xvEPA
United States
Environmental Protection
Agency
Office of Research
and Development
Washington, DC 20460
EPA/600/R-08/010
January 2008
Review of Sewer Design
Criteria and RDM Prediction
Methods
-------
EPA/600/R-08/010
January 2008
Review of Sewer Design Criteria and RDII
Prediction Methods
By
Fu-hsiung (Dennis) Lai
In support of:
Development of Capacity Analysis Tools and Associated Technical Documents for SSO
Control Planning
Cooperative Research and Development Agreement
between
EPA National Risk Management Research Laboratory
and
Camp Dresser & McKee Inc.
CRADA216-02
Project Officer
Dr. Fu-hsiung (Dennis) Lai
Water Supply and Water Resources Division
National Risk Management Research Laboratory
Edison, New Jersey 08837
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
Notice
The U.S. Environmental Protection Agency (EPA) through its Office of Research and Development
performed and managed the research described here. It has been subjected to the Agency's peer and
administrative review and has been approved for publication as an EPA document. Any opinions
expressed in this report are those of the authors and do not, necessarily, reflect the official positions and
policies of the EPA. Any mention of products or trade names does not constitute recommendation for use
by the EPA.
II
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the Agency strives to
formulate and implement actions leading to a compatible balance between human activities and the ability
of natural systems to support and nurture life. To meet this mandate, EPA's research program is
providing data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect
our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation
of technological and management approaches for preventing and reducing risks from pollution that
threaten human health and the environment. The focus of the Laboratory's research program is on
methods and their cost-effectiveness for prevention and control of pollution to air, land, water, and
subsurface resources; protection of water quality in public water systems; remediation of contaminated
sites, sediments and ground water; prevention and control of indoor air pollution; and restoration of
ecosystems. NRMRL collaborates with both public and private sector partners to foster technologies that
reduce the cost of compliance and to anticipate emerging problems. NRMRL's research provides
solutions to environmental problems by: developing and promoting technologies that protect and improve
the environment; advancing scientific and engineering information to support regulatory and policy
decisions; and providing the technical support and information transfer to ensure implementation of
environmental regulations and strategies at the national, state, and community levels.
This document has been produced as part of the Laboratory's strategic long-term research plan. It is made
available by EPA's Office of Research and Development to assist the user community and to link
researchers with their clients.
Sally Gutierrez, Director
National Risk Management Research Laboratory
III
-------
Abstract
Rainfall-derived Infiltration and Inflow (RDII) into sanitary sewer systems has long been recognized as a
source of operating problems in sewerage systems. RDII is the main cause of sanitary sewer overflows
(SSOs) to basements, streets, or nearby streams and can also cause serious operating problems at
wastewater treatment facilities. SSOs usually contain high levels of pathogenic microorganisms,
suspended solids, toxic pollutants, floatables, nutrients, oxygen-demanding organic components, and oil
and grease. There are serious concerns of potential health and environmental risks associated with these
discharges.
The Nation's sanitary sewer infrastructure is aging, with some sewers dating back over 100 years.
Nationwide, there are more than 19,500 municipal sanitary sewer collection systems serving an estimated
150 million people and about 40,000 SSO events per year. To assist municipalities in developing plans to
mitigate SSO problems, the U.S. Environmental Protection Agency (EPA) in 2002 signed a cooperative
research and development agreement (CRADA) with Camp Dresser & McKee Inc. (COM) to develop a
public-domain Sanitary Sewer Overflow Analysis and Planning (SSOAP) Toolbox. It contains a suite of
computer software tools to facilitate the analysis of RDII and performance of sanitary sewer systems. In
addition, the CRADA includes a recently published technical guide (EPA/600/R-07/111) and a SSOAP
user's manual being prepared to guide the application of the Toolbox. A beta version of SSOAP is
planned to be released to the public in 2008.
This report primarily provides a literature review of the RDII quantification methods (Chapter 4) to
support the development of the SSOAP Toolbox under the CRADA. The literature review is centered on
the 1999 WERF report in which eight methods are thoroughly assessed using real data from three
sewerage agencies. While there is no single RDII method that is universally applicable, the RTK method
was chosen to be implemented in the Toolbox as it is probably the most widely accepted one. The
method has long been an option in the EPA Storm Water Management Model (SWMM) and is
extensively used. Other RDII methods can be included in future expansion of the Toolbox.
Since RDII is closely associated with the structural conditions of sewers and the hydrologic/hydraulic
criteria used to design them, Chapters 1 to 3 are included to provide background information. Chapter 1
summarizes the nation's wastewater infrastructure conditions and problems, origins of and problems
caused by infiltration and inflow, and EPA regulatory approaches to address the aging systems and SSOs.
Chapter 2 presents the components of wastewater flows that form the basis for sanitary sewer design.
Historical and current sewer design practices and flow design standards of selected states and local
sewerage agencies are described in Chapter 3.
IV
-------
Table of Contents
Notice II
Foreword Ill
Abstract IV
Table of Contents V
List of Acronyms and Abbreviations VII
Chapter 1 Introduction 1-1
National Wastewater Infrastructure Condition and Problem 1-1
Problems Caused by I/I 1-2
Causes of I/I 1-2
EPA Approach 1-3
Cost-effective I/I Control 1-3
Proposed SSO Rule 1-4
Cooperative Research and Development Agreement 1-5
Chapter 2 Components of Wastewater Flows for Sanitary Sewer Design 2-6
Base Wastewater Flow (BWF) 2-6
Sanitary Flow Contribution 2-6
Groundwater Infiltration (GWI) 2-7
Rainfall-Derived Infiltration/Inflow (RDII) 2-7
Chapter 3 Past and Current Sewer Design Practices to Provide Capacities for I/I 3-9
Historical Practices 3-9
Current Practices 3-9
V
-------
The "Ten-State" Standards (1997Edition) 3-9
State and Local Design Requirements 3-10
Chapter 4 RDII Prediction Methods 4-13
Introduction 4-13
Unit Hydrograph 4-13
Literature Review 4-13
WERF Report (1999) 4-13
Review by Crawford, et al. (1999) 4-15
Review by Wright et al. (2001) 4-15
Recommended RDII Prediction Method for SSOAP 4-16
References 4-19
VI
-------
List of Acronyms and Abbreviations
ASCE
BWF
COM
CMOM
CRADA
DWF
EPA
GWI
I/I
LP
MCES
NRMRL
O&M
POTW
RDII
RTK
R-value
SSES
sso
SSOAP
SUH
SWMM
UH
WEF
WERF
WWF
WWTP
American Society of Civil Engineers
Base wastewater flow
Camp Dresser & McKee Inc.
Capacity, Management, Operation and Maintenance
Cooperation Research and Development Agreement
Dry-weather flow
U.S. Environmental Protection Agency
Groundwater infiltration
Infiltration/inflow
Linear programming
Metropolitan Council Environmental Services
National Risk Management Research Laboratory
Operation and maintenance
Publicly owned treatment works
Rainfall-derived infiltration/inflow
The use of R,T,K parameters in the RTK method for RDII prediction
Percent of rainfall volume
Sewer system evaluation survey
Sanitary sewer overflow
Sanitary Sewer Overflow Analysis and Planning
Synthetic unit hydrograph
EPA Storm Water Management Model
Unit hydrograph
Water Environment Federation
Water Environment Research Foundation
Wet-weather flow
Wastewater treatment plant
VII
-------
Chapter 1 Introduction
National Wastewater Infrastructure Condition and Problem
Municipal sanitary sewer collection systems play a critical role in protecting public health in our cities
and towns. They are designed to convey wastewater from their sources to a wastewater treatment plant
(WWTP). Collection systems consist of sewers, pumping stations, force mains, manholes, and all other
facilities used to collect wastewater from individual residential, industrial, and commercial sources. The
performance of these systems can significantly influence the performance of the WWTP. The 1998 Clean
Water Needs Survey (EPA, 2001) identifies more than 19,500 municipal sanitary sewer collection
systems nationwide serving an estimated 150 million people and comprising about 500,000 mi of
municipally owned pipes in publicly owned systems and probably another 500,000 mi of privately owned
pipes that deliver wastewater into these systems. The replacement costs of the nation's sanitary collection
systems are estimated to be from 1 to 2 trillion dollars. Another source (NCPWI, 1988) estimates that
wastewater treatment and collection systems represent about 10-15% of the total infrastructure values in
the United States.
Much of the nation's sanitary sewer infrastructure has been installed over a long time period, with some
sewers dating back over 100 yr. An American Society of Civil Engineers (ASCE) study under
cooperative agreement with the U.S. Environmental Protection Agency (EPA) (Black & Veatch, 1999)
conducted a survey that involved 42 wastewater utilities serving about 26 million people across the
continental U.S. of various sizes and population. The survey indicated that the age of collection systems
ranged from new to 117 yr, with an average age of 33 yr. About 18% of sewers were built in the last 10
yr, 41% in the last 20 yr, 82% in the last 50 yr, and 98% in the last 100 yr. The average sewer density in
this survey is 21 ft of sewer/capita, or about 10 mi/mi2 of service area.
Older sewers were constructed mainly of vitrified clay, brick, and concrete, while modern sewers were
constructed of plastic, ductile iron, steel, and reinforced concrete. A survey conducted for EPA (Arbour
and Kerri, 1998) which included 13 sanitary sewer systems indicated that material distribution of gravity
sewers are: vitrified clay 61%, plastic of all types 20%, reinforced concrete 7%, unreinforced concrete
7%, and other 5%. Over 50% offeree mains use ductile iron. Clay pipes are generally under 915 mm (36
in.) and larger pipes use reinforced concrete to obtain strength especially for resisting vertical loadings.
Sewer joints are probably the most susceptible component of a sewer system for infiltration. The joints in
older clay sewers were made of cement mortar which may have been initially water-tight and root-
resistant, but tend to deteriorate over time because of their rigidity and the potential corrosive conditions
associated with hydrogen sulfide.
Infiltration/inflow (I/I) problems have long been the primary focus related to public sewer lines of
collection facilities (Lai and Field, 2001). As the integrity of a sewer system starts to deteriorate because
of a variety of factors such as old age, traffic load and overburden, poor design, lack of maintenance, the
system's ability to transport wastewater to treatment facilities is impaired. Sewer pipeline stoppages and
collapses are increasing at a rate of approximately 3% per yr. Roots that puncture and grow inside pipes
1-1
-------
cause over 50% of the stoppages, while a combination of roots, corrosion, soil movement, and inadequate
construction are the cause of most structural failures (ASCE, 1994). According to a study conducted by
the Urban Institute (1981), approximately 50 major main breaks and 500 stoppages occur per 1,600 km/yr
(1,000 mi/yr), amounting to an estimated 50,000 breaks and 500,000 stoppages annually in the nation.
Deterioration of joint materials, force main pressure surges, disturbance by construction or direct tapping,
and seismic activity also contribute to collection line failures. These problems result in approximately
75% of the nation's piping systems functioning at 50% of capacity or less (ASCE, 1994).
Problems Caused by I/I
Besides stoppages and collapses, the problems caused by aging and deteriorating collection systems
include excessive I/I that robs capacity in a sanitary sewer system and negatively affects operation of the
entire sewerage system. This I/I problem eventually comes to the attention of the general public in the
form of sewer overflows, sewer backups, equipment failures, facility expansion needs, permit violations,
and increases in operating costs and user fees.
I/I can greatly increase flows and cause unnecessary burdens on the treatment plant and contribute to
sanitary sewer overflows (SSOs). SSOs are untreated sewage overflows from sanitary sewer collection
systems to streets, private property, basements, and receiving waters. SSOs occur when flows exceed the
capacity of a sewer and sewers surcharge. Usually, SSOs are most prevalent during and immediately
after wet weather when flows are high due to I/I. In addition to I/I, contributing factors to SSOs include
sewer blockages from root intrusion, grease build-up, sedimentation, and debris, all of which are not wet
weather related. SSOs usually contain high levels of pathogenic microorganisms, suspended solids, toxic
pollutants, floatables, nutrients, oxygen-demanding organic components, and oil and grease. SSO effects
are many, including: (1) closing of beach and recreational areas; (2) prevention of fishing and shellfish
harvesting; (3) public health risks associated with raw sewage in roadways, drainage ditches, basements,
and surface waters; (4) inhibition of potential development from sewer connection moratoriums; and (5)
financial liability of a community from public relation problems. In San Diego, CA, SSOs threatened
drinking water supplies, creating the potential for serious adverse public health impact (Golden, 1996).
Available data indicate that essentially all large collection systems experience periodic SSOs and between
one-third and two-thirds of the nation's sanitary sewer systems have problems with SSOs or peak flows at
the WWTP. It is estimated that there are about 40,000 SSO events per yr nationwide. The national cost
estimate to mitigate SSOs for the next 20-yr period is $164 billion (2007 dollars) to attain almost zero
overflows per yr per municipality and about $119 billion and $96 billion (2007 dollars), respectively, to
attain one and two overflows per yr per municipality (EPA, 1996a).
Causes of I/I
I/I is caused by stormwater and groundwater entry into faults in the sewer system. The causes of I/I are
not the same. Infiltration is the water entering a sewer system and service connections from the ground
through defective pipes, pipe joints, damaged house lateral connections, or manhole walls. Infiltration
most often is related to a high groundwater table that is observed during a wet season or in response to a
severe storm. Because sewers are underground, signs of accelerated deterioration and capacity limitations
are not readily apparent until there is a major failure. Sewer pipe failures start with cracking, lateral
deflection, crown sag and offset joints, as well as deteriorated mortar and exposed reinforcing caused by
hydrogen sulfide corrosion. Factors that can contribute to deterioration and lead to structural failure
include the size of defect, soil type, trench bedding characteristics, sewer hydraulic regime, sewer
sedimentation, root intrusion, groundwater level and fluctuation, internal and external corrosion, method
1-2
-------
of construction, and loading on a sewer. When a defect is present, outside water may breach the sewer
wall and soil can enter with it. Loss of soil and reduced side support eventually causes sewer collapse by
a random event adjacent to the area, such as a near-by excavation, storm events, or traffic loads.
The production and release of hydrogen sulfide (H2S) gas in sewers contributes to the destruction of
sewer pipes and manholes. The process begins with the biological reduction of sulfate by the anaerobic
slime layer residing on pipe and sewer sediment surfaces below the water in a sewer pipe. The anoxic
bacteria utilize the oxygen in the sulfate ion as an electron acceptor in the 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, aerobic bacteria (Thiobacillus) which reside on sewer walls and
surfaces above the water line consume the H2S gas and secrete sulfuric acid (H2SO4). In severe instances,
the pH of the pipe can reach as low as 0.5. This will cause severe damage to unprotected sewer surfaces
and may eventually result in the total failure of the pipe (Fan et al., 2000; Fan et al., 2001, Fan et al.,
2003).
Inflow is the water discharged into a sewer system and service connections from various sources, e.g.,
roof leaders, sumps, yard and area drains, foundation drains, cooling water discharges, manhole covers,
cross connections from storm sewers and combined sewers, catchbasins, surface runoff, street washwater,
or drainage. All of these inflows, mostly associated with building connections, are usually unauthorized.
They can contribute as much as 70 to 80% of the I/I load (Field and Struzeski, 1972). Public sewers and
private building sewer connections are both becoming older and are subject to I/I. However, house
connection sewers may be in worse conditions because they are ignored and receive little or no
maintenance and inspection. WEF (1999) performed an in-depth study using questionnaires to gather
information on wastewater collection systems, private building sewer connections, and the way
municipalities are addressing I/I problems. A vast majority (83%) of those sewer agencies surveyed have
aggressive programs to detect illegal or unauthorized connections and, once detected, most would
exercise their authority through enforceable regulations to remove them. I/I associated with private
building sewer connections are either caused by the property owner's intention to prevent property from
water damage, such as illegal sump pump connections, or by causes (e.g., cracked or broken lines) that
are beyond the control of property owners. A small sump can pump as much as 50,000 gal/d and can
have a significant, cumulative effect when considered along with the high number of sump pumps in a
sewer system.
Once the infiltrated and inflow waters are combined within the sewer system, their net effect is the same:
robbed sewer capacities and usurped capacities of system facilities such as pumping, treatment, and
overflow regulators.
EPA Approach
Cost-effective I/I Control
For about two decades, from the early 1970s to the late 1980s, the impetus for sewer infrastructure work
in a municipality was dictated by the Federal Water Pollution Control Act Amendments of 1972 which
focused on publicly owned treatment works (POTW) and their discharges. It required all applicants for
federal construction grants for POTW to verify that sewer systems discharging into POTW were not
subject to "excessive I/I," which is the quantity of I/I that can be economically eliminated from a sewer
system by rehabilitation. In other words, the Act required potential grantees to compare the cost of sewer
rehabilitation (including replacement) with the cost of transportation to and treatment at a POTW. Sewer
systems determined to have "excessive I/I" were eligible for rehabilitation in the construction grant while
systems with "non-excessive I/I" were not. Cesareo and Field (1975) described the procedures to perform
1-3
-------
a cost-effectiveness analysis of I/I control of a sewer system as it is related to transport and treatment.
From 1978-1989, the total outlay of construction grant funds required for the identification of sewer I/I
problems was about $2 billion (EPA, 1991). These costs included costs for preliminary I/I analysis or a
detailed sewer system evaluation survey. The total cost associated with replacement and/or major
rehabilitation of existing sewer systems was about $1.8 billion. Major rehabilitation was defined as
extensive repair of existing sewers on the verge of collapse, structurally unsound, or beyond the scope of
normal maintenance programs.
In the early 1990s, as much of the Nation's collection system infrastructure continued to age and
deteriorate, there was a growing concern over the health and environmental risks of SSOs. In response,
EPA organized a National SSO Policy Workgroup in 1993 to institute and implement a national SSO
control policy. In 1995, EPA sponsored a National Conference on SSOs which resulted in many
important papers on SSO control (EPA, 1996b). In May 1999, EPA was directed to develop and issue a
strong national regulation in one year to prevent SSOs from contaminating our Nation's beaches and
jeopardizing the health of our Nation's families. But after an extended review of the draft Rule by
various EPA regional offices and stakeholder organizations, the Rule is currently on hold. Though its
official release status is unclear, its major component, namely, the Capacity, Management, Operation and
Maintenance (CMOM) program, has been adopted as a good operation and maintenance tool by the SSO
communities.
Proposed SSO Rule
The proposed SSO Rule will consist of the following components that are proposed to be included in all
National Pollutant Discharge Elimination System (NPDES) permit requirements for POTW served by
sanitary sewer collection systems:
1) Capacity, Management, Operation and Maintenance (CMOM) program for municipal sanitary
sewer collection systems
2) Prohibition on municipal sanitary sewer system discharges
3) Reporting, recordkeeping, and public notification requirements for municipal sanitary sewer
collection and SSOs
4) Remote treatment facilities
The parts of the proposed SSO Rule that are likely to require substantial engineering analysis efforts are
contained in the CMOM program. This program requires that a NPDES permittee must:
1) Properly operate and maintain all parts of the collection system
2) Provide adequate capacity to convey base and peak flows for all parts of the collection system
3) Take all feasible steps to stop and mitigate the impact of SSOs
4) Provide public notification of overflow events
Requirement (1) of the CMOM program specifies that municipalities conduct inflow elimination or
reduction, cost-effective sewer rehabilitation, and collection system inspection with associated clean out
and repair. As building connection lateral sewers contribute as much as 70 to 80% of I/I, a significant
amount of I/I will not be abated even after proper operation and maintenance (O&M) and rehabilitation of
street sewers. This remaining I/I would be included in meeting CMOM requirements (2) and (3), which
require municipalities to develop a capacity assurance plan to convey peak wet-weather flows (WWF)
and a plan to mitigate SSOs. To develop a capacity assurance plan, it is necessary to know the flow
conveyance capacity at various parts of the collection system under normal dry-weather and "stressed"
1-4
-------
wet-weather conditions. To capture the system response to the dynamic nature of WWF generation and
baseflow variations, a system wide hydraulic evaluation using dynamic models (e.g., EXTRAN in
SWMM4 [Roesner et al, 1988]) would be advantageous and often necessary for identifying the causes of
the SSO problem to allow the development and evaluation of the most cost-effective engineering
solutions (Lai et al., 2000; Lai et al., 2001).
Cooperative Research and Development Agreement
To assist SSO communities in developing SSO mitigation plans, EPA signed a cooperative research and
development agreement (CRADA) in 2002 with Camp Dresser & McKee Inc (CDM) to develop a public
domain computer analysis and modeling toolbox. The Toolbox is named Sanitary Sewer Overflow
Analysis and Planning (SSOAP). It contains a suite of computer software tools to facilitate the analysis
of rainfall-derived infiltration/inflow (RDII) and performance of sanitary sewer systems. It is designed to
provide technical support for complying with the "C" for Capacity in the CMOM program. In addition, a
reference document and a SSOAP user's manual to guide the application of the Toolbox for performing
capacity analysis of a sanitary sewer system and developing SSO control plans will be prepared as part of
the CRADA.
1-5
-------
Chapter 2 Components of Wastewater Flows for Sanitary Sewer Design
Sanitary sewers are constructed primarily to collect and transport the wastewater of a community to a
treatment facility before eventual disposal to a receiving water body. Traditionally, two main criteria in
the design of a sanitary sewer are to carry the peak discharge for which it is designed and to transport
suspended materials to prevent deposition in the sewer (ASCE, 1982). Wastewater flows for design of
sanitary sewers can be divided into two categories: (1) base wastewater flow (BWF) component
associated with flows during dry-weather periods, and (2) extraneous flow component associated with
flows from wet-weather events.
The BWF component primarily includes sanitary flow contribution from residential, commercial,
industrial, and institutional users. BWF rates typically vary throughout the day, with the peak flow
generally occurring during the morning hours. It also includes some amount of groundwater infiltration
(GWI), particularly in areas where groundwater table is high.
Extraneous water enters the sewer system during wet-weather periods through cracks and open joints in
sewer mains, manholes, and building laterals, as well as through direct connections between storm drains
and sanitary sewer and from illegal drainage connections on private property. These extraneous flows,
termed RDII, can cause significant increases in peak flows in the system. In designing sanitary sewer
systems, an engineer must estimate the current and future BWF and RDII to insure that sewers have
adequate conveyance capacities for the determined design horizon.
Estimation of wastewater flows involves determining the amount of each of the wastewater flow
components (BWF, GWI, and RDII), as well as the time variations of flow associated with each
component.
Base Wastewater Flow (BWF)
Sanitary Flow Contribution
Sanitary flows are largely a function of population, population density, water consumption, and land uses.
Hence, sanitary flow estimation usually involves a study of existing and projected land uses and
demographic data, and water consumption from which to estimate per capita daily water consumption.
The per capita wastewater flows are usually expressed as percentages (or "return ratio") of per capita
water consumptions. The "return ratios" for residential and non-residential uses may vary. These ratios
can be determined from a careful analysis of a long-term (e.g., a full year) monitored sewer flow data and
water consumption data taking into consideration of groundwater infiltration flows. In estimating
probable future per capita wastewater contributions, changes in indoor and outdoor water use habits from
the growing water conservation ethic and associated requirements should be considered. Special
consideration needs to be given to industrial contributions as the rates vary with the type, size, and
operation of the industry. Furthermore, peak discharges may be the result of flows contributed over a
short operation time of the industry. Flows important to the design of sanitary sewers are daily minimum
and maximum, daily average, and peak flow. Peak flows estimated for the end of the design period
2-6
-------
usually determine the desired hydraulic capacities of sanitary sewers, pumps, and some treatment plant
conduits (ASCE, 1982).
Groundwater Infiltration (GWI)
GWI refers to that portion of wastewater embedded in the monitored dry-weather flow (DWF) data at a
sewage pump station or a treatment facility. The rate of GWI depends on the number and size of defects
within a sewer and the hydraulic head available and hence is greater in wet spring high groundwater table
season than in other seasons. As sewer pipes age and deteriorate, GWI is expected to increase in amount
and scope. GWI problems can be particularly severe in areas where sewer systems are installed below the
groundwater table. This unwanted flow results in the need for larger sewerage and increased O&M costs
for pumping and treatment. Hence, in the design of sewers, it is customary to include a rate of about
1,000 gal/acre-d, some to require as high as 2,000 gal/acre-d (ASCE, 1982).
GWI rates can vary a great deal in a sewer system and local flow monitoring data are needed for
determination of GWI rates that meaningfully reflect the gross site-specific conditions. If permanent
monitoring stations do not already exist, temporary monitoring stations are installed for flow data
collection. The best time for flow monitoring for BWF and GWI determinations would be in early spring
when groundwater table is high and outdoor water uses are low. During this time, wastewater from a
residential area may be assumed to be the same as the billed water use and the GWI can be calculated as
the difference between the measured DWF and the wastewater flow determined from the billed water use.
Rainfall-Derived Infiltration/Inflow (RDII)
RDII is that portion of a sewer flow hydrograph above the normal dry-weather base flow pattern. It is a
sewer flow response to rainfall or snowmelt in a sewershed. The term of "I/I" probably first appeared in
the 1960s. Prior to that time, the main focus was the determination of "infiltration" in DWF. "Inflow,"
referred to as the "stormwater" contribution, was recognized but was only casually addressed (ASCE,
1962; HES, 1968). With the influx of federal money and enforcement of federal requirements in mid-
1970s, the amount of "I/I" data surged as did the interest of flow prediction of RDII from monitoring data
particularly since late 1980s.
RDII has long been recognized as a major factor in the sizing of sewer pipes and treatment plants. It was
found by tests that as much as 150 gal/min may leak through a manhole cover as stormwater inflow
(Rawn, 1936). With manholes placed 300 to 500 ft apart, this would amount to 3.5 to 2.0 Mgal/mi-d of
sewer pipe (Babbitt, 1947). It was recognized in the early 20th century that improper connections of roof
drains, street inlets, and foundation and cellar drains to sanitary sewers, in combination with poor quality
of house lateral construction, depleted the reserved sewer capacities that were usually built-in for the
future growth of an area. These unwanted entries of stormwater to sanitary sewers were considered a
"misuse" or "abuse" of the system that should be "prohibited" (Metcalf & Eddy, 1928).
The amount of this extraneous water had never been adequately reported in general terms because of the
difficulty in quantifying the flows accurately. To cope with this problem in the early design of sewers,
various provisions for design flow rates were adopted by various State's Board of Health. For example,
the Illinois State Board used 300 to 350 gal/d for each person to be served. The Missouri State Board
specified the use of per capita flows ranging from 500 to 1,000 gal/d depending upon infiltration,
anticipated stormwater connections, and possible future development (Babbit, 1947). For the design of
the sanitary sewer system for the Beargrass Interceptor District in Louisville, KY, the maximum sewage
rates of 875, 500, and 400 gal/capita-d were used respectively for areas draining 10, 250, and 1000 acres
(Metcalf & Eddy, 1928). These rates included GWI of about 2,000 gal/acre-d. Earlier, the design flows
for the City's sewers were based on a per capita flow rate of about 300 gal/d.
2-7
-------
Flow monitoring data such as from the City of Houston, TX shows that wet-weather peaking factors
(peak WWF to average DWF ratio) of 30 are commonly recorded, and factors reaching 50 have been
recorded in individual basins (Jenq et al., 1996). Hence, better flow prediction methods with parameters
calibrated with site-specific data must be used to insure that sewers are provided with adequate
conveyance capacity throughout the design life of the system. A reliable estimate of RDII is critical in
developing an effective and cost-effective plan to control SSOs.
-------
Chapter 3 Past and Current Sewer Design Practices to Provide Capacities for
I/I
Historical Practices
Human beings have built sewers since 3500 BC (Babylonian time) but these sewers (in the form of open
ditches) were initially built for drainage of storm water away from populated areas (Schladweiler, 2001).
As time passed, people began to realize the need of getting human wastes away from their home. Storm
water drainage pipes were the most convenient means of resolving the problems. Smaller connector pipes
were built from home to transport waste via these storm sewers to streams and rivers. The original storm
sewers then became combined sewers. While combined sewers continued to be built until about the
1940s, people began to understand in the mid-1800s (Dr. Snow of England) that "filth", when mixed with
their water supply, often resulted in disease (cholera outbreak) and death. Later, Dr. Louis Pasteur
discovered that germs convey disease (Schladweiler, 2001). Since then, cities and municipalities in
Europe and the United States began installing public sewer and treatment systems to address health and
aesthetic concerns. A lot of "standards" for design and maintenance of sewers set in late 1800s through
early 1900s are basically still in use today. The changes since then are primarily in pipe materials and
installation methods, and better tools for monitoring and detection of sewer defects and for maintenance.
For instance, J. W. Bazalgette, Chief Engineer of the Metropolitan Commission of Sewers (of London) in
and around 1852 adopted a mean velocity of 2.2 ft/s as the sewer design velocity for preventing siltation
in intercepting and outfall sewers running half full (Metcalf and Eddy, 1928). This is still considered a
good design practice today. Metcalf and Eddy (1928) had a detailed historical count of earlier
engineering practices.
Current Practices
The "Ten-State" Standards (1997Edition)
The design of new sewer systems must conform with state and local regulations that provide minimum
design standards, acceptable methods for estimating peak design flow, and minimum design criteria to
ensure that flow is conveyed through the system with enough velocity to scour out materials that settle in
pipes during periods of low flows. In 1947, 10 states (Illinois, Indiana, Iowa, Michigan, Minnesota,
Missouri, New York, Ohio, Pennsylvania, and Wisconsin) in the Great Lakes-Upper Mississippi River
region formed a committee to review existing design standards for sewage works and published joint
standards in 1951. Several revisions were made since and the 2004 edition is the latest. While the earlier
editions used per capita daily contributions for estimating sewer design flows that include I/I (400
gal/capita-d for laterals, 250 gal/capita-d for trunks and outfall sewers, and 3.5 times of average DWF for
interceptors), the more recent edition of the "Ten-State" standards (HES, 1997) uses less numerical and
more qualitative design criteria.
The Standards state that wastewater collection systems are to be sized based on an average of 100
gal/capita-d plus that from industrial plants and major institutional and commercial facilities. The 100
gal/capita-d figure is assumed to cover "normal" infiltration but an additional allowance should be made
3-9
-------
where conditions are unfavorable, such as when rain water from roofs, streets, and other areas, and
groundwater from foundation drains cannot be fully eliminated from the system. In determining the peak
wastewater flow that includes normal infiltration for design, the 100 gal/capita-d is multiplied by a
peaking factor (peak hourly/design average) computed from a formula that relates to the square root of
population in thousands developed by Harmon (1918) from flow data in Toledo, OH. For the new sewers
serving existing development, the "Ten-State" Standards, without indicating how, caution that the
likelihood of I/I contributions from existing service lines and non-wastewater connections to those service
lines shall be site-specifically evaluated and quantified. Many states and municipalities in the nation,
whether or not in the Great Lake-Mississippi River region, make a reference to the "Ten-State" Standards
in the approval of plans for new sewer systems or modifications to existing ones.
State and Local Design Requirements
Table 3-1 presents a summary of selected state and local design standards for sizing of new sanitary
sewers and rehabilitation of existing ones. The table is prepared from the Rules and Regulations of the
respective states and local communities cited and from the available references (Parsons Engineering
Science, 2000; Mauro, 2001)
Table 3-1 shows that states and local municipalities use two distinctly different criteria for granting
construction permits of new sanitary sewers and for reviewing control plans to mitigate SSOs in existing
sewer systems. For design of new collection sewers, the prevalent practice is to use the Ten States
Standards or similar, which use standard engineering peaking factor to account for I/I. Massachusetts
uses the criteria in TR-16 (NEIWPCC, 1998) that provides guides to develop per capita flow, reasonable
amounts of I/I, and appropriate peaking factors. Oregon uses similar methods but, instead of allowing a
minimum infiltration of 250-500 gal/in.dia-mi-d of sewer pipe as in Massachusetts, the infiltration amount
is limited to a maximum of 2,000 gal/acre-d in order to limit the treatment capacity required. New Jersey
provides an elaborate table showing average flow rates for various types of establishments and
measurement units. New sewers are to be sized to carry twice the average flow rate at half full. No
additional provisions of I/I are required in computing the average flow.
EPA had proposed a SSO Rule to mitigate SSO problems associated with existing sanitary sewer systems.
The CMOM program in the Rule emphasizes O&M and capacity assurance of all system components, and
is widely accepted by states and sanitary sewer communities. States prevailingly regulate SSO activities
through their respective but generally similar SSO policies, and the policies are enforced through NPDES
permits. The driving forces to eliminate or reduce SSOs are from environmental and health concerns.
The target is to achieve the desired performance of no overflow in the entire existing system that includes
collection sewers and treatment facilities. In other words, the focus is on the collection and treatment
system as a whole rather than the individual sewer lines. Hence different performance criteria are used.
The engineering approaches practiced by states and local municipalities are similar. The "rainfall
method," that uses "design" storms, is generally used to perform the hydraulic analysis of a system for
developing sewer rehabilitation and SSO mitigation plans. This reflects an understanding that the prime
culprit of SSO problems is the RDII associated with severe wet-weather events. However, the return
frequency of the design storms required by states for system analysis varies, as does the levels of risk
management. Massachusetts requires an I/I removal study if infiltration exceeds 4,000 gal/in.dia-mi-d or
I/I from 1-yr, 6-hr storm exceeds sewer carrying capacities in segments of the sewer system. To abate
potential bacteria pollution, Oregon prohibits SSO in winter time from a storm event equal to or less than
the 5-yr, 24-hr duration storm, and in summer months, from the storms equal to or less than the 10-yr, 24-
hr storms. Michigan adopts a more aggressive criterion that requires the sanitary sewer system to retain
the flow generated from a 25-yr, 24-hr storm event, or 3.9 in. of rain in a 24-hr period. One common
3-10
-------
requirement among the states is that all states require the collection of flow and rainfall monitoring data
and use of a hydraulic model for sewer system evaluation and identification of rehabilitation locations.
Table 3-1. Summary of State and Local Standards for Sizing of Sanitary Sewers
State/
Municipality
Design Flows/Peaking
Factors for New Sewers
(construction permits)
Wet-weather
Performance Criteria for
Existing System w/SSO
Remarks
Illinois
Use Ten States Standards
Design storm not specified
as SSO not allowed
- Use water pollution control
permit processes to require
compliance reporting and cost-
effectiveness to enforce case-by-
case compliance of no SSO
- Sewer ban enforced when 2 SSOs
occur in 5 years
Oregon
- Use 50-100 gal/capita-d for
domestic flows, peaking
factors 1.8 - 4.0, and allow a
maximum of 2,000 gal/acre-d
for infiltration
- Use "rainfall method" to
derive baseline flow rates
from STP flow data to
project future design flow
rates Detailed guidelines
provided.
- No overflow allowed
from:
Winter (Jan-May) - 5-yr,
24-hr event (effective 2010)
Summer(Jun-Dec) - 10-yr,
24-hr event
Systems currently experiencing
SSO due to I/I are usually subject to
enforcement action and must
prepare and submit plans to assure
compliance by 2010, earlier if
sensitive receiving streams are
involved
North Carolina
- Generally follow Ten States
Standards with the minimum
peak flow at 2.5 of the
average daily flow
- Infiltration/exfiltration rate
not to exceed 100
gal/in.dia-mi-d (200
gal/in.dia-mi-d in Ten States
Standards)
Design storm not specified
as SSO is not allowed
- To complete issuing first time
holistic collection system permits
by 2005
- Use operation/maintenance based
permit to enforce individual
compliance of system performance
Michigan
Use Ten States Standards
The State SSO policy
requires sewer correction
actions based on 25-yr, 24-
hr storm or 3.9 in. of
rainfall
Equivalent to less than one
overflow per 10 years period
Kentucky
Use Ten States Standards
- Design storm not
specified as SSO is not
permitted
- Trickle I/I study when
- SSO communities are required to
develop a SSO plan and schedule
(guidelines provided) to eliminate
overflows
3-11
-------
State/
Municipality
Design Flows/Peaking
Factors for New Sewers
(construction permits)
Wet-weather
Performance Criteria for
Existing System w/SSO
Remarks
annual ave. daily flow >
120 gal/capita-d or 24-hr
flow > 275 gal/capita-d at
STP
- if failed, enforce through
moratorium on sewer extension and
connection and consent decree
Massachusetts
Use TR-16 "Guides for the
Design of Waste water
Treatment Works"
Sewer design based on peak
hourly sewage rate plus I/I
based on 1-yr, 6-hr storm
Cost-effectiveness I/I removal study
is required if infiltration exceeds
4,000 gal/in.dia-mi-d or inflow from
1-yr, 6-hr exceeds carrying capacity
in segments of the sewer system
Minnesoda
Use Ten States Standards
- All bypasses are
regulated by permit and are
prohibited
- Trickle I/I study when
average daily flow > 120
gal/capita-d or peak flow >
275 gal/capita-d at STP
- Require to follow bypass
notification procedures and take
timely corrective action
- Failure will result in enforcement
actions including monetary
penalties
New Jersey
- Use the projected design
flow rates (I/I included)
specified for various types of
establishments and facilities
inN.J.A.C.7:14A-23.3
- Sewers to be sized to carry
twice the projected flow at
half full
- Design storm not
specified
- Use 80% permitted
treatment capacity as a
threshold for requiring the
development of I/I
reduction plan
Like other states, permitting tools
(sewer connection ban, penalties)
are used to enforce compliance of
capacity assurance and prevention
of system overloading causing
permit violation
Cincinnati,
OH
- Use 100 gal/capita-d to
compute the average sewage
flow
- Peaking factors decreasing
from 4.0 to 3.3 for population
increasing from 750 to 5,000
- Additional I/I allowance of
1000 gal/acre-d added to the
peak flow
- Design capacity is based
on flow monitoring and
model-projected peak flow
rate from a design storm
- Design flow comprised of
BWF, GWI, and RDII
- BWF is the greater of
peak diurnal flow projected
or monitored
- GWI is the greater of
observed or projected one-
year max. rate
- RDII based on SCS Type
II rainfall distribution
assuming full build out
conditions
Like many other SSO communities,
the sewer rehabilitation program
was developed in response to
directive from Ohio State DEP to
eliminate SSOs as dictated in the
NPDES permit that prohibits SSOs
3-12
-------
Chapter 4 RDII Prediction Methods
Introduction
A reliable estimate of RDII is critical in the development of a cost-effective SSO control plan. This
chapter provides a summary of selected literature articles and reports that review and evaluate RDII
prediction methods. The RDII prediction method that is included in this phase of CRADA effort will be
presented. It should be kept in mind that RDII is very much site-specific. No single method is likely to
be universally applicable. The goal is to select and implement one RDII prediction method that is most
likely to be accepted and used by practicing engineers. Other methods can be included later outside this
CRADA by independent software providers.
Unit Hydrograph
A unit hydrograph is defined as the direct runoff hydrograph resulting from unit depth of excess rainfall,
say, 1 in. or 1mm, produced by a storm of uniform intensity and specified duration over a watershed. It
was first proposed by Sherman for flood estimation and has since found wide-ranging application for the
estimation of actual floods where a hydrograph is required. Unit hydrographs are generally derived from
streamflow data and estimates of rainfall excess. The unit hydrograph is applied to the hyetograph of
rainfall excess to estimate the surface runoff hydrograph. A flood hydrograph is a combination of the
surface runoff hydrograph and baseflow.
Literature Review
WERF Report (1999)
A Water Environment Research Foundation (WERF) publication (Bennett et al., 1999) and a conference
paper (Schultz et al., 2001) provided an expanded review of RDII prediction methods in the literature
going back to 1984. Their literature search included online catalogs at the University of Wisconsin at
Milwaukee and Madison. Additional references were identified through contacts with engineering firms
and municipal agencies. A total of 42 documents was compiled and reviewed.
Eight broad categories of RDII quantification methods were identified and three cooperating municipal
agencies were involved in testing these methods against the monitored rainfall-flow data. The three
agencies are the Metropolitan Council Environmental Services, St. Paul, MN; Bureau of Environmental
Services, Portland, OR; and the Montgomery Water Works and Sanitary Sewer Board, Montgomery, AL.
The eight categories are:
1) The constant unit rate method
2) The percentage of rainfall volume (R-value) method
3) The percentage of stream flow method
4) The synthetic unit hydrograph (SUH) method
4-13
-------
5) The probabilistic method
6) The rainfall/sewer flow regression method
7) The synthetic stream flow regression method
8) Methods embedded in hydraulic software
The constant unit rate method calculates RDII as a fixed constant (e.g., gal/acre-in.rainfall) multiplied
by measurements of tributary sewershed characteristics (e.g., area, land use, population, pipe diameter,
pipe length, and pipe age). The percentage of rainfall volume (R-value) method calculates RDII
volume as a fixed percentage of the rainfall amount. The percentage of streamflow method is similar to
the previous rainfall method, but it uses streamflows in nearby watersheds as an independent variable.
This method recognizes that streamflows inherently account for the effects of antecedent moisture
conditions that consequently influence the groundwater levels. A relationship can be developed between
sewer flow and streamflow data.
The SUH method assumes that RDII in a sewer responding to rainfall is similar to stormwater runoff in a
watershed and calculates RDII hydrograph from a specified "unit" hydrograph shape that relates RDII to
unit precipitation volume and specified time duration, and sewershed characteristics. The simplest
synthetic unit hydrograph has a triangular shape and many formulations like that in EPA's Storm Water
Management Model (SWMM) (Huber and Dickenson. 1988; EPA 2006) use multiple unit hydrographs to
account for fast, medium and slow RDII responses. Probabilistic method calculates RDII of a given
recurrence interval from long-term sewer flow records using probability theory. The method establishes
the relationship of peak RDII flow to recurrence interval. Rainfall/sewer flow regression method
calculates peak RDII flows from rainfall data through a relationship between rainfall and RDII flows.
This regression, expressed as an equation, is derived from rainfall and flow monitoring data in sewers
using multiple linear regression methods and considering dry and wet antecedent conditions.
In the synthetic streamflow regression method, RDII is calculated from synthetic streamflow records
and sewershed characters using regression equations derived from multiple regression techniques to
correlate hydrologic responses to sewer flow responses. The synthetic streamflow records can be
generated by calibrated hydrologic simulation models. Finally, methods embedded in publicly or
commercially available hydraulic software uses one or more of the previous seven prediction methods
discussed above. The most notable one is probably the EPA's SWMM that programs the synthetic unit
hydrograph method into the codes.
The 1999 WERF study concluded that in practice any of these RDII prediction methods should be used
with the site-specific database of rain and flow observations during both wet and dry periods. However,
no one method was likely to be universally applicable due to a variety of site conditions and analysis
application needs. The study identified criteria (listed below) to test the alternative RDII prediction
methods using flow and rainfall data supplied by the three previously identified cooperating sewer
agencies. Specifically, the methods should be able to:
1) Predict peak flows for individual storms
2) Predict volume for individual storms
3) Predict the hydrograph timing, shape, recession limb
4) Predict peak flows for multiple storms
5) Predict volume for multiple storms
6) Operate on commonly available data
The 1999 WERF study also concluded that the SUH and rainfall/sewer flow regression methods were the
4-14
-------
most accurate at predicting peak flows and event volumes both for single storm event and multiple storm
events. While the probabilistic method can be accurate for predicting peak daily RDII flows, it will
require long-term data.
The constant unit rate method is simple to apply and can provide a good prediction of the event volume of
a single storm. But it is difficult to develop reasonable estimates of unit rate constants, which may vary
temporally, for all storm events. Percent of streamflow methods are only good for rare cases where
stream gauge data are available in watersheds with similar basin characteristics as the sewersheds being
analyzed. Besides, a sewershed is usually smaller than a watershed. The synthetic streamflow regression
method was successfully applied in Milwaukee for sewerage system improvement planning. The
prerequisite for applying this method is the availability of a calibrated watershed runoff model which,
more often than not, does not exist for the case at hand.
The WERF study emphasized that, in an actual application, the objective, intent, and purpose of the
studies as well as the availability of data, time, and staff should be considered in selecting the most
appropriate flow estimation method.
Review by Crawford, et al (1999)
Crawford et al. (1999) focused their review on three RDII prediction methods: constant unit rate,
rainfall/flow regression, and percent of rainfall volume (R-value). The merits and limitations of these
methods were evaluated from applications to two collections of the City of Salem, Oregon (with a
population of about 160,000) and the City and County of Honolulu, Hawaii serving a population of about
1,000,000. They concluded from the data that peak hourly RDII rates per acre from a 5-yr storm increase
significantly as the average age of sewer pipes increases from 10 to 30 yrs. Hence, use of a single rate for
new sewer areas is easily proved not to be realistic. To overcome this limitation, they suggest that unit
RDII rates should appropriately increase with the age of the sewer system.
The regression method provided a means of determining the shape and magnitude of a RDII hydrograph.
Crawford et al. (1999) were able to use the regression equations derived from data under winter
conditions to get a good match between the monitored and simulated hydrographs from other winter
storms. However, when they applied the regression equations to summer and early fall storms, large
discrepancies between observed and simulated flows were noted. To overcome this limitation, they
suggest that a separate series of regressions should be developed to represent the seasonal nature of the
rainfall-I/I processes. Hence, adequate and representative rainfall and flow data are pre-requisite for a
successful application of regression method.
In applying the "R-Value" method to estimating RDII rates for greater intensity and less frequent storms
than that used to derive the "R" value, Crawford et al. (1999) cautioned that the "R" values should be
appropriately tapered to account for the upper limit of peak flows that the leaky sewers can take in. This
is to recognize that there is a limit to the ability of the flow connections, leaky manholes, and damaged
laterals to take in water.
Review by Wright et al. (2001)
Wright et al. (2001) reviewed literature of RDII estimation techniques that appeared since 1993. The
methods are grossly classified into three groups: volume-based "Rational" method (or R-value method),
unit hydrograph method, and physical processes modeling method. The volume-based method does not
provide temporal information that is needed for sewer conveyance capacity assessment. Like the volume-
based method, the unit hydrograph methods are empirical methods based on observations of rainfall and
flow. The paper discusses various unit hydrograph methods, including: the SUH where the shape of the
4-15
-------
hydrograph is pre-defined; the data-derived unit hydrograph using multiple regressions to directly derive
the ordinates of a unit hydrograph from measured rainfall and RDII flow data; and, the conceptually-
derived unit hydrograph using a system of cascading linear reservoirs.
The so-called "RTK method", included as an option in SWMM Runoff Block (Huber and Dickenson,
1988), is probably the most popular SUH method. This method uses three triangular hydrographs to
estimate the wide range of response times associated with the effect of fast inflow and slower GWI. The
R parameter is the fraction of rainfall volume entering the sewer system as RDII, T is the time to peak,
and K is the ratio of time of recession to T. Since the three unit hydrographs distinctively represent the
quantitative contribution of inflow and infiltration to the overall RDII hydrograph, the paper stated that
the "RTK method" can be used to estimate RDII reduction from selected rehabilitation methods by
applying a reduction factor to the RDII and GWI hydrographs.
Data-derived unit hydrograph (UH) derives the ordinates of a unit hydrograph directly from measured
rainfall and RDII flow data using multiple linear regression or linear programming techniques and is not
based on calibration methods like the "RTK method." Instead of beginning with an assumed shape
characteristic as the "RTK method," the data-derived UH is a linear transform function completely
derived from measured data. The goal is to find a vector of unit hydrograph ordinates that minimize the
difference between the time series of measured flow and the estimated flows. The paper summarized two
approaches in determining RDII responses. One approach is a regression on the time series of measured
rainfall and flow and is often labeled in the literature as "the Regression Method." The other approach
estimates unit hydrograph ordinates using optimization techniques such as unconstrained or constrained
least squares regression or a liner programming technique. Wright et al. (2001) proved that the "the
Regression Method" is exactly equivalent to the traditional unit hydrograph analysis.
Unit hydrograph methods may also be derived using a system of cascading linear reservoirs where a unit
pulse of precipitation is routed through reservoirs characterized by a linear storage-discharge relationship.
The cascading reservoir approach provided an important conceptual link between purely empirical
methods and more physically based conceptual models, like SWMM which uses the non-linear reservoir
approach and the continuity and momentum principles. As with the data-derived UH method, the
reservoir parameters are found using some optimization techniques such as linear least squares regression
or linear programming (LP) technique. The reservoir parameters may be constrained to derive physically
realistic values when LP is used. But physically unrealistic values (i.e., negative UH ordinates) may be
derived when an unconstrained ordinary regression method is used.
The physically based model advocated by Wright et al. (2001) is the SWMM Runoff Block with
modifications by Kadota and Djebbar (1998). The primary modification included the use of water
elevation in a sewer pipe instead of pipe invert elevation to better represent the driving head of
groundwater for computing the rate of infiltration entering. They also promoted the use of effective RDII
contributing area as a parameter to quantify RDII reductions for pipeline and manhole rehabilitation.
RDII rates and volumes are conceptualized as coming from an effective area. With these modifications,
SWMM, a physically based model, will have physical parameters that can be used to reflect the reduction
of RDII associated with the extent of sewer rehabilitation work completed.
Recommended RDII Prediction Method for SSOAP
From the above literature review, it can be concluded that no single RDII prediction method is universally
applicable due to a wide variety of site-specific database of rain and flow observations during both wet
and dry periods. All methods require monitored data to evaluate and validate their predictive capabilities
but the amount of data required varies. If flow data are not available, but sewershed data (e.g., area, land
4-16
-------
use, population, and sewer pipe length) are, the constant unit rate method (e.g., gal/acre) may apply for
estimating peak I/I rates for sewer design. On the other hand, the probabilistic method will require long-
term flow data to perform meaningful frequency analysis for deriving the relationship of peak RDII flow
rates or event volumes to recurrence interval. For other methods, flow data of a few months duration are
usually sufficient. But when daily flow data are used, several years of data are needed to obtain a good
RDII prediction (Bennett et al., 1999).
A flow monitoring program in a sanitary sewer system can never monitor flow under future sewershed
and sewer conditions. Sewer routing models are needed to expand the limited monitored data to existing
and future build-out conditions where no data are available. These models rely on the RDII prediction
methods in conjunction with appropriate DWF and GWI projections to develop representative inflow
hydrographs at various entries of a sewer system. Hence, the selected RDII prediction method must be
amendable for estimating current sewer flows in all parts of a sewer system and projecting how sewer
flows will change in response to sewer system expansion and aging, and RDII control measures. As
stated earlier, the 1999 WERF study concluded that SUH and rainfall/flow regression methods were
preferred for predicting flows for single as well as multiple storm events. Good multiple-storm peak and
volume prediction is important in extrapolating data beyond the calibration events for a prolonged period
simulation to evaluate the effect of RDII on storage and treatment.
Both SUH and rainfall/flow regression methods are empirical methods with parameters calibrated by
observed rainfall and sewer flow data. Both have been widely applied and are successful in the RDII
source identification and quantification (peak, volume, and time series) for the development of sewer
system/treatment improvement plans. The rainfall/flow regression methods will be attractive if there are
at least two years of extensive (both temporal and spatial) rainfall and flow data to develop several sets of
equations to reflect seasonal influences for dry and wet antecedent and groundwater conditions. Since
regression equations only relate the RDII rate to the preceding rainfall amounts corresponding to various
time periods (e.g., 1 hr, 2-3 hrs, 4-6 hrs, 7-12 hrs, 12-24 hrs, 1-2 d, 4-7 d, and 7-15 d) through a series of
coefficients, antecedent moisture and groundwater elevations are implicitly embedded in the coefficients
determined by the regression analysis. It is difficult to quantify the individual contributing flow
components and identify if the RDII problems are caused by inflow, infiltration or both. A quantitative
knowledge of flow contributing sources would help identify SSO control options.
On the other hand, the RTK method, one kind of the SUH method, uses three triangular unit hydrographs
to represent the various ways that precipitation contributes to RDII. The RDII volumes of three unit
hydrographs are designated as RI, R2, and R3. A high RI value indicates that the RDII is primarily inflow
driven. If more of the total R-value is allocated to R2 and R3, this indicates that the RDII is primarily
infiltration driven. This knowledge is useful during a sewer system evaluation survey (SSES) to
determine the best SSES approach to use in an area and whether a point repair or a comprehensive
rehabilitation approach is more suitable.
The UH approach used in the RTK method is a common method for generating a hydrograph from a
rainfall record based on linear response theory. One benefit of using a UH technique to determine rainfall
responses in a sewer system is that the technique can be applied to analyze RDII flow from storms that
have complex patterns of rainfall intensities and durations. The RTK method has been included as an
option in SWMM4 and SWMM5 and has been widely used and proven to be a valuable method in
separate sanitary sewer system analysis associated with storm events.
The RTK method was developed by COM staff members (Giguere and Riek, 1983; COM et al., 1985;
Miles, et al., 1996; Vallabhaneni, et al., 2002). It has been used on sewer system master planning projects
throughout the country by COM engineers and client staff members since the mid-1980s. This CRADA
4-17
-------
is intended for CDM to share their rich and long experience in applying the RTK method for SSO
analysis and control planning. Hence, the SSOAP Toolbox will initially incorporate the RTK method
only. Other RDII prediction methods may be included in future effort to expand SSOAP by EPA or
private sectors.
4-18
-------
References
Arbour, R. and K. Kerri. (1998). Collection Systems: Methods for Evaluating and Improving
Performance, California State University - Sacramento, USEPA Grant No. CX924908-01-0.
American Society of Civil Engineers (ASCE). (1962). Design and Construction of Sanitary and Storm
Sewers, Manuals and Reports on Engineering Practice - No. 37.
American Society of Civil Engineers (ASCE). (1982). Gravity Sanitary Sewer Design and Construction,
Manuals and Reports on Engineering Practice - No. 60.
American Society of Civil Engineers. (1994). Federal Infrastructure Research & Development
Technology Transfer: Meeting State and Local Public Works Needs, Civil Engineering Research
Foundation.
Babbitt, H.E. (1947). Sewerage and Sewage Treatment, 6th Edition, John Wiley & Sons, Inc.
Bennet, D., Rowe, R., Strum, M., Wood, D. (1999). Using Flow Prediction Technologies to Control
Sanitary Sewer Overflows, Water Environment Research Foundation (WERF), Project 97-CTS-8.
Black & Veatch. (1999). Optimization of Collection System Maintenance Frequencies and System
Performance, ASCE, EPA Cooperative Agreement #CX 824902-01-0.
Camp Dresser & McKee Inc. (COM), F.E. Jordan Associates, Inc., James M. Montgomery, Consulting
Engineers. (1985). East Bay Infiltration/Inflow Study Manual for Cost-Effectiveness Analysis. East Bay
Municipal Utility District (EBMUD), Oakland, CA.
Cesareo, D. J., and Field, R. (1975). "Infiltration-Inflow Analysis." Journal of the Environmental
Engineering Division, ASCE, Vol. 101, No. EE5, Proc. Paper 11645, 775-785.
Crawford, D., Eckley, P., and Pier, E. (1999). "Methods for Estimating Inflow and Infiltration into
Sanitary Sewers." In James, W. (ed.). New Applications in Modeling Urban Water Systems, Monograph
7, CHI, Guelph, Ontario, Canada.
Fan, C., Field, R, and Lai, F-h. (2003). "Sewer-Sediment Control: Overview of an EPA Wet-Weather
Flow Research Program." Published in a special issue "Hydraulics of Sediment Movement in Urban
Drainage Systems" of the ASCE Journal of Hydraulic Engineering. 129(4): 253-259, May.
Fan, C.-Y., Field, R., Lai, F.h., and Sullivan, D. (2000). "Sewer-Sediment Control: Overview of an EPA
Wet-Weather Flow (WWF) Research Program," Proceedings of the 2000 Joint Conference on Water
Resources Engineering and Water Resources Planning & Management, Minneapolis, MN, August.
Fan, C.-Y., Field, R., Pisano, W. C., Barsanti, J., Joyce, J.J., and Sorenson, H. (2001). "Sewer and Tank
4-19
-------
Flushing for Corrosion and Pollution Control" Published in a special issue "Mini-Symposium on Urban
Drainage" of^hsASCE Journal of Water Resources Planning and Management, 127(3): 194-201.
Field , R., and E.J. Struzeski, Jr. (1972). "Management and Control of Combined Sewer Overflows,"
Journal of the Water Pollution Control Federation, Vol. 44(7), p. 1393-1415, July.
Giguere, P.R. and Riek, G.C. (1983). "Infiltration/Inflow Modeling the East Bay (Oakland-Berkeley
Area) I/I Study." Proceedings of the 1983 International Symposium on Urban Hydrology, Hydraulics and
Sediment Control., University of Kentucky, Lexington, KY, July 25-28.
Golden, J.B. (1996). "An Introduction to Sanitary Sewer Overflows." National Conference in Sanitary
Sewer Overflows, Seminar Publication, EPA/625/R-96/007, EPA, Washington, D.C.
Harmon, W. G. (1918). "Forecasting Sewage at Toledo under Dry-Weather Conditions." Engineering
New Records, 80, 1233.
Health Education Service (HES). (1997). Recommended Standards for Wastewater Facilities. A Report
of Committee of the Great Lakes - Upper Mississippi River, Board of State and Provincial Public Health
and Environmental Managers, Albany, NY 12224. www.hes.org.
Huber, W.C., and Dickinson, R.E. (1988). Storm Water Management Model Version 4, User's Manual,
EPA/600/3 88/OOla (NTIS PB88 236641/AS), EPA, Athens, GA.
Jenq, K., Bagstad, M.J., and Chang, J. (1996). "New Collection System Modeling Techniques Used in
Houston." National Conference in Sanitary Sewer Overflows, Seminar Publication, EPA/625/R-96/007,
EPA, Washington, D.C.
Lai, F.-h., Field, R, Fan, C.-Y., and Sullivan, D. (2000). Collection System Modeling for
Planning/Design of Sanitary Sewer Overflow (SSO) Control, Proceedings of the 2000 Joint Conference
on Water Resources Engineering and Water Resources Planning & Management, Minneapolis, MN,
August.
Lai, F.-h., Claytor, R, and Field, R. (2001). "Planning for SSO Control: Henrico County, VA- Case
Study." Proceedings of the World Water and Environmental Resources Congress, May 20-24, Orlando,
FL.
Lai, F.-h. and Field, R. (2002). "USEPA Research on Infiltration/Inflow Control. "Proceedings of the
WEF2001 Collection System Odyssey, Bellevue, WA, July 8-11.
Mauro, J. T. (2001). "Review of Design Standards and Criteria for Sizing Sanitary Sewers." Proceedings
of 2001 A Collection Systems Odyssey: Combining Wet Weather Solutions, July 8-11, Bellevue, WA.
Metcalf and Eddy. (1928). American Sewerage Practice. Vol.1 Design of Sewers, 2nd Edition, MaCraw-
Hill Book Company.
Miles, S.W., Dorn, J.L., and Tarker, RE. (1996). "An I/I Analysis and Prediction Method to Help Guide
Separate Sanitary Sewer Improvement Programs." In Urban Wet weather Pollution: Controlling Sewer
Overflows and Stormwater Runoff. Water Environment Federation, Alexandria, VA.
National Council on Public Works Improvement (NCPWI). (1988). Fragile Foundations: A Report on
4-20
-------
America's Public Works, Final Report to the President and Congress. Washington, B.C.
New England Interstate Water Pollution Control Commission (NEIWPCC). (1998). Guides for the
Design of Wastewater Treatment Works, Technical Report #16
Parsons Engineering Science, Inc. (2000). Review of Design Standards/Criteria for Sizing Sanitary
Sewers: New Installation and Rehabilitation Efforts. A draft report prepared for USEPA Office of Water,
personal communication with Joseph Mauro.
Rawn, A.M. (1936). "What Cost Leaking Manholes." Waterworks and Sewerage, December.
Roesner, L.A., Aldrich, J.A., and Dickinson, R.E. (1988). Storm Water Management Model User's
Manual, Version 4: Extern Addendum, EPA/600/3-88/00Ib (NTIS PB 88-236658/AS), EPA, Athens,
GA.
Schladweiler, J. C. (2001). "Tracking Down the Roots of Our Sanitary Sewers." Proceedings of 2001 A
Collection Systems Odyssey: Integrating O&M and Wet Weather Solutions, Water Environment
Federation, Bellevue, WA, July 8-11.
Schultz, N., Wood, D., Adderly, V., and Bennet, D. (2001). "RDII Quantification Research Results."
Proceedings of the World Water and Environmental Resources Congress, May 20-24, Orlando, FL.
U.S. Environmental Protection Agency (EPA). (1991). Handbook, Sewer System Infrastructure Analysis
and Rehabilitation, Report No. EPA/625/6-91/030, Office of Research and Development, Center for
Environmental Research Information, Cincinnati, OH.
U.S. Environmental Protection Agency (EPA). (1996a). Sanitary Sewer Overflow Cost/Benefit Analysis,
prepared for EPA by M&E/ERG/CDM.
U.S. Environmental Protection Agency (EPA). (1996b). National Conference on Sanitary Sewer
Overflows (SSOs), Seminar Publication, Report No. EPA/625/R-96/007, EPA, Washington, D.C.
U.S. Environmental Protection Agency (EPA). (2001). Preamble of the Draft Proposed SSO Rule.
U.S. Environmental Protection Agency (EPA). (2006). Stormwater Management Model Redevelopment
Project, http://www.epa.gov/ednnrmrl/models/swmm/index.htm.
Urban Institute. (1981). Capital Infrastructure Investment Needs and Financing Options, HUD-0003404,
Department of Housing and Urban Development, Washington, D.C.
Vallabhaneni, S., Koran, J.M., Moisio, S.E., and Moore, C.I. (2002). "SSO Evaluations: Infiltration and
Inflow using SWMM RUNOFF and EXTRAN." In James, W. (ed.). Best Modeling Practices for Urban
Water Systems, Monograph 10, CHI, Guelph, Ontario, Canada.
Vallabhaneni, S., Chan, C., Burgess, E.H. (2007). Computer Tools for Sanitary Sewer System Capacity
Analysis and Planning. Report No. EPA/600/R-07/111. U.S. Environmental Protection Agency,
Washington, DC. http://www.epa.gov/ORD/NRMRL/pubs/600r0711 l/600r07111 .pdf
Water Environment Federation (WEF). (1999). Control of Infiltration and Inflow in Private Building
Sewer Connections, Monograph.
4-21
-------
Wright, L., Dent, S., Mosley, C., and Kadota, P. (2001). "Comparison of Rainfall Dependent Inflow and
Infiltration Simulation Methods." In James, W. (ed.) Applied Modeling of Urban Water Systems,
Monograph 9, CHI, Guelph, Ontario, Canada.
4-22
------- |