I-11_City of Durham Street Sweeping Study March 2021

Water Quality Report #15-004
PUBLIC WORKS March 2021
CITY OF DURHAM
EVALUATION OF POLLUTANTS IN STREET PARTICULATE MATTER
COLLECTED BY A MUNICIPAL STREET SWEEPING PROGRAM IN DURHAM, NC
EXECUTIVE SUMMARY
The City of Durham, NC is subject to several regulatory forces requiring reduction of pollutant loads
entering local surface waters. Among these are local water supply nutrient reduction goals for two
drinking water reservoirs, a turbidity TMDL, multiple stream segments impaired due to heavy metals,
and a NPDES municipal stormwater permit mandating pollution prevention and good housekeeping
practices for municipal operations. One of several management practices implemented by the City in
service to those duties is a municipal street sweeping program. Between 2014 and 2019, City crews
swept over 127,000 miles of curb -lined streets using modern regenerative air sweeper trucks, removing
and disposing of over 15,000 tons of material. The purpose of this study was to quantify pollutant loads
removed from City streets by our municipal street sweeping program. In this way, we may arrive at an
approximation of pollutants prevented from entering the MS4 and surface waters. To evaluate the
street particulate matter and quantify pollutant loads, sweeper truck loads collected from across
multiple regions of the City were sampled on two dates in November 2018 and June 2019. These
sweeper composite samples were analyzed for total Kjeldahl nitrogen (TKN), nitrate+nitrite, total
nitrogen, total phosphorus, total barium, total copper, total chromium, total lead, total zinc, total
selenium, total organic carbon, polycyclic aromatic hydrocarbons, and particle size distribution.
Sampling results identified high concentrations of many pollutants of concern in the sweeper composite
samples. These concentrations were used to calculate pollutant load removal rates, which may be
applied to past and future street sweeping operations data in the City of Durham to estimate pollutant
removal on a monthly or annual basis. This study aims to further inform the public, other municipalities,
and regulatory agencies about the effectiveness of a municipal street sweeping program as an urban
stormwater quality management practice.

Table of Contents
EXECUTIVESUMMARY.................................................................................................................................. 1
ABBREVIATIONS............................................................................................................................................
3
INTRODUCTION.............................................................................................................................................
3
RegulatoryDrivers....................................................................................................................................
3
Purposeand Scope....................................................................................................................................4
METHODS......................................................................................................................................................
4
SiteSelection.............................................................................................................................................4
Sample and Field Data Collection.............................................................................................................
5
SampleAnalysis.........................................................................................................................................
6
Quality Assurance and Quality Control Measures....................................................................................
7
Equipmentand Field Blank...................................................................................................................
7
DuplicateSamples.................................................................................................................................
7
STUDYRESULTS.............................................................................................................................................
7
Quality Control Sample Results — Field and Equipment Blanks................................................................
7
Quality Control Sample Results — Field Duplicates...................................................................................
9
Sampling Location and Conditions..........................................................................................................
10
SPAM Laboratory Analysis Results..........................................................................................................
11
November Nutrients and Metals Results............................................................................................
11
June Nutrients and Metals Results.....................................................................................................13
Study Mean and Median Concentrations...........................................................................................16
November -June Nutrients and Metals Comparison...........................................................................17
Comparison to Other Sweeping Studies.............................................................................................
22
PAHAnalysis........................................................................................................................................
24
ParticleSize Distribution.....................................................................................................................25
Pollutant Loads Removed by Durham Municipal Street Sweeping............................................................
26
Pounds per Mile Removal Calculation....................................................................................................
26
Removal Rate Application to Historical Data..........................................................................................
29
Estimated Pollutant Removal Costs............................................................................................................30
Nutrient Cost Comparisons.....................................................................................................................
31
SUMMARY...................................................................................................................................................
31
REFERENCES................................................................................................................................................ 33
APPENDIX A — Laboratory Qualifier Definitions and Notes........................................................................ 34
2

APPENDIX B—Sweeping Route Maps.........................................................................................................35
ABBREVIATIONS
• SPAM — Street Particulate Matter
• MDL— Method Detection Limit
• MS4 — Municipal Separate Storm Sewer System
• NPDES— National Pollutant Discharge Elimination System
• NCDEQ— North Carolina Department of Environmental Quality
• PQL— Practical Quantitation Limit
• QAPP — Quality Assurance Project Plan
• TKN—Total Kjeldahl Nitrogen
• TMDL—Total Maximum Daily Load
• TSS—Total Suspended Solids
INTRODUCTION
The City of Durham Department of Public Works Street Maintenance Division operates a street
sweeping program responsible for removing debris on all City -maintained curb -lined streets. This
program provides multiple benefits, including aesthetics, pavement preservation, traffic safety,
prevention of MS4 obstruction, and removal of surface water pollutants.
A fleet of eight TYMCO Model 600 regenerative air sweeper trucks are operated on weekdays by drivers
assigned to specific geographical routes across the City. This schedule allows for each City street to be
swept approximately twice per month. High -traffic streets are swept overnight approximately once
daily. During the spring and summer months, each sweeper truck tends to collect one load per day,
composed mostly of fine grit debris. During the fall and winter months, each sweeper truck tends to
collect between three and five loads per day, composed mostly of bulky leaf material and other woody
debris. Once collected, sweeper truck loads are emptied at the City Public Works Operations Center
(PWOC) located at 1100 Martin Luther King, Jr. Parkway in Durham. Collected sweepings are disposed of
in a landfill as municipal solid waste.
In November 2018 and June 2019, joint project teams comprised of Public Works Street Maintenance
and Stormwater Quality staff sampled SPAM collected across multiple regions of the City by street
sweeper trucks. Laboratory analysis of those samples showed elevated concentrations of many
pollutants of concern.
Regulatory Drivers
The City was issued a NPDES Permit (Number NCS000249), allowing discharge from the MS4 to surface
waters, so long as the City acts consistently with the Permit's requirements. Among the Permit
requirements is a mandate for the City to implement pollution prevention and good housekeeping
practices for municipal operations. One of the several pollution prevention Best Management Practices
(BMPs) implemented by the City is a program of street sweeping.
The City is further subject to two approved Total Maximum Daily Loads (TMDLs) with wasteload
allocations. A fecal coliform bacteria TMDL for Northeast Creek was approved in 2003. A turbidity TMDL

for Third Fork Creek (which drains to Jordan Lake) was approved in 2005. The pollutant reductions for
this TMDL are expressed as pounds per year of total suspended solids (TSS). A maximum TSS load of 0.75
tons per day is required in Third Fork Creek.
Falls Lake, located in the Neuse River Basin, is the water supply reservoir for the City of Raleigh and has
the most stringent nutrient reduction goals in North Carolina. The nutrient reduction goals for Falls Lake
require a 40% reduction in nitrogen loads and a 77% reduction in phosphorus loads to the lake. Jordan
Lake, located in the Cape Fear River Basin, is the water supply reservoir for Cary, Apex, and other
communities in Durham and Orange County. The nutrient reduction goals for Jordan Lake differ
according to the lake arm. The Upper New Hope arm has a nitrogen reduction goal of 35% and a
phosphorus reduction goal of 5%.
Additionally, multiple stream segments within the City of Durham have been identified by the Clean
Water Act Section 303(d) Program as impaired due to excess copper or zinc.
• Northeast Creek (three segments impaired for copper or zinc)
• Third Fork Creek (two segments impaired for copper, zinc, or both)
Purpose and Scope
The objective of this study was to evaluate SPAM collected by the City's municipal street sweeping
program and quantify pollutant loads prevented from entering the City's MS4 and surface waters.
METHODS
Site Selection
A representative sample was collected from the material picked up by each sweeper truck during a
single day of sweeping. Each truck represented a different geographical area of the City.
All curb -lined streets to be swept are divided into eight routes by the Public Works Street Maintenance
Division. One of these routes is composed of heavily -trafficked streets, such as Miami Boulevard and
Fayetteville Road, and is swept overnight. The remaining seven routes are composed of all other City -
maintained streets and are swept between the hours of 7:30 a.m. and 3:30 p.m. These daytime routes
are arranged by geographic area of the City.
4

Table 1. Sweeping route characteristics.
Route Durham Region
Notable Streets and Areas Swept
Barbee Chapel, Scott King, Farrington, Renaissance,
lxxx South Massey Chapel, Woodcroft Area, Parkwood Area, Revere
Road, Sedwick
Garrett, South Briggs, Lakeland, Sima, Person, South
2xxx Southeast Alston, Pickett, Ward, South Roxboro, Hillside Park area,
Parkside
3xxx
South -Southwest
Northwest
4xxx
5xxx
Central Durham
6xxx
Northeast
7xxx
East
Night Multiple
Hope Valley, Old Chapel Hill, Cornwallis, South Roxboro,
Swarthmore, Laurel Oaks
American, Cole Mill, Carver, Stoneybrook, Croasdale area,
Old West Durham, Trinity Park area
Forest Hills, Jackson, Swift, East Umstead, Linwood, North
Elizabeth, Glendale, Trinity, Dacian, North Guthrie, West
Markham, West Knox
West Murray, Stadium, North Pointe, Carver
Lumley, Mineral Springs, Highway 98
Miami, Liberty, Main, Ramseur, Angier, Pettigrew, Alston,
Highway 55, Fayetteville, Cornwallis, Highway 54, NC
Central Univ., Duke Univ., Chapel Hill, Morehead,
Academy
Sample and Field Data Collection
This study was performed in accordance with the Street Sweeping Credit Study Quality Assurance Project
Plan (City of Durham, 2018).
Prior to the sample dates, tare weights (full fuel and water tanks) were collected for each sweeper truck,
using the scales located at the City Solid Waste Disposal and Recycling Center.
Data for each sweeper truck were collected and each sweeper load was sampled using the following
procedure on both sampling days.
1. All sweeper trucks departed from the Public Works Operations center to sweep their routinely
assigned routes using standard practices. Each driver recorded their total curb miles swept.
2. When the truck's sweepings capacity became full, the truck was brought to the City of Durham
Fleet Maintenance fuel station to refill the truck's fuel and water tanks to full capacity.
3. The sweeper truck was then weighed at the Solid Waste truck scales. This full load weight was
recorded by the driver.
4. The sweeper truck returned to the Public Works Operations Center to dump its load of
sweepings in the project sampling area, which was prepared with a fresh plastic drop cloth. (See
Figure 1)
5. All sampling rakes, shovels, and compositing containers were freshly cleaned with Liqui-Nox
detergent and triple -rinsed in preparation for each load.
6. A project team member photographed the load and recorded tallies of observable trash items.
5

7. A composite sample for each load was assembled by collecting subsamples of the load pile. The
pile was sampled from each of the four corners and center. These subsamples were placed into
a pre -cleaned compositing container for mixing. A separate compositing container was prepared
for each truck.
8. After each load was sampled, it was pushed aside using heavy machinery and the area was reset
for the next truck load.
9. The above steps 1-8 were repeated for all trucks that dumped multiple loads during the
sampling day. In that event, each of an individual truck load's composite samples were mixed to
create one sample for that truck.
10. After mixing, composite samples were aliquoted into labeled sample containers and stored on
wet ice until delivery to the laboratory.
Figure 1. Sweeper truck preparing to dump load onto a clean plastic ground cloth. (November 2018)
Sample Analysis
November 2018 samples were submitted to Meritech Inc., a North Carolina certified laboratory. June
2019 samples were submitted to ENCO Labs, a North Carolina certified laboratory. Each sediment
sample was analyzed according to the below table.
0

Table 2. Sediment sample analytes and corresponding standard analysis methods and limits.
Analyte
Method Detection
Limit
Practical
Quantitation Limit
Standard Method
Dry weight
0.01%
0.10%
WETS-072
Particle Size Distribution
N/A
N/A
ASTM 1140
Copper
0.1 mg/kg
0.2 mg/kg
EPA SW-846 6010
Zinc
0.2 mg/kg
1 mg/kg
EPA SW-846 6010
Lead
0.1 mg/kg
1 mg/kg
EPA SW-846 6010
Chromium
0.1 mg/kg
0.5 mg/kg
EPA SW-846 6010
Barium
0.1 mg/kg
0.5 mg/kg
EPA SW-846 6010
Selenium
0.1 mg/kg
1 mg/kg
EPA SW-846 6010
Total Phosphorus
1 mg/kg
2 mg/kg
EPA 365.4
Total Organic Carbon
100 mg/kg
500 mg/kg
Walkley Black
Total Kjeldahl Nitrogen
5 mg/kg
20 mg/kg
EPA 351.2
Inorganic Nitrogen
(NO2+NO3)
5 mg/kg
10 mg/kg
EPA 353.2
Polycyclic Aromatic
Hydrocarbons
variable
33 µg/kg
EPA SW-846 8270
Quality Assurance and Quality Control Measures
The quality control samples consisted of one laboratory -blind duplicate composite sample and one field
blank sample per sampling day.
Equipment and Field Blank
The purpose of this sample was to assess the adequacy of the equipment decontamination process and
assess contamination from field conditions during sampling, sample handling, storage, transport, and lab
analysis. One blank was collected immediately before the first sample collection on each day.
Ottawa Sand was chosen as a clean control material for this blank, as suggested by the US EPA's
Hazardous Waste Test Methods (US EPA, 2020). The Ottawa Sand used in this project was purchased as
lab -certified 100% crystalline silica quartz.
Duplicate Samples
Duplicate samples quantify the total variability for each parameter due to all sources: environmental,
sampling, handling/storage, transport, and lab analysis. One truck load was selected at random on each
sampling day. A double -size composite was built from each selected load and split into two samples.
Duplicates were laboratory blind to evaluate laboratory performance.
STUDY RESULTS
Quality Control Sample Results — Field and Equipment Blanks
The project sampling team followed the QAPP to minimize equipment and field blank contamination on
each sampling day.
7

In November, five analytes were measured in concentrations greater than the associated reporting
limits. As seen in Table 3, blank results for total copper, total lead, total phosphorus, total zinc, and total
organic copper were all detected above the laboratory reporting limit. As a general rule, acceptable
sample concentrations should be greater than five times the blank concentrations. Accordingly, seven
November analyses were rejected and not used in data analysis for the study:
• Route 1xxx: total lead
• Route 2xxx: total copper and total lead
• Route 3xxx: total lead
• Route 4xxx: total lead
• Route 2xx: total zinc
• Route 7xxx: total lead
In June, three analytes were measured in concentrations greater than the associated PQL. As seen in
Table 3, blank results for total copper, total lead, and total phosphorus were all detected above the PQL.
Using the above general rule, two June analyses were rejected and not used in data analysis for the
study:
• Route 1xxx: total lead
• Route 6xxx: total lead
Staff suspect that blank contamination could have occurred during setup of the plastic drop cloth. The
sample area is partially unpaved and customarily used for sweepings disposal and other heavy
machinery traffic. Heavy metals such as lead, zinc, and copper are commonly found deposited in areas
subject to these uses. Gusts of wind likely stirred up dust from both the paved and unpaved surfaces,
depositing trace amounts onto the sample drop cloth.
Table 3. Field and equipment blank sample results with associated Reporting Limit (RL), Method Detection Limit (MDL), and
Practical Quantitation Limit (PQL).
Analyte
November (Meritech)
June (ENCO)
RL
Result
MDL
PQL
Result
Qualifier
Total Kjeldahl Nitrogen (mg/kg)
20
5.21
35
38
35
U
Nitrate+Nitrite as Nitrogen
(mg/kg)
3.07
< 3.07
2.0
10
2.0
U
Barium, total (mg/kg)
0.050
< 0.050
0.100
0.500
0.100
U
Chromium, total (mg/kg)
0.050
0.075
0.100
0.500
0.100
U
Copper, total (mg/kg)
0.050
1.92
0.190
0.500
3.53
Lead, total (mg/kg)
0.0125
1.32
0.120
0.500
2.66
Phosphorus, total (mg/kg)
0.500
1.00
2.9
8.0
23
Zinc, total (mg/kg)
0.125
22.4
1.10
2.50
1.47
J,B
Carbon, total organic (mg/kg)
0.0100
5890
500
500
500
U
Orange cells: Sample measurements that exceeded PQL/RL.
Yellow cells: Sample measurements between the MDL and PQL.
J: Qualifier indicates the estimated reported value is between the MDL and PQL.
8: Qualifier indicates the analyte was detected in the associated method blank.
U: Qualifier indicates the analyte was not detected at a concentration greater than the method detection limit
W

Quality Control Sample Results - Field Duplicates
Duplicate samples were collected on both sampling days. In November, the randomly chosen route was
5xxx, which was split into sample IDs 5 and 10. In June, the randomly chosen route was 6xxx, which was
split into sample IDs 6 and 8.
As a general rule, field duplicate results differing no greater than 10% suggest a well -mixed composite
sample with high reproducibility and reasonable laboratory precision. As seen in Table 4, for the
November sampling day, only one duplicate pair were within this 10% benchmark: total phosphorus. For
the June sampling day, only three duplicate pairs were within the 10% benchmark: nitrate+nitrite, total
selenium, and total organic carbon.
Table 4. Duplicate composite sample results for November and June sample days.
Analyte
November
June
Result
Duplicate
% difference
Result
Duplicate
% difference
TKN (mg/kg)
200
243
21.5%
4500
3300
26.7%
Nitrate+Nitrite, as Nitrogen
(mg/kg)
2.12
4.29
102.4%
1.2
1.2
0.0%
Barium, total (mg/kg)
47.5
54.5
14.7%
70.4
81.2
15.3%
Chromium, total (mg/kg)
4.03
6.52
61.8%
23.9
30.1
25.9%
Copper, total (mg/kg)
8.45
13.6
60.9%
59.1
39
34.0%
Lead, total (mg/kg)
6.18
11.3
82.8%
8.3
10.5
26.5%
Phosphorus, total (mg/kg)
477
497
4.2%
760
570
25.0%
Selenium, total (mg/kg)
0.084
0.153
82.1%
0.639
0.619
3.1%
Zinc, total (mg/kg)
60.5
80.5
33.1%
128
160
25.0%
Carbon, total organic (mg/kg)
117000
145000
23.9%
78000
71000
9.0%
These duplicate sample results suggest that the reliability and replicability of the sampling protocol
could be improved with more thorough sub -sampling of each sweepings pile. Though each pile was sub -
sampled in five locations, then thoroughly mixed together in a clean 16-gallon container, the nature of
the pile composition made it physically difficult to sample. Leaves, grass, pine straw, and other large
debris (visible in Figure 2 and Figure 3) often interfered with the ability to scoop and grab sub -samples
also containing finer particulates.
0

Figure 2. Photograph of sweepings collected on Route 5xxx in November 2018.
Figure 3. Photograph of sweepings collected on Route 6xxx in June 2019.
Sampling Location and Conditions
Each truck route is generally characterized in Table 1 above, and more specifically diagrammed in the
maps presented in APPENDIX B — Sweeping Route Maps. The project field team set up in the same
location at the City of Durham Public Works Operations Center (PWOC) on both sampling days. Sampling
was conducted during dry weather, as street sweeping operations do not occur during rain events. On
each sampling date, the most recent rain event was approximately five days prior.
10

Table 5. Summary of weather conditions and previous rainfall for each sampling date.
Sampling Date
20-November-2018
5-June-2019
Location
PWOC
PWOC
Weather Conditions
50° F, partly cloudy
77° F, overcast
Previous Rain Event
15-November-2018
31-May-2019
Previous Rainfall Amount
0.9 inches
0.36 inches
SPAM Laboratory Analysis Results
November Nutrients and Metals Results
Eight sweeper trucks participated in the November sampling day. Two of these trucks were assigned to
overnight routes (2xx and 3xx) and six were assigned to daytime routes (1xxx, 2xxx, 3xxx, 4xxx, 5xxx,
7xxx). The overnight trucks delivered one load each for sampling. Daytime routes 1xxx, 2xxx, 3xxx, 4xxx,
and 7xxx each delivered two loads. Daytime route 5xxx delivered three loads. Trucks delivering multiple
loads were composited together for final sampling. According to City street sweeping personnel, it is
common to deliver multiple loads per day per truck in the fall -winter seasons due to the great amount
of leaves on streets and in curbside gutters. Truck load capacities fill up quicker due to the extra bulk of
these leaves.
Table 6. SPAM analysis of samples collected on November sampling day. No lab qualifiers were attached to the below data.
20-November-2018 Sampling Day
Nitrate
Carbon,
+
Nitrogen,
Barium,
Chromium,
Copper,
Lead,
Phosphorus,
Zinc,
November
TKN
total
o
� dry
Route
(mg/kg)
Nitrite,
total
total
total
total
total
total
total
organic
material
Nitrogen
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
lxxx
280
2.69
282.69
74.8
7.86
17.2
R
551
77.5
140000
46.1%
2xxx
318
4.65
322.65
44.9
4.38
R
R
457
59
156000
46.2%
3xxx
255
5.16
260.16
57.9
4.53
12
R
516
39.2
154000
47.5%
4xxx
404
6.51
410.51
65.5
6.05
9.6
R
496
60.5
175000
40.1%
2xx
214
9.55
223.55
58.9
12.2
21.7
22.1
354
R
123000
59.4%
7xxx
210
3.99
213.99
66.2
4.43
10.2
R
385
72.7
165000
42.3%
3xx
327
7.37
334.37
43
6.35
13.1
14.5
354
97.4
167000
44.9%
5xxx
222
3.205
224.705
51
5.275
11.025
8.74
487
70.5
131000
46.1%
Mean
279
5.39
284.08
58
6.38
14
15.11
450
68
151375
46.6%
StdDev
68
2.30
68.32
11
2.64
4.4
6.70
76.3
18
18384
5.72%
Median
268
4.91
271.43
58
5.66
12
14.5
472
71
155000
46.1%
R: denotes results rejected due to possible blank contamination. See the discussion above on page 7 for detail.
A summary of November sample analyses can be viewed in Table 6. Nitrogen and phosphorus
concentrations are summarized graphically below in Figure 4. Total nitrogen concentration, calculated
by adding TKN concentrations to nitrate+nitrite concentrations, ranged from 213.99 mg/kg to 410.51
mg/kg. Total phosphorus concentrations ranged from 354 mg/kg to 551 mg/kg.
11

November SPAM Nutrient Concentrations
600
Ln
Ln
Ln
Ln W
00
rn 00
500
v
Lq
0
a Ln
00
400 Cn
Ln rn
-he Ln Ln
M Cn
ho N M
C N �1
m
C N �
ci
300 o o
N Ln Cn I�
N M N
U N c N
� N
O
U 200 .
100
0
Nitrogen, total (mg/kg) Phosphorus, total (mg/kg)
■ lxxx ■ 2xxx — 3xxx ■ 4xxx ■ 2xx ■ 7xxx ■ 3xx ■ 5xxx
Figure 4. Nutrient concentrations for November SPAM samples.
Metals concentrations are summarized graphically below in Figure 5. Due to potential contamination
suggested by blank analysis, the following results were rejected: 1) one total copper; 2) five total lead;
and 3) one total zinc. Total barium concentrations ranged from 43 mg/kg to 74.8 mg/kg. Total chromium
concentrations ranged from 4.38 mg/kg to 12.2 mg/kg. Total copper concentrations ranged from 9.6
mg/kg to 21.7 mg/kg. Total lead concentrations ranged from 8.74 mg/kg to 22.1 mg/kg. Total zinc
concentrations ranged from 39.2 mg/kg to 97.4 mg/kg. Total selenium was also analyzed, but not
detected above method detection limits for any sample. Total organic carbon concentrations ranged
from 123,000 mg/kg to 175,000 mg/kg.
12

November SPAM Metals Concentrations
120
100
80
tin N
E ID
I
c n co m ID
60
+�
c O1
a) M
� a N
O r
u 40
..j N
N ry N
n
20 N "
00Oo
IA V1 � e'I ti
0
Barium, total (mg/kg) Chromium, total Copper, total (mg/kg) Lead, total (mg/kg) Zinc, total (mg/kg)
(mg/kg)
■ lxxx ■ 2xxx - 3xxx ■ 4xxx ■ 2xx ■ 7xxx ■ 3xx ■ 5xxx
Figure 5. Metals concentrations for November SPAM samples.
June Nutrients and Metals Results
Seven sweeper trucks participated in the June sampling day. An eighth truck was in service, but a driver
was unavailable to operate it on the sampling day. Two of these trucks were assigned to overnight
routes (1xx and 3xx) and five were assigned to daytime routes (1xxx, 2xxx, 4xxx, 5xxx, 6xxx). All trucks
delivered a single load each for the full day of sweeping.
According to City street sweeping personnel, it is common to deliver one load per day per truck in the
spring -summer seasons because these loads tend to be mostly composed of fine grit and smaller debris
such as grass clippings and pine straw.
13

Table 7. SPAM analysis of samples collected on June sampling day. Yellow highlighted values are imputed by software to replace
values U-qualified by the lab, indicating non -detection above the method detection limit.
5-June-2019 Sampling Day
June
Route
TKN
(mg/kg)
Nitrate+
Nitrite,
Nitrogen
(mg/kg)
Nitrogen,
total
(mg/kg)
Barium,
total
(mg/kg)
Chromium,
total
(mg/kg)
Copper,
total
(mg/kg)
Lead,
total
(mg/kg)
Phosphoru
s,total
(mg/kg)
Zinc,
total
(mg/kg
)
Carbon,
total
organic
(mg/kg)
%dry
material
lxx
2600
3.7
2603.7
56.2
21.9
48.4
42.1
380
116
42000
89.6%
2xxx
3500
1.2
3501.2
56.1
8.14
20.8
13.9
500
86.8
53000
74.8%
3xx
2300
3.4
2303.4
60.0
18.9
42.7
22.8
350
151
33000
85.5%
4xxx
9000
4.8
9004.8
77.5
8.83
47.3
38
860
136
45000
89.3%
1xxx
2700
1.2
2701.2
49.3
11.5
31
R
500
81.7
41000
89.9%
5xxx
2700
1.2
2701.2
45.5
11.5
278
23.5
510
127
37000
91.6%
6xxx
3900
1.2
3901.2
75.8
27
49.05
R
665
144
74500
65.2%
Mean
3814
2.4
3817
60
15
74
28
538
120
46500
83.7%
StDev
2354
1.5
2355
12
7
91
12
175
27
13853
9.9%
Median
2700
1.2
2701.2
56.2
11.5
47
24
500
127
42000
89.3%
R: denotes results rejected due to possible blank contamination. See the discussion above on page 7 for detail.
Yellow shaded values were imputed using ProUCL 5.1 due to the prevalence of results less than RL/PQL.
A summary of June sample analyses can be viewed in Table 7. Nitrogen and phosphorus concentrations
are summarized graphically below in Figure 6. Four nitrate+nitrite sample results were flagged by the
laboratory as undetected above the method detection limit. These values were imputed using ProUCL
5.1 software. Total nitrogen concentration, calculated by adding TKN concentrations to nitrate+nitrite
concentrations, ranged from 2,300 mg/kg to 9,000 mg/kg. Total phosphorus concentrations ranged from
350 mg/kg to 860 mg/kg.
14

June SPAM Nutrient Concentrations
10000 °Q
a
O
O
9000
8000
7000
on
own
6000
E
C
°
5000
N
ca
r;
N
O
C
u
4000
0
Ln
rn
m
C
^ m
N N
O
U
m
0 0
n n
3000
N mo
N N
m
2000
. A
UO Ln
1000 00 O o 00 O �o
= M = M M
Nitrogen, total (mg/kg) Phosphorus, total (mg/kg)
■ 1xx ■ 2xxx ■ 3xx 4xxx ■ 1xxx ■ 5xxx ■ 6xxx
Figure 6. Nutrients concentrations for June SPAM samples.
Metals concentrations are summarized graphically below in Figure 7. Due to potential contamination
suggested by blank analysis, two total lead results were rejected. Total barium concentrations ranged
from 45.5 mg/kg to 77.5 mg/kg. Total chromium concentrations ranged from 8.14 mg/kg to 27 mg/kg.
Total copper concentrations ranged from 20.8 mg/kg to 278 mg/kg. Total lead concentrations ranged
from 13.9 mg/kg to 42.1 mg/kg. Total zinc concentrations ranged from 81.7 mg/kg to 151 mg/kg. Total
selenium was also analyzed, but not detected above method detection limits for any sample. Total
organic carbon concentrations ranged from 33,000 mg/kg to 74,500 mg/kg.
15

300
250
200
50
0
June SPAM Metals Concentrations
Ln Du-
r` Ui
N -1
L6.6
Ln Ln
M�
m
O1 rn r`
N 06 Ln lJ7 N
Barium, total (mg/kg)
Chromium, total
(mg/kg)
00
N
�t m o
r`
v v� v v
00
ap oo m Ln
p m �i M
I � 1
Copper, total (mg/kg) Lead, total (mg/kg)
■ 1xx ■ 2xxx ■ 3xx ■ 4xxx ■ lxxx ■ 5xxx ■ 6xxx
Figure 7. Metals concentrations for June SPAM samples.
rn
rn
N
L
t1
i
r`
00
00
Zinc, total (mg/kg)
Study Mean and Median Concentrations
To generalize observations of this study, the mean and median were calculated for each SPAM
constituent across the November and June analytical results. A summary of results may be seen in Table
8. Graphic representations are provided in Figure 8 and Figure 9.
Of note are the results for total phosphorus, total barium, total lead, and total chromium. The median
and mean for each of these datasets are similar, suggesting normal distribution of these data.
Table 8. Summary of data across the November and June sampling dates.
Nitrate
Carbon,
+
Nitrogen,
Barium,
Chromium,
Copper,
Lead,
Phosphorus,
Zinc,
TKN
total
Nitrite,
total
total
total
total
total
total
total
(mg/kg)
Nitrogen
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
organic
Img/kg)
(mg/kg)
Mean
1929
4.0
1933
59
11
44
23
491
94.2
102433
Median
404
3.7
410.5
58
8.1
21
22
496
84.3
123000
Minimum
210
1.2
213.99
43
4.38
9.6
8.74
350
39.2
33000
Maximum
9000
9.6
9005
77.5
27
278
42.1
860
151
175000
16

2500
2000
1500
1000
Mean & Median Nutrients Concentrations
1933
500 410.5
Em
Nitrogen, total (mg/kg)
■ Mean I Median
491 496
M M
Phosphorus, total (mg/kg)
Figure 8. Column plot of mean and median nutrients concentrations across November and June datasets.
Mean & Median Metals Concentrations
100
94.2
90
84.3
80
70
59 58
60
50
44
40
30
21
23 22
20
10
11 8.1
■-
Barium, total Chromium, total Copper, total
Lead, total Zinc, total
(mg/kg) (mg/kg) (mg/kg)
(mg/kg) (mg/kg)
■ Mean r Median
Figure 9. Column plot of mean and median metals concentrations across November and June datasets.
November -June Nutrients and Metals Comparison
Due to the observation that SPAM loads varied in physical texture and composition and allowing for
seasonal differences in the accumulated matter, a statistical comparison of the November and June data
is necessary.
Assuming that environmental datasets are generally nonparametric and that all observations are
independent, a Wilcoxon-Mann-Whitney test was used to assess the similarity of the November and
June sample results. A summarized comparison overview of this analysis, with p-values, may be viewed
in Table 9.
17

The November and June datasets were statistically similar at the 95% confidence level for:
• Barium, total (p-value 0.7789)
• Lead, total (p-value 0.1429)
• Phosphorus, total (p-value 0.3969)
The November and June datasets were statistically different at the 95% confidence level for:
• Total Kjeldahl Nitrogen (p-value 0.0003)
• Nitrate+Nitrite, nitrogen (p-value 0.0205)
• Nitrogen, total (p-value 0.0003)
• Chromium, total (p-value 0.0037)
• Copper, total (p-value 0.0012)
• Zinc, total (p-value 0.0023)
• Organic Carbon, total (p-value 0.0003)
Table 9. Comparison of November and June mean analyte values. Orange -shaded values denote statistically significant
difference between data sets (95% confidence level).
Analyte
Units
November
June
p-value
Total Kjeldahl Nitrogen
mg/kg
279
3814
0.0003
Nitrate+Nitrite, Nitrogen
mg/kg
5.39
2.4
0.0205
Nitrogen, total
mg/kg
284.08
3817
0.0003
Barium, total
mg/kg
58
60
0.7789
Chromium, total
mg/kg
6.38
15
0.0037
Copper, total
mg/kg
14
74
0.0012
Lead, total
mg/kg
15.11
28
0.1429
Phosphorus, total
mg/kg
450
538
0.3969
Zinc, total
mg/kg
68
120
0.0023
Organic Carbon, total
mg/kg
151375
46500
0.0003
Notably, the datasets for total barium, total lead, and total phosphorus are both statistically similar and
have nearly normal distributions (as discussed above and seen in in Figure 8 and Figure 9).
Graphical representations of dataset distributions are presented in Figure 10 through Figure 17. These
box -and -whisker plots also support the above analysis of similarity and difference between the datasets.
18

i0000
OAK
Y 6000
z
w
0 4000
FIGIO
850
750
450
350
June Novemher
Month
Figure 10. Box -and -whisker plot for June and November total nitrogen datasets.
June November
Month
Figure 11. Box -and -whisker plot for June and November total phosphorus datasets.
19

so
70
Y
�r
C
—' 60
E
_
m
m
50
40
211
24
20
Y
E
June T November
Month
Figure 12. Box -and -whisker plot for June and November total barium datasets.
E 16 1
E
0
L
12
S
4
June November
Month
Figure 13. Box -and -whisker plot for June and November total chromium datasets.
20

300
250
200
m
Y`
rm
C
—' 150
v
a
a
0
V
100
50
0
45
35
Y
E 25
d
J
June November
Month
Figure 14. Box -and -whisker plot for June and November total copper datasets.
15
5
June
Month
November
Figure 15. Box -and -whisker plot for June and November total lead datasets.
21

60
30
175000
150000
125000
Y
E 100000
u
O
H
75000
50000
25000
June November
Month
Figure 16. Box -and -whisker plot for June and November total zinc datasets.
June
November
Month
Figure 17. Box -and -whisker plot for June and November total organic carbon datasets.
Collectively, this statistical analysis suggests that the fall -winter sweeping season (represented by the
November sampling day) is different enough from the spring -summer sweeping season (represented by
the June sampling day) that when using the results to estimate pollutant load removal, seasonality
should be taken into consideration.
Comparison to Other Sweeping Studies
A summary of this study's results compared with literature values can be viewed below in Table 10. The
mean TKN value for November was an order of magnitude less than literature values, while that for June
slightly exceeded the greatest value found. No comparable nitrate+nitrite data were found in the
22

literature. Both mean total nitrogen values for this study fell within the range of values found in a similar
study performed by Charlotte -Mecklenburg. A range of total phosphorus values were found across the
literature, from 50 mg/kg to 3,309 mg/kg. This study's mean total phosphorus values fall within the
above range.
A single study from Pensacola, FL provided a comparison point for total barium. This study's mean
barium values were approximately double that of the Pensacola study. Total chromium values in the
literature ranged from 9.7 mg/kg to 87 mg/kg. This study's mean chromium values overlap at the low
end of that range. Total copper values in literature ranged from 12 mg/kg to 215 mg/kg. This study's
mean copper values are within the low end of that range. Total lead values in literature ranged from 5
mg/kg to 407 mg/kg. This study's mean lead values are within the low end of that range. Total zinc
values in literature ranged from 22 mg/kg to 464 mg/kg. This study's mean zinc values are within the low
end of that range. Finally, total organic carbon values were found in one other study, ranging from 800
mg/kg to 1200 mg/kg. This study's mean total organic carbon values are several orders of magnitude
greater.
Table 10. SPAM analysis results for this study compared with literature values.
Study/Locale
Total
Nitrogen
Kitrogeln
Nitrate
+
Nitrite,
Nitrogen
Nitrogen,
total
Barium,
total
Chromium,
total
Copper,
total
Lead,
total
Phosphorus,
total
Zinc,
total
Organic
Carbon
total
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
November
279
5.39
284.08
58
6.38
14
15.11
450
68
151375
Durham, NC
June
3814
2.4
3817
60
15
74
28
538
120
46500
Mean
1929
4.0
1933
59
11
44
23
491
94.2
102433
CDOT
156
17
52
5
119
22
WWTP
1227
14
49
10
364
73
Charlotte, NC1
P&R
806
16
12
6
50
113
LM
4843
13
215
26
304
141
USEPA2
1571
87
143
407
464
Baltimore, MD'
1,477 to
3,067
1,033 to
3,309
West
3090
24.1
34.6
51.5
648
180
1000
Seattle, WA4
Southeast
3170
33.3
37.6
63.5
633
176
1200
Duwamish
3540
62
72.6
192
516
211
800
Pensacola, FL5
27.35
9.7
9.62
19.68
38.48
1 (University of North Carolina, Charlotte, 2017)
2 (Sartor, 1974)
3 (Center for Watershed Protection, 2008)
4 (Seattle Public Utilities, 2009)
5 (Liebens, 2001)
In summary, while this study suggests total nitrogen and total organic carbon concentrations in Durham
street particulate matter are relatively high compared to other urban areas, other nutrient and metals
concentrations are on the low end of comparable ranges.
23

PAH Analysis
All samples from both sampling days were additionally submitted for polycyclic aromatic hydrocarbons
(PAHs) analysis.
In November, PAH analysis included nineteen different chemicals across nine samples, plus the
field/equipment blank. Only three detections were found, all from the Route 2xxx sample. No detections
were found in the blank sample. These detections, below in Table 11, amount to only 1.7% of the total
sample pool for November.
Table 11. PAH detections for November samples. All detections were found in Route 2xxx samples.
PAH Chemical
Result (mg/kg)
MDL (mg/kg)
Fluoranthene
0.7
0.167
Phenanthrene
0.717
0.167
Pyrene
0.557
0.167
In June, PAH analysis included eleven different chemicals across eight samples (difference due to lab
availability), plus the field/equipment blank. Except for naphthalene, which was not detected in any
sample, low concentrations of PAH chemicals were detected in each load sample. No detections were
found in the blank sample. A summary of PAH means in the June samples is below in Table 12.
Table 12. PAH sample analysis summary for June samples.
PAH Chemical
Concentration (mg/kg)
Mean
Median
Std. Dev.
Min.
Max.
Anthracene
0.16
0.068
0.17
0.035
0.51
Ben zo(a)anth race ne
0.34
0.15
0.36
0.056
0.093
Benzo(a)pyrene
0.41
0.19
0.45
0.067
1.1
Benzo(g,h,i)perylene
0.43
0.24
0.43
0.090
1.1
Chrysene
0.57
0.31
0.58
0.12
1.5
Dibenzo(a,h)anth racene
0.22
0.058
0.34
0.022
0.96
Fluoranthene
1.1
0.50
1.1
0.23
2.7
Indeno(1,2,3-cd)pyrene
0.32
0.17
0.33
0.054
0.81
Phenanthrene
0.50
0.24
0.52
0.12
1.3
Pyrene
0.89
0.43
0.91
0.18
2.2
June total PAH (IPAH) concentration in Durham SPAM can be estimated at 4.9 mg/kg, as a sum of means
across each sample. Menzie et al. (1992) present several ranges for soil PAH content across multiple
geographic regions. That article, which surveys road dust PAH analysis results across the United States,
presents a range of 8 mg/kg to 336 mg/kg. Results from other cities are below in Table 13 alongside
Durham's results. According to the results from this City of Durham study, 4.9 mg/kg FPAH appears to be
on the low end of observations for urban streets in other localities.
24

Table 13. Total PAH content in soils of different locales (Menzie CA, 1992).
Location
Mean EPAH concentration
Durham, NC
4.9 mg/kg
New Orleans, LA
6 mg/kg
Boston, MA
19 mg/kg
London, UK
18 mg/kg
Torino, Italy
2 mg/kg
Particle Size Distribution
SPAM samples from both the November and June sampling days were assessed for particle size
distribution. Standard sieve sizes and methods were used for both sample groups. All organics (leaves,
twigs, grass, pine straw, etc.) were initially removed from samples using a 3/8" sieve. A summary of
particle size analysis can be seen below in Table 14.
Table 14. Particle size distribution for November and June SPAM samples.
Route
Fine
Gravel %
Coarse Sand
%
Medium
Sand %
Fine
Sand
Fines
1xxx
0.1
16.5
60.7
21.4
1.3
2xxx
0.1
30.8
50.5
16.3
2.3
3xxx
0.1
20.7
58.7
19.83
0.67
4xxx
0.3
32.8
56.8
8.1
2.0
v
5xxx
1.1
28.2
49.4
20.0
1.4
0
z
2xx
0.1
15.5
54.2
29.2
1.0
7xxx
0.5
19.6
63.7
13.2
3.0
3xx
0.1
22.8
60.2
16.33
0.57
(blank)
0.0
0.0
98.5
0.68
0.82
1xx
19.5
2.7
20.1
46.3
11.4
2xxx
23.3
6.1
25.8
35.8
9.1
3xx
18.6
4.8
31.5
35.2
9.9
4xxx
42.7
4.0
24.8
23.1
5.5
=
1xxx
20.9
6.0
34.9
30.7
7.4
6xxx
54.6
2.8
15.1
19.0
8.5
5xxx
27.0
6.3
25.7
30.9
10.1
(blank)
0.0
1 0.0
1 99.7
0.2
0.1
The sand category accounted for the majority of particles in November and June samples. In November,
medium sand particles generally accounted for 50%-60% of each sample. However, June samples
tended to be more distributed between fine gravel, medium sand, and fine sand. June samples also
generally had a greater proportion of fines (silt/clay).
2S

These findings are consistent with observation of the physical characteristics of truck loads on both
sampling days.
According to a review of the literature, nutrients and metals have an affinity for finer particles in soil and
street dust (Center for Watershed Protection, 2006). This supports Durham's study results generally
finding greater concentrations of each pollutant in the June samples, which were finer in grain
composition than the November samples.
Pollutant Loads Removed by Durham Municipal Street Sweeping
Pounds per Mile Removal Calculation
To estimate the City-wide impact of the municipal street sweeping program, the following calculation
was used to determine pounds removed per curb mile swept:
Lp = (C * M * 2.20462 * 10-6)/D
where:
Lp = pollutant load removed per curb mile swept (Ibs./mi)
C= pollutant concentration (mg/kg)
M = dry mass of sweepings collected (kg)
2.20462*10-6 = unit conversion factor (mg to Ibs.)
D = distance swept (mi)
26

Table 15 below shows the results of the above calculations, performed for each pollutant based on
mean November and mean June values. November and June values were not combined because the
statistical analysis discussed above suggests the datasets to be too different.
27

Table 15. Summary of calculated pollutant removal rates via municipal street sweeping.
November
June
Mean
Concentration
(mg/kg)
Removed
(Ibs.)
Removed
per mile
swept
(Ibs./mi)
Mean
Concentration
(mg/kg)
Removed
(Ibs.)
Removed
per mile
swept
(Ibs./mi))
Total Kjeldahl Nitrogen
279
0.653
0.067
3814
12.70
0.75
Nitrate+Nitrite, Nitrogen
5.39
0.013
0.0013
2.4
0.0079
0.00047
Nitrogen, total
284.08
0.665
0.068
3817
12.71
0.75
Barium, total
58
0.135
0.014
60
0.20
0.012
Chromium, total
6.38
0.0150
0.0015
15
0.051
0.0030
Copper, total
14
0.032
0.0032
74
0.25
0.014
Lead, total
15.11
0.0354
0.0036
28
0.093
0.0055
Phosphorus, total
450
1.1
0.11
538
1.79
0.11
Zinc, total
68
0.16
0.016
120
0.40
0.024
Organic carbon, total
151375
354.6
36
46500
155
9.1
Mean Load Dry Weight (Ibs.)
2342
3330
Mean Miles Swept per Day
9.8
17
The above results in
28

Table 15 show the calculated estimates for total pounds of each pollutant removed during a single day
of sweeping in November and in June.
November removal per mile rates ranged from 0.0013 lbs./mi to 36 lbs./mi. Total organic carbon was
the most efficiently removed pollutant, at 36 lbs./mi. Total phosphorus was the most efficiently
removed nutrient, at 0.11 lbs./mi. Among the nutrients, total phosphorus was removed at
approximately 1.6 times the rate of total nitrogen. Among the metals, total zinc was also the most
efficiently removed, at 0.016 lbs./mi.
June removal per mile rates ranged from 0.00047 lbs./mi to 9.1 lbs./mi. Similar to the previous sampling
session, total organic carbon was the most efficiently removed pollutant, at 9.1 lbs./mi. The total
nitrogen removal rate was greater in June, at 0.75 lbs./mi. and nearly seven times the total phosphorus
removal rate, which was again 0.11 lbs./mi. Zinc was again the most efficiently removed metal in June,
at 0.024 lbs./mi.
Removal Rate Application to Historical Data
The City of Durham Department of Public Works routinely maintains records of curb miles swept and
SPAM tonnage disposal. These records are tabulated monthly, so seasonal variability in removal rates
may be adjusted easily.
The removal rates calculated above were applied to five years of historical records. Each sampling
event's calculated rates were applied separately to the historical data to preserve the differences in
seasonality. If monthly breakdowns of the historical data are obtained, a more precise estimate may be
calculated. Table 16 contains calculations of estimated pounds removed by the City's street sweeping
program from 2014 to 2019. Estimates were calculated by multiplying removal rate by total curb miles
swept.
Table 16. Application of November and June removal rates applied to five years of historical street sweeping data.
November
Removal rate
(lbs/mi)
Estimated Pounds Removed
2014-2015
2015-2016
2016-2017
2017-2018
2019-2019
Total Kjeldahl Nitrogen
0.067
1925
2109
1407
1368
1693
Nitrate+Nitrite, Nitrogen
0.0013
37
41
27
26
33
Nitrogen, total
0.068
1962
2150
1434
1394
1726
Barium, total
0.014
399
437
292
284
351
Chromium, total
0.0015
44
48
32
31
39
Copper, total
0.0032
94
103
68
66
82
Lead, total
0.0036
104
114
76
74
92
Phosphorus, total
0.11
3108
3405
2271
2208
2734
Zinc, total
0.016
470
515
344
334
414
Organic carbon, total
36
1045401
1145555
764077
742838
919520
Curb miles swept
28,892
31,660
21,117
20,530
25,413
June Removal
rate(lbs/mi)
Estimated Pounds Removed
2014-2015
2015-2016
2016-2017
2017-2018 1
2019-2019
29

Total Kjeldahl Nitrogen
0.75
21589
23657
15779
15340
18989
Nitrate+Nitrite, Nitrogen
0.00047
14
15
10
10
12
Nitrogen, total
0.75
21602
23672
15789
15350
19001
Barium, total
0.012
340
372
248
242
299
Chromium, total
0.0030
87
95
64
62
77
Copper, total
0.014
418
458
306
297
368
Lead, total
0.0055
159
174
116
113
140
Phosphorus, total
0.11
3044
3336
2225
2163
2678
Zinc, total
0.024
681
746
498
484
599
Organic carbon, total
9.1
263187
288402
192362
187015
231496
Curb miles swept
28,892
31,660
21,117
20,530
25,413
Over this five-year period, total nitrogen removal can be estimated between 81665 lbs. and 95,414 lbs.
Total phosphorus removal can be estimated between 13,446 lbs. and 13,726 lbs. Total barium removal
can be estimated between 1,501 lbs. and 1,762 lbs. Total chromium removal can be estimated between
195 lbs. and 385 lbs. Total copper removal can be estimated between 413 lbs. and 1,847 lbs. Total lead
removal can be estimated between 461 lbs. and 701 lbs. Total zinc removal can be estimated between
2,078 lbs. and 3,009 lbs. Finally, total organic carbon removal can be estimated between 127,612 lbs.
and 1,162,462 lbs.
Estimated Pollutant Removal Costs
The above analyses, in addition to cost data maintained by the City of Durham Department of Public
Works, allows for calculation of per -pound removal costs. These estimated per -pound costs allow
comparison between various methods of controlling pollutants or treating surface or drainage waters.
Table 17 below contains calculated pollutant removal costs for Durham's street sweeping program
based on a $55.00 per mile sweeper operation cost, a figure provided by the Public Works
Administrative Analyst.
Table 17. Estimated costs of per -pound pollutant removal achieved by Durham municipal street sweeping program. Cost
estimates are based on $55/mile operation cost.
November
June
Removed per
mile swept
Cost per pound
Removed per
mile swept
Cost per
pound
ML
Total Kjeldahl Nitrogen
0.067
$ 825.65
12.70
$ 4.33
Nitrate+Nitrite, Nitrogen
0.0013
$ 42,684.74
0.0079
$ 6,922.47
Nitrogen, total
0.068
$ 809.98
12.71
$ 4.33
Barium, total
0.014
$ 3,982.65
0.20
$ 274.99
Chromium, total
0.0015
$ 36,040.71
0.051
$ 1,072.70
Copper, total
0.0032
$ 16,985.84
0.25
$ 223.50
Lead, total
0.0036
$ 15,224.80
0.093
$ 588.56
Phosphorus, total
0.11
$ 511.33
1.79
$ 30.71
Zinc, total
0.016
$ 3,378.11
0.40
$ 137.22
30

Organic carbon, total 36 $ 1.52 1 155 1 $ 0.36
Cost of operation $55.00 per mile
Based on the above calculations, total organic carbon, total nitrogen, and total phosphorus are the most
cost-effective on a per -pound removal basis. This is true for both the November and June sampling days.
Metals are all generally more expensive to remove, but far less so in the June sample than in November.
Again, this is likely due to particle size distribution differences between the sampling events. Metals
generally have more affinity for smaller SPAM particles. Since the June samples tended to be more
predominately finer particles, the June efficiencies for metals tend to be more favorable.
Nutrient Cost Comparisons
Literature data for nutrient removal costs were compared to the results of this study. For the City of
Durham's street sweeping program, total nitrogen estimated removal costs range from $4.33 to $809.98
per pound. Total phosphorus estimated removal costs range from $30.71 to $511.33 per pound.
Compared to the literature values summarized in Table 18 below, these costs measure favorably against
most other types of stormwater BMPs. Also, Durham's sweeping program appears to be more cost-
effective than the average street sweeping costs in the literature.
Table 18. Nutrient removal cost per pound comparisons of various stormwater BMPs.
Dewberry
CWP - James Myer
IVIDEI
Average
TN
TP
7N
7P
TN
TP
7N
TP
6ioretention (new)
$ 254.00
$2,120.00
$ 603.75
$ 3,934,84
$ 254.00
$ 2,980-00
$ 370.58
$ 3,211.61
Bioswale
$ 225.00
$ 309.13
$ 2,653.91
S 250.00
$ 279.57
$ 1,439.46
Dry Detention Pond
$4,597.20
521,143.16
54,250.00
$4,423.60
$21,143.16
Dry Retention Pond
$ 773.00
$2,901.00
$ 773.00
$ 2,901.00
Filtering (sand, above and below ground
averaged)
$ 227.00
$1,022.40
S 4,741,26
$ 540.00
$ 3,136.00
$ 596.47
S 3,938.63
Forest Buffer
$ 55.00
5 759.00
$ 105.86
$ 1,851.00
S 106.00
$ 759.00
$ 88.95
$ 1,123.00
Grass Buffer
S 3.00
$ 3,00
COrrettiOn of crass -connections
$ 17.70
$ 70.79
$1,525.00
S 771.35
$ 70.79
Sewer repair
$ 8-86
$ 35-43
$1,525-00
$ 766.93
$ 3S.43
Impervious Surface Reduction
$2,439.05
$ 7,534.09
$ 972.00
$ 4,118.00
$1,705.53
$ 5,826.05
Infiltration (with and without sand,
averaged)
$ 1D0.00
$2,643.00
$ 492.64
$ 3,325.22
$ 275.00
$ 299.21
$ 2,994.11
Permeable Pavement {with and without
5,and, averaged}
$1,289-00
$9,351-00
$6,045.20
$33,161,90
$1,288-00
$ 9,351-00
$2,873,73
$17,297.93
Street Sweeping
$ 35.00
$ 350.00
$1,389.99
5 3,474,98
$4,737.00
522,006.00
$2,06d.04
$ a,610.33
Tree Planting
$ 523.00
$7,377.00
$ 657.58
S 9,621.48
S 523.00
S 7,377.00
$ 567.86
$ 8,225.16
Vegetated Open Channels
$ 796.43
$ 7,325.8D
$ 796.43
$ 7,325.80
Wet Ponds and Wetlands (new)
$3,28&00
$2,700.00
$ 696.63
$ 2,947,91
$ 448.00
$ 2,330.00
$1,476.88
$ 2,625.97
Wetlands (retrofit)
$ 448-00
$2,330-00
$1,160.28
S 6,670,36
$ 8ad_14
$ 4,500.18
1 (Leinhart, 2014)
2 (Center for Watershed Protection, 2013)
3 (Maryland Department of the Environment, 2013)
SUMMARY
The City of Durham conducted this study to quantify the amounts of various stormwater pollutants
removed from City streets via its municipal street sweeping program. With the assistance of the Public
Works Street Sweeping Crew, composite samples of street sweepings were collected on two separate
31

days. A November sampling date was selected as a representative of the fall and winter seasonal SPAM
accumulations. A June sampling date was selected as a representative of the spring and summer SPAM
accumulations. These samples were analyzed for nutrients, metals, total organic carbon, hydrocarbons,
and particle size distribution.
The statistical analysis supports that total phosphorus, total lead, and total barium datasets were
similarly distributed across the November and June sampling dates. The opposite holds true for the
remaining pollutants. This seasonal variability, supported by the data, suggests that calculating pollutant
load removals should be done on a per -season basis, rather than averaging results over an entire year.
In addition to seasonality, other variables such as street pavement condition, sweeper vehicle operation
speed, sweeper vehicle maintenance level, and the amount of rainfall prior to sweeping can all impact
pollutant removal efficiency. Furthermore, though this study design sought to collect samples from all
across the City on each sampling day, a relatively small number of samples were collected and analyzed.
These results are a snapshot of a program that operates daily, weather permitting.
Based on the removal rates calculated by this study, average annual pollutant removals between 2014
and 2019 are estimated below. These ranges reflect the seasonal variability observed in this study's
datasets.
• Total nitrogen removal ranged from 1,733 lbs. to 19,083 lbs.
• Total phosphorus removal ranged from 2,689 lbs. to 2,745 lbs.
• Total barium removal ranged from 300 lbs. to 352 lbs.
• Total chromium removal ranged from 39 lbs. to 77 lbs.
• Total copper removal ranged from 83 lbs. to 369 lbs.
• Total lead removal ranged from 92 lbs. to 140 lbs.
• Total zinc removal ranged from 416 lbs. to 602 lbs.
• Total organic carbon removal ranged from 232,492 lbs. to 923,478 lbs.
In comparison to other stormwater BMPs, municipal street sweeping is among the most efficient (cost
per pound removed) strategies. This is in addition to the many other benefits provided by a municipal
street sweeping program, such as pavement preservation, infrastructure integrity, aesthetics, public
safety, and property safeguarding.
32

REFERENCES
Center for Watershed Protection. (2006). Research in Support of an Interim Pollutant Removal Rate for
Street Sweeping and Storm Drain Cleanout Activities. Center for Watershed Protection.
Center for Watershed Protection. (2008). Deriving Reliable Pollutant Removal Rates for Municipal Street
Sweeping and Storm Drain Cleanout Programs in the Chesapeake Bay Basin. Center for
Watershed Protection.
Center for Watershed Protection. (2013). Cost -Effectiveness Study of Urban Stormwater BMPs in the
James River Basin. Center for Watershed Protection.
City of Durham. (2018). Street Sweeping Credit Study Quality Assurance Project Plan.
Leinhart, S. R. (2014). How much and how long will it take to achieve NNC compliance? Retrieved from
www.dewberry.com.
Liebens, J. (2001). Heavy metal contamination of sediments in stormwater management systems: The
effect of land use, particle size, and age. Environmental Geology(41), 341-351.
Maryland Department of the Environment. (2013). Cost Efficiency and Other Factors in Urban
Stormwater BMP Selection.
Menzie CA, P. B. (1992). Exposure to carcinogenic PAHs in the environment. Environmental Science and
Technology(26), 1278-1284.
Sartor, J. G. (1974). Water Pollution Aspects of Street Surface Contaminants. Journal of the Water
Pollution Control Federation, 46(3), 458-467.
Seattle Public Utilities. (2009). Seattle Street Sweeping Pilot Study. Seattle, WA: Seattle Public Utilities.
University of North Carolina, Charlotte. (2017). Examination of the Pollutant Removal Efficiency and the
Cost Effectiveness of an Active Parking Lot Sweeping Program, Charlotte, NC. Charlotte, NC: UNC
Charlotte.
US EPA. (2020, November 20). The SW-846 Compendium. Retrieved from US EPA Website:
https://www.epa.gov/hw-sw846/sw-846-compendium
33

APPENDIX A— Laboratory Qualifier Definitions and Notes
Category
Qualifier
Code
Description
B
Results based on colony count outside of the acceptable range and should be used with caution.
B1
Countable membranes w/ less than 20 colonies. Reported value is estimated or is a total of the
counts on all filters reported per 100mL.
Counts from all filters were zero. The value reported is based on the number of colonies per
B2
100mL that would have been reported if there had been one colony on the filter representing the
largest filtration volume ( reported as "<" value).
B: Microbial tests
B3
Countable membranes with more than 60 (or 80) colonies. The value reported is calculated using
the count from the smallest volume filtered (reported as a ">" value).
membrane filter (MF)
counts
B4
Filters have counts of both >60 (or >80) and <20. Reported value is a total of the counts from all
countable filters reported per 100mL.
Too many colonies were present / too numerous to count (TNTC). The numeric value represents
B5
the max number of counts typically accepted on a filter membrane (60 or 80), multiplied by 100
and then divided by the smallest filtration volume analyzed (reported as a ">" value).
B6
Estimated value. Blank contamination evident.
D: Dilution
D
Sample analyzed at dilution
G: Depletion
G7
<1 mg/L DO remained for all dilutions set. The reported value is an estimated > value and is
calculated for the dilution using the least amount of sample
G8
Oxygen Usage <2 mg/L for all dilution sets. The reported value is an estimate < value and is
calculated for dilution using the most amount of sample.
J
The reported value is less than the laboratory Method Reporting Limit (MRL) but greater than or
equal to the Method Detection Limit (MDL)
J1
Surrogate recovery limits have been exceeded
J2
The reported value failed to meet the established QC criteria for either precision or accuracy
J3
The sample matrix interfered with the ability to make any accurate determination.
J: Estimated
J4
The data are questionable because of improper laboratory or field protocols
J5
Temperature limits exceeded (samples frozen or > 4 C) during transport
J6
Laboratory analysis was from an unpreserved or improperly chemically preserved sample.
J7
Blank contamination evident, value may not be accurate
The reported value is determined by a one -point estimation rather than against a regression
J9
equation. The estimated concentration is less than the practical quantitation limit and greater
than the method detection limit.
P: Elevated PQL
P
Elevated PQL due to matrix interference and/or sample dilution
Q
Holding time exceeded
Q1
Holding time exceeded prior to receipt by lab
Q: Holding Time
Q2
Holding time exceeded following receipt by lab
R: Rejected
R
Sample rejected due to blank contamination exceeding value reported
U: Non -detect
U
The contaminant was not detected at a concentration greater than the detection limit
X1
Sample not screened for this constituent
X: Not Analyzed
X2
Sampled, but analysis lost or not performed - field error.
X3
Sampled, but analysis lost or not performed - lab error
Y: Elevated PQL
Y
Elevated PQL due to insufficient sample size
Z: Rejected
Z
Sample results are rejected due to serious deficiencies in the ability to analyze the sample and
meet QC criteria. The presence or absence of the analyte can't be verified.
34

APPENDIX B —Sweeping Route Maps
Route 1000
35

Route 2000
36

Route2001
RauteAlow Route:
aP
pi(Y
LN
--
-"' - Route: 1093 5 Hit O 140 w uT px.
T t .er1s
,oQ KENTIHGTO nR R M 7ER 55� r� 1i7 i 9v
`3s C ap i ¢N a' O�
T,
Route: 1092 4L °¢ " Route: 1007� k y er< NC
l l'YMW 1Rn�[. WR.WO .�R010 RJ�2 3Q 5YfNl !6
Street Sweeper Routes . „a,,,,. , iT
.••. asRpl Roy! ,6 Y�I1NYF]� � 16Th- 1 � lo- M~
public Works ©w Rft .KOA 7P da Mr.1 o jso. emei. xM..
ew? Dam 5ned: 3I112(71] All+M PM M Aalhor PNttni
37

Route 4000
KNIGHT DR
SC,PNFN ;Pl�PCm�7,PC%HAM7 DEERFIELD
BI OCAIN AVE WILSON �OMIRAMONT D
ZRoute:6009
RD ERp�
G DR— Tnn
DIN
N LN�PasF m
RD DY 3n umi a rn 4
^a\i� 5 3' p�Ssv 'ASHY Y Route: 6006 "� o < w ik
( \�' 0 SIBLING DR �4 �� i LEAN C2
V PL p TACY DR 157 O ¢ \ 40
P-AiS� INE DR �.OR <o WILQWO
DR 7 ENE 4� °o DER 0 a �
3 00 m�O
pgBE F D G 'Off'
r Route. 400.3 Inc
WOODSIDE Oy SF� SOP
�Mq�'�J�Aecp `tGt�T PARK LN rSRO� 'F 0' STNCOEK
LINDEN S / CR s` q,� RSTq`J/,A•Ay'_ RD �W�o L ��
NADyLN lry I ? OSF :.Ory� l" 57 �(9 x 2 0J a::.6005� ¢-
RENF EW ?O bG0 y <�1Jy� '!09 o y RtI o
D .,i- M !F m oue: bLI03
' TODD ST SARAT
6 iO kRO yeP��a�C+ 't�O'P �'l1TE OG� z THE FAIRFIELD X MEDICAL pROGq
O`h SN Route' 4009 O w FUryE a KDR
P � DR ��pQ' N DR RET RD O� PAR m
y �
9 O _ P RO o c
pOUGERO EWIR'f_l'
� ° Qy Om OV \< EN WS M m
HOWE ST r2 0 u•4 iON• lYp ROu 5 ROUte:-6002 3T P�JE O c CI VEN
1 C P A ST
OUCHO 1o<RD NE4LILc3. BERINI DR C ,.6q tm7 J R •^ ; KIRK
0.,4SF+FN w�?JO RD m c oute600��0.0
Na1p P�2• u�im ute: 4002 o�`O
I g°
� 751 E�EOR o.�Rolite: 6001 m ��2Jp g HAIR x
• 294P`�HPD O ai �y 2' �` IN INN —DRAKE �F g�. AVE
HU Sp
(Y' 3 0� r- SUNSET VE R SST C�F1, e� �P
Route:40L 10 w Lz
SS
W CNgSFLcR p 4 N D m 0��� c RDutel 4010 = F H gt
C `F 6 y g Q c h R- r.- E R, III���■
P�EKON �NkAL�s z
cT RD LEI Ew 3iLL q. o Route:5
iO Z ST
PINE LASALLESr w a A
p N CONE LN AqS TO WI ST ADR 3
�c,'��o R�1FWR0 / Ro to 01 pD P g Route 50a3 S o E �
BURCHAVE
OR �i -. DAY
•K O TREETO BERGI MAN 00 LILY DR BASSE pR V CEY O m
RDG Rpl-s'O „ ST ff'
„,;,,ti, e7g Dute:5
® 01 Raute �04 Route 07 ROule 10 ROule 30- Shareel 60-Jennings
Street Sweeper Routes 02 Raub OS Roue 09 ROuta�10. Munamma0 U 10 Tnomsa 0- NORM 1 0.8 ]
Public Works Q03 ROub=06 Route 09 Raub ZO SauMedeM 50. erotlls ^r '"as
Date Saved: 3/132013 1:27:09 PM INAuttgf FWard
38

Route 5000
39

Route 6000
40

Route 7000
------ ---- --- -
Route.11
�<6003
m s
Route: 5002t EahN<< GT . Route: 7010'
�wA, V•ME URRmA NU�SUND-•I.ALCI UU
CO '/
PECANPTINIX�1
aNI
-0 N
G Rout0014
S
Rn� tiq�
HILL fk G
Route:500a Route:5009
NP�E E Route.5010F'{
E � rn ;iA/7l�p SUIT' r pIR
M CON I O''Pe • G � y z1 _ RYA •�- - 'a qa A O
ST FP r11/C!I�CC/—� R
�BOWI�FR.D ;Sm
Roule:5003 � o � � Dr Q'
ROUte. 51104 pST 8�0� DANIEL•'RDWA`!`!'IAn\
p LN•I r< o .r� o _v <F4T �< cn TT tR �b OR 11
OU :5 LIBSTTM 007 1"0�2. VERNILL1� nos n (;y i °j -�P.rS Sy.RY
A m 2 s Ht@ zaotu w:p e�
r oute 5 02 o GT�N�2 , 7��`
outer 001 TE Kk^M y VAIR sTou Route: 7609
C12. r N
DUPREE , PRD� �_ !sl 1 O �O I WAKE �O
Route 7004 �C EK�VN 2m� .1 F00�N FORESTWAKE FOREST HWY
v1 5! Route 7,OO6r, _ p
R.ygNr I R
A GIOSON D �uter2010
147 >24� 4_Rollte:7003 P�Q�' iJ. IA ATW . C•- �.
'np yTO OPAST �i�I'JiA S
AO'1l�Q SF,Q�I!� 'MAcy!'RODI:FRRp vE1v2 �P� JG -- R`DUEI.MAvo I ut\\e700
2,UR
L.
V-ANA T OyAT[
ORo\e: o
E�t�q7 'PO y
CYO�FM
KEMPP20009
VE R
M.
PAQDOCK
!'�!pFN �2 KINARD
Ro\te 3010 Ro Q� \J !•rnky <2`��. R_Dop
__ Q9 D
Y rn WYNI NE RrD,_�. _ � �� �0 ?
GHER �CONWAY '� -- y .� - -- SUMMER
OR� DR ,I. �>� WFg1 i1'/ONEPA ROSE LN �4
/r LuoP _WAY Route: 7002 —r T[E p� op BEY:S
PEEDRR^K WAN
Cl� F e. 3088 Rp LANG. -j-
R 1 J ARDEN OYI�'�q/
R0ut%•`3009 %.' (Route:-7001 EN� uQ- ` C°j^ir yls ac g�
9O�yeD N EX .ERy�%N
CT F O'C P�� BLuAY SiW /�O� pl
"= RD)HIEL-DR ::.�. 1 -I OR SS RDpIDO' VL(1fi� �tSF���`RD
01R..to �a/Rriuo o]Roulo®lOftot 00 Sllarealj 60-JenRlnOs q
St�Publlc
outes o2 R d5 Ro1la UU Roula 1 0 0.75 16
s I I10-rnm mmad �4oTnomas P 70.Nedon W\
0] Roule OB Roulo U9 Roule 20 Ig64 O , Soumodendl I60-Rrodle �Mq°4
--- Dale Saved: 3/1420132:31:36 PM N Au11wrFWerd
41