v/EPA
United States
Environmental Protection
Agency
Office of Research
and Development
Washington, DC 20460
EPA/600/R-07/111
October 2007
Computer Tools for
Sanitary Sewer System
Capacity Analysis and
Planning
-------
EPA/600/R-07/111
October 2007
Computer Tools for Sanitary Sewer System
Capacity Analysis and Planning
by
Srinivas Vallabhaneni, P.E., BCEE
Carl C. Chan, P.E.
Camp Dresser & McKee Inc.
Indianapolis, IN 46204
and
Edward H. Burgess, P.E.
Camp Dresser & McKee Inc.
Cincinnati, OH 45249
In support of:
Cooperative Research and Development Agreement
CRADA 216-02
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.
-------
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 groundwater; 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
in
-------
Abstract
A properly designed, operated and maintained sanitary sewer system is meant to collect and convey all of the sewage
that flows into it to a wastewater treatment plant. However, occasional unintentional discharges of raw sewage from
municipal sanitary sewers - called sanitary sewer overflows (SSOs) - occur in many systems.
Rainfall-derived infiltration and inflow (RDII) into sanitary sewer systems has long been recognized as a major
source of operating problems, causing poor performance of many sewer systems. RDII is the main cause of SSOs to
customer basements, streets, or nearby streams and can also cause serious operating problems at wastewater treatment
facilities. There is a need to develop proven methodologies and computer tools to assist communities in developing
SSO control plans that are in line with their projected annual capital budgets and provide flexibility in future
improvements.
To accomplish this goal, EPA entered into a cooperative research and development agreement (CRADA) with Camp
Dresser & McKee, Inc. (COM) to develop public-domain software tools to support SSO control planning. These
tools, named the Sanitary Sewer Overflow Analysis and Planning (SSOAP) Toolbox, are accompanied by this
technical document that describes how to use the Toolbox in analyzing infiltration/inflow, performing capacity
analyses of sanitary sewer systems, and developing SSO control plans.
It is the intent of this report to provide the Toolbox users with technical information needed for its effective use for
analysis and mitigation of SSO-related problems. The technical report is not intended to serve as the user manual for
the SSOAP Toolbox, which is a separate document included with the software package. Instead, it provides an
introductory hydrologic approach and identifies a RDII methodology for initial incorporation into the SSOAP
Toolbox; an overview of the required sewer system hydraulic analysis; and data collection requirements to support
SSO planning and analysis using the SSOAP Toolbox. In addition, the report describes the tools and their functions
for performing a sanitary sewer system capacity assessment.
The report also includes a description of the application of EPA's Storm Water Management Model Version 5
(SWMM5) application within the SSOAP Toolbox for assessing the baseline hydraulic conditions of the system and
quantifying capacity improvements of various identified improvement scenarios. Guidance is provided for
establishing system improvement objectives, screening potential options for improvements, developing improvement
scenarios, and using SWMM5 model output to evaluate alternatives. Finally, the report provides a case study that
demonstrates how the RDII methodology described in this technical report has been effectively used in SSO planning
and analysis.
IV
-------
Table of Contents
Computer Tools for Sanitary Sewer System Capacity Analysis and Planning i
Notice ii
Foreword iii
Abstract iv
Table of Contents v
List of Tables viii
List of Figures ix
List of Figures ix
Abbreviations and Acronyms x
Acknowledgements xi
Chapter 1 Introduction 1-1
1.1 Background and History 1-1
1.2 Regulatory Framework 1-1
1.3 Technical Report Organization 1-2
Chapter 2 Hydrologic Approach 2-4
2.1 Overview of Sanitary Sewer System Hydrology 2-4
2.2 RDII Prediction Methodologies 2-6
2.2.1 Literature Review Studies 2-7
2.2.2 Overview of RDII Prediction Methodologies 2-8
2.2.3 Recommendation of RDII Method for SSOAP Toolbox 2-12
2.3 RDII Unit Hydrograph Method 2-14
Chapters Hydraulic Analysis 3-17
3.1 Overview of Sanitary Sewer System Hydraulics 3-17
3.2 Analysis Methods: Static, Kinematic, and Dynamic 3-17
3.3 Pipe Flow Resistance 3-18
3.4 Surcharged Pipe Flow 3-19
3.5 Hydraulic Analysis of Special Structures and Appurtenances in Sanitary Sewers 3-19
Chapter 4 Data Collection 4-20
4.1 Introduction 4-20
4.2 Data Requirement for SSO Planning and Analysis 4-20
4.3 Sewer System Data 4-20
4.3.1 Physical Sewer System Data 4-21
4.3.2 Spatial Sewer System Data 4-21
4.4 Sewer Flow and Rainfall Data 4-22
4.4.1 Flow Monitoring Overview 4-22
4.4.2 Flow Monitoring Implementation 4-24
4.4.3 Equipment Selection 4-25
4.4.4 Equipment Installation and Maintenance 4-25
4.4.5 Data Collection and Quality Control Using SSOAP 4-26
4.5 Rainfall Data 4-29
Chapters Sanitary Sewer Overflow Analysis and Planning Toolbox 5-30
5.1 Introduction 5-30
5.2 Database Management Tool 5-30
5.2.1 Interfacing with External Data Sources 5-31
Flow monitoring data 5-31
Rainfall data 5-32
Hydraulic analysis data 5-32
5.2.2 Interacting with Other Tools in the SSOAP Toolbox 5-33
-------
5.2.3 DMT Utilities 5-33
Data quality control utility 5-33
Rainfall data analysis utility 5-33
Scenario management utility 5-34
5.3 RDII Analysis Tool 5-34
5.3.1DWF Analysis 5-35
5.3.1.1 BWF and DWF Adjustment 5-35
5.3.2 WWF Analysis 5-35
5.3.3 Hydrograph Decomposition and Unit Hydrograph Curve Fitting Analysis 5-36
5.3.3.1 Statistical Analysis of RDII Parameters 5-39
5.3.3.2 MedianR-Value Method 5-39
5.3.3.3 Average R-Values Method 5-40
5.3.3.4 Linear Regression Method 5-40
5.4 RDII Hydrograph Generation Tool 5-41
5.5 SSOAP-SWMM5 Interface Tool 5-42
5.5.1 Pre-processing RDII Hydrographs 5-43
5.5.2 SWMM5 Simulation 5-43
5.5.3 Post-processing Model Results 5-43
Chapter 6 Sewer System Model Development and Capacity Assessment 6-44
6.1 Introduction 6-44
6.2 Sewer Model Development 6-45
6.2.1 Model Input Development 6-46
6.2.1.1 Determine Model Network Extent 6-47
6.2.1.2 Collect System Configuration and Attribute Data 6-49
6.2.1.3 Develop Sanitary Sewer Model Network 6-49
6.2.1.4 Develop Sewershed Delineations 6-51
6.2.1.5 Develop DWF Components 6-53
6.2.1.6 Develop RDII Characteristics 6-54
6.2.2 Model Calibration and Verification 6-54
6.2.2.1 DWF Calibration 6-55
6.2.2.2 WWF Calibration and Verification 6-56
6.3 Capacity Assessment 6-58
6.3.1 Capacity Assessment Steps 6-58
6.3.1.1 Capacity Assessment Goals 6-58
6.3.1.2 Baseline Hydraulic Performance Assessment 6-58
Chapter 7 Development and Analysis of System Improvement Alternatives 7-63
7.1 Establishing Planning Objectives and Improvements Criteria 7-63
7.1.1 System Improvement Planning Objectives 7-63
7.1.2 System Planning Criteria 7-64
7.2 Options for Improving Collection System Performance 7-65
7.2.1 Sewer System Rehabilitation 7-65
7.2.2 Storage 7-66
7.2.2.1 On-Line Flow Equalization Storage 7-67
7.2.2.2 Off-Line Flow Equalization Storage 7-67
7.2.3 Conveyance 7-67
7.2.3.1 Trunk Sewer System Improvements 7-67
7.2.3.2 Pump Station Improvements 7-68
7.2.4 Treatment 7-69
7.2.5 Real-Time Control (RTC) 7-69
7.2.5.1 Overview of In-System Storage 7-69
7.2.5.2 Overview of Dynamic Flow Diversion 7-69
7.2.5.3 Overview of Enhanced Control Logic 7-69
7.3 Strategies to Develop Improvement Alternatives 7-70
7.3.1 The Conveyance Improvement Alternative 7-70
7.3.2 The Storage Improvement Alternative 7-71
vi
-------
7.3.3 The I/I Reduction Improvement Alternative 7-72
7.3.4 The No RDII Reduction Improvement Alternative 7-73
7.3.5 The No Conveyance Improvement Alternative 7-73
7.3.6 The No Storage Improvement Alternative 7-74
7.3.7 Additional Improvement Alternatives 7-74
7.4 Applying the SSOAP Toolbox to Evaluate Improvement Alternatives 7-74
7.5 Developing a Wet-Weather Management Plan 7-75
Chapters A Case Study - Ann Arbor, Michigan 8-78
8.1 Introduction 8-78
8.2 Hydrology and Hydraulics 8-79
8.3 Data Collection 8-80
8.4 Development of System Response Parameters 8-80
8.5 Capacity Assessment 8-82
8.6 SSO Control Program 8-82
8.7 Public Outreach 8-87
8.8 Summary 8-88
Chapter 9 References 9-89
VII
-------
List of Tables
Table 4-1. Sensitivity of sewer service area on R-value estimates 4-22
Table 5-1. Example of rainfall data analysis results 5-34
Table 5-2. Ranges of values for unit hydrograph parameters 5-38
Table 5-3. Determination of median R-value/distribution of Rvalues using "median R-values method." 5-40
Table 6-1. Example DWF capacity assessment results under different conditions 6-59
Table 6-2. Sewer surcharge and manhole flooding summary 6-62
Table 8-1. RDIITK parameterization 8-81
Vlll
-------
List of Figures
Figure 2-1. Three components of wet-weather wastewater flow 2-5
Figure 2-2. Pathways of infiltration and inflow into sanitary sewer systems 2-6
Figure 2-3. Example of a triangular unit hydrograph 2-14
Figure 2-4. Summation of three unit hydrographs 2-15
Figure 2-5. Summation of synthetic hydrographs 2-16
Figure 4-1. Example flow monitoring data review and analysis 4-28
Figure 5-1. Overview of tools within the SSOAP Toolbox 5-31
Figure 5-2. External data sources 5-32
Figure 5-3. Scenario management 5-34
Figure 5-4. DWF hydrograph derived from RDII analysis tool 5-35
Figure 5-5. Hydrograph decomposition in the RDII Analysis Tool 5-37
Figure 5-6. Unit hydrographs curve fitting using the RDII Analysis Tool 5-38
Figure 5-7. Example of a linear regression analysis 5-41
Figure 5-8. User interface of the RDII Analysis Tool 5-42
Figure 6-1. Capacity assessment steps for atypical sanitary sewer system 6-45
Figure 6-2. Model development, calibration, and verification 6-46
Figure 6-3. Model development plan view connectivity data check 6-50
Figure 6-4. Model development sewer profile check 6-50
Figure 6-5. Example of service area delineation 6-52
Figure 6-6. SWMM5 thematic map example 6-61
Figure 7-1. Triangular universe of possible wet-weather improvement solutions 7-70
Figure 7-2. Example results showing benefits of storage in reducing overflow frequencies 7-72
Figure 7-3. Required storage volume with and without sewer rehabilitation 7-74
Figure 7-4. Alternative evaluation to involve stakeholders 7-76
Figure 8-1. Ann Arbor sanitary collection system 8-79
Figure 8-2. Seasonal breakpoint 8-81
Figure 8-3. Seasonal responses to rainfall relationships 8-82
Figure 8-4. System capacity limitations coincide with basement flooding incidents 8-83
Figure 8-5. Curb drains convey disconnected footing drain flow to the storm water system 8-84
Figure 8-6. Testing pumping rate to support monitoring of a disconnected footing drain 8-84
Figure 8-7. Footing drain monitoring provides evidence of flows being removed 8-85
Figure 8-8. Footing drain disconnection progress 8-86
Figure 8-9. Neighborhood meeting with property owners 8-88
IX
-------
Abbreviations and Acronyms
AGO
BWF
CMOM
CSO
DMT
DWF
FAC
FDD
GIS
GWI
HGL
I/I
MDEQ
MG
MOM
NPDES
NWS
OF
O&M
PSO
RDI
RDII
RTC
R,T,K
R-value
SCS
SSD
SSES
SSOAP
SSO
SUH
SWMM
SWMM5
UH
WERF
WWF
WWTP
Administrative Consent Order
Base Wastewater Flow
Capacity, Management, Operation and Maintenance
Combined Sewer Overflow
Database Management Tool
Dry-Weather Flow
Federal Advisory Committee
Footing Drain Disconnection
Geographical Information System
Groundwater Infiltration
Hydraulic Grade Line
Infiltration and Inflow
Michigan Department of Environmental Quality
Million Gallons
Management, Operation and Maintenance
National Pollutant Discharge Elimination System
National Weather Service
Overflow
Operations and Maintenance
Pump Station Overflow
Rainfall-derived Infiltration
Rainfall-derived Infiltration and Inflow
Real Time Control
The R, T, and K parameters in the RTK method for RDII prediction
Fraction of rainfall volume entering the sewer system as RDII
U.S. Soil Conservation Service
SSOAP System Database
Sewer System Evaluation Survey
Sanitary Sewer Overflow Analysis and Planning
Sanitary Sewer Overflow
Synthetic Unit Hydrograph
Storm Water Management Model
Storm Water Management Model Version 5
Unit Hydrograph
Water Environment Research Foundation
Wet-Weather Flow
Wastewater Treatment Plant
-------
Acknowledgements
The authors from Camp Dresser & McKee Inc. (CDM) acknowledge the assistance of EPA Project Officer - Dr.
Fu-hsiung (Dennis) Lai and Project Advisor - Mr. Rich Field in successfully executing the Cooperative Research
and Development Agreement (CRADA) 216-02. Special thanks are extended to Dr. Lai for his technical
participation and valuable suggestions during the development of the SSOAP Toolbox and his detailed review of
this report. A significant part of the write-up in Chapter 2 on a literature review of RDII prediction methods was
derived from his draft report in support of this CRADA.
Several technical experts within CDM contributed to the preparation of this technical report and the SSOAP Toolbox.
CDM's Project Manager/Principal Investigator - Mr. Srini Vallabhaneni has provided oversight and guidance to the
team of CDM technical experts and the computer programmers. Mr. Carl Chan provided lead programming support
to the tool development. Mr. Ted Burgess provided quality assurance and project advice. In addition, the following
CDM technical experts supported the development of the report and SSOAP Toolbox: Mr. Phil Brink, Mr. Terry
Meeneghan, Mr. Rod Moeller, Mr. Wayne Miles, Mr. Chuck Moore, Ms. Barbara Moranta, Mr. Ben Sherman, and
Mr. Derek Wride. It has been a tremendous effort by each and every one of these individuals in preparing the project
work products.
The project team would like to thank the City of Ann Arbor, Michigan, for allowing us to showcase its experiences in
SSO analysis and planning. The City's experience demonstrates how the hydrologic and hydraulic methodologies
described in this technical report have been effectively used in SSO analysis and planning.
The report content and Toolbox functionality was improved by the review comments of an external technical panel.
The project team would like to acknowledge the contribution of five members: Dr. James P. Heaney, University of
Florida in Gainesville; Dr. Ken Kerri, California State University at Sacramento; Richard E. Nelson, CH2M HILL;
Nancy U. Schultz, CH2M HILL; and Dr. Larry Roesner, Colorado State University.
XI
-------
Chapter 1 Introduction
This technical report accompanies the Sanitary Sewer Overflow Analysis and Planning (SSOAP) Toolbox, a software
package developed by Camp Dresser & McKee Inc. (CDM) under a cooperative research and development agreement
between between the National Risk Management Research Laboratory of the U.S. Environmental Protection Agency
(EPA) and CDM. The SSOAP Toolbox is a suite of computer software tools used to predict rainfall-derived
infiltration and inflow (RDII) in sanitary sewer systems and to facilitate capacity analysis of these systems, using EPA
Storm Water Management Model Version 5 (SWMM5) (EPA, 2007).
RDII causes operational problems in sanitary sewer systems across the United States. Although sanitary sewer
systems are generally designed to accommodate RDII flows during wet weather, these flows often exceed the design
allowances. When this occurs, operational problems such as basement floodings, manhole overflows, bypasses to
storm sewers and direct discharges to receiving waters from sanitary sewer overflows (SSOs) often result. EPA
regards these problems, especially SSOs, as a high priority for corrective action by system owners and operators.
To support the analysis of RDII in sanitary sewer systems and planning of corrective actions to address SSOs, EPA
will make the SSOAP Toolbox publicly available free of charge. To those who might find this software package
useful, this report provides a technical foundation to understand the intended use of the tool. It also provides an
understanding of its underlying technical approach and the role of the software in broader efforts to evaluate sanitary
sewer systems and plan improvements. This report is not intended to serve as the user manual for the SSOAP
Toolbox, which is included with the software package to be separately released.
1.1 Background and History
A sanitary sewer system is a wastewater collection system, typically owned by a municipality, authority or utility
district, which is specifically designed to collect and convey only sanitary wastewater (domestic sewage from homes,
and wastewaters from industrial and commercial facilities). Storm water is typically conveyed through a "separate"
system. Sanitary sewer systems are not designed to overflow. However, sanitary sewer systems can overflow when
the system capacity is exceeded due to wet weather (as a result of RDII), when normal dry-weather flow is blocked
for any reason, or when mechanical failures prevent the system from operating properly.
On August 26, 2004, EPA delivered to Congress a report (EPA, 2004) on the impacts and control of combined sewer
overflows (CSOs) and SSOs. The report indicates that the occurrence of CSOs and SSOs is widespread, and EPA
estimates that between 23,000 and 75,000 SSOs occur each year in the United States, resulting in releases of between
3 billion and 10 billion gallons of untreated wastewater. These events occur throughout the United States and cause
or contribute to environmental and human health impacts. Further, the report indicates that there are many existing
structural and non-structural technologies that are well-suited for SSO control.
1.2 Regulatory Framework
The 2004 EPA Report to Congress (EPA, 2004) states that: "SSOs that reach waters of the United States are point
source discharges, and, like other point source discharges from municipal sanitary sewer systems, are prohibited
unless authorized by a National Pollutant Discharge Elimination System (NPDES) permit. Moreover, SSOs,
including those that do not reach waters of the United States, may be indicative of improper operation and
maintenance of the sewer system, and thus may violate NPDES permit conditions."
1-1
-------
Despite the Clean Water Act prohibition on SSOs cited above, the need for additional regulatory clarification has
been suggested, often in the form of an "SSO Rule." To that end, in Fall 1994, EPA initiated a stakeholder process to
"address factual and policy issues related to SSOs," and a Federal Advisory Committee (FAC) was formed for this
purpose.
In 1995, EPA re-formed the FAC as the Urban Wet Weather Flows Federal Advisory Committee to address SSO
issues from a broader water quality perspective. Two subcommittees were established under this committee - one
addressed separate storm water issues, and the second addressed SSO issues, which was known as the "SSO
Subcommittee."
The SSO Subcommittee held 12 meetings between 1995 and 1999 to discuss SSO policy issues and recommended a
number of actions for municipal sanitary sewer collection systems, including development of capacity, management,
operation, and maintenance (CMOM) programs and a "closely circumscribed framework for raising a defense for
unavoidable discharges."
Since 1999, EPA has focused on SSO problems with compliance assistance and enforcement activities according to
the "Compliance and Enforcement Strategy Addressing Combined Sewer Overflows and Sanitary Sewer Overflows,"
issued April 27, 2000 (EPA, 2000). In addition, EPA was or has been evaluating options for improving NPDES
permit requirements for SSOs and municipal sanitary sewer systems.
Although there is no national regulatory program specific to SSOs, a number of EPA regions and state agencies have
initiated programs to address SSOs. EPA Region 4's Management, Operation, and Maintenance (MOM) Program is
one example, along with state programs such as those in California, Oklahoma, and North Carolina. As these and
other programs are implemented, along with updated NPDES permit requirements placed on sanitary sewer systems,
it is EPA's intent that the SSOAP Toolbox will become a useful analysis and planning tool for system
owners/operators to aid them in addressing SSOs.
1.3 Technical Report Organization
The remaining sections of this document include:
Chapter 2 Hydrologic Approach - An introduction to the sanitary sewer system hydrologic approach, including
RDII pathways and mechanisms, and RDII prediction methodologies. This chapter also identifies the initial RDII
methodology for incorporation into the SSOAP Toolbox.
Chapter 3 Hydraulic Approach - An overview of sewer system hydraulics and type of sewer hydraulic capacity
analysis needed to support SSO analysis and planning.
Chapter 4 Data Collection - A description of data collection needs to support SSO planning and analysis using the
SSOAP Toolbox. Guidelines are provided for the collection of information related to system configuration data, and
flow and rainfall monitoring data to support RDII analysis and model development.
Chapter 5 Sanitary Sewer System Analysis and Planning (SSOAP) Toolbox - An overview of the SSOAP
Toolbox and descriptions of its tools and functions in performing a sanitary sewer system capacity assessment.
Chapter 6 SWMM5 Sewer Flow Routing and Capacity Evaluations - Describes the use of SWMM5 within the
SSOAP Toolbox to perform sanitary sewer system capacity assessments. This chapter provides general guidelines on
sewer model development and application, and the steps typically used in establishing baseline conditions and
performing capacity analysis.
Chapter 7 Development and Analysis of System Improvement Alternatives - Guidance for establishing system
improvement objectives, screening potential options for improvements, developing improvement scenarios, and using
SWMM5 model output to evaluate alternatives.
1-2
-------
Chapter 8 Case Study - Provides a case study that demonstrates how the RDII methodology described in this
technical report has been effectively used in SSO planning and analysis.
Chapter 9 References - List of references cited in the report.
1-3
-------
Chapter 2 Hydrologic Approach
This chapter provides an overview of sanitary sewer system hydrology and summarizes alternative hydrologic
approaches to quantify and characterize the response of sanitary sewer systems to wet weather. Methodologies for
RDII prediction and analysis are discussed along with their advantages and disadvantages. This chapter also
identifies the specific RDII methodology included in the current version of the SSOAP Toolbox. Finally, a detailed
description of the selected RDII methodology is presented.
2.1 Overview of Sanitary Sewer System Hydrology
Sanitary sewer system hydrology is so closely related to urban drainage hydrology that it has effectively become a
sub-specialty of what can be termed simply "urban hydrology." Urban hydrology is primarily a study of rainfall
runoff in urbanized or urbanizing areas and how much and how often rainfall is captured by a collection system,
infiltrates into the soil, or runs off the land surface to receiving waters. Once these questions are answered, engineers
turn to the sister science of hydraulics to evaluate resulting water surface elevations and flow velocities - namely,
how high water levels will rise, how fast they will flow, and how full sewers will become.
Urban hydrology, as commonly practiced, is an inexact science. It seeks to balance needs for accuracy in making
infrastructure decisions against the costs of data collection and continued model calibration. This is also evident with
hydrology affecting sanitary sewer collection systems. Therefore, it is important that infrastructure designers have a
fundamental understanding of urban hydrology to minimize potential uncertainties and recommend solutions in a
cost-effective manner.
Although closely related, the response of a sanitary sewer system to wet-weather conditions differs from that of a
storm sewer system. The hydrologic processes for a storm sewer system are more straightforward and understood in
which surface runoff creates the predominant response. The hydrologic processes in the sanitary sewer system are not
as well understood, nor are they as accurately simulated with the degree of reliability needed for infrastructure
improvements. As a result, empirical methods derived from actual flow data are commonly applied to estimate the
hydrologic response in the sanitary sewer system, rather than the use of physical processes that are difficult to
characterize.
Vallabhaneni et al. (2002) and Wright et al. (2001) discussed various practices in simulating RDII using rainfall-
runoff tools. Attempts to simulate the response of a sanitary sewer system to rainfall using rainfall-runoff analysis
tools typically yield less reliable results due to inherent differences in the physical process compared with those
driving a storm sewer system response. Hence, a customized RDII methodology is developed for analyzing sanitary
sewer systems as part of the SSOAP Toolbox.
The three major components of wet-weather wastewater flow into a sanitary system - base wastewater flow (BWF),
groundwater infiltration (GWI), and RDII are illustrated in Figure 2-1 and are discussed below.
2-4
-------
Time
Figure 2-1. Three components of wet-weather wastewater flow.
BWF, often called base sanitary flow, is the residential, commercial, institutional, and industrial flow discharged to a
sanitary sewer system for collection and treatment. BWF normally varies with water use patterns within a service
area throughout a 24-hour period with higher flows during the morning period and lower during the night. In most
cases, the average daily BWF is more or less constant during a given day, but varies monthly and seasonally. BWF
often represents a significant portion of the flows treated at wastewater treatment facilities.
GWI represents the infiltration of groundwater that enters the collection system through leaking pipes, pipe joints, and
manhole walls. GWI varies throughout the year, often trending higher in late winter and spring as groundwater levels
and soil moisture levels rise, and subsiding in late summer or after an extended dry period.
GWI and BWF together comprise the dry-weather flow (DWF) that occurs in a sanitary sewer system. Because the
determination of GWI and BWF components of DWF is not an exact science, various assumptions related to the
water consumption return rates and wastewater composition during early morning hours are typically used to help
estimate these flows components.
RDII is the rainfall-derived flow response in a sanitary sewer system. In most systems, RDII is the major component
of peak wastewater flows and is typically responsible for capacity-related SSO and basement backups. Snowmelt
may also cause RDII flows. RDII flows are zero before a rainfall event, increase during the rainfall event, and then
decline to zero some time after the rain stops. For cases with less than saturated antecedent moisture conditions,
surfaces and soils may take up some of the rainfall early in the event before a response is observed and, if the event is
small enough, there may not be a sanitary system response. The maximum amount of rainfall that does not produce a
response in the system is termed the "initial abstraction."
Figure 2-2 depicts various pathways of RDII into sanitary sewer systems. "Inflow" is the water that enters the
sanitary sewer system directly via depressed manhole lids and frames, downspouts, sump pumps, foundation drains,
area way drains and cross-connections with storm sewers. Although direct connections such as downspouts, sump
pumps, foundation drains, and areaway drains are no longer common design practices, they still exist and contribute
2-5
-------
to inflow in many older sanitary systems. Inflow typically occurs shortly after a rainfall starts and stops quickly once
it stops. Inflow is typically the major component of the RDII peak flow.
Rainfall-derived infiltration (RDI) refers to rainfall runoff that filters through the soil before entering a sanitary sewer
system through damaged pipe sections, leaky joints or poor manhole connections. These defects can occur in both the
public right-of-way portions of the sanitary sewer system or in individual service laterals on private property.
Infiltration processes typically extend beyond the end of rainfall and takes some time to recede to zero after the storm
event. A system may experience a fast RDI response, a slow RDI response, or both.
Figure 2-2. Pathways of infiltration and inflow into sanitary sewer systems.
In areas characterized by soils with high percolation rates, RDI can quickly enter shallow service laterals and sewer
system defects, contributing significantly to the peak wet-weather response. RDI is typically the major component of
the total RDII volume, especially during periods of high antecedent soil moisture conditions when the recession limb
of the wet-weather response can last for several days after the wet-weather event.
The response of a sanitary sewer system is quite complex. Various factors control RDII responses in addition to the
rainfall (volume, intensity, and duration) and antecedent moisture conditions, including depth to groundwater, depth
to bedrock, land slope, number and size of sewer system defects, type of storm drainage system, soil characteristics,
and type of sewer backfill. Further, RDII responses can vary greatly due to spatial rainfall distributions over a
sewershed.
2.2 RDII Prediction Methodologies
The ability to estimate RDII flows reliably is critical for developing SSO control plans. The methods presented in this
section focus on estimating RDII hydrographs based on flows observed at a point in the sanitary sewer system as
monitored by a flow meter. These methods do not require that the number and severity of the defects that allow
rainfall runoff to enter the system be determined, though the results can sometimes be used to infer the relative
severity of structural defects.
2-6
-------
RDII quantification methods require information on the rainfall that fell over the sewershed. The required
information varies with specific methodology and may include total rainfall depth, peak intensity, and a complete
rainfall hyetograph. The successful application of these methods relies on accurate rainfall and wastewater flow data
measurements.
This section also discusses the strengths and weaknesses associated with each RDII prediction method. The goal is to
provide an understanding of the available analysis methods and to support the choice of the RDII predictive
method(s) for inclusion in the SSOAP Toolbox.
The RDII process and associated data are very much site-specific. No single flow prediction method is likely to be
universally applicable. Also, the RDII prediction methods described in this section do not directly account for snow
fall and snow melt processes. If analysis of RDII response to snow melt is needed and this is a major concern in
defining sewer system improvements, then special procedures may need to be developed to meet project needs.
Typically, RDII response to the peak rainfall events establishes the basis for the system improvements rather than
snow melt events. In an actual application, however, the objective of the study, as well as the availability of data,
time, staff and funding, should be considered when selecting the most appropriate method.
2.2.1 Literature Review Studies
The following summary descriptions of literature review studies and overview of RDII prediction methods are
primarily derived from a report prepared by EPA in support of this SSO CRADA that reviewed sewer design
practices and RDII flow predictions (Lai, 2007).
A Water Environment Research Foundation (WERF) publication (Bennett et al., 1999) and a conference paper
(Schultz et al., 2001) reviewed RDII prediction methods in literature dating back to 1984. Their literature searches
included online catalogs at the University of Wisconsin at Milwaukee and Madison. Additional references were
identified through contacts with engineering firms and municipal agencies. Forty-two documents were compiled and
reviewed in preparing the WERF publication. A detailed discussion on this WERF study is presented in the next
section.
In parallel to the 1999 WERF study, Crawford et al. (1999) reviewed three RDII prediction methods: the constant unit
rate, rainfall/flow regression, and percent of rainfall volume (R-value). They evaluated merits and limitations of these
methods using applications for the City of Salem, Oregon (with a population of about 160,000) and the City and
County of Honolulu, Hawaii (population of about 1 million).
Wright et al. (2001) reviewed literature on RDII estimation techniques that have appeared since 1993. They classified
the methods into three main groups: the volume-based "rational" method (or R-value method), the unit hydrograph
method, and the physical processes modeling method. The paper discussed various unit hydrograph methods,
including:
• Synthetic unit hydrograph, in which the shape of the hydrograph is pre-defined
• Data-derived unit hydrograph, which uses multiple regressions to derive the ordinates of a unit hydrograph
directly from measured rainfall and RDII flow data
• Conceptually derived unit hydrograph, which uses a system of cascading linear reservoirs.
The 1999 WERF study identified eight broad categories of RDII quantification methods. Three cooperating
municipal agencies tested these methods against monitored rainfall-flow data: the Metropolitan Council of
Environmental Services, St. Paul, Minnesota; the Bureau of Environmental Services, Portland, Oregon; and the
Montgomery Water Works and Sanitary Sewer Board, Montgomery, Alabama. The eight described methodologies
were:
1. Constant unit rate method
2. Percentage of rainfall volume (R-value) method
2-7
-------
3. Percentage of stream flow method
4. Synthetic unit hydrograph method
5. Probabilistic method
6. Rainfall/flow regression method
7. Synthetic stream flow regression method
8. Methods embedded in hydraulic software
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 because of a variety of site conditions and analysis application needs. The study
identified criteria that are used to test the alternative RDII prediction methods using flow and rainfall data supplied by
the three cooperating sewer agencies. Specifically, the methods should be able to:
• Predict peak flow for individual storms
• Predict volume for individual storms
• Predict the hydrograph timing, shape, and recession limb
• Predict peak flows for multiple storms
• Predict volume for multiple storms
• Operate on commonly available data
Because the characteristics of the available data vary among the cooperating sewer agencies, the collected data were
not applicable to all eight RDII prediction methods considered in 1999 WERF study.
Another important criterion is the adaptability of the analysis and RDII prediction methods to guide rehabilitation
programs, and to develop and assess alternative corrective measures. This criterion was not explicitly considered in
the 1999 WERF study, but it is included when selecting the RDII methodology to be included in the SSOAP Toolbox.
2.2.2 Overview of RDII Prediction Methodologies
The following section discusses the strengths and weaknesses of each of the alternative RDII prediction
methodologies primarily derived from a companion literature review report prepared by EPA in support of this
CRADA (Lai, 2007).
The constant unit rate method calculates RDII as a constant unit rate based on sewershed characteristics. The unit
RDII rates (e.g., gallons per inch of rainfall per acre; gallons per acre land use; gallons per capita; gallons per inch-
diameter-mile) from sewersheds are estimated based on flow and rainfall monitoring and sewershed characteristics.
RDII is calculated by multiplying RDII unit rates with respective to tributary sewershed characteristics (including
area, land use, population, and pipe diameter/length/age) to derive RDII. The constant unit rate method is simple to
apply and can help predict RDII volume for unmonitored conditions. However, it is difficult to develop reasonable
estimates of unit rate constants, which may vary by storm, antecedent moisture, and season. The method also lacks
the capability of developing hydrograph timing and its shape.
Crawford et al. (1999) based their evaluation of the constant unit rate method on data applications for the City of
Salem, Oregon, and the City and County of Honolulu, Hawaii. They concluded that peak hourly RDII rates per acre
from a five-year storm increase significantly as the average age of sewer pipes increases from 10 to 30 years. To
overcome this limitation they suggested that unit RDII rates should increase with the age of the sewer system.
The constant unit rate method is suitable for simple applications, such as sizing conveyance facilities for relatively
frequent storms, provided there is available rainfall and wastewater flow data from a similar sewershed or sewersheds.
If data are not available, precipitation gauges and wastewater flow monitors can be installed with a reasonable level of
effort to obtain the needed information.
-------
However, for more complex applications that require multiple storms in multiple sewersheds of varying ages to be
analyzed, for the evaluation of major trunk sewer system where the timing causes peak flows to not be additive, or for
applications requiring the development and assessment of flow equalization facilities where hydrograph timing and
shape are needed, this method is insufficient. It also would not provide adequate information for sewer rehabilitation
programs in which sewersheds and predictions of potential RDII flow reduction must be prioritized.
A subset of this method is the method where peak flows are correlated against peak rainfall intensities for observed
events. Through this correlation, it may be possible to estimate peak flows of a given return period based on the
rainfall frequency. Caution must be used in correlating the return periods of peak rainfall and peak flows. The data
will typically be widely scattered due to various factors including antecedent moisture conditions, rainfall distribution,
errors in peak flow measurement, and the inability to accurately determine the rainfall intensity over the whole
sewershed. Furthermore, unit rates developed for relatively small and frequently occurring storms observed in a
typical flow monitoring program cannot be reliably extrapolated to estimate peak flows for infrequently occurring
storms such as the two-, five- or 10-year design storms.
The percentage of rainfall volume (R-value) method calculates RDII volume as a fixed percentage of the rainfall
amount. This method is more adaptable than the constant rate method, in that it accounts for seasonal effects and
multiple sewershed types and ages as long as corresponding monitored rainfall and wastewater flow data are
available. The R-value is relatively easy to calculate from flow monitoring and rainfall data, and can be used as a
guide to determine the relative number and size of RDII defects within a particular sanitary sewer system.
However, this method is unable to estimate peak flows or hydrograph timing and shape. Thus, it should be cautioned
when this method is used to calculate the peak flow capacity of sanitary sewer facilities or to model peak flows in
sanitary sewer systems. As in the case of all other RDII methodologies, one needs to understand that this method
makes simplified assumptions about the complex physical processes that affect RDII volume and the rate at which it
enters sanitary sewers.
Regression analysis plots for observed RDII volumes and rainfall depths often show a wide scattering of the data
points. Antecedent moisture conditions significantly affect the RDII response to rainfall events. When attempts are
made to extrapolate the calculated R-values for frequent storms that are typically occur in monitoring period to an
infrequent storm such as a 10-year design storm, even the slightest change in the slope of the regression line can
significantly impact the predicted R-value. A monitoring program typically captures flows from low-intensity, low-
volume storm events that are less than the one-year return period rainfall volume. Care must be taken when
extrapolating these values to estimate flow volumes for high-intensity, high-volume design storm events, such as a
five- or 10-year storm.
In using the R-value method to estimate RDII rates for high intensity and less frequent storms, Crawford et al. (1999)
cautioned that the R-values should be appropriately tapered to account for the upper limit of peak flows that leaky
sewers can take in. This consideration recognizes that there is a limit to the peak flow capacity of inflow connections,
leaky manholes, and other sewer system defects. The inflow rate to the system cannot exceed the capacity of the
defects. The conveyance capacity of the upstream sewers also limits the observed volumes.
The R-value method is suitable for applications requiring the determination of volume estimates for multiple storms,
multiple sewershed types and ages, and seasonal impacts as long as there is corresponding precipitation and
wastewater data. This method is also useful in identifying the relative magnitude of infiltration and (through
inference) the extent and severity of defects between metered areas. Note that to develop an accurate R-value
estimate, only the sewered portion of the flow meter tributary area must be taken into account. When making
comparisons of R-values among the metered areas, it is important that a consistent measure of "sewered" area be used
in estimating the R-value. This method readily uses existing available rainfall and monitored sewer flow data. If
additional monitoring data are needed, supplemental gauges and meters can be installed. Rainfall data can also be
supplemented with radar rainfall estimates to better determine the rainfall that fell over the upstream sewershed.
2-9
-------
This method is insufficient for sizing conveyance and storage improvements where hydrograph timing, peak flows,
and shape are needed.
The percentage of stream flow method is similar to the previous rainfall method, but it uses gauged stream flows in
nearby watersheds as an independent variable. This method recognizes that stream flows inherently account for the
effects of antecedent moisture conditions that influence groundwater levels and resulting GWI in sewers. Therefore, a
relationship can be developed between stream flow and sewer flow data. Realistically, the percentage of stream flow
method is only applicable for relatively rare cases in which stream gauge data are available in watersheds with basin
characteristics similar to the sewersheds being analyzed. Establishing a new stream gauging station is much more
demanding and labor intensive than installing a wastewater flow monitor in a sewer. Another problem with this
method is that a sewershed may be much smaller than a gauged streamshed, and the scaling factors can produce non-
representative analytical results. Again, care should be applied when extrapolating relationships between sewer and
stream flow to less frequent rainfall events as the stream has a higher capacity to accept and convey flows than a
sanitary sewer system. In addition, where sewer system problems are predominantly driven by the inflow portion of
RDII, direct correlation of stream flow and sewer flow data offer limited benefit in characterizing and predicting total
RDII.
One variation on this approach is to use a stormwater hydrologic model to simulate stream flows and then correlate
these modeled stream flows to the flows observed in the sanitary sewer, as described in the synthetic stream flow
regression method.
The synthetic unit hydrograph method assumes that RDII in a sewer responding to rainfall can be quantified and
characterized via classical unit hydrograph techniques used to analyze storm water runoff in a watershed. The method
calculates the RDII hydrograph from a specified unit hydrograph shape that relates RDII to unit precipitation volume,
specified time duration and sewershed characteristics. In classic watershed analyses, a unit hydrograph is defined as
the direct runoff hydrograph resulting from a unit depth of excess rainfall (e.g., 1 inch or 1 mm, produced by a storm
of uniform intensity and specified duration over a watershed). It was first proposed by Sherman (1932) for flood
estimation, and has since found wide-ranging applications in surface water hydrology. For surface water applications,
unit hydrographs are generally derived from stream flow data and estimates of rainfall excess. The unit hydrograph is
applied to the hyetograph of rainfall excess to estimate the surface runoff hydrograph. The method is adapted for
sanitary sewershed analyses in which the unit hydrograph is related to the sewer system RDII response to a unit depth
of rainfall over the sewershed.
There are three families of unit hydrographs: the synthetic unit hydrograph (SUH), in which the shape of the
hydrograph is pre-defined; the data-derived unit hydrograph, which uses multiple regressions to derive the ordinates
of a unit hydrograph directly from measured rainfall and RDII flow data; and the conceptually derived unit
hydrograph, which uses a system of cascading linear reservoirs.
The simplest SUH has a triangular shape and many formulations, such as that used in SWMM5 (EPA, 2007). Huber
and Dickenson (1988) used three unit hydrographs to account for fast, medium, and slow RDII responses. SWMM
uses what is known as the RTK curve-fitting method, where R is the fraction of rainfall volume entering the sewer
system as RDII during and immediately after the rainfall event, T is the time to peak, and K is the ratio of the time of
recession to T. The R, T, K (RTK) method was first developed for an RDII study for the East Bay Municipal Utility
District in Oakland, California (Giguere and Riek, 1983). The method, which was included as an option in the
SWMM Runoff Block (Huber and Dickenson, 1988), is probably the most popular SUH method due to its
applicability to analyze RDII response to rainfall. Since the three unit hydrographs distinctively represent the
quantitative contribution of inflow and infiltration to the overall RDII hydrograph, 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. The method also provides information on the relative magnitude of rapid infiltration and long-term
infiltration in determining the peak flows and total RDII volumes.
The data-derived unit hydrograph (UH) is not based on calibration methods like the RTK method. Instead of
2-10
-------
beginning with an assumed shape characteristic, the data-derived UH is a linear transform function completely
derived from measured data. Namely, it derives the ordinates of a unit hydrograph directly from measured rainfall
and RDII flow data using multiple linear regression or linear programming techniques. The goal is to find a vector of
unit hydrograph ordinates that minimizes the difference between the time series of measured flow and the estimated
flows.
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 provides an important conceptual link between purely empirical methods and physically based
conceptual models such as SWMM, which uses the non-linear reservoir approach and the continuity and momentum
principles. As with the data-derived UH method, the reservoir parameters use some optimization techniques, such as
linear least squares regression or linear programming. The reservoir parameters may be constrained to derive
physically realistic values when linear programming is used. Wright et al. (2001) reported that physically unrealistic
values (i.e., negative UH ordinates) may be derived from an unconstrained ordinary regression method.
In summary, the unit hydrograph method is suitable for complex sewershed analysis applications requiring multiple
storms, multiple sewershed types and ages, and system response hydrograph timing and shape. The method also
accounts for the seasonal response of sewer systems to wet-weather conditions as long as corresponding precipitation
and wastewater data are available. The method can be used to rank relative sewershed leakiness and prioritize sewer
system rehabilitation efforts, and can estimate RDII reduction from selected rehabilitation methods by applying
appropriate reduction factors to the RDII and GWI hydrographs. All three unit hydrograph methods can be adapted to
a wide range of applications and needs. However, the three methods differ greatly in the quantity of monitored data
required and the level of effort and skill needed to apply them.
An analyst can learn to calibrate a series of three triangular unit hydrographs using monitored rainfall and wastewater
flow data in a relatively short period of time. However, because the data-derived UH and cascading reservoir
methods require larger quantities of monitored data, the analyst must apply multiple linear regression or linear
programming techniques. Experience has shown that the additional level of complexity, effort, and cost required by
the latter two UH methods typically are not justified by a limited increase in the level of accuracy or precision in the
analysis.
With the SUH method, the analyst should exercise care when extrapolating the sewer system RDII response that is
determined during a limited flow monitoring period to an unmonitored, large, and infrequent storm, such as a two-,
five- or 10-year design storm. Statistical methods complementing the RTK approach are available to extrapolate the
RDII response quantified from monitored conditions to unmonitored conditions (natural and synthetic). Vallabhaneni
et al. (2002a) reported a successful statistical model with multi-variable regression developed using the results of a
RTK analysis for a range of monitored flow conditions. This statistical model was developed based on the observed
R-value relationship with event rainfall depth, antecedent one-month precipitation, and GWI. The resulting
regression equation was then applied to predict RDII for unmonitored conditions. The statistical model yields reliable
results only when the input to the model was consistent with the observed interdependency among the variables
during the monitoring conditions. Analysts must ensure that extrapolated R-values take into account the physical
reality that there is an upper limit to the peak flows that leaky sewers can accommodate.
The probabilistic method uses probability theory to calculate the RDII of a given recurrence interval from long-term
records of peak wastewater flows. Various statistical approaches that analyze peak stream flows and rainfall
intensities can be used in the statistical analysis of peak wastewater flows. This method is valuable because it is
becoming quite common for municipalities to assess sanitary sewer system hydraulic performance using a large and
infrequent synthetic storm event, such as a two-, five- or 10-year design storm. While the probabilistic method can
predict peak daily RDII flows for large, infrequent storms, the method's accuracy deteriorates rapidly if a long-term
dataset in the order of 10 to 25 years is not available. In general, many sanitary sewer systems can not afford to
collect monitoring data for such extended periods.
2-11
-------
The probabilistic method is well suited for applications that require the characterization and quantification of peak
sanitary sewer flow associated with a specific recurrence interval. However, it does not provide flexibility in
assessing multiple storms, multiple sewershed types and ages, and seasonal impacts. The method is also not suitable
for applications that require hydrograph timing and shape.
The rainfall/flow regression method calculates peak RDII flows from rainfall data and provides a means of
determining the shape and magnitude of a RDII hydrograph. 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. Regression equations have stated limits in their parameter ranges and will have even
larger errors for sanitary sewersheds that are atypical in construction, slope, age or condition.
Crawford et al. (1999) used regression equations derived from winter data to produce a good match between the
monitored and equation predicted hydrographs from other winter storms. However, when they applied the regression
equations to summer and early fall storms, they observed large discrepancies between observed and equation
predicted flows. To overcome this limitation, they suggested developing a separate series of regressions to represent
the seasonal nature of RDII processes. Hence, adequate and representative rainfall and flow data are prerequisites for
a successful application of the regression method.
This method provides an alternative to the probabilistic method for projects requiring analysis of sanitary sewer
collection systems for specific design storms. It has a greater flexibility for a wider range of applications than the
probabilistic method, but it requires the same long-term record of monitored precipitation and wastewater flow to be
effective.
In the synthetic stream flow regression method, RDII is calculated from synthetic stream flow records and
sewershed characteristics using regression equations derived from multiple regression techniques to correlate
watershed hydrologic responses to sewer flow responses, specifically infiltration portion of RDII. The synthetic
stream flow records typically are generated using calibrated hydrologic simulation models. The 1999 WERF study
reported that this method was successfully applied in Milwaukee for sewerage system improvement planning.
However, it requires a calibrated watershed runoff model, which often does not exist for many sanitary sewer system
improvement projects unless special efforts are invested.
Finally, methods embedded in publicly or commercially available hydraulic modeling software use one or more
of the prediction methods discussed previously. The most notable is the SWMM program, which incorporates the
synthetic unit hydrograph method in its codes. A number of commercial software packages use the SWMM
computational engine and therefore use the RTK method. Further, the RTK method has been incorporated into non-
SWMM based commercial modeling packages.
Many of the model applications in the sanitary sewer system studies are derived from modeling tools and methods
originally developed to simulate stormwater runoff. In many cases, the models used for land surface runoff have been
applied to model RDII. As an example, the hydrologic model parameters such as area, width, and roughness are used
to simulate flows using kinematic wave procedures as programmed in SWMM and other commercially available
models. These models have been used to simulate observed RDII flows. The total runoff area and the runoff capture
coefficient can be adjusted to match the observed RDII volumes. The width, slope, and roughness can also be
adjusted to calibrate the model to the observed flow pattern. This method is related to the synthetic stream flow
method, except that the model parameters are calibrated directly to the observed wastewater flows. One drawback of
this method is that it is not possible to conceptually correlate the model parameters to the physical characteristics of
the sewers and sewer defects that produce RDII. As with any of the models, caution must be used when extrapolating
the models calibrated to frequently occurring observed events to design storms that occur less frequently.
2.2.3 Recommendation of RDII Method for SSOAP Toolbox
From the previously discussed literature review, it can be concluded that there is no single RDII prediction method
that is universally applicable. Rainfall and flow observations include a wide variety of site-specific characteristics.
2-12
-------
All methods require monitored data to evaluate and validate predictive capabilities. However, the amount of data
required varies.
Sewer routing models are used to extrapolate RDII predicted from the limited monitored data to predict flows under
future build-out conditions. 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 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 the SUH and rainfall/flow regression methods were preferred for predicting flows
for single as well as multiple storm events. The ability of good multiple-storm peak and volume prediction is
important in extrapolating data beyond the calibration events for a prolonged period simulation, which is essential for
evaluating the effect of RDII on storage and treatment requirements.
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) in developing a sewer system/treatment improvement plan. The
rainfall/flow regression methods will be attractive if there are two or more 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 relate the RDII rate to the preceding rainfall amounts corresponding to various time
periods (e.g., one hour, two to three hours, four to six hours, seven to 12 hours, 12-24 hours, one to two days, four to
seven days, and seven to 15 days) 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 and
identify if the RDII problems are caused by inflow, infiltration or both.
On the other hand, the RTK method (one kind of the SUH method) uses up to 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 rapidly responding and presumably inflow
driven. If more of the total R-value is allocated to R2 and R3, this indicates that the RDII is more slowly responding
and presumably infiltration driven. This knowledge is useful during a sewer system evaluation survey (SSES) to
determine the best SSES approach to use in a particular area, as well as whether a point repair or a comprehensive
rehabilitation approach may be 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 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 it has been widely used
and proven as a valuable method in separate sanitary sewer system analysis associated with storm events (Giguere
and Riek, 1983; COM et al., 1985; Miles, et al., 1996; Vallabhaneni, et al., 2002a).
In conclusion, the synthetic unit hydrograph method, or RTK method, was selected as the primary hydrologic analysis
approach for the SSOAP Toolbox during its initial development. The method is widely used and can analyze and
predict RDII in sanitary sewer systems. It has demonstrated flexibilities for a wide range of project needs. As long as
corresponding precipitation and wastewater data are available, the method is capable of complex analyses involving
peaks and volumes of multiple storms and multiple sewershed types and system ages. The RTK method can account
for system response hydrograph timing and shape and seasonable responses of sewer systems to wet-weather
conditions. The method can also be used to guide sewer system rehabilitation efforts and estimate the RDII reduction
from selected rehabilitation methods.
The probabilistic method and rainfall/flow regression method show promise as supplemental approaches for the future
enhancement of the SSOAP Toolbox. These alternatives are well suited for analysis applications requiring the
2-13
-------
characterization and quantification of sanitary sewer flow associated with a specific recurrence interval event. As
long as a sufficient record of rainfall and monitored sewer flow data is available, these methods are viable future
options for future SSOAP releases.
2.3 KDII Unit Hydrograph Method
The SSOAP Toolbox uses the RTK method to derive the sanitary sewer system RDII response using the associated
rainfall and flow monitoring data. The Toolbox enables users to estimate R,T,K parameters for each rainfall/flow
monitoring event and generate corresponding RDII hydrographs. The user has the choice of exporting the RDII
hydrographs directly into SWMM5 from the Toolbox or inputting R,T,K values into the SWMM5 input file and
generating RDII hydrographs within SWMM5. Alternatively, the user can export the RDII hydrograph from the
Toolbox to incorporate into other hydraulic analysis tools. Both SWMM5 and the SSOAP Toolbox have similar
computational routines to develop RDII hydrographs using the RTK method. Some commercially available sewer
hydraulic modeling programs now allow users to simulate flows using the RTK unit hydrographs, and they can use
the RDII hydrograph generation tool in the SSOAP Toolbox.
The RTK method is similar to unit hydrograph methods that are commonly used to simulate flows in storm water
runoff analyses. This method is based on fitting three triangular unit hydrographs to an actual RDII hydrograph
derived from flow meter data. A unit hydrograph is defined as the flow response that results from one unit of rainfall
during one unit of time. Figures 2-3 through 2-5 depict the RTK method and how RDII hydrographs are generated.
Figure 2-3 depicts the triangular unit hydrograph in response to one unit of rainfall over one unit of time. This unit
hydrograph is described by the following parameters:
R: the fraction of rainfall volume that enters the sewer system and equals the volume under the hydrograph
T: the time from the onset of rainfall to the peak of the unit hydrograph in hours
K: the ratio of time to recession of the unit hydrograph to the time to peak
A: sewered area
P: rainfall depth over one unit time
Volume: volume of RDII in unit hydrograph
Qp: peak flow of unit hydrograph
VOLUME=R*P*A
TIME-
T+ K*T
Figure 2-3. Example of a triangular unit hydrograph.
2-14
-------
This RTK hydrograph generation method performs has two basic steps, and are illustrated in Figures 2-4 and 2-5.
The first step is to define R,T,K parameters in response to one unit of rainfall over one unit of time. Three unit
hydrographs are typically used because the shape of an RDII hydrograph is too complex to be well represented by a
single unit hydrograph as shown in Figure 2-3. The RDII hydrograph can be generated using less than three sets of
R,T,K. However, experience indicates that it often requires three unit hydrographs to adequately represent the various
ways that precipitation becomes RDII. The first triangle represents the most rapidly responding inflow component,
and has a T of one to three hours. The second triangle includes both rainfall-derived inflow and infiltration, and has a
longer T value. The third triangle includes infiltration that may continue long after the storm event has ended and has
the longest T value. In this first step, the R,T,K parameters for each of the three triangles are defined for each unit
rainfall over one unit time frame. The sum of the R values for each of the three unit hydrographs (i.e., RI, R2, and R3)
must equal the total R value for the rainfall event. Figure 2-4 below depicts a summation of three unit hydrographs
into a total RDII hydrograph in response to one unit rainfall over one unit time frame.
SUMMATION OF THREE UNIT HYDROGRAPHS
FIRST UNIT
HYDROGKAPHOF
AND 1
TOTAL RDII HYDROGRAPH
RESULTING FROM
RAINFALL, P
SECOND UNIT
HYDROGRAPH OF
R2,T2,ANDK2
THIRD UNIT
HYDROGRAPH OF
R,,T,,ANDK,
0 Ti T2 Ti+Ki*Ti Ts T2+K2*T2 TIME'
T3+K3*T3
Figure 2-4. Summation of three unit hydrographs.
The unit hydrograph parameters in Figure 2-4 can be described as follows:
T], T2, and T3: time to the peak of respective unit hydrographs
KI, K2, and K3: recession coefficients of the respective unit hydrographs
TI +Ki * TI : last time unit of the first unit hydrograph
T2+K2*T2: last time unit of the second unit hydrograph
T3+K3*T3: last time unit of the third unit hydrograph and the total unit hydrograph
RI, R2, and R3: R-values of respective unit hydrographs; R = Ri+R2+R3
The second step of the unit hydrograph methodology is to sum all of the RDII unit hydrographs that were developed
for each unit of time within a rainfall event to develop a total event RDII hydrograph. Figure 2-5 illustrates the
summation of three hydrographs as an example rainfall event. This would represent the hydrograph from a rainfall
event lasting three unit time steps. If a rainfall event has rainfall duration is two hours with a 15-minute unit time
2-15
-------
step, then the hydrograph developed by this method would be the summation of the 24 unit hydrographs that resulted
from each 15-minute rainfall increment.
A description of parameters shown in Figure 2-5 is as follows:
PI, P2, and P3: successive rainfall depths over each unit time step
Total Hydrograph: summation of synthetic hydrographs for each unit rainfall
The SSOAP Toolbox automatically sums the unit hydrographs to derive the total RDII hydrograph for a sewershed
and selected rainfall events.
The toolbox provides graphical tools and statistical comparisons of predicted and observed peak flows and flow
volumes to assist the users in identifying the combination of R, T, and K values that best match the simulated
hydrograph with the observed RDII hydrographs. This is accomplished by a curve fitting procedure. In this
procedure, the flow monitoring data is first decomposed into the DWF and RDII components. Then the DWF
component is subtracted from the total hydrograph to derive the RDII component. The best combination of the R, T,
and K values for each of the three triangular unit hydrographs is determined iteratively until the derived RDII
hydrograph closely approximates the observed RDII hydrograph.
Pi
TOTAL HYDROGRAPH FROM 3 UNITS OF RAINFALL
RDII HYDROGRAPH FOR P,
RDII HYDROPGRAPH FOR R
TIME
Figure 2-5. Summation of synthetic hydrographs.
Once developed, the RTK unit hydrograph parameters and the rainfall hyetograph of interest can be used to define the
RDII inflow hydrograph for sanitary sewer system flow evaluations using hydraulic simulation methods described in
Chapter 3.
2-16
-------
Chapter 3 Hydraulic Analysis
This chapter provides an overview of sanitary sewer system hydraulics. The SSOAP Toolbox includes linkage to the
SWMM5 software to provide the hydraulic analysis function in support of sewer system analysis and planning.
The SSOAP Toolbox functionality is applicable to analyzing wet-weather flow in sanitary sewer systems as opposed
to combined sewer or storm sewer systems, for which the wet-weather response mechanisms are very different.
However, despite different wet-weather response mechanisms, many of the hydraulic concepts that apply to sanitary
sewers also apply to combined and storm sewers.
Because of the closely related nature of the SSOAP Toolbox analytical capabilities and the hydraulics of sewer
systems, this chapter is included as background. However, it is only intended to introduce the concepts and provide
references to other sources for more detailed information on these topics.
3.1 Overview of Sanitary Sewer System Hydraulics
Sanitary sewer systems are generally constructed as a network of pipe conduits ranging in size from 8 inches in
diameter up to 8 ft (or even larger) at the downstream end of large networks. Most systems drain by gravity to the
terminus at a wastewater treatment plant. However, where system configuration and topography do not allow for
gravity flow conditions, pump stations and force mains are used to deliver flow to the plant or to a point in the system
where gravity drainage is available. In some relatively rare cases, such as typically very flat areas and areas where
subsurface conditions preclude the use of gravity sewers, pressure systems are used for wastewater collection and
transport.
Flow conditions in sanitary sewers vary and are unsteady and non-uniform. During dry-weather conditions, flow in
gravity flow portions of sanitary sewer systems generally are designed with the water surface at less than pipe crown,
i.e., free-surface flow. This flow may be either sub-critical or super-critical. During wet-weather, flows typically
increase, often significantly. Free-surface pipe flow may give way to surcharge flow conditions where pipes are full
and under pressure. After the wet-weather event, surcharge flow conditions typically transition back to free-surface
flow conditions. The three basic flow regimes - sub-critical, super-critical, and surcharge and the transitions between
them - are depicted graphically and described in detail in Yen (1986).
Analyzing surcharge flow conditions is particularly important to SSO planning, as it is typically surcharge conditions
that give rise to SSOs and other operational problems in sanitary sewer systems. Also during wet weather, it is
common for backwater conditions to develop in sanitary sewers. Backwater conditions often significantly influence
the water surface profile in the sewer, which can lead to SSOs and other problems when the water surface rises
beyond critical elevations (e.g., SSO weir crests). The analysis of backwater conditions is aided by using steady flow
backwater curves to classify backwater surface profiles as mild, steep, critical, horizontal and adverse. A detailed
discussion of backwater curves can be found in Chow (1959).
3.2 Analysis Methods: Static, Kinematic, and Dynamic
Flow in sanitary sewer systems is unsteady, but may in some cases and for some purposes be treated as steady flow.
However, in most cases the rapid changes in sanitary sewer flows that occur in response to wet weather require
dynamic routing rather than static (steady-state) analysis methods.
3-17
-------
Dynamic (unsteady flow) analysis may be performed at varying levels of complexity, ranging from relatively simple
methods using kinematic wave approximations to those that apply the full dynamic wave (St. Venant) equations for
continuity and momentum. Kinematic wave and other simplified forms are derived by dropping terms from the full
dynamic wave equation. A complete discussion of the unsteady flow equations can be found in a number of
references. One of the most often cited is Lai (1986).
The discharge (conservative) form of the momentum equation is commonly written (Yen, 2004) as:
1 dQ 1 d (PQ2 \ . dh „ „
— + (/ ^ ) + cos 9 S0 + S f = 0
gA dt gA dx ^ A J dx ° f
where:
x = longitudinal direction of sewer
A = flow cross-sectional area normal to x
y = coordinate direction normal to x on a vertical plane
h = depth of flow of the cross-section, measured along y-direction
Q = discharge through A
S0 = channel slope, equal to sin 9
0 = angle between sewer bottom and horizontal plane
Sf = friction slope
g = gravitational acceleration
t = time
(3 = Boussinesq momentum flux correction coefficient for velocity distribution (Yen, 2004).
The continuity equation can be written as:
dt dx
The solution for the full dynamic wave equation is computationally intensive. Because of this, excessive model
runtimes have historically been experienced, especially for large pipe systems and for long simulation periods.
However, models applying simplified forms of the dynamic wave equation neglect pressure and/or acceleration terms
that may be important in some cases. This has created a dilemma for engineers seeking to balance accuracy
requirements with practical considerations. A good discussion of the comparative trade-offs between the various
simplified forms of the dynamic wave equation can be found in Akan and Yen (1981). Fortunately, the rapid
advances in computer technology that have occurred in recent years have significantly reduced the runtimes required
for a solution to the full dynamic wave equation, which has greatly facilitated the use of models that apply this
approach.
3.3 Pipe Flow Resistance
All forms of the dynamic flow equation retain the friction and gravity terms. The friction term represents the friction
slope using a semi-empirical formula, usually Manning's formula, which applies a roughness coefficient (i.e.,
Manning's n-factor) to compute pipe friction losses.
An accurate definition of the pipe roughness coefficients throughout the modeled sewer network is important to
3-18
-------
accurately simulate hydraulic conditions. Coefficients vary according to a number of factors, including pipe material
and condition. There are a number of references available to help define reasonable values for different pipe
materials. Brater and King (1976) provides values that account for pipe conditions.
Grit deposition within the sanitary sewer system influences both the effective pipe roughness coefficient and the
actual cross-sectional area of flow. Grit deposition is common in sanitary sewer systems, especially in portions of the
network where flow velocities are relatively low, allowing solids to settle.
Because of the difficulty obtaining complete information about grit deposition conditions, which can vary overtime,
and because pipe roughness cannot be measured directly, hydraulic model calibration efforts typically focus on the
pipe roughness coefficient as a key calibration parameter.
3.4 Surcharged Pipe Flow
As previously noted, surcharge conditions can be very important to sanitary sewer system analysis, as surcharging
typically is associated with SSOs and other operational problems. Surcharge conditions can be modeled in two ways:
1. Standard closed-conduit algorithms for pressure flow with an incompressible fluid.
2. The use of a hypothetical (extremely narrow) slot at the crown of each surcharged pipe to maintain free
surface conditions in the model even when the computed water surface exceeds the pipe crown.
The latter approach, credited to Preissmann (Cunge and Wegner, 1964), simplifies the computational strategy, as it
eliminates the need to switch between the St. Venant equation and a surcharge equation. It also eliminates the need to
define and test for surcharge criteria, and track the pipes passing the test, to invoke the computational switch from one
equation to the other.
There are disadvantages to using the Preissmann slot, which include accuracy problems (dampening of flow peaks) if
the slot is too wide and numerical stability problems if the slot is too narrow. Some hydraulic routing models employ
one surcharge solution approach or the other, while other models incorporate both and allow the user to define the
surcharge solution.
3.5 Hydraulic Analysis of Special Structures and Appurtenances in Sanitary Sewers
There are a number of special structures and appurtenances in sanitary sewer systems that are of particular concern in
the hydraulic analysis of these systems. SSOs can occur in the physical system in a number of ways. In some cases,
particularly where constructed overflow points exist, the overflow discharges through an orifice or over a weir. In
some cases the structure may operate as a weir until the structure itself is surcharged, at which point the structure
operates as an orifice. In other cases, the SSO may occur as an overflow discharged through the manhole opening at
the ground surface (often known as a flooded manhole). Weirs and orifices are well-described in the literature. A
good discussion of weir characteristics can be found in French (1985), while Brater and King (1976) cover both
subjects in excellent detail.
Many sanitary sewer systems include pumps and force mains. These structures are generally relatively simple to
model, and each sewer modeling software typically has a means for representing the control rules that define the
pump operation. In some cases, moveable gates (e.g., sluice gates, inflatable dams), either manual or automatic, are
installed in sanitary sewer systems, operation of which can be modeled using control rules. Storage facilities can be
represented in most models using a stage-volume (area) relationship.
3-19
-------
Chapter 4 Data Collection
4.1 Introduction
This chapter discusses the data requirements to support SSO planning and analysis using the SSOAP Toolbox. One
of the primary functions of the Toolbox is to assist users with predicting and estimating RDII. The primary data
needs for RDII prediction tools in SSOAP are flow and rainfall data. This chapter focuses on providing guidelines for
establishing a flow monitoring and rainfall data collection program. Data such as sewer network attributes, pump
stations, treatment plants, and gate operations are also needed for hydraulic modeling using SWMM5 or other sewer
models. The data needed for hydraulic modeling are more intensive than the data needed for the RDII prediction
tools within the SSOAP Toolbox. More explicit descriptions of the hydraulic modeling data needs are provided in the
user manual for the selected hydraulic model, SWMM5 (EPA, 2007).
4.2 Data Requirement for SSO Planning and Analysis
Data collection can be time consuming and expensive and should be carefully defined to meet project objectives and
answer specific questions using the SSOAP Toolbox. When planning a data collection, it is important to understand
what data are needed and how the SSOAP Toolbox can use this data to provide answers.
The SSOAP Toolbox can be applied to various SSO planning and analysis related activities (e.g., confirming
basement flooding reports, determining sewersheds with excessive RDII, and developing SSO elimination plans).
The scale and details of data collection depend on project objectives. For example, a municipality that is conducting a
macro-level study to determine the relative RDII in different sewersheds within the service area does not require the
same level of data collection efforts as one whose primary objective is to conduct a detailed capacity assessment of
the system at the individual sewer segment level. In addition, one must consider the practical constraints, such as
resources and schedule, in the development of a specific data-collection program.
Data used in the Toolbox to support SSO planning and analysis can be divided into two major types:
1. Sewer system data:
a. Hydraulic data (e.g., sewer, manholes, pump stations, and other hydraulic components in the sewer
system)
b. Hydrologic/sewershed data (e.g., sewer service area, sewershed delineation)
2. Sewer flow and rainfall data
The ability to predict RDII is the most significant functionality offered in the SSOAP Toolbox to support SSO
planning and analysis. The Toolbox utilizes flow and rainfall monitoring data to define an empirical relationship
between rainfall and the collection system response. Physical sewer system data are primarily needed for hydraulic
modeling using SWMM5 or other commercially available models, which are also briefly discussed in this chapter.
4.3 Sewer System Data
This subsection provides a brief discussion on collecting adequate sewer system data, such as sewersheds, sewer
networks, pump stations, and treatment plants. Knowledge of the sewer system configuration is imperative for
designing a flow monitoring program, as well as for defining the sewered area which contributes to the RDII
hydrologic response.
4-20
-------
4.3.1 Physical Sewer System Data
Physical sewer system data, such as sewer segment and manhole attributes, system operational data, and sewer
conditions, are needed for hydraulic modeling. The level of detail needed can range from minimal to extensive.
Chapter 6 includes a discussion on which factors should be considered when determining the level of detail needed
for hydraulic modeling. The level of effort depends on the completeness, organization, and quality of the existing
data. In some communities, sewer system information is already available in digital form, such as in a GIS while
others in different formats, such as paper records or CAD.
After the physical sewer system data is collected, data reliability should be assessed. The sources of the sewer system
data need a thorough review and confirmation. It is common for elevation data to change over the period a sewer
system has been constructed because of the use of different vertical reference datum and/or because manhole rim
elevations have been raised. The completeness, organization, and quality of the existing data would help estimate the
assessment effort and allocate the resources for the data collection.
When information is suspect, effort must be made to ensure data accuracy. This may include efforts such as cross-
checking GIS attributes against original as-built drawings, and field investigation. At a minimum, suspect
information should be flagged as a source of uncertainty. In some cases, it may be appropriate to apply an estimate of
sewer attribute data in locations where some degree of uncertainty is acceptable (e.g., ground elevation from
topographic contours may substitute for missing or suspect rim elevation). How suspect information is addressed
depends on the significance of the location in question. For example, a small pipe at the upstream end of the system
is typically less important than a flow split in a trunk sewer. This data collection effort and refinement serve multiple
needs. One is to provide the data required to model the system. Other needs include an accurate and up-to-date
description of the sanitary sewer system and GIS or other readily available accurate data on the existing sewers. A
more explicit description of the physical sewer data required for the development of a hydraulic model can be found
in the user's manual of the selected hydraulic model (e.g., SWMM5 user's manual).
4.3.2 Spatial Sewer System Data
Spatial sewer system data is needed for RDII analysis as well as developing a sewer system model such as using
SWMM5. The service area within a community is typically delineated into sub-service areas (i.e., sewersheds) to
facilitate information organization and to support day-to-day engineering and operational functions. Typically, many
of these communities have GIS- or CAD-based mapping that can be used to develop the service area delineations.
These service area delineations must be collected and reviewed. If necessary, the delineations may need to be refined
to support the RDII analysis and model development. Chapter 6 provides a detailed discussion on the sewer service
area delineation into smaller areas. These delineated small areas are the building blocks of SWMM5 (or other)
hydrologic model.
The SSOAP Toolbox requires actual sewered areas within the sewersheds to accurately determine the RDII volume
(i.e., R-value). In many cases, service area and sewershed delineation may include areas that are not sewered, such as
cemeteries, park land, highway rights-of-way, stream valleys, golf courses, undeveloped areas, or areas on septic
systems. The existing GIS mapping may need adjustments by subtracting the unsewered areas to allow a more
accurate estimate of R-values required for model input. Chapter 6 also provides additional information on adjusting
the collected sewer system spatial information.
In determining the R-value for a metered sewershed, it is important that actual sewered areas are accurately measured.
If the sewered area is underestimated, the R-value will be over predicted. On the other hand, if the sewered area is
over estimated, the R-value will be under predicted. Accurate and consistent definition of sewered areas allows more
meaningful comparison of RDII results. Table 4-1 presents an example that shows the sensitivity of sewered area
estimate used on the R-value prediction. This example shows that R-value estimate is directly proportional to the
sewered area used for the RDII analysis using SSOAP Toolbox. Hence, the sewer system delineation data must be
reviewed thoroughly and adjusted appropriately.
4-21
-------
Table 4-1. Sensitivity of Sewer Service Area on R-value Estimates
Sewershed area (acre)
Rainfall (in.)
Rainfall Volume (MG)
Measured RDII Volume (MG)
R-Value
Estimate 1
100
1
2.7
0.25
9%
Estimate 2
300
1
8.1
0.25
3%
4.4 Sewer Flow and Rainfall Data
This sub-section provides a general guidance on collecting the sewer flow and rainfall data from existing sources and
establishing a focused flow monitoring and rainfall collection program to obtain adequate flow and rainfall data.
Predicting RDII and hydraulic modeling requires highly reliable flow monitoring and rainfall data. Additional
references are available on flow monitoring and rainfall data collection (U.S. EPA, 1999; USEPA, 2005; and
Vallabhaneni et al., 2003).
A sewer flow monitoring program typically involves installing a network of meters within the sewer system for a
specific duration. Data from these meters are used to develop flow characteristics under dry- and wet-weather
conditions at the installed locations.
Rainfall data are usually available from different sources, such as temporary or permanent rain gauge network
maintained by the utility within the service area and the nearest National Weather Service (NWS) station or other
public agency's rain gauge location. These rainfall sources can be supplemented by the calibrated NWS radar
images. The rainfall data support the identification of wet-weather periods to critical input to RDII analysis and
sewer system models.
Many communities have invested in flow and rainfall monitoring to support various sewer system management
functions either on permanent or temporary basis. These data sources should be reviewed and assessed to verify data
adequacy for the RDII and modeling analysis and to determine if additional data is needed.
4.4.1 Flow Monitoring Overview
The flow monitoring program should be designed to meet the broader objectives established for the SSO control
planning and analysis. The most cost-effective way to implement a flow monitoring program is to achieve multiple
objectives with a single properly planned flow monitoring program. In addition to the broader objectives of SSO
control, specific goals established for RDII analysis and sewer system capacity assessment may require refining of the
flow monitoring program. The monitoring program, which can be the most costly undertaking in the overall system
data collection process, must be tailored to the RDII analysis needs (such as spatial resolution of RDII assessment,
number and type of events to monitor, monitoring period, data accuracy) and the modeled system extents (i.e., portion
of the service area sewers included in the model and calibrated). Chapter 6 discusses common capacity assessment
goals and needs for modeling and RDII analysis.
The following discussion presents primary factors that must be considered in establishing flow monitoring programs.
Permanent vs. temporary
A flow monitoring program can be established as temporary, permanent, or a combination thereof. A temporary flow
monitoring program is a "snapshot" of the sewer system over a short duration, typically lasting from weeks to several
months or over a longer term lasting one year or longer. More permanent monitoring might be used for determining
long-term trends in wastewater flows in major sewer systems or to evaluate before-and-after conditions of a portion of
a sewer system for which infrastructure improvements and/or rehabilitation were performed.
4-22
-------
The temporary flow monitoring should be established with consideration of historical rainfall patterns and sewer
system responses so that the probability of obtaining the most beneficial data is improved. Data from multiple events
will improve the ability to assess a range of RDII conditions.
A permanent flow monitoring program offers insight into sewer flow behavior for a wide range of weather and
operational conditions over a long period of time. Permanent metering at strategic locations within the collection
system is an excellent management tool for the municipality in maintaining and operating collection system and
wastewater treatment plants. Permanent metering is often used for flow measurements between jurisdictions and
billing purposes. It also provides multiple benefits, including:
1. Historical trends of the flow patterns and overflows/sewer backups/flooding frequency.
2. A sound basis for seasonal variation of RDII response to support developing an effective SSO mitigation
plan.
3. The effectiveness or affect of system changes, e.g., sewer rehabilitation, capacity improvements, improved
operation practices, and development.
A good flow monitoring strategy must properly select temporary and permanent monitoring locations. Ideal
combinations include: temporary monitoring at relatively large number of locations primarily to provide high
resolution RDII data from contributing sewersheds upstream of trunk sewer system; and relatively small number of
permanent monitoring locations along trunk sewers, upstream of wastewater facilities, and priority sewersheds with
known operational problems. In many instances, experience with temporary monitoring helps refine and enhance the
permanent metering strategy.
Flow monitoring duration
The duration of a temporary flow monitoring program must be long enough to allow a desired range of dry- and wet-
weather flow behavior in the collection system to be determined. The duration will depend on the precipitation
characteristics of the region. Generally, a four- to six-month duration under normal precipitation conditions should be
adequate to provide flow characteristics for a range of dry- and wet-weather conditions to support the RDII analysis
using the SSOAP Toolbox. Where longer-term metering data is available, the uncertainties in the RDII quantification
and prediction will be reduced and reliability improved.
Timing of temporary flow monitoring
The temporary monitoring should be conducted during the time of the year where RDII levels are highest. Historical
flow records at the wastewater treatment plant can help determine the months with highest observed RDII and can
guide establishing the temporary monitoring program. RDII is more pronounced when the groundwater table and
inter-event soil moisture are high. Monitoring in the wet season may increase probability of better characterizing the
RDII impacts on the sewer flows. Collecting adequate dry-weather data to establish the dry-weather flow conditions
is equally important. Beginning the temporary flow monitoring just before the wet season may also help data
collection during periods of dry-weather conditions to establish the base flow conditions adequately. Finally, the
program should include flexibility to shorten or extend the program depending on the storms and flows captured.
Flow monitoring resources
Planning ahead and securing the required resources is a key for a successful flow monitoring program. Once the
general framework for type of monitoring, duration, and the timing for flow monitoring are determined, users must
first secure resources (equipment, staffing, and financing). Required field staff must be available for installation,
preventative and corrective maintenance, on-site data retrieval, and metering removal activities. Office staff required
for regular quality checking of data must be committed to ensure corrective actions are made in a timely manner.
With sufficient early planning, proper equipment can be available during the best monitoring periods. Starting the
resource planning two to four months ahead of the preferred start day for data collection is a good practice. This lead
time allows for securing flow monitoring personnel and equipment, and performing pre-installation activities such as
site selection, equipment installation, and confirming that all the meters are installed correctly before starting the data
collection.
4-23
-------
4.4.2 Flow Monitoring Implementation
Once the general framework for a flow monitoring program is determined, a comprehensive flow monitoring protocol
comprised of the following common elements should be developed:
1. General criteria for flow monitoring
2. Site selection
3. Equipment selection
4. Equipment installation and maintenance
5. Data collection and quality control
In general, there are three phases in implementing a flow monitoring program: mobilization, data collection, and
demobilization. The mobilization phase includes site selection, equipment selection, installation, and calibration.
The data collection phase includes operation and periodic on-site maintenance of the flow monitoring equipment, data
collection and data review/quality control. The demobilization phase includes removing the equipment and preparing
a report documenting the flow monitoring efforts.
Site selection is critical for successful data collection in any monitoring program. Given the usually limited data
obtained during a short-term monitoring program (e.g., a 3 to 4-month period), it is very important to collect high
quality data to provide reliable RDII prediction and the model calibration. Vallabhaneni et al. (2003) presented a
comprehensive case study of a large-scale flow monitoring program in Metropolitan Sewer District of Greater
Cincinnati. They concluded that the rigorous site selection process is imperative to gathering information where it is
most needed and gathering information that is of higher quality. This case study also suggested that developing a site
rating system based on initial flow data quality is useful in making the final site selections and in helping data users in
properly interpreting the data.
The general location for flow monitoring typically focuses on isolating the flow in each major contribution area.
Depending on the specific study circumstances, some may require detailed monitoring of the inflows from upstream
sewersheds and the outflows to the trunk sewers. The consideration factors for determining the general locations may
include:
1. Thorough understanding of the system layout - A good understanding of sewer system features is critical
to identifying the proper locations for flow monitors. These features include active SSOs, flooding reports,
pump stations, pump station overflows, sewer network, and treatment plants.
2. Determination of sewershed discharge points to the trunk sewers - The confluence of major tributary
sewersheds with trunk sewers provides the primary locations for the flow monitors. These locations are
critical for both RDII analysis, and hydraulic model development and calibration.
3. Upstream of known SSOs and flooding locations - Monitors should be located upstream of known SSOs
or flooding locations to allow determination of RDII. In some cases, SSOs and flooding locations may have
tributary areas sufficiently small that monitoring is not required.
4. Pump Stations - The RDII prediction tool in the SSOAP Toolbox requires flow data with minimum flow
fluctuations. Hence, it is a good practice to avoid locations close to the pump stations. However, for
hydraulic modeling purpose, pump stations may need to be monitored if the pump station records and pump
operation are not available at the desired resolution. Each pump station records should be reviewed to
determine if additional flow monitoring is required.
5. Trunk sewers - Flow monitors should be located at critical points along the trunk sewer, including points of
major confluence and upstream of interconnection points between parallel trunk sewers. Meters located in
series along a trunk sewer where flows must be subtracted to compute flows from the intermediate
contributing area should be avoided. The meter error, which can be as great as +/- 20 percent, can surpass the
4-24
-------
flow from the incremental area between the meters. This may result in negative or very large RDII and GWI
flows for the incremental areas. It is best to place meters on major, or minor, inflow points to the trunk sewer
system and avoid meters in series wherever possible.
6. Treatment plants - Flow monitors should be located on all influent lines to the treatment plants.
7. Priority sewersheds - Flow monitors should be located in the high priority sewersheds with known
operational problems.
In addition to the criteria described above, effective use of GIS data, if available, will help identify proper flow
monitoring locations.
The final site selection process includes performing field investigations to review candidate manholes for each desired
location of interest and select appropriate site for meter installations. Typically, several factors must be considered
during these field investigations, including sewer hydraulics, structural conditions of the manhole, access for meter
installation, maintenance, data collection, and other safety concerns such as presence of hazardous gas and vehicle
traffic. Based on these field investigations and from preliminary data collected for up to one week, final site selection
is made. Vallabhaneni et al. (2003) reported detailed procedures for field investigation protocols and the final site
selection.
4.4.3 Equipment Selection
The objective of the equipment selection is to obtain the best quality data at selected sites to meet project goals.
There are many flow monitoring technologies available, such as ultrasonic sensors, pressure sensors, bubbler sensors,
and float sensors. Newer technologies/techniques continue to emerge. Descriptions of the flow monitoring
technologies were included in previous literature, such as the EPA's guidance document: "Combined Sewer Overflow
Guidance For Monitoring and Modeling" (EPA, 1999) and WEF's Manual of Practice: "Prevention and Control of
Sewer System Overflows" (WEF, 1997).
Users should evaluate the hydraulic conditions at selected sites and determine the most suitable flow monitoring
equipment/technology to obtain the best quality data possible. Users should consult with the flow monitoring
manufacturers and/or flow data service providers when selecting the equipment/technology for each specific flow
monitoring site. In some cases, one technology can meet the needs for all locations in a flow monitoring program and
in other cases several technologies may be needed to match the system hydraulics. It is a good practice to balance the
need for quality data with the available technologies and resources.
4.4.4 Equipment Installation and Maintenance
Proper installation and accurate site calibration of equipment is key to collecting accurate data. Users must adhere to
the manufacturer's recommendations and industry standards when installing and calibrating the installed meters. The
meter installation must be performed by qualified personnel. The following general procedures would assure the
proper meter installation and collect information needed for data processing:
1. Record both the measured and monitor-reported water levels.
2. Adjust level setting of monitor as necessary per manufacturer's recommendations.
3. Record position of probe installation with respect to the bottom of the pipe.
4. Record depth of sediment (if present).
5. Measure the distance upstream from the back of probe to the butt of pipe.
6. Take photographs of the completed installation.
It is a good practice to perform a site check three to seven days after initial installation to confirm the meter
performance and site suitability for flow monitoring. Based on the initial site observation after installation, users
should make necessary adjustments for the meter installation and, if required, relocate the meter to an alternate site.
4-25
-------
Proper maintenance procedures are required to assure consistent meter performance throughout the monitoring
duration. These procedures are designed to minimize meter downtime, maintain data quality, and produce reliable
data to support the RDII analysis and model calibration.
The frequency of the maintenance visits are based on metering type (temporary vs. permanent) and study objectives.
Typically, weekly visits are performed in short-term temporary monitoring program. In some cases, the sites are
visited twice in a week when they are concluded to be problematic due to factors such as sediment deposition.
Usually, in a permanent monitoring program, monthly or quarterly site visits are made. It is quite common that the
permanent metering program has a remote data acquisition system, which allows the checking of meter performance
in lieu of frequent site visits. Site visits will include the following activities:
1. Flow data collection.
2. Meter operation check - the real time operating status of a monitor must be checked per manufacturer's
recommendations. Typically, level and velocity readings are checked. In addition, other operating
parameters such as signal strength, temperature, clock, and battery voltage are also checked, depending on the
specific meter technology used. This information should be recorded. Desiccant and batteries should be
replaced as required.
3. Installation inspection - the meter probe and all portions of the equipment installation should be visually
inspected by entering the manhole or with a remote inspection camera. If necessary, meter installation is
adjusted and re-calibrated as needed. It is a good practice to keep a photographic record of the installation
inspection. In the case of permanent meters, if required by the manufacturer, they should be removed and
recalibrated annually.
4.4.5 Data Collection and Quality Control Using SSOAP
To ensure that the flow monitoring program provides quality data, users should perform periodic on-site inspections
of all flow meters as described previously. These site visits can reduce equipment fouling and other unforeseen
conditions, such as clock errors, and will reduce meter down time. Data collection should be recorded in a
sufficiently fine temporal resolution to meet project needs. A common interval is five minutes, but 10 or 15-minutes
and hourly records are also commonly used. Five-minute data can always be averaged to hourly or daily, but not the
other way. In addition to on-site inspections, users should perform weekly QA/QC review of the collected data.
Figure 4-1 displays an example from the SSOAP Toolbox data review utility that can be used to assess the quality of
the collected flow monitoring data. This Data Review Tool graphic consists of three charts: flow hydrograph with
rainfall, flow depth and velocity with rainfall, and depth vs. velocity scatter graph. Users can use the combination of
these charts on a weekly basis to review the measured data quality and prompt the field crew if site checks are needed
to troubleshoot data inconsistencies or missing data. Once the data collection is complete, this data review tool can
assess whether or not data meets the flow monitoring goals. This tool can also determine the portions of the data that
are useful in supporting RDII analysis and model development. The following are general guidelines on performing
data review using the data review utility.
Data gaps
Flow hydrographs can be used to identify any missing data. Flow meters measure depth and velocity at the desired
monitoring location. The flow hydrograph is the product from the measured depth and velocity data. When there is
any missing data shown in the flow hydrograph, users can identify the source of the missing flow data (such as depth,
velocity, or both) using the depth and velocity data plot. Users may be able to use the available depth/velocity data
even if there is missing flow data for certain engineering analysis. In the example shown in Figure 4-1, some velocity
readings drop to zero resulting in drops of flow. However, no noteworthy missing data is observed. The analyst may
ask the field crew to investigate the causes for these occasional velocity reading drops.
DWF patterns
Flow hydrographs, as well as depth and velocity data plots, can be used to review the consistency of DWF data. The
rainfall shown in the flow hydrograph can help users identify the dry-weather periods. Under normal dry conditions,
4-26
-------
users should observe a similar pattern from the measured flow, depth and velocity data. For example, Figure 4-1
shows that under dry-weather conditions, the depth is consistently near 5-inches, the velocity around 1-1.5 ft/s, and
the flow near 0.25 MOD. If a portion of the measured data differs from the normal DWF pattern, analysts should
investigate the reason for the data pattern shift. In some cases, DWF patterns may vary because of failed flow
monitoring equipment, changes of sewer system flow patterns due to operational changes such as temporary
blockages or changes in pump/flow control gate operation in the proximity of the meter. Field investigation and
confirmation may be needed to verify the change of measured data pattern. Depending on the duration of the flow
monitoring period, longer-term evaluation of data consistency may indicate seasonal variations of the average DWF
or monitoring problems from depth sensor's gradual drift.
WWF response
Flow hydrographs and the depth and velocity data plot can also help review the consistency and reliability of the
measured data under wet-weather conditions. Analysts should first observe how a meter responds to the rainfall
events and note the magnitude of peak flows and shape of hydrographs. A pattern should be found between wet-
weather response and the total rainfall. The depth and velocity data plot can also help users gain insight of the sewer
system behavior under various wet-weather conditions. Analysts can establish the hydraulic patterns at a specific
meter site based on depth and velocity relationships and then look for consistent behavior during various rainfall
events. If a portion of the data being reviewed is out of character, the analyst should alert the field crew to
investigate.
Depth vs. velocity scattergraphs
The depth-velocity scattergraph is a useful tool to understand the site hydraulic behavior and to assess data
consistency and reliability. The depth-velocity scattergraphs can be used to determine if the measured data agrees
with the Manning's pipe flow theory. The Manning's theory describes uniform and steady state flows in open
channel and points that the depth of flow in an open channel increases as the velocity increases under free flow
conditions. Once the pipe goes into surcharge conditions because of capacity limitation at the point of flow
measurement or in downstream sewers, this depth to velocity relationship does not hold. The scattergraphs in Figure
4-1 illustrate both depth versus velocity relationships under free and surcharge conditions. In this example the flow
monitor is located inside a 12-inch diameter sewer, which shows when the sewer flow is under the capacity, free flow
conditions prevail and the meter consistently tracks the depth and velocity relationship. When the flow reaches sewer
capacity and cause surcharge conditions (i.e., depth is higher than 12-inch in this example), velocity reduces and
another depth-velocity pattern is developed. A scattergraph with uncharacteristic depth and velocity does not always
reflect bad meter data and may reflect the actual hydraulic condition at the flow monitoring location. Sewers at
capacity constraints, near SSOs, and pump stations will not behave in uniform flow condition, and thus, do not have
the similar scattergraphs under free flow conditions shown in Figure 4-1. Sands and Stevens (1995) offered more
detailed descriptions of the scattergraphs pattern and how it can be used to gain insight of the sewer system.
At the conclusion of the monitoring period, a final report should be prepared summarizing the efforts. It is a good
practice to include the following key elements in the report:
1. Summary of data collection efforts and findings.
2. Map showing flow monitoring locations.
3. Technical summary including descriptions of equipment, Installation notes/records.
4. Calibration and field maintenance records.
5. Hydrographs of five-minute (or other interval) flows and rainfall for the entire monitoring period. In
addition, daily, weekly, and monthly summaries of flow depth, velocity, and rates should be developed.
6. Summary of depth-velocity scattergraphs.
7. Data loss and known data limitations.
A well-designed flow monitoring approach will ensure the collection of quality data to successfully apply the RTK
approach in the SSOAP Toolbox to predict RDII. Additional information on flow monitoring can be found in the
references cited.
4-27
-------
2.5
2.0
1.5
O)
o
» 1.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
3.5
3.0
2.5
£ 2.0
>,
+j
'o
-------
4.5 Rainfall Data
Rainfall monitoring is an integral component in atypical flow monitoring program. The precision, accuracy, and
resolution of rainfall data are critical for RDII analyses and sewer modeling. Inadequate rainfall data introduces
errors in RDII analyses and model calibration, and hence reduces the overall reliability of the SSOAP Toolbox
application.
Many communities have an established permanent rain gauge network, which can be used to support RDII analyses.
Available rain gauge data sources should be evaluated to determine if the supplemental rainfall monitoring is needed.
The following discussion assesses the data adequacy and methods for collecting the required rainfall data.
The accuracy of precipitation data is typically a function of the equipment, its location, and maintenance. Precision is
a function of the type of equipment used. Accuracy and precision are usually less problematic than resolution when
working with the precipitation data. In recent years, it has become widely recognized that RDII prediction, and the
calibration and application of ever-more precise sewer system models, is significantly hindered by the limitations of
precipitation data resolution. These limitations are caused by several factors, including the spatial resolution, poor
gauge sitting, equipment malfunctioning, and data collection/transformation errors.
As a first step, one should determine the performance of the existing rain gauge network to identify the needs for
improvements. These improvements may include increasing the rain gauge density to provide adequate coverage for
the service area, and modify/replace or relocate the existing rain gauges. It is becoming popular to use the radar
technology as a viable means to enhance the spatial coverage of precipitation gauge data, specifically in large
communities that span several square miles of area. When the radar technology is used, it is imperative to field verify
using the data from the installed gauges..
The location and performance of individual rain gauges should be thoroughly reviewed by qualified personnel. The
factors for evaluation include rain gauge spatial distribution, number of gauges and gauge setting. Each site must be
visited and site conditions (such as possible obstructions from trees and buildings, wind effects created by
surrounding buildings) documented that may adversely affect accurate rainfall measurements. Performance testings
should be conducted for each rain gauge to identify any refinement/enhancement necessary to the existing rain gauge
network with respect to gauge sitting; spatial coverage and number of gauges; and other relevant changes that will
support the data accuracy needs. Vallabhaneni et al. (2002b) presented a case study that described historical data
analyses supplemented by various field tests to effectively confirm the rain gauge performance. It is also common
practice to follow the manufacturer's recommendation in installing the rain gauges and periodically conduct
maintenance and calibration of the equipment.
If a service area spans for several square miles and spatial/temporal variation of rainfall is significant, a combination
of aerial estimates of rainfall from radar and point estimates from a rain gauge network will produce a better estimate
of the spatial distribution than either system alone. Vallabhaneni et al. (2003) discussed radar-rainfall integration into
hydrologic and hydraulic modeling and offered step-by-step approaches. Wride et al. (2003) presented a case study
which combines the rain gauges and radar-rainfall technology to improve the characterization of the RDII prediction.
They concluded that using only the nearest rain gauge-based analyses can lead to inappropriate R-values. Analysts
are strongly encouraged to use the combined knowledge of rainfall characteristics and their potential impacts on the
RDII analysis to assess the need for using the radar rainfall technology to supplement the rain gauge network.
In summary, proper flow and rainfall data collection efforts are critical for RDII analyses in sanitary sewer system.
Careful planning of the spatial and temporal extent of this data collection program is a key component of a successful
SSO analysis and planning efforts.
4-29
-------
Chapter 5 Sanitary Sewer Overflow Analysis and Planning Toolbox
5.1 Introduction
This chapter provides an overview of the SSOAP Toolbox, including introductory descriptions of its tools and
functions in performing a sanitary sewer system capacity assessment. It is not intended to serve as the user's manual,
which is included with the software package.
The SSOAP Toolbox is a suite of computer software tools used to predict RDII in sanitary sewer systems and
facilitate capacity analysis. The toolbox includes the option to use SWMM5 for the hydraulic analysis of the sanitary
sewer system. The program architecture allows efficient sewer system capacity analysis and planning by a linking of
various external data sources, RDII analyses tools and SWMM5. The SSOAP Toolbox is programmed using Borland
Delphi® (2006 edition) and operates within the Microsoft Windows® environment.
Figure 5-1 depicts the SSOAP Toolbox organization. There are five tools in the SSOAP Toolbox:
1. Database Management Tool
2. RDII Analysis Tool
3. RDII Hydrograph Generation Tool
4. SSOAP-SWMM 5 Interface Tool
5. SWMM5
The green boxes in Figure 5-1 represent the tools within the SSOAP Toolbox listed above. The grey boxes represent
the external data sources for the Toolbox. In addition, the toolbox allows the use of external software tools to perform
RDII and hydraulic routing analyses. This chapter provides an overview of the tools within the SSOAP Toolbox.
5.2 Database Management Tool
The purpose of the Database Management Tool (DMT) is to help users manage and organize all data required for
performing of RDII analysis and system capacity assessment. DMT also includes useful utilities that help users
efficiently perform quality checks of the external data sources and their analysis.
DMT stores and organizes data in a Microsoft Access® database, referred to as the SSOAP System Database (SSD).
It serves as the command center and transfers data between the tools in the SSOAP Toolbox. It organizes data in the
SSD and serves as a data exchange agent for other tools in the SSOAP Toolbox. As shown in Figure 5-2, the external
data sources may include sewer system GIS databases, data from flow monitoring programs, data from rainfall
monitoring programs or radar rainfall analyses, and hydraulic modeling analysis results. These external data sources
are discussed in greater detail in Section 5.2.1. Since DMT can access SSD directly, the SSOAP users do not require
the installation of Microsoft Access® in their computers.
The utilities in DMT perform: (1) rainfall and flow data quality control; (2) data analysis/queries; and, (3) scenario
management to support SSO analysis and planning. These functions are often performed using various external, non-
integrated tools (e.g., spreadsheets, flow meter software). These tools and capabilities are integrated into the SSOAP
Toolbox.
5-30
-------
Sewer System
GIS Database
Flow Monitoring
Data
Database
Management
Tool
RDII Analysis
Tool
(Unit Hydrograph
Parameters)
RDII
Hydrographs
Generation
Tool
SSOAP-SWMM 5
Interfacing Tool
Figure 5-1. Overview of tools within the SSOAP Toolbox.
5.2.1 Interfacing with External Data Sources
DMT manages the import of data from external databases into SSD and also the export of data stored in SSD. Data
such as flow monitoring, rainfall, and sewer system characteristics (described in Chapter 4) are stored in SSD using
DMT. DMT is specially programmed to support several generic text formats. Users can modify the DMT import
function to customize the imported data format. However, in some cases, DMT may not be able to support particular
proprietary formats used by flow meter manufacturers. In that case, data may require pre-processing to convert data
to standard formats used by the Toolbox prior to importing. In many cases, software provided by flow meter
manufacturers may have the ability to export data to one of the standard text formats used by DMT.
Typical data used in the SSOAP Toolbox and their collection methods have been previously described in Chapter 4.
How these data are stored and processed is presented below. Examples of importing data is described in the separate
user's manual within the SSOAP software.
Flow monitoring data
Users can store a large amount of time-series flow meter data in SSD and manage these data through DMT. Time-
series data for a location, including the date, time, and flow, are the minimum flow-related information for predicting
RDII using SSOAP. Depth and velocity data can also be stored and utilized to support data quality control as to be
described in Section 5.2.3. Users may need to pre-process the flow monitoring data before using DMT to import data
as previously described. If the users already have the RDII parameters (i.e., R, T, and K) determined, flow data will
not be needed for the RDII analysis.
5-31
-------
External Data Sources
Sewer System
Sewer System
GIS Database
Time Seres
Flow
Velocity
Depth
Time Series
Rair-a
Flow Monitoring
Data
SSO Volume
Capture Flow Volume
Overflow Frequency
Flooding Locations
Pipe Capacity
Rainfall
Data
Hydraulic
Analysis
Data
Sewer System
Flow Data
Rair-all Data
Database
Management
Tool
SSOAP
System
Database
in MS-ACCESS
Internal Data Sources
RDII Analysis
Tool
D'WFaralysis resi.Its
Wet-weaker selection results
WWF analysis resu Its
RDM rest Its
E'.'ert base-: RTK pararrefers
RTK predictive analysis result
RTK parameters
Rainfall Data
Ss-A'sr Systsrr
RDII
Hydrograph
Generation
Tool
RDM
SSOAP-SWMM5
Interfacing
Tool
EvVVV S I'C'-tF =
. SWM M 5 In put File with RDII Hydrograph
SWMM 5
Figure 5-2. External data sources.
Rainfall data
Users can store a large amount of time-series rainfall data either from rainfall gauges or from radar rainfall estimates
using SSD and manage these data through DMT. Rainfall data corresponding to the flow monitoring period can be
imported as either precipitation volume or intensity. As with the flow monitoring data, rainfall data must include the
date, time, and location and amount or intensity. Once stored in the SSD, these data are used for RDII analysis and
hydrograph generation. DMT allows custom data formats. As with the flow monitoring data, users may need to pre-
process the rainfall data outside of the SSOAP environment. Examples of data custom formats are provided in the
user's manual.
Hydraulic analysis data
Hydraulic modeling results such as overflow volumes and frequencies, capture flow volumes, flooding locations, and
pipe capacity can be stored in SSD. DMT can be used to directly import a SWMM5 binary output file into SSD.
Users can use the scenario manager function in DMT to analyze the SWMM5 result and to support SSO planning and
analysis.
DMT also provides options to export data stored in SSD to text files for other sewer system management functions
besides capacity assessment. For example, processed flow and rainfall data, and hydraulic simulation results can be
exported from SSD to support routine operation and maintenance of a sewer system.
5-32
-------
5.2.2 Interacting with Other Tools in the SSOAP Toolbox
DMT can also manage data produced by tools within the SSOAP Toolbox, as described below.
RDII Analysis Tool
DMT exchanges information from and to SSD to perform RDII analysis and store the analysis results. DMT uses
information such as flow monitoring, rainfall, and sewer system data from the SSD as input to the RDII Analysis Tool
(described in Section 5.3). The RDII analysis results, such as R,T,K parameter values, RDII event start/end time, and
DWF and GWI estimates are automatically fed back to SSD and stored.
RDII Hydrograph Generation Tool
DMT provides information stored in SSD such as sewer system data, rainfall and RDII analysis results to the RDII
Hydrograph Generation Tool.
5.2.3 DMT Utilities
DMT is equipped with software utilities to perform data quality control, data analysis, and scenario management in
support of sanitary sewer system capacity analysis and planning.
Data quality control utility
This utility has three routines to assist users to perform rainfall and flow data review and assess the data usability for
RDII analysis:
• Rainfall data review - This routine provides a tabular and graphical (i.e., hyetograph) presentation of the
imported rainfall data. In addition, the routine allows a comparison of multiple hyetographs to assess any
data inconsistencies and identify data gaps. It summarizes the rainfall data, such as total volume for the
period of record, and identifies the largest rainfall event in the record, for each rain gauge. These reviews
allow users to assess the overall reliability of the rainfall data before proceeding to the data analysis step.
• Missing flow data analysis - The raw flow data imported into DMT can be reviewed efficiently using this
routine to determine missing data. This routine also helps users filter out the missing flow data from the
record. It also offers an option to fill in the missing flow data values for small time steps by interpolating
them from the recorded values. Note that this method of determining missing flow data for a prolonged
period can be unreliable and users should exercise caution.
• Flow data review - This routine offers users a flow data review capability by generating plots from flow data
stored in DMT, and data quality and reliability assessment as discussed in Chapter 4. The relationships
between measured depth and velocity scatterplots can be generated for selected periods from flow data
records. In addition, similar plots can be generated for review of inter-relationships among depth and flow
values. Depth and velocity scatterplots can then be used to determine data consistencies and reliability. The
scatterplots also help develop knowledge of the sewer system hydraulic behavior at the flow metering
locations. The assessment of this flow data review capability within SSOAP offers users an efficient data
re view tool.
Rainfall data analysis utility
DMT offers this utility for analyzing the imported rainfall data to identify wet-weather events of interest and perform
analysis to determine the rainfall characteristic such as the rainfall event start time, rainfall duration, rainfall volume,
peak rainfall intensity, and number of dry days prior to each rainfall event. Table 5-1 shows an example of the
analysis result. It lists all rainfall events that occurred with relevant characteristics of each event.
5-33
-------
Table 5-1. Example of Rainfall Data Analysis Results
Event No.
1
2
3
4
5
Start Date/Time
7-9/2002/ 0630
7-9/2002/1705
7-17/2002/1045
7-19/2002/1835
7-29/2002/1410
Total Volume
(inches)
4.6
0.44
0.24
0.33
0.17
Duration
(hours)
4.6
0.9
3.8
1.9
3.8
Peak Intensity
(inches/hour)
0.3
0.4
0.2
0.3
0.2
# of Dry Days
Prior to Event
11.0
0.3
7.7
2.2
9.3
Scenario management utility
The Scenario Management Utility in DMT offers users the ability to organize the sewer system flow routing results
from SWMM5. Users can store the modeling results from various model scenarios and apply this utility to compare
scenarios to support sewer system evaluations under various precipitation conditions and system configurations.
Figure 5-3 provides an example of scenario comparison.
In summary, SSD is a data bank for storing data from flow monitoring, rainfall monitoring, RDII analysis, and
SWMM5 modeling. DMT manages data input to and retrieval from SSD. DMT also provides simple utilities that
facilitate data review and analysis, and scenario comparisons. Because data is stored in a Microsoft Access database
file, users can develop additional custom utilities using Access or other programs to perform more in-depth analysis to
meet project-specific needs, without interfering with the core computational routines of SSOAP.
Scenario
Database
Management
results
SSOAP Scenario Manaaement
Scenario
A
B
C
SSO Volume
17MG
1MG
OMG
OF Frequency
9
1
0
Capacity of Pipe A
1 00%
100%
70%
Figure 5-3. Scenario management.
5.3 RDII Analysis Tool
The RDII Analysis Tool implements the RDII unit hydrograph methodology described in Chapter 2. This tool has the
ability to support four major analyses:
• DWF analysis
• Wet-weather flow (WWF) analysis
• RTK unit hydrograph curve fitting analysis
• RTK unit hydrograph parameter predictive analysis for unmetered and design precipitation conditions
These analyses are described in the following sections.
5-34
-------
5.3.1 DWF Analysis
The RDII Analysis Tool assist users in performing DWF analyses based on flow monitoring and rainfall data. The
RDII Analysis Tool helps users select dry-weather days automatically based on flow data and filter out the days with
influence of rainfall events. This tool also allows users to automatically select dry-weather days in a graphical
environment, and manually add to or remove from the automatically selected days. The RDII Analysis Tool
calculates the average DWF from the selected dry-weather days and the average DWF for weekdays and weekend
days. Figure 5-4 depicts the results of a DWF analysis indicating weekday and weekend DWF for a sample
sewershed.
fl? Dry Weather Hydrographs
Graph Print Edit Help
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hours
Figure 5-4. DWF hydrograph derived from RDII analysis tool.
5.3.1.1 BWF and DWF Adjustment
The RDII Analysis Tool can address seasonal variations in DWF by accounting for the influence of GWI. The RDII
Analysis Tool calculates the difference of the average flow during the given day and the average flow of all DWF
days in the period of record. This difference is defined as DWF adjustment in SSOAP Toolbox. The DWF
adjustment can be a positive or negative value. This adjustment allows users to set the proper DWF conditions prior
to rainfall events to determine the rainfall event specific RDII hydrograph. The user's manual contains an expanded
discussion and graphical presentation on the event-specific DWF adjustment.
The RDII Analysis Tool can also decompose DWF into two components: BWF and GWI. For example, when users
enter the percentage of the DWF as GWI, the computational routines within the Tool will divide the DWF hydrograph
into BWF and GWI.
5.3.2 WWFAnalysis
The RDII Analysis Tool offers users various automatic functions to analyze RDII flows. These functions are
5-35
-------
designed to:
• Identify individual RDII events based on flow monitoring data and rainfall data. The RDII event is defined as
the time period during the flow pattern that varies from the DWF because of the influence of RDII. It starts
when the rainfall begins and ends when the flow pattern returns back to the pre-rainfall level.
• Summarize the RDII event related data, including:
o Rainfall (i.e., volume, duration, peak, start time, end time)
o Flow data (i.e., peak total flow, peak RDII flow, RDII Volume)
o Percentage of rainfall entering sewer system (i.e., R-value)
The RDII Analysis Tool determines RDII event start and end times by analyzing the RDII flow hydrograph and
corresponding rainfall records. Users can also manually define a RDII event or adjust the event automatically. A
RDII event is characterized by the peak rate, duration, and volume. These characteristics are needed because users
can define threshold limits of RDII events based on a minimum rainfall depth and a minimum peak RDII flow rate.
Users can then review and refine the RDII events selected by the tool and use these events for further analyses,
including hydrograph decomposition and unit hydrograph curve fitting.
5.3.3 Hydrograph Decomposition and Unit Hydrograph Curve Fitting Analysis
The RDII Analysis Tool uses the RTK unit hydrograph methodology described in Chapter 2 to determine the
relationship between rainfall and RDII for each metered sewershed. Parameters in a unit hydrograph are developed
through a systematic analysis of measured flow and rainfall. Once developed, these unit hydrograph parameters and
rainfall hyetographs are used to define RDII hydrographs for collection system modeling and evaluation using
SWMM5 or other hydraulic modeling tools.
The RDII hydrograph of a rainfall event must be first derived using hydrograph decomposition procedures included in
the RDII Analysis Tools. Hydrograph decomposition considers a range of parameters, including rainfall depths,
sewershed area, antecedent moisture conditions, and groundwater elevations to quantify the individual wastewater
flow components in the system.
The RDII Analysis Tool helps users decompose the monitored flow hydrograph automatically and graphically.
Figure 5-5 depicts decomposition of the observed flow into DWF and RDII components during a rainfall event. The
RDII Analysis Toolbox allows users to adjust the pre-event DWF to change the GWI on a given day compared with
the average DWF for the period of record. The RDII flow is the remaining flow after subtraction of adjusted DWF
from the observed flow.
After RDII flow hydrograph is defined, the Tool is used to calculate the total R-value, using the rainfall data and
sewershed area information.
Users can further decompose the RDII flow using the unit hydrograph approach described in Chapter 2. For each unit
rainfall input the RDII Analysis Tool generates unit hydrographs for the corresponding sets of R, T, K, with values
defined by the user. Figure 2-4 in Chapter 2 illustrates how the tool generates three unit hydrographs based on the R,
T, K parameters for a given unit rainfall input. It also demonstrates that the total RDII unit hydrograph is the
summation of three individual unit hydrographs. The three unit hydrographs can be related with fast (first unit
hydrograph), medium (second unit hydrograph), and slow (third unit hydrograph) RDII responses typically observed
in the sanitary sewer system. In some cases, only one or two unit hydrographs are required to adequately define
observed RDII hydrographs.
5-36
-------
ROM Graph
firaph Print E* View Help
Evsrt 1
Start D«le 2-9-2001 16:55
•nd Dale 2-12-2001 11 35
: • / -r, -> - •
| DWF Adjustment
Figure 5-5. Hydrograph decomposition in the RDII Analysis Tool.
R, T, K parameters can be determined by graphically comparing the total RDII hydrographs generated by users'
defined R, T, K parameters with the RDII hydrographs from the monitored data. This visual curve fitting is
accomplished by iterations. The iteration process continues until a good visual comparison between the calculated
RDII hydrograph and the observed RDII hydrograph is achieved. Numerical comparison of total RDII volume with
the sum of volume under each of the unit hydrographs will confirm the success of curve fitting. This simple,
interactive and visual approach will facilitate determination of the unit hydrograph parameters rather than relying on
complicated numerical techniques.
Figure 5-6 shows an example of the hydrograph curve fitting using three unit hydrographs. This analysis can lend
understanding of the fast, medium, and slow RDII response in the sewer system tributary to the point of metering.
The following general guidelines should be followed in selective the R, T, K parameters to ensure that the calculated
RDII hydrograph meets the goal of visual curve fittings:
• Total R value = RI + R2 + RS, if all three unit hydrographs used.
• The T and K parameters should be similar for rainfall events for a given sewershed tributary to the flow
monitor since they depend on the geometry and sewer system layout.
• In all cases, TI < T2< T3
• In most cases, KI < K2 < K3
• The necessity to change T and K significantly for a particular event to match the observed flows is often a
sign that the rainfall data being used is not representative of the rainfall that fell over the basin for the event or
the system experienced operational challenges resulting in an altered shape of the hydrograph.
5-37
-------
• The event specific R-values will vary, generally being higher for wet antecedent moisture conditions and
lower for dryer antecedent conditions. Similarly, R-values will typically be higher in a wet season.
• T and K for the three triangular unit hydrograph should generally be within the ranges shown in Table 5-2.
•- RDM Graph
graph Pii* Edit View Help
Evert I
Start tftte 2-9-2001 16:55
•rid Cute 2-12-2001 11:35
Duraum: 66.67 rtrs
Calculated RDII
(yellow), curve fitted
with observed RDII
Flow (red)
Figure 5-6. Unit hydrographs curve fitting using the RDII Analysis Tool.
Table 5-2. Ranges of Values for Unit Hydrograph Parameters
Curve
1
2
3
T (hours)
0.5-2
3-5
5-10
K
1-2
2-3
3-7
Once the R, T, K parameters are defined, the RDII Hydrograph Generation Tools would be used to generate RDII
hydrographs for selected events from monitoring data, desired design storm events, or continuance multiple events.
5-38
-------
5.3.4 Statistical Analysis of RDII Parameters
The R, T, K parameters developed using the RDII Analysis Tool are specific to a rainfall event and observed flow
data. These parameters are used to develop the RDII hydrographs for selected monitored events to support calibration
of hydraulics of the sewer system model.
When the calibrated hydraulic model is used to perform hydraulic capacity assessment for a non-monitored condition
or a design storm, one can estimate the R, T, K parameters condition using statistical methods. The sophistication of
these statistical methods varies depending on the quantity of RDII data available. For example, a short-term
temporary monitoring program that spans over four months may produce a limited amount of suitable events for RDII
analysis compared to a long-term (>1 year) temporary or permanent monitoring. Typically a longer-term monitoring
provides more events (samples) to do more sophisticated analysis such as multi-variable regression as compared to a
smaller sample set from a short-term monitoring. Vallabhaneni et al. (2003) presented a case study that applied a
multi-variable regression method to predict RDII for non-monitored conditions. In addition, long-term monitoring
will allow for the development of the seasonal RTK characteristics for the monitored sewershed. A published case
study (Loehlein et al, 2004) has a more detailed description on using monthly varied R, T, and K parameters for a
RDII study.
The RDII Analysis Tool includes basic statistical analysis functions for developing the correlations between observed
precipitation conditions and system RDII responses. The correlation relates RDII responses with several factors such
as rainfall characteristics (i.e., depth and intensity) and antecedent moisture conditions. The basic statistical methods
included can support application where only limited data is available. It is important that the effects of antecedent
moisture conditions are taken into account in deriving the ultimate R,T,K values. When adequate RDII data is
available from long-term monitoring, advanced statistical methods such as multi-variable regression can be applied
using commercially available statistical packages.
The RDII Analysis Tool is coded to perform the following three basic statistical analyses:
• Median R-value method
• Average R-value method
• Linear regression of R-value method
The RDII predictive analysis provides better results when long-term data are available. The accuracy and
applicability of the predicted R,T,K parameters depend on the statistically significant number of wet-weather events
available for analyses, the length of the flow monitoring period, the quality of the rainfall and flow data, and the
nature of soil moisture conditions that prevailed during the monitoring period.
5.3.4.1 Median R-Value Method
As the name describes, this method selects the median total R-value from all analyzed wet-weather events at a
selected flow meter as the representive RDII response. The median total R-value is then broken down into RI, R2, and
R3 parameters. The general procedures are:
• Computing the RI, R2, and R3 distribution of all wet-weather events
• Averaging the distributions of all wet-weather events
• Applying the average distribution to the median total R values
Table 5-3 depicts the determination of median R-value, and the distribution of R values using the "median R-value
method."
The tool also determines the remaining RDII parameters (i.e., T and K) by averaging their values for all wet-weather
events at a selected flow meter (Table 5-4).
5-39
-------
Table 5-3. Example of "Median R-value Method" Application
RDII Analysis Results
Ri
1%
3%
2%
R2
3%
6%
4%
R3
6%
11%
7%
Median Total R-value
Total R
10%
20%
13%
13%
Average R-value Distribution
Representive RDII Response
13%
Distribution of
Ri
10%
15%
15.4%
13.5%
1.8%
R2
30%
30%
30.8%
30.3%
3.9%
R3
60%
55%
53.8%
56.3%
7.3%
Table 5-4. Example of RDII Analysis Results
Event 1
Event 2
Event 3
Average
TX
1.5
2
1.3
1.6
T2
4
4.3
3.9
4.1
T3
7
7.5
6.7
7.1
K!
1
2
1
1.3
K2
2
3.2
2.3
2.5
K3
3
4
3
3.3
It is important to recognize the limitations of applying median values of R, T, K parameters for a sewershed
considering wide variation of RDII responses due to varying rainfall and antecedent moisture conditions. For
example, if there were only three rainfall events during a temporary flow monitoring period under a much below-
normal precipitation condition, users should be cautious when using any kind of statistical approach discussed in this
chapter to determine a statistically representative R-value distribution.
5.3.4.2 Average R-Values Method
Instead of using the median value to determine total R value, the method simply takes the average of all available R-
values. The distributions of total R, T, and K values are determined using the same procedures as the median R-value
method. Limitations should be recognized in applying the method to a sewershed.
5.3.4.3 Linear Regression Method
This method develops a linear relationship between total R value and total event rainfall depth or other system
variables such as rainfall intensity and volume of events occurred a week prior to the currently analyzed wet-weather
event. Figure 5-7 shows an example of the linear regression results. The results show that as the total rainfall volume
increases, the total R value increases but with a wide variation. The tool determines the correlation using the linear
regression method and provides the best fit relationship between total R-value and the total rainfall volume.
Users must be cautious when applying linear regression to predict RDII responses based on limited observed data or
when there is a poor correlation of R value with a single variable such as total event rainfall depth. RDII is not
dependent on the total rainfall depth alone. Other factors, such as GWI, antecedent moisture, rainfall intensity, and
season, may contribute to the variability in the RDII response. As mentioned earlier, users may consider applying
multi-variable regression methods available outside of SSOAP to determine the R,T,K parameters for unmetered and
design conditions. The SSOAP user's manual provides guidance on how to apply multi-variable regression methods
for a determination of R, T, K parameters.
5-40
-------
*r Linear Regression Analysis
Options
Total R Value vs. Total Rainfall Volume
2.0
0.1
0.2 0.3
Total Rvalues
0.4
Figure 5-1. Example of a linear regression analysis.
5.4 RDII Hydrograph Generation Tool
This tool generates RDII hydrographs using R,T,K parameters determined by the RDII Analysis Tool using rainfall
and sewershed area stored in SSD. To use the tool, users may select:
• Rainfall-event-specific R, T, K parameters from SSD, or
• R, T, K parameters from statistical analysis in RDII Analysis Tool, or
• Parameters derived from other analyses external to the SSOAP Toolbox.
This tool can export RDII hydrograph directly to an external file in SWMM5 or text file. Figure 5-8 shows the user
interface of this tool:
5-41
-------
|North_Gauge_2002_15
Events
|4/20/2002 2:30:00 PM to 4/30/2002 3:45:00 PM
Start Date/Time: | 4/20/2002 -j- | 2:30:00 PM -;
End Date/Time: | 4/30/2002 -j- \ 3:45:00 PM -r-
Figure 5-8. User interface of the RDII Analysis Tool.
Within this tool, users can view the RDII hydrograph before exporting the data using the following options:
1. SWMM5 RDII Interface File - A SWMM5 inflow interface file containing RDII and DWF hydrographs.
2. Time-series file in SWMM 5 format - A text file with RDII hydrograph in SWMM 5 [TIMESERIES] format.
3. SWMM5 [RDII] and [Hydrograph] - A portion of SWMM5 input file with the RTK unit hydrograph
parameters and sewer system information. The RDII hydrograph will be generated within SWMM5 based on
the R,T,K input from the SSOAP Toolbox.
4. Time-series file - A time-series hydrograph in plain text. This format allows users to apply the RDII
hydrographs to other hydraulic simulation models and other sewer system management analyses.
After generating the RDII hydrograph file using this tool, users can: (1) go to SWMM5 to arrange these hydrograph
files manually before performing sewer routing; or, (2) use the SSOAP-SWMM5 Interface Tool to organize these
hydrograph files within SSOAP.
5.5 SSOAP-SWMM5 Interface Tool
The SSOAP-SWMM5 Interface Tool performs three primary functions:
1. Incorporates the hydrographs generated by the RDII Hydrograph Generation Tool into SWMM 5 input files
(SWMM5 pre-processing).
2. Performs SWMM5 simulation.
3. Delivers SWMM5 model simulation results to DMT where it organizes the model results in SSD (SWMM5
post-processing).
5-42
-------
5.5.1 Pre-processing RDII Hydrographs
The Interface Tool allows users to integrate RDII hydrographs into a SWMM5 input file. The Tool first prompts
users to specify the RDII hydrographs and corresponding flow loading locations within the model network and the
location of a SWMM5 input file on the computer local hard drive or a network drive. Then the Interface Tool
accesses the selected SWMM5 input file for integration of the RDII hydrographs at appropriate locations within the
model network.
5.5.2 SWMM5 Simulation
The Interface Tool allows users to run SWMM5 within SSOAP for performing sewer hydraulic routing simulations.
This is accomplished by selecting the SWMM5 engine from the SSOAP Toolbox window, and then clicking the
appropriate menu option to start the model simulation. The actual SWMM5 model development and improvement
are not part of the SSOAP Toolbox development effort. The SWMM5 user manual is the primary reference for the
model. Chapter 6 outlines the model and calibration steps and discusses sewer system capacity assessments using
model simulations.
5.5.3 Post-processing Model Results
After the SWMM5 model simulation is successfully completed and checked, users then can use the SSOAP-SWMM5
Interface Tool to import the model results into SSD. This interface tool can read a SWMM5 binary output file and
extract pertinent data (e.g., flooding/overflow locations) to assess sewer system capacity. DMT will organize these
data through scenario management functionalities previously described.
In summary, the SSOAP Toolbox includes a suite of useful tools to assist users in analyzing monitored data for
predicting RDII in sanitary sewer systems and in running dynamic flow routing of sewers for the development of
needed information that facilitates capacity analysis and planning. Users must have a good understanding of their
sewer systems in order to make decisions with the help of the SSOAP Toolbox.
5-43
-------
Chapter 6 Sewer System Model Development and Capacity Assessment
6.1 Introduction
This chapter describes the use of SWMM5 within the SSOAP Toolbox environment to perform a sanitary sewer
system capacity assessment. References to the SWMM5 user's manual (EPA, 2007) are included throughout the
chapter for users needing detailed assistance with model network development, input and output processing, and
model execution.
General guidelines are provided on model development and application, and the steps typically used in establishing
baseline conditions and performing capacity analysis. The specifics that can be applied to any study depend on many
factors, including project objectives, data availability, funding, schedule, and needs and expectations of the
stakeholders.
Before discussing the model development and capacity assessment guidelines, it is useful to reiterate the steps
required to perform a capacity assessment for a typical sanitary system. These steps are shown in Figure 6-1 and
summarized.
1. Define capacity assessment objectives
2. Collect data and review
a. Review sewer system data/information and develop a data gap analysis to identify data needs to meet the
capacity assessment objectives.
b. Conduct field surveys/investigations to collect sewer system attribute data.
c. Collect rainfall and flow monitoring data.
3. Develop RDII parameters and hydrographs using the SSOAP Toolbox
a. Perform rainfall and flow data analysis using the Toolbox to derive RDII parameters and hydrographs.
4. Develop, calibrate, and verify model
a. Develop a sewer system representation for SWMM5.
b. Develop DWF model input and perform DWF model calibration.
c. Select model calibration and verification rainfall events from the flow and rainfall monitoring data.
d. Perform WWF model calibration and verification for the selected rainfall events.
5. Perform capacity assessment of the existing system
a. Apply the calibrated model under various DWF and WWF assessment conditions.
b. Perform assessment based on model simulation results to identify baseline sewer system capacity
problems and develop SSO characteristics (frequency, duration, and volume).
Chapters 1 through 5 of this technical report addressed many aspects related to steps 1 through 3 listed above. Model
development is described in Section 6.2. Capacity assessment approaches are described in Section 6.3.
6-44
-------
Stepl
Step 2
StepS
Step 4
StepS
Define Capacity
Assessment
Objectives
Collect Data
and Review
Develop RDM
Parameters, and
Hydrographs
in SSOAP Toolbox
Model
Development
Capacity
Assessment
Figure 6-1. Capacity assessment steps for a typical sanitary sewer system.
6.2 Sewer Model Development
A properly developed hydrologic and hydraulic computer model can provide an effective means of evaluating the
hydraulic capacity of a sanitary sewer system under DWF and WWF conditions. Model development includes data
collection (discussed in Chapter 4), RDII analysis (discussed in Chapter 5), model input development, and model
calibration and verification.
Figure 6-2 presents the breakdown of Step 4 of the sewer system capacity assessment steps discussed in Section 6.1.
6-45
-------
This figure shows the sequence of activities required to develop model input data, calibrate, and verify a model based
on RDII parameters estimated using the SSOAP Toolbox. As illustrated in Figure 6-2, Step 4 contains two
components: (1) model input development; and (2) model calibration and verification.
Q.
0
Model Input
Development
Determine
Model Extent
Collect Sewer
System Configuration
and Attribute Data
Develop Sewer
Network
Hydraulic Model Input
Develop DWF
Model Input
Develop RDII
Model Input
Perform DWF
Model Calibration
Perform WWF
Model Calibration
Model
Calibration &
Verification
Perform Model
Verification
Figure 6-2. Model development, calibration, and verification.
6.2.1 Model Input Development
The goal of model input development is to develop an adequate model in SWMM5 that can fulfill the pre-determined
objectives of the capacity assessments (Step 1). The upper part of Figure 6-2 depicts five key procedures in
6-46
-------
developing a model input for a sanitary sewer network:
• Determine model extent - Section 6.2.1.1 discusses how to determine the extent of a sanitary sewer network
to fulfill the objectives of the capacity assessment.
• Collect sewer system configuration and attribute data - Section 6.2.1.2 describes the collection of sanitary
sewer system characteristics and configuration for model development. Chapter 4 provides more in-depth
guidelines and discussion on this topic.
• Develop sanitary sewer network model input - Section 6.2.1.3 outlines how to develop a sanitary sewer
network model input in SWMM5 using the collected sewer system configuration and attribute data. It
provides guidance on how to assure more accurate representation of a system's configuration. This section
also illustrates ways to identify data inconsistencies and provides examples of data confirmation needs.
• Develop RDII model input - Section 6.2.1.4 discusses the hydrologic aspects of a sanitary sewer network.
It provides guidance and examples on how to divide the study sanitary sewer service area into sewersheds and
sub-sewersheds based on flow monitoring, RDII responses, and points of contribution to the sewer network.
• Develop DWF model input - Section 6.2.1.5 presents the two components of DWF: BWF and GWI, as well
as suggestions on how to develop DWF data for a study area. In areas with significant projected growth,
estimating future DWF is necessary to evaluate the impact of increases in wastewater flows on sewer system
capacity.
6.2.1.1 Determine Model Network Extent
The initial step in developing model inputs is to determine the extent of the sanitary sewer network to be modeled.
"Model network" refers to the hydraulic representation of a sewer system. Once the project objectives for the study
area are defined, one should determine the level of details to be included in the model network extent. The level of
details in the model is no longer, as in the past, limited by computer speed and memory. It is now more influenced by
the following factors.
1. Capacity assessment objective
2. Size of the sanitary sewer system
3. Model boundary conditions
4. Availability of reliable sewer system data and existing flow monitoring program
5. Funding and schedule
A model network is often selected based on the previously mentioned criteria, but there are cases where some areas of
the collection system are modeled in greater detail than other areas due to specific project objectives. For example, if
the partial objective is to address customer flooding problems, the model may include a more explicit representation
of the sewer system that directly receives the flow from the house or commercial property laterals. The following
describes each of the above factors in more details:
Capacity assessment objectives
The objective of the capacity assessment is an important factor when determining the extent of a model network. One
should determine what questions need to be answered. For example: what is the hydraulic performance of the entire
sanitary system and the occurrence of SSOs? What are the hydraulic constraints in trunk sewers system? Which
particular sewer reach has capacity constraints? What causes the sewer backups into basements? Does the existing
system have adequate capacity for the anticipated future growth? What is the sewer performance during the dry-
weather conditions?
After the study's objective is defined, it is a common practice to select threshold pipe sizes to include in the study. In
general, the greater the extent of the model network, the greater its data needs will be and the greater the funds needed
to develop the model. Typically, the location of SSOs in a sewershed will determine the upstream model extent. It is
6-47
-------
recommended that the model extents include sewers known to have SSO baseline characterizations of the system and
to develop improvements to address the capacity limits. It must be emphasized that the smaller sized pipes (i.e., 8-
and 10-in. -diameter) typically found in the upstream reaches of the system often represent the majority of the sanitary
sewer system. Many of these 8- and 10-in. diameter sewers may have more than sufficient capacity and thus do not
typically need to be included in the model network as a general rule.
The capacity assessment study often includes sewer performance assessments and consideration of future growth
conditions that cause increases in baseflow. In this case, users should consider appropriate model extents within the
existing system where future growth is projected to occur.
Collection system size
System sizes and complexity should be considered when determining the model extents. Large and complex systems
may need a more focused approach to optimize modeling efforts and resource needs. Phased model development is
one way to develop models for larger, more complex systems. In some cases, it may be appropriate to build a
"coarse" model, initially, and then refine the model within areas of concern. In other cases, it may be appropriate to
subdivide the system study area into separate smaller models that can be constructed sequentially or in parallel by
multiple model development teams.
Typically, in large collection systems, the model extents do not include all sewers. The cost of developing a system-
wide model increases when smaller diameter pipes are included, and can increase greatly if an electronic database of
high quality attributes of the system is not readily available. A typical standard option in modeling the complete
system is to include 8- and 10-in. diameter sewers in the modeling network, only in areas of known capacity related
problems or otherwise needed to connect the system. On the other hand, the capacity analysis may focus only on
trunk sewers. In this case, only larger pipe sizes (15, 18, or 21- in. and larger) are included, which could greatly
reduce the modeling effort.
Model boundary conditions
Both upstream and downstream flow boundary conditions should be considered in determining the model extents.
Effects from downstream capacity constrictions may impact the evaluations in upstream segments. The model should
extend far enough below the area of interest so that the effect of downstream conditions is minimized. Alternatively,
the model should terminate at a pump station or other locations where the downstream boundary conditions are
known for all simulated conditions. The availability and knowledge of boundary conditions often dictate when and
where a modeled sewer system can be divided into multiple models.
Availability of reliable sewer system attribute data and existing flow monitoring program
The amount of reliable system data that is readily available can influence the determination of the extent of the
network model. Analysts should consider the magnitude of field investigation and survey work needed when
determining the model extent. There may be cases where high quality attribute data exist in an electronic format that
may allow inclusion of more sewer network than would otherwise be considered if attributes must be obtained from
hard-copy drawings, field investigation, and surveys.
If the existing data from permanent or temporary meters are used for RDII analysis and model calibration, then the
model network should include the locations where these meters are. For example, if the meters are located on a major
trunk sewer, the model extent becomes coarse and is limited to the trunk sewers. The model network can be
expanded upstream of the trunk sewers when flow monitoring data can be collected from the contributing sewershed
areas along the trunk sewers. Alternatively, if the detailed existing flow data is available on both trunk sewers and the
sewer network within the sewershed areas, the model network can be extended.
Funding and schedule realities
The available funds to perform a collection system capacity assessment study can significantly influence the extent of
model development. In some cases, schedule constraints may determine model extents. The more extensive a model,
the greater the need for field surveys and flow monitoring effort. In some cases where a high quality of sewer
6-48
-------
attribute database is available, the needs of field surveys will be significantly reduced. These field activities are time
consuming and cost intensive. Again, phasing the modeling efforts should be considered in large and complex
systems. The objectives for the capacity assessment should be balanced with funding and schedule realities.
6.2.1.2 Collect System Configuration and Attribute Data
Once the extent of the sewer system model has been determined, the system configuration and attribute data must be
collected from existing sources. Depending on completeness and quality of the existing data, field investigation and
survey efforts may be required. In addition, rainfall and flow monitoring efforts may be required to supplement
existing data. Chapter 4 describes typical data collection efforts in support of SSO planning and analysis.
6.2.1.3 Develop Sanitary Sewer Model Network
The sanitary sewer system attributes, including connectivity, are compiled as input to SWMM5. The SWMM5 user
manual and software tutorial provide step-by-step instructions for developing a model.
In SWMM5, a user can represent the sanitary sewer network in a graphical environment by dragging and dropping the
pre-defined sewer system components (e.g., pipes, manholes, pump stations, and weirs) in the software's Study Area
Map window. The user can then manually assign attributes to each component with real or approximate coordinates
and define the connectivity of the sewer system components. A background image of the service area can be
imported to guide the setup of the sewer network.
In addition to SWMMS's network generation capabilities, software developers have created commercial interface
products that provide tools to transfer existing sanitary sewer system data needed from a GIS/CAD package directly
to the SWMM5 input file format.
Perform quality control of network data
The accuracy of the physical sewer system representation in the model directly impacts the accuracy of the model
results. After sewer system attributes and configuration data for SWMM5 are prepared, quality control procedures
are essential to assure that the system configuration is accurately represented and identify data inconsistencies and
data confirmation needs. Identifying data gaps and resolution early will minimize adverse schedule and budget
impacts while assuring the overall quality of the resulting model.
A typical quality check of a model network includes a sewer plan and profile from SWMM5 and a detailed review of
other parameters not readily plotted such as pipe roughness, weir configuration, and pump operational settings. The
plan views of the model are used to detect any gaps in connectivity data as shown in Figure 6-3.
The profile views help identify questionable sewer system attribute data (e.g., pipe invert, manhole invert and rim
elevations) or data gaps as shown in Figure 6-4.
A review of other parameters not readily reviewed in plan or profile should be conducted by querying input data for
values that fall outside a reasonable range. For example, the pipe roughness parameter can be sorted and reviewed to
identify abnormally high or low values to investigate further if necessary.
When data problems are found, the modeler should first review the GIS and other data sources and attempt to resolve
any data transfer problems. If the problem is not related to data transfer, then the problems should be resolved via
additional intensive data collection efforts such as map/drawing collection, field survey, and field investigation. In
some cases, it may be appropriate to fill in data based on engineering judgment. Missing manhole rim elevations may
be estimated by topographic data, if the pipe diameter is inaccurate or missing, the diameter can usually be estimated
from the diameters of the upstream and downstream sewers. Records should be kept where data are assumed. If
these sections are later determined to be critical for sewer system capacity evaluation, additional field collection
efforts should be conducted.
6-49
-------
SWUU 5 . DWF _115_1 2BJMJ:
Efc 6* WM £rt.«t
nwyct
**•
T Connectivity Issue
Plan View Check
Figure 6-3. Model development plan view connectivity data check.
r^ Pi of ii
:Profile -Node 225043 - 220091
Water Elevation Profile: Node 225043 - 220091
71*} Down stream -<-
7H--
712
g 710-
I 703-
796
704-
70?-
700
698
^Ms_sjngMianhote Depth
Upstream
1
Check Manhole frtverr
800
700
600
500 400
Distance Cft)
300
200
100
Figure 6-4. Model development sewer profile check.
6-50
-------
For large complex models, additional quality control checks may be justified. For example, pipes with adverse
(negative) slopes should be identified and verified. Similarly, locations where a larger diameter sewer enters a
smaller diameter sewer should be reviewed, as this often identifies inaccurate model data. Other checks can be
performed, such as where there are large drops at manholes and where the outgoing sewer invert of a manhole is
higher than the influent sewers. Many of the commercially available modeling software include tools to perform
these checks automatically.
Other system features may also need to be modeled in addition to the manholes and sewers. These include:
• Wastewater pumping stations
• Control structures (e.g., weirs, dams or gates used to control the splitting of flows in downstream sewers)
Once the network connectivity is confirmed using SWMM5 plan and profile views and general reasonableness of
parameter values, additional quality control checks can be made to assure the flow routing through the model network
makes sense in general terms. This is accomplished by performing a model simulation with assumed flow inputs at
the upstream flow loading manholes within the model network. If the model simulation is completed successfully and
flow balance is preserved, and hydrodynamic profiles in SWMM5 are appropriate, then the model network is ready
for the next steps in model development.
6.2.1.4 Develop Sewershed Delineations
The first step in preparing the hydrologic model for a sanitary sewer system is to divide the service area into smaller
areas tributary to the flow monitoring locations. This process is called sewershed delineation. The flow monitoring
data and rainfall data, described in Chapter 4, and the delineation results, are the three key parameters to determine
RDII using the SSOAP Toolbox. Also, these delineated small areas are the building block of the hydrologic model in
SWMM5.
The level of delineation depends on the extent of the sanitary sewer network model. In general, a model may contain
up to four working levels of delineation:
• Service Area - all areas tributary to a wastewater treatment plant (WWWT)
• Sewershed areas - subdivisions of a service area, delineating sewers directed to point of treatment or a major
trunk sewer
• Sub-sewershed areas - subdivisions of a sewershed, delineating sewers directed to second tier points of
interest such as metering locations, districts, or SSOs
• RDII catchment areas - subdivisions of subsewersheds, delineating sewers directed to model hydraulic
loading nodes
These levels of delineation will be used to organize and manage the hydrologic model datasets. Figure 6-7 illustrates
the application of these three levels in a single service area.
The sewershed area is the coarsest level of delineation within the service area tributary to a WWTP. The service area
can include one or many sewersheds. Figure 6-5 shows the entire service area and sewersheds for a hypothetical
sanitary sewer system.
A sewershed may be subdivided into sub-sewersheds. Typically, a sub-sewershed is delineated at flow monitoring
locations. Each flow monitor may have different RDII responses. Therefore, the hydrologic modeling work is
typically organized at this level.
The sub-sewersheds can be further delineated into RDII catchments, the finest level of delineation. RDII catchments
are drainage areas associated with each flow loading point (i.e., manhole) in the modeled sewer network. A sub-
6-51
-------
Sewer Service Area
Undeveloped Area
Flow Monitor
Model Extent
Unmodeled Sewer
Figure 6-5. Example of service area delineation.
sewershed can contain a number of RDII catchments tributary to the manhole where the flow meter is located.
The delineation will typically be made according to the location of sewers, parcel boundaries and topographic data.
In many cases to expedite delineation, the identified RDII catchments, sub-sewersheds, and sewersheds may include
areas that are not sewered, such as cemeteries, park land, highway rights-of-way, stream valleys, golf courses,
undeveloped areas, or areas on septic systems. Given the delineated area, the sewered area can be determined by
subtracting an estimate of the unsewered areas to allow a more accurate estimate of R-values required for model
input. For example, GIS procedures can be used to estimate unsewered areas for each RDII catchment based on
parcels that are not developed or based on percentages of predominantly sewered versus non-sewered land-use
definitions. Alternatively, aerial photography or mapping can be used to subtract large unsewered areas from the
6-52
-------
sewered area calculation.
6.2.1.5 Develop DWF Components
There are two components of DWF: BWF; and GWI. In many studies, the observed DWFs will be decomposed into
these two components. This decomposition process is not exact. Often, various assumptions are required. Typically,
DWF, BWF, and GWI are determined from metering data either installed in the collection system or located at
wastewater pumping stations or wastewater treatment plants. Alternative procedures can be used to estimate BWF
from water consumption or demographic data, but these will typically be adjusted to match the flows observed at
these flow meters.
It is often not necessary to separate a DWF into its GWI and BWF components in sanitary sewer system modeling
evaluations. The SSOAP toolbox only requires a lumped DWF (GWI and DWF) to determine the RDII flow.
However, it is useful to estimate GWI for prioritizing areas with large values for more detailed sewer system
evaluations and rehabilitation. Further, understanding GWI contributions lends insight into the seasonal variation
observed in DWF. This seasonal variation in DWF used to develop monthly values for input to SWMM5.
While DWF should be accurately determined, DWF is typically a smaller portion of the total peak flows that can be
generated in many wastewater collection systems under WWF conditions. Accurately determining the WWF
component is typically more critical in performing a successful sanitary sewer system capacity evaluation.
BWF Component
BWF refers to the sanitary flow component, which includes commercial, industrial, and residential flows discharged
to the sanitary sewers for treatment. This is sometimes referred to as base sanitary flow. There are many approaches
to developing BWF. One approach uses population data, generally derived from land use or census data, together
with an assumed unit wastewater flow rate (gallons per day per capita). Flow monitoring data within the system, as
well as flow data collected at the WWTP, are then used to define the composite base flow (BWF plus GWI). Finally,
the difference between the observed DWF and the computed BWF is attributed to GWI. This procedure normally
requires allocating WWTP flows to individual modeled sewersheds, which introduces uncertainty, as the actual BWF
from a basin may be very different from the allocated amount. For this reason, other approaches, such as using water-
use records with an estimated rate of return, have recently been used more frequently in lieu of population data as a
basis for estimation of BWF.
Water use is a more reliable basis for BWF estimates, and the data is generally available at a better resolution. In fact,
the resolution is so fine (at individual parcel level) that processing and storing the large datasets has only become
feasible recently with the advent of GIS capabilities. As the use of automated meter reading (AMR) becomes more
widespread, more detailed water use data will be available, thus further increasing the benefits and accuracy of this
approach.
If a water usage method is chosen, the winter-month water use is recommended for estimating the BWF. During
winter months, it is expected that only a small percentage of the total water use will be for lawn irrigation.
Alternatively, land use and population data may be used to support the DWF input generation. The diurnal flow
patterns can then be established using the DWF analysis of the observed flow data at the nearest downstream location.
These patterns are then applied to the average BWF from each sub-sewershed and RDII catchment that are estimated
based on the average water consumption rates.
Vallabhaneni et al. (2002c) published detailed procedures for using water billing records for DWF model
development and calibration. The procedures involved compiling water billing records geographically within the
sanitary sewer service area, developing BWF for each RDII catchment based on actual water consumed, and
estimating GWI and diurnal variations based on the downstream flow meter data. Water meter data should be
reviewed to detect reading errors.
While water use data is one potential source for develping GWI and BWF rates, another approach may be used to
6-53
-------
obtain the minimum nighttime flows at the permanent or temporary flow meters. All of the minimum nighttime flows
are expected to be attributable to GWI. These minimum flows include a nominal and constant BWF.
No matter which approach is used to distribute the DWFs, large sources of wastewater flow, such as from industries,
water treatment plants, commercial establishments, should be taken into consideration in distributing the wastewater
flows within the upstream sewersheds.
GWI Component
GWI can be estimated from four sources of data described below. Each data source provides increasingly less precise
estimates, but covers increasingly larger areas.
1. Direct GWI measurement - Flow isolation studies can be used to estimate GWI flows during a nighttime
metering when the BWF is small. GWI estimates are developed via direct measurement using temporary
weirs installed at the desired sites. Seasonal variations of GWI for the test sites is established and used
(together with long-term WWTP or other meter flow records) to project seasonal variations across the entire
system. These data can be combined with data from installed pieziometers, which provide long-term
groundwater level data within the study area to improve the understanding of these seasonal groundwater
variations and their influence on GWI rates.
2. Inferred GWI measurement - The network of flow monitors for model calibration provides data that can be
used to estimate GWI throughout the system. This can be accomplished in the smaller basin areas where an
assumed fraction (usually between 80 and 90 percent in predominantly residential areas) of the minimum
diurnal low flows can be attributed primarily to GWI. However, the fraction can go as low as 50 percent
depending on the demographics, such as college towns, business districts, and industrial areas.
3. Water consumption data - BWF can be estimated as a percentage of water consumption. This percentage is
often referred to as the rate of return. The difference between DWF and BSF can be used to compute the
GWI flows. The initial BWF estimates are developed by assuming that 80 to 90 percent of the water
consumed in a predominant residential area is eventually discharged into the wastewater collection system.
4. WWTP flow-based estimates - At the WWTP service area level, GWI is attributed to the difference
between observed DWF and the estimated BWF for the service area. WWTP plant influent flow records can
be used to develop DWF estimates. In smaller sewer systems this approach can yield better results.
Taken together, the above sources of data will support seasonally adjusted GWI estimates at the modeled-sewershed
level.
The estimated BWF, together with appropriate diurnal patterns and GWI estimates, should be used as flow inputs to
the model and then calibrated using the observed flow monitoring data during dry periods. The estimated wastewater
return rates and GWI are adjusted during the model calibration.
GWI rates determined at a meter will typically be distributed to the upstream sub-sewersheds and RDII catchments
based on the sewered area or sewer inch-miles. Sewer inch-miles equal the sum of the sewer length in miles times the
sewer diameter in inches for the sewer segments within an RDII catchment or sub-sewershed.
6.2.1.6 Develop RDII Characteristics
Rainfall and flow monitoring data should be analyzed using the SSOAP Toolbox to develop an understanding of the
system RDII characteristics for each sewershed/sub-sewershed/catchment. Chapter 5 describes how to perform flow
and rainfall analysis, develop R,T,K parameters, and generate RDII hydrographs. The R,T,K values or RDII
hydrographs for specific precipitation conditions will then be input into SWMM5 for hydraulic routing.
6.2.2 Model Calibration and Verification
Model calibration and verification is a critical step in ensuring that the model properly simulates the prototype system
6-54
-------
over a range of storm events under existing conditions. As described in Chapter 4, rainfall and flow data collection
programs are critical for a successful model calibration and verification. During a model calibration, the selected
rainfall data are used as model input, and the observed flow data are compared with the model simulation results.
This is followed by successive applications of the model, during which initial estimates of calibration parameters are
adjusted until the simulated results reasonably match the observed data.
A numerical goal for the model calibration should be established to assure that the model simulation results are
adequately correlated with the observed flow data. The goal of the model accuracy should be established based on the
quality achieved in the flow monitoring data. The factors for consideration include the specific objectives for the
capacity assessment, model extents and detail, rainfall data accuracy and precision, the spatial resolution, and data
accuracy and precision of the flow monitoring program. In general, a greater degree of model accuracy can be
achieved with a detailed model representation of the sewer system, extensive field surveys/investigations, flow
monitoring and rainfall monitoring at strategic locations, focused model calibration and verification efforts. This can
be very time/cost intensive. The calibration goal must achieve a balance between the capacity assessment objectives
and resources and schedule constraints. Flow characteristics, and the shape of flow hydrographs for comparison
between the simulated model results and the observed flow data, include peak flow depth, peak flow rate, and volume.
It must be emphasized that only reliable portions of observed flow monitoring data should be used in making the
correlation analysis.
Model calibration and verification is a three-step process, involving:
1. DWF calibration
2. WWF calibration
3. WWF verification
A brief discussion of these steps follows.
6.2.2.1 DWF Calibration
The purpose of the DWF calibration is to ensure the model is able to properly represent sewer characteristics under
existing DWF conditions. This process will also help develop a proper understanding of the sewer network
hydraulics without the influence from a rainfall-derived hydrologic response.
Once DWF is routed through the collection system, the simulated DWF hydrographs can be compared with the
observed DWF response at each flow meter location. The difference between the observed flow and the computed
BWF is attributed to GWI and uncertainties in the wastewater return rates assumed for initial BWF estimates. DWF
model calibration involves adjusting the lumped parameter, which includes the initial estimate of wastewater return
rates and the magnitude of the GWI component. In addition, minor adjustments may be necessary to improve
correlation of the timing of the simulated diurnal flows with the observed DWF data.
Groundwater elevations, and thus GWI rates, may vary considerably over the course of the year in response to
seasonal rainfall trends, the growing season, and other factors. The modeler should decide whether a winter/spring
high groundwater condition is the desired condition to be simulated. This decision should also consider the selection
of the wet-weather rainfall condition that will be simulated as the rainfall frequency statistics for many parts of the
country are driven by short-duration, high intensity rainfall events that occur in the summer or fall. Model
simulations should not combine GWI rates for one season with rainfall conditions from another season without
considering the joint probabilities of these conditions occurring at the same time.
The "Time Series" feature in SWMM5 allows the analyst to view the simulated hydrograph and the observed DWF
response. The following calibration metrics should be checked for correlation between the simulated and observed
data:
• Peak water depths
6-55
-------
• Peak flow rates
• Volume
• Hydrograph shape
Where significant differences occur in the above metrics, the appropriate model parameters should be adjusted while
keeping the parameters within an acceptable range. Users should focus on the following elements in DWF
calibration.
• Estimates of DWF (BWF and GWI.)
• Manning's roughness coefficient (n-value) and friction loss in the system. This parameter can affect depths
of flow in the system and available system capacity.
• Physical system components involving uncertainty (e.g., unknown flow dividers, hydraulic blockage, and
sediment.)
• Possible large water dischargers, such as from industries that could affect estimated GWI rates.
Model calibration is an iterative process. The target is to achieve reasonable correlations between the model results
and observed data. First, the analysis should match flow depth [or hydraulic grade line (HGL)], then match peak flow
and volume, and subsequently go back and confirm that the depth correlation is still valid. Typically, the observed
depth data is more reliable than the velocity data which is used to estimate flow rates. In cases where the velocity
data is questionable, the model calibration goal can still be achieved by calibrating to the depth data only.
DWF model calibration can be useful in identifying possible system operation problems due to major root intrusion or
major sediment deposition in trunk sewers. For example, an unusually high pipe roughness coefficient needed to
achieve model calibration is an indication of some blockages in the sewer.
6.2.2.2 WWF Calibration and Verification
The goal of WWF calibration and verification is to ensure the accuracy of the sewer system model in estimating
WWF response (i.e., peak HGL, peak flow rates, and surcharge/overflow). Selection of rainfall events that generate a
desired range of RDII responses is critical for model calibration and verification. At least two storms from the flow
monitoring data should be selected for model calibration and one independent event for verification. Additional
events may be necessary in some cases when a higher degree of confidence is needed and adequate time and
resources are available. Reliable flow and rain data must be available for as many of the installed flow and rainfall
monitors as possible for the selected event.
The storm events selected for WWF calibration will produce sewer-system responses under varying antecedent
moisture conditions. Where resources are permitted, continuous rainfall records may be considered for calibration
and verification. The storm event selection should also consider the proposed use of the model. If the model will be
used to simulate peak flows for a design storm event, then the model calibration should focus on large events as close
as possible to the design storm. If, on the other hand, the model is going to be used for continuous simulations to
determine peak flows and flow volumes to be stored, then the calibration events should include a range in events.
Model simulation
Once users have identified the WWF calibration events and obtained appropriate RDII hydrographs using the SSOAP
Toolbox, they can perform model simulations using SWMM5. (See Chapter 5 and the SSOAP Toolbox user manual
for guidance on how to generate RDII input hydrographs and link them to SWMM5.) This section provides guidance
on selecting typical model simulation parameters, such as solution technique, simulation time step, and simulation
duration. The selection varies depending on a sewer system's configuration and complexity. The following provides
three considerations in performing model simulation using SWMM5.
Solution technique - It is recommended that the Dynamic Wave Routing option (full Saint Venant equations) be
used for sanitary system modeling to perform capacity assessment that includes surcharge and overflow analysis.
6-56
-------
Please refer to Chapter 3 for a discussion of the dynamic wave flow routing option, and to the SWMM5 user manual
or online SWMM5 help for specific information on this option. Other solution techniques are discussed in the
SWMM5 user manual.
Simulation time step - It is a good general practice to use 5 seconds or less time steps for single event simulation.
For complex systems, a smaller time step is usually required to maintain stability during the computations and to
achieve continuity errors no more than 2 percent. Users may also use the variable time-step option in SWMM5,
where the program will automatically decrease the simulation time step when needed. For simpler system networks,
the simulation time step can be larger.
Simulation duration - The model simulation should last long enough to capture the collection system's full response
to a rainfall event. The duration is determined based on the sewersheds RDII response characteristics. For a sewer
system that exhibits a drawn-out response to rainfall (i.e., prolonged infiltration), simulation duration beyond the last
time step of rainfall event may be warranted. In some cases, this could be several hours to days depending on the size
of the sewershed. A review of the observed flow data at downstream locations which reflect the flow travel time
through the system will provide a basis to estimate the proper simulation duration. For continuous simulation, the
duration depends on the rainfall period of record selected and the rainfall distributions at the end of the selected
period.
WWF calibration
The model calibration efforts should first include the adjustment of DWF to reflect pre-rainfall event GWI conditions.
This adjustment is necessary to establish pre-event specific conditions to allow calibration to the wet-weather
response only without introducing error due to DWF temporal variation. This adjustment must be performed because
the DWF in the SSOAP Toolbox is determined through averaging the flow rates for multiple dry days during the flow
monitoring period. The GWI varies gradually through seasons. Therefore, the DWF before each wet-weather event
must be adjusted to reflect the changes in GWI. The variation in GWI is typically of lower magnitude for shorter-
term periods (one to four months) and becomes more significant in longer-term (annual and year-to-year.) DWF
returns to DWF-calibrated conditions following the wet-weather calibration for use in subsequent analyses. Once
appropriate pre-rainfall event antecedent DWF is established, model simulations are performed to calibrate the wet-
weather parameters.
Model calibration and verification are performed with SWMM5 for the selected rainfall events using the estimated R,
T, and K parameters and RDII hydrographs derived from flow monitoring data. During wet-weather model
calibration, the estimated hydraulic characteristics of the sewers such as pipe roughness pump station characteristics
and flow diversion structures (such as weirs, dams, orifices) may need additional adjustments beyond refinements
through the dry-weather calibration. This is primarily due to the full range of flow rates that occur during wet-
weather conditions compared to low flow rates observed during dry weather.
In some cases, when the model extends sufficiently far upstream of the flow meter location with multiple flow
loading points, the T and K parameters may need adjustment to account for the flow-travel times from the model flow
loading points to the downstream flow meter location where the initial R, T, and K parameters were derived.
Like the DWF calibration, the following calibration parameters are commonly used to compare the simulated and
observed data:
• Peak water depths
• Peak flow rates
• Volume
• Hydrograph shape
WWF verification
After the model is adequately calibrated based on the reliable data from the flow monitoring program, rainfall
6-57
-------
collection program, and understanding of the sewer network system, it must be verified using at least one additional
independent rainfall event. To confirm the model parameters, verification can also be performed using a continuous
simulation for an extended period with multiple rainfall events. Typically, the extended period can last a few weeks
or a few months depending on the data availability. Continuous simulations will ensure that the model can accurately
predict the dry- and wet-weather flow under different antecedent moisture conditions, and confirm the model
performance under back-to-back wet-weather conditions.
6.3 Capacity Assessment
The framework for the capacity assessment should be consistent with overall SSO control objectives and stakeholder
expectations. It is critical to develop capacity assessment goals specific for each study. The sanitary sewer system
wet-weather flow management plan requires a sophisticated hydraulic capacity analysis. The flow monitoring data is
used for understanding of the system behavior at the point of measurement during the monitoring time period and for
calibration of the sewer system hydraulic model. The calibrated model is then applied for selected combinations of
precipitation, antecedent moisture, and groundwater conditions to assess the system performance. Through the
model application, the sewer system capacity is assessed and the baseline hydraulic conditions are established.
6.3.1 Capacity Assessment Steps
There are two key steps in capacity assessment:
1. Establish specific capacity assessment goals.
2. Determine baseline system hydraulic performance under existing and future projected growth conditions.
6.3.1.1 Capacity Assessment Goals
Capacity assessment goals must be established early in a SSO control analysis and planning. These goals must
consider the system's unique configurations, hydraulic capacity challenges, types of overflows, stakeholders'
expectations, and socio-economic conditions. Some common goals for capacity assessments include:
1. Identify locations and causes of sewer system hydraulic constraints.
2. Assess pump station hydraulic performance.
3. Assess the WWTP's ability to handle RDII.
4. Assess sewer system hydraulic constraints in the context of inadequate maintenance, hydraulic capacity, and
excessive RDII.
5. Assess the performance of the existing system under future population growth scenarios.
6. Assess customer sewer back-ups, manhole-flooding locations, and overflows to waterways.
7. Assess how existing system performance will be improved by currently planned sewer rehabilitation and
improvement projects.
8. Define performance expectations.
6.3.1.2 Baseline Hydraulic Performance Assessment
Baseline hydraulic conditions are defined as the DWF and WWF hydraulic responses under the current sanitary sewer
system configuration. This baseline assessment provides the basis for a system-wide characterization of the problem,
identification of problem areas, and ranking of improvement needs. In addition, baseline conditions are used to
measure a municipality's progress in improving system performance as projects are implemented to address the
capacity problems.
DWF capacity assessment.
The first step is to perform model simulation with DWF input to assess sewer capacity during peak diurnal flow
conditions, identify capacity constraints, and assess GWI impacts on the DWF capacity. Existing sewers are then
assessed for the projected future DWF. This assessment will provide insights of the existing sewer network's
capability to accommodate future flows from population growth and service area extensions. The goals established
for the capacity assessment may require DWF evaluations for a range of future growth conditions (two, five, 10 or 20
years.) Running the model for multiple future growth scenarios allows the development of priorities and
6-58
-------
implementation schedules for identified projects to address future capacity needs. Table 6-1 shows an example of
DWF capacity assessment results. In this example, most of the sewer sections have the ability to perform under
existing and up to five-year growth conditions. However, some sections of the system would not have the capacity
for the 20-year growth conditions. The whole sewer system is flowing more than half full during the peak DWF
under 20-year growth conditions, indicating the potential needs of system improvements 20 years into the future.
Table 6-1. Example DWF Capacity Assessment Results under Different Conditions
Interceptor
Section 1
Section 2
Section 3
Section 4
Section 5
Section 6
Pipe
Dia.
(in.)
36
42
36
36
60
60
Existing DWF Conditions
Peak
Depth
(in.)
13
17
23
19
23
23
Cap.
Taken
36%
40%
64%
53%
38%
38%
Cap.
Remain
64%
60%
36%
47%
62%
62%
5 -Year Growth DWF
Conditions
Peak
Depth
(in.)
15
25
28
21
29
31
Cap.
Taken
42%
60%
78%
58%
48%
52%
Cap.
Remain
58%
40%
22%
42%
52%
48%
20-Year Growth DWF
Conditions
Peak
Depth
(in.)
20
35
37
25
40
62
Cap.
Taken
56%
83%
100%
69%
67%
100%
Cap.
Remain
44%
17%
0%
31%
33%
0%
WWF capacity assessment
Once the existing sewer system's ability to manage the current and future DWF flows are understood, a WWF
capacity assessment can be performed to establish baseline conditions. The WWF capacity assessment should be
comprehensive and consistent with the capacity assessment objectives established. A vast number of possibilities
exist in defining the assessment scenarios depending on unique problems of a sewer system. The following general
scenarios are commonly considered:
1. WWF capacity assessment with existing DWF during the rainfall/flow monitoring conditions - This
assessment characterizes system behavior during the period of observed flow data and further validates model
reliability. If resources are available, conducting continuous simulation for the duration of a monitoring
period and assessing the magnitude and recurrence of capacity constraints and resulting surcharge and
overflow conditions is recommended. The result is a clear definition of the system WWF capacity under
monitored conditions throughout the sewer network modeled.
2. WWF capacity assessment with future DWF projections with selected rainfall events during the
monitoring program - This assessment provides insight into potential system problems with the existing
system configuration if the projected increase in DWF occurs.
3. WWF capacity assessment with existing DWF and selected design storm conditions - This assessment
applies to an unmonitored design storm (synthetic or natural) and a model with the existing system
configuration and existing DWF. In some cases, this may be based on a historic storm with rainfall volume
scaled up or down to meet desired design storm intensity or volume. Some studies have used a design storm
based on a standard distribution such as the Natural Resources Conservation Service (formerly U.S. Soil
Conservation Service -SCS) Type II rainfall distribution. Caution is advised when selecting a design storm.
The design storm definition should be consistent with the performance expectations for the individual sewer
system. It should be recognized, however, that the simulated peak flows in most cases will not have the same
return period as the rainfall event because of the many other conditions that influence peak wastewater flows
6-59
-------
(i.e., antecedent moisture conditions, groundwater elevations.)
This scenario establishes the baseline conditions that characterize the sewer system capacity constraints and
overflows. For the unmonitored design storm conditions, there are different approaches available for deriving
the RDII parameters. Chapter 5 introduces four statistical/regression approaches to determine R,T,K
parameters for those conditions.
4. WWF capacity assessment with existing DWF and selected long-term historical precipitation period -
This assessment allows the user to characterize system behavior over a long-term historical period (in the
order of one to 10 years) and a range of rainfall conditions and antecedent moisture conditions. It supports a
comprehensive review of average annual system overflow frequency, volume, and duration characteristics.
This work will be time-consuming and cost-intensive, but it would yield high confidence in establishing
baseline conditions.
5. WWF capacity assessment with future DWF and selected design storm conditions - This assessment
provides insight into potential system problems with the existing system configuration if the projected
increase in DWF occurs under design storm conditions.
Review of model results
The following assessment parameters are usually used to review model results for DWF and WWF capacity
assessment simulations. The graphical capabilities is SWMM5 facilitate the system wide assessment of these
parameters:
• Sewer capacity use
• Sewer surcharge levels
• Manhole flooding
• Overflows
A sewer surcharge occurs when the HGL is in excess of the sewer's crown elevation and a manhole is flooded is
when the HGL is in excess of the manhole ground elevation. It is generally uncommon to experience manhole
flooding, sewer back-ups,and overflows under DWF conditions. However, they do occur because of temporary
blockages in the system or operational problems at pump stations and WWTP facilities.
Sewer capacity use: This is a very common assessment parameter, as sewers are designed to use only a portion of
full-pipe capacity under dry-weather conditions. Pipe capacity use can be evaluated based on flow rate and flow
depth. The sewer flow capacity use is calculated by the ratio of the simulated peak flow to the design flow of a
conduit in percentage. The sewer depth capacity use is calculated by the ratio of the simulated peak depth to the size
of the pipe in percentage.
Full-flow sewer capacity based on the pipe slope is conservative in that it does not recognize the fact that many sewer
systems can be allowed to surcharge to some degree during infrequent storm events as long as surcharges are not
excessive to cause manhole overflows or basement flooding.
Furthermore, the simulated peak flow to design flow ratio can yield some results that do not represent the actual
capacity of the pipe segment in question. This is due to the design flow typically being calculated using Manning's
equation based on the slope of the pipe segment in question only. There are often cases where a short segment of pipe
will have a mild adverse or near-zero slope and result in a very low design flow calculated by Manning's equation
which would result in a very high simulated peak flow to design flow ratio. In many of these cases, the actual design
capacity to convey flow depends on the conditions upstream and downstream of the pipe segment in question and
should be considered for these cases they may not actually be system constraint.
The intervals for the percent capacity can be defined to suit a particular study. A common set of intervals to evaluate
system capacity, which is also the default SWMM5 color for each interval, are listed below.
6-60
-------
Interval 1: 0-25 percent (blue)
Interval 2: 25-50 percent (light blue)
Intervals: 50-75 percent (green)
Interval 4: 75-100 percent (yellow)
Interval 5: > 100 percent (red)
SWMM5 allows users to display the system capacity thematic maps. Figure 6-6 depicts such a map. Users can also
develop thematic maps illustrating sewer capacity utilization in standard GIS or CAD packages. The red highlighted
conduits in Figure 6-6 represent sewers with no capacity left.
ft SWMM 5 - AnyTown_10yr.lNP
File Edit View Project Report Tools Window Help
D fi£ y (» Sfe *4 ? ?{J tid S !* •
H
Themes
Subcatchments
Animator
H < I- >
I Q~~
f^t Frequency Plot
Frequency/Contour
O Subcatchments
O Nodes
O Links
1 Uurrent Selection
> Study Area Map
D Auto-Length Flow Units: CFS
Zoom Level: 100%
X,Y: 1430260.114, 445818.E Current Item:
Figure 6-6. SWMM5 thematic map example.
6-61
-------
Sewer Surcharge Level: Reviewing sewer surcharge levels is an important step in assessing system hydraulic
performance and for investigating basement flooding problems. The SWMM5 simulation results contain data of all
conduits/manholes that surcharge including duration and surcharge depth. Table 6-2 shows an example of surcharge
summary, showing one manhole flooded and others surcharged. Using the thematic map similar to Figure 6-6, users
can present the same information in SWMM5 or in GIS environment.
Table 6-2. Sewer Surcharge and Manhole Flooding Summary
Interceptor
Section 1 (36 in.)
Section 2 (42 in.)
Section 3 (36 in.)
Section 4 (36 in.)
Section 5 (60 in.)
Section 6 (60 in.)
Design
Capacity
(MGD)
51
31
25
12
41
55
Simulated
Peak
Depth
(in.)
143
21
42
45
41
50
Surcharge
Depth
(in.)
107
N/A
6
9
N/A
N/A
Manhole
Depth (in.)
143
120
120
120
120
120
Capacity
Used
100%
50%
100%
100%
68%
83%
Capacity
Remain
0%
50%
0%
0%
32%
N/A
Duration
Surcharged
(h)
0
0
0
Notes
Flooding
N/A
Surcharged
Surcharged
N/A
N/A
Manhole flooding: Manhole flooding is another consideration for assessing the hydraulic performance of a sanitary
sewer system. Manhole flooding information is essential when investigating potential street flooding, and overflows
to area waterways. Manhole flooding information, such as manhole overflow volume contained in the SWMM5
output, can be extracted for analysis.
Overflows: Overflows from sanitary systems occur in several ways: at previously constructed outfall pipes within the
sanitary collection system, at the WWTP, or by overflows to the customer basements via house lateral sewers. A
solid understanding of overflow location, frequency, volumes, duration, and causative factors is critical to mitigating
overflows. Some communities may have historically constructed SSOs, which are typically represented by an
overflow pipe at a manhole that discharges to a receiving water body or storm sewer. SSOs typically have free
outfalls but can be influenced by the stage elevations of receiving waters. The receiving water is considered a
boundary condition in SWMM5 and it is important to include it in the calibrated sewer system model.
In summary, the steps described above are expected to yield a baseline characterization that fulfills the goals
established for the capacity assessment. A properly developed and calibrated hydrologic and hydraulic computer
model can provide an effective means to assess the hydraulic capacity of a sanitary sewer system for the existing and
future growth conditions. The knowledge gained through the system characterization combined with the SSO control
objectives and system model can then be used to screen a range of alternatives to determine a recommended plan that
can cost-effectively mitigate SSOs. The resulting WWF management plan can then be integrated into an overall
system capital improvement program, as described in Chapter 7.
6-62
-------
Chapter 7 Development and Analysis of System Improvement Alternatives
This chapter presents guidelines for developing and analyzing improvement alternatives to mitigate SSOs. Analysis
techniques for simulating and developing improvements needed to meet the identified planning criteria are described.
These techniques use the various system analysis approaches using the SSOAP Toolbox. A general approach for
developing a targeted SSO control plan within funding and schedule constraints is outlined. This chapter also
discusses ways to integrate a wet-weather management plan or sewer system improvement plan into a municipality's
capital improvement program.
7.1 Establishing Planning Objectives and Improvements Criteria
7.1.1 System Improvement Planning Objectives
The development of system improvements requires an integrated hydrologic and hydraulic analysis of the sewer
system of concern, including boundary conditions or limitations from downstream pumping constraints, to gain a
comprehensive understanding of existing capacity, DWFs, and WWF responses. The system improvements discussed
in this chapter focus primarily on controlling capacity-related SSOs that occur during wet weather. SSOs caused by
system operations and maintenance (O&M) are briefly addressed. For more information on O&M related SSOs,
useful references include the Guide for Evaluating Capacity, Management, Operation, and Maintenance (CMOM)
Programs at Sanitary Sewer Collection Systems published by the EPA's Office of Enforcement and Compliance
Assurance (EPA, 2002), Optimizing Operation, Maintenance, and Rehabilitation of Sanitary Sewer Collection
Systems (NEIWPCC, 2003), and Wastewater Collection Systems Management (WEF, 1999.) Data regarding sewer
system maintenance, sewer surcharging and resulting basement backups, and customer complaints should be
reviewed to understand if SSOs are the result of capacity limitations and/or O&M issues such as obstructions and
blockages.
Sewer system improvement programs are established, in general, to meet the following capacity-related objectives:
• Provide sufficient transport and treatment capacity for existing and future flows during both dry- and wet-
weather conditions.
• Comply with regulatory requirements for capacity assurance and SSO avoidance.
• Meet the level of service expected by customers to avoid system surcharging that may lead to basement or
service backups.
These objectives are usually developed in response to system flow conditions identified by the utility through
customer interactions, regulatory discussions, and/or their combinations. Increasingly, these capacity objectives are
geared to control SSOs in terms of frequency of occurrence, such as no more than once every two, five or 10 years on
average.
In addition to the above capacity-related objectives, the following objectives should also be considered during final
selection and design of sewer system improvements:
• Control wet-weather effect on operation of system facilities, such as WWTP and pumping stations.
• Provide long-term system structural integrity.
7-63
-------
• Control O&M costs.
• Maintain the asset value of the system infrastructure and ensure that they will continue to operate with
minimized risk of failure.
• Minimize the effect on the community and environment from construction, odors, and noise.
A systematic evaluation based on system-wide hydrologic and hydraulic modeling has become the main stream
approach for developing a good understanding of a sewer system's capacity status, especially when the evaluation is
supported by performance data collected through flow and rainfall monitoring and SSES . Utilities throughout the
world have used this approach to gain a better understanding of their current conditions and future wastewater
collection needs. In turn, such an understanding is essential for making cost-effective, proactive, equitable, and
efficient decisions concerning sanitary sewer system improvements.
7.1.2 System Planning Criteria
There are two types of planning criteria in developing and implementing sewer system improvements. The first
criterion is to define a threshold of system deficiency based on system model simulations to alert that system
improvements are needed. Examples of these criteria include:
• SSOs shall not be predicted to occur during the specified wet-weather planning condition.
• Peak DWF shall use no more than a given percentage of the system capacity (usually 50 to 75 percent,
depending on system age and level-of-service expectations.)
• Sewer surcharges shall not exceed a given limit, such as no more than three feet above the crown of the pipe
or no less than three feet below the manhole rim elevation.
The second criterion is to define the sizing requirements of new improvements to correct the identified system
deficiency. In many cases, regulatory agencies governing sewer system improvements spell out the system design
criteria that must be met, such as minimum wet-weather to dry-weather peaking factors or minimum design velocities
to maintain solids flushing. Examples of these criteria can be found in the Recommended Standards for Wastewater
Facilities, also known as the Ten State Standards (HES, 1997.)
Criteria to size system improvements for facility planning studies are typically based on a combination of future
projected DWFs and a maximum frequency of SSO occurrence. Predicted wastewater flows are generated for these
conditions using calibrated hydrologic and hydraulic models. Based on these simulations, system improvements are
sized to meet identified criteria, which may include the following considerations:
• No system surcharging will occur for the future base wastewater flows (e.g., 20 to 30 years in the future), in
combination with a maximum RDII recurrence interval of once every two, five, or 10 years.
• Maximum levels of surcharge will be limited to some criteria for future flows in combination with a less
frequent RDII recurrence interval.
• Sewer system overflows will not occur more frequently than a given recurrence interval.
• The selected recurrence interval may change within a sewer system service area depending on the
environmental impact of the overflow, and the impact and cost of the facilities required to reduce the
frequency.
• Any system storage facilities will not hold wastewater for more than 72 hours and more frequently than once
every two, five,or 10 years to control odor production.
• A maximum sustainable wet-weather treatment capacity will not be exceeded for more than 24 hours and
more frequently than once every two, five, or 10 years.
These criteria will be system-specific based on customer level-of-service preferences, O&M considerations, and
wastewater treatment processes and operations. Therefore, they should be considered prior to developing specific
improvement alternatives, as they will influence the size of improvements and thus may influence the outcome of the
alternatives analysis.
7-64
-------
Alternatively, sensitivity analyses should be performed to identify the facilities required to meet selected levels of
service. The cost and impact of these improvements should be evaluated so that informed decisions on the
improvements can be made.
7.2 Options for Improving Collection System Performance
Improvement alternatives should account for the impacts of improvements on the whole system, including the impact
on downstream facilities and the ability of the treatment facilities to handle the additional flows. System
improvement options generally include:
• Sewer rehabilitation as a means of increasing existing sewer capacity and/or reducing infiltration and
inflow.
• Equalization storage to reduce downstream peak flows, including the possible use of real-time controls
(RTC) to optimize the use of existing system storage.
• Increased conveyance capacity through gravity sewer construction or pump station/force main upgrades.
• Increased wastewater treatment capacity through facility expansion, or the use of other wet-weather
treatment processes, such as high-rate clarification.
I/I reduction generally relies on rehabilitation of sewer lines and manholes that are often the sources of groundwater
or WWFs in sewers. However, in some areas of the country such as Ann Arbor, Michigan (which is described more
fully in Chapter 8), footing drains or other private system connections to the sanitary system have been identified as
major sources of RDII. In these situations, the contributions of RDII and the cost of removing these connections
should be evaluated as an alternative.
Initial evaluations may consider each of these alternatives individually. However, as to be described in Section 7.3,
improvement alternatives that combine these improvements to meet system performance objectives should be
developed. The final alternatives chosen should meet the utility's level of service objectives and regulatory
requirements, as well as budget and time constraints.
Descriptions of each improvement strategy is provided in the following sections, followed by a discussion of the
development and evaluation of system-wide alternatives to improve sewer system performance using combinations of
these improvements.
7.2.1 Sewer System Rehabilitation
It is typical to perform a focused SSES to determine the scope of sewer rehabilitation required. This often involves a
dense flow monitoring network to determine the sources and severity of extraneous flows. The monitoring is
generally followed by detailed manhole inspection, close-circuit television inspections, smoke testing, flow isolation,
and other techniques to gather adequate information of sewer conditions to guide rehabilitation.
Once an area is identified as a contributor of high RDII and thus designated as a rehabilitation priority, there are three
general sewer rehabilitation approaches to proceed:
• Rehabilitate all sewers including service laterals located within the public right-of-way and on private property.
• Rehabilitate only sewers located in public rights-of-way.
• Repair structural defects in pipes and manholes and remove major inflow sources identified.
The first and second approaches are considered "comprehensive rehabilitation." A comprehensive rehabilitation
approach consists of rehabilitating every foot of sewer to eliminate all potential points of entry for RDII.
Experience has shown that the greatest cost/benefit ratios can be achieved by comprehensive rehabilitation of those
sewersheds area witth the greatest level of deterioration. Benefits may be reduced significantly for sewersheds with
lower levels of extraneous flow.
7-65
-------
The third approach is point rehabilitation, which repairs localized defects identified from inspection and focuses on
SSOs resulting from structural and maintenance problems rather than RDII. However, potentials are there to identify
some specific defects that are significant sources of RDII. This approach does not include rehabilitating the laterals,
and thus is not as effective in reducing RDII as the comprehensive approach.
Because of the time and cost required, and the uncertainty in peak flow reductions provided, sewer system
rehabilitation is best used as one part of an overall program that also includes other capacity improvement options,
such as relief sewers and pumping station upgrades. However, rehabilitation is an important part of all utilities'
ongoing O&M programs to prevent high levels of RDII and to ensure that the sanitary sewer system continues to
operate as designed.
When considering rehabilitation, lifecycle costs and benefits should be considered. Sanitary sewer systems
continually deteriorate over time. While generally accepted design life for the materials used to construct sewers is
on the order of 20 to 30 years, these sewers are called on to provide service for 50 years and longer. While
comprehensive rehabilitation approaches previously described have higher initial costs, the collection system is
revitalized both structurally and hydraulically, and the service life of the sewers can be extended significantly. A
point-repair approach is less costly, but it may not adequately control system deterioration. In addition, migration of
infiltration from the repaired defect to defects not addressed by the point repair approach may significantly reduce the
effectiveness of this approach in reducing RDII. The potential need for a continuing series of spot repairs may be
more costly and less effective than a comprehensive rehabilitation approach. The best approach will vary by systems,
and pilot rehabilitation projects that include pre- and post-rehabilitation flow monitoring to determine the RDII
reduction success of different approaches within each system are recommended. The validity of these RDII reduction
assumptions are critical to the success of the recommended sewer improvements program.
7.2.2 Storage
Flow equalization facilities can reduce SSOs by storing peak flows in excess of sewer capacity. They can effectively
reduce localized overflows as well as upstream and downstream overflows (by reducing the hydraulic grade line
elevation upstream, and by reducing downstream peak flow rates.) Flow equalization facilities can be constructed
within the sewer system, at pump stations and at wastewater treatment plants. Equalization basins sited at plants can
also be used for dry-weather diurnal equalization to dampen daily flow fluctuations and improve treatment
performance.
Storage is most effective when it is located immediately upstream from the portion of the system with insufficient
capacity. For this reason, many existing storage facilities are located at pump stations or treatment plants. The
principle of WWF control is relatively simple: flows that exceed the capacity of the downstream facilities are diverted
directly to the storage facility. Controlling flows using facilities located far upstream from the flow constraints is
often less effective due to time lags and the inflows that enter the system between the facility and the flow constraints.
The key concern in designing a storage facility is to determine its size. The size should account for the range in wet-
weather events that affect the system. The storm that produces the largest peak flow may not be a design storm;
rather it may be long-duration, low-intensity events that occur back-to-back. Long-term flow monitoring data and/or
continuous model simulations using long-term rainfall data are recommended to evaluate system performance when
sizing storage.
There are other factors to consider when locating and designing storage facilities, including odor control, solids
handling and cleaning the facility after system operation. The operating costs of the facilities must also be
considered. Flow equalization analyses usually do not address continuing system deterioration and RDII flow
increases overtime. As collection systems age and extraneous flows increase, and as DWFs increase due to
development, the equalization facility may become undersized. Conversely, if rehabilitation is conducted in the
upstream sewers, flow equalization volume required may actually decrease or be eliminated altogether.
Flow equalization storage facilities are designed and operated either on- or off-line, as discussed next.
7-66
-------
7.2.2.1 On-Line Flow Equalization Storage
In on-line flow equalization facilities, flow is continuously routed through the storage system during both dry and wet
weather. On-line storage can be achieved by replacing a portion of an existing sewer with a larger conduit (pipe,
tunnel, or culvert), or by constructing a parallel conduit to provide additional storage capacity. Flows enter and exit
the on-line storage system by gravity, and wet-weather storage can be regulated by the downstream hydraulic grade
line or by a physical control device. Physical control devices include rate-of-flow control valves, regulators, orifices,
and inflatable dams. A low-flow channel may be constructed in the facility to ensure cleansing velocities are
maintained during DWF conditions.
7.2.2.2 Off-Line Flow Equalization Storage
Typical off-line flow equalization facilities include equalization basins (either open or covered), above- or below-
grade tanks, tunnels, and culverts sized to store peak WWFs that the sewer system cannot accommodate. Flow
diversion chambers or pump stations are required to divert peak flows from sewers to a flow equalization tank. A
good design and operating practice is to segment the tank into multiple cells and allow it to fill one cell at a time.
This approach minimizes the area to be cleaned after a wet-weather event, and can expedite tank draining by gravity
or by pumping. The basins can be covered and equipped with odor control systems to reduce the potential public
nuisance. Tank mixing systems are also frequently provided (mixers, blowers, pumps) to keep solids in suspension to
minimize cleanup effort and odor.
Costs for new flow equalization basins vary significantly depending on the method of tank construction, equalization
volume, and site-specific conditions. Construction of an above-ground, open-topped tank typically offers the lowest
costs, while a below-ground, covered tank the highest.
Facilities that require pumping of peak flows into the storage tanks would incur the largest cost. Facility costs can be
significantly reduced if site conditions allow influent flows to enter via gravity and pumping requirements are limited
to effluent flows, since return flow rates can be much smaller than the uncontrolled, storm-induced peaks. The
greatest cost efficiencies are achieved if site conditions allow gravity flow for both influents and effluents.
Flow equalization basins require inspection and cleanup after each storm event; routine testing and maintenance;
power for wastewater pumping, blowers, and mixers (if applicable); and chemicals for odor control (if applicable.)
The annual O&M costs for WWF equalization can vary significantly, depending on volume and number of excess
flows.
Tunnels have been successfully used in CSO applications to provide storage during storm events, with subsequent
pumping of stored flows to treatment facilities. Tunnels can be constructed in soft ground (shallow tunnels) or deep
rock formations. Soft-ground sewer tunnels are commonly constructed by microtunneling, pipe jacking, or installing
pipes inside an excavated tunnel (conventional method.) Box culverts can be effectively used for either on- or off-
line flow equalization facilities. This option typically entails construction of a below-ground covered culvert aligned
so that existing trunk sewers can overflow into it by gravity.
7.2.3 Conveyance
If rehabilitation or source controls are not expected to sufficiently reduce GWI or RDII, an increase in conveyance
capacity may be necessary. Increased conveyance capacity by larger or parallel sewers can provide the capacity
required for increases in sanitary flows due to population growth or system expansion. It may also prove to be a cost-
effective approach to address WWF requirements when combined with other improvement options.
In many cases, enhancement of conveyance capacity can be the least costly alternative in terms of capital and O&M
costs. However, potential environmental and community impacts may make this approach impractical, as it will
significantly impact the operation of downstream facilities such as pump stations, force mains, interceptor sewers and
wastewater treatment plants.
7.2.3.1 Trunk Sewer System Improvements
7-67
-------
Trunk sewer system improvements can effectively relieve pipelines prone to surcharging and overflows by increasing
sewer capacity. These improvements also have the benefit of providing additional dry-weather wastewater
conveyance capacity to accommodate future growth in a service area. Trunk sewer improvement alternatives include:
(1) replacement and relief sewers, and (2) sewer pressurization. Because trunk sewer system improvements result in
increased downstream wet-weather peak flows, downstream sewer system improvements (additional trunk sewer
capacity, plant equalization, and plant improvements) may be required in conjunction with upstream improvements.
Replacement and relief sewers
Replacement and relief sewers convey DWFs that exceed the existing trunk sewer capacity. Relief sewers may be
constructed to parallel an existing trunk sewer, or along an independent route designed to bypass hydraulically limited
areas. In some cases, relief sewers may also be used to divert flows to another branch of the collection system. They
may be designed as on- or off-line systems. On-line relief sewers, which convey both dry- and wet-weather flows,
should be designed to ensure that cleansing velocities are maintained to prevent solids deposition, odor, and
maintenance problems. In contrast, off-line relief sewers are only used during wet-weather conditions. Flow into off-
line relief sewers can be controlled hydraulically via a fixed weir or junction box, or mechanically using a power-
operated gate or similar device. In addition to providing additional wet-weather conveyance capacity, relief sewers
also provide sewer maintenance flexibility by allowing one sewer line to be removed from service without bypass
pumping.
Replacement sewers allow an existing sewer to be abandoned or removed. This approach may be preferable to relief
sewer construction if the existing trunk sewer is in poor condition, or if construction easement limitations and/or land
acquisition requirements preclude cost-effective relief sewer construction. However, the material costs for
replacement sewers are generally higher than those for relief sewers since the replacement sewers are sized larger to
offset the loss of capacity served by the existing sewer that will no longer be used. In addition, the need to maintain
sewer flow during replacement sewer construction may necessitate special construction procedures (e.g., bypass
pumping) that can significantly increase costs.
Sewer pressurization
Sewer pressurization can increase the hydraulic and storage capacity of existing trunk sewers by increasing the
hydraulic grade line in the reach until sewers are surcharged. Typically, manholes along the reach are either sealed or
raised to allow the sewer to be surcharged during peak wet-weather conditions without flooding.
Sewer pressurization is not a conventional improvement option, and its potential impacts should be carefully
considered on a case-by-case basis. The structural integrity and design of the sewer in question must be carefully
checked to ensure that it can withstand the increase in pressure. Equally important, the hydraulics should be carefully
examined to ensure that the higher water level does not cause sewage backups into homes or other connected systems,
and that the backwater does not reduce upstream carrying capacity. If manhole inverts are formed using conventional
methods to convey flow from one-half of the pipe depth, then pressurization may not increase hydraulic capacity
because of significant entrance and exit losses. To achieve this benefit, the channel must be reconstructed for
conveyance of flows that will fill the pipe.
If these considerations are adequately addressed, sewer pressurization can be one of the most cost-effective means of
eliminating localized overflows and increasing hydraulic carrying capacity and in-line storage.
7.2.3.2 Pump Station Improvements
Capacity upgrades to existing pump stations or the construction of new pump stations may be required to convey
WWFs and prevent overflows upstream of the pump station. A traditional approach would be to parallel the existing
force main with a new force main and either retain or abandon the existing force main, depending on its condition. In
some cases, this force main can convey increased flow by using pumps with higher discharge heads to overcome the
greater friction losses generated by the increased velocities. If the force main receives flow from other pump stations,
one must also weigh the impacts of increased system heads on the pumping capacity of the other stations. Also, over
time, velocities in excess of 10 ft/s can scour the force main interior, causing premature structural failure.
7-68
-------
Another alternative to consider is constructing a new pump station and force main to divert flows from a sewer reach
with insufficient capacity to a location that can adequately convey the additional flow. The pump station can be
constructed to divert all flows, dry- and wet-weather, from a portion of the sewershed. Alternatively, the pump
station can operate intermittently to divert wet-weather flows that exceed the existing downstream conveyance
capacity.
7.2.4 Treatment
Pump station and pipe capacity upgrades will produce increased wastewater flows to WWTPs. If these flows exceed
the wastewater treatment plant capacity, the flows are temporarily stored in the system until capacity is available, or
improvements must be made at the plant to handle the additional flows. The methodology for evaluating treatment
plant alternatives is discussed in detail in the 2006 WEF document (WEF, 2006.) To evaluate impacts to a WWTP,
analysis must be performed to quantify flows to the WWTP, summarize the service area's characteristics, determine
the capacity of the WWTP system and its components, and evaluate the secondary treatment processes.
Alternatives for improving the WWTP's performance during wet weather include on-site storage, in-plant flow re-
routing, high-rate treatment processes, blending, and full secondary treatment. Stakeholders should select the best
option based on the defined criteria and the cost of each alternative.
7.2.5 Real-Time Control (RTC)
RTC enables wastewater collection systems to capture increased wet-weather flow by: (1) maximizing the use of
available in-system storage; (2) maximizing available system conveyance capacity by diverting flow dynamically;
and, (3) enhancing control logic for off-line storage. In-system storage reduces conveyance capacity during wet
weather and allows flow to back up into and be stored within the otherwise unoccupied pipe. Dynamic flow diversion
enables the system's hydraulic conveyance capacity to be fully used during wet weather by shifting flows from
overloaded lines to those with capacity. Enhanced control logic at off-line flow equalization facilities allows storage
volumes, as well as influent and release rates, to be adjusted based on hydraulic conditions at critical locations along
trunk sewers that may be a considerable distance from the storage facilities. Hydraulic models are used to simulate all
three RTC strategies described below for increased flow capture and determine which approach or combinations of
approaches are most beneficial for system performance improvements.
7.2.5.1 Overview of In-System Storage
In-system storage is the most common objective of RTC in combined sewer systems, but the principles can be applied
to sanitary sewer systems in some cases. This approach generally provides greater benefits in larger systems. Since
much of the storage capacity is typically used to convey flows, this approach would typically provide relatively small
benefits. The ideal condition for its application is whenever the system includes large-diameter sewers that flow
partially full during large wet-weather events.
7.2.5.2 Overview of Dynamic Flow Diversion
Branched and parallel interceptor sewers enable a sewer system to employ RTC for dynamic diversion of flows
during wet weather. RTC facilities are located at critical system junctions of major trunk sewers, and diversion of
flows within the interceptor network is controlled through real-time monitoring of flows and levels (and sometimes
precipitation-based predictions of flows) to maximize conveyance capacity. A detailed study using hydraulic analysis
models can simulate RTC scenarios and identify RTC opportunities.
7.2.5.3 Overview of Enhanced Control Logic
For many flow equalization facilities, the quantity of flow entering the storage tanks and the release rate to the
downstream trunk sewer is determined by local hydraulic conditions. Many control structures are static, such as a
fixed side-spillway weir for influent flow and a fixed orifice or pumping rate for effluent flow. System operators can
achieve more effective and flexible use of available storage if static controls are replaced with more complex RTCs.
Using variable release rates from flow equalization facilities can adjust decision control logic to seasonal groundwater
conditions and/or ensure that storage tanks are emptied as quickly as possible and available for back-to-back storms.
7-69
-------
Remote sensors can set the release rates from upstream storage tanks according to the downstream hydraulic
conditions and ensure that the hydraulic capacity along critical trunk sewers is fully used at all times. It is very
important that system operators are properly educated and trained to meet the challenge of increased complexity in
system operations.
7.3 Strategies to Develop Improvement Alternatives
An important principle in developing alternatives is for each alternative to provide the same level of service and the
same level of infrastructure renewal. For example, in evaluating a conveyance alternative that increases peak WWFs
to the WWTP beyond its existing capacity, some provision for how the flow will be handled at the plant must be
included. As an alternative, storage and/or RDII reduction measures may reduce peak flows to the WWTP. To
properly compare these alternatives, the conveyance alternative should include either wet-weather storage at the plant
and/or increases in the WWTP capacity so the level-of-service of the alternatives is equal.
Given this principle, the following approach to develop improvement alternatives, as illustrated in Figure 7-1, is
recommended. The first three alternatives suggested for simulation with the system model are alternatives that
completely rely on: (1) conveyance (Alternative Al); (2) storage (Alternative A2); and, (3) I/I reduction (Alternative
A3.) While these alternatives may or may not be the practical solutions in meeting the planning criteria, hydraulic
model simulation of each of these three alternatives will be useful to develop understandings of the relative benefits of
each improvement type in various parts of the system. These alternatives provide the outer boundaries of the
triangular universe of possible improvement solutions. The best solution will reside somewhere within this triangular
universe, and will likely include a combination of all three improvement alternatives. Guidance in simulating each of
these types of improvements follows.
01-
No I/I Reduction
A3- / \/ \ la-
in Reduction O <> <> Storage
B2
Ma Conveyance
Figure 7-1. Triangular universe of possible wet-weather improvement solutions.
7.3.1 The Conveyance Improvement Alternative
The intent of the conveyance improvement alternative is to determine the sizes of replacement or parallel gravity
sewers and/or pump stations and force mains needed to convey all wet-weather flows for the planning storm to the
existing treatment facility. If this alternative results in peak WWFs at the wastewater treatment plant that exceeds its
existing capacity, then either new storage at the plant or the addition of equalization storage at the plant would be
required as part of this alternative. This alternative is typically simulated with the system model assuming that the
7-70
-------
basic system configuration remains the same. For example, gravity sewers would be installed at the same slopes as
the existing system and pump stations would discharge to the same locations as currently used. However, these
assumptions may be changed as the alternative is refined.
Pipes can be sized by first routing WWFs through a model system where pipe sizes are increased to a size that is
larger than that required to convey flows. Peak WWFs at each modeled pipe may then be exported to a spreadsheet or
other design tool to compute replacement and/or parallel pipe sizes needed. In developing cost estimates for this
alternative, the condition of the existing sewers should be considered to determine whether a complete replacement, a
parallel sewer, or a combination of parallel sewer and rehabilitation of the existing sewer lines is appropriate.
7.3.2 The Storage Improvement Alternative
The storage improvement scenario, in its purist definition, would consist of locating a flow equalization storage
facility in the vicinity of every SSO or basement backup. In practicality, storage facilities may be oversized in
upstream locations to provide downstream benefits. On the other hand, downstream storage facilities can reduce
hydraulic gradelines sufficiently to eliminate upstream surcharging and overflows.
As described in Chapter 6, continuous simulation of sewer system performance over a long-term period of rainfall
records that considers the impacts of back-to-back storm events is particularly important when sizing storage
facilities. Storage facilities may take several days to drain after a storm event, and subsequent storms may occur
during this period. These back-to-back storms may cause more impact on the system than one single event, even if
the single event is larger. Hence, continuous simulation is recommended for sizing storage facilities to control
WWFs.
Figure 7.2 shows an example of how the result of a continuous simulation is presented for evaluating the benefits of
storage facilities. A series of continuous simulations is performed with various sizes of storage to reduce overflow
frequency. The results are expressed in terms of an average annual overflow frequency. After a number of
simulations that provided results above and below the target annual overflow frequency, the results were plotted and a
curve was fit to the points.
In this example, the addition of 7MG of system storage at this location would reduce the predicted annual average
overflow frequency to one-half overflows per year, or one overflow every two years.
Similar to the situation discussed in the conveyance improvement alternative, the analyst must consider if any
additional improvements are needed to achieve the equal level of service as other improvement alternatives. For
example, if a conveyance alternative included replacing an aging trunk sewer that has a history of structural problems,
then the storage alternative may require rehabilitation of this trunk sewer. This would be done using a structural liner
to provide an equal level of service of the final solution. Including all common improvements needed to provide an
equal level of service is an important principle in alternatives analysis. This will allow a reasonable comparison of
alternatives, even though the individual components in the selected alternative may be implemented in phases (e.g.,
the storage facility is constructed in year two of the plan and the trunk sewer is rehabilitated in year 10.)
7-71
-------
5-,
A *;
1 4
<&
>
tu
§Q c _
o.o
t
0
£ ^ -
_
3
c 9 c _
<
o
i- 0 _
QJ «1
£
i 1 c -
•O
Of
I 1 -
£
Q.
Oc
•
0 1
\
\
\
\
(
\
*
'••--
2 3 4 5 6 7 E
'- —
_
_
9 10 11 12 13 14 15 16 17 18 19 20
Storage Volume (MG)
Figure 7-2. Example results showing benefits of storage in reducing overflow frequencies.
7.3.3 The I/I Reduction Improvement Alternative
The I/I reduction improvement alternative would conceptually consist of rehabilitating only the sewers required to
reduce I/I sufficiently to eliminate all overflows and/or basement back-ups. This is difficult to determine, especially
at the early planning level of a wet-weather improvements program. Therefore, in practicality, the analyst typically
must make assumptions of how a rehabilitation program would be implemented. The available flow monitoring and
sewer inspection data are generally used to identify sewers where rehabilitation would be effective.
The R-value is a useful parameter in setting priorities for sewer rehabilitation. In the hydrologic system model, each
sewershed is assigned an R-value or a range of R-values to represent RDII contributions under various seasonal or
antecedent moisture conditions. These R-values are determined from directly available flow monitoring data at a
monitored sewershed or from model development and calibration processes for non-monitored sewersheds. The latter
would assign R-values from surrogate areas predicted to have similar RDII responses as the monitored sewersheds
because of their similarities in characteristics, such as sewer age, construction practices, soils characteristics, and/or
maintenance and repair histories. The rehabilitation work would begin in sewersheds that have the highest R-values,
and are located upstream from overflows and/or basement backups.
Sewer rehabilitation may be simulated in the hydrologic model by reducing the R-values. Ideally, the analysis can
base R-value reductions on system-specific flow monitoring data that is available for other similar areas having
undergone rehabilitation and monitoring during pre- and post-rehabilitation conditions. In reality, these data are often
not available. If this type of data is not available, the utility must be encouraged to begin a pilot sewer rehabilitation
program and collect both pre- and post-sewer rehabilitation flow monitoring data. While reasonable assumptions can
be made in planning stages for the purposes of alternatives analysis, actual results can be highly system specific, and
assumptions should be validated as early in the program implementation as possible so that the direction of program
improvements can be confirmed.
R-values may be reduced in the hydrologic model using a variety of approaches. The following represent the most
commonly used methods and assumptions:
7-72
-------
• R-values are reduced by a given percentage, usually ranging from 30 to 70 percent. To achieve a higher
percentage reduction would require a comprehensive sewer rehabilitation effort, including a program to
address RDII from lateral sewers. A comprehensive sewer rehabilitation effort typically consists of
rehabilitation of every foot of sewer, every manhole, and lateral connections at a minimum. Unit hydrograph
parameters (Ri, R2, and R3) may be reduced evenly, or more weight placed on reducing Rj. The available
documented results of sewer rehabilitation programs in systems similar in characteristics to the system being
analyzed would be helpful in making the determination. Typically, T values and K values in the unit
hydrograph model are not changed.
• R-values are reduced from the current value to a target value. Typically, target values may be an R-value of 2
percent, representing the results of a comprehensive sewer rehabilitation program that includes removing
private-server RDII contributions. A target of 5 percent may represent the results of a point repair approach
aimed at reducing major inflow sources. As before, unit hydrograph parameters may be reduced evenly or
more weight placed on reducing R^ T values and K values are typically not changed.
In analyzing future scenarios, R-values may be increased in some areas to reflect deterioration of sewers. The amount
of increase should be judged based on the utility's willingness to commit to an ongoing, annually funded sewer
rehabilitation program that targets RDII reduction on a system-wide basis.
The results of the RDII reduction scenario will help determine where in the system wet-weather problems can be
addressed with a sewer rehabilitation alternative and where this may not be practical. For example, if wet-weather
problems cannot be addressed with RDII reductions of more than 70 percent or to an R-value of 2 percent, then sewer
rehabilitation alone is likely not a practical wet-weather improvement solution.
7.3.4 The No RDII Reduction Improvement Alternative
The above discussions of three 'A' alternatives provided good perspectives on the relative benefits of individual
conveyance, storage, and RDII reductions in developing an overall improvements strategy. The next three
improvement alternatives exclude one of the three major improvements and include a pairing of the other two. The
alternatives are referred to as "B" alternatives (see Figure 7-1) and include no RDII reduction alternative (alternative
Bl), no conveyance alternative (alternative B2) and the no storage alternative (alternative B3.) This combination of
alternatives is more commonly practiced by municipalities.
The no RDII reduction alternative is important to consider. This is because measurable RDII reduction in large
portions of a sewer system is difficult to achieve in a short period. In addition, there are potentially significant
variabilities and uncertainties in the actual quantification of RDII reduction through sewer rehabilitation. Thus,
alternatives relying on RDII reduction have some degree of risk in not being successful. Therefore, other alternatives
have a higher probability of success in meeting the intended goals in a shorter period.
For this reason, utilities tend to choose conveyance and/or storage as the primary alternatives, to meet a given
planning criteria (e.g., controlling wet-weather flows up to a two-year return frequency.) Utilities also include
targeted sewer rehabilitation for RDII reduction. If RDII reduction is successful, facilities originally designed will
fail less frequently and are able to handle larger storm events.
7.3.5 The No Conveyance Improvement Alternative
The no conveyance improvement alternatives represent a situation where it is difficult to locate or construct new
pipelines because of development or environmental issues. Because storage will be a significant part of this
alternative, use of continuous simulation for assessing improvements is recommended.
RDII reduction through sewer rehabilitation in strategic sewersheds upstream of the proposed storage facility can
reduce the required size. As illustrated in Figure 7-3, sewer rehabilitation in one upstream sewershed reduced the
amount of storage required to meet a given overflow return frequency. In this case, a storage facility of
approximately 9 MG was required to meet an overflow frequency of 0.5 overflows per year (once every two years),
7-73
-------
on average. With sewer rehabilitation in one sewershed, assuming that R-value is reduced to 2 percent, the required
storage to meet 0.5 overflows per year was reduced to 2.5 MG.
I
I
.5
14
12
10
T3
£
a
.
Storage (MG)
-Without Rehab -*-With Rehab in 34A2
Figure 7-3. Required storage volume with and without sewer rehabilitation.
7.3.6 The No Storage Improvement Alternative
A no storage improvement alternative consideration will help quantify the relative benefits of storage as compared to
conveyance and RDII reduction alternatives. In general, for RDII reduction to be cost-effective it must be sufficient
to eliminate the need for other conveyance improvements. Usually, developing the cost-effective no storage
alternative includes a combination of conveyance and RDII reduction. This would be based on where in the system
that each of these improvements was found to be the most cost-effective.
7.3.7 Additional Improvement Alternatives
From the analysis results of the six previously described improvements alternatives, the relative benefits and
feasibility of conveyance, storage, and/or RDII reduction throughout the various problem locations of the sewer
system can be determined. With this information, an optimized combination of conveyance, storage, and/or RDII
reduction alternatives can be developed, considering other cost and non-cost factors.
7.4 Applying the SSOAP Toolbox to Evaluate Improvement Alternatives
The sewer system modeling and the establishment of baseline conditions are critical in identifying and screening
potential system improvement alternatives. Combined with system-specific knowledge developed during the baseline
characterization under existing and future growth conditions, the applicability of each alternative can be properly
assessed. The SSOAP Toolbox is designed to help screen potential alternatives and identify the most promising ones
to improve system performance cost-effectively.
7-74
-------
SWMM5 is used to evaluate the hydraulic consequences of each scenario. Model input must be changed to reflect the
specifics of each alternative. As an example, SWMM5 was applied to a hypothetical community with three
sewersheds and a downstream WWTP that experiences system surcharges and overflows. One of the selected
scenarios may include the following improvements:
1. Sewershed A: Increase the pump station and downstream force main capacity.
2. Sewershed B: Disconnect the foundation drains to sanitary sewers, rehabilitate the selected manholes and
sewers, and construct parallel sewers at selected locations that experience hydraulic bottlenecks.
3. Sewershed C: Construct a storage facility upstream of a pump station to attenuate the peak flow to
downstream sewer system.
4. Construct a storage facility at the WWTP to equalize the flows delivered from the collection system prior to
discharging to the plant head works.
These facilities are initially sized and refined using the model until the desired system performance is achieved.
Quality checks must be performed to assure that the proposed improvements are properly represented in the model
network.
The model also provides a means to determine the impact of an upstream system improvement on downstream
facilities. For example, constructing a relief sewer or adding pump station capacity may simply transfer the problem
to a downstream sewer reach.
For RDII reduction (i.e., sewer rehabilitation and inflow reduction) projects, users must estimate the potential
reduction in RDII and reflect that in model input for the R-value, keeping in mind that the R-value is not overly
reduced.
SWMM5 model applications may be driven by a design storm or selected continuous precipitation records to evaluate
the performance under various conditions. The simulations and their results must be reviewed carefully for accuracy
to judge the effectiveness of a given alternative in achieving the performance objective. The model results for each
scenario A are then compared with the baseline conditions. Similarly, other potential improvement scenarios can be
developed and analyzed.
Model scenarios for various improvements can be developed and tracked using the SSOAP Toolbox. Once all the
potential scenarios are analyzed in SWMM5, the results from all the scenarios can be imported to the SSOAP System
Database and compared with the baseline conditions. These comparisons can also be performed outside SSOAP
using spreadsheets or other means. Chapter 5 and the SSOAP Toolbox User Manual describe the use of this feature.
7.5 Developing a Wet-Weather Management Plan
The ultimate outcome of a sanitary sewer system wet-weather management plan study is a set of system
improvements for the municipal planning area that meet the system performance objectives under both current and
future flow conditions. These improvements should then be incorporated into the utility's overall capital
improvement program. Sewer system owners may need to break the improvements into phases if the plan cannot be
implemented as a single project. Project phasing must consider the order and priorities of improvements, so that they
do not decrease the level of service. The plan should estimate the costs of the improvements and provide an
implementation schedule that considers factors such as funding, permitting, design, bidding, and construction. The
future budget may be prepared considering an inflation factor. The assumed inflation factor should be documented so
that adjustments may be made when the actual inflation rate differs from the one used.
Wet-weather management plans should incorporate a process to evaluate the full range of improvement alternatives
using the categories discussed in this chapter. The evaluation process should include a facilitated workshop process
for, at a minimum, representatives from the utility's engineering, operations, management, financial, and customer
service divisions. In some cases, it may be appropriate to include other stakeholders: environmental groups,
regulators and other customer interests, such as homeowners, commercial interests, and major industrial
7-75
-------
representatives. This inclusive approach may lead to new and innovative ways of solving problems.
Before convening a broad stakeholders group, it will be necessary to perform system diagnostics and analysis for a
full range of planning scenarios to develop practical, effective alternatives for the group's consideration. Stakeholder
involvement is beneficial in the development of alternative evaluation criteria (see Figure 7-4), which should be
considered in the final selection of alternatives. Examples of such criteria include:
Alternative Evaluation Criteria
D ::•;:: I Effccttraia J
• ConfiuinRv Interests Q Regulatory Issues
a EnvironmentsiImpacts • ImpienientaHan Icsuex
Figure 7-4. Alternative evaluation to involve stakeholders.
• Capital and operational costs
• Effectiveness
• Community interests (e.g., safety, aesthetics, odors, public disruption)
• Environmental impacts
• Reliability of improvements
• Ease of operation and maintained
• Ease of expansion
• Compatibility with regulatory requirements
• Constructability/implementation issues (e.g., utilities and easements)
Evaluation, criteria can be weighted according to stakeholders' input. The following describes the steps commonly
used in evaluating alternatives.
Step 1: Determine magnitude of the problem using the SSOAP Toolbox and modeling analyses.
Step 2: Review and screen applicable alternatives.
Step 3: Prepare matrix of feasible scenarios.
Step 4: Use model to determine the level of effort required for each scenario, including their downstream impacts.
Step 5: Estimate cost of each scenario.
Step 6: Determine most cost-effective scenario for the problem area/design storm.
Step 7: Repeat procedure until a series of cost-effective scenarios for varying design storms are developed for
each problem area and their costs and benefits are assessed.
Step 8: Consider other non-cost evaluation criteria.
Step 9: Compare and evaluate alternatives.
Step 10: Identify the recommended alternative.
7-76
-------
It is important to recognize that interrelationships and differences among problem areas in a sanitary sewer system
can be very complex. The most cost-effective solution for a certain area may not address other problem areas. The
evaluation process can address this complexity by structuring the alternatives analysis in an iterative process that
allows specific problem areas to be revisited in a broader context. This process allows the recommended
improvements to be customized to meet unique needs.
One of the most difficult aspects of developing a sanitary sewer improvements plan is the question of sewer
rehabilitation benefits. If a sewer system is subject to wet-weather capacity problems, then the reduction in RDII and
GWI through sewer rehabilitation should be considered as one element of the overall improvement plan. At issue,
however, is the effectiveness of sewer rehabilitation in reducing RDII and GWI, as well as the cost of implementing
the rehabilitation. Unfortunately, rehabilitation benefits are highly system-specific and depend on a number of
factors, including system age, type of rehabilitation, construction practices, relative importance of private-side RDII
and GWI sources, and local economic conditions. Therefore, the best way to understand the benefits of a sewer
rehabilitation program is to implement a program and measure its results. This may be accomplished by evaluating
the results of past or current sewer rehabilitation work. If that is not available, the planning process may include
implementing pilot sewer rehabilitation projects in parallel with other planning activities. At the very least, the
municipality should find one or more case studies from other municipalities that have implemented sewer
rehabilitation and use their results as assumptions in the planning process until actual local projects can be completed
and documented.
Generally, a sewer system capital improvement program needs to include sewer rehabilitation so that the
infrastructures may be sustained over a long term. The approach used to perform sewer rehabilitation depends on
whether the primary objectives of the rehabilitation are to increase or restore capacity, to reduce RDII and GWI, to
correct structural deficiencies, to reduce maintenance costs, or a combination of these objectives. The amount of
capital available for sewer rehabilitation will vary considerably by system, but financial analysis methodologies are
available to determine the possible return from a given level of investment. This type of analysis stems from the EPA
construction grants analysis procedure (EPA, 1975), which compares the cost for performing sewer rehabilitation to
the costs of transporting and treating the RDII and GWI entering the system. This method has been updated and
enhanced through a recent project by WEF, scheduled for completion in 2008. This WEF study considers the reduced
risks and maintenance costs achieved through sewer rehabilitation. Through these analyses, the right combination of
sewer rehabilitation and other system improvements, such as capacity upgrades and system storage, can be developed.
7-77
-------
Chapter 8 A Case Study - Ann Arbor, Michigan
This chapter provides a case study to demonstrate how the RDII methodology described in this technical report has
been effectively used in SSO planning and analysis. This case study serves to illustrate considerable challenges in
operating sanitary sewer systems and in planning and analysis for SSO control. More detailed descriptions of the case
study and engineering analyses are available in the following references: CDM (2001), Sherman, et al. (2002),
Sherman, et al. (2005), Sherman, et al. (2006), and Stonehouse, et al. (2005.)
8.1 Introduction
Controlling SSO and basement backups is important to the City of Ann Arbor, its customers, and its regulators. SSOs
are addressed by reducing flows into the system, offsetting any new flow entering the system through a developer
mitigation program, and increasing conveyance capacity for strategic reaches of sewer. Inflow from private
properties, attributed primarily to footing drains, was determined to be the single largest source of flow under wet-
weather conditions. Footing drain flow monitoring provided a greater understanding into the variability and
magnitude of this flow. A system-wide model was developed to characterize the hydrology and the collection system
hydraulics. This model allowed a system-wide determination of where improvements were needed; prioritization of
the footing drain disconnection (FDD) program; and application of a developer mitigation program during which
flows from new development are offset by reduced flow in other locations. The approach taken in the City of Ann
Arbor is somewhat unique in that:
1. The City recognized that footing drains are a major source of the extraneous flows entering the system and
there was real benefit to all users of the system to spend public monies to address the footing drains on private
property.
2. The City passed an ordinance requiring that footing drains be disconnected as a legal basis. However, the
City's greatest success in compliance was through its public outreach and educational efforts that resulted in
an informed and participating citizenry.
3. The City also requires that developers of new projects that would add flow to the sanitary sewer system,
offset these flows by reducing existing flows or provide additional capacity to the collection system. The
sewer model was reconfigured to specifically represent the development and estimate the additional flow
added. Then, flows were removed by disconnecting the appropriate number of footing drains from the
collection system at the developer's expense.
Figure 8-1 shows the City's sanitary sewer system and location of its WWTP. The system has the following
characteristics:
• Population Served - 114,000
• Tributary Area - 21,900 acres
• Citywide Model - 11,000 pipes (390 miles)
• Base Flow-19 MOD
SSO occurs during large storm events at the WWTP. Furthermore, precipitation-induced stresses on the sanitary
sewer system result in a number of properties experiencing basement backups. A study investigating five areas of
concentrated basement backup incidents was performed in 2000 and identified footing drains as a significant
8-78
-------
contributor to the wet-weather response observed in the sanitary sewer system. In 2003, the City of Ann Arbor,
Michigan and the Michigan Department of Environmental Quality (MDEQ) agreed to an Administrative Consent
Order (AGO) that serves as the regulatory basis for the City's SSO control program.
Pump Stations
Non-Tributary Area
Tributary Area
Figure 8-1. Ann Arbor sanitary collection system.
8.2 Hydrology and Hydraulics
The computer model of the City's sanitary sewer system was developed during the third and fourth quarters of 2003.
The model helps City staff plan for infrastructure changes and improvements by realistically simulating sewer flows
under both existing and future conditions. The model drew information from a GIS database to support development
of the hydrologic portion of the model. It also provides collection system attributes to characterize the hydraulic
portion of the model. The project team gathered and organized extensive data on the structure of the City's sanitary
sewer system. It also collected data on the response of the system, that is, the quantity of flow occurring in its various
pipes, and their ability to handle that flow, during dry-weather and storm conditions.
The GIS of the collection system, the DWF data and the wet-weather response information were combined to provide
the basis for a computer model of the sanitary sewer system. The modeling software selected for this project was the
8-79
-------
SWMM version 4.4h with some minor modifications to array sizes. DHFs Mike SWMM, a commercial package,
provided some of the pre- and post-processing functionality. The model has since been converted to SWMM5.
The model development included an initial setup phase, during which connectivity and pipe attributes underwent a
thorough review of every profile. Those cases with suspect attributes were investigated further by researching GPS
field records (including photos) and where necessary, additional field inspection. DWFs were then incorporated into
the model. Finally, wet-weather response parameters were added. These parameters were adjusted to provide a
consistent model response during the calibration and validation processes.
8.3 Data Collection
To determine DWF and evaluate the response of the collection system to wet weather, a flow metering network was
installed. Flow and water-level data were collected during a six-month period from April to September 2003. The
network consisted of 40 flow meters installed throughout the City. Some of these meters were installed to evaluate
the response from specific tributary areas, while others were installed in the interceptor sewers to allow overall
measurement of flows generated in the areas of the City located north and south of the Huron River.
To understand the water levels in other areas throughout the sanitary collection system, 40 inexpensive peak-stage
recorders were also installed to support the six-month monitoring program. At the end of the project, the monitors
were left in place to support analysis in the event that a very large storm would occur at a later date. Periodic data
were gathered on the peak water elevations reached at these locations to enhance understanding of the wet-weather
response of the system. Additionally, a network of 10 tipping bucket rain gauges was used to measure the amount of
rainfall to which the collection system was responding. Of these rain gauges, five existed as permanent gauges and
five additional gauges were installed (two of which were permanent.) Doppler radar-based information (1 km x 1 km
resolution) was used to supplement the rain gauge network with more detailed geographic coverage of rainfall data to
support model calibration and verification. Doppler radar-based information (2 km x 2 km resolution) was also
obtained for the large June 24-26, 2000 storm event to support analysis of the relatively widespread basement
flooding produced by this event.
8.4 Development of System Response Parameters
Collection system response was defined for dry- and wet-weather conditions for dormant and growth seasons. Once
the initial framework of the model was established, the next step was to determine typical DWF throughout the City
for both dormant and growth seasons. Metering data were analyzed using CDM's Sewer Hydrograph Analysis
PackagE (SF£APE) program to establish DWF and RDII days. The SF£APE program allows the decomposition of
metered flow into various components, including GWI, BWF, and three unit hydrographs used for defining the shape
and volume of the RDII. SHAPE functionality has since been integrated into the SSOAP Toolbox. Measurements of
flows at the wastewater treatment plant, supplemented by a comparison to water billing records, provided the initial
estimate of DWF. More importantly, they provided a basis on which DWF from metering could be disaggregated to
upstream subsewersheds. That is to say that DWFs determined for 40 metering locations were disaggregated to more
than 2,600 upstream subsewersheds in proportion to water billing records.
Flow information was also compared to the observed rainfall to determine wet-weather response at each meter
location. Response curves created for each meter showed seasonal changes, consistent with previous flow analyses in
Ann Arbor and elsewhere. The curves clearly indicated two different seasonal characteristics in the system's
response to rainfall, with a marked change from one to the other season. Figure 8-2 illustrates a clear break between
seasons occurring in late May 2003. A second transition was not monitored as part of this project but would occur in
the fall. These two types of collection system behavior are referred to as the dormant and growth season responses.
During the dormant season, with its lack of vegetation, the system responds to wet weather with a significantly greater
volume of sewer flow than in the growth season, when vegetation reduces the amount of rainfall entering the
collection system. The MDEQ defined growth season from April 1st through October 31st, so for regulatory
simulations, parameters were adjusted to appropriately represent these requirements.
8-80
-------
250000
10 15 20 25
-50000
Cummulative Rainfall (in): March to October 2003
Figure 8-2. Seasonal breakpoint.
The general hydrologic model centers on using the RDII parameters, R, T, and K, for up to three component
hydrographs. The T and K parameters are listed in Table 8-1, whereas, the area-weighted average R-values varied
from 1.6 to 4 percent for the growth and dormant seasons, respectively. For this work, two component hydrographs
were found to be adequate: one defines the more direct response, or inflow; and the other defines the delayed
response, or infiltration. The R values quantify the response volume and shape and the T and K parameters further
refine the response shape. Another aspect of RDII is illustrated in Figure 8-3, that of the initial abstraction. Initial
abstraction is the ability of surfaces and soils to hold water with little or no response observed in the collection
system. The maximum initial abstraction (V0) noted in Figure 8-3 represents the amount of rainfall below which the
collection system will not respond if dry antecedent moisture conditions apply. Initial abstraction varies between zero
for very wet antecedent moisture conditions and V0 for dry antecedent moisture conditions. The maximum value also
varies seasonally with growth season often as much as twice that of dormant. Furthermore, the growth season
response volume can be markedly lower than dormant season conditions for the same volume of rainfall. The
seasonal variation in initial abstraction and response volume is similar to that observed in many sanitary systems.
Table 8-1. RDII TK Parameterization
Areas Affected
General Application (except
specific areas noted below)
General Application (except
specific areas noted below)
*03B, 17C
*06A
*07A, 07B, 09D
*13A
Footing Drains
Yes
X
X
X
X
X
No
X
X
X
X
X
Component Hydrograph Shape Parameters
TX
2
2.4
1.2
5
3
2
K!
1.4
1
1
1
1
2
T3
5
5
5
10
5
8
K3
18.2
18.2
18.2
8.6
18.2
3
* Areas where flow monitoring data were available but not shown in Figure 8-1.
8-81
-------
Q
or
0.04
0.03
0.02
0.01
Dormant
Growth
Dormant
V0 Growth
Rtntal Growth
Rainfall (in.)
Figure 8-3. Seasonal responses to rainfall relationships.
8.5 Capacity Assessment
Once calibrated, the model was used to analyze Ann Arbor's sanitary sewer system. Several scenarios were created
and modeled to determine where the collection system needed improvements. The analyses performed included
modeling strategies to prevent SSOs in the Swift Run trunk and assessing the effect of the FDD program to determine
when the collection and treatment systems would comply with SSO requirements. Figure 8-4 illustrates surcharging
in terms of depth from ground surface to the HGL. High HGL relative to the ground, as determined by the model,
coincides well with the locations of basement flooding.
A second part of the modeling analysis work was to review the system from a level-of-service perspective. This
effort included simulating two large historical storm events in August 1998 and June 2000, which caused widespread
basement backup problems. This analysis resulted in recommendations for system improvements to provide
protection to City residents from future similar events. The model allowed various improvements under various FDD
program conditions to be evaluated. For example, the model is used to evaluate the effect of new development on the
collection system and the effect of removing flows as part of the developer-mitigation requirement.
8.6 SSO Control Program
The City-wide model was used to evaluate the FDD program to reduce RDII in the sanitary sewers. By adjusting the
model to account for the number of properties for which FDD has been performed or will be performed in the future,
the model helps determine the resulting effects on the wastewater collection system. This capability assists in
planning current and future phases of that program.
8-82
-------
Figure 8-4. System capacity limitations coincide with basement flooding incidents.
To eliminate SSOs, flow must be removed from the sanitary sewer system or the capacity of the system be increased.
The requirements for compliance with the ACO include:
• Disconnect footing drains as the primary method of reducing flow. Figures 8-5 and 8-6 show some of the
exterior and interior work associated with FDD.
• Monitor a representative sampling of disconnected footing drain flow to confirm the removal of flows from
the sanitary sewer system. A hydrograph of footing drain flow is shown in Figure 8-7.
• Monitor sewer flow and model system-wide hydrology and hydraulics to certify that the corrective action
plan meets the control objective.
• Implement a mitigation program to offset flow produced by new construction.
• Submit annual report to document compliance with the above requirements.
Figure 8-5 shows work on the curb drain. A curb drain is required in most cases to serve as a collector pipe by which
all disconnected footing drains for properties along one side of the street can connect. This curb drain then taps into
the existing stormwater system, often at the back-of-curb side of existing catchbasins as shown in Figure 8-5. Figure
8-6 shows a finished sump, to which the footing drain disconnected from the sanitary sewer is reconnected prior to
pumping from the basement through a pipe that discharges into the curb drain. Figure 8-6 also shows two persons
performing a drawdown test to assess the average pumping rate at one of many locations instrumented with data
loggers that record pump ON/OFF cycling. The pumping rate and the ON/OFF data can be used to determine the
flow directed to the sump from the footing drains. An example of this footing drain flow is shown in Figure 8-7 and
8-83
-------
is a useful means of quantifying flows removed at the source, in this case as a response to a storm event. Many
footing drains also produce flow under dry-weather conditions, that once removed, also provide a benefit in the form
of reduced treatment costs.
Figure 8-5. Curb drains convey disconnected footing drain flow to the storm water system.
Figure 8-6. Testing pumping rate to support monitoring of a disconnected footing drain.
8-84
-------
E. U
- ~ 02
E .9-
"*~ i" n «
f
1
r r •
1
i
- r T-
5/21/04 5/22/04 5/23/04 5/24/04
Figure 8-7. Footing drain monitoring provides evidence of flows being removed.
Prioritizing FDDs relied heavily on modeling results. Approximately 1,000 properties have been disconnected.
Figure 8-8 illustrates FDD progress. In the areas where the FDD schedule is yet to be determined, there are properties
that do not have connected footing drains. These properties are primarily identified by the year constructed or other
factors, such as structures on slabs instead of basements.
Many factors influence the prioritization of work. Homes selected for disconnection were based on incident reports,
model results, visual surveys for properties at similar elevations, and where curb drain construction made sense for
both engineering and cost. The severity of backups and concentrations of effort were also factors. Modeling was
used in some areas to confirm that properties were not at risk for basement backup. In general, properties with
connected footing drains, which are at risk for basement backup, will have check valves installed. The intent is to
ultimately reduce system stresses to the point that the check valves are no longer necessary. Therefore, after focusing
on properties at risk for basement backup, footing drain flows will be removed to reduce hydraulic grade line in
8-85
-------
critical areas.
FDD Complete
FDD Scheduled
FDD To Be Determined
Non-Tributary Area
Tributary Area
Figure 8-8. Footing drain disconnection progress.
As part of this subsequent phase, the City found that multi-family residential buildings offered the potential to remove
footing drain flow at less cost per gallon removed due to a greater length of footing drain (related to building
perimeter) per disconnection made. All flow removed has the added benefit of not requiring treatment and further
reduces the risk of SSOs at the WWTP.
For large storm events on the order of 3.6 to 3.9 in. during 24 h, estimates based on extrapolating footing drain flow
monitoring data suggest approximately 4 gpm peak rate is removed for each FDD on average. With approximately
1,000 disconnections thus far, this equates to between 5 and 6 MGD removed during a large storm event. This is a
significant improvement for a system with peak treatment capacity of 29.5 MGD and flow equalization of 16.8 MG.
Although less significant, DWF analysis of footing drain flows indicate that there has been approximately 14
MG/year removed from treatment due to base footing drain flow alone. This represents 75 percent of one day's
wastewater treatment on a typical DWF day.
8-86
-------
In addition to supporting the FDD program, the SSO control program requires that the collection system be managed
to ensure improvements are made and that additional development does not exacerbate existing problems. The
sanitary sewer system model has been used to identify and support specific infrastructure improvements, including:
• Increased capacity for the Swift Run trunk system.
• Optimized pumping at the Lakewood Pumping Station, including pump upgrades.
• Modeling results were provided in GIS and hardcopy format to the City's planning department to allow
consideration of sewer improvements during street repaying and prioritized work. Sewer upgrades are
performed concurrently with street repaying efforts where necessary.
• The model is being used to support an ongoing evaluation of flow control between the north and south Huron
River interceptors to better optimize system operations to effectively use the existing conveyance capacity.
• Developers are required to mitigate flows added to the collection system by performing FDDs at existing
properties. To date, there have been 58 developer projects requiring mitigation of 8.8 MGD based on
expected sanitary flow - multiplied by a factor of 4 - to represent potential peak rate added.
• Evaluate impact of University of Michigan football stadium expansion.
8.7 Public Outreach
The City and its residents participated in a joint task force that oversaw the study, the development of the program,
and effectiveness of solutions implemented. The task force was the first step in reaching out to the public and
involving them in the process. The FDD program included work in five areas in Ann Arbor that have historically
experienced basement backups. These areas primarily comprise the most dense sewer network areas shown in Figure
8-8 that lie in the northeast and southwest of the map. For these properties, the facilities located in basements, which
include floor drains, laundry, showers, and bathrooms, are being protected from future basement backups using check
valves.
As part of construction management services, the program manager scheduled groups of properties to be disconnected
each month. The program manager also scheduled group discussions as part of neighborhood meetings (see Figure 8-
9), individual meetings with homeowners during preconstruction inspections, reviewed estimates from prequalified
contractors, and performed post-construction inspections to ensure that the work met the standards of the City. The
public outreach work included periodic presentations to the City Council and weekly video broadcasts on the local
cable access channel that explained the need for the program and the required steps for success. In addition, the
outreach program consisted of an informational website that explained the project activities, as well as informational
materials provided in the reference areas of all of the local libraries.
The work on private properties included installing new sumps in basements, installing check valves, disconnecting the
footing drains from the sanitary piping, and installing a sump pump and discharge line. This program also included a
flow verification step. For this work, individual sump pump discharges were monitored to better understand the flow
volume and peak rates to be expected once these footing drains were disconnected. This information is useful in
guiding future implementation phases of the program and providing information to homeowners on the reductions in
flows that resulted from performing this disconnection work. Participating property owners rate the program
"excellent" year after year based on feedback from a survey provided to owners post-construction.
8-87
-------
Figure 8-9. Neighborhood meeting with property owners.
8.8 Summary
A computer model of the entire publicly owned sanitary collection system was developed. The model includes the
physical characteristics of the City's sanitary sewers and the wet-weather response characterization for different
seasons. The wet-weather response, RDII, was modeled using SWMM and included the R, T, and K parameters for
all model tributaries. The model was calibrated and validated with available flow and water-level data. It provides a
reasonable estimate of the actual responses expected in the sanitary collection system under a variety of conditions.
The FDD program is underway with approximately 1,000 properties disconnected. This RDII removal program, with
work on private property, requires an effective public outreach program to ensure participants are well informed and
receptive to the changes needed to address system capacity problems and associated basement backups.
The model was used to determine how the sanitary collection system responds to various storm scenarios, and to
understand how the FDD program affects the operation of the system during such storms. The analysis has helped
improve the City's understanding of specific deficiencies in its collection system, so that corrections can be made to
provide a consistent level of service to all customers.
Moreover, Ann Arbor's City-wide model provides City staff with an ongoing planning tool that can be used to
evaluate sewer service for new developments and the impact of such development on the existing system and its
customers. It also helps the City develop effective strategies for continued compliance with environmental
regulations and policies.
8-88
-------
Chapter 9 References
Akan, A.O. and Yen, B.C. (1981). "Diffusion-Wave Flood Routing in Channel networks" Journal of Hydraulics
Division, Proceedings of the American Society of Civil Engineers Vol.107, No HY6, p.719-732, June, 1981.
Brater, E.F. and King, H.W. (1976). "Handbook of Hydraulics", 6th ed. McGraw Hill Book Company, NY.
Bennett, 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.
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.
Camp Dresser & McKee Inc. (COM). (1989). East Bay I/I Correction Program Certification Procedures Manual.
East Bay Municipal Utility District (EBMUD), Oakland, CA.
Camp Dresser & McKee Inc. (COM). (2001). SSO Prevention Study - Final Report, prepared for the Ann Arbor City
Council, June 2001. http://www.cdm-mich.com/aa-sso/FinalReport.htm
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.
Chow, V.T. (1959). Open Channel Hydraulics. McGraw Hill Book Company, NY.
Cunge, J.A. and Wegner, M. (1964). "Integration numerique des equations d'ecoulement de Barre de St-Venant par
un schema implicite de differences finies." La Houille Blanche, No. 1, 33-39.
French, R.H. (1985). Open Channel Hydraulics. McGraw Hill Book Company, NY.
Giguere, P.R and Riek, G.C. (1983). "Infiltration/Inflow Modeling for 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, July 25-28, 1983, Lexington, KY.
Health Education Service (HES). (1997). Recommended Standards for Wastewater Facilities. A report of Committee
of the Great Lake-Upper Mississippi River Board of State and Provincial Public Health and Environment
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.
Lai, C. (1986). "Numerical modeling of unsteady open-channel flow" in: Advances in Hydroscience, V.T. Chow and
B.C. Yen, eds.,Vol. 14, Academic Press, Orlando, FL., 163-333.
9-89
-------
Lai, F.-h. (Dennis). (2007). Review of Sewer Design Criteria and RDII Prediction Methods. A companion CRADA
report by EPA, publication pending.
Loehlein, M.D. and Meeneghan, T.J. (2004). "A Continuous Simulation Approach for Separate Sewer Areas." CHI
Effective Modeling of Urban Water Systems, 2005, Monograph 13, pp. 435-450.
Meeneghan, T.J., Prevost, T.D., and Khan, K.N. (2004). "Impacts of Rainfall Data on Model Refinement in Greater
Pittsburgh" Proceedings of the International Conference on Stormwater and Urban Water Systems Modeling;
Toronto, Canada. (ISBN: 0-9683681-9-0)
Merrill, M.S., Lukas, A., Roberts, C., Palmer, R.N., and Van Rhennan, N. (2003). "Reducing Peak Rainfall-Derived
Infiltration/Inflow Rates - Case Studies and Protocol." Water Environment Research Foundation (WERF) Report
No. 99-WWF-8, Alexandria, VA.
Miles, S.W, Dorn, J.L., and Tarker Jr., R.E. (1996). "An I/I Analysis and Prediction Method to Help Guide Separate
Sanitary Sewer Improvement Programs. " Proceedings of the Water Environment Federation (WEF) Conference
on Urban Wet Weather Pollution: Controlling Sewer Overflows and Stormwater Runoff. June 1996, Quebec City,
Quebec.
Miles, S.W., Shirley, K.J., Tarker Jr., R.E. and Dorn, J. (1999). "How to Defensibly Document I/I Reduction: Two
Carolinas Success Stories." Proceedings of the Water Environment Federation Annual Conference (WEFTEC).
October 1999, New Orleans, LA.
Miles, S.W., Moyer, J. and Ridge, J. (2004). "Building a Better Business Case for Funding Sewer Rehabilitation."
Proceedings of the Water Environment Federation (WEF) Collection System 2004: Innovative Approaches to
Collection Systems Management. August 2004, Milwaukee, WI.
New England Interstate Water Pollution Control Commission. (NEIWPCC). (2003). Optimizing Operation,
Maintenance, and Rehabilitation of Sanitary Sewer Collection Systems."
Sands, H. and Stevens, P. (1995). "Sanitary Sewer Overflows Leave Telltale Signs in Depth-Velocity Scattergraphs."
ADS Environmental Service, Inc.
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.
Sherman, B. J., Rolfe, C., Perala, P.P, Hupy, C., and TenBroek, M.J. (2002). "Implementing SSO Control - Technical
Aspects." Proceedings of the Water Environment Federation WEFTEC 2002 Conference. October 2002, Chicago,
IL.
Sherman, B. J., Perala, P.P, Stonehouse, M.C., Fan, H.P., and TenBroek, M.J. (2005). "Allocating RDII Using
Footing Drain Flow and Other Information." International Stormwater and Urban Water Systems Modeling
Conference, February 24-25, 2005, Toronto, Ontario, Canada.
Sherman, B. J., Perala, P.P., and Stonehouse, M.C. (2006). "Making Sense of Private RDII From Footing Drains."
Alternate Paper, 2006 WEF Collection Systems Conference - Infrastructure Stewardship: Partnering for a
Sustainable Future. August 6-9, 2006, Detroit, MI.
Stonehouse, M. C., Sherman, B.J., Perala, P.P., Wilburn, D. and TenBroek, M.J. (2005). "Modeling Every Sanitary
Pipe in the City: Unique Aspects and Benefits for Ann Arbor, Michigan." International Stormwater and Urban
Water Systems Modeling Conference, February 24-25, 2005, Toronto, Ontario, Canada.
9-90
-------
U.S. Environmental Protection Agency. (EPA). (1975). Handbook for Sewer System Evaluation and Rehabilitation,
Report No. EPA/430/9-75/021, EPA, Washington, D.C.
U.S. Environmental Protection Agency. (EPA). (1999). Combined Sewer Overflows: Guidance for Monitoring and
Modeling. Report No. EPA/832/B-99-002, EPA, Washington, D.C.
U.S. Environmental Protection Agency. (EPA). (2000). "Compliance and Enforcement Strategy For CSOs and
SSOs."
U.S. Environmental Protection Agency. (EPA). (2004). Report to Congress: Impacts and Control of CSOs and SSOs.
Report No. EPA/833/R-04-001, EPA, Washington, D.C.
U.S. Environmental Protection Agency. (EPA). (2005). Guide for Evaluating Capacity, Management, Operation, and
Maintenance (CMOM) Programs at Sanitary Sewer Collection Systems. Report No. EPA/305/B-05-002, EPA,
Washing, D.C.
U.S. Environmental Protection Agency (EPA). (2007). "Storm Water Management Model User's Manual Version 5."
Report No. EPA/600/R-05-040.
Vallabhaneni S, Koran, J., Moiso, S, and Moore, C. (2002a). "SSO Evaluations: Infiltration and Inflow using SWMM
RUNOFF and EXTRAN". Proceedings of the International Conference on Stormwater and Urban Water Systems
Modeling; Toronto, Canada. (ISBN: 0-9683681-6-6)
Vallabhaneni, Vieux E. V., Donovan, S., Moisio S. (2002b), Interpretation of Radar and Rain Gauge Measurements
for Sewer System Modeling. Proceedings, 9th International Conference on Urban Drainage, International Joint
Committee on Urban Storm Drainage of the IWA and the IAHR, Portland OR, Sept. 8-13, 2002.
Vallabhaneni, S., Wride, D, Koran, J., Moisio, S, and Burgess, T. (2002c). Integration of Water Billing Records in
Estimating Base Wastewater Flow Rates to Support Large-Scale Sewer System Computer Modeling. Proceedings
WEFTEC 2002.
Vallabhaneni, S., Donovan, S., and Barton, J. (2003). "Case Study: Flow Monitoring and Site Selection - Cincinnati
MSD System Wide Model Project." Proceedings of the Collection Systems Conference 2003: cMOM Go!, Water
Environment Federation.
Vallabhaneni, S; Vieux, B; Meeneghan T. (2004). "Radar-Rainfall Technology Integration into Hydrologic and
Hydraulic Modeling Projects". Proceedings of the International Conference on Stormwater and Urban Water
Systems Modeling; Toronto, Canada. (ISBN: 0-9683681-9-0)
Wang, E. (2005). "Infrastructure Rehabilitation Management Using Life Cycle Cost Analysis." ASCE Computing in
Civil Engineering, ASCE, New York, NY.
Water Environment Federation (WEF). (1997). Prevention and Control of Sewer System Overflows. Report No:
WEF/MOP FD-17, WEF's Manual of Practice.
Water Environment Research Foundation. (WERF). (2003). Reducing Peak Rainfall-Derived Infiltration/Inflow Rates
- Case Studies and Protocol. Alexandria, VA.
Water Environment Foundation. (WEF). (1999) Wastewater Collection Systems Management. WEF Manual of
Practice No. 7, Alexandria, VA.
Water Environment Foundation. (WEF). (2006). Guide to Managing Peak Wet Weather Flows in Municipal
9-91
-------
Wastewater Collection and Treatment Systems. Alexandria, VA.
Wride, D, Vallabhaneni, S, Chen, M, Johnstone, R, and Moisio, S. (2003). "Using Radar-Rainfall to Improve the
Characterization of the Inflow/Infiltration (I/I) Response." Proceedings of the WEF/WEAT Collection Systems
2003 Conference, cMOM Go!; June 1 - 4, 2003; Austin, TX.
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.
Yen, B.C. (1986). "Hydraulics of Sewers", in: B.C. Yen, ed., Advances in Hydroscience, Vol. 14, Academic Press,
Orlando, FL.
Yen, B.C. (2004) "Hydraulic of Sewer Systems", in: Urban Stormwater Management Tools (Chapter 4: L.W. Mays
Ed.). McGraw-Hill, New York.
Vieux, B.E. and Vieux, J.E. (2004). "Rainfall Accuracy Considerations Using Radar and Rain Gaguge Networks for
Rainfall-Runoff Monitoring." Proceedings of the International Conference on Stormwater and Urban Water
Systems Modeling; Toronto, Canada. (ISBN: 0-9683681-9-0)
Zawacki, J. R., Kotlyar, I., Sherman, B.J. and TenBroek, M.J. (2005). "Effective I/I Removal with Footing Drain
Disconnection as a Solution for SSO and Basement Flooding." Alternate Paper, 2005 WEF Collection Systems
Conference - Sustaining Aging Infrastructure: System, Workforce, and Funding. July 17-20, Boston, MA.
9-92
------- |