before and after the rain events.
These loading corrections are based on the previously described street
surface pollutant accumulation rates. Values smaller than one in the ratios
between the street surface washoff and the observed creek yields possibly
signify that more of the pollutants originated from non-street surface areas.
Values greater than one possibly indicate that most of the material that
originated from the street surfaces accumulated in the storm sewerage. These
ratios appear to vary for the different pollutants as a function of the rain
storm characteristics (magnitude of the rain quantity and accumulation
interval between rains). The storms of greater intensity and lower street
surface loading conditions appear to result in more material originating from
surrounding areas and carried onto the street surfaces during the storms
through erosion. Storms of relatively small intensities and long accumula-
tion intervals, however, showed almost complete removals of street surface
contaminants with little additional erosion yield for some of the pollutants.
On some of these smaller rains, the flows in the sewerage were not capable of
preventing the material from depositing in the sewerage. The small number of
data points available for this analyses prevents a detailed model from being
developed. However, the eight data points observed were used to develop
44
-------
Table 15 OBSERVED STREET AND CREEK YIELDS
Days
Since Avg. Total Solids
Total Last Days Initial Storm
Storm Rainfall S1gni f. Since Load Yield Lead Zinc COO
Date (In.) Rain Cleaned (Lbs.) (Lbs.) Load Yield Load Yield Load Yield
Jan 8
1.24
1
47
66,300
112,000
245
--
33
70
10,600
20,000
Jan. lO-
ll, 14
4.01
1
49
15,400
65,600
33
—
4.9
41
1,750
—
Feb.13-
14
1.20
27
57
37,200
8,000
63
19
9.8
9.6
3,070
7,200
Feb. 15-16
18,19,20,
21 S. 22
2.80
2
59
27,100
119,000
39
80
7.0
54
2,160
22,600
Feb. 28-
Mar. 1
0.58
5
40
29,200
' 12,300
50 9.6
8.1
9.9
2,370
3,710
Mar.15-17
0.75
15
39
43,800
6,990
69
6.7
12
4.5
3,740
2,490
Mar.26-28
1.98
10
12
45,000
26,500
65
21
12
14
3,930
7,400
Apr. 26
0.37
31
17
33,500
4,970
47
8.2
8.5
4.1
2,850
2,020
Storm
Date
Total P 0-P04
Load Yield Load Yield
Arsenic Copper Chromium
Load Yield Load Yield Load Yield
Kjeldahl Volatile
Sulphur Nitrogen Solids
Load Yield Load Yield Load Yield
Jan. 8
0.66
130 0.51
—
2.7
- 10.6
-- 30.0
—
140 --
77
470 12 t-
Jan. lo-
ll, 14
0.11
-- 0.085
—
0.44
1.7
--
4.9
--
22 -
12
— 2.1 *-
Feb. 13
14
0.19
23 0.
14
29
0.73
0.30 3.3
2.3
8.7
0.55
43 -
23
83 1.9 --
Feb. 15-16, 0.13
18,19,20,
21 & 22
130 0.
10
260
0.54
1.8 2.3
13
6.2
5.7
29 --
16
470 1.3 --
Feb. 28- 0.15
Nar. 1
13 0.
12
8.9
0.62
0.27 2.5
2.2
6.9
0.86
33 -
18
97 1.5 --
Mar. 15-
-17 18
6.7 1
.9
2.9
0.43
0.10 3.1
1.2
12
0.63
35 -
17
18 3.1 —
Mar. 26-28 18
38 2.0
15
0.45
0.50 2.9
2.9
13
1.9
36 -
18
135 3.2 «
Apr. 26
14
5.8 1
.6
0.93
0.34
0.042 2.1
0.71
9.5
0.25
27 --
13
41 2.8 -
45
-------
rough approximations between street cleaning effectiveness and improving
the creek runoff yields. The additional monitoring effort during the second
year of the demonstration program will yield much more information for this
analytical procedure.
EFFECTIVENESS OF STREET CLEANING IN IMPROVING RECEIVING WATER QUALITY IN
CASTRO VALLEY
Street cleaning can be effective in reducing the quantity of some
pollutants in urban runoff and therefore the effects of urban runoff on
receiving waters. Most of the materials removed by a street cleaner on
smooth streets would wash off during a rain and contribute to the pollution
of a receiving water. Figures 14 through 18 relate street surface pollutant
loadings before the storm to the change in creek pollutant yield between the
two monitoring stations.
In almost all cases, two specific groupings of data were observed.
One grouping was associated with short accumulation intervals before the
monitored storm (one or two days) with storm quantities of values greater
than 1-inch (up to 4-inches). These storms showed (for almost all the
pollutants) that other sources besides the street surface are more important
in urban runoff pollutant yields. Street cleaning therefore has little
effect in reducing the pollutant loads from these storms. These storms have
sufficient energy to remove the small amounts of street surface pollutants on
the street before these storms and to erode significant quantities of pollu-
tants from other known pollutant sources (such as roof tops, parking lots,
vacant lots, landscaped areas, and other areas.
The other group of storms observed had relatively long accumulation
intervals between storms (5 to 31 days) and had relatively small amounts of
rainfall (0.37 to 1.98-inches of total rain). The ratios of number of days
since last significant rain to total rain, in inches, for this category was
always greater than about 5 (a maximum ratio of 84 was observed). In contrast,
the ratio between the accumulation interval, in days, and total rainfall, in
inches, for the other group of rain conditions observed were always less than
one. These lighter rains were capable of removing as much of the street
surface contaminants as possible, with little additional contributions for
additional pollutant source areas. Street cleaning activities could be very
important in affecting the runoff water quality and receiving water condi-
tions for these storms.
Figures 14 and 15 show the street surface loading and runoff yield
relationships for total solids and chemical oxygen demand (COD). For the
lighter rains, with dirtier streets, (Figure 14) much of the street dirt
accumulated in the sewerage or in the creek or was not completely washed off
the street surface. Rains having heavier intensities and shorter accumulation
intervals (Figure 15) showed that other sources of total solids were much
more important than street dirt. These heavier rains were much more capable
of creating significant erosion which would quickly dominate the urban runoff
when compared to the street dirt yields.
46
-------
FIGURE. 14
total soups anp COP CREEK YltLPS FOR
STORMS Of RELATIVELY SMALL RAINFALL QUANTITIES ANP L0N6 PKY ACCUMULATION PERIODS
l0-l I I I I 1 I I-1 I I I I I I T
<•000 2,000 5,000 IOOOO KLOOO $0000
STUPY AREA CREEK POLLUTANT FLOW (L&s/iTORM EVENT)
47
-------
FIGURE 15
TOTAL SOUPS ANP COP CREEK YieLPS FOB
STORMS OF RELATIVELY LAR6E RAINFALL QUANTITIES ANP SHOCT PRY ACCUMULATION PERIOCS
I " I I ~l
mnnwi
°~i 1 1 1 1—i—I i—I I i r~ i"~i—
1MB ZOOO s.ooo 10000 8*000 *1090
STUPY AREA CREEK POLLUTANT FLOW (L&s/sTOKM EVttfI-)
48
-------
A similar relationship was observed for lead (Figures 16 and 17)
either not all the lead was removed from the streets or it was deposited in
the sewerage or the creek before the Knox monitoring Station for the lighter
rains. For the heavier rains with shorter accumulation intervals, more lead
was observed in the creek flow than was on the street before the rain occurred.
Other sources of lead (besides the street surfaces) were probably the storm
sewerage and adjacent areas where lead was blown by winds or traffic induced
turbulence off the street surfaces or parking lots and driveways could also
have contributed lead under these conditions.
The zinc and COD relationships were slightly different. The heavier
rains showed that much of the zinc and COD occurred in erosion material and
not street surfaces. However, the lighter rains with the dirtier streets
showed that almost all of the zinc or COD observed in the creek flow was due
to the street surface loadings.
The nutrients Kjeldahl nitrogen, total phosphorous, and orthophoshate
showed that almost all of the nutrients were associated with non-street
surface sources, except during the lightest rains when the rains would not be
capable of removing much of the non-street surface materials. Analysis of
arsenic, copper, and chromium levels indicated that more of these metals were
on street surfaces before the storms than in the creek flows. There was no
pattern of runoff yield compared with rain magnitude or accumulation period
for these metals. More variability in the runoff yields were observed than
in the street loadings for these metals. Copper and chromium results showed
that larger rains produced greater erosion yields.
Clearly, street cleaning can significantly affect the quality of receiv-
ing waters. The preliminary relationships between street cleaning programs
and the quality of receiving waters are presented in Figure 1. The more
common street cleaning program changes are summarized in Table 16. The
receiving water conditions in this table are related to mass flow values
between the two monitoring stations. The effects would not be as large if
"base flow," especially if polluted by upstream construction or agricultural
practices, was considered. Therefore, the effects relate only to receiving
waters that are totally urban runoff. These values, however, are quite
large. Lead, total solids, and zinc mass flows could be reduced 65 percent,
if street cleaning programs were increased from "no effort" to a "maximum"
effort of three times a week. These pollutants could even be reduced 40
percent by changing street cleaning programs from a monthly to weekly
schedules. Changing from none to quarterly, quarterly to monthly, or weekly
to several times per week would not be nearly as cost effective for the
conditions monitored. If an area is currently cleaned weekly, it would be
much better to leave that schedule alone and start weekly cleaning in other
areas that currently receive monthly, or less cleaning (such as industrial
areas). Similarly, if an area is being cleaned several times a week, that
cleaning effort should be reduced to weekly cleaning and other areas should
be cleaned using the newly available resources.
49
-------
FIGURf 16
LEAP, ZIUC, CHROMIUM,OOTMOPWOSPWATe AMP TOTAL PWOiPHOWJS CKUX VICLPS POK
STORMS OF CELATIVELY SMALL RAINFALL QUANTITIES AMP LONG PRr ACCUMULATION PBZIOBS
10
S
STUPY AB£A CR6BK POLLUTAMT FLOW (ubi/STOZM 6VtNT)
50
-------
Pi6uee 17
Le-AP,IINC,CHe0MIUM,0KTW)FH05PHATE ANP TOTAL PHOSPHORUS CKE£VC YlfcLBS. FOE
STOKMS OP KfcLATIVELY URGE OUNPALL OUANTITieS AND SHORT DBV ACCUMULATION PtRJORS
STUPY AK&A CKtVC POLLUTANT FLOW (LM«/sTORM eVBNT)
51
-------
FIGURE 18
1.0 —
0.5-
0.2-
0.1-
aj
_j
2
i
&
&£
o 005-
¦O
VD
| (X02--
<
o
o oo':
w
3
^ 0.005-
V—
i) I
ID '
ti
)
o.ooz-
0.001->
0.01
ARSEWIC(C0PP£E ANP KJELPAHL MITR06EM CREEK FLOWS F0£
VARIOUS STREET SURFACE LOAPlNG VALUES
0.02
0.05
KJELPAHL MITE06ES1
% I r 1 I I I I
I » I I I t I I
I I i | I I I i |
I I I I I I I I
0.J 1-0 2 5 10 20 50 100 200
STUPY AEEA CREEIC POLLUTANT FLOW (LBS/STOeM EVENt)
i t i i i i i i i
500
1,000
-------
TABLE 16. ESTIMATED EFFECTIVENESS OF STREET CLEANING IN IMPROVING
RECEIVING WATER QUALITY
Percentage change in receiving water lead* yield of current
street cleaning program is:
and if 3 to 7
new program times a
is:
None
Quarterly
Monthly
Weekly
week
none
0
15
25
¦ 100
200
quarterly
10
0
10
80
200
monthly
20
10
0
60
150
weekly
50
40
40
0
0
3 to 7
65
60
60
30
0
times a
week
~Adjustment factors for other parameters include:
• Approximately the same as for lead: total solids and zinc
for most storms and orthophosphate and total phosphorous
for the initial rains of the season only.
• Approximately one-half as sensitive to street cleaning operations
as lead: chromium, copper, arsenic and COD for most storms.
• Not sensitive at all to street cleaning oeprations: Kjeldahl
nitrogen for all storms and orthophosphate and total phosphorous
for most storms.
53
-------
Chromium, copper, arsenic and chemical oxygen demand would also be
affected by street cleaning, but only to about one-half the level as lead.
Changing from monthly to weekly cleaning schedules would, therefore, corre-
spond to a maximum mass flow reduction of about 20 percent for these parameters.
The nutrients Kjeldahl nitrogen, orthophosphates, and total phosphorus would
hardly be affected by street cleaning, except for the initial hard rains when
the phosphorus and orthophosphate street surface loadings and storm runoff
yields may be substantially decreased by street cleaning.
These preliminary results should be viewed only as indicative of impacts
that street cleaning operations can have on the quality of receiving waters.
The results of the second year will include more storm-related data points.
These extra data points will consider a wider variety of storm conditions,
especially smaller storms, than were monitored during the first year. There-
fore, these preliminary results will most likely be significantly different
in the final report.
54
-------
SECTION 6
REFERENCES
Ellefsen, R. and R. Raburn. A Remote Sensing Method For Mapping Urban
Land Cover In A Water Quality Assessment Program. CARTREMS Laboratory,
San Jose State University. Unpublished, 1978.
Farmer, J. and T. Lyon. Lead in Glasgow Street Dirt and Soil.
The Science of the Total Environment 8:89-93, 1977.
McMullen, T. and R. Faoro. Occurrence of Eleven Metals in Airborne
Particulates and Surficial Materials. Journal Air Pollution Control
Association 27:12:1198-1202. December, 1977.
Olson, K. and R. Skogerboe. Identification of Soil Lead Compounds
from Automotive Sources. Environmental Science and Technology
9:3:227-230. March, 1975.
Pitt, R. Demonstration of Non-Point Pollution Abatement through
Improved Street Cleaning Practices. EPA Grant No. S-804432,
289, pp. 1979.
Pitt, R. and G. Amy. Toxic Materials Analyses of Street Surface
Contaminants. EPA-R2-734-383. U.S. Environmental Protection
Agency, Washington, D.C., 135 pp, August, 1973.
Pitt, R. and M. Bozeman. Water Quality and Biological Effects of
Urban Runoff on Coyote Creek, EPA Grant No. R805418010, 66 pp,
June, 1979.
Pitt, R. and G. Shawley. Workplan Demonstration of Non-Point
Pollution Management on Castro Valley Creek. Alameda County
Flood Control and Water Conservation District. Unpublished,
January, 1979.
Sartor, J. and G. Boyd. Water Pollution Aspects of Street Surface
Contaminants. EPA-R2-72-081, U.S. Environmental Protection
Agency, Washington, D.C., 1972.
Shawley, G. First Technical Progress Report, Castro Valley
Demonstration Project. Alameda County Flood Control and
Water Conservation District. Unpublished, May, 1979.
Sylvester, M. and W. Brown. Relation of Urban Land-use and Land-
surface Characteristics to Quantity and Quality of Storm Runoff
in Two Basins in California. Geological Survey Water-Supply Paper
2051, 1978.
55
-------
SECTION F
APPENDIX
The raw data (i.e., street surface loading values, water quality values
and calculations) is not included in this annual report. It is available at
the Alameda County Flood Control and Water Conservation District Office in
Hayward, CA. The raw data will be included in the final report.
56
------- |