GPO's CGP - Record View
Page 1 of2
Search the CGP / BASIC / ADVANCED / EXPERT / BROWSE / NewTrrtES
CATALOG OF
U.S. GOVERNMENT PUBLICATIONS
Catalogs to Search:
Congressional SerialSet
Congressional Publications
GPO Access Publications
Internet Publications
Periodicals
Serials
My Options:
Bookshelf
Preferences
Hfci.H / ABOUT
Results list
Previous Searches
Catalog of U.S. Government Publications Home Page > National Bibliography of U.S. Government Publications
Add to Bookshelf
Choose Record View: Standard | Short | MARC
Record 1 out of 1
E-mail
Title
Author
Publisher Info.
Internet Access
Sewer sediment and control a management practices reference guide /
Fan. Chi-Yuan.
[Edison, NJ]: United States Environmental Protection Agency, Office of Research
and Development, National Risk Management Research Laboratory, Water Supply
and Water Resources Division, Urban Watershed Management Branch, [2004]
http://gurl.access.gpo.gov/GPO/LPS68041
SuDoc Number
Item Number
Description
General Note
Bibliography
System Details
Subject -LC
Added Entry
US EPA Headquarters Library
U'S- Mail Code 3404T
1200 Pennsylvania Avenue, NW
Washir^ton DC 20460
202-566-0556
EP11.8:600/R-04/059
0431 -Y-01 (online)
1 v (various pagings): digital, PDF file.
Title from title screen (viewed on Mar. 23, 2006).
"January 2004."
"EPA600/R-04/059."
Includes bibliographical references.
Mode of access: Internet from the EPA web site. Address as of 3/23/2006:
http://www.epa.gov/ord/NRMRL/pubs/600r04059/600r04059.htm; current access
available via PURL.
Sewage sludge precipitants -- Government policy - United States - Handbooks.
manuals, etc.
National Risk Management Research Laboratory (U.S.), Urban Watershed
Management Branch.
Holdings All items
Locate in a Library (online) http://cataloa-web2.Qpo.gov/LocateLibraries/locate.)sD?ltemNumber=0431-Y-
01&SYS=000586937
OCLC Number
System Number
(OCoLC)65193039
000586937
http://catalog.gpo.gOv/F/N3KMI353DUHBBT67H5D58QM7SKTCU8SQQYYH39AI2M... 7/18/2006
-------
GPO's CGP - Record View
Page 2 of2
End Session - Preferences - Feedback - Help - Browse - Search - Results List - Previous Searches - Catalogs to Search - Bookshelf
A service of the U.S. Government Printing Office. 2005-2006.
http://catalog.gpo.gOV/F/N3KMI353DUHBBT67H5D580M7SKTCU8SQQYYH39AI2M... 7/18/2006
-------
Chapter 1
Introduction
Background
This report provides an integrated approach to manage the solids generated in urban wet weather flows (WWFs),
both on the land surface and in drainage sewer systems. Threats to the quality of receiving waters by discharges
from urban-storm-generated WWF, including combined sewer overflow (CSO) and polluted runoff from urban
streets, are well known. During many storm events, large volumes of stormwater are drained via street inlets into
the urban sewer system. The storm runoff washes off street dust-and-dirt and pollutants from catchment surfaces
into the sewer system. Furthermore, the unsteady-state storm inflow resuspends sewer sediment that has settled in
the sewer bottom, causing it to be transported downstream. Recently, researchers reported that sewer sediment
deposited from prior storms contributed a significant amount of pollutants into receiving waters. This often creates
a highly concentrated pollutant load. In most cases, CSO carries resuspended sewer sediment and generates a highly
concentrated pollutant load sometimes associated with the "first-flush" phenomenon (Saget et al., 1996; Arthur and
Ashley, 1998; Krebs et al., 1999). In most cases, CSO carries this resuspended sewer sediment into local
waterways. Sewer sediment deposited during dry weather flow (DWF) contributes between 30% to 80% of
pollutants into receiving waters (Ahyerre et. al., 2001).
The Problem
One of the underlying reasons so much sediment is deposited in combined sewers is hydraulic design. Combined
sewers are sized to convey many times the anticipated peak DWF; they can carry up to 1,000 times the expected
background DWF. Ratios of peak to average DWF usually range from 2 to 10 for interceptor sewers. The oversized
and mildly sloping combined sewer segments possess a substantial potential for sedimentation during dry weather
periods. DWF velocities are typically inadequate to maintain settleable solids in suspension and a substantial
amount of sewer solids accumulated in the pipes. During rain storms, the accumulated solids may resuspend,
because of the limited hydraulic capacity of the interceptor sewers, overflow to receiving waters. Suspended solids
(SS) concentrations of several thousand parts per million are common for CSOs. This can produce shock loadings
detrimental to receiving waters. Accumulation of sewer solids in sewers also result in a loss of flow-carrying
capacity that may restrict/block flow and cause an upstream surcharge, local flooding, and enhanced solids
deposition. Sewer-solid accumulation in urban drainage systems also creates septic conditions that pose odor, health
hazards, and corrosion problems.
During low-flow periods, sanitary wastewater solids deposit in combined sewers because the flow velocity is usually
less than the particle-settling velocity. Estimates of solids deposition range from 5% to 30% of the daily SS
pollution loading (Pisano et al., 1998). The average dry period between storm events is about four days for many
areas of the United States, especially along the eastern seaboard. If 25% of the daily pollution loading accumulates
in the collection system, an intense rainstorm causing a two hour CSO, after four days of antecedent dry weather,
will wash the equivalent of one-day's flow of raw-sanitary wastewater to the receiving waters. In Europe, average
1
-------
deposition rates have been measured to range from 30 to 500 g/m/d (Ashley and Crabtree, 1992). Even sewers
supposedly designed to be 'self-cleansing' will have transient sediment deposits, and part of the load in transport
will move near the bed (May et al., 1996). Furthermore, a one-day equivalent of raw-sanitary wastewater,
discharged within a two-hour period, is twelve times the rate at which raw-sanitary wastewater enters the collection
system.
Sewer sediments contain high concentrations of pollutants (80,000 mg/L of BOD5; 200,000 mg/L of COD; and 200
mg/L of NH3-N) (Arthur et al., 1996). As storm-flow intensities increase, lesuspension of sewer sediment will
occur. In combined sewers this occurs when they hydraulically overload and discharge as CSOs. The sewer-
sediment layer contains organic materials and sulfides that can generate toxic, corrosive, and hazardous gases (e.g.,
hydrogen sulfide (H2S) and methane, under anoxic conditions). Sulfates are reduced to H2S and then oxidized to
sulfuric acid by biochemical transformation; the acid attacks the sewer, thereby weakening its structural integrity.
Extensive corrosion of concrete and their reinforcing bars results in cracks and infiltration and exfiltration of raw
wastewater, causing overflow and WWTP overloading and groundwater contamination, respectively. Thus, control
of sewer sediment not only protects urban receiving-water quality it also prevents hazardous conditions in sewerage
systems and maintains the structural integrity of the sewer.
One of the challenges in protecting urban waterways lies in effectively managing contaminated sediments in both
the sewer system and the receiving water. To enable urban communities to develop better plans to reduce the risks
associated with WWF, research is needed to develop tools for a better understanding and assessment of the fate and
transport of sediment solids and associated pollutants.
This report serves as a reference for the user community faced with the challenges to combat urban wet-weather-
induced point and diffused sources of water pollution. It covers the gamut of engineering requirements, from
pollution problem assessment tools for both desktop analysis and field investigation to determine extended problems
of sewer sediment. It includes the following six chapters:
Sources of sewer sediment and impacts (Chapter 2)
Estimation of urban watershed solids loading (Chapter 3)
Methodology for quantifying sediment-solids in sewer system (Chapter 4)
Methods for field sampling and monitoring of hydrogen sulfide in sewer (Chapter 5)
Sewer sediment control: sewer flushing (Chapter 6)
A case study of sewer flushing system design and operation (Chapter 7)
The sewer sediment solids and associated pollutants found in combined sewers are mainly resulted from sanitary
wastewater solids deposition during dry weather. These solids account for the majority of WWF pollution. All
sources of WWF pollutants and their impacts are explained in Chapter 2. Methods for estimating solids mat are
washed-off from land surface during a storm event are presented in Chapter 3. A set of generalized procedures for
estimating pollutant loadings associated with dry weather wastewater solids deposition in combined sewer systems
is described in Chapter 4. Once the sewer segments most prone to sewer sedimentation are identified, sediment
sampling is needed to determine actual sewer sediment and sedimentation characteristics. Chapter 5 describes
sampling and monitoring procedures for measuring both the gas phase and dissolve phase of H2S concentrations.
In order to reduce solids and associated pollutants entering in sewer systems, one must start with pollution
prevention and source control by the best management practices (BMPs). Information on the urban stormwater
BMPs implementation and evaluation are available in published literatures (ASCE, 2001; USEPA, 2002; Streckeret
al., 2001; Strecker, 2002). Thus, Chapter 6 addresses only the management practices for in-sewer sediment solids
control. The last chapter, Chapter 7 - Sewer Sediment Rushing, brings together information on the most recent case
in planning and implementing in-sewer sediment control technologies in a large urban sewer-catchment. In this
chapter, also includes estimated operation and maintenance costs as well as capital costs based on the Engineering
News Record (ENR) Construction Cost Index of 6389 as of August 2001.
-------
Chapter 2
Sewer Sediment
Sources of Sewer Sediment
The sediment-solids and associated pollutants found in combined sewer overflow result from resuspension of
deposited DWF sanitary wastewater solids and wash-off from land surfaces during storm events. A review of the
sources (Heaney et al., 1999) shows that directly-connected impervious areas contribute a high pollutant loading
in separate storm sewers. For combined sewers, the largest solids and pollutant loads are likely to originate from
sanitary wastewater input during dry weather. Ashley and Hvitved-Jacobsen (2002) categorized sources of
sediment solids as shown in Table 1.
Table 1. Sources of sewer sediment
Source
Winter gritting/salting
Road surfacing
Flow from ground
Industrial wastewater
Construction sites
Flooding
Runoff from impervious
areas
Sanitary wastewater
Soil erosion
Wind-blown from
sand/soil/litter
Particle Characteristics
Salt 1000 mm possible
> 1000 mm possible
Typically solids < 250 \un enter
sewer.
Up to 100 mm
Typically <1 mm
Large organics possible.
Inorganics <5 mm
Description
Sand used up to 30% of total mass annually.
Primarily inorganic.
Depending on sewer condition.
Pretreatment removal of toxic solids.
All sizes organic, inorganic possible.
All sizes organic, inorganic possible.
These solids may be up to 40% by mass of total.
Roof surface up to 30% of total.
Largest organic solids source - typically 97% of
these solids. All enter sewer.
Organic and inorganic.
Entry via catchbasins/inlet. Size reduced when
discharged into sewer.
-------
Overland Surface Runoff Solids
The paniculate and associated pollutants in urban stormwater come from atmospheric deposition, roof tops,
parking lots, and streets/highways. Other sources include construction sites, commercial and industrial parking
lots, automobile maintenance operations, leaking sewer infrastructure, accidental spills, and runoff from lawn
irrigation.
Atmospheric Deposition
In the United States, each year millions of tons of pollutants are emitted into the; troposphere zone of the
atmosphere; this has the potential to redeposit in the urban and terrestrial watershed and be subsequently
transported downstream to receiving waters. The factors affecting atmospheric deposition include wind speed and
direction, dry dust fall, site temperature and precipitation (snow and rainfall), elevation and slope of the land, land
use, and sources of air pollution (automobile, industrial, and residential emission). Pollutants in the atmosphere
contribute significantly in urban WWF contamination through dustfall and by wash out. As reported by Cotham
and Bidleman (1995) and Hilts (1996), enormous amounts of certain toxic pollutants contained in urban storm
runoff are associated with atmospheric deposition.
Rooftops, Roadways, and Parking Lots
One of the major sources of pollutants in urban drainage catchments are runoff from: urban streets (Sansalon,
1996; Sansalone and Buchberger, 1996), highways (Shaheen, 1975; Montrejaud-Vignoles et al., 1996), building
rooftops (Sakakibara, 1996; Forster, 1996; and Wada et al., 1996), and parking areas (Pitt et al., 1995;
Nowakowska-Blaszezyk and Zakrzewski, 1996). In some cases, treated wood has been identified as a potential
source of arsenic (As), cadmium (Cd), copper (Cu), chromium (Cr), and zinc (Zn) in stormwater (Weis and Weis,
1996). Table 2 depicts relationships between toxic pollutants in solids and urb:m land use.
Table 2. Toxic-pollutant concentrations from land use In France (Bertrand- Krajewekl, 1993)
Land Use
Residential area
Commercial area
City downtown
Industrial area
Parking lot
Street
Highway
Toxic Pollutant Concentration
Cd
(•S/g)
0.04-10.7
0.02-1.06
2.6-7.0
0.7-3.4
1.0-14.6
0.22-3.90
0.6-4.3
Cu
(•g/g)
14-221
10.4
143 - 390
228
206
22-200
90-281
Pb*
(*g/g)
120-1,000
160-220
1,880-2,550
488-1,410
2,000-15,000
--
1304,800
Zn
Cg/g)
47-1,170
53-1,065
470-534
655-1,445
1,600
44^80
250-336
TPHs
(mg/g)
15.7-59.8
16.4-34.0
8.8-51.8
61.9-507.0
--
-
--
PAHs
(•E/g)
-
~
-
-
--
0.2-20
--
* Pb relates to the use of leaded gasoline.
Legend: Cd-cadmium, Cu-copper, Pb-lead, and Zn-zinc.
TPHs-total petroleum hydrocarbons, and PAHs-polycyclic aromatic hydrocarbons.
Distributions of heavy metals and hydrocarbons in urban stormwater are associated with their particulate fractions
and the relative size of SS. Particles finer than 250 • tn contain higher concentration of heavy metals and total
petroleum hydrocarbons (TPHs) than particles larger than 250 • m and about 70% of the heavy metals are attached
to particles finer than 100 • m (Ellis and Revitt, 1982). Vignoles and Herremans (1995) examined the heavy metal
associations with different particles sizes in stormwater samples from Toulouse, France and discovered that the
vast majority of the heavy meta! loadings in stormwater were associated with particles less than 10 • m in size.
These results are shown in Table 3.
-------
Table 3. Concentration of metals by size fraction
Panicle
Size
Range
(Mm)
>100
50-100
40-50
32-40
20-32
10-20
<10
Metal Concentration
Cd
Hg/g
13
11
11
6
5
6
14
%*
18
11
6
5
5
9
46
Co
H8/g
18
16
25
20
18
22
53
%*
9
5
4
6
6
10
60
Cr
Hg/g
21
25
26
50
23
39
134
%*
5
4
2
6
3
9
71
Cu
Mg/g
42
62
57
46
42
81
171
%*
7
8
3
4
4
11
63
Mn
Mg/g
86
59
70
S3
54
85
320
%*
8
4
3
3
4
7
71
Ni
Mg/g
31
27
31
31
27
39
99
%*
7
5
7
7
5
10
59
Pb
Mg/g
104
129
181
163
158
247
822
%*
5
7
10
9
8
14
46
Zn
Mg/g
272
419
469
398
331
801
1232
%«
7
11
12
10
9
20
31
Legend: Cd-cadmium, Co-cobalt, Cr-chromium, Cu-copper, Mn-manganese, Ni-nickel, Pb-lead, and Zn-zinc.
* Distribution of metal pollutant weight among the different particle size range.
Snowmelt runoff is much greater in volume than typically considered in drainage designs, resulting in greater
winter flooding than during the summer; however, there is still a notable lack of experience about urban runoff
during the winter season (Thorolfsson and Brandt, 1996). Saxton et al. (1996) conducted a study to characterize
the pollution of snow versus snowmelt runoff at Eielson Air Force Base, Alaska and reported that snow was more
contaminated than snowmelt runoff and that snowmelt runoff appeared to be representative of what reached
surface water. Sansalone (1996) investigated the forms of stormwater and snowmelt heavy metals and reported
that Zn, Cd, and Cu were mainly dissolved in stormwater, while only Cd was mainly dissolved in snowmelt.
Sanitary Wastewater Solids
According to Ashley and Hvitved-Jacobsen (2002), solids originating from sanitary wastewater sources can be
categorized into the following types:
1. Fine fecal and other organic particles.
2. Large fecal and other organic matter.
3. Paper, rags, and miscellaneous sewage litter.
These categories also apply to commercial and other workplaces, where other substances may be added, subject to
effluent controls. Industrial sources are also important, but due to the diversity of the inputs from industrial
sources, they will not be considered further here. Garbage grinders, that are installed in many residential areas for
disposal of kitchen wastes generate higher organic solids loading. Pollutant loads and concentrations from
residential sources discharging to sewers are shown in Table 4.
Table 4. Pollutant loads and concentrations from residential sources (EPA 1992)
Parameter
BODj
SS
Nitrogen
Phosphorus
Garbage Grinders
gpcd
11-31
16-44
0.2 - 0.9
0.1
mg/L
2380
3500
79
13
Toilets
gpcd
7-24
13-37
4.1 - 16.8
0.6-1.6
mg/L
260
450
140
20
Basins, Sinks, Appliances
gpcd
25-39
11-23
i.1-2.0
2.2 -3 4.
mg/L
260
160
17
26
Results from the Jefferies and Ashley (1994) study of gross solids discharge in combined sewers can be
interpreted to give a rate of 0.05 visible items /capita/day. The average disposal rate reported by Friedler et al.
(1996) was 0.15 refuse items/capita/day, 72% of which was due to female toilet usage. The most common item of
-------
refuse (23% of those reported) was the tampon. According to Ashley et al. (2000) some 2.5 million tampons, 1.4
million sanitary towels, and 700,000 panty liners were found to be flushed into sewers in the United Kingdom
every day. These items become floatable solids in CSO. Thus, the accumulation of trash on beaches and along
shorelines of impacted waterways is the most obvious impact of floatable pollution. It is not only in the United
Kingdom that the toilet is being used as a rubbish bin. A limited questionnaire survey was undertaken of the
items disposed in 72 countries. Results indicated that some 33% of respondents claimed that sanitary items, other
than feces and toilet paper, were regularly flushed, and in some countries 'disposable' napkins were also put into
the toilet (Ashley et al, 1999). There will not likely be any significant reduction in these items found in sewers in
the near future necessitating expensive screens and transport systems for their control and disposal (Ashley et al,
2000).
Impacts
In general, sewers will not maintain self-cleansing velocities at all times. The diurnal pattern of DWF and the
temporal distribution and nature of sediments found in sewer flows may result in the deposition of some
"juvenile" sediments at times of low flow. The subsequent erosion and transport of these sediments at times of
higher flow during a storm-flow event, either as suspended load or bedload, contribute to the "first-flush"
phenomena or polluted segment in CSO (Saget et al., 1996; Arthur et al., 1996; Arthur and Ashley, 1998; Krebs et
al., 1999). During low flow dry weather periods, sanitary wastewater solids deposited in combined sewer systems
can generate H2S and methane gases due to anaerobic conditions. Sulfates are reduced to HjS gas that can then be
oxidized to sulfuric acid on pipes and structure walls by further biochemical transformation. Furthermore, these
sediments are discharged to urban streams during storm-flow events and can cause degradation of receiving water
quality. Thus, dry weather sewer sedimentation not only creates hazardous conditions and sewer degradation but
also contributes significant pollutant loads to the urban receiving waters during wet-weather high-flow periods.
Furthermore, broken sewer lines cause direct exfiltration of raw sanitary wastewater and sewer sediment leachate
into subsurface groundwaters.
Structural Deterioration of Sewerage System
The primary cause of odor and corrosion in collection systems is the sulfide ion (S=), which is produced from
sulfate (SO4=) by bacteria residing in a slime layer on the submerged portion of sewer pipes and structures. Once
S= is released from the wastewater as H2S gas, odor and corrosion problems begin. Bacteria utilize H2S gas to
produce sulfuric acid (H2SO*) (Boon, 1995; Boon and Lister, 1975; Thistlethwayte 1972). For sanitary
wastewater the main source of S= is SO4Z. Sulfide generation is a bacterially mediated process occurring in the
submerged portion of combined and sanitary sewers and force mains. Fresh sanitary wastewater entering a
collection system is usually free of S". However, a dissolved form of S= soon appears as a result of low dissolved
oxygen content, high-strength wastewater, low flow velocity, long detention time in the collection system,
elevated wastewater temperature, and extensive pumping (EPA 1985).
The effect of HaSO* on concrete surfaces in the sewer environment can be devastating. Sections of collection
interceptors and entire pump stations have been known to collapse due to loss of structural stability from
corrosion. In severe instances, pipe failure, disruption of service, street surface cave-ins, and uncontrolled
releases of wastewater to surface streams can occur.
Receiving Water
From 40% to 80% of the total annual organic loading entering receiving waters from a city is caused by WWF.
During a single storm event, WWF accounts for about 95% of the organic load as well as high loads of heavy
metals and petroleum hydrocarbons (Field and Turkeltaub, 1981). CSO can have damaging impacts on receiving
waters. The EPA evaluated the distribution and biological impacts of discharged particulates for selected CSO
and storm drain points in the Seattle, Washington region (Tomlinson et al., 1980). The concentrations of SS,
heavy metals, and chlorinated hydrocarbons were greater for the CSO than for the storm drains. Particulate
distributions were influenced by various dispersion processes, including water density layering, near-bottom
offshore streaming and advection along the shoreline Human enteric viruses were also detected in the CSO, but
6
-------
were not found in storm drainage or in any near outfall sediments. However, impacts of discharges on the
freshwater benthos raised concern relative to the feeding success of sport fish due to polluted sediments.
Saul et al. (1999) investigated the production of undesirable solids in CSO as it related to social, economic, and
ethnic factors. The goals of the research were first to determine the differences in sewer solids characteristics
that were ultimately discharged to the receiving water and then to use the solids' characteristics to predict the
efficiency of CSO treatment devices, especially CSO storage basins. St. Michelbach and Brombach (1999)
showed that the nutrient content, especially of dissolved phosphorus, from CSO and existing WWTPs was
endangering the health of Lake Constance. They proposed a simple methodology to estimate the nutrient loads
from CSO to the lake the results of which can be used to determine the cost effectiveness of CSO improvement
versus WWTP improvement.
Sanudo-Wilhelmy and Gill (1999) compared current pollutant concentrations in the Hudson River Estuary, New
York with concentrations measured in the 1970's. The concentrations of Cu, Cd, Ni, and Zn have declined, while
concentrations of dissolved nutrients (namely PO4) have remained relatively constant during the same period.
This suggests that WWTP improvements in the New York/New Jersey Metropolitan area have not been as
effective at reducing nutrient levels within the estuary as heavy metals. Rather than inputs from point sources, the
release of Pb and Hg from watershed soils, and Ni and Cu from estuarine sediments, may represent the primary
contemporary sources of these metals to the estuary. Mason et al. (1999) showed that the Chesapeake Bay was
an efficient trap for Hg. However, in the estuary, methylation of the mercury occurred, the Bay became a source
of methylmercury, and on a watershed scale, only about 5% of the total atmospheric deposition of mercury was
exported to the ocean.
Venkatesan et al. (1999) investigated the potential for using sediment cores to determine the history of chlorinated
pesticide and PCB application in a watershed. They found that the sediment cores accurately reflected the length
of use of these chemicals in the watershed, and that the surface sediment layer, after mixing and resuspension was
accounted for, reflected the reduction in use that had occurred during the last few years. The long-term impacts of
WWF- toxic pollutants to stream habitat are depended on bio-availability and accumulation of the substances by
aquatic life. Herrmann et al. (1999) found that the concentration of ammonia plus urea in CSO was found to be a
significant measure of the likelihood of a fish kill after an overflow event, more relevant than the concentration of
ammonia alone.
Groundwater
In 1999, the EPA conducted a nationwide study to quantify leakage of sanitary and industrial wastewater sewer
systems based on groundwater table elevations. The study indicated low levels of wastewater exfiltration (less
than groundwater infiltration) in much of the midwestern and eastern parts of United States due to relatively high
groundwater tables. However, problems of exfiltration in the western United States seem more widespread
because of lower groundwater table (EPA, 2000). Thus, contamination of soils and groundwater in the vicinity of
a leaking sewer does not appear to occur under conditions favorable to the infiltration of groundwater into sanitary
sewers. Exfiltration events are likely to be more severe than infiltration events at locations where groundwater
fluctuates. Possible groundwater contamination, resulting from sewers that have collapsed or catastrophically
failed and from sewers which are believed to suffer from long-term deterioration, has been noted in groundwater
contamination studies (EPA, 1989).
In those areas having shallow depth of wells and high permeability of soil, any surface contamination could
easily migrate to the groundwater. Thus, a significant amount of groundwater contamination is as attributable to
surface runoff as leaky sewer exfiltration. Squillace, et al. (1996) and Zogorski, et al. (1996) investigated urban
stormwater as a source of groundwater MTBE contamination. Mull (1996) stated that traffic areas are the third
most important source of groundwater contamination in Germany (after abandoned industrial sites and leaky
sewers). The most important contaminants are chlorinated hydrocarbons, sulfate, organic compounds, and
nitrates. Heavy metals are generally not an important groundwater contaminant because of their affinity for soils.
-------
Trauth and Xanthopoulus (1996) examined the long-term trends in groundwatei quality in Karlsruhe, Germany.
Results indicated that the urban land use could cause a long-term adverse influence on the groundwater quality.
The concentration of many pollutants have increased by about 30% to 40% over 20 yrs. In Dortmund, Germany,
an infiltration trench for stormwater disposal caused Zn problems that were associated with the low pH value
(about 4) in the infiltration water (Hiitter and Remmler,1996).
-------
Chapter 3
Storm Runoff Solids Loading
Introduction
This chapter presents methodologies for using relatively simple equations to estimate pollutant loading from
urban drainage areas. The loading equations in this chapter account for floatable solids or street Utter, highway
sand due to winter deicing, street dust and dirt, and eroded soil-sediment from open space. The calculated results
only indicate a magnitude of solids loading. For designing a stormwater pollution control system, field sampling
and monitoring data are required to verify the estimated results.
Litter/FIoatables
A large amount of the litter that enters urban drainage systems reaches receiving waters. Urban litter consists
mainly of manufactured materials, e.g., such as plastic and paper wrapping, shopping bags, cigarettes and their
packets, and items used in public parks, gardens, and fast food outlets. The total amount of material discharged
can vary significantly depending on the degree to which the watershed is littered. Five studies (conducted in the
NY/NJ Metropolitan Area; Jamaica Bay, NYC; Fresh Creek, NYC; Hartford, CT; and Newark, NJ) looked at the
total amount of solid material discharged from CSOs on a dry weight basis and reported between 0.02 and 1.7
Ib/acre/in. of rainfall (Newman and Leo, 2000).
A study in the City of Auckland, New Zealand (Cornelius et al, 1994) indicated that the annual litter loading rates
from commercial, industrial, and residential areas are 1.35, 0.88, and 0.53 kg/ha-yr, dry weight bases (or 0.014,
0.009, and 0.006 m3/ha-yr), respectively. The litter's bulk densities vary with land use (96.4 kg/m3 for
commercial, 97.8 kg/m3 for industrial, and 88.3 kg/m3 for residential areas). Although the commercial and
industrial areas produced higher annual loadings per unit area, the residential areas contribute more total litter than
all other areas combined because residential development occupied the largest area. Armitage and Rooseboom
(2000) developed an empirical equation to determine annual volume of litter for South Africa:
T= Łf*(V, + B,)Ai [[[ 0)
Where: T = total litter load in the waterways (m3/yr)
fsci - street cleaning factor for each land use (varies from 1 for regular street
cleaning to about 6 for no street cleaning services)
V) = vegetation load for each land use (varies from 0.0 m3/ha-yr for poorly vegetated areas to
about 0.5 m3/ha-yr for densely vegetated areas)
B, = basic litter load for each land use (1 .2 m3/ha-yr for commercial; 0.8 m3/ha- yr for
-------
For each storm (>1 mm of rainfall) the litter volume can be estimated by usinf; the following equation (Armitage
and Rooseboom, 2000):
S'lT/Zh (2)
Where: S = storm load in the waterways (mVstorm)
fs = storm factor (varies from 1.0 for storms occurring less that a week after a previous >lmm
storm; to about 1.5 for a storm occurring after a dry period of about three weeks; to
about 4.0 for a storm occurring after a dry period
of more that about three months)
T = total litter load in the waterways (nvVyr)
Zfa = the sum of all the storm factors for all of the storms in the year (since this information is
generally not available, a suggested alternative is to count the average number of
significant storms in a year and multiply by 1.1)
Roadway Sanding for Snow/Ice Events
Roadway sand application is a common practice during the winter snow season for increasing track friction
between highway surfaces and automobile wheels. Sanding is important to public safety because it provides safe
conditions during treacherous winter weather. However, after snow-melt, sand becomes part of highway nonpoint
source pollutants. Guo (1999) developed a method to determine the sand recovery during winter highway
sanding. The sand recovery rate is defined by the ratio of the annual sand amount collected by the highway
drainage system to the annual sand amount applied to the highway. The method is being adopted for estimating
the amount of sand that escapes into the environment.
During winter, the total amount sand application can be estimated as:
s = w,B,L.
•(3)
Where: Ws = sand amount in kg or Ib
w, = annual unit sand amount in kg/m2 or Ib/fr2
B, = width of traffic lanes in m or ft
L = distance of highway between two adjacent culverts in m or ft
A typical highway drainage hydraulic routing system for the snow removal process includes: (1) piling snow on
bom sides along the highway shoulders for snow storage; (2) roadway drainage gutter; (3) highway runoff
collection system for releasing runoff that contains various types of pollutant with different concentrations to
receiving streams. During a snow plowing operation, sand is applied only to traffic lanes. Snow mixed with sand
are removed from the traffic lanes to a storage area which is located along the highway shoulders for compacting
and piling. The captured snow volume can be estimated as
,.(4)
Where: V,; = captured snow volume in m3 or ft3
Hm = maximum height of snow pile in m or ft
B,, = width of storage area in m or ft
L - distance of highway between two adjacent culverts in m or ft
Snow removed from the highway is placed in the storage area along highway with a maximum height of 7.5 ft.
The compacted snow volume between two adjacent culverts can be estimated as
nmP,BL.
10
.(5)
-------
Where: Vs = compacted snow volume in m3 or ft3
P, = equivalent water depth to annual fresh snowfall depth
n = snow compact ratio, defined as 1 ft fresh snowfall equivalent to n ft compacted
snow
m = snow-to-water depth ratio, defined as m ft fresh snowfall to produce 1 ft of water
B = total width of the paved highway area including traffic lanes, shoulder areas, and
snow storage areas in m or ft
L = distance of highway between two adjacent culverts in m or ft
The snow volume capture rate (r) from the highway/paved surface by the storage area is defined by Guo (1999) as
r= VC/V, [[[ (6)
Since snow and sand will be well mixed during the plowing process, the amount of sand captured during this
process and stored in the snow storage area is
Wt*rW, [[[ (7)
Where:
Wr = sand amount in weight captured by the snow storage area
Ws = sand amount in weight applied
r = snow capture rate by storage area, which is the ratio of captured snow volume to
the compacted snow volume in the storage area.
After snow melt, the recovery amount of the sand remaining in the storage area that needs to be recovered by
street sweeping equipment is estimated as follows:
Wa = Rm(Wf-Wb) [[[ (8)
Where:
Wm = sand amount in weight collected by machine
Wb = sand amount in weight transport by runoff
Rm = efficiency of sand collection by machine, such as 0.80 to 0.90, depending on field
operations
The sand amount transported (Wk) through the highway drainage ditch can be estimated by the event mean
concentration method (Urbonas et al., 1996; Mosier, 1996):
(9)
Where: ft - specific weight of sand
Ł„ = empirical value of event mean concentration
V0 = total annual runoff volume
The sand recovery (W,) between two adjacent culverts is:
W,= Wm + eWb [[[ (10)
In which: e = 1 for sand collection with a sand basin at the end of drainage system, or e = 0 for direct release
through a culvert to the receiving stream. Therefore, one may estimate that the annual sand emitted to the
-------
Street Dust and Dirt Accumulation
Sartor and Boyd (1972) reported that the build-up of dust and dirt between street cleanings was non-linear and of
an inverse exponential form over a period of up to 10 days. Huber and Dickinson (1988) used three types of
equations in the U.S. EPA's Stormwater Management Model (SWMM) for estimating the loading of dust and dirt
accumulation:
Power-Linear Equation: DD = DDFACT(T°DPOW) (11)
DD 270
85-910
>750
370 -(> 1400)
Days To
Observed
Maximum
Loading
>25
5-70
>30
10-(>50)
Street Dust and Dirt Washoff
Based on field study by Sartor and Boyd (1972), the washoff can be expressed by the following first-order decay
equation:
•(14)
Where:
N = amount of street dust and dirt washoff, g/curb-m
NO = amount of initial street dust and dirt, g/curb-m
K = washoff coefficient (ranged 0.167 - 1.007 depending on rain
intensity, street dirt loading category, and street texture category)
R = total rain depth, mm
12
-------
Washoff is more efficient for the higher rain energy and smoother pavement (Delleur, 2001).
Soil Erosion
Soil erosion from an open land is considered by many to be a problem to receiving-water quality. The amount of
soil loss can be computed by the Revised Universal Soil Loss Equation (RUSLE). The RUSLE computes sheet
and rill erosion from rainfall and the associated runoff for a landscape profile. The equation is written as (Renard
et al., 1996):
A = R\Kx.LS\C\P (15)
Where: A = annual soil loss from sheet and rill erosion, tons/acre
R = rainfall and runoff factor; ranged 80-94
K = soil credibility factor; depended on soil type and organic matter,
for 2-4% of organic matter, ranged 0.4-0.25
LS = slope length and steepness or slope length-gradient factor
C = cover and management factor; legume, C = 0.005; ryegrass, C = 0.1
P = support practice factor; 0.3-1.0
The slope length-gradient factor (LS) can be determined by Equation 16 below:
LS= [0.065 + 0.0456 (slope) + 0.006541 (slope)2] [(slope length)n2.5]NN .
.(16)
Where: Slope = slope steepness, %
Slope length = length of slope, ft
NN = slope steepness factor, ranged 0.2-0.5
Individual factor values can be entered directly into the formula or calculated from information provided by the
user. The equations given are empirical and can be used for planning purposes. Actual measurement of
pollutants is always the best way to understand and predict pollutant loads specific to any watershed, but it is
often expensive and time consuming. These equations may be used to estimate the total maximum daily loads for
watershed management plans, but for final design, field-monitoring data should be obtained.
Hypothetical-Case Example
A hypothetical urban watershed is presented to illustrate the application of pollutant loading estimation methods
as described in this chapter. The total drainage area in this example is approximately 1,200 ha which consists of a
mixture of land uses. The areal distribution of each land-use category is shown in Table 7.
Table 7. Land use areal distributions for hypothetical-case example
Land Use
Low density residential areas
High density residential areas
School
Commercial areas
Light industrial areas
Parks
Streets, total length = 6 km
Minor arteries, total length = 2 km
Major arteries, total length = 1 km
Total
Area
(ha)
300
100
20
200
100
280
120
50
30
1,200
13
-------
Area Characteristics
Land use parcel characteristics are addressed in terms of land area with roadway right-of-way (RW)
characteristics in terms of width and length. The RWs are measured as assigned widths based upon the following
criteria. Streets within a development have an average RW of 20 m, a minor aitery has a 25m RW, and a major
artery a 30m RW. The profiles for each RW in this case study are shown in Table 8.
Table 8. Roadway right-of-way characteristics for hypothetical-case example
Length
(km)
6
2
1
RW
(m)
20
25
30
Curb*
(m)
2
2
2
Parking*
(m)
4
4
4
Landscaped Strip*
(m)
3
3
6
Sidewalk*
(m)
3
3
3
Traffic Lanes
(m)
8
13
15
* Parameters are summed from both sides of the street.
An aggregated analysis was used for the low density (single family houses) and high density residential areas,
commercial, school, and light industrial land uses because they exhibited multi-parcel characteristics, such as
parking. The lot and aggregated characteristics for residential parcels, commeicial, schools, and light industries
are presented in Tables 9 and 10, respectively.
Table 9. Lot characteristics for residential, commercial, schools, and industries in hypothetical-case example
Land use
Single family houses
Apartment buildings
Commercial buildings
Schools*
Light industries
No. of
parcels
1,200
50
20
2
5
Each Parcel
Area (m2)
2,500
20,000
100,000
100,000
200,000
Roof Area
(m2)
500
6,000
45,000
17,000
100,000
Driveway/
Parking (m2)
300
9,000
35,000
23,000
80,000
Landscaped
Area (m2)
1,700
5,000
20,000
60,000
20,000
Areas include athletic fields
Table 10. Aggregate characteristics for each land use for hypothetical-case examf le
Land Use
Low density residential areas
High density residential areas
School
Commercial areas
Light industrial areas
Parks
Streets, total length = 6 km
Minor arteries, total length = 2 km
Major arteries, total length = 1 km
Total
Total Area
(ha)
300.0
100.0
20.0
200.0
100.0
280.0
120.0
50.0
30.0
1,200.0
Roof
Area
(ha)
24.0
30.0
3.4
90.0
50.0
2.6
0
0
0
200.0
Parking/
Roadway
(ha)
36.0
45.0
4.6
70.0
40.0
27.4
84.0
38.0
21.0
366.0
Landscaped
Area
(ha)
240.0
25.0
12.0
40.0
10.0
250.0
36.0
12.0
9.0
634.0
Utter/Floatable Solids
The empirical equation (Eq. 1) developed by Armitage and Rooseboom (2000) was used to determine annual
litter volume (7) in nrVyr. The estimated litter/floatable solids volume and loading are summarized in Tables 11
and 12, respectively.
14
-------
Table 11. Calculations of litter/floatable solids volume for hypothetical-case example
Land Use
Low density residential
High density residential
School
Commercial areas
Light industrial areas
Parks
Total
A,
(ha)
300.0
100.0
20.0
200.0
100.0
280.0
B,
(m3/ha-yr)
0.01
0.02
0.02
1.20
0.80
0.50
V,
(nrVha-yr)
0.02
0.02
0.03
0.03
0.03
0.01
/*
1
1
1
1
1
1
Litter Volume
Eq. (1)
(mVyr)
9
4
1
246
83
143
486
Table 12. Summary of litter/floatable solids loading for hypothetical-case example
Land Use
Low density residential
High density residential
School
Commercial areas
Light industrial areas
Parks
Total
Litter Volume
GnVyr)
9
4
1
246
83
143
486
Bulk Density
(kg/m3)
88.3
88.3
88.3
96.4
97.8
88.3
Litter Loading
(kg/yr)
795
353
88
23,714
8,117
12,627
45,694
The estimated total annual litter and floatable solids loading is about 45,700 kg.
Road Sand
Sand loading estimates, due to winter sand application, were calculated using the method developed by Guo
(1999), and the results are summarized in Table 13.
Table 13. Amount of sand discharged to receiving water for hypothetical-case example
Type of
roadway
Street
Minor
artery
Major
artery
Total
Length
(km)
6
2
1
Lane
Width
(m)
8
13
15
Road
Surface
Area
(m2)
48,000
26,000
15,000
Sand(l)
Application
Rate
(kg/mVyr)
2
5
10
Sand
Applied
Eq. (3)
(kg/yr)
96,000
130,000
150,000
Sandw
Recovered
Eq. (8)
(kg/yr)
67,200
91,000
105,000
Sand (3)
Transported
Eq. (9)
(kg/yr)
28,800
39,000
45,000
112,800
Notes: (1) Local Dept. of Public Works road services inventory records, assumed average values.
(2) Amount of sand recovered by street cleaning operation.
(3) Amount of sand removed by storm runoff and discharged to receiving water.
The amount of road sand discharged into receiving water is estimated to be 112,800 kg/yr.
15
-------
Street Dust and Dirt
Street Dust and Dirt Accumulation
The Michaelis-Menton equation (Eq.13) was used for estimating the street dust and dirt accumulation between
storm events. Results are presented in Table 14.
Table 14. Street dust and dirt accumulation rates and loading for hypothetical case study
Type of
roadway
Street
Minor
arterial
Major
arterial
Total
Length
(km)
6
2
1
9
Total
Curb Length
(curb-m)
12,000
4,000
2,000
18,000
Maximum
Build Up Limit(1)
(g/curb-m)
250
180
150
Estimated Dust and Dirt
Accumulation^
(kg)
2,760
660
280
3,700
Notes: (1) Selected from published DDLIM values, Delleur (2001)
(2) Between storm events loading calculated from Eq. (13): DDFAC -- 0.9d and T - 10 d
Street Dust and Dirt Washoff
Street dust and dirt washoff loadings were estimated based on the First-order decay equation (Eq.14) and results
are presented in Table 15.
Table 15. Street dust and dirt washoff loadings for hypothetical case study
Type of
roadway
Street
Minor
arterial
Major
arterial
Total
Total
Curb Length
(curb-m)
12,000
4,000
2,000
18,000
Estimated Dust and Dirt
Accumulation (Ay
Between Storm Events
(kg)
2,760
660
280
Washoff
Coefficient
(/O
0.5
0.75
1.0
Solids Washoff
Loadings to Sewer*1*
(kg)
2,530
645
278
3,453
Notes: (1) Loadings per storm calculated from Eq. (14): average rain depth - 5 mm
Each storm carries 3,453 kg solids to the drainage sewer systems. Total solid washoff loadings generated by 20
rainfall (>5 mm) events over a year will be 69,000 kg.
Soil Erosion
The majority of soil erosions are from park open space and landscaped areas. There are no construction activities
in the example, otherwise much higher soil erosion would be generated. The calculations of amount of soil loss
were based on Equation 15 and results are summarized in Table 16.
Table 16. Soil erosion load for hypothetical-case example
Land use
Residential, school, commercial,
and industrial areas
Parks
Total
Landscaped Area
(ha)
384
250
(acre)
950
620
Soil Erosion
ton/yr
25
23
kg/yr
22,600
22,300
44,900
16
-------
The total soil erosion from this urban watershed is estimated about 44,900 kg/yr.
Summary of Solids Loading
A summary of total annual solids loadings of each category is indicated in Table 17.
Table 17, Summary of total annual solids loadings for hypothetical-case example
Solid Category
Litter/floatable solids
Highway/street sands
Street dust and dirt
Soil erosion
Total
Annual Loadings
(kg)
45,700
112,800
69,000
44,900
272,400
A total annual solids loading discharged from the watershed land surface is estimated about 272,400 kg or 272.4
tonnes. However, a highly significant portion of pollution, that in the dissolved solids form is not presented in the
estimated values. Solids falling directly onto the surface of a waterway, such as a large lake, during rainfall is not
accounted for. Sewer sediment contains very high concentrations of organic (oxygen demanding) pollutants and a
significant amount of suspended solids compared to the other categories that are addressed in the Chapter 4.
17
-------
-------
Chapter 4
Sewer Sediment Solids
Introduction
Deposition of sewage solids during dry weather in combined sewer systems has long been recognized as a major
contributor to "first-flush" phenomena occurring during wet-weather runoff periods. Estimation of these loadings
for a given sewer system is an extremely difficult task. Measurement for extended periods is possible but
extremely expensive. Techniques presently available to estimate dry weather deposition in sewerage systems
involve the use of computer models that are both complex and expensive and requiring more effort than
appropriate for preliminary "first-cut" assessments (Sonnen, 1977; Ashley et al., 1999; Bachoc, 1992).
The U.S. Environmental Protection Agency developed a set of generalized procedures for estimating pollutant
loadings associated with dry weather sewage solids deposition in combined sewer systems. It utilizes data and
information from three sewerage systems in eastern Massachusetts (Pisano and Queiroz, 1977) and one in the City
of Cleveland, Ohio (Pisano and Queiroz, 1984). A complete exposition of the analysis procedure, assumptions
and methodologies has been previously given in two aforementioned referenced documents, and will not be
presented here.
The predictive equations developed in the previous study relate the total daily rates of pollutant deposition within
a collection system to physical characteristics of collection systems such as per capita waste rate, service area,
total pipe length, average pipe slope, average diameter and other parameters that derive from analysis of pipe
slope characteristics. Several alternative predictive models were presented reflecting anticipated differences in
the availability of data and user resources. Pollutant parameters include suspended solids (SS), volatile suspended
solids (VSS), 5-day biochemical oxygen demand (BOD5), chemical oxygen demand (COD), organic nitrogen by
the Kjeldahl method (TKN) and total phosphate (TP). Sewer system age and degree of maintenance were also
considered. Factors were presented for estimating the increase in collection system deposition resulting from
improper maintenance. The empirical least squares approach was used to formulate the final equations that are
presented along with summarized results from the previous study.
Overview of Approach
An empirical model relating pollutant deposition loading to collection system characteristics is described in this
Chapter. The approach is to use least squares to fit parameters of a postulated model. The model form is a single-
term power function relating total daily sewage solids deposition over a collection system to simple sewer
catchment characteristics, including service area, length of pipe and average pipe slope. The major steps in the
18
-------
analysis are depicted in Figure 1. Sewer system data including manhole-to-manhole length, slope, size, shape and
age are assembled for entire collection systems in Step A. This information is used in Step B to compare total
daily sewage solids deposited in these collection systems for a wide variety of different operating conditions.
These quantities are estimated using an existing exogenous model that uses extremely detailed information to
compute deposition loadings throughout an entire collection system network.
The simulated deposition loadings for different input conditions constitute the dependent variable data in the
regression analysis. The independent variable data is prepared in Step C as a result of the analysis of assembled
data from Step A and detailed outputs from Step B. The dependent variable data was generated from an
exogenous predictive analysis, while the independent variable data was obtained from primary collection system
data and from a secondary analysis of the exogenous simulation outputs with selected collection system data.
The regression analysis is performed in Step D to prepare the simplified predictive relationships. The entire
process is designed to eliminate using the complicated network model requiring thousands of individual bits of
technical information.
STEP A STEP B
Gather sewer system
atlas data for different
collection systems
STEP C
Simulate detailed network deposition model per
collection system using varying population
density and per capita wastewater rates
Data analysis
Simulation data per
collection system
REGRESSION DATABASE
1. L,A,S,D per collection system
2. Lpo, SPD, SPD/J per collection system
STEP D
Daily deposition loading, 55
per collection system per
discharge (0
Simple predictive relationships
Figure 1. Overview of Method of Approach
Independent and Dependent Variables
The list of variables considered in the regression analysis is the following:
1. Total collection system pipe length (L) ~ ft or m
2. Service area of collection system (A) — acre or ha
3. Average collection system pipe slope (5) -- ft/ft or m/m
4. Average collection system pipe diameter (D) -- in. or mm
5. Length of pipe corresponding to 80% of the solids deposited in the system (LPD) - ft or m
6. Slope corresponding to Lfo (SPO) - ft/ft or m/m
19
-------
7. Slope corresponding to 1/4 of the percentage of pipe length (LPD) below
which 80% of the solids deposit (SPD/4) - ft/ft or m/m
8. Flow rate per capita, including allowance for infiltration (Q) - gpcd or Lpcd
9. Daily total sewage solids deposition loading in collection system (75) -- Ib/d or kg/d
(dependent variable predicted from deposition model)
It was determined in the prior study that the mean pipe slope alone would not be adequate to explain the effects of
the pipe slopes on the variations of the deposition loads. A better characterization of the sewer slopes could be
obtained by defining various parameters at the flatter sewer slope range. Two other pipe slope parameters besides
the mean pipe slope were selected for inclusion into the regression model.
The collection system slope parameters Sfo and SPDM were arbitrarily defined with the sole aim of better defining
the range of the pipe slope distribution function. These collection system slope parameters were defined after
reviewing several plots of the cumulative distribution of pipe slopes for several collection systems. Other choices
could have also been made.
Estimates of collection system pipe length, service area, average pipe slope and average diameter were prepared
from direct inventory and analysis of sewer system alias information. Estimates of LPD, SPD and SPM were
prepared from a detailed analysis of simulated data generated from a complex sewer system network deposition
model. The total daily deposition load, 75, is also computed using this model. Finally, it is clear that the
deposition process is also strongly affected by the wastewater flows in the system. Variations in population
density and the degree of infiltration affects the dry weather flow rates. These effects were incorporated into the
per capita wastewater rates (0 used in the deposition model simulations and in the regression analysis.
Models
Both linear additive and multiplicative models were investigated. Untransformed observed values of the
dependent and independent variables are initially used, leading to a strictly linear regression equation. In another
case the observed values of both the dependent and independent variables were transformed by taking their
natural logarithms, leading to a linear equation in the logarithmic domain which can be put into a non-linear
multiplicative form.
Regression Method
The linear regression program used to empirically establish the relationships of the total daily suspended solids
(7S) deposition within a sewerage collection system with the independent variables is one that operates in a step-
forward manner. At each step in the analysis, the particular variable entered into the regression equation accounts
for the greatest amount of variance between it and the dependent variable, i.e., the variable with the highest partial
correlation with the dependent variable. The program is flexible to allow any independent variable to be: (1) left
free to enter the regression equation by a criterion of the sum of squares reduction; (2) forced into the regression
equation; or (3) kept definitely out of the regression equation in one given selection. The procedure permits
examination of several alternative considerations of the independent variables. It is done by optional selections of
variables to be forced in and out of the regression equation, or to be simply left free to enter the equation using
variance reduction criteria.
Observation of the relative change in the standard error of estimate was used as the stopping rule in the regression
analysis. An increase of the standard error at a given step indicates that the additional information realized by
introducing the variable is off-set by the loss in degrees of freedom. This implies that the particular variable can
be eliminated in the regression equation.
20
-------
The Students T statistics computed for each of the regression coefficients of the final relationships all exceeded
4.0, and averaged about 9.5 using the aforementioned stopping rule. A value of 1.96 is considered statistically
significant at 95 percent confidence limits for large sample size (greater than 100 observations).
Sewer Sediment Solids Loading Models
Introduction
Regression results reported in earlier work (Pisano and Queiroz, 1977 and 1984) are summarized in this section.
Various predictive models are described, relating total suspended solids deposition within a collection system with
independent variables under the assumption of clean pipe conditions. These relationships are therefore applicable
for situations in which the sewer piping system is properly maintained. These equations were developed from
data assembled from three major sewerage systems in Eastern Massachusetts and one in Cleveland, Ohio.
General Description
The physical characteristics of the three major collection systems used in this analysis derived from three prior
studies. The first area, covering portions of West Roxbury In Boston, Dedham, Newton arid Brookline is strictly
separated. The second area covering major portions of Dorchester and South Boston, two neighborhoods of the
Boston metropolitan area, is a mixed, combined and separate area, while the third basin covering a portion of the
City of Fitchburg is served by a combined sewer system. The total pipe footage for all three areas entails 196 mi
of separate and combined sewer systems encompassing a total area of 8.9 mi2.
The Easterly District in the City of Cleveland is bordered on the west by the Westerly District extending along the
Cuyahoga River; on the south by the Southerly District, generally extending along Woodland, Holton, Parkhill
and Abell Avenues; on the east by the communities of Euclid, Cleveland Heights, South Euclid, East Cleveland,
and Shaker Heights; and on the north by Lake Erie. The Easterly District of the City totals approximately 16,000
acres and includes the downtown area, with an additional 25,000 acres tributary from the surrounding areas. The
existing sewerage system within the Easterly District is almost entirely combined. Tributary areas outside of the
city use sewers and drains for conveyance of drainage to downstream water courses. The available topographic
data showed that most of the Easterly District is relatively flat with a ground slope under 2.0% averaging at about
0.5%.
Alternative Model Selections
In this section several regression models are recommended for user application. Alternative forms reflecting the
availability of data and/or user resources will be presented. The simple forms require little data and have the least
predictive reliability, whereas the more complicated models, requiring greater user resources and data availability,
provide estimates with higher reliability.
Equations calibrated with field data collected from Boston and Fitchburg, MA and Cleveland, OH (Pisano and
Queiroz, 1984) are:
Boston and Fitchburg, MA:
Simplest Model:
Intermediate Model:
Elaborate Model:
[R2-0.95]:
TS = 0.001 l(Lu)(S-044)(ff051) (17)
TS = 0.0013 (L'^XD^'X/O^XQ'051) (18)
75 = 0.00073 (LMI)(SPDOK)(SpD/40n)(Q'ost) (19)
21
-------
= 0.0012 (i")(y°43)(e-°54) .................................................. (20)
2X5/»D//15)(e'052) ........................... (21)
Cleveland, OH:
Simplest Model:
[/?2 = 0.88]:
Elaborate Model:
[R2 = 0.94]:
Where:
A = service area of collection sewer system, acre
D = average sewer diameter, in.
L = total sewer length, ft
Lf0 = sewer length corresponding to 80% of the solids deposited in the sewer system, ft
Q = flowrate per capita, including allowance for infiltration, gpcd
S = average sewer slope, m/m
SPD = sewer slope corresponding to LPD, ft/ft
SpD/4 = sewer slope corresponding to W of the percentage of sewer length (LPD) below which 80% of the
solids deposit, ft/ft
TS = daily total wastewater solids deposition loading in collection system, Ib/d
As shown above, all R* values of these regression models are > 0.85. The differences of/?2 values between
Boston and Fitchburg, MA and Cleveland, OH are < 5% for the Simplest Model and < 1% for the Elaborate
Model. However, with all of the uncertainties involved in such calculations, R2 - 0.94 may be as good as R2 =
0.85. With this in mind, using the Simplest Model for a load calculation could be very useful.
Effects of Age and Maintenance
The above regression equations were derived from deposition data computed under the assumption of clean pipes
with no bottom sediments from prior storms. In this section the impact of pooriy maintained systems was
examined by arbitrarily assuming various levels of prior sediment accumulation in the pipes (Pisano and Queiron,
1977). These sediment levels would change the bottom cross-sectional shape of the pipe channel, the depth of
flow, the hydraulic radius, and the shear stress characteristics accordingly.
Two cases simulating different degrees of maintenance other than perfect clean pipe conditions were considered.
In the first case, or the intermediate maintenance category, sediment beds ranging from 1 to 3 in. in depth were
assumed for all pipes with slopes < 0.0075. A sediment bed of 3 in. was assumed for all pipes with slopes <
0.0005. The bed depths then ranged linearly starting at 3 in. for a pipe slope of 0.0005 up to one in. for a pipe
slope of 0.0075. This range was established using judgment and also based on visual inspection of numerous
combined sewer laterals in eastern Massachusetts sewerage systems. In the second category of maintenance, the
zero maintenance case, sediment beds ranging from 3 to 6 in. for the same range of slopes was considered.
Considering the two age and maintenance criteria mentioned here, the deposition model was used to estimate total
deposition loadings for each of the 75 sewer systems for each of the four per capita waste generation rates of 40,
1 10, 190 and 260 gpcd. Before similar regression computations were performed on the deposition results
obtained for pipes with bottom deposits, a comparison was made of the total deposited loads computed under the
assumptions of clean and sedimented pipes.
For each basin the ratios of TS computed for sedimented pipes with sediment beds of 1 to 3 in. and 3 to 6 in. and
the TS values for clean pipes were calculated for all four per capita waste rates considered, i.e., 40, 1 10, 190 and
260 gpcd. The resulting ratios were very stable for a given per capita waste rate for both cases of sediment
deposits. The mean and coefficient of variation of these ratios are presented in Table 18 for both conditions of
bottom deposits.
22
-------
Table 18. Average values of the ratios of computed loads in deposited pipes over clean pipes
Ratios
* ^ 1 -3 in. Diior sediment' 1 O clean Dine
Average Values of Ratios for per Capita Wastewater Rates
(gpcd)
40
1.263(0.18)
1.312(0.14)
110
1.186(0.14)
1.211(0.11)
190
1.128(0.07)
1.151(0.09)
260
1.094(0.12)
1.121 (0.09)
Note: The numbers in parenthesis indicate the coefficient of variation of the ratios.
The results shown on Table 18 suggest that the prediction of TS in sedimented pipes could be accomplished by a
simple functional multiplicative correction of the results given by any of the regression equations for clean pipes.
An equation was fitted using the data of Table 18 for each of the bed deposit conditions. These equations are:
For a system with deposits ranging from 1 to 3 in.:
75.3 „. = l.68Q°°'6TS<:]ean (^ = 0.988) (22)
For a system with deposits ranging from 3 to 6 in.:
7SMta = 1.79 fi*"4 7Scteon (R2 =0.999) (23)
Where: Q = flow per capita, and TSc]em = load of total solids computed from any of the above regression equations
(Eq. 17 to 21).
The tf2 values indicated above refer to the regression of the ratios of TS on the values of flow per capita. The
small difference found between the two conditions of bottom deposits may well be the result of an inappropriate
accounting of these factors by the deposition model. On the other hand it may simply have resulted from the
particular combination of pipe diameters and sediment depths used as data, which may have led to actually small
differences in flow depths above the sediment levels, and therefore small differences in shear stress between the
cases.
Organic Pollutant Loading
A regression was performed between TS and each one of the other 6 indicators, including BOD}, COD, TKN, NHj,
P, and VSS (Pisano and Queiron, 1977). The resulting regression equations arc presented in Table 19, with their
associated correlation coefficients. Estimates of the total daily BOD}, COD, TKN, NH3, P and VSS depositing
loads within a given collection system can be made using the regression equations in Table 19 with the predicted
TS loading calculated from any of the regression equations (Eq. 17 to 22) for clean pipe conditions and the bias
correction factors for pipes with sediment beds given in Eq. 23.
Table 19. Regression of different pollutants an TS
Regression Equation
(Ib/d)
BOD, = 0.344 re1'308
COD = 0.875 re10*
TKN- 0.039 ITS1'135
NH3= 0.0 17 75-0.0336
P = 0.0076 TS -0.006
VSS = 0.689 TS1 'm
Correlation
Coefficient
0.80
0.77
0.67
0.44
0.67
0.97
23
-------
Estimate of Sewer Length and Slope
The total sewer length of the combined sewer system, L, is generally assumed to be known. In cases where it is
not known, crude estimates may suffice and the estimated sewer length, L', can be determined from the service
area of the collection sewer system, A, using the following expressions (Pisano and Queiron, 1977):
For low population density (10-20 people/acre)
L' = 169 (A)093; (R2 = 0.82) (24)
For moderate-high population density (30 - 60 people/acre)
Z,' = 239(A)093; (R2 = 0.82)
..(25)
If data on pipe slope is not available, the average sewer slope, S', can be estimated from the average ground slope
Sg using the following equation (Pisano and Queiron, 1977):
S' = 0.35(5,)°82; (R2=0.96) (26)
Procedure for Estimating TS Deposited
As indicated in Eqs. 17 through 21, the -fi2 values between Simplest Models and Elaborate Models are 0.85 and
0.95 for Boston and Fitchburg, MA and 0.88 and 0.94 for Cleveland, OH, respectively. They are < 5% for the
Simplest Model and < 1% for the Elaborate Model. With all of the uncertainties involved in such calculations, R2
= 0.94 may be as good as R2 = 0.85. With this in mind, using the Simplest Model for a load calculation is
illustrated in the following generalized procedure for estimating TS deposited as shown in Figure 2.
From mean ground slope
determine mean pipe
slope by Eq. 26
1
1
es
r
Determine mean pipe
slope from data
From total area
determine total
pipe length by
Eq. 24 or 25.
Define per-capita
value Q, including
infiltration.
Figure 2. Steps to Determine Deposited Solids (TS).
24
-------
Hypothetical-Case Example
A hypothetical urban watershed is presented to illustrate the application of pollutant loading estimation methods
as described in this chapter. The total drainage area in this example is approxinately 1,200 ha which consists of a
mixture of land uses as described in Table 7 (Chapter 3). The sewer length of e ach land use category is estimated
and summarized in Table 20.
Table 20. Sewered area in each category of land-jse
Land Use
Low density residential areas
High density residential areas
School
Commercial areas
Light industrial areas
Area
(ha)
300
100
20
200
100
The Simplest Model (Eq. 17) was used for calculating sewer sediment solid loading. Results are summarized in
Table 21.
Table 21. Estimated sewer sediment solids loading
Land-use
Low density residential
High density residential
Commercial areas
Light industrial areas
Total
Population Density
(p/ha)
30
150
150
150
Sewer Length(1)
(m)
24,000
12,000
23,000
12,000
Sewer Slope'2'
(m/m)
0.01
0.01
0.01
0.01
Solids Loading"1
(kfi/d)
36.3
17.0
34.7
17.0
105.0
(1} Estimates were based on population density and Equations 24 and 25
(2) Esti mates were based on ground slope and Equation 26
(3) Estimates were based on Equation 17; flowraie per capita, Q = 200 L/d
The estimated total annual sewer-sediment soiids loading is 38,300 kg.
By using the equations listed in Table 19, the organic pollutants associated with TS can be estimated. Results are
summarized in Table 22.
Table 22. Estimated organic pollutant loading
Regression Equation
(Ib/d)
BODS = 0.344 rS1'308
COD ~ 0.875 TS1'04
TKN = 0.039 7/S1 135
NH3 =0.01775 -0.0336
/> = 0.0076 TS -0.006
VSS = 0.689 TSim
Organic Pollutant Loading
Ib/d
152.0
111.0
7.7
1.8
0.8
303.0
kg/d
70.0
50.0
3.5
0.8
0.4
137.0
kg/yr
25,550
18,250
1,280
300
150
50,000
Table 22 results show that the sediment solids contain high level of decomposed human wastes that are the main
source of sulflde (S~) (Nielsen 1991). The root cause of odor and corrosion in collection systems is S", which is
produced from sulfate by bacteria residing in a slime layer on the submerged portion of sewer pipes and
structures. Identification of potential problem areas before structure damage requires field investigation of S~
concentrations in the sewerage system being addressed in Chapter 5.
25
-------
Chapter 5
Hydrogen Sulfide in Sewer
Introduction
Sanitary wastewater solids deposited in combined sewers (CS) systems during dry weather is a major contributor
to the CSO-pollution load. Furthermore, sulfates are released from organic substances contained in the sewer
sediments by bacteria under anaerobic conditions. In the absence of dissolved oxygen and nitrates, sulfates serve
as electron acceptors and are chemically reduced to sulfides and to H2S by bacteria. The hydrogen sulfide is then
converted to sulfuric acid, which disintegrates sewer pipes. Thus, dry weather sewer sediments create odor and
sewer decomposition problems in addition to the CSO pollution (Fan et al., 2001).
The production and release of HjS gas in municipal wastewater collection systems is responsible for numerous
odor complaints and the destruction of sewer pipes and other wastewater facilities. The process begins with the
biological reduction of sulfate to sulfide by the anaerobic slime layer residing below the water surface in
wastewater collection systems. The anaerobic bacteria utilize the oxygen in the sulfate ion as an electron acceptor
in their metabolic processes. The resulting sulfide ion is transformed into H2S gas after picking up two hydrogen
ions from wastewater. Once released to the sewer atmosphere, an aerobic bacteria (Thiobacillus) which resides
on sewer walls and surfaces above the water line consume the r^S gas and secrete sulfuric acid. In severe
instances, die pH of the pipe can reach as low as 0.5. This causes severe damage to unprotected collection system
surfaces and can eventually result in the total failure of the sewer piping and the uncontrolled release of raw
wastewater to the environment.
This chapter describes detailed procedures and methods for conducting field sampling and monitoring procedures
to determine transient levels of H2S generation of agitated beds within the presence of oxygen during simulated
storm conditions.
Background
For domestic wastewater the main source of sulfide (S=) is sulfate (SO*°°). Sulfide generation is a bacterially
mediated process occurring in the submerged portion of combined and sanitary sewers and force mains. Fresh
domestic wastewater entering a collection system is usually free of S". However, a dissolved form of S= soon
appears as a result of low dissolved oxygen content; high-strength wastewater; low flow velocity and long
detention time in the collection system; elevated wastewater temperature; and extensive pumping. The chemistry
of sulfur cycle, microbial process in sewer networks, and mechanisms of corrosion are covered elsewhere
(Thistlethwayte 1972; U.S. EPA, 1985; Hvitved-Jacobsen 2002). This section briefly discusses the relationship
of SCV reduction, biochemical oxidation, and the factors affecting those biotransformations in sewer.
26
-------
Once released from the wastewater as H2S gas, odor and corrosion problems begin. Another type of bacteria
utilizes H2S gas to produce H2SO4 that causes the destruction of wastewater piping and facilities. Operation and
maintenance (O&M) expenditures are required to correct the resulting damage caused by this H2SO4. In severe
instances, pipe failure, disruption of service and uncontrolled releases of wastewater can occur.
The first step in this bacterially mediated process is the establishment of a slime layer below the water level in a
sewer. This slime layer is composed of bacteria and inert solids held together by a biologically secreted protein
"glue" or film called Zooglea. When this biofilm becomes thick enough to prevent dissolved oxygen from
penetrating it, an anoxic zone develops within it. Approximately two weeks is required to establish a fully
productive slime layer or Zooglea film in pipes. Within this slime layer, sulfate reducing bacteria use the sulfate
ion (SO,f), a common component of wastewater, as an oxygen source for the assimilation of organic matter in a
way equivalent to dissolved oxygen is used by aerobic bacteria. Sulfate concentrations are almost never limiting
in normal domestic wastewaters. When SQf is utilized by these bacteria, S= is die by-product. The rate at which
S= is produced by the slime layer depends on a variety of environmental conditions including the concentration of
organic food source or biochemical oxygen demand (BOD), dissolved oxygen concentration, temperature,
wastewater velocity, and the area of the normally wetted surface of the pipe.
As SO<~ is consumed, the S* by-product is released back into the wastewater stream where it immediately
establishes a dynamic chemical equilibrium between four forms of sulfide; die sulfide ion (S1), the bisulfide or
hydrosulfide ion (HS-), dissolved H2S (H2S(liq)), and H2S gas (H2S
-------
process can only take place where there is an adequate supply of H2S(g) (> 2.0 pprn,), high relative humidity, and
atmospheric oxygen. These conditions exist in the most of wastewater collection systems for part of the year. A
pH of 0.5 (approximately equivalent to a 70 mI7L of H2S04 concentration) has been measured on surfaces
exposed to severe f^S^) environments (> 50 ppmv in air).
The effect of H2SO4on concrete surfaces exposed to the sewer environment can be devastating. Sections of
collection interceptors and entire pump stations have collapsed due to loss of structural stability from corrosion.
The process of concrete corrosion, however, is a step by step process that can sometimes give misleading
impressions. The following briefly describes the general process of concrete corrosion in the presence of a sewer
atmosphere:
Freshly poured concrete has a pH of approximately 1 1 to 12, depending upon the composition of mixed
aggregates. This high pH is the result of the formation of calcium hydroxide [Ca(OH)2] as a by-product of the
hydrarion of cement. Ca(OH)2 is a very caustic crystalline compound mat can occupy as much as 25% of the
volume of concrete. A surface pH of 1 1 or 12 will not allow the growth of any bacteria; however, the pH of the
concrete is slowly lowered over time by the affect of carbon dioxide (CC^) and H2S(g). These gases are both
known as "acid" gases because they form relatively weak acid solutions when dissolved in water. COj produces
carbonic acid and H2S produces thiosulfuric and polythionic acid. These gases dissolve into the water on the
moist surfaces above the wastewater flow and react with the Ca(OH)2 to reduce the surface pH. Eventually the
surface pH is reduced to a level that can support the growth of bacteria (pH 9 to 9.5).
The time it takes to reduce the pH is a function of the concentration of CO2 and H2S(g> in the sewer atmosphere. It
can sometimes take years to lower the pH of concrete from 12 to 9, however, in severe situations it can be
accomplished in a few months. Once the pH of the concrete is reduced to about pH 9, biological colonization can
occur. More than 60 different species of bacteria are known to regularly colonize wastewater pipelines and
structures above the water line. Most species of bacteria in the genus Thiobacillus have the unique ability to
convert H2S(8) to H2SO4in the presence of oxygen. Since the production of H2SO4from H2S is an aerobic
biological process, it can only occur on surfaces exposed to atmospheric oxygen.
The color of corroded concrete surfaces can be various shades of yellow caused by the direct oxidation of H2S to
elemental sulfur. This only occurs where a continuous high concentration supply of atmospheric oxygen or other
oxidants are available. The upper portions of manholes and junction boxes exposed to high H2S concentrations
are often yellow because of the higher oxygen content. This same phenomena can be observed around the outlets
of odor scrubbers using hypochlorite solutions to treat high HjS^ concentrations.
Another damaging effect of H2SO4 corrosion concrete is the formation of a mineral called "ettringite" calcium
sulfbaluminate hydrate (3CaO«Al2O»3CaSO4»32H2O) or gypsum (CaSC^HiO) produced by the incomplete
reaction H2SO4 and cement. It forms at the boundary line between the soft CaSO4 layer and the sound,
uncorroded concrete surface. Ettringite is damaging because it is an expansive compound that occupies more
space than its constituents. When ettringite forms, it lifts the corroded concrete away from the sound concrete and
causes a faster corrosion by continually exposing new surfaces to acid attack. Although the rate of concrete loss
depends on a series of factors including ettringite formation, it is not uncommon to see concrete loss of 1 in. per yr
in high sulfide environments.
Sampling and Monitoring
The control of H2S in wastewater systems is of vital importance to the wastewater industry. The biological and
chemical processes resulting in sulfide production in wastewater are well understood, but there are significant
contributing factors about which we know nothing. Settled solids and other debris in sanitary sewers and
wastewater collection systems provide a greatly increased surface area upon which anaerobic sulfate reducing
bacterial slime can grow, thereby increasing the incremental (per ft) sulfide production potential of sewers.
28
-------
Objective of Field Investigation
The objective of field investigation is to determine the change in concentration of dissolved H2S. Subsequent
reduction in H2S(g> can also be measured as a secondary objective. The field monitoring measures the sulfide
production in a sewer containing moderate to heavy settled solids and debris, Siimple and characterize the solids in
the sewer. An additional facet of the field investigation will measure in-situ dissolved sulfide concentrations
inside the interstitial spaces of a typical debris pile in the sewer. From knowledge of the practical pore space
volume and the surface area, the specific sulfide production rate can be determined. The mass of suifide can be
calculated, and H2S could be prevented by cleaning the upstream sewers.
From an analytical standpoint, the primary objectives of field investigation are to determine the following:
The mass of sulfide generated in-situ by a known type and quantity of debris
The reduction of sutfide that can be achieved by removing sewer debris
The sewer headspace HjS concentrations of a clean and dirty sewer
The odor and corrosion reduction achieved by removing sewer debris
The ventilation dynamics of a sewer being flushed
The approximate odor potential of flushing a debris laden sewer
The effect of flushing on the downstream long-term corrosion potential
Hydrogen sulfide and Dissolved Sulfide Sampling and Testing Procedures
Hydrogen sulfide in the gas and dissolved phases are the focus of the field inspections. Hydrogen sulfide gas
testing includes measuring H2S gas concentration at the manholes upstream and downstream from the test area.
Measurements can be taken at the manhole access points before, during and after flushing. Dissolved sulfide
testing includes measuring dissolved sulfide ion concentration in the wastewatcr upstream and downstream from
the test area and at specific points in between. Samples shall be collected before and after flushing. Additional
dissolved sulfide ion measurements can be taken from the debris piles within the sewer, prior to flushing. These
measurements are taken in-situ by a project engineer experienced in sulfide sampling and confined space entry
(CSE). The following procedures shall be used to measure H2S gas and dissolved sulfide ion concentration.
Air samples are analyzed for H2S by three instruments equipped with an extension hose, H2S gas detection tubes,
and a H2S gas detection and recording station. Liquid samples are analyzed using Gastec tubes for dissolved H2S.
Continuous dissolved H2S sampling can be accomplished by an American Sigma Streamline model 800 SL (or
equivalent) automatic portable liquid samplers.
The Industrial Scientific STX70 or TMX412 gas detector and recording station are used to continuously monitor
the H2S gas concentration in the sewer. The Industrial Scientific STX70 and TMX412 are small portable units
enclosed in impact resistant case.
These instruments contain a passive H2S electrochemical diffusion type sensor. The sensor signal is monitored on
a single channel in the range of 0 to 999 ppm at 1 ppm increments. The data is displayed on the LCD digital
display and recorded into memory by the data logger according to preconfigurcd parameters set by the sampling
personnel. The unit has sufficient memory to store 3600 data points and can record H2S gas concentrations at
intervals of 1 s to 5 min. Recording data at S-min interval allows for 12.5 days of continuous data collection.
The STX70 and TMX412 are programmed using a personal computer to record H2S gas samples at specified time
intervals. Typically, data loggers are programmed for 2 minutes, which allows for a total of 5 d. The hydrogen
sulfide gas monitoring station is secured tightly with rope so that it hangs inside the manholes without coming
into contact with wastewater. The gas detectors should be calibrated according to the manufacturers
29
-------
recommended schedule using manufacturer supplied equipment and NBS traceable calibration gas.
Wastewater Sampling
Sensidyne Model 211L and 21 ILL Gastec (or equal) tubes can be used to determine dissolved sulfide
concentrations in wastewater. Gastec tubes draw wastewater by capillary forces through an indicator compound
that reacts with the dissolved sulfide ions and changes from white to light brown color. The range of sulfide
measurements on the 211L and 21 ILL Gastec tubes is 0 to 100 ppm and 0.5 to 20 ppm, respectively. This type of
detection tube can be performed within one minute on freshly collected samples. Immediate field-testing is
required because the sulfide ion is very unstable; it is easily stripped from solution and is easily oxidized by
bacteria and available electron acceptors. The alternative wet chemical methods require that each sample be
preserved and transported to a lab for testing. Both of these actions artificially lower the dissolved sulfide
concentration. Also, the tubes can be carried into the sewer by the project engineer. Liquid grab samples are
collected in 2 in. diameter by 3 in. tail sampling containers lowered to the water surface. The sample is retrieved
and immediately tested for dissolved sulfide concentration. For quality assurance (QA), duplicate measurements
are needed every 10 samples with at least one duplicate sample per day. The QA objective for 211L is +/-15%
(or 0.5 mg/L). The QA objective for 21 ILL is +/-15% (or < 0.25 mg/L). If duplicate samples are not within the
QA objective run duplicate samples for the next 3 samples. If duplicate samples continue not to be within the QA
objective, suspend sampling and contact the manufacturer.
Pore Space Sampling
Liquid samples from within the matrix of the debris layers in the sewer will be collected using a new sampling
apparatus. The apparatus consists of a stainless steel sampling cylinder with an O-ring seat for a Gastec tube and
a sample collection nozzle that consists of a cylinder with perforated walls and a stainless steel screen filter inside.
The entire apparatus is inserted into the debris layer. The cylinder has perforated walls and a stainless steel screen
filter inside. Small debris and water from the void spaces in the debris layer passes through the holes in the
cylinder to the interior screen, which removes the debris and allows water to accumulate for analysis. The driving
force is the pressure differential between the water surface and the sample nozzle in the debris. The device can
also be equipped with a 5 mL syringe to provide additional suction if needed.
Continuous Dissolved Sulfide Sampling
American Sigma Streamline model 800SL automatic portable liquid samplers (or equal) should be used to collect
wastewater samples over a 24-hour period at each desired sampling location. The sampler contains 24 - 575 mL
plastic bottles that can be programmed to collect one 100 mL sample per hour. The ends of the Gastec 21 ILL (or
211L) tube can be broken off and then inserted into each bottle when the sampler is set up (blue end up). As each
bottle is filled, the sulfide reaction and color change occurs. When the sampler is opened the following day, the
dissolved sulfide measurements are read directly off of the scale on the side of the tube. The bottles are cleaned
and re-loaded with fresh Gastec tubes and another day of sampling continues. The sampler can be moved to
another location or left in the same location for additional data collection.
Collecting three samples and testing for dissolved sulfide will confirm die consistency of the continuous sampling
data. The tubes will be read at time zero and read again after 24 hours. Quality assurance and quality control
(QA/QC) from previous continuous sampling projects indicate a change of less than 5% over 24 hours. It is
understood that wastewater streams are different. If the test concentration varies by more than 0.5 mg/L, then the
continuous sampling period will be reduced to a length of time that is within the stated limits.
30
-------
-------
Chapter 6
Sewer Cleaning
Introduction
As stated, the deposition of sewage solids during dry weather in combined sewers has long been recognized as a
major contributor to "first-flush" phenomena. Another manifestation of "first-flush," in addition to the scouring
of materials already deposited in the sewers, is the mobilization of loose solids on the urban ground surface and
transported into the sewerage system by surface storm runoff. These particulates may settle out in the system and
be scoured and resuspended during wet weather periods. Such materials also create "first-flush" loading from
storm drainage systems. Deposition of heavy solids is also a problem in separate sanitary systems.
One underlying reason for considerable sewage solids deposition in combined sewers is the hydraulic design.
Combined sewers are sized to convey many times the anticipated peak dry weather sewage flow. Combined
sewer laterals can carry up to 1000 times the expected background sewage flow. Ratios of peak to average dry
weather flow usually range from 2 to 10 for interceptor sewers. The oversized combined sewer pipes possess
substantial sedimentation potential during dry weather periods. Dry weather flow velocities are typically
inadequate to maintain settleable solids in suspension which tend to accumulate in the pipes. During rainstorms,
the accumulated solids can resuspend and overflow to receiving waters.
Generally, if sediments are left to accumulate in pipes, hydraulic restrictions can result in blockages in flowline
discontinuities. Otherwise, the bed level reaches an equilibrium level. A number of conventional cleaning
techniques are described below, followed by a discussion of various manual and automated flushing methods.
Conventional Sewer Cleaning Techniques
Conventional sewer cleaning techniques include rodding, balling, flushing, poly pigs and bucket machines. These
methods are used to clear blockages once they have formed, but also serve as preventative maintenance tools to
reduce future problems. With the exception of flushing these methods are generally used in a "reactive" mode to
prevent or clear up hydraulic restrictions. As a control concept, flushing of sewers is viewed as a means to reduce
hydraulic restriction problems as well as a pollution prevention approach.
Power Rodding
Power rodding includes an engine driven unit, steel rods and a variety of cleaning and driving units. The power
31
-------
equipment applies torque to the rod as it is pushed through the line, rotating the cleaning device attached to the
lead end. Power rodders can be used for routine preventative maintenance, cutting roots and breaking up grease
deposits. Power rodders are efficient in lines up to 0.30 m (12 in.) in diameter.
Balling
Balling is a hydraulic cleaning method in which the pressure of a water head creates high velocity water flow
around an inflated rubber cleaning ball. The ball has an outside spiral thread and swivel connection that causes it
to spin, resulting in a scrubbing action of the water along the pipe. Balls remove settled grit and grease buildup
inside the line. This technique is useful for sewers up to 0.60 m (24 in.) in diameter.
Jetting
Jetting is a hydraulic cleaning method that removes grease buildup and debris by directing high velocities of water
against the pipe walls at various angles. The basic jetting machine equipment is usually mounted on a truck or
trailer. It consists of water supply tank of at least 3.8 m3 (1,000 gal), a high pressure water pump, an auxiliary
engine, a powered drum reel holding at least 152 m (500 ft) of 1 in. hose on a reel having speed and direction
controls and a variety of nozzles. Jetting is efficient for routine cleaning of snull diameter low flow sewers.
Pigging
Poly pigs, kites, and bags are used in a similar manner as balls. The rigid rims of bags and kites cause the
scouring action. Water pressure moves these devices against the tension of restraining lines. The shape of the
devices creates a forward jet of water. The poly pig is used for large sanitary sswers and is not restrained by a
line, but moves through die pipe segment with water pressure buildup behind it.
Power Bucket
The power bucket machine is a mechanical cleaning device effective in partially removing large deposits of silt,
sand, gravel, and grit. These machines are used mainly to remove debris from a break or an accumulation that
cannot be cleared by hydraulic methods. In cases where the line is so completely plugged that a cable cannot be
threaded between manholes, the bucket machine cannot be used. The bucket machine is usually trailer or truck
mounted and consists mainly a cable storage drum coupled with an engine witli controllable drive train, up to 300
m (1000 ft) of 1.3 cm (1/2 in.) steel cable and various sized buckets and tools. The cable drum and engine are
mounted on a framework that includes aim (36 in.) vertical A-frame high enough to permit lifting the cleaning
bucket above ground level. Typically two machines of the same design are required. One machine at the
upstream manhole is used to thread the cable from manhole to manhole. The other machine is used at the
downstream manhole has a small swing boom or arm attached to the top of the A-frame for emptying cylindrical
buckets. The bottom of the bucket has two opposing hinged jaws. When the bucket is plugged through the
material obstructing the line, these jaws are open and dig into and scrape off the material and fill the bucket.
When the bucket is pulled in the reverse direction, the jaws are forced closed by a slide action. Any material in
the bucket is retained as the bucket is pulled out through the manhole.
Silt Treys
Silt traps (or grit sumps) have successfully been used to collect sewer sediments at convenient locations within the
system with the traps being periodically emptied as part of a planned maintenance program. The design and
operational performance of two experimental rectangular (plan) shaped silt traps in French sewer systems was
reported (Bertrand-Krajewsk et al., 1996). Information on design procedures iind methodology for silt traps is
scarce.
Sewer Flushing Systems
Flushing of sewers either by manual or by automated means is generally meant to reduce hydraulic restriction
problems and infrequently as a pollution prevention approach. Flushing of sewers has been a concern dating back
to the Romans. Ogden (1892) describes early historical efforts for cleaning sewers in Syracuse, New York at the
32
-------
turn of the century. The concept of sewer flushing is to induce an unsteady waveform by either rapidly adding
external water or creating a "dam break" effect by quick opening of a restraining gate. Cleansing efficiency of
periodic flush waves depends on flush volume, flush discharge rate, sewer slope, sewer length, sewer flow rate,
sewer diameter and population density. Maximum flushing volumes at upstream points are limited by available
space, hydraulic limitations and costs. Maximum flushing rates at the downstream point are limited by the
regulator/interceptor capacities prior to overflow.
The relationship between cleaning efficiency and pipe length is important. The goal of flushing is to wash the
resuspended sediment to strategic locations {i.e., to a point where the waste stream is flowing with sufficient
velocity, to another point where flushing will be initiated, to a storage sump that will allow later removal of the
stored contents, or to the wastewater treatment plant). This reduces the amount of solids resuspended during
storm events, lessens the need for CSO treatment and sludge removal at downstream storage facilities, and allows
the conveyance of more flow to the WWTP or to the drainage outlet. Clean sewers provide maximum wastewater
carrying capacity thereby preventing sewer overflows and protecting the environment. There is another benefit to
be gained by maintaining sewers in a clean and free flowing condition—sulfide odor and corrosion reduction.
Manual flushing methods usually involve discharge from a fire hydrant or quick opening valve from tank truck to
introduce a heavy flow of water into the line at a manhole. Flushing removes floatables and some sand and grit,
but is not very effective for removing heavy solids. In recent years, automated flushing equipment has emerged in
France and Germany.
Hydrass®
The Hydrass® flushing system developed in France, and shown in Figure 3, is comprised of a balanced hinged
gate with the same shape of the cross section of the sewer. At low flows the self-weight of the gate holds the gate
in the vertical position and the sewer flow builds up behind the gate. The depth of flow continues to build up
behind the gate until the force created by the retained water becomes sufficient to tilt the gate. As the gate pivots
about the hinge to a near horizontal position, the sewer flow is released and this creates a flush wave that travels
downstream and subsequently cleans any deposited sediment from the invert of the sewer. The gate then returns
to the vertical position and the cyclic process is repeated, thus maintaining the sewer free of sediment. Gates are
positioned in series at intervals dictated by the nature, magnitude and location of the sedimentation problem.
Chebbo et a}. (1995) reported the effective operation of the Hydrass system. This system has been installed on a
segment of the Marseilles Number 13 trunk. A 100 m (328 ft) stretch required about 700 flushes to clean an
initial deposit of about 100 mm {4 in.). Flushing frequency can be reduced if the upstream head can be increased.
For example, the number of flushes with a 0.5 m (1.6 ft) head is 24 times more than that required for a 1.5 m (4.9
ft) head.
r
Figure 3. The Sequence of Hydrass* Sewer Flushing Gate Operates
33
-------
Hydroself
In recent years pollution caused by CSO has become a serious environmental concern. Over 13,000 CSO tanks
have been constructed with over 500 being in-line pipe storage tanks 1.8 to 2.1 m (6 to 7 ft) diameter with lengths
125 to 180 m (400 to 600 ft). Discharge throttles control the outlet discharge to about twice the average dry
weather flow plus infiltration. Many different methods for cleaning these pipes were tried over the years. One of
the most popular flushing systems has been the Hydroself9 system was developed by Steinhardt Wassertechnik,
Taunusstein about 11 years ago (Pisano et al., 1997).
The Hydroself® system is a simple method that uses a wash water storage area and hydraulically operated flap
gates to create a cleaning wave to scour inverts of sewers. This system consist'! of a hydraulically operated flap
gate, a flush water storage area created by the erection of a concrete wall section, a float or pump to supply
hydraulic pressure and valves controlled by either a float system or an electronic control panel. The water level in
the sewer is used to activate the release and/or closure of the gate using a permanently sealed float controlled
hydraulic system. The flushing system is designed to operate automatically whenever the in-system water level
reached a pre-determined level, thereby releasing the gate and causing a "dam break" flushing wave to occur.
Activation by remote control is also possible. This technology does not require an outside water supply, can be
easily retrofitted in existing installations with a minimal loss of storage space, and may operate without any
external energy source. The system consists of a hydraulically operated flap gute, a flush water storage area
created by the erection of a concrete wall section, a float or pump to supply hydraulic pressure and valves
controlled by either a float system or an electronic control panel. The actual arrangement for a given installation
is site dependent. The sewer size, length, and slope determine the flush water volume needed for an effective
single flush of the system.
The Hydroself® system has been used to clean settled debris in sewers, interceptors, tunnels, retention and
detention tanks in Germany and Switzerland. This technology was first used in 1986 for cleaning a tank in Bad
Marienberg (a small town with a population less than 10,000 people, about 100 km northeast of Frankfurt). In
that same year the first two pipe storage projects, using the flushing gate technology, were implemented. This
system has been used extensively in Europe with 284 installations with over 600 units in operation.
Approximately 37% of the projects are designed to flush sewers, interceptors and tunnels ranging from 0.25 m to
4.3 m (0.8 to 14 ft) in diameter and flushing lengths of up to 340 m (1100 ft) for large diameter pipes and up to
1000 m (3300 ft) for small diameter pipes. The balance of flushing gate installations is for cleaning sediments
from CSO tanks. The largest project in Paris, France cleans an underground 120,000 m3 (31.6 Mgal) tank beneath
a soccer field using 43 flushing gates (Pisano et al., 1997).
For large diameter sewers greater than 2 m (78 in.) the flushing system may be installed in the sewer pipe itself.
The required storage volume for the flush water is created by erecting two walls in the sewer pipe to form a flush
water storage area in between the two walls. For the area to remain free of debris, a reasonable floor slope (5 to
20%) must be provided in the storage area. The requirements for the storage area slope will determine, in most
instances, the maximum flushing length possible for a single flush gate. Should the actual flushing length be
longer than this value, then additional flushing gates must be installed to operate in series with the first one. In
order to increase the maximum flushing length it is also possible to build additional flush water storage area by
creating a rectangular chamber in-line or adjacent to the sewer line itself.
BiogestVacuum Flushing System
Biogest® is a system comprised of a concrete storage vault and a vacuum pump system to create a cleaning wave
to scour the inverts of sewers. The system consists of a flush water storage area, diaphragm valve, vacuum pump,
level switches, and a control panel for automatic operation of the system. The water level in the sewer is used to
activate the vacuum pump. The vacuum pump evacuates the air volume from the flush chamber and as the air is
34
-------
evacuated the water is drawn in from the sewer and rises in the chamber. The vacuum pump shuts off when a
predetermined level in the flushing vault is reached. A second level sensor detects the water level in the sewer
and activates the flush wave. The flush wave is initiated by opening the diaphragm valve above the flush chamber
and subsequently releasing the vacuum and vault contents (Pisano et al., 1997).
U.S. EPA Automatic Vacuum Flushing System
A new design of sediment flushing system was developed by the U.S. EPA (2003). The system includes a
flushing-water reservoir that can be installed in either a CSO storage tank or in a combined sewer. The reservoir
has an ingress-egress port through which WWF is received from and discharged and an air release valve that
closes when die reservoir is substantially full to create a water-retaining vacuum. As the system surge passes and
the water level falls, the vacuum seal is broken as air enters the reservoir through an air intake conduit, releasing
the water from the reservoir to flush accumulated sediment solids from the storage tank or combined sewer. The
reservoir defines a box-like receptacle having a top portion and downwardly-extending sidewalls. The floor of the
reservoir is the floor of the storage tank or sewer line flush chamber in which the reservoir is installed. The
ingress-egress port is positioned in one of the sidewalls along the bottom edge thereof. The reservoir opens to the
sewer line flush chamber or storage tank through the ingress-egress port. The opening height of the port is about
2 to 4 in. higher than the historical height of the sediment-solid layer. The air intake conduit extends from an
upper opening in the reservoir to a lower opening along a sidewall, other than the sidewall with the ingress-egress
port. The air intake conduit may be in the form of a rectangular duct defined by a partition wall or in the form of
an air intake tube connected to the reservoir at the upper opening by a tee joint. The lower opening is sized to be
about 30% of the size of the ingress-egress port. The lower opening is about 5 to 8 cm (2 to 3 in.) higher than the
top of the ingress-egress port.
In use during a storm, when the CSO storage tank or sewer line flush chamber downstream of the reservoir is
filling up with W WF during a storm, the flow enters the reservoir through the ingress-egress port in the reservoir.
As the liquid level rises in the reservoir, positive pressure automatically opens the air release valve allowing air to
purge from the reservoir. When the reservoir is full, the air release valve automatically closes. During draining of
the sewer or storage tank (e.g., after a storm), a vacuum is created in the air space of the reservoir that holds the
liquid up in the reservoir. When liquid in the sewer or storage tank is drained to a predetermined level (below the
elevation of the air intake conduit opening), air is drawn into the reservoir via the air intake conduit, breaking the
vacuum inside the reservoir. Thus, water in the reservoir is quickly released through the ingress-egress port to the
downstream storage tank or sewer, resuspending the settled sewer solids and transporting them to a sediment pit
for final disposal.
The reservoir may be installed in an upstream end of the storage tank and/or sewer line with the ingress-egress
port facing the downstream end of the storage tank or sewer line flush chamber. The ends of the reservoir may be
mounted to the floor of the storage tank or sewer line flush chamber. When installed in the WWF storage tank,
the volume of the reservoir will be based on the volume of the storage tank. For sewer line applications, the
reservoir volume will depend on the size and the total length of the sewer line to be flushed.
Hydraulic Laboratory Testing of U.S. EPA Automatic Vacuum Flushing Device
A laboratory hydraulic flume was used to simulate a reach of sewer or storage tank. The flushing device was
fabricated and installed at the head-end of the flume. The removed sediment was collected at the end of the flume
and weighed. Water is held up by vacuum and is released upon dissipation of the vacuum in the vacuum-flushing
device rather than through closing and opening of a mechanical gate in the gate-flushing device. The test results
indicate that sediment removal efficiency of the vacuum-flushing device is close to the gate-flushing device (Guo
etal.,2004).
35
-------
Flushing Small Diameter Sewers
A field research program sponsored by EPA was conducted in the Dorchester area of Boston to determine the
pollution reduction potential of flushing combined sewer laterals. It was concluded that small volume flushing of
a 300 mm (12 in.) diameter pipe at a slope of 0.0049 would transport organics/nutrients and heavy metals
sufficient distances (up to 305 m, 1000 ft) to make the option feasible and attractive (Pisano et al., 1979). The test
segments were flushed three times each on five different days and the effectiveness (i.e., incremental removals at
each downstream manhole by special sampling) of the flushes was empirically estimated based on the observed
results in each field test. Table 23 presents the results of a single 1.4 m3 (375 gal) manual flush to scour, entrain
and transport materials within 30 cm to 46 cm (12 in. to 18 in.) pipes.
Table 23. Percent of pollutant removal by manual flush In small diameter sewers
Pollutant
Organic/Nutrient Deposits
(BOD, TP, TN)
Total Suspended Solids Deposits
Flushing Length
76m (250 ft)
75-90%
75%
Flushing Length
2 13m (700 ft)
65-75'%
55-65%
Flushing Length
305 m (1000 ft)
35-45%
18-25%
Flushing Small Diameter Sewers using a Dosing Siphon
A self-flushing tank, or "dosing siphon", designed to clean small diameter sewers has recently been developed in
Germany as shown in Figure 4 (Pisano et al., 2001). The mechanism is placed in a manhole with an inlet from a
water source such as a catch basin, sump pump, or from infiltrating groundwatcr. When the manhole is filled to a
certain elevation, the mechanism creates a siphon and releases 0.76-1.13 m3 (200-300 gal) of water in the manhole
to the sewer. Since it is designed for smaller pipes, it works with low inlet flows, is less expensive to construct,
and requires less space. It is designed for sewer diameters of 300 mm (12 in.) or less and can clean up to 183 m
(600 ft) of sewer. The dosing siphon type mechanism is a patented device produced by Steinhardt (Pisano et al.,
2001).
Figure 4. Dosing Siphon Top View and External Drum
The siphon mechanism resides in a solid stainless steel external drum open at the bottom to allow fluid within the
manhole storage area to enter the device. Inside the external drum are guides bolted to the drum and attached to
the discharge pipe connected to the sewer being flushed. Within this section is a stainless steel flexible hose
having a solid connection to the sewer at the bottom and oversized section (larger diameter cup) at the top. On
rising water level the flexible hose rises within the drum due to buoyancy force on the cup at the top of the hose.
At a certain level the hose cannot extend any further and is now at maximum elevation. As the water level
continues to rise and then spills over the fixed weir causing an unbalanced force on the top-side of the ring. At
that point the hose collapses inducing the siphon effect, thus rapidly draining die contents of the manhole out the
36
-------
discharge pipe connected to the sewer. The effective volume of the flush equals the product of the height of the
flexible hose and the effective cross section of the manhole.
Tests and Observations of Dosing Siphon
Tests were conducted with the dosing siphon at Cambridge, Massachusetts by the Montgomery Watson Harza
project team on June 8,15, and 21,2001 (Pisano et al., 2001). Pertinent dimensions of the 25 cm {10 in.) vitrified
clay pipe (VCP) segment 82 m (270 ft) test segment, upstream and downstream manholes. The flushing volume
for each test was generated by filling the upstream manhole from a nearby fire hydrant. Seven repetitive flushing
experiments were conducted on each day. Dye was introduced into the flushing waters to note time of arrival of
the flushing wave. Peak velocity was then computed. Sediment characteristics were noted at the end of each
flush. Base flow in the 25-cm (10 in.) segment averaged about 5 cm (2 in.). Volume of flush equaled 0.78 m3
(207 gal) and peak velocities averaged about 1.1 m/s (3.6 ft/s).
Pre-flush and post-flush experiment sediment scrapings within the downstream manhole were performed on June
15 and June 21. All material within a portion of the downstream manhole was removed before the seven
experiments were conducted. After all tests were performed the same area of the manhole was again scrapped.
The following describes an assessment of the results:
Dosing Siphon Testing and Sediment Scraping Results at Museum Street, June 15, 2001
The second round of dosing siphon flush tests was conducted on June 15. The test program consisted of
repeatedly filling with hydrant water end-of-the-line manhole having the dosing siphon directly connected into the
test segment. A fill volume of approximately 0.76 m3 (200 gal) generated a flush wave with a peak velocity noted
82 m (270 ft) at the downstream manhole of 1.1 m/s (3.6 ft/s). The experiment was repeated seven times.
Sediment depths and the nature of sediments in the downstream manhole were visually noted after each flush.
Before starting the flush sequence, a sediment scraping was performed in the downstream manhole from a
predetermined portion of the manhole base and placed in a container. The area was scrapped to the invert. After
the seven experiments were performed, residual sediments in the downstream manhole were again scraped in
exactly the same manner and placed in second container. These samples were retained for visual inspection and
assessment. The following results are noted below.
• Approximately 300 g were collected in the pre-flush sample and about 900 g collected in the post-flush
sample.
• There were 5 stones retained on the # 4 sieve, ranging from 0.64 cm to 1.25 cm ('/Ł in. to Vi in.) in the pre-
flush sample.
• There were 31 stones well retained on the # 4 sieve, ranging from 0.64 cm to 1.9 cm (M in. to % in.) in the
post-flush sample.
The stones were removed and the remaining portions of each sample were visually inspected to approximate
fractions of the residual mass per sieve size with the results presented in Table 24.
Table 24. Approximate fractions of residual mass per sieve size (after rocks removed)
Sieve Size Range
Pre-Flush Sample
Post-Rush Sample
> #10 and < #4
> #50 and < #10
> #200 and < #50
Organic Materials
10%
10%
5%
75%
30%
20%
20%
30%
37
US EPA Headquarters Library
' Mail Code 3404T
202-566-0556
-------
Large grain sand and small gravel are typically retained by the #10 sieve. Material in excess of 0.64 cm (M in.) is
retained on the #4 sieve. Medium grain sand is typically noted as #50 sieve, and very fine sand (i.e., "sugar sand"
found on most Florida beaches) is typically captured by the #200 sieve.
Subsequent to the above observations, the collected rock was then carefully washed. All stones pre-flush and
post-flush were crushed granite and evidently had been inadvertently discharged into the upstream recently
upgraded manhole.
Dosing Siphon Testing and Sediment Scraping Results at Museum Street, June 21, 2001
The third round of dosing siphon flush tests was conducted on June 21. Procedural details were the same as the
round two experiments. A fill volume of approximately 0.76 m3 (200 gal) triggered the device sending flush
water having a peak velocity noted 82 m (270 ft) downstream of 1.1 m/s (3.6 fl/s). The experiment was repeated
seven times. Sediment depths and the nature of sediments in the downstream manhole were visually noted after
each flush.
Before starting the flush sequence a sediment scrapping was performed in the downstream manhole from a
predetermined portion of the manhole base and placed in a container. The area was scrapped to pipe invert. After
the seven experiments were performed residual sediments in the downstream manhole were again scrapped in
exactly the same manner and placed in second container. These samples were then visually inspected.
In the morning of June 22, the two samples were placed in two long plastic garden trays and hand spread for
visual inspection and assessment. The following results are noted below:
• Approximately 500 g were collected in the pre-flush sample and about 1000 g were collected in the post-flush
sample.
• There were 20 stones in excess of #4 sieve, ranging from 0.64 cm to 1.25 cm (V* in. to Vi in.) in the pre-flush
sample.
• There were 38 stones well in excess of #4 sieve, ranging from 0.64 cm to 1.9 cm (V* in. to 3 in.) in the post-
flush sample.
The stones were removed and the remaining portions of each sample were visually inspected to approximate
fractions of the residual mass per sieve size with the results presented in Table 25.
Table 26. Approximate fractions of residual mass per sieve size (after rocks removed)
Sieve Size Range Pre-flush Sample Post-flush Sample
> #10 and < #4 10% 25%
> #50 and < #10 15% 25%
> #70 and < #50 30% 30%
> #200 and < #50 20% 10%
Organic Materials 25%
The #70 sieve gradation for the large amount of small grain sand was added to the observations. The post-flush
sample was far grittier than the pre-flush sample over the entire range. The large oblong rock was a piece of
concrete that had been attached in the sediment bed for a long period as it was discolored and corroded.
Conclusions from Testing
Pisano et al. (2001) concluded that the dosing flushing scheme was capable of transporting large inorganic dense
aggregate by combination of probably bed load movement and perhaps saltation (rising and falling, i.e., bouncing
within the pipe segment). The earlier field research experiments conducted in the 1970s (Pisano et al. 1979)
38
-------
would not have anticipated such a favorable result. It is probable that such favorable transport conditions have
resulted from repeated flushing in short period of time precluding "stickiness" conditions. Since the dosing
siphon device is meant to be filled either by infiltration or inflow mechanisms, repeated operation in a short time
period is a probable design condition.
The other point worth noting is that the residual materials after flushing were more inorganic in nature which is
important from "first-flush" and odor and corrosion prevention perspective. The results although preliminary and
the measures of performance admittedly crude are encouraging.
39
-------
-------
Chapter 7
Sewer Sediment Flushing - A Case Study
Introduction
This chapter describes a case study aimed at assessing the cost-effectiveness of sewer flushing technology from
different performance perspectives. These performance perspectives are minimization of maintenance costs,
reduction of sediments CSO "first-flush", and reduction of sediments to lower H2S levels. This case study uses
information developed from Fresh Pond Parkway Sewer Separation and Surface Enhancement Project in
Cambridge, Massachusetts (Pisano et al., 2001). Grit deposition within both domestic sewerage and storm drainage
systems is a major problem because of general flatness of the area. Presence of several shallow streams that the
sewerage (storm and sanitary) systems must cross under as siphons, and the hydraulic level of the receiving water
body that frequently backwaters the storm systems. To overcome this problem in the area, automated flushing
systems using quick opening (hydraulic operated) flushing gates to discharge collected stormwater will flush grit and
debris to downstream collector grit pits.
Over the last twenty years, the City of Cambridge has enhanced drainage service for improving the water quality in
the Alewife Brook and the Charles River. This area is north and west of Harvard Square and within dense heavily
traveled urban regions.
Background Characteristics
Fresh Pond Parkway Sewer Separation Project
Over the last twenty years, the City of Cambridge has separated old combined systems to sanitary and storm
sewerage systems throughout the city to enhance drainage service and to improve the water quality in the Alewife
Brook and the Charles River. Presently, the City is in the construction phase of separating a 100 ha (250 acre)
catchment North and West of Harvard Square within a highly urbanized and heavily traveled area.
Grit deposition within both existing sewerage and storm drainage systems is a major problem because of general
flatness of the area, presence of several shallow streams that the sewerage (storm and sanitary) systems must cross
then streams under as siphons, and the hydraulic level of the receiving water body that frequently backwaters into
the storm systems. The existing and recently constructed storm drains on Fresh Pond Parkway and Concord
Avenue have invert slopes of approximately 0.0003 to 0.0005. Deposition of any residual stormwater solids not
captured by the surface best management practices (BMPs) that discharge into these conduits would be severe.
Since no chemical salting during winter conditions can be tolerated in the low, flat Fresh Pond Reservation
watershed, heavy winter sanding only exacerbates potential deposition problems. Figure 5 depicts the Wheeler
Street storm drain, which is the wet-weather flow outlet from the catchment area. Sediment deposition was
40
-------
observed up to the spring line of the conduit.
Figure 5. Wheeler Street 2.8 m Storm Drain Sewer Half Filled with Sediments
To overcome this problem, automated flushing systems, using quick opening (hydraulically operated) flushing
gates to discharge collected stormwater, will flush grit and debris to downstream collector grit pits (either sumps
in the flush vault structure or manholes). Grit pits will not be provided on the sanitary systems being flushed.
The storm and sanitary sewer systems to be flushed are within the Fresh Pond Parkway near the Cambridge Water
Treatment Plant (CWTP), continue East to Concord Circle and then northeast to the Fresh Pond Circle. Both
systems then proceed down Wheeler Street. Figure 6 shows the locations of thu sanitary sewer and storm drain
flushing vaults. The piping systems consist of approximately 1000 m (3280 ft) of sanitary trunk sewers, ranging
from 460 mm to 600 mm (18 in. to 24 in.), and approximately 1620 m (5314 ft) of existing storm drains with pipe
sizes ranging from 900 mm to 1.2 m by 1.8 m (36 in. to 4 ft by 6 ft.)
Description of Piping Systems to be Flushed
The storm and sanitary sewer systems to be flushed are located within the catchment area. These systems start on
the Fresh Pond Parkway near the Cambridge water treatment plant, continue Bust to the Concord Circle and then
Northeast to the Fresh Pond Circle. Both systems then proceed down Wheeler Street under the Massachusetts
Bay Transportation Authority - Conrail railroad tracks and terminate near the Alewife Parking Garage. The
piping systems consist of approximately 555 m (1820 ft) of sanitary trunk sewers, ranging from 460 mm to 600
mm (18 to 24 in.), and approximately 1620 m (5314 ft) of existing storm drain:; with pipe sizes ranging from 975
mm (24 in.) to 1.52 m by 1.83 m (5 ft by 6 ft). Figure 6 shows the general locations of the flushing vaults.
Figure 6. Fresh Pond Parkway - Locations of Flushing Vaults
41
-------
Description of Flushing Vaults
Another alternative is to retain pipes with flat slopes, but provide periodic cleaning of these pipes by automatic
passive means to maintain hydraulic capacities. The use of flushing chambers at specific locations, with grit pits
downstream was designed for the Fresh Pond Project. The design utilized quick opening flushing gates
(hydraulically driven) that release stored water to create a "dam break" flush wave to cleanse and move sediments
downstream to a grit pit.
Figure 7 shows a typical storm sewer-flushing chamber with quick opening gate designed for the City of
Cambridge. The Fresh Pond Parkway flushing gate chamber is shown in Figure 8. During a rainfall event,
stormwater from the incoming storm drain fills the sump adjacent to the flush chamber. Then stormwater is
pumped from the sump into the flush chamber. Each flush chamber volume was sized based on the roughness,
slope, size and length of the pipe being flushed. The "flush wave" is designed to have a depth of approximately
75 to 100 mm (3 to 4 in.) and a velocity range between 0.5 to 0.75 m/s (1.6 to 2.5 ft/s) at the end of the pipe
segment being flushed.
SECTION B-B
Figure 7. Flushing Storage Configuration with Flushing Gate Installation
Process water {back wash) from the new Cambridge Water Treatment plant will be pumped to the new sanitary
system and collected in sanitary sewer flushing vaults and used for periodic flushing of the sanitary sewers. This
approach is intended to minimize the daily operation of the system and provide the flexibility of cleaning the
pipes on demand. It would be cost-effective, due to reduced initial capital costs and minimal long term
operational and maintenance costs versus a typical pumping station that requires daily maintenance and power.
42
-------
nut MC GMC
FLUSWHlEfl STWK21 AREA
DRAIN VAULT NO. 1-PLAN
WEKOPENHG
RUSHNrtOATf
RU CT (BETCND)
Figure 8. Fresh Pond Parkway - Flushing Gate Chamber
Physical Site Constraints
The design problems in separating the existing combined sewer system, increasing the level of drainage service
from 1 yr to 10 yrs, and providing a means to routinely flush the sanitary and storm drainage system were
included in the following:
The Fresh Pond Parkway in this area consists of four lanes conveying 30,000 to 50,000 vehicles per day with
several rotaries having multiple directional ingress pathways. The Parkway has been historically a utility corridor
for 17 other major electric, telephone, communication, gas, and water supply conduits. The inverts of both the
sanitary and combined sewers are nearly the same as the sanitary system. Effective sewer separation mandated
re-laying new sanitary trunk systems to permit cross connections. Traffic management was horrific as major
commercial enterprises had direct access to the Parkway and had to be maintained.
43
-------
Design Process
At the onset of the design in mid 1997, the flushing volumes for the storm drain vaults noted in Table 26 were
developed. The design information regarding pipe size, roughness, shape, slope, and the distance between the
proposed flushing vault and the downstream receiving pit (i.e., the flushing length that requires sediment
transport), were provided. These volumes were then adjusted upwards where feasible to account for uncertainty,
expected high amount of sand used during winter operations on Fresh Pond Parkway, and the extreme space
limitations, imposed by other utilities within the Fresh Pond Parkway. It is noteworthy that 17 other utilities share
the same four-lane corridor. Flushing volumes for the sanitary sewers were also upsized. Over the course of the
final design and during construction, pipe sizes, slopes and alignments were field modified due to the complexity
of existing utilities in the streets. At the onset of design, as-built horizontal and vertical alignments of utilities
were only partially known. However, the initial flush volumes remained unchanged. The final piping
configuration and flushing volumes were reanalyzed by the flushing gate vendor in mid 2001 and were deemed
adequate and sufficient.
Table 26. Flush vault design information summary
Location
Downstream Flushed
Pipe Size
(m)
Flushing Pipe Length
(m)
Flush Water Volume
(m3)
Drain Vault #1
Drain Vault #2
Drain Vault #3
Drain Vault #4
Drain Vault #5
Sanitary Vault #1
Sanitary^ Vault #2
0.98, 1.37
0.98,1.07
1.37
1.22, 1. 22 by
1.85 by 1.52,
0.46
0.60
1.83
1.83
393
216
220
343
472
201
350
12.1
12.5
12.2
13.8
44.6
9.6
7.3
Proprietary flushing volume sizing rules have been developed in Germany based on a combination of physical
modeling, mathematical modeling and empirical visual observations of prototype pipe flushing installations using
rapid opening flush gate and other conventional more slowly opening valve schemes. The salient feature of the
flushing gate technology is the ability of the gate to be instantly unlatched, to fully open, and to create flush wave
with initial velocities. Typical gate opening times are 0.1 s with releasing the retained water within 10 s. The
initial gate opening is best characterized as a hydraulic "dam break."
Justifications for providing flushing systems for the new 600 mm sanitary trunk sewer system are provided in
Table 27. Average peak dry weather and peak infiltration flow velocities throughout most of the year excluding
inflow periods will not approach 1 m/s (3.28 ft/s) as a limit. Peak daily velocity and shear stress conditions for the
upstream 450 mm (18 in.) sanitary trunk sewer are less than the estimates provided for the downstream 600 mm
(24 in.) sanitary sewer noted in Table 27.
In addition to the low discharge velocities, the domestic waste tributary to the Fresh Pond Parkway and Concord
Avenue sanitary system is unusual for two reasons. First, the waste contains high quantities of fats, oils and
grease (FOG) discharged into the sewers from the numerous restaurants in the catchment. While a rigorous FOG
program is in place, complete control is not possible. Grease buildups have been a significant problem and are
expected to continue. Second, the new CWTP disposes (by permit) filtration backwash process waste on a daily
basis into the sanitary sewer system. High levels of silt, soils and larger sized inorganic material within a
congealed matrix of coagulants and other flocculation aids wilt be disposed into the sewer system on a daily basis.
Since the new sewers are flat in the area, significant deposition problems exacerbated by the combination of FOG
and CWTP process wastes are expected.
44
-------
Table 27. Design flow and velocity evaluation for 600 mm sanitary trunk sewer
Measured Flows
(11 months)
Flow
(L/s)
Velocity
(m/s)
Shear Stress
(N/m2)
Peak Daily Dry Weather
Flow
Average Yearly Dry
Weather Flow
Average Summer Dry
Weather Flow
79
37
28
0.73
0.58
0.56
1.8
1.3
1.1
The design basis for the self-cleansing of the storm drain system assumed that the peak flow velocities for the 3-
month storm should exceed 1 m/s (3.28 ft/s). The USEPA Stormwater Management Model (SWMM) was used to
simulate system flows for the trunk sewers for the regional 3-month storm having a peak hourly intensity equal to
10 mm/hr (0.4 in./hr) with a total rainfall depth of 50 mm (2 in.). The results indicated that peak velocities for the
new storm drain system consisting of existing drains, rehabilitated combined sewers or new drains (box culvert)
designed to handle up to alO-year storm having peak intensity of 58 mm/hr (2.1!8 in./hr) did not exceed 0.5 m/s
(1.64 ft/s). Flow velocities for lesser, more frequent storms will be even smaller and more problematic with
respect to solids deposition. Automated flushing systems with downstream grit collection were therefore
provided.
Hydraulic Modeling Simulation of Flushing Technology
The Stormwater Management Model (SWMM) with Extended Transport Block (EXTRAN) was used to
investigate the efficiency of the flushing technology. Simulation output takes the form of water surface elevations
and discharge at selected system locations. Computed results are only approximate since EXTRAN does not
model the "dam break" phenomenon inherent to the flush gate technology. EXTRAN was developed by the
USEPA and is described in tola! in the User's Manual (Huber and Dickinson, 1988). The SWMM model was
used to evaluate pipe-flushing facilities in Germany and for the Fresh Pond Paikway Sewer Separation Project in
Cambridge, Massachusetts. The evaluation results for the German facilities have been reported elsewhere
(Pisano, et al., 1998) and have been reported here for completeness.
Evaluation of Systems in Cambridge
The basic conveyance element input data required in EXTRAN are specifications for shape, size, length,
roughness, connecting junctions and ground (rim) and invert elevations. Pipe lengths were discretized into
approximately equal sections. These discretized sections varied from 9 to 60 meters (30 to 200 ft). Pipe sections
were assumed to be circular (equivalent diameters calculated) or rectangular. The following parameters were kept
constant in pipe simulations:
• Computation time increment = 1 s.
• Manning roughness coefficient = 0.013 for new concrete, 0.015 to 0.016 for worn concrete and 0.011
for plastic
• Gate opening time in 6 to 10 s.
• Flow hydrographs at the flushing gate are assumed to increase linearly from zero to a constant flow in
two seconds and also to decrease linearly from the constant rate to ;«ro in two seconds.
• Upstream of the conduit/tank was assumed to be the input and downstream was assumed to be a free
overflow.
45
-------
Table 28 summarizes the hydraulic data and results from the respective flushing gates determined from the
hydraulic modeling evaluations of the Cambridge facilities. The listed results are at the downstream end of the
pipe or channel flushed.
Table 28. Summary of pipe flushing hydraulic modeling simulations in Cambridge, MA
Location
Drain Vault #1
Drain Vault #2
Drain Vault #3
Drain Vault #4
Drain Vault #5
Sanitary Vault #1
Sanitary Vault #2
Flush
Volume
(m3)
12.1
12.5
12.2
13.8
44.6
9.6
7.3
Flush
Length
(m)
393
216
220
343
472
201
350
Pipe
Slope
0.0007
0.0009
0.0008
0.001
0.0001
0.0003
0.001
Pipe
Size
(m)
0.98, 1.37
0.98, 1.07
1.37
1.22, 1.22 by 1.83
1.85 by 1.52,1.83
0.46
0.60
Flow
Velocity
(m/s)
0.39
0.60
0.42
0.34
0.29
0.50
0.42
Flow
Depth
(m)
0.06
0.09
0.06
0.05
0.04
0.08
0.09
Interpolated EXTRAN results
Interpolated EXTRAN results for intermediate locations are noted in Table 29 through Table 35 for each flush
vault. Inspection of the modeling results noted in Table 29 through Table 35 indicates that flushing velocities in
excess of 0.7 m/s at the end of the flushing length are not realized. The flushing gate vendor has reviewed these
results and has noted that EXTRAN does not explicitly model the "dam break" gate opening and release of flush
water within vaults with floor slopes typically at 10% to 20%.
Table 29. Drain vault No.1 EXTRAN results
Distance downstream (m)
0
61
122
183
244
305
366
386
Velocity (m/s)
2.86
0.78
0.67
0.53
0.44
0.39
0.37
0.42
Depth (mm)
470
195
152
110
95
88
85
58
Note: Flush Volume =12.1 m3
Reach 1: Diameter = 150 mm; Length = 175 m
Reach 2: Diameter = 213 mm; Length = 210 m
Table 30. Drain vault No.2 EXTRAN results
Distance downstream (m)
Velocity (m/s)
Depth (mm)
0
61
122
183
208
3.04
1.04
0.81
0.70
0.60
628
210
165
122
95
Note: Flush Volume = 12.5 m*
Reach 1: Diameter = 1067 mm; Length = 17 m
Reach 2: Diameter = 965 mm; Length = 191 m
46
-------
Table 31. Drain vault No. 3 EXTRAN results
Distance downstream (m)
Velocity (m/s)
Depth (mm)
0
61
122
183
220
3.00
0.78
0.61
0.48
0.42
482
195
137
116
64
Note: Flush Volume = 12.2 m3
Pipe Diameter = 213 mm; Flush Length ••
220m
Table 32. Drain vault No. 4 EXTRAN results
Distance downstream (m)
Velocity (m/s)
Depth (mm)
0
61
122
183
244
305
343
2.88
1.52
1.04
0.84
0.37
0.30
0.42
488
210
168
67
61
58
21
Note: Flush Volume = 13.8 m
Reach 1: Diameter = 1219 mm; Length = 175 m
Reach 2: Diameter = 1219 x 1829 mm rectangular; Length = 168 m
Table 33. Drain vault No. 5 EXTRAN results
Distance downstream (m)
0
61
122
183
244
305
366
427
473
Velocity (m/s)
3.17
0.98
0.55
0.55
0.49
0.37
0.29
0.24
0.34
Depth (mm)
851
366
198
168
140
122
116
107
43
Note: Flush Volume = 44.6 m3
Reach 1: Diameter = 1524 x 1829 mm rectangular; Length = 76 m
Reach 2: Diameter = 1829 mm ; Length = 396 m
Table 34. Sanitary vault No. 1 EXTRAN results
Distance downstream (m)
Velocity (m/s)
Depth (mm)
0
61
122
183
201
2.78
0.73
0.62
0.56
0.51
805
213
155
107
76"
Note: Flush Volume = 9.5 nr
Pipe Diameter = 457 mm; Flush Length = 201 m
47
-------
Table 35. Sanitary vault No. 2 EXTRAN results
Distance downstream (m)
Velocity (m/s)
Depth (mm)
0
61
122
183
244
305
352
2.22
1.18
0.92
0.81
0.76
0.40
0.46
457
152
128
113
95
104
58
Note: Flush Volume = 7.2 m3
Pipe Diameter = 610 mm; Rush Length = 352 m
Alternative Sources of Flush Water
Several possible sources of flush waters and collection systems were considered in the initial design of the
flushing systems.
In the initial phase of planning for the new separation system design, several hundred meters of abandoned large
diameter (1-2 m or 3.3-6.6 ft) water distribution conduits were investigated for use as flushing volume collectors
using inputs from nearby catch basins. These conduits would then discharge to downstream flushing gate
structures. These systems were considered feasible, attractive and cost effective given the premium of space and
the high cost of constructing vaults in the parkway system. These systems were not pursued as the design of
major large diameter storm drains was altered, moving the need for the flushing chambers further downstream and
away from the abandoned conduits.
Second, the notion of filling stormwater-flushing chambers from roof drains of the newly constructed Cambridge
Water Department treatment facility was investigated in the initial phase. This concept was also abandoned,
because the need for flushing chambers in this area was obviated as the major new conduits were moved further
downstream.
In the final design, the notion of utilizing pumped stormwater from major drains directly into the storm flushing
chambers was selected. The sanitary systems are planned to be flushed daily using pumped spent filtrate waters
from the new water treatment plant.
Integration of New Conveyance System/Flushing Vaults and Grit Pit Functionalities
As shown in Figure 6, the new sewerage and drainage system piping at the intersection of Fresh Pond Parkway
and Lakeview Avenue. Sanitary Vault #2 and Drain Vault #1 are also depicted in Figure 6. Pumped process
(backwash) filtrate flow from the new Cambridge WTP is daily pumped into Sanitary Vault #1. This vault is
filled and overflow continues 215 m down to Sanitary Vault #2. This scheme is used in lieu of an external water
source to flush the sanitary trunk sewers as the City of Cambridge pays the Massachusetts Water Resources
Authority for disposal of filtrate volumes. Both vaults will be flushed at least once daily. Controls at both vaults
are programmed to flush in sequence once full.
During a rainfall event, stormwater from the incoming storm drain to Drain Vault #1 fills the sump adjacent to the
flush chamber. Then stormwater is pumped from the sump into the flush chamber. A level sensor within the
flushing volume chamber relays water level data to the PLC in the control panel which terminates pump operation
when the chamber reaches a predetermined fill elevation. A level sensor in the downstream storm drain notes
when the water level in down stream drain in sufficiently low to initiate the flushing operation. Activation of the
hydraulic power pack then causes the flush gate to unlatch creating the flush wave. Once the system has been
activated it is possible to repeat the process during a multi-peaked storm event. A generic 24-hr time clock
function adds an additional level of operational flexibility. For example, it is possible to interrogate the system 24
48
-------
hours after the "fist-flush" to unlatch any partially filled flush volumes. This procedure is the same for all other
drain vaults.
An adjustable bottom-acting gate on the side of the entrance to the pump wet w ell controls the depth of storm
flow entry. This feature can be used to ensure that the vault is not filled with base flows and allow bed load
sediment to flow into the sump during storm events. The seven receiving grit pits (either as a manhole or
integrated within the flush vault chamber) have been sized to provide maximum capture volume given the
extraordinary spatial site constraints along the parkway. Average capture volume per pit is about 3 m3.
Inspection of the grit pits is programmed on a quarterly basis with clean out every two years. The flushing
systems will be field tested in the summer of 2001 and will be put into operation when all upstream sewer
separation work has been completed.
Equipment for each pipe flushing system includes an embedded anchoring system, frame, gate, locking and
sealing mechanism, hydraulic cylinder, hydraulic tubing, hydraulic pump, reservoir, valves, mechanical
connections, the electronic control panel, expansion modules, solenoids, motot, relays, timers and level sensing
equipment.
Each flushing gate is fabricated of stainless steel equipped with bronze bushings for the hinges and locking
mechanism. The gate is fixed to the flushing chamber wall stainless steel anchors. The flushing gate hinge
mechanism is designed to allow for full travel and permit manual lifting of gate flap to a minimum of 135° from
the vertical plane (when the gate is fully closed). The hinges are adjustable in two directions. Material for hinges
and the locking mechanism is stainless steel and bronze.
A hydraulic power pack is located in the control panel for each flushing gate. Hydraulic pressure is not used to
lock the gates or to keep the gates closed. The hydraulic cylinder requires no more than 200 psi to release or open
the locking mechanism. There is only one hydraulic cylinder and one hydraulic line for each gate. The hydraulic
cylinder is constructed of stainless steel. There is no sealing material inside or in the operating shaft. The
hydraulic cylinder has no mechanical or friction seals, no piston rings or sealed shaft. The cylinder is sealed and
is leak proof.
There is one manually operated control panel, equipped with a selector switch for LOCAL/OFF/REMOTE
operation. In LOCAL (manual) mode any of the flushing gates linked can be Hushed from the hydraulic power
pack/control panel. In REMOTE mode, a contact closure is the signal to open the flush gate. Water level
indicators are explosion-proof, continuous, flexible, level transmitter type. These devices control the operation of
the flushing system. Water level indicators are located in the flushing volume chamber and in the downstream
pipe to note chamber fill level and the downstream pipe water level. Each level probe is equipped with a stilling
well to protect it from physical damage.
Each control panel controls the local (manual) flushing of the system from a series of cabinet face mounted
pushbuttons and selector switches. Each is equipped with a PLC and a three-position selector switch, which will
allow for LOCAL/OFF/REMOTE operation. Each control panel enclosure houses controls for the hydraulic
equipment and for all of the electronic components. A second enclosure within the first separates the electronic
components from the hydraulic ones. All operational status alarms are manually reset and both alarms and status
lights have dry contacts for future SCADA system connection. Each control panel allows for manual operation
(pushbutton) of the flushing system, so that any of the flush ways may be flushed at random. Once a flush is
started the control panel cannot accept another signal, other than abort, until the flush is completed. It is equipped
with a status light indicating which flushing gate is operating. The panel is also equipped with indicator lights to
show if high water level conditions exist in the storage areas. The PLC is used to control the duration of the flush
by using various internal timers and relays and by taking the water level in the sump into consideration.
49
-------
Drain Vaults and Sanitary Vaults
Table 36 presents the design aspects and functions of each of the flush vaults.
Table 36. Flushing vault functions
Flush Vault Flushing Grit Collection
Function Location
Drain Vault #1 Yes DV#3
Drain Vault #2 Yes D/S Grit
Manhole
Drain Vault #3 Yes DV#5
Drain Vault #4 Yes D/S Grit
Manhole
Drain Vault #5 Yes D/S Grit
Manhole
Sanitary Vault Yes N/A
#1
Sanitary Vault Yes N/A
#2
Notes:
1 . Pumped from local storm drain system manhole.
2. Pumped at vault.
3. Pumped Water Treatment Plant (WTP) filtrate.
4. Vortex separator used to pretreat pumped flow
fromSV#ltoSV#2.
In-line Grit
Collector
No
Yes
No
No
No
No
No
from the WTP by
Flush Water
Source
Stormwater(1)
Stormwater®
Stormwater®
Stormwater®
Stormwater(2)
WTP
WTP
removing heavy grit
Flush Water
Pre-treated
No
No
No
No
No
Yes(4)
Yes
Flushing Vault Used as
Junction Structure
No
Yes
Yes
Yes
Yes
Yes
Yes
from being conveyed by gravity
Operation and Maintenance
In order to maintain effective system operation, routine scheduled maintenance must be performed. Maintenance
procedures for the flush vaults and grit pits are presented below.
Flushing Vaults
A typical flushing gate vault maintenance procedure is outlined below.
Task 1: Check flushing gates including electrical: Each flushing gate vault should be visually inspected on a
monthly basis. The consistent operation of these devices will ensure that the full capacity of the storm
drain is available during the course of a wet weather event. The inspection should include verification of
proper operation of flushing equipment as well as sensing instrumentation and other electrical
equipment.
This task requires two personnel over the course of two hours, as these structures are typically located in
traffic sensitive areas. All equipment operation can be visually verified from the surface and confined
space entry procedures are not required to perform this task.
Task 2: Check instrumentation/controls: Instrumentation and controls will be inspected on a quarterly basis. The
inspection will provide a more detailed analysis of the operation of the flushing vaults. The flushing
vault will be manually activated, as controls and electrical equipment are monitored. This will provide a
direct indication of the status and operation of the equipment.
This task requires two personnel over the course of two hours as these structures are located in traffic
sensitive areas. All equipment operation can be visually verified from the surface and confined space
entry procedures are not required to perform this task.
50
-------
Task 3: Clean pump wet well: Each pump wet well should be cleaned on an annual basis. During the operation
of the wet well pumping units debris and detritus will remain as they 250 |im moving as bed load). These .ire average values from UK
urbanized areas, serviced by catchbasins with little sump volume, with nominal street cleaning. The National
Urban Runoff Program (NURP) reports a median SS value of 180 mg/L (range of median values from 141 mg/L
to 224 mg/L) developed from long term measurements in 21 urbanized catchments (9 across the US). It is
important to note that none of the NURP measurement programs in the early 1980s sampled bed-load as this is
51
-------
extremely difficult to accomplish in practice. European sewer solids research initiatives in the 1990s noted the
importance of the particle size distribution. Pisano and Brombach (1996) reported the results of several hundred
solids settling curves for a wide variety of waste types (dry weather flow, CSO, storm water, street solids,
sediment scraping, pipe slime) collected across North America and Germany over the last two decades.
Using this collective body of information, an assumed mass solids distribution of stormwater solids including both
grit and lighter particles is considered. The distribution is presented in Table 39. Settling velocities noted reflect
worn angular particles at 10 degrees Celsius. Inspection of Table 39 indicates that the greatest preponderance of
materials is associated with solids particles in the 16-62 jjm range associated with settling velocities between 0.02
to 0.25 cm/s (0.008 to 0.1 in./s). As a matter of note, mass settling velocities determined from most settling
column tests of stormwater, which have excluded bed-load materials, are generally within the lower end of the
range noted above. An overall SS concentration, including grit and suspended load, equal to 300 mg/L is assumed
for the heavily urbanized catchment tributary to the Fresh Pond Parkway system.
Table 39. Assumed stormwater runoff solids characteristics
Category
Very fine gravel
Very coarse sand
Coarse sand
Medium sand
Fine sand
Very fine sand
Coarse silt
Medium silt
Fine silt
Very fine silt
Size
(iam)
>2000
>1000
>500
>250
>125
>62
>31
>16
>8
>4
Settling Velocity
(cm/s)
30.0
15.0
7.0
2.8
1.0
.25
.06
.02
.01
.005
% mass
per category
1
2
4
5
14
20
26
18
6
2
Sum = 100
Next, solids removal associated with a comprehensive, closely spaced system of catch basins and fairly rigorous
street sweeping program within the area are included to reduce the above assumed distribution of stormwater
solids). In most areas of Cambridge the rule is generally about one catch basin per 1.5 acres. Typically, street
sweeping (mechanical) occurs within the Fresh Pond Parkway catchment about 12-15 times per yr.
Measured NURP results indicate on the average, 15% to 20% SS reductions for urbanized areas occur when street
sweeping is routinely practiced. Ashley (1992) reported European results noting solids removal per solids sizes
for mechanical type street sweeping. His results have been generalized to fit within the 10 particle sizes noted in
Table 40 and are given below.
Table 40. Solids removal per solids size for mechanical street sweeping
Particle Size (Jim)
>2000
>1000
>500
>250
>125
>62
>31
Otherwise zero
Effectiveness (%)
80
70
60
55
45
30
15
52
-------
Pitt (1984) measured the solids removal effectiveness of 100 catch basins and concluded that solids removal is
principally a function of the rate of incoming gutter flow. Removal rates approach 45% when the inflow is
discharging less than 0.05 cfs and is negligible for flow rates in excess of 1.5 els. Using judgment and research
and development experience with catch basin performance conducted in Dorch;ster by Process Research (1976),
Pitt's results are generalized in Table 41 to fit into the overall conceptual solids distribution scheme used so far.
Table 41. Solids removal par solids size for typical Cambridge urban catchment art a
Particle Size (urn) Effectiveness (%)
>2000
>1000
>500
>250
>125
>62
100
90
80
60
40
20
10
Otherwise zero
Using the above formulations, the initial assumed solids distribution of stormwater into the catchment is reduced
to reflect the collective impacts of the surface-related Best Management Practices (BMP's). The final average SS
concentration after this reduction is 145 mg/L that falls within the range of NURP reported values but higher than
CIRA assumptions for clean stormwater that is 60 mg/L. Removal associated with catch basin programs is 23%
while street sweeping accounts for an overall reduction of 18%. Table 42 shows the characteristics of
stowmwater runoff solids in Cambridge, Massachusetts. The median settling velocity noted in the "final" mass
solids distribution in Table 42 is 0.06 cm/s which is consistent with measured stormwater values.
Table 42. Stormwater runoff solids characteristics In Cambridge, MA urban catchment
Category
Very fine gravel
Very coarse sand
Coarse sand
Medium sand
Fine sand
Very fine sand
Coarse silt
Medium silt
Fine silt
Very fine silt
Size
(um)
>2000
>1000
>500
>250
>125
>62
>31
>16
>8
>4
Settling Velocity
(cm/s)
30.0
15.0
7.0
2.8
1.0
.25
.06
.02
.01
.005
Initial / (Final)
(% mass per category)
I/ (0.0)
2 /(O.I)
4/(0.3)
5 / (0.9)
14/(4.6)
20 /(1 1.2)
26 / (19.9)
18/(18.)
6/(6)
2/(2)
Sum = 100 / (63.0)
Initial SS = 300ml/L; Final SS = 145 mg/L
Runoff Volumes
Average annual runoff volumes for the catchment are computed assuming a total of 1000 mm (3.28 ft) of rainfall
per yr and 75% conversion to runoff. The annual volume of runoff for the Fresh Pond Parkway catchment (100-
ha or 250 acre) is approximately 822,000 kL (217 Mgal).
Potential Wet Weather Solids Deposition
The mass of annual solids deposition within the Fresh Pond Parkway catchment is estimated as follows.
Assuming quiescent settling with an average forward flow velocity of 0.3 m/s (1 ft/s), all particles having settling
velocity greater than 0.06 cm/s (0.02 in./s) are expected to deposit. Table 42 indicates that the mass concentration
of particles having settling velocities less than 0.06 cm/s (0.02 in A) equals 78 mg/L. The difference between the
final average SS concentration (145 mg/L) and the mass concentration of particles with settling velocities less
53
-------
than 0.06 cm/s (0.02 inVs) will settle in the storm pipes. This equals 67 mg/L. Annual solids' depositions in the
Fresh Pond Parkway storm drain system resulting from stormwater inputs are shown in Table 43 below.
Estimates for the annual deposited masses equal the total runoff volume times the concentration of deposited
materials above.
Table 43. Annual solids deposition In the fresh pond parkway system
Runoff Volume
(kL)
822,000
Mass Concentration (mg/L)
(Particles less than 0.06 cm/s)
67
Annual Solids Deposition (kg)
55,000
Bulk specific weights of such fine-grained sediments have been noted by the Construction Industry Research and
Information Association (CIRA) to be 1602 kg/m3 (100 Ib/ft3). This specific weight does not reflect any
waterlogged materials that may be entrapped within the sediments. Floatables and trash generated within the
entire catchment will be inordinately high due to the large preponderance of eating establishments, hotels and
malls. While the catch basins in the area will capture much of this material, some material will invariably escape
from the catch basins into the storm drain system. Much of this material will become water logged and sink due
to the very low forward outflow conditions. To account for this, the bulk specific weight is reduced to 1,250
kg/m3 (85 Ib/ft3). On an annual basis, the seven grit pits will be cleaned twice.
Cost Analysis
Operation and Maintenance Costs
This section includes the basis for estimates of annual utility costs, preventive maintenance, parts replacement and
structural repairs as needed.
Annual operation and maintenance (O&M) cost estimates are summarized in Table 44. The costs are developed
from the stormwater management system operation and maintenance recommendations. The actual costs for
O&M are dependent on the amount of operational equipment in service at any given time. The cost estimates are
based on the labor estimates presented in Tables 37 and 38 annual labor requirements for flushing gate vault and
grit pit, respectively.
The unit costs are based upon the number of man hours estimated to perform the given task plus the cost of
specialized equipment for cleaning as well as the facility operation and maintenance cost including electrical,
structural and mechanical upkeep and repair. The rates include overhead and equipment necessary to perform the
required tasks (i.e., maintenance staff at $75/man-h and vactor/flusher truck at $120/h).
Table 44. Annual operation and maintenance cost estimates
Task
Annual Unit Cost
Flushing Vaults - Inspection
Flushing Vaults - Check Controls
Flushing Vaults - Clean
Flushing Vaults - Electrical, Mechanical & Structural
Grit Pits - Clean
$3,600
$1,200
$1,380
$1,000
$2,400
Cost Effectiveness Analysis of Automated Flushing versus Periodic Manual Sediment
Removal
This analysis presents life cycle costs for two alternative systems to clean the major storm and sanitary systems
described above over a thirty-year period. Catch basin cleaning and cleaning of all incidental lateral lines
tributary to both systems were not included. The cost of each alternative system does not include estimates of
materials to be removed and disposed. Notwithstanding this limitation, all costs necessary to remove deposits to
54
-------
street level using either scheme are included. Assumptions for the cost analysis, of both alternatives are listed
below:
1. Pipe cleaning costs assume inflation rate of 3.12% per yr.
2. Stormwater pipes are cleaned every 5 yrs, and sanitary pipes are cleaned every 3 yrs.
3. Flushing costs are based on inflation rate of 3.12% per yr and discount rate of 7.1% per yr.
4. Discount Period = 30 yrs.
5. Maintenance labor cost = $75/man-hr.
6. Sanitary systems will be flushed daily using spent process water from the water treatment plant.
7. Storm systems will be flushed approximately every two weeks depending on rainfall.
Capital costs for the flushing facilities reflect final construction costs with all change orders, and include
excavation and backfill, hauling, pavement, gravel, dewatering, hazardous soil disposal, piping, traffic
maintenance, equipment, structures and mobilization. Since the flushing facilities have been built at piping
system intersection points, total facility construction costs have been adjusted to only include flushing and grit
capture functions.
Grit pits have been included for the storm systems only. Pit volumes average about 3 m3 (793 gal). Small
diameter 75 mm-100 mm (3-4 in.) force mains from the Cambridge Water Treatment plant to the sanitary flushing
systems are included in the capital cost estimates. Operation and maintenance costs for flushing sites include
hydraulic oil, routine inspection and servicing, power, and removal of collected sediments from the storm system
vaults on a semi-annual basis. Trucking and disposal costs are not included.
Capital Costs for the Automated Flushing Systems
The capital costs of the flushing systems include the flushing vaults, the grit sumps/manholes (storm only), small
above ground vaults to house the hydraulic power pack units to trigger the flushing systems and electrical
pumping controls, and chambers as appropriate to pump storm water into the flushing chambers.
No additional sewage treatment costs associated with added " flush water" is included for the two sanitary sewer
chambers at Fresh Pond Circle as this volume. It is already paid for as spent filtrate from the City of Cambridge
new water treatment plant. No such costs are included for the storm system, as collected Stormwater will be used
to flush the storm drain pipes. Incidental costs of pumping storm water to flushing vaults are included. It is
assumed that on a quarterly basis all vaults will be cleaned of collected materials. Trucking and disposal costs are
not included. Pertinent cost summary details of the flushing systems are given in Table 45 and Table 46.
Table 45. Flushing system capital costs (ENR Construction Cost Index = 6389, August 2001)
Location Gross Construction Cost Apportioned Flushing System Cost
gj {$}
Drain Vault #1 210,000 170,000
Drain Vault #2 290,000 240,000
Drain Vault #3 325,000 265,000
Drain Vault #4 335,000 275,000
Drain Vault #5 771,000 661,000
Downstream DV#5 Grit Pit N/A 80,000
Sanitary Vault #1 187,000 147,000
Sanitary Vault #2 158,000 132,000
Force Mains for Sanitary Vaults N/A 82,000
Totals 2.276,000 2.052.000
55
-------
Table 46. Flushing system operation and maintenance costs
Flushing System Average Annual Cost Present Value Cost
($} i$J
Storm Drain 17,600 1,475,000
Sanitary 7,040 236,000
Total 24.640 1,711,000
The overall present worth cost including capital and operation and maintenance costs over a 30 year period for the
automatic storm and sanitary sewer flushing systems is estimated to equal $3,766,000. Average capital cost of
flushing volume is approximately $18,000/m3.
Costs for Manual Cleaning
It is assumed that the sanitary systems will be cleaned on a three-year cycle and the storm lines cleaned on a five-
year cycle. Unit cleaning costs were obtained from actual contractor bids for the cleaning construction package of
existing storm and sanitary sewers within the project area and then used to estimate cleaning of ail newly
constructed and rehabilitated pipes as follows:
1. 1067 mm (42 in.) Storm Drain -$102.00/m ($34.00/ft)
2. 1219 mm (48in.) Storm Drain -$ 129,00/m ($43.00/ft)
3. 1372 mm (54 in.) Storm Drain -$163.50/m ($54.50/ft)
4. 1829 mm (72 in.) Storm Drain- $267.00/m ($89.00/ft)
5. 1.22 m x 1.83 m (4 ft by 6 ft) Storm Drain -$232.50/m ($77.50/ft)
6. 1.85m byl.52 m (5 ft by 6 ft) Storm Drain - $312/m ($96.00/ft)
7. 457 mm (18 in.) - 610mm (24 in.) Sanitary -$49.50/m ($15.00/ft)
8. Storm Sewer Cleaning Mobilization $55,000
9. Sanitary Sewer Cleaning Mobilization $5,000
Present worth costs for cleaning the storm drain system at 5-yr intervals for a 30-yr period equals $4,692,000.
Similarly, the present worth cost for manually cleaning the sanitary sewer system at 3-year intervals equals
$920,000. Total present worth costs for the 30-yr period equals $5,612,000. No trucking and sediment disposal
costs for either alternative are assumed. On a life cycle basis, the automated flushing scheme is more cost
effective than periodic manual cleaning with savings of $1,850,000. The reader must also be aware that the
avoidance of potential real and societal costs of flooding caused by surcharged and clogged drains and sewers is
not reflected in this cost estimate. In addition, the nuisance level costs associated with traffic disruption on Fresh
Pond Parkway (4 lanes with 50,000 vehicles per day) are also not reflected.
56
-------
-------
References
Ahyerrc, M., G. Chebbo, and M. Saad (2001). Sources And Erosion Of Organic Solids In A Combined Sewer.
Urban Water (G.B.), 2(2).
American Society of Civil Engineers (2001). Guide for Best Management Practice (BMP) Selection in Urban
Developed Areas. Urban Water Infrastructure Management Committee's Task Committee for Evaluating Best
Management Practices, American Society of Civil Engineers, Reston, VA.
American Society of Civil Engineers and Water Environment Federation (1998). Urban Runoff Quality
Management, ASCE, Manuals and Report of Engineering Practice No. 87, Reston, VA. WEF, Manual of
Practice No. 23, Alexandria. VA.
Armitage, N. and A. Rooseboom (2000). The Removal of Urban Litter from Stormwater Conduits and Streams:
The Quantities Involved and Catchment Litter Management Options. Water South Africa, 26 (2), 181.
http://www.wrc.org.za
Arthur S. and R.M. Ashley (1998). The Influence Of Near Bed Solids Transport On First Foul Flush In Combined
Sewers. Wat.Sci.Tech. (G.B.) 37 (1).
Arthur S., R.M. Ashley, and C. Nalluri (1996). Near Bed Solids Transport In Sewers. Wat.Sci.Tech. (G.B.) 33 (9)
Ashley R M., T. Hvitved-Jacobsen, and J-L. Bertrand-Krajewski (1999). Quo Vadis Sewer Process Modelling?
Wat.Sci.Tech. (G.B.) 39 (9).
Ashley, R. M. and R.W. Crabtree (1992). Sediment Origins, Deposition And Build-Up In Combined Sewer
Systems. Wat. Sci. Tech. (G.B.) 25 (8).
Ashley, R. M. and T. Hvitved-Jacobsen (2002). Management of Sewer Sediments. In Wet-Weather Flow In The
Urban Watershed Technology and Management, Field, R. and D. Sullivan (Eds.) CRC Press/ Lewis
Publishers, Boca Raton, FL.
Ashley, R. M., N. Souter, D. Butler, J. Davies, J. Dunkerley, and S. Hendry (1999). Assessment of the
Sustainability of Alternatives for the Disposal of Domestic Sanitary Waste. Wat.Sci.Tech. 39,5.
Ashley, R. M., N. Souter, DJ. Blackwood, S. Hendry, J. Moir, and J. Dunkerley (2000). The Relative
Sustainability Of Rushing or Binning Sanitary Waste. Submitted toJEnvi. Eng American Society of Civil
Engineers, Reston, VA.
Ashley, R.M. and R.W. Crabtree (1992). Sediment Origins, Deposition and Build-up in Combined Sewer
Systems. War. Sci. Techn. 25 (8), 1-12.
Bachoc, A. (1992). Location And General Characteristics Of Sediment Deposits Into Man-Entry Combined
Sewers. Wat.Sci.Tech. (G.B.) 25,(8).
Bertrand-Krajewski J-L., H. Madiec, O. Moine (1996). Two Experimental Sediment Traps: Operation and Solids
Characteristics. Wat.Sci.Tech. 33,9.
Bertrand-Krajewski, J.L. (1993). Solid Transport Modeling in Combined-Sewer Systems: The HYPOCRAS
Model (France). HOUILLE BLANCHE, No. 4, ISSN 0018-6368.
57
-------
Boon, A..G. (1995). Septicity in Sewers: Causes, Consequences and Containment. Wat. Set. Techn. 31 (7),
237-253.
Boon, A..G. and A.R. Lister (1975). Formation of Sulphide in a Rising-Main Sower and its Prevention by
Injection of Oxygen. Prog. Wat. Technol. 7(2), 289-300.
Chebbo, G.( D. Laplace, A. Bachoc, Y. Sanchez, and B. Le Guennec (1995). Technical Solutions Envisaged in
Managing Solids in Combined Sewer Networks, International Conference on Sewer Solids-Characteristics,
Movement, Effects and Control, Dundee, Scotland.
Cornelius, M.; T. Clayton, G. Lewis, G. Arnold, and J. Craig (1994). Utter Associated with Stormwater
Discharge in Auckland City, New Zealand. Island Care New Zealand Trust, Auckland.
Gotham, W. and T. Bidleman (1995). Polycyclic Aromatic Hydrocarbons and Polychlorinated Biphenyls in Air at
an Urban and a Rural Site near Lake Michigan, Environ. Sci. & Technol. 29: 2782-2789.
Delleur, J.W. (2001). Sediment Movement in Drainage Systems. Chapter 14. Stormwater Collection Systems
Design Handbook. Mays, L.W. (Ed.) McGraw-Hill, New York, NY
Ellis, J.B.and DJvl. Revitt. (1982). Incidence of Heavy Metals in Street Surface Sediments: Solubility and Grain
Size Studies. Water, Air, and Soil Pollution, 17 (1).
Fan, C-Y., R. Field, W.C. Pisano, J. Barsanti, JJ. Joyce, and H. Sorensen (2001). Sewer and Tank Flushing for
Sediment, Corrosion, and Pollution Control. Water Resour. Plan. & Managt. 127 (3).
Field, R., and R. Turkeltaub (1981). Urban Runoff Receiving Water Impacts-Program Overview. /. Environ. Eng.
Div. 107,1,83.
Forster, J. (1996). Heavy Metal and Ion Pollution Patterns in Roof Runoff. Proc. 7th Int. Conf. on Urban Storm
Drainage, IAHR/IAWQ Joint Committee on Urban Storm Drainage, Hannover, Germany, 241.
Guo, J. C-Y. (1999). Sand Recovery for Highway Drainage Designs, J. hrig. & Drainag. Eng. 125 (6).
Guo, Q., C-Y. Fan, R. Raghaven, and R. Field (2004). Gate and Vacuum Flushing of Sewer Sediment: Laboratory
Testing. J. Hydulic Eng. ASCE (In Press)
Heaney J P., L. Wright, D. Sample, R. Field, and C-Y. Fan (1999). Innovative Wet-Weather Flow Collection,
Control, and Treatment Systems for Newly Urbanizing Areas in the 21sl Century. In: Sustaining urban water
resources in the 21" century (Ed. Rowney C., Stahre P., Roesner L A.) ASCE. ISBN 0-78440-0424-0.
Herrmann, T., C. Bruns, B. Gatje, and G. Ilgen (1999). Urea - The Reason for Fish Kill after Overflow Events
from Combined Sewers? Proc. The Eighth International Conference on Urban Storm Drainage. August 30 -
September 3, 1999, Sydney, Australia. Edited by I.B. Joliffe and J.E. Ball. The Institution of Engineers
Australia, The International Association for Hydraulic Research and The international Association on Water
Quality. 1296.
Hilts, P. (1996). Fine Particles in Air Cause Many Deaths, Study Suggests. The. New York Times, May 9,1996.
Huber, W.C. and R.E. Dickinson (1988). Storm Water Management Model User's Manual, Version 4, EPA/600/3-
88/OOla; NTIS PB88-236641/AS, U.S. Environmental Protection Agency. Athens, GA.
Hiitter, U. and F. Remmler (1996). Stormwater Infiltration at a Site with Critical Subsoil Conditions:
Investigations of Soil, Seepage Water, and Groundwater. Proc. 7* Int. Conf. on Urban Storm Drainage,
IAHR/IAWQ Joint Committee on Urban Storm Drainage, Hannover, Germany, 713.
Jefferies, C. and R.M. Ashley (1994). The Behaviour Of Gross Solids In Sewer Systems. EWPC, 4 (5).
Hvitved-Jacobsen, T. (2002). Sewer Process: Microbial and Chemical Process Engineering of Sewer Networks.
CRC Press, Boca Raton, Florida.
Krebs, P., P. Hoizer, J.L. Huisman, and W. Rauch (1999). First-Flush of Dissolved Compounds. Wat. Sci. Tech.
39,9,55.
Mason, R.P., N.M. Lawson, A.L. Lawrence, JJ. Leaner, J.G. Lee, and G.R. Shew (1999). Mercury in the
Chesapeake Bay. Marine Chemistry. 65,1-2, 77.
May, R.W.P., J.C. Ackers, D. Butler, and S. John (1996). Development Of Design Methodology For Self-
Cleansing Sewers. Wat. Sci. Tech., (G.B.) 33 (9).
Montrejaud-Vignoles, M., S. Roger, and L. Herremans. (1996). Runoff Water Pollution Of Motorway Pavement
In Mediterranean Area. Proc. 7* Int. Conf. on Urban Storm Drainage, IAHR/IAWQ Joint Committee on
Urban Storm Drainage, Hannover, Germany.
Mull, R. (1996). Water Exchange Between Leaky Sewers and Aquifers. Proc. 7* Int. Conf. On Urban Storm
Drainage, IAHR/IAWQ Joint Committee on Urban Storm Drainage, Hannover, Germany, 695.
58
-------
Newman, T. L. and W.M. Leo (2002). Characterization of Urban-Source Floatables. In Proceedings ofWEF
Specialty Conference 2000: Collection Systems Wet Weather Pollution Control.
Nielsen, P.H. (1991). Sulfur Sources for Hydrogen Sulfide Production in Biofilm from Sewer Systems. Water Sci.
Tech. 23(1).
Nowakowska-Blaszczyk, A. and J. Zakrzewski (1996). The Sources and Phases of Increase of Pollution in Runoff
Waters in Route to Receiving Waters. In: Proc. 7* Int. Conf. on Urban Storm Drainage, IAHR/IAWQ Joint
Committee on Urban Storm Drainage, Hannover, Germany.
Pisano, W.C. and C.S. Queiroz (1977). Procedures for Estimating Dry Weather Pollutant Deposition in Sewerage
Systems, Report No. EPA-600/2-77/120, NTIS PB 270 695, U.S. Environmental Protection Agency,
Municipal Environmental Research Laboratory, Cincinnati, OH.
Pisano, W.C. and C.S. Queiroz (1984). Procedures for Estimating DryWeather Sewage Inline Pollutant
Deposition-Pliase Ii, Report No. EPA-600/2-84/020, NTIS PB 84141 480, U.S. Environmental Protection
Agency, Municipal Environmental Research Laboratory, Cincinnati, OH.
Pisano, W.C., D. Connick, C. Queiroz, and J. Aronson (1979). Dry Weather Deposition and Flushing for CSO
Pollution Control, EPA/600/2-79-133; NTIS PB 80 118 524, U.S. Environmental Protection Agency,
Cincinnati, OH.
Pisano, W.C. and H. Brombach (1996). Solids Settling Curves. Water Environment & Technology, 8(4). Water
Environment Federation, Alexandria, VA.
Pisano, W.C., G. Novae, and N. Grande (1997). Automated Sewer Flushing Large Diameter Sewers. Collection
Systems Rehabilitation and O&M: Solving Today's Problems and Meeting Tomorrow's Needs, Section 12.
Wade, M. (Ed.) Water Environment Federation, Alexandria, VA, 12-9-12-20.
Pisano, W.C.; Barsanti, J.; Joyce, J.; Sorensen, H. (1998). Sewer and Tank Sediment Flushing: Case Studies.
EPA-600/R-98/157. NTIS PB 99-127839INZ, U.S. Environmental Protection Agency, National Risk
Management Research Laboratory, Cincinnati, OH.
Pisano, W.C., J.R. Barsanti, and FJ. Ayotte (2001) Sewer Sediment Flushing Evaluation and Design: A Case
Study. Draft interim report under EPA Contract No. 68-C98-157, Urban Watershed Management Branch,
Water Supply and Water Resources Division, National Risk Management Research Laboratory, U.S.
Environmental Protection Agency, Edison, NJ.
Pitt, R., R. Field, M. Lalor, and M. Brown (1995). Urban Stormwater Toxic Pollutants: Assessment, Sources, and
Treatability. Wat. Environ. Research 67(3): 260.
Process Research, Inc., (1976). A Study of Pollution Control Alternatives for Dorcester Bay, Metropolitan District
Commission, Boston, MA.
Renard, K.G., G.R. Foster, G.A. Weesies, O.K. McCool, and D.C. Yoder (1996). USDA-ARS Agriculture
Handbook No. 703; Predicting Soil Erosion by Water - a Guide to Conservation Planning with the Revised
Universal Soil Loss Equation (RUSLE). Soil and Water Conservation Society, Ankeny, IA.
Saget, A., G. Chebbo, and J-L. Bertrand-Krajewski (1996). The First-Flush Sewer Systems, Wat. Sci. Tech. Vol.
33,9,101-108.
Sakakibara, T. (1996). Roof Runoff Storm Water Quality. Proc. 7* Int. Conf. on Urban Storm Drainage,
IAHR/IAWQ Joint Committee on Urban Storm Drainage, Hannover, Germany, 157.
Sansalone, J. (1996). Immobilization of Metals and Solids Transported in Urban Pavement Runoff. North
American Environ. Congress '96. Anaheim, CA. America Society of Civil Engineers, Reston, VA.
Sansalone, J.J. and S.G. Buchberger (1996). Characterization of Solid and Metal Element Distributions in Urban
Highway Stormwater. In: Proc. 7* Int. Conf. on Urban Storm Drainage, IAHR/IAWQ Joint Committee on
Urban Storm Drainage, Hannover, Germany, 253.
Sanudo-Wilhelmy, S. A. and G.A. Gill (1999). Impact of the Clean Water Act on the Levels of Toxic Metals in
Urban Estuaries: The Hudson River Estuary Revisited. Evniron. Sci. Technol. 33,20, 3477.
Sartor, J.D. and G.B. Boyd (1972). Water Pollution Aspects of Street Surface Contaminants. EPA-R-2-72-081;
NTIS PB , U.S. Environmental Protection Agency, Municipal Environmental Research Laboratory,
Cincinnati, OH.
Saxton, G., K. Siftar, and P. Fowler (1996). Quantifying the Pollutant Characteristics of Snow Versus Snowmelt
Runoff in a Reliable and Economical Manner at Eielson Air Force Base, AK. In: Proc. WEFTEC'96,
Annu. Conf. & Expo., Dallas, TX. Water Environment Federation, Alexandria, VA.
59
-------
Saul, A.J., J.K. Houldsworth, J. Meadowcroft, D.J. Balmforth, C. Digman, D. Ilutler, and J.W. Davies (1999).
Predicting Aesthetic Pollutant Loadings from Combined Sewer Overflows. Proc. The Eighth International
Conference on Urban Storm Drainage. August 30 - September 3,1999, Sydney, Australia. Edited by l.B.
Joliffe and J.E. Ball. The Institution of Engineers Australia, The International Association for Hydraulic
Research and The International Association on Water Quality. 482.
Shaheen, D.G. (1975). Contributions of Urban Roadway Usage to Water Pollution. EPA-600/2-75/ 004; NTIS PB
245 854, U.S. Environmental Protection Agency, Municipal Environmental Research Laboratory, Cincinnati,
OH.
Sonnen, M. (1977). Abatement of Deposition and Scour in Sewers, Report No. EPA-600/2-77/212, NTIS PB 276
585, U.S. Environmental Protection Agency, Municipal Environmental Research Laboratory, Cincinnati, OH.
Squillace, P.J., J.S. Zogorski, W.G. Wilber, and C.V. Price (1996). Preliminary Assessment of the Occurrence and
Possible Sources of MTBE in Groundwater in the United States, 1993 - 94. Environ. Sci. & TechnoL, 30,5,
1721.
St. Michelbach, G.W. and H. Brombach, (1999). Nutrient Impact from CSOs on Lake Constance. Proc. The
Eighth International Conference on Urban Storm Drainage. August 30 - September 3, 1999, Sydney,
Australia. Edited by l.B. Joliffe and J.E. Ball. The Institution of Engineers Australia, The International
Association for Hydraulic Research and The International Association on Water Quality. 474.
Strecker, E.W., M.M. Quigley, B.R. Urbonas, J.E. Jones, and J.K. Clary (2001). Determining Urban Storm Water
BMP Effectiveness, ASCE Journal of Water Resources Planning and Management, Vol. 127, No. 3, p. 114.
Strecker, E. (2002). Low-Impact Development (LID): Is it Really Low or Just l^ower. In Proceedings of an
Engineering Foundation Conference Linking Stormwater BMP Designs and Performance to Receiving Water
Impacts Mitigation; Snowmass, CO, American Society of Civil Engineers, Reston, VA..
Thistlethwayte, D.K.B. (1972). Control of Sulphides in Sewerage Systems. Ann Arbor Science, Ann Arbor, MI.
Thorolfsson, S.T. and J. Brandt (1996). The Influence of Snowmelt on Urban Runoff in Norway. Proc. 7th Int.
Conf. On Urban Storm Drainage, IAHR/IAWQ Joint Committee on Urban Storm Drainage, Hannover,
Germany, 133.
Tomlinson, R.D., B.N. Bebee, A.A. Heyward, S.G. Munger, and R.G. Swartz (1980). Fate and Effects of
Particulates Discharged by Combined Sewers and Storm Drains, EPA-600/2-80- 111, U.S. Environmental
Protection Agency, Municipal Environmental Research Laboratory, Cincinnati, OH.
Trauth, R. and C. Xanthopoulos (1996). Non-point Pollution of Groundwater in Urban Areas. Proc. 7th Int. Conf.
on Urban Storm Drainage. IAHR/IAWQ Joint Committee on Urban Storm Drainage, Hannover, Germany,
701.
U.S. Environmental Protection Agency (1985). Design Manual Odor and Corrosion Control in Sanitary
Sewerage Systems and Treatment Plants, EPA/625/1-85/018, U.S. Environmental Protection Agency,
Washington, D.C.
U.S. Environmental Protection Agency (1985). Design Manual: Odor And Corrosion Control In Sanitary
Sewerage Systems And Treatment Plants. EPA/625/185/018, U.S. Environmental Protection Agency,
Cincinnati, Ohio, USA
U.S. Environmental Protection Agency (1989). Results of the Evaluation Groundwater Impacts of Sewer
Exfiltration, NTIS No. PB95-158358, Municipal Facilities Division, Office of Water, U.S. Environmental
Protection Agency, Washington, D.C.
U.S. Environmental Protection Agency (1992). Manual For Wastewater Treatment/Disposal For Small
Communities. EPA/625/R-92/005, U.S. Environmental Protection Agency, Cincinnati, Ohio, USA.
U.S. Environmental Protection Agency (2000). Exfiltration in Sewer Systems. Report No. EPA/600/R-01/034.
U.S. Environmental Protection Agency, National Risk Management Rese;irch Laboratory, Cincinnati, OH
U.S. Environmental Protection Agency (2002). Considerations in the Design o/Treatment Best Management
Practices (BMPs) to Improve Water Quality. EPA publication number pending. Urban Watershed
Management Branch. Water Supply and Water Resources Division, National Risk Management Research
Laboratory, Edison, NJ. (http://www.epa.gov/ednnrmrl/).
U.S. Environmental Protection Agency (2003). System and Method for Vacuum Flushing Sewer Solids. United
States Patent No. 6,655,402. U.S. Patent and Trademark Office, U.S. Department of Commerce, Washington,
D.C.
60
-------
Vcnkatcsan, M.L, R.P. de Leon, A. van Geen, and S.N. Luoma (1999). Chlorinated Hydrocabon Pesticides and
Polychlorinated Biphenyls in Sediment Cores from San Francisco Bay. Marine Chemistry. 64,1-2,85.
Vignoles, M. and L. Herremans (1995). Metal Pollution of Sediments Contained in Runoff Water in the Toulouse
City. NOVATECH 95, Second International Conference on Innovative Technologies in Urban Storm
Drainage, May 30-June 1,1995. Lyon, France, Organized by Eurydice 92 and GRAIE. p.611.
Wada Y., H. Miura, and K. Hasegawa (1987). Model Building And Analysis Of Runoff Water Quality Of Flush
From The Street Gully Pots. Proceedings of the 4th International Conference on Urban Storm Drainage,
Lausanne, Switzerland, 60-65.
Wada, Y., H. Miura, and O. Muraoka (1996). Influence of Discharge Pollutants from the Highway at Rainfall on
Water Quality of the Public Water Body. Proc. 7k Int. Conf. on Urban Storm Drainage, IAHR/IAWQ Joint
Committee on Urban Storm Drainage, Hannover, Germany, 461.
Weis, P. and J.S. Weis (1996). Leaching from Chromated-Copper-Arsenic (CCA) Treated Wood and Effects in
Chesapeake Bay. Abstract Book: SETAC 17th Annu. Meeting, Washington, DC., 88.
Zogorski, J.S., A.B. Morduchowitz, A.L. Baehr, B.J. Bauman, D.L. Conrad, R.T. Drew, N.E. Korte, W.W.
Lapham, J.F. Pankow, and E.R. Washington (1996). Fuel Oxygenates and Water Quality: Current
Understanding of Sources, Occurrence in Natural Waters. Environmental Behavior, Fate, and Significance.
Executive Office of the President, Office of Science and Technology Policy, Washington, DC.
61
-------
------- |