vvEPA
United States
Environmental Protection
Agency
Region 5
Great Lakes National
Program Office
536 South Clark Street
Chicago, Illinois 60605
EPA-600/2-79-031a
July, 1979
Section 108(a)
Combined Sewer
Overflow Abatement
Program
Rochester, NY
Volume 1
Abatement Analysis
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and governmental concern about the dangers of pollution to the
health and welfare of the American people. Noxious air, foul water, and
spoiled land are tragic testimony to the deterioration of our natural
environment. The complexity of that environment and the interplay between
its components require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in the problem
solution and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems for the prevention,
treatment and management of wastewater and solid and hazardous waste
pollutant discharges from municipal and community sources, for the preser-
vation and treatment of public drinking water supplies and to minimize the
adverse economic, social, health, and aesthetic effects of pollution. This
publication is one of the products of that research; a most vital communi-
cations link between the researcher and the user community.
The Great Lakes National Program Office, through Section 108(a) of
PL 92-500, enters into grants for the demonstration of new methods and
techniques and for the development of preliminary plans for the prevention,
reduction or elimination of pollution within all or any part of the water-
sheds of the Great Lakes. The Great Lakes National Program Office has
joined with the Municipal Environmental Research Laboratory in carrying
out this research and demonstration project to assist the Rochester Pure
Waters District to eliminate an urban drainage pollution problem to
Lake Ontario.
The deleterious effects of storm sewer discharges and combined sewer
overflows on the nation's waterways have become of increasing concern in
recent times. Efforts to alleviate the problem depend on characterization
of these flows as to both quantity and quality. This report describes
the general approaches available for the control of combined sewer overflows
and their application to the Rochester, New York combined sewer system.
Francis T. Mayo
Director
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Madonna F. McGrath
Director
Great Lakes National Program Office
U.S. Environmental Protection Agency
Chicago, Illinois 60604
-------
EPA-600/2-79-031a
COMBINED SEWER OVERFLOW ABATEMENT PROGRAM, ROCHESTER, N.Y.
Volume I: Abatement Analysis
by
Frank J. Drehwing
Cornelius B. Murphy, Jr.
David J. Carleo
Thomas A. Jordan
O'Brien & Gere Engineers, Inc.
Syracuse, New York
Grant No. Y005141
Project Officers Grant Officer
Richard Field/Anthony Tafuri Ralph G. Christensen
Storm and Combined Sewer Section Great Lakes Demonstration Program
Wastewater Research Division Great Lakes National Program Office
Municipal Environmental Research Region V, Chicago, Illinois 60604
Laboratory (Cincinnati)
Edison, New Jersey 08837
Lawrence Moriarty
Region II
Rochester, New York 10007
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
and
GREAT LAKES NATIONAL PROGRAM OFFICE
U.S. ENVIRONMENTAL PROTECTION AGENCY
CHICAGO, ILLINOIS 60604
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory and the Great Lakes National Program Office, U.S. Environmental
Protection Agency, and approved for publication. Approval does not signify
that the contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names or com-
mercial products constitute endorsement or recommendation for use.
n
-------
ABSTRACT
Pollution abatement analyses, conducted in conjunction with system net-
work modeling studies and supported by combined sewer overflow (CSO) monitor-
ing and sampling, were initiated with the ultimate goal of formulating a co-
hesive and workable Master Plan for CSO reduction and control. The Master
Plan was developed in light of fiscal constraints, sewer system complexities,
necessity for optimized benefits from minimal capital and operating expendi-
tures, and best use policies of the affected receiving waters. The presented
methodology is considered applicable to other urban areas.
Overflow monitoring and sampling data from thirteen CSO locations within
the Rochester, New York Pure Waters District collected during the period
January through December,1975 served as the basis for network modeling
studies-. The USEPA Stormwater Management Model - Version II, Simplified
Stormwater Model, and receiving water models were used to evaluate various
CSO pollution abatement alternatives. Nonstructural , minimal structural,
and structurally intensive alternatives were defined and evaluated by these
models. The nonstructural approach centered around the application of Best
Management Practices (BMPs), e.nd structural alternatives involved evaluation
of conventional storage and treatment options.
Cost benefit analyses of all structurally intensive alternatives were
conducted using optimum treatment process train configurations developed
from pilot plant evaluations, as reported in Volume II of this project's
report.
Preliminary analysis of BMP and minimal structural alternatives indica-
ted that by addressing the major sources of pollution and by eliminating
throttling constraints within the existing sewerage system, a substantial de-
crease in the total annual load of contaminants to the receiving waters from
rainfall induced CSO can be achieved for relatively small capital expendi-
tures. These measures can be initiated within a short period of time, there-
by immediately reducing pollution to the receiving waters while long term
design and construction of more structurally intensive alternatives are un-
dertaken.
This Report was submitted in fulfillment of Grant No. Y-005141 by
O'Brien & Gere Engineers, Inc. under the partial sponsorship of the U.S.
Environmental Protection Agency. Work was completed as of November, 1976.
-------
CONTENTS
Foreword Inside Cover
Abstract ill
Fi gures vi i
Tabl es xi
Abbreviations and Symbols xiv
Acknowledgment xvi
1. Introduction 1
Background 1
Problem definition 1
Purpose of study 3
Project elements 3
2. Conclusions 10
3. Recommendations 15
4. Methodology of Approach to Alternative Analysis 18
General Methodology 18
Technical Procedures 27
5. Overflow Monitoring and Drainage Basin Characterization.. 31
Overflow monitoring 31
Drainage area characterization 37
6. Network Modeling 41
Configuration and type of models used 41
Assumptions in modeling 47
Limitations of available software 49
Model calibration and verification 51
7. Criteria for Alternative Analysis 72
Introduction 72
Rainfall analysis 73
Overflow quality analysis 80
Quality analysis for use in abatement alternative
studies 84
8. Description and Evaluation of Alternatives 93
General 93
Nonstructural Alternatives 93
Minimal Structural Alternatives 106
Structural ly Intensive Al ternatives 124
9. Cost Effective Analysis of Structurally Intensive
Alternatives 141
Alternatives Evaluated 141
Development of Cost Relationships 142
Cost Optimization 142
Cost-Effective Relationships 143
-------
Contents (Con't)
10. Description of Master Plan Configuration and
Abatement Schedule 151
References
-------
FIGURES
Number Page
1 Study Area 2
2 Program Elements and Their Relationship to the Development
of the Master Plan 4
3 Overflow Monitoring and Characterization Effort 6
4 Combined Sewer Overflow System Modeling Effort 7
5 Pilot Plant Effort 8
6 Evaluation Process for Abatement Alternative Analysis 9
7 Diagram for Model Utilization 28
8 Overf 1 ow Locati ons 32
9 Typical Head and Velocity Probe Installation 34
10 Overall Monitoring and Telemetry System 35
11 Concept of Storage/Treatment-SSM Analysis ^
12 Calibration Overflow Hydrograph for Maplewood Park 10/9/75. 54
13 Verification Overflow Hydrograph for Maplewood Park
9/11/75 55
14 Verification Overflow Hydrograph for Maplewood Park
9/20/75 56
15 Calibration Overflow Hydrograph for Carthage 8/29/75 57
16 Verification Overflow Hydrograph for Carthage 8/24/75 58
17 Verification Overflow Hydrograph for Carthage 9/11/75 59
18 Calibration Overflow Hydrograph for Thomas Creek
6/9/75 60
19 Verification Overflow Hydrograph for Thomas Creek
6/5/75 61
vii
-------
FIGURES (continued)
Number Page
20 Verification Overflow Hydrograph for Thomas Creek
6/12/75 62
21 SWMM Node-Conduit Representation of Sewerage Network for
City of Rochester 63
22 SWMM Sewer Network Configuration for QQS-SWMM Comparison... 71
23 Example Curve - Storm Magnitude vs Frequency 76
24 Example Curve - Storm Intensity vs Frequency 76
25 Example Curve - Storm Duration vs Frequency 77
26 Example Curve - Percent of Storms Having Maximum 1-Hour
Intensity vs Hour After Start of Storm 77
27 Rainfall Intensity-Duration-Frequency Curves for Rochester,
NY 79
28 Synthetic Design Storm Hyetographs 81
29 Present-Proposed Tunnel Interceptor System 87
30 Runoff BOD5 Profiles for Various Storms and Street
Cleaning Frequency 96
31 Runoff TSS Profiles for Various Storms and Street
Cleaning Frequency 96
32 Runoff Hydrograph for Various Storms 98
33 Runoff Hydrograph for Various Land Uses 98
34 SWMM Runoff Qua! ity Equation 99
35 Runoff BOD5 Profile for Various Land Uses 101
36 Runoff TSS Profile for Various Land Uses 101
37 Overflow Volume as a Function of Land Use 104
38 Location of Drainage Areas Suitable for the Application of
Nonstructural Alternatives 105
39 Conveyance System Locations Suitable for In-system
Regulation 110
vi i i
-------
FIGURES (continued)
Number Page
40 Location of Constraints in the Existing St. Paul
Boulevard Interceptor Ill
41 Minimal Structural Alternatives Modeling Results (SWMM)
Overflow Hydrograph 113
42 Minimal Structural Alternatives Modeling Results
Overflow Hydrograph Area No. 22 Regulator 119
43 Minimal Structural Alternatives Modeling Results
Flow Projections at the Van Lare STP 121
44 Minimal Structural Alternatives Modeling Results
Total River Overflow vs Regulator Capacity 123
45 Schematic of Structurally Intensive Alternate No. 1 -
Storage First-Flush/Treatment Post First-Flush 125
46 Schematic of Structurally Intensive Alternate No. 2 -
Overf 1 ow Treatment Al ong Ri ver 126
47 Storage vs Treatment Requirements for Varying Overflow
Volumes for Centrally Located River STP 128
48 Storage Volume vs Tunnel Diameter for Structurally
Intensive Alternate No. 2 129
49 Mass Inflow Curve for the River Treatment Facility 131
50 Storage vs Treatment Rate for Varying Overflow
Quantities 132
51 Schematic of Structurally Intensive Alternate No. 3 -
Overflow Treatment at Van Lare STP 133
52 Storage vs Treatment for Varying Overflow Volumes for
Treatment Plant Located at F.E. Van Lare 134
53 Storage Volume vs Tunnel Diameter for Structurally
Intensive Alternate No. 3 135
54 Schematic of Structurally Intensive Alternate No. 4 -
Storage of First-Flush 137
55 Schematic of Structurally Intensive Alternate No. 5 -
Local Overflow Treatment 138
56 Schematic of Structurally Intensive Alternate No. 6 -
Conveyance of Overflow by East-West Interceptor 140
ix
-------
FIGURES (continued)
Number Page
57 Typical Storage-Treatment Cost Optimization 144
58 Cost-Effectiveness Relationships for Subalternatives of
Alternate No. 2 - River Overflow Interceptors with
Overflow Treatment Facility Centrally Located 146
59 Cost-Effectiveness Relationships for Subalternatives of
Alternate No. 3 - River Overflow Interceptor with
Overflow Treatment Facility at Van Lare STP 147
60 Cost-Effectiveness Summary for the Structurally
Intensive Alternatives 148
61 Cost-Benefit Summary for the Structurally Intensive
Al ternati ves 150
'»
62 Rochester Combined Sewer Overflow Abatement Program
Preliminary Implementation Plan 152
63 Total System Overflow Reduction Response to Implementation
of Abatement Measures 155
64 Total System Pollutant Loading Reduction to the Genesee
River Response to Implementation of Abatement Measures...156
65 Genesee River Reduction in Potential Dissolved Oxygen
Contravention Response to Implementation of Abatement
Measures 157
66 Reduction in Potential Ontario Beach Closing Days Response
to Imp! ementati on of Abatement Measures 158
-------
TABLES
Number Page
1 Overflow Monitoring Installations 33
2 Overflow Analysis Schedule 36
3 Summary of Drainage Area Characteristics 39
4 Summary of Drainage Area Characteristics 40
5 Rainfall Characteristics 52
6 Overflow Peak Flowrates - Volumes. 64
7 Regression Analysis - Quality Parameters 66
8 Rainfall Data for Model Comparison 67
9 Subcatchment Data 69
10 Runoff Volumes for Model Comparisons 68
11 Overflow Volumes for Model Comparisons 70
12 Total Rainfall 74
13 Number of Raindays 74
14 Rain Per Storm in Inches Per Rain Day 75
15 Rainfall Intensity-Duration Equation Constants 78
16 Summary of Drainage Area CSO Quality 82
17 Zi/Z2 Confidence Intervals 83
18 Drainage Area Overflow Analysis for 2 Yr-2 Hr Design
Storm 85
19 Ranking of Drainage Areas Per Mass Loadings ,. 84
20 Overflow Qua!ity by Time Increment 86
xi
-------
TABLES (continued)
Number Page
21 Overflow Volume by Time Increment 2 Yr-2 Hr
Design Storm ............................................. 88
22 Overflow Pollutant Loadings by Drainage Area and Time
Increment For 2 Yr-2 Hr Design Storm ..................... 88
23 Pollutant Loadings by Time Increment For 2 Yr-2 Hr Design
Storm East Side Overflows ................................ 90
24 Average Annual Loading to Genesee River Projected by
Simplified Stormwater Model .............................. 90
25 Number of Required Swirl Units by Drainage Area ............ 91
26 Overflow Storage Capacity Required to Retain First-Flush
2 Yr-2 Hr Storm ........................................... 91
27 Swirl Unit Solids Removal Efficiency of First-Flush
2 Yr-2 Hr Storm ........................................... 92
28 East Side Pollutant Loadings for 2 Yr-2 Hr Storm ............ 92
29 Influence of Sewer Maintenance .............................. 100
30 River Overflow Projected by Simplified Stormwater Model
For the Proposed Tunnel Interceptor System ................ 112
31 Effect of Upgrading the St. Paul Interceptor For 2 Yr-2 Hr
Design Storm .............................................. 114
32 River Loading Tunnel Interceptor System- Projected
by Simpl if ied Stormwater Model ............. . .............. 115
33 Selective Blockage of Overflows Proposed Tunnel Inter-
ceptor System For 2 Yr-2 Hr Design Storm, .................. 117
34 Selective Weir Analysis-Proposed Tunnel Interceptor System
For 2 Yr-2 Hr Storm ....................................... 120
35 Total River Overflow vs Regulator Capacity Based on
Simplified Stormwater Model
36 Runoff and Overflow Volumes By Drainage Area As Projected
By the SWMM for the 2 Yr-2 Hr Storm ....................... 127
37 Primary Swirl Solids Reduction Treating Entire Overflow For
2 Yr-2 Hr Design Storm .................................. 136
xii
-------
TABLES (continued)
Number Pages
38 West-East Storage-Treatment Balance Simplified Stormwater
Model 139
39 Process Train Configurations 141
40 Process Cost Model s 143
xm
-------
ABBREVIATIONS
LIST OF ABBREVIATIONS AND SYMBOLS
5
ac
ave.
BMP
BOD or BOD
C
cfs
CLFREQ
CSO
D, dia
D2
DD
DIS
DMF
DMHRF
D.T.
f. coliform
F-F
F/S
f. strep
ft2
fir
G
-- acre
G/S
ha
hr
in
K
kg
Ib
nu
nu
mj
MA7CD10
mi2
mil
mgd
mg/1
min
gal
average
Best Management Practice
5 day biochemical oxygen demand at
concentration, units as specified
cubic feet per second
street cleaning frequency
combined sewer overflow
diameter
chamber diameter
dust and dirt
disinfection
dual-media filtration
dual-media high-rate filter
detention time
fecal col iforms
first-flush
flocculation/sedimentation
fecal streptococci
feet
20°C
(sec"1)
square feet
velocity gradient
grams
gallons per day
swirl degritter
hectares
hour
inch
rate constant or thousand
kilograms
Ib
meters
square meters
cubic meters
minimum average 7 consecutive day flow occurring
every 10 years
square miles
million gallons
million gallons per day
milligrams per liter
minutes
kiv
-------
ABBREVIATIONS (Con't)
MPN
NF
no.
O&G
OR
P/S
psf
QD
QQS
r
sec
Set. S
SOR
SS
SSM
std dev
STP
SWMM
t. coliform
TIP
TKN
trtmt
TSS
VSS
v 3
yd3
yr
most probable number
Froude number
number
oil and grease
overflow rate
swirl primary separator
pounds per square foot
design flowrate
quantity-quality simulation
ratio of hour of maximum rainfall to total duration
second
settleable solids
surface overflow rate
suspended solids
Simplified Stormwater Model
standard deviation
sewage treatment plant
USEPA Stormwater Management Model - Version II
total coliform bacteria
total inorganic phosphorous
total Kjeldahl nitrogen
treatment
total suspended solids
volatile suspended solids
velocity
cubic yards
year
SYMBOLS
C12
C102
Q
$ mil
chlorine
chlorine dioxide
flowrate, units as specified
million dollars
xv
-------
ACKNOWLEDGMENTS
O'Brien & Gere Engineers, Inc., gratefully acknowledges the cooperation
of the Monroe County Division of Pure Waters. Appreciation is expressed to
Dr. Gerald McDonald, Director, Thomas Quinn, Chief of Technical Operations,
Robert Hallenbeck, former Chief of Technical Operations, and James Stewart
of the Monroe County Division of Pure Waters for their cooperation and
assistance.
The support of this effort by the Storm and Combined Sewer Section,
Edison, New Jersey of the USEPA Municipal Environmental Research Laboratory,
Cincinnati, Ohio and of the Office of the Great Lakes Coordinator, Region
V, USEPA, Chicago, Illinois and especially of Richard Field, Chief, Anthony
Tafuri, Project Officer, Storm and Combined Sewer Section, Ralph Christensen,
Chief, Section 108a Great Lakes Demonstration Program, and Larry Moriarty,
Project Officer, USEPA, Region II for their guidance, suggestions and
contributions is acknowledged with gratitude.
Appreciation is also expressed to Donald Geisser and Gary Lade for the
operation of the pilot plant, and to Steven Garver for his assistance in the
alternative analysis portion of the project.
This report has been prepared by O'Brien & Gere Engineers, Inc.,
Syracuse, New York under the direction of Frank J. Drehwing, Vice President,
and Cornelius B. Murphy, Managing Engineer.
xvi
-------
SECTION 1
INTRODUCTION
BACKGROUND
The significance of pollution resulting from urban storm runoff is well
documented. A large portion of this pollution is attributable to combined
sewer overflow (CSO) or discharges from relief points within a combined
sewer system.
Urban stormwater management is a continuous process. Essential to its
success is constant innovation, demonstration, assessment, implementation
guidance, and active program feedback (1). Based upon 1980 dollars, the
cost of abating pollution from CSO is estimated to be approximately $31
billion (2). Such a program, therefore, must be based on proven capabili-
ties, practical methodology and assessment criteria, an expanding data
base, and a continuous technology transfer.
PROBLEM DEFINITION
All programs adopted by Monroe County for the Rochester Pure Waters
District are directed at meeting water quality standards as established by
state and federal legislation for the Rochester, New York area. These
programs specifically address the problem of receiving water quality degra-
dation due to urban storm runoff and subsequent CSO. Figure 1 shows the
study area including the major receiving water bodies, intercepting sewer
network, and significant overflow relief points.
Present CSO results in the direct contravention of established water
quality standards for the Genesee River, imposes heavy nutrient and chemical
loadings on Irondequoit Bay, and causes bacterial contamination of the public
bathing beaches along the Rochester Embayment of Lake Ontario (3). The
latter has resulted in periodic beach closing days during the summer months.
Aside from the direct impacts of CSO, such as objectionable floating
material and high bacteria loadings, the overflows contribute to the
contravention of stream standards in the lower reaches of the Genesee
River under dry-weather conditions. This is caused by the settling of
oxygen demanding materials discharged during overflow events.
Previous studies of the District's combined sewer system (4,5,6) cited
major deficiencies in the existing sewer system and identified the effects
of CSO on the area receiving waters. These studies have all recognized the
-------
-------
need for measures to collect, control, and treat CSO within the City of
Rochester.
PURPOSE OF STUDY
In response to the transient water quality problems induced by periodic
overflows from the Rochester Pure Waters District combined sewer system, a
Research and Demonstration project was undertaken jointly by the U.S.
Environmental Protection Agency (USEPA) and Monroe County, New York. The
primary purpose of the project was to develop an abatement and management
program necessary to achieve a cost effective solution to the CSO induced
impairment of the Genesee River, Irondequoit Bay, and Lake Ontario. In
developing such a program, a practical methodology of approach was formu-
lated to analyze, assess, and evaluate CSO pollution and abatement alter-
natives.
A second purpose was to demonstrate the usefulness of mathematical
models simulating both the urban rainfall-runoff process and the subsequent
stormwater flows within a large combined sewer system. Specific models
that were evaluated include the Simplified Stormwater Model (SSM) developed
by Metcalf & Eddy, Inc. (7) and the USEPA Stormwater Management Model -
Version II (SWMM).
This Report also presents the merits associated with the implementation
of a Best Management Practices (BMP) program to abate CSO. Both source and
collection system management options were evaluated. The developed manage-
ment program including the methodology of approach, the urban stormwater
mathematical modeling, and abatement alternative analysis, involving both
structural and BMP measures, lead to formulation of a master plan for CSO
pollution abatement within the Rochester Pure Waters District. In the
process, it was hoped that the resulting developments in the area of treat-
ment technology, CSO monitoring, overflow sampling and characterization,
urban runoff modeling, and pollution abatement methodology may help other
municipalities define and solve their urban runoff problems.
PROJECT ELEMENTS
To achieve the stated objectives of this study, the Research and
Demonstration project was divided into three basic elements: a CSO
monitoring and assessment program, a CSO mathematical modeling program,
and a pilot plant demonstration program. Figure 2 presents the three
basic elements of the project and their relationship to the development
of the Master Plan. The monitoring and modeling programs are described
in detail in Volume I, whereas, the pilot plant studies are described in
detail in Volume II (8). These three elements evolved from the formulation
of a methodology of approach to the development of a Master Plan.
As part of the overflow monitoring program, an intensive CSO flow
recording and sampling system was implemented to define the frequency,
volume, and pollutant characteristics associated with the District's CSO
-------
Overflow Monitoring Program
- rainfall
- flow
- analytical characteristics
Pilot Plant Evaluations
- optimized process parameters
- removal efficiencies
Drainage Study
- network logic development
- definition of drainage
area characteristics
- definition of receiving-
water impacts
Defini tion of
Abatement Alter-
natives to be Evalua-
ted
Development and
Calibration of the
CSO Network Models"
- SSM
- SWMM
Water Quality Model
, Development and
Calibration
Ranking of Abatement
and Management
Alternatives and
Combinations Thereof
Al ternati ve
Analysis External
to the Network
Model
- implementability
- cost effective-
ness
- political accept-
ability
- environmental
considerations
Preparation
of Master
Plan
FIGURE 2. Program Elements and Their Relationship to the
Development of the Master Plan
-------
discharges. A drainage basin field investigation was conducted to define
those basin parameters that affect the urban stormwater runoff process.
These two programs provided the necessary data sets, including representa-
tive CSO hydrographs and pollutographs, drainage basin characteristics, and
sewer system inventory, to facilitate model calibration and verification.
Included in the modeling effort was the refinement and verification of
the Genesee River water quality model previously developed by Limno-Tech,
Inc. (9). An important task in the modeling program was the preliminary
definition of pollution abatement alternatives. Figures 3 and 4 present
the tasks associated with the monitoring and modeling programs, respectively.
The pilot plant program involved the design and construction of a
pump station and pilot treatment facilities necessary to evaluate the
effectiveness of eight unit processes. The evaluated processes included
high-rate flocculation/sedimentation, swirl degritting, swirl primary
separation, high-rate dual-media filtration, granular activated carbon
adsorption, high-rate disinfection using chlorine and chlorine dioxide,
and microscreening. The pilot plant results were utilized to develop pro-
cess models and associated cost effectiveness relationships. Figure 5
presents the detailed tasks associated with the pilot plant efforts.
It must be emphasized that the three major program elements are not
independent of each other but are closely interrelated and oriented towards
the evaluation of applicable abatement alternatives and the development of
the master plan. Figure 6 presents the relationship of each of the outlined
program elements to the master plan development for the Rochester Pure
Waters District.
-------
PROGRAM
SUBCLEMENT
ESSENTIAL SUBELEMENT
EFFORTS
INDIVIDUAL
TASK DEFINITION
Overflow Monitoring
Program
Rainfall, Flow
and
Quality Info.
en
Measured Hydrograph
Preparation
Time Variable Definition
of CSO Loading
Measured Pollutograph
Preparation
Definition of Time
Variable Constituent
Data Base
Calibration of Hydraulic
Component of Model
Input for SWMM and
Simplified Stormwater Model
Input for Receiving
Model Calibration
Ranking of Drainage Area
Loading
Calibration of SWMM
and Simplified Model
Real Time Definition of
Solids Handling Problems
Provide Hourly Based
Geometric Means and
Regression Relationship
for Simplified Model
Application
FIGURE 3. Overflow Monitoring and Characterization Effort
-------
PROGRAM
SUBELEMEHT
ESSENTIAL SUBELEMENT INDIVIDUAl
EFFOKTS TASK DEFINITION
1 | Land Use Definition |
CSO Svstem
Modeling
1 rnllprtinn nf nrainarjo
Area Characteristics
j SWMM
L
r
L
Simplified Model)-- - 1
Definition of 1
Network System |
L_
Surface Characterization
Network Characterization
I Multiple Drainage Area
and Storm Calibration
Sensitivity Analysis |
Output Correlation
with SWMM Results
Multiple Drainage Area
and Storm Calibration
Detailed Network
Definition
Truncation of Network |
PrPliminary Definition of 1 Lan° use «'««""* |
Abatement Alternatives >
_ Prepa
Recei
Li
ration of
ving Stream Model
Max. Use of Existing Facilities]
Capital Intensive Alternatives]
Steady State Model Definitions!
Verification of Steady
State Model
FIGURE 4. Conbined Sewer Overflow System Modeling Effort
-------
PROGRAM
SUBELEMENT
ESSENTIAL SUBELEMENT
EFFORTS
INDIVIDUAL
TASK DEFINITION
00
Design and Construction of Pump Station
and Pilot Plant Facilities
Operation of Pilot Plant to Define
Response to Variable Hydraulic Loadings
Factorial Design of Pilot
Plant Program
18 Storm Event Operation
Mathematical Modeling of
Individual Pilot Plant Processes
Definition of Possible
Treatment Trains
Cost Effective Treatment
Train Analysis
Evaluation of Specific
Process Train Potential
Problem Areas
Incorporation in SWMH |
External Model Evaluation [
1J
External and Inclusive of
SWMM
Operating and Capital Cost
Evaluations
I Sludge Hand!ing|
[ Operating Problems |
Instrumentation Require-
ments
[ Land Area Requirements |
FIGURE 5. Pilot Plant Effort
-------
Genesee River and
Embayment Model Sensitivity
Analysis
Preliminary Definition of
Alternatives Consisting of
1. Land Use Alternatives
2. Maximum Use of Exist-
ing Facilities
3. Capital Intensive
Alternatives
Preliminary Screening of
Land Use Alternatlves
Utilizing SWHM Sensitivity
Analysis
Preliminary Screening of
Capital of Storage, Treat-
ment and Hydraulic Alter-
natives Using SSM
Detail CSO Network Balancing
of Treatment and Storage
Definition of Receiving
Water Quality Requirements
Treatment Process Train
Cost/Effective AnalysIs
Definition of Implementsbillty
of Alternatives
Environmental, Energy and
Resource Requirements
Political Acceptability of
the Alternatives
Development of
Implementation
Timetable
FIGURE 6. Evaluation Process for Abatement Alternative Analysis
-------
SECTION 2
CONCLUSIONS
Based on the results of this investigation, the following conclusions
relative to CSO management and abatement are presented:
General
A rigorous definition of the existing system of CSO and
stormwater facilities is fundamental to the development of
an abatement program. This definition includes the identi-
fication of major drainage basins, major trunk and inter-
cepting sewers, and CSO and stormwater relief points.
2. The installation and proper maintenance of overflow moni-
toring instrumentation are essential for both receiving
water problem definition and any subsequent sewer network
and water quality model calibration and verification.
Accurate rainfall data collection and subsequent statis-
tical analyses, including design storm definition and
formulation, are essential in evaluating the response of
the existing system as well as the effectiveness of various
abatement alternatives.
Development of a methodology of approach and definition of
applicable abatement alternatives early in the program insure
that the purpose of the study is not lost and all data col-
lection activities are conducted according to the required
analyses.
5. SSM is capable of providing a preliminary screening of
potential abatement alternatives involving a balance between
storage and treatment.
10
-------
6. SWMM (USEPA Stormwater Management Model-Version II) can pro-
ject the urban storm runoff and quantities within acceptable
confidence limits but is presently limited in its ability to
simulate overflow quality.
7. Overflow quality can be better simulated through the appli-
cation of statistical techniques using actual monitored over-
flow data.
8. The ability to abate CSO pollution resulting from an infrequent
design storm event will require the implementation of struc-
turally intensive facilities, minimal structural improvements
to the existing sewer system, or the implementation of non-
structural abatement alternatives, known as BMPs.
Specific to Rochester. New York
1. Hydraulic simulations on the existing sewer system, based on
the application of a simplified stormwater model, indicate
that overflows occur, on the average, 66 days annually. The
total volume associated with these overflows discharged to
all receiving waters within the Rochester Pure Waters District
amounts to 1.9 billion gallons.
2. Because most CSO are generated within highly developed urban
areas, implementation of large volumes of surface storage
that could alleviate some of the problems is economically
impractical.
Point source treatment for CSO is infeasible due to size
requirements and field limitations. The only point source
treatment alternative investigated was the application of
swirl primary separators installed at the discharge point
of the combined sewers. Results obtained from the Pilot
Plant Studies indicate that for the high rates of flows
associated with these overflows, relatively large diameter
separators (approximately 100 ft) would be required. Avail-
able land adjacent to the overflow locations for implementa-
tion of these large separators is nonexistent.
4. Additional studies relative to both soil conditions and
solids removal are required to properly assess the feasi-
bility of underground storage facilities.
11
-------
More intensive street cleaning can alleviate a portion of
CSO induced water quality degradation. Although many of
the surface quality routines associated with SWMM could
not be verified, sensitivity analyses conducted on the
model indicate that increased street sweeping can reduce
the total load of contaminants reaching the sewer system.
More intensive sewer maintenance can help alleviate CSO
induced deterioration of receiving water quality. Review
of sewer maintenance programs recently adopted by Monroe
County indicates that the number of flooding complaints
received during a storm event can be reduced by addi-
tional sewer cleaning.
Additional solids investigations are necessary to determine
the feasibility of using porous pavement instead of conven-
tional asphalt to decrease the rate of urban stormwater
runoff. Preliminary modeling results indicate that the
rate of runoff can be substantially reduced upon appli-
cation of porous pavement throughout the City. Institu-
tional constraints, however, may limit the implementability
of porous pavement on a large area-wide basis.
Disconnecting roof drains from the sewer system can signifi-
cantly mitigate the volume and frequency of flooding in
selected urban areas. The effect on CSO has not yet been
determined. Institutional constraints and public nonparti-
cipation may be strong deterrents to implementing roof down-
spout disconnection ordinances.
The minimal structural abatement alternatives that can sig-
nificantly reduce CSO pollution include: selective upgrading
of the present interceptor system by removing three throttling
constraints, selective overflow weir/regulator modifications,
and blockage of high-impacting overflows.
The three throttling constraints are along the St. Paul
Boulevard Interceptor and specifically include:
(1) the siphon across the Genesee River conveying
flows from the Lexington Avenue and West Side
Trunk sewers to the interceptor,
12
-------
(2) the section of interceptor from the Carthage
Avenue junction chamber to the junction chamber
at Norton Street and St. Paul Boulevard, and
(3) the siphon that is part of the interceptor
conveying flows from the west side to the east
side of the Genesee River.
Selective overflow weir/regulator modifications include
raising of present weir level settings and altering present
gate and flow settings within several overflow regulators.
Blockage of several high-impacting overflows can reduce the
total pollutant load presently being discharged to the Genesee
River. Additional detailed hydraulic modeling is necessary
to determine the possibility of any adverse surcharge condi-
tions existing within these sewers after blockage that could
cause flooding problems.
10. On the basis of projections using a simplified mathematical
stormwater model, the non- and minimal structural abatement
alternatives are expected to reduce significantly the existing
volume of CSO, and the average annual BOD5 and TSS loadings
to the Genesee River.
11. The capital costs associated with the recommended CSO pollu-
tion abatement measures involving non and minimal structural
alternatives are estimated at 12 million dollars (1976
dollars).
12. In Rochester, one structural alternative involves grit
removal, in conjunction with the optimized operation of
the F. E. Van Lare Treatment Facility.
13. The very low assimilation capacity of the Genesee River,
as indicated from previous water quality investigations,
especially during the months of June, July, August, and
September, precludes the construction of a centralized
overflow treatment facility along the Genesee River.
14. Additional control structures within the present sewer sys-
tem would be necessary to utilize the large volume of po-
tentially available in-system storage capacity.
13
-------
15. In the interim, during which the structurally intensive
abatement alternatives are designed and constructed,
application of various BMPs can significantly reduce the
annual CSO pollutant loading to the Genesee River. These
measures include selective overflow weir and regulator
modifications or adjustments, removal of several convey-
ance constraints within the St. Paul Boulevard Interceptor,
and implementation of inflow restriction regulations such
as use of porous pavement in select areas.
14
-------
SECTION 3
RECOMMENDATIONS
It is recommended that:
1. Familiarization with the various mathematical models for urban
storm runoff simulations be made before application of a specific
model. This will insure that the model best suited for a defined
objective will be selected.
2. Network models, such as SWMM (USEPA Stormwater Management Model-
Version II), be relied upon for determining runoff and overflow
volumes for selected storm events and less for estimating the
quality of the runoff. Estimates of surface runoff quality should
be made using statistical analyses of actual field monitored data.
3. Initial screening, planning, and design of storage and treatment
abatement alternatives be made with a simplified continuous
simulation model. Computer costs associated with many detailed
hydraulic models can be prohibitive in many cases; therefore,
only a model that will satisfy the objectives of a study at the
least possible cost should be utilized.
4. Hydraulic analysis and design of sewer systems be conducted with
a detailed network model such as SWMM.
5. Rainfall characterization be based primarily on the use of his-
torical precipitation data, although the design storm approach
may have to be applied in certain situations. More research
must be conducted on the concept of design storms to establish
design rainfall hyetographs that could be applied with mathe-
matical network runoff models.
6. Models not be used to predict runoff/overflow quantities or
qualities without proper field calibration and verification.
A relatively detailed field monitoring program is essential
in providing the background data for proper model calibration
and verification.
7. Continued research and development be made in the area of flow
measuring devices for sewer applications, especially under wet-
weather flow conditions. Field conditions should dictate the
type of monitoring device to be installed.
15
-------
8. More statistical analyses be conducted to better establish the
correlation between runoff quality and parameters such as rainfall
and land use characteristics.
9. Field monitoring systems, regardless of the type and sophistication,
be frequently inspected and verified on a weekly basis to insure
proper data collection.
10. Additional detailed hydraulic analyses be conducted to better
define the interceptor throttling constraints, regulator/weir
modifications, and control structure locations. The SWMM is
capable of providing the required analyses.
11. Additional studies be conducted on the identification of dissolved
oxygen impacts on the Genesee River resulting from CSO discharges.
Previous water quality investigations have indicated that the lower
reaches of the river became most severely impacted by CSO discharges
relative to dissolved oxygen concentrations. An automated water
quality monitor should be located within the lower reach of the
Genesee River and operated over a period of one year to best define
water quality degradation due to these CSO discharges.
12. BMPs be considered when conducting any CSO abatement program. In
many instances implementation of BMPs, possibly in conjunction
with minimal-structural alternatives, can alleviate many problems
associated with frequent CSO discharges. Failure to investigate
their effect could severely limit establishing cost effective
abatement solutions.
13. A program to investigate the effectiveness of increased street
sweeping on reducing the pollutant loadings to the sewer system be
initiated. Such a program should include provisions for corre-
lating the effects of increased street sweeping operations, types
of equipment used, street parking use and restrictions, and pro-
gram costs,1'with reduction of surface pollutants available for
washoff during storm events.
14. Closely associated with the street sweeping program should be a
study to investigate the cost effectiveness of increased sewer
flushing and sewer system maintenance in abating CSO pollution.
15. Control structures be installed at selected locations within the
existing and proposed sewer system. The exact location of these
structures should be determined by conducting additional detailed
hydraulic modeling with the use of a simulation model such as the
SWMM. The type of structure to be implemented should be deter-
mined by actual field conditions in conjunction with results of
the network modeling.
16. The development and implementation of a Master Plan for CSO pol-
lution abatement follow a sequence of phasing of required system
improvements according to their projected cost-effectiveness.
16
-------
17. Scheduled reviews be included in any CSO abatement program to
periodically evaluate the effectiveness and cost/benefits associa-
ted with alternative implementation. This periodic review will
insure that previously defined objectives are being met and, if
not, changes to the program can be made to better solve the initial
problems.
17
-------
SECTION 4
METHODOLOGY OF APPROACH TO ALTERNATIVE ANALYSIS
GENERAL METHODOLOGY
In the process of conducting the Rochester CSO study a methodology for
the analysis of CSO abatement alternatives was developed. The abatement
alternative analysis involved the logical development of a comprehensive
data base and the subsequent analysis activities involving fifteen major
activity categories herein described.
A number of points were inherently stressed in the development of
the methodology sequence. The first of these guiding points was the
need to define the applicable abatement alternatives very early in the
program. This process helped insure that the purpose of the study was
frequently refocused and all data gathering activities would be conducted
according to the required analyses.
The importance of installing an accurate, reliable, and easily
maintained field monitoring and sampling system cannot be overemphasized.
The collection of accurate characteristic overflow data is paramount
in defining the CSO pollutant loadings to the receiving waters and in
establishing the required model calibration and verification data sets.
Another important step is model calibration and verification.
The models selected for application were based on the urban area and
collection system requirements as well as on the objectives defined in
the initial phase of the program. The Simplified Stormwater Model was
used in the initial screening of alternatives, allowing extensive system
analyses to be conducted utilizing minimal computer time. The final
analysis was conducted on a select number of abatement alternatives with
the use of SWMM*. In conducting the network modeling activity it became
apparent that one should not use models which are more complex than the
analysis requires. One should utilize the simplest model which will attain
the defined objectives. It is for this reason that the Simplified Storm-
water Model was utilized in the initial screening of alternatives while
the use of SWMM was utilized in the final analysis of a select number of
abatement alternatives.
The following major tasks were considered essential for the develop-
ment of the Master Plan:
*The exact modeling framework is discussed in detail in Section 6.
18
-------
1. Definition of Program Objectives
a. minimize the wet-weather discharge of combined sewer overflows
to the receiving waters
b. maximize the effective utilization of existing facilities
c. develop cost-effective alternative solutions
d. develop solutions which minimize the energy and resource consumption
It is essential in conducting a comprehensive combined sewer overflow
abatement alternative analysis program that the program objectives be well
defined early in the analysis. It is also important that the program
objectives be frequently reviewed in the course of conducting the analysis
so that the tools and their attendant complexity do not become the objective
rather than the means of attaining the objective.
It is also of tremendous importance that the optimum utilization of
existing conveyance and treatment facilities be identified early on as a
fundamental objective. There are many collection and treatment system
management options that can be employed to minimize the discharge of combined
sewer overflows. In seeking cost-effective solutions to wet-weather discharges,
the full and optimum utilization of the existing facilities should be a high
priority program objective.
2. Definition of Existing Conveyance and Treatment Systems
a. dry-weather conveyance and treatment capacity
b. dry-weather treatment efficiency
c. wet-weather conveyance and treatment capacity
d. wet-weather treatment efficiency
e. system wet-weather relief points
f. system control measures
g. separate sewered areas
The development of an extensive data base and understanding of the
operation and response of the existing conveyance and treatment system is
required before any system modeling can be conducted. For that matter a
basic understanding of the function and present utilization of the existing
facilities is required before the appropriate collection system model can be
selected.
It is particularly important that the available drawings for the con-
veyance system, regulating structures, and control structures be reviewed
and compared to the conveyance system as built. A field survey of all
wet-weather relief points on the system is most important. The Monroe County
Division of Pure Waters had maintained good documentation; however, this is
found not to be the case with many metropolitan areas.
A beneficial activity in defining the available treatment capacity
involves the analysis of the treatment efficiency as a function of plant
inflow and total precipitation. Conducting this type of analysis fully
introduces the investigators to the limitations of the available data base
as well as the facilities.
19
-------
3. Definition of Drainage Areas
a. division of service district into relief point catchment areas
b. determination of area and land use characteristics of the
defined catchment areas
c. determination of percent imperviousness
d. establishment of dry-weather quantity/quality characteristics
The delineation of the combined sewered service area into discrete
hydrologic catchment areas is required for even the most basic form of system
analysis. This can be accomplished by reviewing the available collection
system and topographic mapping for the study area. Each of the delineated
catchment areas must be associated with the system overflow points to which
it is tributary.
Once the catchment areas have been delineated, basic surface definition
is required. The land use, population density, and percent imperviousness
must be established for each catchment area. Associated collection system
dry-weather flow and wastewater characteristics are also required as input to
the network modeling activity.
4. Review of the Meteorological Data Base
a. extent of historical precipitation record
b. interval of precipitation records
c. determination of basic precipitation frequency and magnitude
statistics
The most important factor in the generation of pollutant loadings from
combined sewer overflow discharges is rainfall; and as such, rainfall data
in the form of historical and current precipitation records should be one
of the first items obtained and analyzed. Depending on the level of
refinement which might be employed in the definition of transient stormwater
loads, substantially different rainfall data are required. For preliminary
abatement analyses utilizing a simplified stormwater model, precipitation
records that may be obtained from the National Weather Records Center in
Asheville, North Carolina are usually sufficient. These data are hourly
rainfall and snowfall records and collected over long periods of time,
usually 20 years or more. Utilization of more sophisticated hydraulic
models will require rainfall data on a much shorter time interval, usually
on the order of minutes.
Since the basic driving force in the generation of stormwater pollutant
loads is rainfall, determination of precipitation frequency and magnitude
statistics is essential for analyzing such stormwater loads, their impacts,
and various alternative control strategies. A statistical assessment of
such rainfall parameters as intensity, duration, volume, and time between
successive events should be developed. This determination can most easily
be accomplished through the application of hourly rainfall records supplied
by the U.S. Weather Bureau.
20
-------
5. Selection of Detailed Network Model to be Utilized in the Hydraulic
Analysis of the Existing System and the Proposed Alternatives
a. evaluate control strategies on the basis of single event and
multiple event network models
b. determine need to evaluate surcharge conditions
c. determine the input data set requirements of the selected model
The various identified control options must be evaluated on the basis
of both single event and multiple event network models. Preliminary design
of certain structurally intensive alternatives such as overflow treatment
plants and storage facilities requires the utilization of single event models
to properly size the particular structure. In general, these facilities will
have to be designed on the basis of one particular storm event such as a
5 year rainfall. In order to properly evaluate the effectiveness of any
implemented structurally intensive alternative, it is essential that such a
structure be stressed under both varying rainfall conditions and under the
application of successive events. Multiple event models are best suited to
accomplish this task.
There are many stormwater models available that evaluate surcharge
conditions within the collection system. Depending on the sewer system in
question, surcharge conditions may or may not exist within the various
sewer conduits within a drainage area. Whether they exist can best be
determined from sewer surveys and an overflow monitoring program. The
need to utilize a model with surcharge analysis capability is only neces-
sary if field investigations support the presence of surcharge conditions.
One should be careful not to utilize a more sophisticated hydraulic model
than is actually needed to satisfactorily evaluate the existing conditions.
Many of the available detailed hydraulic models require very large and
intensive input data sets for proper utilization. In applying any of these
models it is important to initially determine all such requirements before
selecting a particular model.
6. Initiation of Overflow and Meteorological Monitoring Program
Necessary to Augment the Existing Data Base
a. hydrograph measurement
b. pollutograph measurement
c. characterization of precipitation event intensity/duration
d. measurement of overflow data on 25 percent of the overflows
for six storm events
e. establishment of statistical relationships for quality parameters
In general, municipalities experiencing a combined sewer overflow
problem lack the necessary overflow and meteorological monitoring data. Thus,
in order to assess the impact of urban runoff on the receiving waters and to
evaluate various control strategies, it is necessary to augment the existing
data base. The most important parameters that should be locally monitored
include: in-system flows at various locations, major overflow locations, and
rainfall data. In addition to these measurements it is essential that the
21
-------
concentrations associated with the overflow. In this manner the total mass
loading of contaminants discharged to the receiving waters can be computed.
Flow monitors installed on the overflow conduits indicate the storm
induced overflow hydrograph. Knowing the pollutant concentrations developed
from the sampling program, pollutographs can be established. Local raingauges
located throughout the study area characterize the rainfall in terms of
duration, intensity, and volume. This information provides some of the
necessary input data required for all hydraulic network models. Statistical
relationships can then be developed that relate rainfall and the overflow
volumes and quality. These relationships are essential in evaluating various
control strategies and abatement alternatives.
Since monitoring of the overflows and rainfall can continue indefinitely
upon initial installation of the monitoring equipment, the question becomes
how many monitored storm events constitute a statistically reliable data base
on which to project future loadings resulting from implementation of various
abatement measures. It is believed that reasonable overflow data collected
for six storm events from 25 percent of all major CSO discharge locations
can provide sufficient data for subsequent abatement alternative analyses. The
resulting data base should allow for an adequate span in monitored land use
characteristics and population densities for the total study area. Variations
from this rule of thumb are frequently encountered in smaller systems and
in systems having complex interconnections.
7. Establish Relevant Abatement Alternatives
a. Best Management Practices
Source Management
street cleaning
sewer flushing
catchbasin cleaning
improved waste collection
stock pile covering
limitation of percent imperviousness through zoning
porous pavement application
Col lection System Management
detention basins
playground or parking lot temporary storage
roof top storage
optimization of in-system storage
addition of polymers
improved system regulation
inflow/infiltration reduction
b. Structural Intensive Alternatives
Storage
in system storage
off-line storage (caverns, lagoons, structures)
Treatment
high-rate physical/chemical processes
multiple use dry-weather/wet-weather facilities
22
-------
c. Optimization of the Existing Facilities
removal of conveyance system throttling constraints
optimization of treatment processes
Although no single abatement alternative is uniquely appropriate for
all municipalities with CSO problems, it is necessary to establish and identify
those relevant alternatives early in any urban stormwater assessment. The
methodology to be applied in such a study is presented herein and is
dictated by many factors. These factors include:
a) local conditions
land use
conveyance system
rainfall patterns
b) availability of local data
receiving water quality
collection system definition
rainfall
demographic
c) type of receiving water body
d) specific water quality problems
dissolved oxygen
bacteria
solids
floatables
e) time scale of receiving water problems
f) decisions involved
feasibility
economic risk
environmental risk
g) study constraints
time
costs
The early identification and preliminary assessment of proposed abate-
ment alternatives can limit the amount of time and money subsequently
devoted to specific alternative evaluations that are not feasible for one or
several reasons. It is advisable that in any abatement program the applicabil-
ity of implementing Best Management Practices (BMPs) be investigated. BMPs
can almost be universally applied at minimal cost within a relatively short
period of time. In many instances, evaluation of collection system management
BMPs can be accomplished with the use of simplified stormwater models as
opposed to the use of more detailed and costly hydraulic models which are
necessary to properly evaluate storage and treatment options.
23
-------
Optimization of existing facilities should also be one of the first
alternatives considered under any CSO abatement program. Many CSO
induced receiving water impacts result from relatively minor problems
within the overall conveyance and treatment system of large urban centers.
Generally monies expended to correct these minor deficiencies are well
spent in that significant CSO loading reductions can be attained at
minimal expense. Detailed evaluation of structurally intensive alternatives
involving storage and treatment on a massive scale need be conducted
only upon the conclusion that implementation of BMPs and optimization of
existing facilities cannot fully realize stated receiving water quality
objectives.
8. Use of a Preliminary Screening Model to Evaluate the Existing System -
Long Term Simulation
a. calibration of model on the existing data base
b. prediction of the annual runoff volume
c. determination of the annual overflow volume
d. determination of the frequency of overflow events
The initial step in evaluating any proposed abatement alternative is the
determination of existing conditions on an average annual basis. Use of a
simplified stormwater model is under long term simulation ideally suited for
such an analysis. If no reasonably accurate data base exists, then real
monitored data must be obtained for the required calibration of the model. Once
the annual volume and frequency of CSO discharges are determined, then the
magnitude of the problem and necessary methodology can be established.
9. Initial Evaluation of Storage/Treatment Capital Intensive Alternatives
Using a Preliminary Screening Model
a. establish the preliminary capital intensive alternative con-
figurations for the range of alternatives being evaluated
b. determination of expected reduction of annual overflow volume
Since most of the detailed hydraulic models are prohibitively costly for
long term abatement evaluations, initial evaluation of structurally intensive
storage/treatment options should be conducted using a simplified stormwater
model. If these evaluations indicate the feasibility of the storage/treat-
ment option in terms of overflow reduction, then more detailed analyses can
be conducted to better define the exact option to be implemented.
10. Application of the Detailed Network Model to the Preliminary Evaluation
of Structural Alternatives
a. determine sensitivity response of CSO loading to the application
of selected structural alternatives
b. determine sensitivity response of drainage area characteristics
to the application of selected structural alternatives
Due to the cost involved in computer analyses of conveyance systems
using a detailed hydraulic model, it is important to initially determine
the applicability of these models to the proposed structural alternatives
from the standpoint of sensitivity responses to various critical design
parameters. An initial evaluation of the modeled response to increases in
24
-------
conveyance capacity, in-line storage capacity, off-line storage capacity or
treatment rates can be used to conduct the preliminary abatement alternative
analysis. A generalized sensitivity analysis of this nature can allow a
number of alternatives to be deleted from additional costly evaluation.
11. Calibration and Verification of the Detailed Network Model
a. 25 percent of all drainage areas
b. three to six storm events
c. verify interceptor flow
d. compare model projections to those obtained using the Simplified
Model output
It is absolutely essential that any applied detailed network model be
calibrated and verified before accepting any overflow projections as reasonable.
Without having a verified model, there is no way in which to insure the accuracy
of the results. Since for any large urban area, calibration and verifica-
tion of detailed network model runs for all defined drainage areas would be very
costly in terms of computer time and preparatory work to define input
parameters, it is advised that approximately only 25 percent of all such
areas for three to six storm events be evaluated. The selection of those
drainage areas should be made on the basis of incorporating a representative
sample of all varying land uses, population, percent imperviousness, and
acreage associated with these drainage areas. It is important that model
projected overflow volumes and frequencies be compared to any existing data
base and also to projections from a simplified stormwater model.
12. Selection and Verification of Wet-Weather Quality Predictive Models
a. selection of water quality models
b. measurement of wet-weather loading and receiving water quality
response
c. calibration and verification of water quality models
d. determination of sensitivity of water quality response to wet-
weather loading
Since the basic problem in assessing urban stormwater runoff is not
specifically the volume, frequency and quality of the runoff but rather the
impact it has on receiving water quality, selection and verification of
appropriate water quality models are absolutely essential. The subsequent
abatement alternative evaluations and final alternative selection will depend
almost entirely on the accuracy and reasonableness of the water quality model
uti1ized.
13. Pilot Plant Evaluations of Applicable Treatment Processes
a. establish the significant process variables
b. establish process efficiency mathematical models
c. determine the cost-effective relationship for the preferred
process treatment trains
In order to properly assess varying treatment options it is necessary to
develop mathematical models that accurately reflect various treatment processes.
If existing data are not available for applicable selected treatment options
25
-------
to be investigated, then pilot plant studies should be developed to evaluate
all such processes. For the Rochester CSO study such pilot plant studies and
developed performance relationships are described in detail in Volume II (8).
14. Detailed Analysis of the Prime Alternatives
a. determine the optimum balance between storage and treatment
under the application of a range of design storms
b. establish the effectiveness of a potential BMP Program
c. establish the preferred system relief volumes and their impact
on the receiving water body as a function of the design storm
frequency
d. establish the marginal cost to marginal benefit relationship for
the prime alternatives
e. determine the institutional and environmental constraints which may
inhibit the implementation of any of the prime alternatives
Since it is extremely difficult, if not impossible, to establish a
design storm hyetograph that is utilized by all detailed hydraulic models,
it is important that a range of storms be evaluated during abatement alter-
native analysis. The selection of these storms can be made on the basis of
selected historical storm events or on application of several hydrologic
methods presently available. The important aspect is that in the process
of developing design storm events, consistent reasoning is used.
Important points that are sometimes overlooked are the institutional
and environmental constraints associated with any of the prime abatement
alternatives. An excellent alternative that is either politically infeasible
or which is environmentally unsound cannot be seriously considered for
implementation. The institutional and environmental constraints must be
considered early in the planning process if a successful program is to be
developed. This can be accomplished by bringing the environmental assessment
and public participation activities into the mainstream of the planning
process.
15. Development of the CSO Abatement Master Plan
a. develop implementation schedule for the cost-effective solutions
b. emphasize initial optimization of the existing system
c. develop cash flow requirements
d. project receiving water benefits associated with implementation
Any combined sewer overflow study must involve the full development of
each of the preceding 14 steps in order to arrive at a cost-effective
abatement program which is represented by the formulation of a Master Plan.
The developed Master Plan should present the implementation schedule to be
followed, emphasize the advantages of early implementation of BMPs and
minimal improvements to the existing system's conveyance and treatment. The
successful Master Plan should also present cash flow requirements for funding
needs, and project the favorable receiving water quality improvements as a
result of alternative implementation.
The fifteen tasks presented as essential for the development of a
Master Plan for combined sewer overflow abatement have been completed under
26
-------
the various activities which constitute the combined sewer overflow abatement
planning for Rochester. The set of guidelines represent a composite of
activities developed from both hindsight and foresight, successful initial
planning and necessary backtracking. The listing is presented as a working
practical guide which would have to be adapted for the individual needs of
each planning area.
TECHNICAL PROCEDURES
There is a wide range of technical procedures which have been defined
as necessary to develop a combined sewer overflow Master Plan. Many of these
procedures have been outlined in the previous section. It is the expressed
purpose of this section to outline the use and interrelationships of a
number of the mathematical models utilized in the development of the combined
sewer overflow abatement program for the City of Rochester. These are also
considered applicable to other large urban centers.
Figure 7 shows the relationship between the various models utilized in
conducting the required analysis. It is recommended that a simplified
stormwater model such as the Simplified Stormwater Model (SSM) be utilized
as the screening network model prior to the utilization of a detailed
hydraulic model such as the SWMM. The required input data sets to the SSM
involve land use information, historical rainfall data, CSO wastewater
characterization information, and a defined network configuration.
Upon initial screening of storage and treatment alternatives utilizing
the SSM, the use of the more complex SWMM is recommended to obtain a more
refined fix of the optimum alternatives. In the Rochester application, the
Water Resources Engineers, Inc. (WRE) transport routine was utilized in place
of the normal SWMM transport routine. The WRE transport routine has the
capability to define areas of potential flooding by means of the backwater
and surcharge analysis. These new routines are included in the November 1977
release q'f the SWMM.
Input data sets required for the application of the SWMM in addition
to that utilized by the SSM include dry-weather and industrial flows as well
as infiltration/inflow (I/I) data. A more detailed definition of land use
as well as a more finely defined sewerage network is also required for the
optimum use of the SWMM.
Several process cost/benefit models are recommended to be utilized in
the evaluation of the treatment alternatives. These models are extensively
described in the pilot plant evaluations for the Rochester, New York CSO
Abatement Program (8). Required input information includes the treatment
train definition, development of area specific weighting factors, and
critical CSO wastewater characterization information. The storage and
treatment cost benefit models are then meshed with a cost optimization
model to yield the optimized alternative configuration.
The network model output in conjunction with the process cost/benefit
model must be analyzed in concert with the output from the water quality
models to establish the program cost-effectiveness. The output of the
27
-------
r-o
CO
CSO
Wastewater
Characterization
Abatement
Alternative
Definition
Network
Definition
Development
of Area Specific
Weighting
Factors
Treatment
Train
Definition
CSO
Wastewater
Characterization
Process
Cost Benefit
Mode!
Dry Weather
and Industrial
Flows
Alternative
Cost
Optimization
Model
Receiving
Water Quality
Mode!
Hydrologic
Stream
Definition
Program
Cost Effectiveness
Analysis
FIGURE 7. Diagram for Model Utilization
-------
modeling effort then assists in the selection of an optimum solution or
series of solutions necessary to eliminate the discharge of untreated
wastewater from the combined sewer system under investigation. It must
be understood that the modeling activity is not an end in itself, but
rather an aid in the complex process of developing a Master Plan.
Following is a brief description of each of the modeling tools and
their operational requirements:
EPA Stormwater Management Model (SWMM)
A comprehensive mathematical model, capable of representing urban
stormwater runoff, has been developed to assist administrators and engineers
in the planning, design, evaluation, and management of overflow abatement
alternatives (10).
Hydrographs and pollutographs (time varying quality concentrations
or mass values) are generated for real storm events and systems from points
of origin to points of disposal (including travel in receiving waters) with
user options for intermediate storage and/or treatment facilities. Both
combined and separate sewer systems may be evaluated. Internal cost routines
and receiving water quality output assist in direct cost/benefit analysis of
alternate programs of water quality enhancement (11). Efforts on modification
and improvement of the SWMM have continued since its initial release.
Simplified Stormwater Model (SSM)
A simplified stormwater model was developed under this program to
provide an inexpensive, flexible tool for planning and preliminary sizing
of stormwater facilities (7). The model delineates a methodology to be
used in the management of stormwater. It consists of a series of interrelated
tasks that combine small computer programs and hand computations. The model
successfully introduces time and probability into stormwater analysis,
promotes total system consciousness on the part of the user, and assists in
establishing size effectiveness relationships for facilities.
The following five tasks are performed in this model:
Data preparation
Rainfall characterization
Storage treatment balance
Overflow quality assessment
Receiving water response
The relative success of stormwater control alternatives may be checked
at two major points in the simplified model. After the initial storage/
treatment balance has been performed, the duration, volume, and frequency of
overflows may be checked to determine the impact of a control alternative
on the receiving water quality.
29
-------
Process Cost/Benefit Models
Several cost/benefit models developed under the Rochester, New York
CSO research and development project were utilized to evaluate the effective-
ness of the various treatment process trains. Models were developed which
relate an expected treatment efficiency to process and wastewater dependent
variables.
As an example, the performance model for disinfection relates the
effectiveness of the process to the disinfectant utilized, dose applied,
velocity gradient in the disinfection chamber, influent bacterial level,
wastewater BOD5 and TKN levels, and the disinfection chamber contact time.
To develop this disinfection performance model, a statistical analysis was
conducted on the pilot plant data.
Using capital, operation and maintenance, material and supply, power,
and chemical cost expressions, a cost performance model was developed.
Using a cost optimization subroutine, the capital and operating costs
associated with facilities having different disinfection requirements were
developed, optimizing the operational and design parameters so as to
minimize annual costs.
Similar cost optimization models are available for the analysis of a
wide range of treatment processes. These tools were utilized in the cost/
benefit analysis of the wet-weather treatment facilities in conjunction
with application of the Runoff and Transport routines of SWMM Version II.
The reader is referred to Volume II of this Report for a complete
description of the treatment process performance models used in the alterna-
tive analysis. A detailed description of the process cost models used in
the CSO abatement alternative analysis is presented in Section 9.
30
-------
SECTION 5
OVERFLOW MONITORING AND DRAINAGE BASIN CHARACTERIZATION
OVERFLOW MONITORING
General
The Rochester wastewater collection system is, for the most part, a
combined system whereby sanitary sewage, industrial wastes, and stormwater
are all collected and conveyed via the same conduits. The existing major
trunk and main intercepting system serving the Rochester Pure Waters
District is shown on Figure 1. The District is essentially divided by
the Genesee River into western and eastern components. Overflows from the
western portion discharge to the Genesee River, whereas, overflows from
the eastern portion discharge to Irondequoit Bay. As stated in Section 2,
the first important task in conducting a CSO pollution abatement program
is identifying significant overflow volume and pollutant loadings from
the existing sewer system.
Flow Monitoring
Thirteen major overflows to the receiving waters were monitored under
this program as shown in Figure 8. Within each of these overflow discharge
conduits an electronic flow measuring system was installed to determine the
hydrograph and thus the quantity of combined sewer overflow discharged
from the tributary drainage area. The primary flow measuring device was
an ultrasonic level and/or velocity probe system manufactured by Badger
Meter, Inc. In order to determine which ultrasonic system would be best
suited to the existing conditions within the overflow conduits, intensive
field evaluations were conducted.
Table 1 lists the location of the thirteen major combined sewer
overflow monitoring installations and the installed equipment associated
with each.
Besides measuring these thirteen overflows, four trunk sewer flow
measurements were made. One was located at the point where the Genesee Valley
Canal Sewer (GVCS) empties into the Clarissa Street Tunnel. The GVCS, for
the most part, is a stone box sewer which is badly deteriorated. A replace-
ment, the Genesee Valley Interceptor, is presently under construction. It
31
-------
Culver-Goodman
Tunne!
( Proposed)
Genesee
South we
(Propos
SCALE l"= 7,OOO
- RAIN GAUGE
- OVERFLOW
31 - REFERS TO DRAINAGE AREA
DESIGNATION 8 OVERFLOW NO.
FIGL'RE 8, Overflow Locations
32
-------
TABLE 1. OVERFLOW MONITORING INSTALLATIONS
Ultrasonic Unit
No. Location Installation Type
7
8
9
16
17
18
21
22
25
26
28
29
31
Maplewood Park
Lexington Avenue
West Side Overflow
Mill and Factory
Plymouth and Railroad
Brooks and Agnew Court
Norton at Seth Green
Carthage Avenue
Central Avenue
Court Street
Norton Screenhouse
Denstnore Bypass
Thomas Creek
. , , ._..--.. y j
Head and Velocity
Head
Head
Head
Head
Head
Head
Head
Head
Head
Head and Velocity
Head and Velocity
Head
was felt that any actual flow records that could be obtained would be helpful
in verifying the design capacity requirements for the new interceptor.
Two other interceptor flow measurements were made on the East Side
Trunk Sewer (ESTS) at locations that provided information as to how the
interceptor is operating at the same time that two regulators (tributary
to Overflow No. 31) are diverting flow to overflow conduits.
The fourth interceptor flow measurement was made at the Van Lare STP.
The existing influent flow measuring device is a venturi meter that has a
maximum recording capacity of 200 tngd. At times of intense rainfall the
flow to the Van Lare STP exceeds this amount, therefore, a meter with a
higher range was necessary.
A typical head and velocity probe installation is depicted in Figure 9.
The head probe installation was basically the same except for the
deletion of the velocity probes. Each system generated a 4 to 20 mini-
amperes (ma) signal which was linearly proportional to the primary measure-
ment (head and/or velocity). This signal was then converted at the overflow
location to a tone signal and sent via telephone line to a central
data acquisition station. The conversion and transmission of the tone
signal was accomplished with telemetry instrumentation manufactured by
Bristol Instruments, ACCO Bristol Division.
This tone signal was received and reconverted to a 4 to 20 ma output
again with Bristol telemetry instrumentation. This output was then
converted to a 0-10 v signal which was required for input into the Digital
Equipment Corporation PDP8/E computer. The computer was programmed to
translate the 0-10 v input into a flow value based on hydraulic curves
established for each particular overflow site. The overall telemetry
system is depicted in Figure 10. If the flow measurement involved head,
a curve representing Manning's equation stored in the computer was used
33
-------
to compute the discharge (Q). However, if the monitoring installation
involved both head and velocity, a curve reflecting Q^VA was stored where A,
the flow area within the discharge conduit, was calculated from the water
level sensing head probe.
FIGURE 9. Typical Head and Velocity Probe Installation
The resultant flow values were printed and punched at five minute
intervals. This interval was determined to be reasonable and could be
changed by a simple command at the computer console. The punched paper
tape was collected, translated, and the resultant overflow information
stored on magnetic tape on the Xerox 560 computer located at the offices
of O'Brien & Gere Engineers, Inc.
Sampling
Samplers were installed at each of the overflow locations in Figure
8. The samplers were refrigerated-sequential units that were equipped
with 1/4 in. sampling hose. The pump and sampling apparatus were manu-
factured by Sigmamotor, Inc. and were very similar to their Model WM-5-
24. However, modifications to the sampler package were necessary to
meet the varying head conditions at each location.
For those overflow conduits that were more than 20 ft below grade,
the sampling pump was installed in a waterproof box within 10 ft of the
minimum overflow level. The pump then delivered the sample to an above
grade location where the sample bottles were stored under refrigeration.
In order to coordinate the initiation of the sampler with the first
measureable amount of overflow, a relay system was incorporated into the
instrumentation already existing at each of the overflow locations. This
34
-------
I
CENTRAL
RECEIVE
LOCATION
FIELD
STATION
POP 8/E
COMPUTER
FLOW VALUES
POP 8/E
CONVERTER
Of
L
BRISTOL TELE-
METRY RECEIVER
r
UJ
z
o
I
Q.
UJ
UJ
UJ
co
co
CO
a:
i
MOD.7FMT55C
SAMPLER
SIGMAMOTOR
BRISTOL TELEMETRY
TRANSMITTER
CVJ
OVERFLOW
CONDUIT
I
r-
LU
a:
cc
^
o
o:
4-20 MA
ACROMAG
RELAY
; I
I
BADGER
PROBE
FIGURE 10. Overall Monitoring and Telemetry System
35
-------
relay accepted the 4 to 20 ma signal from the Badger probe output and
initiated the sampler when the 4 ma signal was increased to approximately
4.2 ma.
Samples were collected at approximately 15 min intervals unless the
pumping distance dictated suction and purge times greater than 15 min. At
the completion of an overflow event the samples were collected and transported
to the laboratory for analyses. A portion of the analyses were
performed by the Monroe County Division of Pure Waters laboratory and the
remainder by O'Brien & Gere. Table 2 represents the schedule of analyses
that were performed on the sequential and composite samples.
TABLE 2. OVERFLOW ANALYSIS SCHEDULE
Laboratory
Analyses
Type of Sample*
Monroe County Division
of Pure Waters
O'Brien & Gere
BOD5, Set. S,
SS, TS, TVS, TDS,
VDS, t. col iform,
f. col iform, f. strep
Hg, O&G, Cl-Hydro.
pH, COD, TOC, TKN,
NH3-N, Org. N, NOs,
N02, TIP, Cl, Fe, Cu,
Zn, Pb, Cd, Cr, Ni
S - Sequential; C - Composite
Rain Gauging
A system of 12 rain gauges was installed within the Rochester Pure
Waters District. The location of each gauge was selected by dividing the
District into 12 units of equal area. Within each of the units, a building
of significant height (school, firehouse, pump station) with no
observable influences from adjacent buildings or trees was selected for
installation. The total system of gauges allowed an evaluation of dif-
ferences of rainfall patterns and their possible effect on the intercepting
sewer system. Figure 8 shows the location of each gauge.
All rain gauges were Fischer-Porter bucket-type instruments, Model
No. 35B1558. These gauges recorded the amount of accumulative total rain-
fall at five-minute intervals on punch paper tape. The punch paper tape
was collected approximately every two weeks, translated, and the information
stored on magnetic tape.
The flow, analytical, and rainfall data that were monitored and recorded
served two main purposes. First, the data were used directly in the cali-
bration and verification process associated with the application of the SWMM.
Secondly, much of the data was used for establishing valuable relationships
36
-------
between rainfall and overflow quantity and quality developed through the
use of statistical procedures.
DRAINAGE AREA CHARACTERIZATION
Introduction
In the overall effort to develop an abatement and management program
for the combined sewer overflows, the physical characteristics of the
entire study area were defined. The collected data served several
purposes: (1) it established a framework which allowed the successful
utilization of the available network mathematical models, (2) it provided
the necessary drainage area parameters for determining regression equations
relating the pollutant loading of combined sewer discharges to tributary
area characteristics,and (3) it provided the data base of surface character-
istics that can be reasonably well correlated to that of other metropolitan
areas.
Data Collection
A brief description of the scope of work outlined for the drainage
area surveys follows:
1. Each drainage basin was divided into a maximum of sixty
subcatchments ranging in size from ten to fifty acres.
Division was made on the basis of land slope and zoning
classifications such as residential, commercial, industrial,
or open.
2. Each drainage area subcatchment was characterized by determining
the average ground slope, percent of the area that is pervious
and impervious, total area, average drainage width, total
Tenth of gutters, total number of catchbasins and stored
volume of each, the BODs contained in the volume of stored water
in those catchbasins, and any available surface storage capacities.
3. The estimation of dry-weather flow (DWF) for each subcatchment
was made by collecting statistics as to the number of dwelling
units, average number of people per unit, market value of
average unit, average family income, percent of units with
garbage grinders, and information as to industrial process flows.
4. Other information required to complete the characterization of
the drainage basin included total population, street cleaning
frequency, average number of sweeper passes, and location of
inlet manholes.
5. Information was obtained to characterize the sewer network
including length of sewejvconduits, conduit slope, type, size
and roughness, and manhole invert and ground elevations.
37
-------
The drainage basins identified in Figure 8 were initially defined
using the District's sewer maps in conjunction with previous reports
( 4> 5) and selective field investigations. In addition, the level of
detail for the drainage basin characterization resulted from initial
interpretations of the data base requirements of the mathematical network
models. The level of effort required for drainage basin characterization
is largely dependent on the purpose of the study and should be identified in
the methodology of approach. Tables 3 and 4 summarize the field determined
drainage basin characteristics conducted under this program. These data
were used in conjunction with the mathematical network modeling effort.
38
-------
TABLE 3. SUMMARY OF DRAINAGE AREA CHARACTERISTICS
co
Drainage Area
No.
6 (separate)
7
8
9
16
17
18
21
22
25
26
28
29
31
Area
ac
569
729
988
2682
826
235
541
810
569
348
393
778
1353
1671
Tryon Park (separate) 1720
Charlotte
1390
SFR*
247
610
339
1407
348
196
507
644
434
0
118
506
910
820
1118
695
MFR*
16
7
22
0
10
9
3
0
0
91
40
78
83
155
344
139
Land Use
Commercial
24
53
32
109
318
6
19
72
38
257
175
78
140
320
86
278
- ac
Industrial
130
18
465
994
50
0
0
55
27
0
20
38
70
240
86
139
Open
152
41
130
172
100
24
12
39
70
0
40
78
150
136
86
139
Ave. Land
Slope
0.0074
0.0118
0.0066
0.0060
0.0059
0.0067
0.0073
0.0073
0.0070
0.0080
0.0100
0.0100
0.0100
0.0100
0.0150
0.0150
% Imp*
53.2
48.2
46.4
49.2
66.1
41.6
41.6
44.9
52.9
80.3
43.3
46.6
54.4
49.9
40.0
35.0
*Note: SFR = single family residential
MFR = multi-family residential
% Imp. = percent imperviousness
-------
TABLE 4 . SUMMARY OF DRAINAGE AREA CHARACTERISTICS
Drainage Area
No.
Dwelling Units Population Catchbasins
Gutter
Length-lOQ Ft.
No. of Conduits &
Subcatchments in
Original SWMM Network
6
7
8
9
16
17
18
21
22
25
26
28
29
31
Tryon Park
Charlotte
1300
4653
2016
9919
4397
1979
3965
6333
4596
0
(3500)
5287
(10700)
(9200)
-
-
3948
18476
5522
27296
13687
6408
11980
17540
14332
(12700)
(16700)
16467
29700
25500
-
-
527
911
574
2357
(826)
335
736
1339
-
-
-
-
-
-
-
-
1015
2022
1353
6481
(2800)
743
1730
2895
-
-
-
-
-
-
-
-
139
133
72
136
32
61
114
128
95
24
118
256
122
-
-
29
60
32
60
28
28
43
54
52
11
40
83
48
-
-
NOTE: Values in ( ) indicate extrapolated or assumed data based on the other drainage areas,
No value indicates parameter not used as SWMM input.
-------
SECTION 6
NETWORK MODELING
CONFIGURATION AND TYPE OF MODELS USED
Introduction
Until recently, the most common methods for analyzing and designing
storm and combined sewer systems have been based on the rational formula
and various steady-state flow equations for sewer network analyses.
Mathematical models have been developed in response to the recognized
limitations of previous methods for analyzing highly complex systems under
dynamic flow conditions. The models consider a number of factors:
(1) Spatial nonuniformity of rainfall
(2) Randomness of rainfall events
(3) Time variable runoff
(4) Spatial and temporal variations in dry-weather flows
(5) Overland surface runoff travel times
(6) Sewer conduit flow routing
(7) Operation of diversion structures and storage facilities
(8) Pollutant quality characteristics of surface runoff and dry-
weather flow.
Minimum requirements of the selected network models include the
capability to consider several representative rainfall events, compute
runoff from several subcatchments or drainage basins, and the ability to
route flows in a branched conveyance network.
The two models selected for demonstration and use in the evaluation of
various combined sewer overflow abatement alternatives proposed under this
program were the SWMM (11) and the Simplified Stormwater Model (7). The
Simplified Stormwater Model was developed by Metcalf & Eddy, Inc., and
demonstrated by O'Brien & Gere Engineers, Inc. under this program. Each
model was selected for a specific purpose. Preliminary screening of various
abatement alternatives was accomplished with the use of the Simplified
Stormwater Model, whereas, the SWMM was used to evaluate specific alter-
natives under the application of varying design storms.
Stormwater Management Model (SWMM)
In 1971, a joint venture of Water Resources Engineers, Inc., Metcalf
and Eddy, Inc., and the University of Florida developed a mathematical model
41
-------
capable of simulating urban stormwater runoff and combined sewer overflows.
The model accepts rainfall and drainage basin characteristics as inputs and
calculates the quantity and quality of stormwater and combined sewer over-
flow.
Basically, the model consists of four components: runoff, transport,
storage, and receiving water. The runoff component computes direct runoff
for a given rainfall event for each subcatchment using a kinematic wave
formulation with Manning's equation.
The generated runoff then serves as input to the sewer system network. The
transport component routes the flow through the sewer system using a modified
kinematic flow approximation. This analysis combines Manning's equation and
the conservation of mass. The storage component modifies the sewer flow
within the system at selected points where such facilities have been provided.
Finally, the receiving water component computes the effects of the stormwater
and combined sewer overflow discharges on the receiving waters.
More specifically, the basic elements of the SWMM program are built on the
following major routines (the capitalized, underlined names represent
computer language subroutines as identified in the EPA documentation of the
SWMM).
(1) Input Sources
RUNOFF generates overland surface runoff based on input rainfall
hyetographs and surface characteristics such as imperviousness,
land use, and infiltration capacity. Surface quality character-
istics of urban runoff are also generated in this routine and are
based on factors such as antecedent rainfall conditions, street
cleaning practices, and land use.
FILTH generates the quantity of sanitary and industrial flow
based upon such factors as land use, population density, and
process flows.
INFIL generates infiltration into the modeled sewer system network
based on groundwater and sewer conduit conditions.
(2) Central Processing
TRANS combines and routes the generated surface runoff quantity
and quality values through the sewer system. However, due to
possible surcharge and backwater conditions existing in the
Rochester sewer system, a different version of this particular
routine had to be used to better represent actual flow conditions
in the sewers.
For the network modeling of the Rochester Pure Waters District sewer
system with SWMM, several modifications had to be made to the SWMM
program. First, the Water Resources Engineers, Inc. (WRE) modified version
of transport block had to be used due to possible surcharge conditions within
42
-------
the sewer system (12). The primary differences between the original SWMM
transport and the WRE version of the transport block can be summarized as
follows:
(a) Conceptualization of the transport system with the WRE
model uses a more refined mathematical representation of
the sewer system totally unlike the original SWMM. This
allows for more accurate modeling and analysis of the
actual system.
(b) The basic flow equation in the WRE model includes the inertia!
terms of the Saint-Venant equation in the solution, whereas,
the original SWMM is based on a kinematic wave approach. The
inclusion of these other terms allows for a better dynamic
mathematical simulation of the sewer system with existing surcharge
and backwater conditions.
(c) The WRE model includes the necessary programming to simulate
sewer surcharge, looped systems, weirs, and other related
features that were unavailable with the original SWMM.
It should be noted that the SWMM computer program that was originally
developed in 1971 has been undergoing constant review and improvement. Many
routines that were initially unavailable are now part of, or are included as,
options to the SWMM program. The specific SWMM version used for the
Rochester system analyses is Version II dated March 1975. More recent
versions also include as an option the Water Resources Engineers transport
routine, frequently referred to as WRE Transport, extended transport, or
simply EXTRAN. At the time this study was being conducted this particular
option was not available as part of SWMM Version II.
It should also be noted that in the SWMM Version II runoff subroutines,
the dust and dirt pollutant build-up factors are considered to be concen-
trated over the paved street areas and not uniformly over the entire
drainage area.
The quality routing section of the WRE model in which individual
pollutants are transported by advection in each modeled conduit,
theoretically provides a better representation of actual processes than
the original SWMM (12). In the present link-node version it is assumed
that all transport of constituents occurs by advection in the system
conduits, while dilution of pollutant concentration occurs at the system
nodes. This means that only those pipes which enter a node from upstream
can advect pollutant material into the node. Thus, the pollutant
concentration at a given node is unaffected by quality events downstream
from the node. This is valid for gradually varied flow in a conduit
system assumption.
Although the original storage/treatment routine of the SWMM was
intended to be used to evaluate unit treatment operation and capital
intensive storage systems, many of the mathematical formulations related to
process definition and efficiencies are internal to the model and are therefore
43
-------
not user controlled. Pilot plant evaluations conducted as part of the
original scope of work for this study indicated that many of the inherent
model equations were not representative of the response of a number of
unit processes to the combined wastewaters characteristic to the Rochester
system. It became more feasible to develop empirical relationships
relating flow and treatment efficiency and system costs from these pilot
plant studies than to attempt to modify the equations that are presently
incorporated in the model.
The SWMM incorporates a receiving water simulation routine that can
be used to predict water quality impairment from combined sewer overflow
discharges. Since little prior work had been done with this portion of
the SWMM and because of the availability of a verified receiving water
model, the LIMNOSS model (which is a modification of the EPA steady-state
AUTOQUAL model) was adopted for use (13). The LIMNOSS model is a one-
dimensional, second order, finite-difference simulation model. Projections
of water quality effects on the Genesee River from the present system}
and various abatement alternatives were made using this model.
Machine requirements for the entire SWMM involves a high speed
computer with approximately 90K words of storage. Available core storage
on.the Xerox 560 located at the offices of O'Brien & Gere Engineers is
approximately 30K. Through the use of offline disc storage and retrieval,
the deletion of some routines in the SWMM, and the extensive editing of
the program, the SWMM program was adopted to be executed on the Xerox
560. The system dictated program modifications involved the addition of
a paging algorithm allowing the large arrays associated with the original
SWMM to be compatible with the available storage.
Simplified Stormwater Model (SSM)
A simplified stortnwater management model has been developed by
Metcalf & Eddy, Inc. to provide an inexpensive, flexible tool for planning
and preliminary sizing of stormwater and combined sewer overflow abatement
facilities (7). It successfully introduces time and probability, through the
use of long-term historical rainfall data, to the estimation of the
frequency, duration, and quantities of stormwater during rainfall events.
The use and application of the simplified model involves five main
tasks:
(1) Data preparation
(2) Rainfall characterization
(3) Storage/treatment balance
(4) Overflow quality assessment
(5) Receiving water response
Data preparation consists of synthesizing a schematic of the sewer
system network to be evaluated and subsequent collection of data on both
overflow quantity and quality. This task provides the necessary information
for input to Tasks 3 and 4. The rainfall characterization is easily
44
-------
accomplished through a small computer program introduced as part of the
model and can be used to evaluate various design storms and to produce a
historical perspective on the storm events and their impact on the sewer
system. The storage/treatment portion of the model can be used repeatedly
to analyze various combinations of storage capacities and interceptor
conveyance capacities to determine the optimum conditions required to
minimize overflow.
Strictly speaking, the overflow quality assessment and the receiving
water response (Tasks 4 and 5) are not an integral part of the SSM. Therefore,
modeling of the Rochester sewer system with the SSM involved only tasks 1
through 3. Evaluation of overflow quality and receiving water response were
made using statistical procedures and alternate models that are not integral
to the SSM.
The main purpose of the SSM was to provide a preliminary evaluation of
storage and treatment facilities necessary for combined sewer overflow
control. The conceptual basis for Task 3, storage/treatment, is shown
schematically in Figure 11.
Runoff-
Land Surface
Overflow
Maximum
"Storage
Treatment Rate
or
Interceptor
Capacity
Figure 11. Concept of Storage/Treatment-SSM Analysis
45
-------
The computer program can accept rainfall from either a long-term
historical record or from an individual storm event. Rainfall is then
converted into runoff using a gross runoff coefficient. The generated
runoff is stored in a defined storage volume (which can be zero) which is
drained at a specified treatment rate. The defined treatment rate could be
represented by an actual treatment facility or by available interceptor
capacity. When the runoff rate exceeds the storage capacity and the treat-
ment rate, an overflow occurs.
The major data required to run the storage/treatment portion of the
model involve the catchment area, the gross runoff coefficient of the
overall drainage area, the in-system storage capacity, and the treatment
rate or interceptor capacity. It must be noted that the mathematical model
does not involve a complex dynamic hydraulic mathematical simulation. It
does not route generated runoff volumes through a sewer network according
to established fluid mechanic flow equations, but rather accounts for rainfall
and subsequent runoff on a steady state mass balance - volumetric basis.
The SSM output is presented in the form of a record of the time and
volumes of rainfall, runoff, and overflow, and a monthly summary of these
parameters. Minor programming modifications made to the basic SSM for this
study include a statistical analysis of the generated output, which provides
the user with the average, standard deviation, confidence limits, and other
statistical information associated with the output parameters.
Water Quality Models
Any combined sewer overflow abatement program has, as an ultimate
objective, the establishment of acceptable receiving water quality conditions
and the preservation of the best usage of the aquatic resources. To achieve
this goal, it is necessary to project the effects of both present combined
sewer discharges to receiving water bodies as well as the discharges
resulting from the implementation of various abatement alternatives. As
previously mentioned, the output of the LIMNOSS receiving model generated
under a concurrent program (13) was used to provide the necessary background
receiving water definition as well as the required water quality projections
under varying loading conditions for the abatement alternative analyses.
Three LIMNOSS receiving water quality models were formulated for the
purpose of evaluating the impact of existing and projected combined sewer
overflow pollutional loads to the Genesee River and to the Rochester Embay-
ment of Lake Ontario. The models include a steady-state fecal col iform
and assimilation capacity simulation program for the impacted region of the
Genesee River and a steady-state fecal coliform model for the Rochester
Embayment. The input required for preparation of the above defined models
was acquired through dry and wet-weather sampling. The structure of the model
is based on the simple concepts of continuity, including terms for advection,
dispersion, reaction, and point or non-point sources or sinks. The model
can calculate the steady state spatial distribution of any water quality
parameter whose kinetic behavior can be described by zero or first order
kinetic formulations or described as conservative. In the case of the
46
-------
Genesee River, the parameters modeled included chloride, fecal coliform,
carbonaceous oxygen demand, nitrogenous oxy.gen demand, and dissolved oxygen.
The model requires detailed data input of physical parameters (width,
depth, length, etc.), loads, flows, reaction coefficients, and dispersion
coefficients. The model assigns specific values to each model segment,
sets up a water quality matrix according to the mathematical formulation as
described above, and then solves the system of differential equations using
an algebraic inversion of the equation matrix. The model has been carefully
verified for theoretical and numerical accuracy using a number of test
cases where direct analytical solutions are available for comparison (13).
ASSUMPTIONS IN MODELING
Runoff Projections
An intensive effort was made in the collection of field data that was
established as necessary for drainage basin characterization for subsequent
use in mathematical modeling utilizing the SWMM. Much of the collected data
pertained to input information for the runoff block of the SWMM. Initially,
the various drainage basins were broken down or discretized into a large
number of smaller catchment areas each with its own input parameter character-
istics, such as percent imperviousness, area, overland flow, detention storage
and resistance factors, the number of catchbasins and the length of curbing.
Sensitivity analysis conducted on the runoff block of the SWMM indicated that
the model was quite insensitive to many of the input parameters in relation
to the resultant runoff hydrograph. These input parameters include ground
slope, resistance factors for impervious and pervious areas, and width of
subcatchment. Parameters which significantly changed the resulting runoff
hydrograph included area, percent imperviousness, retention storage and
infiltration rates. In addition, the analysis showed that for specific
drainage basins the resultant runoff hydrograph obtained from either a
coarse or a detailed discretized area was essentially the same. The
coarsely discretized area required much less computer execution time
thereby resulting in significant savings. A coarse network further
allowed for less manhours in computer set-up time. The average savings was
estimated to be by a factor of ten considering both computer costs and labor.
It must be noted, however, that the runoff from a particular drainage
area was not monitored during this study. Only the overflow from the sewer
system was measured. The runoff projections from the SWMM were refined based
on the comparison of projected to measured overflow volumes.. Calibration of
the model is also described in this section.
Application of the runoff block of the SWMM involved several assump-
tions. The default values (values that the model assumes if not user
supplied) for several catchment area parameters were considered applicable
for the Rochester area. The specific parameters involved were the surface
runoff resistance factors, surface depression storage, and soil infiltration
rates. For most of the storm events that were considered, the default values
were considered sufficiently accurate and thus were utilized.
47
-------
In addition to the volume of runoff, the runoff block of the SWMM accounts
for the quality of surface runoff. Initially, the model estimates the amount
of pollutants accumulated on the surface prior to rainfall. Then, accounting
for rainfall intensity, land use, and soil credibility, the model projects
the quality of runoff. The output is in the form of a loadograph or the
time varying pollutant concentration profile. Input parameters include the
number of antecedent dry days, street cleaning frequency and number of passes,
number of catchbasins and their stored volume and 8005 content, erosion
related soil factors, and the length of gutters within the catchment area.
For this study erosion was not considered a factor due to the large percentage
of paved and rooftop areas. In addition, on the basis of small samplings, the
number of catchbasins and the total length of gutters on a per acre basis
were assumed to be essentially uniform throughout the drainage basins. The
problems associated with the runoff block quality subroutine will also be
discussed in this section.
Overflow Projections
It was determined that the most important factor in applying the WRE
Transport routine is the selection of the integration step. Due to the
mathematical representation of the physical system, the optimal length of
the integration step is related to the length and diameter of the sewer
element modeled. The length of time-step required is directly proportional
to the length and inversely proportional to the diameter of the sewer conduit.
Thus, short, large diameter conduits require very short time steps to maintain
mathematical model stability.
As previously mentioned, coarse discretization of the drainage areas
allowed for the development of a coarse yet reliable representation of the
sewer network based on comparisons between model predicted values and
monitored quantities. This resulted in relatively long lengths of conduits
between manholes. Even with a coarse network, the relative magnitudes
of length and diameter of conduits necessitated the selection of a small
integration step. For most model runs a five second time step was
considered sufficient. Any additional accuracy obtained from runs with
a shorter time step was not considered justifiable.
Another important aspect of applying the WRE Transport routine
was the manner in which the actual sewer system was mathematically rep-
resented for model input. By including all elements of the sewer system
such as manholes, conduits, weirs, orifices, pump stations, and other
appurtenances, accurate model overflow projections would result but very long
computer simulation times would be incurred. Consequently, a coarse
representation of the network was selected and modeled. It was assumed that
the average slope of a sewer conduit over a long distance would be equivalent
to modeling many intermediate lengths between manholes. In addition, only
pipes larger than 12 in. in diameter were modeled. No pumps were included
and only those control or regulator structures which diverted combined
sewage to the receiving waters were modeled. Furthermore, although usually
there are varying shaped sewer conduits within a large urban area, almost all
the conduits for this study were modeled as equivalent circular pipes
48
-------
because of the limited number of shapes analyzed by the WRE Transport routine.
If the main concern is the overflow hydrograph, then a network composed of
hydraulically equivalent conduits would be sufficient.
Inasmuch as the modeled network is a combined system, most of the flows
associated with overflow events result from rainfall induced runoff. Dry-
weather flow and infiltration during a storm event played a minor role. In
fact, infiltration during a storm event was considered negligible and
therefore not modeled. Regarding dry-weather flow, although there
are many model input parameters that require definition, in general, model
default values were considered sufficient. The only concern was to match total
dry-weather flow in the modeled system to that being recorded at the Van Lare
STP.
LIMITATIONS OF THE AVAILABLE SOFTWARE
Runoff
In the overall rainfall-runoff SWMM simulation,all wastewater flows
and pollutants, except for the contribution of dry-weather flow and ground-
water infiltration, are generated within the runoff routine of the model.
Therefore, accurate projections of combined sewer overflow hydrographs and
pollutographs depend to a large extent on the accuracy of the runoff block-
generated hydrographs and pollutographs. Actual runoff hydrographs from
individual drainage areas, however, were not monitored. Only the system
overflow from the area was monitored during a storm event. For a mass
balance check at each regulator or diversion structure, the incoming flow
originating from surface runoff entering the sewer system, the wastewater
flow in the trunk or intercepting sewer, and the overflow must be known.
From the monitoring data obtained on the overflows and limited interceptor
measurements the accuracy of the model runoff projections could not be
assessed directly. From other runoff calculations performed and from the
fact that the overflow projections generally corresponded to those monitored
it is believed that the calibrated SWMM runoff routine computes the surface
runoff with sufficient accuracy, given a specific rainfall event.
Combined sewer overflow wastewater quality was also modeled as part
of the overall overflow monitoring effort. The SWMM incorporates two
separate routes for the calculation of overflow quality characteristics.
The surface washoff contribution is calculated as part of the runoff routine
of the model. The impact of pollutants contributed from the dry-weather
flow and industrial discharges is accounted for within the transport routine
of the model. An extensive effort was undertaken to correlate model overflow
quality projections to those monitored. Because it is difficult to separate
the storm-generated pollutants from the dry-weather flow contribution,
the actual accuracy of the runoff quality portion of the model could not be
ascertained. It is believed, however, that various parameters such as the
dust and dirt accumulation rate and the associated fraction of BODs and SS
accounted for within the runoff routine need refinement.
49
-------
It must be further noted that, according to the SWMM runoff block, all
urban contaminants are assumed to derive from street surfaces only; they are
removed by runoff up to the amount of pollutant remaining on the ground.
WRE Transport
This portion of the SWMM is very complex and a major effort is required
to implement it. Limited documentation at the time of the study makes the
model very difficult for the user to implement without exhaustive efforts
in the interpretation of the theoretical base and the definition of the
input data set requirements.
Improvements are also needed in the model output format. The arrangement
of several tables makes it difficult for the user to interpret the overflow
projections. In addition, complete hydrographs and pollutographs may be
obtained for only twenty selected locations within the sewer system. If
additional points need to be evaluated, repeated model runs of the same data
are required to obtain sufficient information.
Specific limitations of this routine include:
(1) Inability to model various shaped sewer conduits other than the
six provided for within the computer program ,
(2) Inability to account for water loss from the system as the
result of surface flooding from surcharged manholes ,
(3) Model is event-specific and does not include provisions for
antecedent moisture conditions or water quality accounting between
rainfall events, and
(4) Inability to accurately define the necessary time interval or
integration step necessary for mathematical simulation based on
the defined differential equations for flow definition.
Most of the above limitations have been recognized as serious problems
and model improvements are currently in progress, but no reasonable
explanation or interpretation has been found for the integration step
problem.
In general, overflow projections provided by the SWMM appear to be
correlated reasonably well with actual monitored field data relative to
overflow discharge rates and volumes. The accuracy of wastewater quality
predictions may be expected, however, to have only order-of-magnitude accuracy.
Considerable model improvement is required in terms of its ability to
simulate overflow quality before the model can be used with confidence for
such projections.
50
-------
Storage and Treatment
A serious limitation of the model was encountered during application
of the treatment routine. The model does not accurately predict BODs and
TSS removal efficiencies of various treatment process devices.
Most of the programming for this portion of the SWMM does not appear
consistent with the results of the pilot plant studies which evaluated
numerous process trains. In order to expedite development of water quality
projections, it was established that entirely new equations had to be
developed to accurately analyze treatment of combined sewer overflows.
These relationships were developed based on the regression analysis of
performance data as described in Volume II of this Report.
Receiving Water
Although the theoretical and numerical accuracy of the previously
described steady state water quality models that were used in the overall
combined sewer overflow abatement program have been thoroughly verified,
their suitability for describing the impact of transient loading phenomenon
may still be questioned. A steady state water quality framework
can be used to approximate the magnitude of the water quality impact; however,
the framework is quite limited in simulating the real time response of the
Genesee River to stormwater discharges. The experienced time variations
in the discharge loadings and the river flow simply cannot be accurately
described by a steady state model. The use of this model was intended
only to provide a relative comparison for the river response to stormwater
discharges, and to bracket the expected resulting water quality.
The one-dimensional assumption of the model also imposes some limitations
on modeling the Genesee River. However, these efforts are not considered
overly significant because intrusion from Lake Ontario and vertical
stratification are not pronounced. In general, the model framework is
considered satisfactory to describe dry-weather conditions in the Genesee
River if, and only if, perturbations resulting from varying loads and
photosynthetic activity are not considered important. Its suitability to
describe wet-weather conditions in the river is adequate only for assisting
in qualitative evaluations and to bracket the projected water quality.
Establishing accurate numerical predictions of pollutant concentrations is
in general difficult because of transient conditions induced by stormwater
and combined sewer discharges.
MODEL CALIBRATION AND VERIFICATION
Overflow Volume
In order to expedite verification and model utilization in abatement
analysis, the SWMM was calibrated and verified for three drainage areas and
then applied to the remaining areas. The areas selected for calibration
and verification were Maplewood Park, Carthage, and Thomas Creek which
represent drainage areas numbers 7, 22, and 31, respectively. The
51
-------
decision to use these areas was based on the extensive amount of measured
flow and quality data obtained from the overflow monitoring program for these
sites. In addition, Carthage was selected because its overflow exhibited
extremely high pollutant mass loading rates relative to the volume discharged.
Furthermore, the three drainage areas exhibited the range of land use
categories characteristic of the entire study area.
The storms involved in the SWMM calibration process were of both varying
intensities and durations. In general, the rainfall data were obtained from
the rain gauges associated with the drainage areas as part of the total
overflow monitoring program. The exception is for Carthage where the
required rainfall data were obtained from the U.S. Weather Bureau at the
Monroe County Airport in addition to rain gauges located in adjacent areas.
It must be remembered that under the monitoring program, only the
resulting overflow discharge rate was measured, not the runoff from a
particular drainage area. As such, model predicted overflow hydrographs
were compared to actual monitored data. Therefore, the WRE transport routine
was used for calibration and verification. It is the transport routine of
the SWMM that generates the overflow quantities based on the sewer
network characteristics. The runoff routine estimates the rainfall-
runoff relationship. This was indirectly verified.
For each of the three drainage areas selected for verification and cali-
bration, three storms were identified for modeling purposes. The three
selected storm events exhibited the range of rainfall intensities, durations,
and volumes likely to be encountered during an average year. From the
rain gauge data, rainfall hyetographs were constructed for model input.
Table 5 shows the selected storms and their important characteristics for
each of the three drainage areas.
These hyetographs provided the required rainfall input to the SWMM
runoff routine which in turn generated runoff hydrographs for each drainage
area. From the SWMM transport routine, which accepted the runoff hydrographs
and added in domestic and industrial wastewater discharges, the combined
sewer overflow volumes from each drainage area were generated.
TABLE 5. RAINFALL CHARACTERISTICS
Drainage
Area
No. Storm
7
7
7
22
22
22
31
31
31
10/9/75
9/11/75
9/20/75
8/29/75
8/24/75
9/11/75
6/19/75
6/ 5/75
6/12/75
Rain-
Gauge
No.
3
3
3
5
Airport
5
2
2
2
Peak
Intensity
(in./hr)
0.12
0.24
0.20
0.60
0.42
2.40
0.60
1.20
1.20
Duration
(hr)
4.33
2.92
1.00
7.92
8.00
3.17
3.68
1.36
2.56
Total
Volume
(in.)
0.40
0.40
0.20
0.80
0.71
0.80
0.50
0.60
0.80
Note: 1 in. = 2.54 cm
52
-------
The most significant characteristics of the predicted overflow hydro-
graphs for calibration and verification purposes were the peak flowrate,
the total overflow volume, and the general shape of the hydrograph.
Figures 12 through 20 show the predicted and measured overflow hydrographs,
plotted as cfs against clock time, for each of the three selected drainage
areas for each of the three storms. The first figure presented for each
drainage area represents calibration runs while the other two represent
verification runs. As indicated, Thomas Creek exhibited the best correlation
between predicted and measured overflow hydrographs. The predicted total
volume for each storm was within 10 percent of the measured value while the
predicted peak flowrate was within 2 percent of the measured value. The
correlation for Carthage also was very good, although the results for the
storm of 9/11/75 showed a variation in peak flowrate and time frame between
measured and predicted volumes. The predicted total volume for each storm
was within 16 percent of the measured quantity while predicted peak flowrates
were within 34 percent of the measured values. The largest discrepancies
between the SWMM predicted volumes and measured overflow quantities occurred
for the Maplewood Park area. The calibration and verification show very
good correlation between model projections and measured field data.
Table 6 shows the total volumes and peak flowrates for the predicted and
measured overflows by site and storm.
The SWMM was calibrated and verified for the purpose of analyzing the
effectivensss of various abatement alternatives in reducing the quantity
of combined sewer overflow being discharged from the collection system. For
this type of analysis, a simplified sewer network of the entire Rochester
Pure Waters District was characterized for model evaluation. Figure 21
shows the conduits and nodes representing the entire combined sewer system.
Trunk sewers from the individual drainage areas discharge to the main
interceptor at rates governed by system regulators. It is important to
note that the principal combined sewer overflows occur on trunk sewer systems
prior to discharge to the interceptor. There are only two overflow relief
points on the main interceptor system leading to the Van Lare STP. One is at
Front Street and the other is at the siphon under the Genesee River where the
St. Paul Boulevard Interceptor crosses from the west to the east side. These
two locations are designated as nodes 250 and 251 in Figure 21.
As with any model simulation of actual conditions, it is impossible to
exactly represent the sewer network as it occurs in the field. The simplified
network represented by Figure 21 is hydraulically equivalent to the actual
sewer network consisting of hundreds of manholes (nodes) and pipes (conduits).
That is, the conduits represented in Figure 21 have similar flow characteristics
to the actual conduits on an average overall basis. Because of the good
correlation between predicted and measured overflow hydrographs for the
three selected drainage areas, application of the same technique to the
remaining drainage areas was justified. It is for this reason that the total
overflow volumes from the entire system are believed to be within 20 percent
of actual.
53
-------
179
Overflow - cfs
OJ
o
£D
CZ
O
B)
cr
CJ
O
<
n>
-s
-h
Q.
-~>
O
OQ
-s
BJ
-a
o
-s
OJ
a
CO
o
o
O-
CO
-5
7T
en
ro
-s
o
fD
-s
re
o -z.
I' O O
O 0.
VO "O ^
-**J 5
CO
c.
-s
Q-
-s
03
a.
^j.
n
r+
0)
Q.
-------
Overflow - cfs
o
ro
o
CO
O
en
O
en
o
on
CD
ID
o
rn
ro
-s
o
Q>
c-f-
j«
O
3
O
ro
-h
O
CL
O
UD
-S
EU
TD
o
-s
B)
a
ro
o
o
Q.
CD
-s
x
co
O '
-s ro (
3 S
OJ
O ^
a. o
to
CD
-a
QJ I
O en
-o
-s
ro
a.
o
r+
ro
O-
-------
730
Area No. 7
Maplewood Park
Storm 9/20/75
Rain Gauae No. 3
Predicted
Measured
900 930
Time (hr of day)
FIGURE 14. Verification Overflow Hydrograph for Maplewood Park 9/20/75
1000
-------
O)
120
110
100
90
80
70
60
50
40
30
20
10
0
Area No. 22
Carthage
Storm 8/29/75
Rain Gauge No. 5
Predicted
Measured
260 2300 2WO 10
1700 1800 1500 2DOO 2100 2200 2300 2400 100 200 300
Time (hr of day)
FIGURE 15. Calibration Overflow Hydrograph for Carthage 8/29/75
-------
CD
l/l
<4-
O
M-
70
60
50
40
30
20
10
0
Area No. 22
Carthage
Storm 8/24/75
Airport Rain Data
Predicted
Measured
430
550 600 630
Time (hr of day)
830
FIGURE 16. Verification Overflow Hydrograph"for Carthage 8/24/75
-------
en
160
140
120
100
co
O
, 80
2
o
t 60
O)
o
40
20
0
1900
Area No. 22
Carthaqe
Storm 9/11/75
Rain Gauge No. 5
Predicted
Measured
2100
2130 2200
2230 2300
2400
Time (hr of day)
FIGURE 17. Verification Overflow Hydrograph for Carthage 9/11/75
-------
en
o
o
I
O
240
200
160
120
80
40 ..
Area No. 31
Thomas Creek
Storm 6/19/75
Rain Gauge No. 2
Predicted
Measured
0630
0730
0830
0930
1030
1130
1230
Time (hr of day)
FIGURE 18. Calibration Overflow Hydrograph for Thomas Creek 6/19/75
-------
CTv
on
M-
O
I
3
O
480
400
320
240
Ol
o 160
80
1330
A
*
1400
Predicted
Measured
Area No. 31
Thomas Creek
\ Storm 6/ 5/75
X Rain Gauge No. 2
\
\
X
\
X
\
v
1430 1500
Time (hr of day)
1530
1600
FIGURE 19. Verification Overflow Hydrograph for Thomas Creek 6/5/75
-------
o
o 200
o
160 . .
80
40
0
Area No. 31
Thomas Creek
Storm 6/12/75
Rain Gauge No. 2
Predicted
Measured
0800 0830 0900
0930 1000
1030
1100
Time (hr of day)
FIGURE 20. Verification Overflow Hydrograph for Thomas Creek
6/12/75
62
-------
AREA 9,8
AREA 18 AREA 17
26) 139
AREA 25 AREA 22
VAN LARE
STP
LINKED INTERCEPTOR SYSTEM
AREA 31,29,28,21
FIGURE 21. SWMM Node-Conduit Representation of Sewerage Network
for City of Rochester
63
-------
TABLE 6. OVERFLOW PEAK FLOWRATES - VOLUMES
Drainage Area
No. Storm
7
7
7
22
22
22
31
31
31
10/9/75
9/11/75
9/20/75
8/29/75
8/24/75
9/11/75
6/19/75
6/ 5/75
9/12/75
Peak
(cfs)
33
78
65
115
118
124
258
490
350
Site
Measured
Total Vol.
(mil gal)
0.86
2.20
1.00
2.30
3.16
3.07
8.17
13.98
4.36
Peak
(cfs)
27
50
72
123
120
82
252
495
350
Model Predicted
Total Vol.
(mil gal)
0.85
1.69
2.18
2.51
3.40
3.56
8.02
14.12
3.98
Note: 1 mil gal = 3785 m^
1 cfs = 0.00283 m /sec
Overflow Quality
The combined sewer overflow pollutants, discharged to the receiving
waters during periods of rainfall are generated primarily from surface
runoff, although the scouring of accumulated sediments within the sewer
conduits may contribute significant amounts of pollutants. In general,
the volume of domestic and industrial wastewaters is small in relation
to that of surface runoff; however, the associated contribution of bacteria
and BOD is large. Projection of overflow quality using the SWMM involves
both the runoff and transport routines. In the runoff routine the
surface runoff pollutant concentrations are generated based on antecedent
rainfall conditions, street cleaning frequency and efficiency, land use,
number of catchbasins and length of gutters within the drainage basin.
Domestic and industrial wastewater contributions are accounted for in
the transport routine of SWMM and are based on land use and population
characteristics of the drainage basin. Domestic and industrial wastewater
contributions are accounted for in the transport routine of SWMM and are
based on land and population characteristics of the drainage basin. The
pollutants from both sources are then routed through the sewer system by
the transport routine of the model.
Although the overflow quantity verification was quite satisfactory,
the quality verification proved to be unsatisfactory. The SWMM runoff
routine generates surface pollutants by initially estimating the accumula-
tion of dust and dirt on streets within the drainage area. These accumu-
lated solids also generate an organic demand, which is estimated as a
fraction of the dust and dirt present. The pollutant loading rates used in
the model were based on studies conducted by the American Public Works
Association (APWA) in Chicago (14). From estimates of average daily
64
-------
traffic and average daily litter production, the APWA developed dust and dirt
accumulation factors for various types of land use. These loading rates
are expressed in terms of pounds of dust and dirt (DD) per day per 100 ft
of curb length. The organic demand parameters are in turn generated as a
fraction of the DD present. These factors are expressed in terms of
milligram of pollutant per gram of DD. More recent versions of SWMM better
account for the build-up of surface pollutants.
Since the pollutant loading to a sewer system is largely determined by
surface runoff, which is extremely variable with respect to both quantity
and quality, modeling efforts to characterize such elements have been
difficult at best. The pollutional load from domestic and industrial
contributions is relatively easy to identify and quantify. It is, however,
the urban surface runoff that largely determines the character of the storm-
water or combined sewer overflow discharges.
Poor quality verification has resulted from the use of the SWMM dirt
and dust accumulation factors which were not Rochester specific. Although
the SWMM program was modified to allow the user to input the values for
the DD parameters, it was difficult to establish the correct set of
factors for good verification. No extensive studies have been conducted
to date in the Rochester area to determine the exact nature of the
surface pollutants. The applicability of the SWMM cannot be determined
until actual build-up rates have been calculated for these pollutants.
Even then, the variability of the contaminants from one location
to another could make accurate modeling of runoff quality extremely difficult.
Extensive work on quality model calibration provided no reasonable correlation,
In summary, assessment of combined sewer overflow quality can theoreti-
cally be addressed by mathematical models such as SWMM. The basic
assumption of these models is that the pounds of pollutants washed off
in any time interval during a rainfall event is proportional to the pounds
remaining on the ground and the runoff rate. In addition, the amount of
pollutant accumulation is based on limited studies for a specific area.
A second method, which is more direct and also consistent with the
concept of simplified modeling, is the application of regression techniques
on data sets collected as part of a combined sewer overflow monitoring
program. Under this program the statistical analysis of field observations
was used to extrapolate the required quality parameters, after several
unsuccessful attempts to modify the quality portion of the SWMM in order to
adequately predict the quality characteristics of an overflow. For this
study, the linear regression equations were of the form:
where Q represents the dependent variable quality parameter, the subscripted
X values represent the independent variable quality parameters, and the
coefficients a through z represent the calculated regression coefficients.
By using a suitable logarithmic transformation of the data, regression
analysis can develop equations of the form:
65
-------
Ql = X!3 x2b-..z
A measure of the ability of the regression equation to predict adequately
the dependent variable is the correlation coefficient. The closer the
absolute value of the coefficient is to one, the better the correlation
between the dependent and independent variables.
From the literature available on the quality of combined sewer over-
flows, it is known that the quality characteristics of an overflow are
extremely variable and are dependent on many factors. The most significant
of these factors are assumed to be the intensity of rainfall, the number of
antecedent dry days, and street cleaning practices. Land use and population
density can also affect overflow quality (15).
Regression equations for the prediction of total suspended solids (TSS)
and biochemical oxygen demand (BODs) were developed based on several vari-
ables including rainfall characteristics and antecedent dry days. The data
base consisted of selected measured analytical values from several overflow
sites for various storms occurring during 1975. Winter overflow events were
excluded because of the possibility of a snowmelt induced overflow independ-
ent of precipitation. The data set consisted of the composite TSS and BOD5
values, the number of dry days (herein defined as the number of elapsed days
since the last day having a total of 1.0 in. of rainfall as determined from
the rain gauges located throughout the study area), the total duration and
average intensity of the rainfall event, the population density, and the
percent imperviousness of the drainage area.
Many combinations of variables were selected for regression analysis.
Presented in Table 7 are the regression equations having the highest
correlation coefficients for TSS and BOD- The determined level of correlation
TABLE 7. REGRESSION ANALYSIS - QUALITY PARAMETERS
Total Suspended Solids:
TSS - 163.66 (RAIN)-0-424(DD)0-084(DUR)-0-589
Multiple Correlation Coefficient = 0.47
Biochemical Oxygen Demand
BOD5 = 37.83 (RAIN)"0'616(DD)°- 159(DUR)~'639
Multiple Correlation Coefficient = 0.61
Where TSS = Total suspended solids - mg/1
RAIN = Average rainfall intensity throughout storm - in./hr
DD = Antecedent dry days in which accumulated rainfall <1.0 in.
_ PUR _= Duration of storm - hrs.
Note: Number of data points = 30
Significance levels on independent variables = 0.95
66
-------
is limited because of the available data base and the inherent vari-
ability of urban stormwater runoff. It is interesting to note that the
correlation for BOD5 is considerably better than that for TSS. Further-
more, the regression coefficients indicate that both TSS and BODs are
proportional to the number of antecedent dry days. These relation-
ships were anticipated except for the negative regression coefficient for
the rain intensity variable. It would seem that a higher intensity rainfall
would impact more energy to surface washoff thereby generating higher con-
centrations of both TSS and BOD5.
The samples used in the regression analysis were composited from
individual samples taken at uniform intervals from the beginning of the
overflow event. A flow-proportional composite sample would correlate
quality better with the flowrate and therefore would reflect quality
variations related to the flow and would be a more representative approach
to quality analysis.
Comparison of SWMM Output To Other Models
A comparison of results from several computer models that simulated the
rainfall-runoff relationship for the City of Rochester indicated that
relatively simple models can accurately estimate the volume of runoff from
a given drainage basin. It was demonstrated that these same models can
also reasonably predict the total sewer system overflow volume. Sophisticated
models requiring many man-hours of data collection and manipulation and long
computer simulation times may not be necessary to closely approximate the
volume of runoff and combined sewer overflow discharged from a given drainage
area.
The three models used for runoff comparison were the Quantity-Quality
Simulation (QQS) by Dorsch Consult (16), SWMM, and SSM. Total runoff volumes
generated by the QQS and the one, two, and five year design storm hyetographs,
which supplied the rainfall data input, were provided by others (3). The
synthetic rainfall hyetographs were formulated using a method developed by
Keifer and Chu (17) in conjunction with established rainfall-frequency-
duration curves for the Rochester area as obtained from the U.S. Weather
Bureau located at the Monroe County Airport. The peak intensity and the
total depth of rainfall for each year are shown in Table 8. Furthermore,
the same drainage basin characteristics that were used for QQS input were
also used as input to the other two models. Runoff volumes generated by
the SWMM and SSM were then compared to the QQS results.
TABLE 8. RAINFALL DATA FOR MODEL COMPARISON
Year Storm
1
2
5
Total Depth
(in.)
0.84
1.37
1.92
Peak Intensity
(in./hr)
2.99
3.98
5.00
Total
Duration (min)
210
210
210
Note:
1 in. = 2.54 cm
67
-------
Although application of the SWMM runoff routine involves estimating
seven drainage basin characteristics for data input, only the area in
acres, the percent imperviousness, and the maximum and minimum infiltration
rates for each drainage area were common to both the QQS and SWMM input
data base. Parameters such as subcatchment width, average ground slope,
infiltration decay rates, and surface resistance and storage factors were
either assumed to be accurately represented by the default values in the
SWMM or were estimated from drainage area maps. The entire City of
Rochester was modeled using 18 separate drainage areas comprising a total
of 12,694 acres. These 18 individual areas were initially further sub-
divided into smaller subcatchments; however, for the runoff comparison the
land characteristics of each subcatchment were area-weighted to provide a
single average value for each of the 18 drainage areas. Table 9 lists the
basin characteristics for runoff data input for the SWMM for each of these
drainage areas.
Application of the SSM involved only three parameters: total depth
of rainfall, size of drainage area in acres, and gross runoff coefficient.
For this comparison, the runoff coefficient was assumed to be equal to
the percent imperviousness of the drainage areas. The runoff determination
of the SSM is based on a volumetric balance of total rainfall. Although
this model is more useful in analyzing many years of rainfall records, it
can be run for a particular rainfall event. For a singular storm event the
runoff volumes can be calculated easily without the aid of a computer.
Table 10 lists by recurrence interval the runoff volumes as generated by
each of the three models.
TABLE 10. RUNOFF VOLUMES FOR MODEL COMPARISONS
Total Runoff Volume (106 ft3)
Year Storm QQS SWMM SSM
1
2
5
17
29
47
16
29
46
17
28
39
As indicated, the runoff volumes generated by the QQS and SWMM are
essentially the same. Volumes generated by the SSM are similar to those
from the other models except for those associated with less frequent storm
events; that is, events less frequent than the two-year storm.
On verification of the QQS and SWMM projected runoff volumes, both
models were compared on the basis of total system overflow volume to the
Genesee River and Irondequoit Bay. System overflow as determined by the
SSM was not simulated because of the degree of simplification and assumptions
that are required to make the QOS sewer network compatible with SSM input
requirements. Because the two-year design storm criteria was to be used in
the abatement alternative analysis, only system overflow induced from runoff
generated during the two-year storm event was compared.
68
-------
TABLE 9. SUBCATCHMENT DATA
en
10
Infiltration
Subcatch- Gutter Width Area Percent Slope Resistance Factor Surface Storage Rates Decay Rate
Ment No. Or Inlet (ft ) (ac) Imperv. (ft /ft ) Imperv. Perv. Imperv. Perv. Maximum Minimum (I/sec)
(in.) (in./hr)
7 5 19600.0 721. 31.5 .030 .013 .250 .062 .184 2.00 .50 .00115
8 210 18000.0 974 42.4
9 196 45000.0 2172. 40.6
10 375 14000.0 344. 46.9
16 303 28000.0 811. 46.7
15 331 10000.0 511. 25.9
18 304 5000.0 448. 30.0
17 316 6000.0 660. 30.0
19 606 13000.0 57. 85.0
25 257 15000.0 360. 80.4
26 3 5000.0 393. 43.3
310 33 25000.0 1501. 47.8
311 38 3000.0 170. 68.9
29 47 23000.0 1353. 54.4
280 86 10000.0 193. 34.4
281 90 18500.0 627. 47.2
81 16 10500.0 838 38.6
22 128 14000.0 561. 52.5
Total Number of Subcatchments, 18 \
1
' \
r \
I ,
r i
r ^
i
t
1
Total Tributary Area (Acres), 12694
-------
The primary difference between the sewer network as modeled with the
QQS and the SWMM is the degree of detail. The application of the QQS model
involved the modeling of hundreds of manholes and sewer conduits throughout
each drainage area, whereas, the SWMM, as applied, involved modeling only the
main trunk and intercepting sewers. Thus, it has been shown that an
accurate determination of total system overflow quantities can be made
using a less detailed sewer network.
The sewer network as used in the SWMM simulation is shown in Figure 22.
This network was based on data collected in the course of the QQS character-
ization. Analysis utilizing the SWMM involved 2880 integration steps of
10 seconds each. Table 11 shows the total system overflow volumes as
predicted by the two models for the two-year storm.
TABLE 11. OVERFLOW VOLUMES FOR MODEL COMPARISONS
Overflow Volume (106 ft3)
Year Storm
2
QQS
26
SWMM
23
Good comparison between the model output for the two-year storm
precluded the simulation of the one- and five-year storms. Choosing a smaller
time step than 10 seconds, thereby increasing the computer simulation, might
possibly increase the accuracy of the SWMM results. Additional runs
utilizing a shorter time step seemed unwarranted based on the results for the
two year storm.
Computer simulation time for the SWMM runoff block was 0.9 min for
each design storm. Modeling of the relatively detailed trunk and intercept-
ing sewer network as shown in Figure 22 required approximately 45 min of
CPU computer time on a Xerox 560. Further simplification of the sewer network
would allow for shorter computer execution times. If done judiciously,
little loss in accuracy would result from such simplifications.
70
-------
NOTE: X INDICATES OVERFLOW
A INDICATES INLET MANHOLE
982) Area 28
FIGURE 22. SWMM Sewer Network Configuration for QQS-SWMM Comparison
71
-------
SECTION 7
CRITERIA FOR ALTERNATIVE ANALYSIS
INTRODUCTION
Extensive data on urban storm runoff loadings that were collected over
the past several years have been reported in the literature. These provide
an immediate source of information and can, when there is a lack of local
monitoring and sampling data, provide a first level estimate of the impact
of runoff from an urban drainage area. The use of literature derived data
is useful in estimating the average yearly stormwater and CSO pollutant
loadings. Literature reported data are not, however, a substitute for local
data collection and analyses specific to the drainage basin under investiga-
tion. Further definition of transient stormwater loads leads to various
levels of refinement in pollutant loading projections. The level of refine-
ment generally proceeds as follows:
Level 1 - Average yearly stormwater and combined sewer overflow load
Level 2 - Actual event distribution
Level 3 - Variation within events
The justification for more detailed analyses involving higher levels of
sophistication is directly related to the environmental and economic risks
involved in decisions made from preliminary analyses involving lower levels of
precision and refinement.
The need for adequate local data above that provided by the literature
is based on the following factors:
1. Differences in rainfall patterns, frequency, intensity, and
duration, in conjunction with surface drainage area characteristics,
significantly influence storm runoff volumes and pollutant loadings.
2. Significant variations in storm-generated loads from similar storms
occurring at different times throughout the year are observed
within the same study area.
3. studies which attempt to characterize loading data generally cover
relatively short periods of time relative to the long term rain-
fall patterns in a given area.
72
-------
The most important factor in the generation of storm-induced CSO volumes
and pollutant loadings is rainfall. All methods for estimating these storm-
induced loadings rely on rainfall data as fundamental input for reasonable
and meaningful projections. Much effort has been expended in the field
of network modeling and the collection of stormwater and CSO monitoring and
sampling. It appears that little effort, however, has been spent in the
area of rainfall characterization for abatement analysis. A very difficult,
and yet unsolved problem, is the accurate and meaningful development of
rainfall hyetographs for various return period frequencies from Weather Bureau
frequency-intensity-duration curves or actual local rainfall data.
RAINFALL ANALYSIS
Daily and hourly precipitation records from the U.S. Weather Bureau
station at Rochester were used to conduct the network modeling studies.
Rainfall records, covering a period from 1948 to 1975, were obtained from
the National Climate Center, Asheville, North Carolina. Precipitation
patterns and durations for the Rochester area are highly variable. High-
intensity, short-duration events are usually associated with thunderstorms
occurring during the summer months; whereas, low intensity, long duration
events are usually associated with cyclonic activity occurring during the
spring and fall months.
Data from local rain gauges located within the study area also provided
rainfall information for specific drainage areas which did significantly
enhance the developed rainfall hyetographs. A system of twelve rain gauges
was installed within the study area of the Rochester Pure Waters District as
discussed in Section 5. On the basis of 19 storms recorded between
January and August of 1975 the Weather Bureau gauge indicated an average
of 0.44 in. per storm while the local gauges recorded an average 0.51
in. The Weather Bureau gauge also indicated an average storm duration
of 8.05 hr compared to 5.5 hr, based upon the local gauges. The difference
in duration is probably due to the lower sensitivity of the local recording
rain gauges. These gauges must accumulate 0.10 in. of rain before
indicating the start of a storm event; whereas, the Weather Bureau gauges
record to the nearest 0.01 in. Although the local gauges did indicate a
rainfall variation pattern throughout the study area for various storm
events, they did indicate that rainfall across the District is generally
relatively uniform.
Two statistically valid and accepted methods (7, 17) for rainfall char-
acterization were used for abatement alternative analysis. Actual rainfall
records for a 20 yr period from January, 1954 to December, 1973 obtained
from the U.S. Weather Bureau were used as input to the SSM. Synthetic
hyetographs for various return periods derived from Weather Bureau rainfall
frequency-duration-intensity curves were used as input to the SWMM.
Based on the 22 years of precipitation records from January, 1954 to
December, 1975, Table 12 summarizes the monthly and annual rainfall statistics.
In addition, Tables 13 and 14 summarize the number of rain days and the
rain per rain day associated with period of record.
73
-------
TABLE 12. TOTAL RAINFALL (in.)
Jan.
Feb.
Mar.
Apr.
May
Jim.
Jul.
Aug.
Sept.
Oct.
Nov.
Dec.
ANNUAL
Average
2.11
2.45
2.36
2.58
2.65
2.81
2.30
3.31
2.29
2.52
2.76
2.50
30.63
Std. Dev.
0.87
1.10
1.01
0.83
1.20
1.61
1.23
1.26
1.14
1.81
1.12
1.10
4.43
Coefficient of
Variation
0.41
0.45
0.43
0.32
0.45
0.57
0.54
0.38
0.50
0.72
0.41
0.44
0.14
95% Confidence
Level Interval*
0.39
0.49
0.45
0.37
0.53
0.71
0.55
0.56
0.51
0.80
0.50
0.49
1.97
TABLE 13.
NUMBER OF RAINDAYS
Jan.
Feb.
Mar.
Apr.
May
Jim.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
ANNUAL
Average
15.64
14.77
13.82
12.91
11.59
9.41
9.23
10.18
10.00
11.05
15.18
17.73
151.50
Std. Dev.
3.16
3.70
3.62
2.74
3.42
3.22
3.12
2.13
3.80
3.12
2.92
3.15
12.91
Coefficient of
Variation
0.20
0.25
0.26
0.21
0.29
0.34
0.34
0.21
0.38
0.28
0.19
0.18
0.09
95% Confidence
Level of Interval1
1.40
1.64
1.61
1.22
1.52
1.43
1.38
0.94
1.69
1.39
1.30
1.40
5.73
74
-------
TABLE 14. RAIN PER STORM IN INCHES PER RAIN DAY
ANNUAL
Average Std. Dev
0.21
0.02
Coefficient of
Variation
0.11
95% Confidence
Level Interval ±
Jan.
Feb.
Mar.
Apr.
May
Jim.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
0.14
0.17
0.17
0.21
0.23
0.29
0.26
0.33
0.23
0.21
0.18
0.14
0.05
0.07
0.05
0.07
0.09
0.12
0.13
0.11
0.10
0.12
0.06
0.06
0.40
0.40
0.31
0.33
0.37
0.41
0.51
0.33
0.43
0.55
0.34
0.43
0.20
0.03
0.02
0.03
0.04
0.05
0.06
0.05
0.04
0.05
0.03
0.03
0.01
From Tables 12 through 14, it can be seen that monthly precipitation in
Rochester is fairly uniform throughout the year. However, as expected, the
months of June, July, and August exhibit the higher intensity-shorter dura-
tion storm events. During the period of 1948 to 1973, the greatest total
amount of rainfall presented by a storm event was 3.91 in. over 87 hr. The
greatest maximum hourly intensity was 1.35 in. associated with a duration of
8 hr and a total rainfall depth of 2.15 in.
A rainfall characterization routine in the SSM defines discrete storm
events and ranks design parameters associated with each storm. The input
data to this program are the hourly rainfall record from the U.S. Weather
Bureau. For this study, a discrete storm event has been defined as starting
with the first measurable rainfall after a minimum interval of six hours with
no rainfall and ending when a gap in measured rainfall of at least six hours
is first encountered. For each event in the historical record, the following
parameters are calculated: date, starting hour, duration, total rainfall,
maximum hourly rainfall and the hour in which it occurred, elapsed hours since
the previous storm, occurrences of excessive precipitation, snowfall, and the
ratio of hour of maximum rainfall to total duration (r value). Figures 23
through 26 are examples of frequency curves based on ranked rainfall data as
provided by the model. Design storm parameters can be determined from these
curves. The validity and usefulness of any of these curves is directly
related to the length of record available.
In any storm pattern, the three most important characteristics affecting
the peak runoff rate for a specific duration are:
(1) volume of water falling within the maximum period,
75
-------
1 0
e
2
o a
o B
- o 4
0. I
01 Q.2 040
1 2 4 e 8 1 0 20
OCCURRENCES PER YEAR
40 60 80 100
FIGURE 23. Example Curve - Storm Magnitude vs Frequency
5 ° ' 0-2 0406 1
4 6810 20 40 60 BO 100
OCCURRENCES PER YEAR
FIGURE 24. Example Curve - Storm Intensity vs Frequency
76
-------
1 0
LU B
UJ
O
X
ae
o
0 6
0 2
0 1
0 1
0. 2
0.4 0
t 2 4 6810
OCCURRENCES PER YEAR
20
40 GO BO 100
FIGURE 25. Example Curve - Storm Duration vs Frequency
30
20
1 0
> 10 12 14 16
HOUR AFTER START OF STORM
20
FIGURE 26. Example Curve - Percent of Storms Having
Maximum 1-Hour Intensity vs Hour After Start of Storm
77
-------
(2) amount of antecedent rainfall, and
(3) location of the peak rainfall intensity.
The application of the Keifer and Chu method (7) to produce a given design
storm hyetograph requires the consideration of the statistical average of
these three characteristics.
For subsequent abatement alternative analyses involving the SWMM, initial
runs would involve the one, two, and five year return period design storms
utilizing the synthetic hyetographs. A two-hour duration storm event was
selected in order to simulate a high-intensity storm. It must be noted that
the selection of this specific storm duration was consistent with the
selection of the design storm frequency utilized in the ongoing facilities
planning activity. Rainfall events of different durations would be expected
to differ significantly in their effect on the sewer network. The location
of the peak rainfall was determined from a statistical analysis of approxi-
mately 180 storms that occurred in the Rochester area between 1933 and 1973 (5),
The equation relating rainfall intensity to duration for any return
period can be represented mathematically by: i = a (tb + c)"1 where
i = rainfall intensity (in./hr)
t = duration time (min)
a,b,c = constants, dependent on specific region and terrain.
Table 15 summarizes the equation constants for Rochester, New York for various
return periods.
TABLE 15. RAINFALL INTENSITY-DURATION EQUATION CONSTANTS
Return Period in Yearsabc
1 28
2 42
5 68
0.85
0.88
0.90
7.0
7.7
10.0
Figure 27 shows the rainfall intensity-frequency-duration curves for
Rochester, New York.
The general equation used for generating synthetic hyetographs is as
follows:
i =
where the variables t, a, b, and c represent the same parameters as in the
intensity-duration equation. The term r represents that portion of a storm
duration occurring before the most intense moment, expressed as a ratio to the
entire duration. A value of r = 0.5 was estimated for the Rochester study
78
-------
20.0
15.0
;8ase Curves Obtained From United States Weather
-Service Technical Paper No. 25
15 20
ii
0 15 20 30 4050 60
Minutes
DURATION
Hours
FIGURE 27. Rainfall Intensity-Duration-Frequency Curves
For Rochester, N.Y.
79
-------
area based on the statistical analysis conducted on the recorded 180 storms.
Figure 28 shows the generated synthetic hyetographs for the one, two, and
five year design storms.
SWMM Version II is a specific storm event model with its simulation
dependent on single input rainfall hyetographs. To assess the various CSO
abatement alternatives on a return period basis,it was necessary to
characterize rainfall events on the same basis. The SWMM overflow
projections are based on the one, two, and five year design storms. All
average annual overflow volumes predicted through the use of the SSM are
based on 20 years of daily rainfall data. The use of hourly rainfall
data provides greater accuracy in predicting overflow quantities; however, the
SSM is not ideally structured to analyze an entire sewer system using
hourly data. Several runs were made to compare the results obtained
from daily and hourly rainfall data. A reasonable first-cut analysis of
a CSO system can be made with the SSM using daily data.
For all subsequent analyses involving overflow quality and abatement
alternative evaluations a 2 Yr-2 Hr design storm was selected. The use
of the 2 Yr rainfall return frequency was consistent with other ongoing
abatement programs for Monroe County, New York (3). The 2 Hr duration was
selected because the times of concentration for all tributary drainage
areas to the main interceptor are less than 2 hr and thus all of the
associated area would be contributing runoff at its maximum rate which
was important for SWMM runoff projections. In addition, it was felt that
the subsequent abatement analyses would be most effective when alternatives
are based on average storm durations; that is, alternatives that reduce CSO
pollution for most storm events.
OVERFLOW QUALITY ANALYSIS
A summary of all the CSO monitored quality data for 6005, SS, and
t. coliform collected at thirteen overflow sites during 1975 is presented
in Table 16. This table also presents the basic CSO quality statistical
data which are basic to the following discussion.
The thirteen sites represent those monitored overflow locations shown
in Figure 8. The relationship of the overflow sites to each of the major
trunk sewers and interceptors of the sewer system for the Rochester Pure
Waters District is also shown.
Although the magnitudes of the arithmetic and geometric means vary con-
siderably, relative to the entire drainage area of the Rochester Pure Waters
District, quality parameters of 6005, SS, and t. coliform are reasonably well
defined by either of the two means. As an example, the ratio of the arithmetic
mean of BOD5 for drainage area no. 22 with respect to the average BODs of all
drainage areas is 4.21; whereas, the same ratio using geometric means is 4.17.
It can be reasonably inferred that the average BOD5 concentration of drainage
area no. 22 is approximately 4.2 times greater than the overall average
drainage basin BODs mean. The other drainage areas exhibit similar consis-
tency between their arithmetic and geometric means. Expressed mathematically,
80
-------
S-
3
O
7.0 -r
6.0 ..
5.0 -
4.0 ..
3.0
2.0 ..
1.0 -
0.0
Return Period
O 5 Year
A 2 Year
1 Year
60
10 2050 40 50 60 70 30 90 100
Time Measured from Beginning of Rainfall in Minutes
50 40 30 30 10 0 10 20 30 40
Time Measured from Peak Rainfall in Minutes
110 120
50
60
FIGURE 28. Synthetic Design Storm Hyetographs
81
-------
TABLE 16. SUMMARY OF DRAINAGE AREA CSO QUALITY
CO
1X5
BOD5(mq/l)
Drainage Arithmetic Geometric
Y
Z
*
**
***
Area Mean
7 134 (+28)**
16 97 (+31)
22 478 (+49)
8 69 (+18)
9 130 (+29)
17
18 63 (+11)
21 136 (+86)
25 61 (+21)
26 15 (+ 4)
28 78 (+ 9)
29 31 (+ 9)
31 69 (+10)
Y 113
= Average of Arithmetic
= Mean Value/Y
= Refer to Figure 8 for
= Values in parenthesis
Mean
79
74
308
47
118
44
75
49
11
57
25
61
Means
location
Z
1.18
0.86
4.21
0.61
1.15
0.56
1.20
0.54
0.13
0.69
0.27
0.61
Ari
449
474
591
169
220
194
511
258
644
331
235
177
351
SS (mg/1)
thmetic
Mean
(+ 80)
(+264)
(+ 65)
( + 15)
(1 43)
(+ 48)
(+338)
( + 67)
Geometric
Mean
247
306
438
111
200
106
269
212
(+790)*** 121
( + 52)
(+ 57)
( + 17)
193
197
137
(MPN/100
Arithmetic
Z
1.28
1.35
1.68
0.37
0.63
0.55
1.46
0.74
1.83
0.94
0.67
0.50
5.1
7.2
8.3
2.0
10.2
7.1
14.1
2.0
0.3
5.9
5.2
6.1
Mean
(+1.4)
(+2.9)
(+1.3)
(+0.5)
(+6.6)
(+1.4)
(+7.7)
(+1.5)
(+0.1)
(+1.2)
(+0.8)
ml)x!0 6
Geometric
Mean
1.7
4.5
3.9
1.0
4.4
4.0
9.0
0.8
0.2
2.4
3.1
Z
0.84
1.18
1.36
0.33
1.67
1.16
2.31
0.33
0.05
0.97
0.85
of drainage areas
represent 95%
= Value not used in analysis of
confi
arithmentic
dence i
ntervals
and geometric mean
comparison
-------
ZI»N=XI>N/YI and
7 = Y / Y
Z2>N ~ X2'N/ 2
where X = arithmetic mean of a specific quality parameter
Y = average of all drainage area arithmetic means
(for a particular quality parameter)
Z = characteristic ratio of drainage area mean
to overall drainage basin mean
N = number of the drainage area
1 = arithmetic mean
2 = geometric mean
The 95% confidence intervals for the ratio Zj/Z2 are shown in Table 17.
TABLE 17. ZT/ZQ CONFIDENCE INTERVALS
1 £
Quality Parameter 95% Z,/Z2 Confidence Intervals
BODs 0.96 ± 0.06
TSS 0.88 ± 0.10
t. coliform 1.07 ± 0.14
Several conclusions were derived from analyzing the overflow monitoring
data:
1. Drainage areas nos.. 7, 21, and 22 have annual mean BODs and SS
concentrations exceeding the mean for the overall drainage
has i n.
2. Drainage areas nos 21 and 22 have annual mean concentrations for
all three quality parameters exceeding the basin average.
3. Pollutant concentrations from overflows discharging to
Irondequoit Bay are significantly lower than most of those
overflows discharging to the Genesee River.
4. River overflows nos. 8 and 25 similarly exhibit lower concentrations
than those of the remaining River overflows.
5. Most overflows except nos. 8, 9, and 22 exhibit SS concen-
trations on the order of twice the 8005 concentrations.
83
-------
6. In general, those drainage areas with population densities of
approximately 18 people per acre and greater, show a t. coliform
concentration mean greater than the overall basin average.
The data presented in Table 16 in conjunction with the overflow volumes
predicted by the SWMM under the application of the synthetic 2 Yr-2 Hr
design storm lend insight to the pollutional impact that the various
overflows have on the Genesee River and Irondequoit Bay. Table 18 lists the
volume and pollutional load placed on the receiving waters by each overflow.
The total mass of pollutants discharged to the Genesee River was computed
from the volume of overflow over 30 min periods, as determined by the
model projected hydrographs, and the concentration of these pollutants
averaged over the same time periods, as determined from actual monitored data.
This procedure is described in the following section. The overflow identified
as the Siphon represents a relief point on the main interceptor. The
ranking of the individual drainage areas with respect to pollutant loads per
volume of overflow and per acre associated with each individual drainage
area is presented in Table 19. These rankings indicate that the predominately
residential areas contribute a lower areal mass loading of pollutants than
industrial and commercial areas. The relatively large pollutional loads
imposed by drainage area no. 22 result from the location of a large brewery in
this area discharging directly into the District sewerage system without
pretreatment.
TABLE 19. RANKING OF DRAINAGE AREAS PER MASS LOADINGS
Drainage Area
No.
8,9
7
25
16
21
22
Predominant
Land Use
Residential
Residential
Commercial
Commercial
Residential
Residential
Ranking - %
of River Overflow
1
2
3
4
5
6
Ranking - BODs
Loading Ib/ac
4
5
3
2
6
1
Ranking
Loading
5
4
2
3
6
1
- TSS
Ib/ac
QUALITY ANALYSIS FOR USE IN ABATEMENT ALTERNATIVE STUDIES
Before an abatement alternative can be evaluated, the pollutional load
imposed on the receiving waters must be determined. Analytical data from
the sampling of the overflows generated in 1975 were statistically analyzed
to yield geometric mean concentrations of selected pollutant parameters
within pre-determined time intervals after the start of an overflow event.
The time interval selected for analysis was 30 min which yielded three
intervals for the first 90 min of the overflow and a final interval
extending from 90 min to the end of the entire overflow event.
A 30 min interval was selected on the basis of the 15 min sampling
frequency conducted under the monitoring program in conjunction with the
84
-------
TABLE 18. DRAINAGE AREA OVERFLOW ANALYSIS FOR 2YR-2HR DESIGN STORM
00
CJ1
Drainage
Area
No.
7
16
22
8 & 9
Front St.*
Siphon*
21
25
28
29
31
Totals
Overflow
(mil gal)
9.1
8.5
5.8
41.4
6.5
13.0
8.4
8.5
10.1
20.9
10.5
142.7
I of Total
Overflow
6.4
6.0
4.1
29.0
4.6
9.1
5.9
6.0
7.1
14.6
7.4
% of River
Overflow
(101.2 mil gal)
9.0
8.4
5.7
40.9
6.4
12.8
8.3
8.4
Total
BODn
Obi
6,000
5,250
14,900
25,710
2,390
4,770
5,250
3,470
67,740
BOD5
(Ib/acre)
6.7
10.1
24.0
8.1
3.8
8.5
Total SS
(lb)
18,750
21,690
21,190
51,450
5,750
11,490
18,850
15,030
164,200
SS
(Ib/acre)
20.4
28.7
41.9
16.3
6.4
36.8
*Drainage areas identified asnos. 17 and 18 in Table 16 have been deleted since overflows
from these two basins have been eliminated due to a new interceptor. Front St. and Siphon
represent overflows/locations that were not field monitored under this program. BOD5 and SS
loadings based on concentrations (geometric means) presented in Table 16 and volumes based on
projections by the SWMM.
-------
assumption that large concentration changes can be best represented by such
an interval. This statistical analysis provided average concentration
profiles for 6005 and TSS to be expected for each overflow event at each
point of discharge. Due to the large amount of analytical data collected
during 1975, it is assumed that for all the storm event evaluations con-
sidered in this study, the calculated concentrations for the selected time
intervals remain reasonably constant. Table 20 shows the pollutant con-
centrations as determined for each drainage area.
TABLE 20. OVERFLOW QUALITY BY TIME INCREMENT
Drainage Area
Concentration (mg/1)
No.
7
8
9
16
17
18
21
22
25
26
28
29
31
0-30
BOD5*
141
135
146
124
131
112
67
342
47
24
51
19
152
min
TSS*
316
211
200
228
114
211
119
586
238
238
194
121
191
30-60
BOD5*
83
58
176
173
-
104
38
268
33
20
41
27
47
min
TSS*
294
133
228
822
-
291
83
476
189
353
181
104
168
60-90
BOD5*
43
49
147
105
129
54
35
213
47
*
28
27
71
min
TSS*
123
129
198
508
357
154
62
392
189
198
178
120
174
90+ mi
BOD5*
44
43
109
38
22
35
50
210
39
*
40
*
*
n
TSS*
95
144
200
241
60
88
70
332
84
75
154
81
100
BOD5 and TSS are geometric means in mg/1 from the 1975 analytical data.
*Indicates zero values(s) in data set, thus geometric mean not valid.
From Table 20, the conditions known as first-flush, in which a dis-
proportionately high pollutional load is carried in the first portion of an
overflow, is generally observed in each of the drainage areas. Based on
an analysis of the data, the first-flush portion of the overflow event has
been defined as the first 60 min. From the concentrations and volumes of
overflow the total pollutant load to the receiving waters from each overflow
site can be determined. Table 21 shows the overflow volumes within the 30
min time intervals determined by the SWMM under the application of the
2 Yr-2 Hr design storm.
The sewer system network modeled included the existing sewer system
with the new Genesee Valley Interceptor (and subsequent overflow from
drainage areas nos. 17 and 18), the Genesee River Southeast Interceptor
Tunnel, and the Culver-Goodman and Cross-Irondequoit Tunnel complex as
identified in Figure 29. Since sampling and monitoring were not con-
ducted at the Front Street and Siphon overflow sites, overflow volumes
by time interval were unnecessary. In addition, since all the overflow
discharging from drainage areas nos. 28, 29, and 31 will be conveyed
to treatment facilities by the Cross-Irondequoit Tunnel, it was considered
86
-------
AREA *
ISO
«-i Clarissa St Tunnel
Genesee Valley
Interceptor
System
'^23
y JG
Genessee River
SouTneast
Interceptor
263
Culver-
Goodman
12 Tunnel -Complex
AREA~8
Van Lore
STP
RS.
Cross -Irondequoif
Tunnel
Node * Conduit Linked Network
for the SWMM Input
AREA '31
FIGURE 29. Present-Proposed Tunnel Interceptor System
87
-------
sufficiently accurate to use mean concentration values for the entire storm
event and total overflow volumes to determine their pollutant loading for
further abatement alternative analyses.
TABLE 21. OVERFLOW VOLUME BY TIME INCREMENT
2 YR - 2 HR DESIGN STORM
Drainage Area
No.
0-30 min
Overflow Volume (mil gal)
30-60 min 60-90 min
90+min
7
8,9
16
21
22
25
28
29
31
0.48
2.88
2.38
0.54
2.21
1.80
1.21
6.88
2.09
3.43
10.01
3.62
1.89
2.36
3.28
4.64
10.71
5.89
3.26
9.85
2.00
2.06
0.86
1.54
2.29
2.65
1.80
1.92
18.67
0.51
3.94
0.37
2.28
1.91
0.64
0.72
Table 22 shows the BOD5 and TSS loadings, by time intervals, discharged
to the River from overflows induced by application of the design storm.
Since drainage areas nos. 8 and 9 have a common overflow discharge conduit,
the loadings from each area were combined by using the geometric mean of the
individual area means for the pollutant concentrations. Total loadings
were given for Front Street and the Siphon, based upon concentrations of
30 mg/1 BODs and 89 mg/1 TSS, and 65 mg/1 BODs and 273 mg/1 TSS, for each
overflow site, respectively. The concentrations for Front Street represent
approximate means of the concentrations from overflows from drainage areas
nos. 17, 18, and 26. With the elimination of the overflows from areas nos.
17 and 18, upon completion of the Genesee Valley Southwest Interceptor,
the first relief point on the system will then be at Front Street. Con-
centrations for the Siphon overflow were assumed similar to those from
overflow no. 16.
TABLE 22. OVERFLOW POLLUTANT LOADINGS BY DRAINAGE AREA AND TIME INCREMENT
2YR-2HR DESIGN STORM
Loadings (Ib)
0-30 min
30-60 min
60-90 min
90+ min
Drainage Area
No.
TSS BQDs TSS
BODs
TSS BODs
TSS
7
8,9
16
21
22
25
Front Street
Siphon
560
3360
2460
300
6300
710
Loadings
1270
4920
4530
540
10800
3570
in text
2370
8430
4020
600
5270
900
8410
14530
9720
1310
9370
5170
1170
6900
1730
600
1530
600
3340
13140
8470
1070
2810
2430
700
10590
160
1640
650
740
1520
26310
1030
2300
1020
1600
-------
Drainage Area
No. Total Loadings (1b)
BO DC; TSS
7
8,9
16
21
22
25
Front St.
Siphon
4800
29280
8370
3140
13750
2950
1620
7020
Totals 70930
14540
58900
23750
5220
24000
12770
4790
29480
173450
Presented in Table 23 are the loadings to the Cross-Irondequoit Tunnel
from the East Side Trunk Sewer overflows for the defined design storm.
These loadings are based on data from Tables 20 through 22. The total
combined sewer overflow loading to the Genesee River for the 2 Yr-2 Hr design
storm is 70,930 Ib BOD5 and 173,450 Ib TSS. Pollutant loading representing
the first-flush is 43,920 Ib BOD5 and 74,140 Ib TSS. The overflow volume
associated with the first-flush is 34.88 mil gal.
For a level one analysis associated with subsequent screening of
possible abatement alternatives, it was advantageous to determine the average
annual BODs and TSS loadings to the Genesee River. The Simplified Stormwater
Model is ideally suited to calculate such quantities. Twenty years (1954-
1973) of daily precipitation records formed the basis for model simulation.
The sewer network on which the model was applied was similar to that
analyzed using the SWMM in terms of in-system storage volumes and interceptor
conveyance capacities. Pollutant concentrations that were associated with
overflow volumes establishing mass loadings were based on the 1975 analytical
data base. Geometric means of samples taken throughout the duration of the
overflow event were assumed sufficiently accurate to be coupled with overflow
volumes to yield the mass loadings of the pollutants. Sensitivity analyses
conducted on the SSM using both daily and hourly rainfall data indicated
that the total overflow volume discharged to the Genesee River as a result
of daily rainfall data should be increased by 60 percent to more accurately
represent actual system rainfall-runoff-overflow response as determined by
application of hourly data. Table 24 shows the results of the simulations
and subsequent projections made with the SSM.
Assuming a total urban tributary drainage area to the Genesee River
of 9172 acres and assuming a gross runoff coefficient of 0.431, the average
annual rainfall for Rochester of 30.67 translates into approximately
3300 mil gal of runoff. Therefore, based upon the projections using the SSM,
one-third of all rainfall becomes combined sewer overflow with subsequent
discharge to the River while only 10 percent of the total rainfall remains
in the sewer system for possible treatment under present conditions.
89
-------
TABLE 23. POLLUTANT LOADINGS BY TIME INCREMENT 2YR - 2HR DESIGN STORM
EAST SIDE OVERFLOWS
Drainage Area
No.
28
29
31
Loadings (Ib)
0-30 min.
BOD5 TSS
510 1960
1090 6940
2650 3330
30-60 min.
BOD5 TSS
1590 7000
2410 9290
4760 8250
60-90 min.
BODR TSS
730 3400
600 2650
1070 2610
90+ min.
BODR TSS
640 2450
130 430
600 600
Total Loadings (Ib) Total Overflow (mil gal)
28
29
31
BOD5 TSS
3470 14810
4230 19310
9080 14790
16780 48910
10.05
20.88
10.50
41.43 ;
TABLE 24. AVERAGE ANNUAL LOADING TO GENESEE RIVER SIMPLIFIED STORMWATER MODEL
Drainage Area
No.
Overflow
(mil gal)
BOD5*
(mg/1)
TSS*
(mg/1)
BODs
(Ib)
TSS
(Ib)
7 35 63
8,9 1370 85
16 445 135
21 51 45
22 277 284
25 285 40
Front St. 2£ 30
Totals 2489
Avg Daily Loading:
192
171
547
74
496
114
89
18390
971190
501030
19140
656090
95080
6510
2267430
6210
56050
1953810
2030080
31480
1145850
270970
19300
5507540
15090
*Note: These values represent average pollutant concentrations based on
volumes given in Table 21 and concentrations given in Table 20.
Treatment Considerations
To further aid in assessing various abatement alternatives, analyses
were conducted on the use of primary swirl concentrator units installed as
site treatment facilities at consolidated overflow locations. The size,
efficiency, and performance of the swirl units used for the analyses were
determined from the pilot plant studies (8). Based upon work done at the
LaSalle Hydraulic Institute and the pilot plant work conducted under this
90
-------
study there appears to be a design Froude number for optimum unit performance.
For this study a Froude number of 0.167 was established. The design Froude
number in conjunction with the established design overflow rate, defined as
that flowrate occurring at the end of the first hour of overflow (previously
defined as the first-flush),was used to develop the size of the primary
swirl units for each overflow location. Table 25 shows the number of swirl
units required for each site based on sizing requirements presented in
Volume II of this report.
TABLE 25. NUMBER OF REQUIRED SWIRL UNITS BY DRAINAGE AREA
Drainage Area Number
No. of Units
7 4
8,9 6
16 4
21 3
22 3
25 3
Front St. 2
Siphon 2
The efficiency of the primary swirl unit is related to the influent
solids concentration and the Froude number which is determined from the
actual influent flowrate. Table 26 shows the overflow storage capacity
required to capture the first-flush volume for the 2 Yr-2 Hr design storm.
These volumes were determined by application of the SHMM. The
overflow in excess of the first-flush volume under this alternative was
to be applied to the primary swirl units for solids removal. Based on
the size and efficiency of each swirl unit, Table 27 indicates the amount
of solids removed by the design swirl units treating post first-flush flows.
TABLE 26. OVERFLOW STORAGE CAPACITY REQUIRED TO RETAIN FIRST-FLUSH
2YR-2HR STORM
Drainage Area Volume
No. (mil gal)
7 3.91
8,9 12.89
16 6.00
21 2.43
22 4.57
25 5.08
91
-------
TABLE 27. SWIRL UNIT SOLIDS REMOVAL EFFICIENCY OF FIRST-FLUSH
2YR-2HR STORM
Drainage Area
No.
TSS Influent
Ob)
TSS Removed
Ob)
Efficiency
(*)
7
8,9
16
21
22
25
Front St.
Siphon
4860
39450
9500
3370
3830
4030
4790
29480
Totals 99310
Avg. System Efficiency
760
11900
6500
0
2430
950
0
14480
37020
16
30
68
0
63
24
0
49
0.37
Therefore, with capture of the first-flush and primary swirl unit
treatment on the post first-flush flows, the total TSS load to the Genesee
River would be reduced to 62,290 Ib; that is, the swirl units remove an
additional 37,020 Ib of TSS. The removed solids could be pumped back into
the St. Paul Boulevard Interceptor and conveyed to wet-weather treatment
facilities at the Van Lare STP. In addition, the solids removed from the
overflows from the east side of Rochester which discharge to the Cross-
Irondequoit Tunnel and are conveyed to the Van Lare STP, must be determined.
Following a similar procedure for calculating BOD5 and TSS loadings associated
with the River overflows, Table 28 presents the pollutant loadings and flows
projected for several east side overflows discharging to the Tunnel.
TABLE 28. EAST SIDE POLLUTANT LOADINGS FOR 2YR-2HR STORM
Drainage Area
No.
28
29
31
Sub-Totals
Totals
F-F
5.85
17.59
7.98
31.42
Flow
(mil gal)
Post F-F
4.20
3.29
2.52
10.01
41.43
BOD5
(Ib)
F-F Post F-F
2100 1370
3500 730
7410 1670
13010 3770
16780
TSS
(Ib)
F-F Post F-F
8970 5850
16230 3080
11580 3210
36770 12140
48910
92
-------
SECTION 8
DESCRIPTION AND EVALUATION OF ALTERNATIVES
GENERAL
The 1972 Council of Environmental Quality Third Annual Report
indicated that in 80 percent of the urban areas studied, downstream
water quality was not controlled by point sources. The urban contribution to
these water quality contraventions of defined standards is in the form
of stormwater runoff and combined sewer overflow.
There are three general approaches available for the control of
stormwater and combined sewer overflow discharges. A combination of storage
and treatment can be applied through the use of structurally intensive
measures to attain defined water quality standards. Secondly, the sources
of the pollutants to the sewer system can be addressed through the
application of the Best Management Practices (BMP) concept. Thirdly,
a combination of the two can be applied. In general, the most cost-
effective abatement alternative will be the application of structural
and non-structural (BMP) measures.
In light of the very significant and substantial capital and operating
costs associated with the implementation of structurally intensive alterna-
tives, the application of Best Management Practices offers itself as a very
attractive alternative to the solution of urban runoff problems. Optimized
source and collection system management can provide a more quickly facilita-
ted and less costly alternative for the abatement of wet-weather induced
quality impairment of the receiving waters.
For this Report the alternatives that were evaluated can be classified
as nonstructural, minimal structural, and structurally intensive. The
first two classes of abatement alternatives are those generally considered
in developing a BMP program.
NONSTRUCTURAL ALTERNATIVES
Land Use Policies
Land use policies and sewer system maintenance procedures can sig-
nificantly affect both the quantity and quality of combined sewer overflows.
Building, plumbing, and zoning codes directly influence the percent imper-
viousness and dry-weather flow characteristics of a drainage area.
93
-------
Zoning ordinances are usually based on the concept of the best use of
land, that is, the use that will most significantly benefit the community as
a whole. Generally, increases in the relative imperviousness of an area
adversely alter runoff and overflow characteristics; therefore zoning which
limits development of impervious areas is beneficial from the aspect of storm-
water management. Urban renewal initiated changes in zoning generally tend
towards the development of open spaces or towards lower imperviousness areas.
Citizen and commercial initiated changes are often opposite to those
initiated by urban renewal, especially when land is bought on speculation for
later development involving commercial shopping centers and apartment com-
plexes. Therefore, because urban land has a high market value, zoning changes
favoring open spaces, and thus stormwater management measures, are difficult
to implement. There are, however, zoning alternatives that favor overflow
reduction.
Planning boards and reviewing agencies may establish codes benefiting
overflow reduction. Restrictive ordinances, which eliminate direct entry
of sump pumps and roof drains into the sewer system, have the same effect as
reducing the imperviousness in a drainage area because runoff and peak
runoff entering the sewer are necessarily reduced. Specifying the use of
porous pavement in those areas with suitable subsoil conditions is another
method that also has the effect of increasing perviousness. Planned open
space within developed areas, which can be used for both recreation and
surface storage, is multifunctional and maximizes land use.
Street Cleaning Practices
Street cleaning in the City of Rochester includes the cleaning of any
debris that may be found on the streets. This operation not only includes
street sweeping but also the picking up of bulk material which may be left
over by refuse collection and tree care operations.
The City of Rochester is divided into 42 residential and seven central
business district cleaning routes. The normal street cleaning schedule for
the residential routes is every sixth working day; that is, every street
will be cleaned on both sides every sixth working day. This schedule is
followed to avoid the alternate parking problem. The seven central business
district routes are cleaned each working day between the hours of 5 AM and
1 PM. Scheduled cleaning operations may be delayed several days at times
due to adverse weather conditions.
The vehicle used by the City of Rochester is an Elgin & Wayne Sweeper
having an approximate cost of $40,000. The City owns 17 such vehicles.
Only nine sweepers are, however, employed during normal operations. The
efficiency of these sweepers depends to a large degree on the operator.
According to City maintenance personnel estimated efficiencies during the
spring months average between 60 and 70 percent while during the summer
months the efficiencies may average between 80 and 90 percent. Efficiency
of the street cleaning operation is also dependent on the number of cars
parked on the street. Sweeper maintenance is very high; however, the
average vehicle repair time is usually less than a day.
94
-------
Immediately following the winter months, a major spring cleanup
program is implemented for a period of six weekends. At this time,
additional equipment from other divisions is used on weekends to help
remove bulk materials from the street. During this period, the weekday sweeper
operations are concentrated in those areas where bulk materials are
removed. During spring cleanup, the cleaning of the central business
district's routes is increased from five to six days per week, weather
permitting. A program similar to spring cleanup is also implemented for
a period of 8 to 10 weekends in the fall to remove leaf accumulations.
The estimated capacity of the Elgin & Wayne Sweeper is about 3 cu
yd. Material that is picked up by the sweeper is taken to a transfer
station where the material is transported to a sanitary landfill.
Presently, there are about 564 miles of street being cleaned in the
City of Rochester, covering an area of about 36 sq mi. Cost estimates
for the Rochester street cleaning operations are about $0.37 per foot of
street per year or approximately 1.1 $ mil per year. This figure represents
the total costs involved in the entire street cleaning operations for
the City of Rochester Department of Public Works. Included therefore,
are manpower, vehicle purchase, operation and maintenance, and hauling
away of the collected solids.
Increased frequency and efficiency of street cleaning, up to a point,
results in a noticeable decrease in the solids that accumulate during
periods of no rainfall. However, increased volumes of solids must
necessarily be processed and transported to a suitable landfill area.
Also, there may be more toxic concentrations or amounts of leachates in
the additional solids removed.
Although the SWMM quality routines did not simulate to any reasonable
degree measured overflow pollutant concentrations, the benefit of increased
street cleaning as predicted by the SWMM model is shown in Figures 30
and 31 for the Thomas Creek drainage area. Thus, the following discussion
actually represents a sensitivity analysis of the quality portion of the
SWMM. In the following seven figures CLFREQ refers to the number of days
between street cleanings and land use classifications 1, 3 and 5 refer to
single family residential, commercial, and park lands, respectively. Using
the June 19, 1975 storm as a base, decreasing the street cleaning frequency
from 5 to 25 days resulted in a corresponding increase in pollutant
loading to the river of 52 percent for BODs and 29 percent for TSS. For
this simulation the number of sweeper passes has been set at one and the
number of antecedent dry days has been set at 30. Antecedent dry days
represent the total number of days when the cumulative rainfall is less
than one inch of rain. On the average, a rainfall event occurs every
four days in the Rochester area with an average ten day rainfall of
about 0.9 inches. Therefore, the 30 day antecedent conditions represents
an extreme situation, but previous model simulations indicate that
little change occurs in runoff quality by increasing the number of dry
days beyond 15 days which is about the number of days required to accumulate
one inch of rain in Rochester.
95
-------
1001-
80 -
\
\
Area No. 31
Antecedent Dry Days =30
Sweeper Passes = I
CLFREQ = 25
Land Use Changed From I to3
5 to I
2 Yr-2 Hr storm
w
^
o
QD
60
40
20
0
' // \ 20124 lb\
/ / "A
' \
/ / \
/' s
1 !
1 /
- j/J
//' | ,
0600 0700 08OO
Time
\ 6/19/75 si
\ CLFREQ = !
\x " 2 Yr-2 Hr
\ V^-7736lb
\ x
\ ^^
\ \.
\ ^\
V-O^^
Ih i-^
Iw ^ ^^^-«^. ^^**NBII^
1 1 J
0900 1000 1100
(hr of day)
FIGURE 30. Runoff BOD5 Profiles for Various Storms and
Street Cleaning Frequency
lOOo -
800- 231,660 Ib-
V
6OO -
\
Area No. 31
Antecedent Dry Days - 30
Sweeper Passes = I
CLFREQ * 25
Land Use Changed From I to 3
5 to I
2 Yr-2 Hr storm
500
200
0
(X
M / \ \v w *~" '
\ l^f \ X CLFREQ
1 \ \\ 9 Yr-?
1 1 / G7525 Ib ^ >
1 I / 117 ?T7 Ih ^~~ku ._ N.
sJfl 1 1 1
500 0700 0800 0900 DOO
J 3 Ul
= 5
Hr
\
1
1100
Time (hr of day)
FIGURE 31. Runoff TSS Profiles for Various Storms and
Street Cleaning Frequency
96
-------
Figures 32 & 33 give a street cleaning comparison for drainage
area no. 31 for both the June 19, 1975 storm and the 2 Yr-2 Hr storm.
Figure 32 shows that the total overflow from area no. 31 is 14.3 mil gal
for the June 19, 1975 storm and 39.5 mil gal for the two year storm.
Cleaning frequency and efficiency have no effect on overflow quantities.
The effect on overflow quality of increasing the cleaning frequency
under the 2 Yr storm is similar to that projected under application of
the June 19, 1975 storm; decreasing the frequency from 5 to 25 days
results in a decrease of approximately 50 percent in 6005 and TSS loading
from the combined sewer overflows. Despite the differences between
storms (0.6 in./hr peak for the June 19, 1975 storm and 1.15 in./hr peak
for the 2-Yr storm) the overflow concentrations for BODs and TSS
follow very closely in both magnitude and frequency when the cleaning
frequency is 25 days. This observation is verified by the fact that the
total BODs and TSS loadings for the June 19, 1975 storm are in direct
proportion to the 2-Yr storm loadings.
The efficiency of street sweeping depends not only on the frequency
but also on the efficiency of the equipment used. From the previous
discussion it would appear that the average efficiency is about 50 percent
which is very similar to the default value applied in the SWMM for
similar cleaning conditions. The effect of cleaning efficiency in
reducing annual pollutant loading is not nearly as significant as cleaning
frequency. The effectiveness of cleaning efficiency in improving overflow
quality is not improved when the frequency is increased beyond the 5-day
interval; (See Figure 34 which presents the initial time interval extraction
of the runoff quality equation from the SWMM). In fact, it is seen that
cleaning frequency is as important as dirt and dust build-up and total
length of street to be cleaned in determining surface runoff quality.
In conclusion, the impact of street cleaning is significant.
Because drainage area no. 31 is representative of all other areas in the
city, it would appear that increasing the frequency of sweeping five-
fold results in about a 50 percent reduction in BODs and TSS loadings.
Because the pollutant build-up is greater in commercial areas, these
areas require more frequent street sweeping than residential areas (by
about a factor of four ) to maintain an equivalent overflow loading.
Note that the effect, as presented, of street sweeping operations
on the quantity of BODs and TSS discharged from the Rochester system has
been predicted solely from the application of the SWMM. The analysis,
however, does indicate a trend that can reasonably be expected. Actual
reductions can only be determined through field demonstrations.
Increased Sewer Maintenance
Sewer maintenance has potential for not only reducing system flooding
but also minimizing first-flush effects. The first-flush phenomenon
observed from River overflows is presented in Table 9 and discussed at
length in other sections. From that table, it is seen that nearly 65
percent of the total solids loading to the River occurs in the first
hour when applying the 2 Yr-2 Hr storm. Similarly about 45 percent of
97
-------
800
600
o
400
200
0
0600
Area No. 31
Antecedent Dry Days = 30
Sweeper Passes = I
CLFREQ = 25
Land Use Changed From I to3
5tol
2yr-2hr Storm
6/19/75 Storm
39.5 mil gal
0700 0800 0900 1000
Time (hr of day)
FIGURE 32. Runoff Hydrograph for Various Storms
1100
OfO
300
* 225
*-
^o
g
C 150
75
0
Area No. 31
N.% Storm 6/19/75
j\ Dry Days = 30
/ \ Sweeper Passes = 1
i '
/ ^ \
' ' \ ' CLFREQ = 5 & 25 land
' ' \ \ CLFREQ = 25 land use
use unchanged
changed to
/ ' \ \ commercial and residential
// \ \ imperviousness = 60%
I \ ^^
i . \ xx
// 10.25 \ V~ 14.31 mil gal
' / mil gal \ "* "^^^
' ) N - ^_
// X-v^ ^^"^^
/ ^"~ -~^,^^^
// ~~~ ~~ ^~~^
A i i i i
0600 0700 0800 0900 1000 1100
Time (hr of day)
FIGURE 33. Runoff Hydrograph for Various Land Uses
98
-------
DRYDAY
to
pshed "
No. of
Subcatchmen
z
i = 1
QFACT * DDFACT * GQLEN * fciFREQ + 1 + (1-REFF) + ... + (1 - REFF)CLFREQ 1 *
1 STREET SWEEPING
f AVFLOW 1'1 ~1 f -(4.6 * AVFLOW/ 12 * DELT)~|
* Ofi + °finn * nvrt-vm + 10 +1 A 1 * 1 i Q UADCA
uo f oouu * uflppA *-e- i.H-j ^ i i-e WAKLA I
ts
SUSPENDED SOLIDS AVAILABILITY FACTOR
+ [CBFACT * CBVOL * BASINS * 28.3]* [T - e -(AVFLOW * DELT/1'6 * CBVOL * BASINS[]
CATCH BASIN B.O.D. CONTRIBUTION
Note: Equation applicable for initial time-step only
DDFACT = Dirt and Dust Buildup rate (lb/Day/100 ft of Curb)
QFACT = Percentage of Dirt and Dust which is TSS or BOD5
GQLEM = Curb Length (100 ft)
CLFREQ = Cleaning Frequency (Days)
REFF = Cleaning Efficiency (%)
DRYDAY = Number of Days where accumulative rain < 1.0 in.
Pshed ~ Pounds of Pollutant Washed Off Land Surface During Runoff
FIGURE 34. SWMM Runoff Quality Equation
-------
the total 6005 loading occurs in the first 60 min of the overflow event;
however, in terms of volume of overflow, the first 60 min represents
only about 37 percent of the total River overflow. The runoff effect
for the 2 Yr storm is evident for about 10 hours; therefore the dispropor-
tionally high initial solids loading is probably due to solids deposited
in the interceptor during the antecedent dry days. Periodic flushing
of interceptors where dry-weather flow is less than 3 fps would therefore
help to decrease the effect of first-flush by scouring solids under
controlled conditions.
Data documenting major sewer complaints before and after a major
cleaning program have been compiled in the City of Rochester and are
shown in Table 29. The cleaning program conducted to date consisted
primarily of sewer flushing but also included catchbasin cleaning. As indicated,
the major identified sewer complaints were those associated with flooding
problems. For the drainage areas in which cleaning occurred, the complaint
calls were down by about 25 percent; for those in areas which cleaning
did not occur, the total complaint calls were down by less than 10
percent for the same study period. Although insufficient data is available
to say that sewer cleaning conclusively reduces flooding, it can be said
that intensive sewer cleaning during the period of March 1975 through
April 1976 had a significant impact on reducing local flooding problems,
particularly since no major sewer network changes were made during this
period.
TABLE 29 INFLUENCE OF SEWER MAINTENANCE
Drainage No. of Major Sewer No. of Major Sewer
Area Complaints Before Sewer Complaints After Sewer
No. Cleaning Program Cleaning Program
6 11 9
7 7 6
9 47 30
10 5 3
21 7 6
31 li 22.
Totals 96 76
Study Period: March 1975 - April 1976
+ Each major sewer complaint is assumed to constitute a flooding problem
Land Use Planning
As previously discussed, increased imperviousness adversely affects the
quantity and quality of stormwater and combined sewer overflow. Figures 35
and 36 show this impact. For the same cleaning frequency and associated dry
days, increasing the land use in drainage area no. 31 from its present resi-
dential zoning (relative imperviousness equals 40 percent) to commercial and
residential (imperviousness equals 60 percent) increases the total
runoff by nearly 50 percent and increases the total overflow loading by
100
-------
100
80
60
in
Q
o
m
20
0
Area No. 31
Storm 6/19/75
Dry Days =30
Sweeper Passes « I
CLFREQ = 25 Days
CLFREQ = 5 Days
CLFREQ = 25 Days Land Use
Change to 1 and 3
7736.4 Ibs
1100
0600 0700 0800 0900 1000
Time (hr of day)
FIGURE 35. Runoff BODs Profile for Various Land Uses
750
600
o>
E
450
300
150
0
Area No. 31
Storm 6/19/75
Dry Days = 30
Sweeper Passes = I
CLFREQ = 25 Days
CLFREQ = 5 Days
CLFREQ = 25 Days Land
Use Change to 1 and 3
^67,575.6 Ib
/// I
0600 0700 0800 0900 1000
Time (hr of day)
MOO
FIGURE 36. Runoff TSS Profile for Various Land Uses
101
-------
85 percent for the 6/19/75 storm. Because loadings for the 6/19/75 and
the 2 Yr design storm follow in nearly direct proportion to the
ratio of overflows, similar percentage increases in overflow volume and
loadings would occur using a 2 Yr storm for analysis. Restrictive
zoning, use of porous pavements, and eliminating direct inflow from
building roof and footing drains are effective land-use practices which
are countermeasures to increased imperviousness.
One option available to planning boards and reviewing agencies is
to require that each land use be associated with a certain percentage of
open green areas and/or certain relative imperviousness values. Legislating
green area development is easier than relative perviousness because the
former has a visible impact on the public's aesthetic awareness. The
latter, because of technological consideration, is more difficult to legislate;
therefore, building codes should be established or altered to act as
guidelines for attaining relative imperviousness values.
Building and plumbing codes are valuable and necessary tools which
potentially have a great impact on alleviating and attenuating stormwater
overflows. Enacting ordinances requiring porous pavement in parking
lots and driveways could virtually eliminate runoff from impervious
nonroad areas especially if roof drains are removed from the stormwater
system and the runoff placed directly on the driveways and/or parking
lots. Previous work (18) involving porous pavement indicates that
commercial parking lots and driveways would require about 5 in. of
porous pavement. In Rochester, a subbase of 2 ft would be required to
prevent freeze-thaw damage to pavements. This configuration of porous
pavement and subbase gives a storage capacity of about 10 in. of rain
which is far in excess of the 2 Yr-2 Hr design storm total rainfall
of 1.15 in. Because the roof area to impervious area ratio is
about 0.5 for commercial areas, the total rainfall that could be put on
parking lots and driveways would be 2.3 in. for a 2 Yr storm providing
roof drains were removed from the sewer system. Furthermore, the porous
pavement has the capacity to store far more than this 2.3 in., since
the average 10-day rainfall in Rochester is about 0.9 in., and the soil
required for percolation is a fine sand and silt mixture which would
require no subdrains. Rochester soils are highly variable and consist of
glacial drift, shale, and limestone that have a much higher permeability
than silt and sand. Therefore, the use of porous pavement in conjunction
with separtion of roof drains from sewers could virtually eliminate
runoff tributary to combined sewer systems from commercial and industrial
areas which constitute about one-fourth of the total city area.
Residential areas amount to one-third of the city area and have a
relative imperviousness of about 40 percent with a roof area to total
impervious area ratio of about 35 percent. Removing roof drains from
sewers in this area would not eliminate runoff from roof areas but would
attenuate this runoff; therefore, the imperviousness of the area would
effectively be reduced to a value between 25 and 40 percent. Using
porous pavement for driveways and placing roof drainage on these driveways
would reduce the percent imperviousness in residential areas to about
15 percent which is similar to the area presently represented by the
102
-------
streets. The annual effect of imperviousness on total annual overflow
in drainage area no. 31 is shown in Figure 37.
Present land use results in an annual overflow of about 22 mil gal
while removing roof drains would decrease overflows by 14 mil gal for
the same land use. Commercialization of drainage area no. 31 using
present building and plumbing codes would result in an annual overflow
from this area of almost 100 mil gal. Using porous pavement and discharging
roof drainage on this pavement in commercial parking lots and driveways
could result, under specific applications, in an annual overflow reduction
of about 90 percent.
Surface Storage
Surface storage is a very useful tool in relieving runoff surges
that would otherwise surcharge storm and/or combined sewers. Because of
the high value of land in urban areas and because of the large areas required,
implementing this concept is extremely difficult. Using available land
such as parking lots, roof tops and open spaces for flow attenuation as
opposed to construction of specific stormwater retention ponds would appear
to be more feasible. Storing the five-year storm on roof tops would amount
to an increased stress of about 9 psf which is well below the 40 psf snow
design loads in this region. Roof storage is only feasible in commercial and
industrial areas and would result in an approximate total runoff reduction
of 10 percent. Open areas and park land contribute about 20 percent of the
total surface area. Flooding park land and open spaces appears to be more
feasible than roof storage with respect to total runoff capture, but the
disadvantage to flooding is post-storm grit and debris clean-up. Because
of this disadvantage, flooding parkland does not appear to be feasible.
Storage ponds are required to effectively handle runoff surges and thus
reduce overflows. Large land areas, as well as separation of sanitary and
storm sewers, are required for depression storage and therefore its applica-
tion is significantly limited in developed areas. In conclusion, surface
storage in the urban areas characteristic of Rochester would be extremely
difficult to implement and therefore is not considered practical.
Each of the nonstructural CSO abatement alternatives has been
evaluated relative to the Rochester study area. The drainage areas
identified as being most suitable for the application of these source
management measures are presented in Figure 38. Increased street cleaning
would be most effective in those areas that receive the highest daily load-
ings such as downtown urbanized areas. Area no. 25 represents the heavily
trafficked downtown portion of the City of Rochester. Area no. 7 represents
another area that receives heavy traffic and thus is subjected to substantial
surface pollutant accumulations. Eastman Kodak Company has several large
industrial plants within this area that employ many hundreds of people.
Thousands of cars and trucks use the main routes through this area daily.
Area no. 16 represents an area essentially undergoing urban renewal. Much
of the area is presently in poor physical condition and large amounts of
debris litter the entire drainage basin. Area no. 22 represents another
section undergoing urban renewal which is presently in poor condition with
103
-------
frOI
OVERFLOW (mil gal)
o
o
01
o
CD
GO
ro
-s
-h
O
fD
DJ
to
n
<-+
o
O
-b
O>
3
Q.
ro
8
P- .
CD
33
O
CO 5
ID
3) rn en
i<
O P O
Z3 S CR
883
m
§pg
m (D
to
r~
>
c
CO
m
ro r\>
O w
O o
H - 1
O
1
> O
33 3)
m >
> z
m
ro >
O 33
O rn
>
-------
N
FIGURE 38.
Land Use Restrictions
Surface Storage
Increased Street
Cleaning
Increased Sewer
Maintenance
Location of Drainage Areas Suitable for the
Application of Nonstructural Alternatives
105
-------
a large volume of surface contaminants.
Relative to increased sewer maintenance, Areas nos. 16, 17 and 18 have
been previously identified as those areas with sewers requiring cleaning and
repair. Such repairs will improve the conveyance capacities of these sewers
which can reduce the frequency and volume of combined sewer overflow.
Areas nos.6 and 8 represent areas that are essentially entirely all
commercial and industrial. With the large number of parking lots and roof tops
in these areas, implementation of surface retention practices such as pond-
ing on roofs and in parking lots can relieve the hydraulic load imposed on
the downstream sewers, that in turn can also reduce the frequency and volume
of CSO.
Areas nos.21, 16, 28, 29 and 31 are essentially residential in nature.
Implementation of varying land use restrictions such as no additional pave-
ment or buildings can possibly relieve the hydraulic load on the East Side
Trunk Sewer thereby reducing CSO occurrences presently experienced along
this major conveyance system.
4
MINIMAL STRUCTURAL ALTERNATIVES
Abandoned Treatment Facilities
Within the study area there are only two possible abandoned treatment
plant or chlorination facilities for use in the abatement of combined sewer
overflow. One is the Joseph Ward Chlorination Station located near the
intersection of Joseph Avenue and Ward Street. The station has not operated
since its construction in 1965. Because the drainage area tributary to this
station is rather small and it is relatively remote from system overflows,
it is doubtful whether this facility could be effectively used in the treat-
ment of such overflows.
The other possible facility is the Norton-Densmore Chlorination Station
located on the north side of Norton Street at the junction with Densmore Creek.
This combination screening and chlorination station was constructed to treat
the combined sewer overflow from the Norton Street Tunnel. In addition to
the overflow from drainage area no. 28 conveyed by the Norton Street Tunnel,
overflow from drainage area no. 29 also discharges to the Densmore Creek
location. Presently, these two overflows discharge to Irondequoit Bay via an
open channel. Under construction, approximately 1000 ft downstream of this
station is a dropshaft into the Cross Irondequoit Tunnel. Use of this
facility for combined sewer overflow pollution abatement could involve
screening and chlorination of the overflow prior to discharge to the tunnel.
The expected peak flow rate through the station resulting from the 2 Yr-
2 Hr design storm is approximately 1800 cfs. Several problems become apparent
when considering such a large flow. It is advisable to remove a large per-
centage of the organics prior to chlorination to substantially lower the
detention time necessary for coliform kill. Even with high-rate treatment
106
-------
facilities, large basins would be required for solids removal for such large
flowrates. Space limitations at this site preclude the use of any large
basins or tanks.
Use of a swirl concentrator at the station is not practicable because
of size requirements. Based on information obtained from the pilot plant
studies, the primary swirl unit would have to be approximately 250 ft in
diameter to adequately treat the peak flowrate.
The problems associated with both the Joseph Ward and the Norton-
Densmore Chlorination Stations preclude their use as facilities for combined
sewer overflow abatement.
Multi-Purpose Upstream Impoundment for Overflow Control
One method for this alternative is underground overflow volume storage
for subsequent release after the storm event. Depending on invert and
ground elevations, the stored volume may be emptied by gravity or may have to
be pumped.
On an annual basis, as predicted by the Simplified Stormwater Model, the
Maplewood Park drainage area no. 7 contributes only two percent of the total
river overflow. Based on measured time-incremented quality data, it is
estimated that area no. 7 contributes one percent of the total average annual
loading of BODs and T'SS. Although overflow from this area does not
represent a substantial pollutant load to the river, it is used as an
example to illustrate the procedure to be followed in determining the
size of facility necessary for upstream impoundment.
Sizing for underground storage must be based on a defined quantity
of wastewater. To remain consistent with the other alternatives the 2 Yr-2 Hr
design storm is used to evaluate the effectiveness of upstream
impoundment for overflow control. Overflow from drainage area no. 7
for this storm event amounts to 9 mil gal. To capture this volume of
overflow completely, an underground storage chamber having dimensions of
300 x 300 x 15 ft would be necessary. The chamber could be constructed
under the park immediately before the overflow discharge outlet to the
river. The two weir overflows located within the drainage area would
continue to function normally. Overflows would occur but would be
intercepted by the detention chamber before the outlet to the river.
Volumes of overflow generated from storms greater than a two-year event
would be allowed to bypass the detention chamber. The stored volume
would be detained until the termination of the storm event and then
pumped back into the interceptor system at a controlled rate.
The described overflow detention chamber could also provide for
equalization of dry-weather flow. During periods of excessively high
dry-weather flow, the wastewater could be temporarily detained and then
pumped back into the system. This would lessen the load on the sewer system
and also would eliminate the possibility of dry-weather overflow or
overloading of the dry-weather treatment facility.
107
-------
Similar considerations for the other drainage areas indicate the
possibility of underground storage chambers for overflow pollution
abatement. Several problems, however, would be encountered with such
structures. One is the sludge or solids removal problem associated with
all types of detention structures. The other is the poor subsurface
conditions encountered throughout the Rochester area.
Considering solids removal, it is estimated that 5500 Ib of solids
are generated from the analyzed Maplewood detention chamber. It is not
practicable to pump these solids back into the interceptor system because
of the possibility of settling in the sewers during the reduced flowrates
after the storm. Therefore, it would be necessary to remove the bulk of
these sludges after the storm event. Additional study with emphasis on
local soil conditions must be conducted to further evaluate this alternative.
Use of In-System Regulators and Control Structures for Increased In-System
Storage
The use of inflatable dams for increased in-system storage can be
effective in reducing the number and volume of combined sewer overflows.
This is especially true considering the number of large tunnels that are
proposed or under construction within the Rochester Pure Waters District.
The Genesee Valley Interceptor (GVI) currently under construction has a
storage volume of approximately 25 mil gal. This interceptor provides
combined sewage drainage relief for the areas presently served by the
Genesee Valley Canal Sewer and will eliminate all combined sewer overflows
along the west side of the Genesee River south of the Upper Falls. The
first overflow relief point is at Front Street. Assuming that the runoff
coefficient is the same over the remaining portion of the drainage area
tributary to the GVI as for areas nos. 17 and 18, the total volume of
runoff from the 2 Yr-2 Hr design storm in 10.2 mil gal. Unless a
control device is installed to limit flow from the interceptor, the avail-
able storage capacity will never be fully utilized. With the use of an
inflatable dam, this interceptor can store as much as a 5-year storm with
an associated total runoff of 14.5 mil gal. Average dry-weather flow accounts
for an estimated 5.8 mil gal. The stored wastewater can be released into
the Clarissa Street Tunnel at a rate of about 25 cfs without overloading
any downstream facilities. At this rate the interceptor would drain in
approximately 32 hr.
Analysis of the proposed Genesee River Southeast Interceptor indicates
the need for some type of control device at the outlet if the large volume
of available storage is to be utilized. This interceptor is to involve
approximately 3 miles of 14 ft dia tunnel. The available storage capacity
is estimated to be 17.3 mil gal. The interceptor relief overflow point
on this system will again be at Front Street. The SWMM estimates surface
runoff from this drainage area resulting from the 2 Yr-2 Hr design storm
to ,,be 21.5 mil gal. Based on the storage capacity and the runoff volume a
relief discharge rate of approximately 52 cfs is required to prevent
surcharging the interceptor. At this rate the interceptor would drain in
approximately 14 hr following the runoff event.
108
-------
Storage capacity is also available in the Cross Irondequoit Tunnel and
the proposed Culver-Goodman Tunnel complex. The control device necessary
for storage and relief discharge will be the pump station located at the
outlet of the Cross Irondequoit Tunnel before discharge to the Van Lare STP.
The number of pumps, pumping capacity, and pumping duration will determine
the extent of storage that is actually used. The Cross-Irondequoit Tunnel
and the proposed Culver-Goodman Tunnel have a total storage capacity of
approximately 85 mil gal. This is sufficient to completely capture all
combined sewer overflows from the East Side Trunk Sewer for the 5 year
design storm (66 mil gal). These overflows originate from drainage areas
nos. 28, 29, and 31.
Regarding the present sewer network, there is insufficient storage
capacity available in the conveyance sewer system of any drainage area to
make the use of inflatable dams or other in-system control structures
feasible for overflow control. Figure 39 identifies system locations where
inflatable dams and/or regulators can be utilized to enhance upstream
storage. Also identified are areas offering potential for minimal surface
and subsurface storage.
Upgarding the Existing St. Paul Interceptor
The present St. Paul Interceptor system contains three major flow
constrictions, one being the lower reach of the interceptor near the
Van Lare STP, another downstream of inlets from drainage areas nos. 8, 9,
and 22, and the last being the River Siphon itself. The location of the
flow constrictions are presented in Figure 40. The first constraint near the
STP consists of three interceptor sections ranging in unsurcharged capacity
from 168 to 260 cfs. Closest to the Van Lare STP is approximately 240 ft
of 5 ft-6 in. pipe having a capacity of nearly 260 cfs, and finally a
section of the same diameter pipe about 1800 ft in length and having a
capacity of 160 cfs. The second constraint is a 5 ft-4 in. x 5 ft-10 in. brick
conduit directly downstream of the interceptor inlets from drainage areas
nos. 8, 9 and 22. The third restriction, the River Siphon, consists
of two cast iron pipes, 24 and 42 in. in dia, each approximately 3400 ft in
length. The combined carrying capacity for the Siphon is slightly more than
70 cfs. Removing these constraints involves increasing the capacities to
the unsurcharged capacity of neighboring conduits. The first constrained
reach is increased to 285 cfs, the second to 250 cfs and the third (Siphon)
to 140 cfs.
Table 30 which presents a tabulation of River overflows on an annual
basis using the SSM would seemingly indicate that removing the constraints
in the St. Paul Interceptor would play a major role in reducing the volume
of combined sewer overflows as well as the frequency of overflow
occurrences.
109
-------
tostff
?19J* 39
-------
N
3200'
5'-6"
Pipe \
Cross \
Irondequoit J^
St. Paul Interceptor JConstraint'5
FIGURE 40.
Location of Constraints in the Existing
St. Paul Boulevard Interceptor
111
-------
TABLE 30. RIVER OVERFLOW PROJECTED BY SIMPLIFIED STORMWATER MODEL FOR THE
PROPOSED TUNNEL INTERCEPTOR SYSTEM
Drainage
Area No.
Average
Annual
Overflow
(mil gal)
Average
Annual
Overflow
(days)
% of Total
River Overflow
St. Paul Interceptor
Constraints Unaltered
7
8 & 9
16
21
22
25
Front St.
Total
St. Paul Interceptor
Constraints Removed
7
8 & 9
16
21
22
25
Front St.
Total
22.4
855.9
278.0
32.6
173.3
178.2
15.9
1556-3
22.4
392.1
0.0
0.0
0.0
9.5
0.2
424.2
7.55
51.73
62.05
10.65
56.25
65.55
7.55
28.60
0.0
0.0
0.0
3.90
1.4
55.0
17.9
2.0
11.1
11.5
5.3
92.4
0.0
0.0
0.0
2.2
Removal of, constraints in the St. Paul Interceptor are projected to
reduce the annual total River overflow by 73 percent. Essentially a
100 percent overflow reduction occurs from drainage areas nos. 16, 21,
25, and Front Street with a 54 percent reduction in overflow from
drainage areas nos. 8 and 9. Drainage area no. 7 is unaffected by these
improvements. A similar analysis using the SWMM under the application of the
2 Yr-2 Hr design storm indicates an approximate 5 percent reduction in
total overflow volume when the St. Paul Interceptor constraints are removed.
This is illustrated in Figure 41 and Table 31.
The SWMM also indicates that removing constraints in the St. Paul
Interceptor sufficiently lowers the hydraulic grade line at the regulators
to effectively reduce River overflow. This action would reduce surcharging
of the relatively low weirs, especially in drainage areas nos. 16, 21, 22 and
25 where the flows are relatively small and where the annual overflow is
essentially eliminated. However, the analysis utilizing the 2 Yr storm does
not project the same level of improvement. This discrepancy between
yearly analysis and design storm analysis is understandable when one
considers that 97 percent of all storms recorded in the twenty-five
112
-------
3
o
300-r
250--
2 00--
I50+
IOO--
50--
0
TOTAL RIVER OVERFLOW 54.0
SYSTEM
onduit No. i, ST RAUL
2 yr - 2 hr STORM
PROPOSED TUNNEL INTERCEPTOR
Overflow in mil gal
No overflow blockage, no regulator changes, no change to St. Paul Interceptor
No overflow blockage, no regulator changes, St. Paul constraints removed
No overflow blockage, regulators increased 150?, St. Paul constraints removed
7, 16, S 22 overflows blocked, regulators increased to carry peak flow, St. Paul
constraints removed
7,16 & 22 overflows blocked, no regulator changes, St. Paul Constraints removed
7, 16, 22, 8 & 9, and 25 blocked, regulators increased to carry peak flow, St.
Paul constraints removed.
1
2
1
3
1
4
Time
1
5
(hr
of
i
6
day)
i
7
1
8
1
9
1
10
1
II
FIGURE 41. Minimal Structural Alternative Modeling Results (SWMM)
Overflow Hydrograph
-------
years of record had a total rainfall less than the 2 Yr-2 Hr design
storm and 87 percent of all storms had a longer duration. On the
annual basis, nearly every storm produces less total runoff than the 2 Yr
storm and distributes that runoff to the interceptor system over a
greater time periodsthus reducing peak flows and subsequent overflows.
The net effect of reducing the St. Paul Interceptor hydraulic gradient
is therefore much more significant on an annual basis than what is
indicated by the analysis of the 2 Yr-2 Hr design storm.
TABLE 31. EFFECT OF UPGRADING THE ST. PAUL INTERCEPTOR
2YR-2HR DESIGN STORM
Present Present
System System
Constraints Constraints
Not Removed Removed
Proposed Tunnel Proposed Tunnel
Interceptor Interceptor
System System
Constraints Not Constraints
Removed Removed
River Overflow
(mil gal)
East Side Over-
flow (mil gal)
Total Overflow
(mil gal)
Percent Reduction
in Total Overflow
With Constraints
Removed
Percent Reduction
in River Overflow
with Constraints
Removed
94.9
105.1
200.0
93.6
97.6
191.2
4.4
101.6
41.4
143.0
96.8
41.2
138.0
1.2
4.7
The overflow results obtained using the SSM were combined with the
annual geometric mean overflow pollutant concentrations to give the annual
river pollutant mass loadings. BODs and TSS results are show in Table 32.
It is interesting to note the magnitude of both BOD5 and TSS loadings. The
total loading projected from this analysis shows more than 2.5 mil Ib of
suspended solids and over 1 mil Ib of BODs entering the Genesee River
annually. Comparison of both flow and quality data indicates that the
2 Yr-2 Hr design storm contributes about 6.5 percent of the total annual
loadings.
114
-------
TABLE 32. RIVER LOADING- PROPOSED TUNNEL INTERCEPTOR SYSTEM-PROJECTED BY
SIMPLIFIED STORMWATER MODEL
Annual BODs
Drainage River Loading
Area No. (1b)
Annual TSS
River Loading
St. Paul Interceptor
Constraints Unaltered
St. Paul Interceptor
Constraints Removed
7
8 & 9
16
21
22
25
Front St.
7
8 & 9
16
21
22
25
Front St.
10,600
550,000
151,000
13,300
332,000
59,000
4.000
1,119,900
10,600
252,000
0
0
0
3,340
40_
265,980
24,000
1,220,000
633,000
20,000
550,000
169,000
12,000
2,628,000
24,000
559,000
0
0
0
9,500
130
592,630
In summary, the annual analysis conducted with the SSM does not consider
surcharge and therefore tends to underestimate the reduction in overflow
volumes. The annual analysis uses daily rainfall data as opposed to hourly
data. Previous work has shown projections made using daily rainfall data to
underestimate the overflow volumes projected using hourly data by about 40
percent. These cancel one another and therefore the SSM reasonably predicts
hydraulic flow conditions along the St. Paul Interceptor.
Selective Blockage of High Impacting Combined Sewer Overflows
Completely removing an overflow from the drainage system can be
accomplished basically by two methods. One method completely captures the
overflow using another collection sewer. The method hydraulically
eliminates the overflow by properly sizing the existing interceptor,
regulators, and weirs to effectively attenuate or remove peak flows. This
method will be discussed here for selected overflows.
Eliminating overflows from drainage areas nos. 7, 16, and 22
would theoretically reduce total River overflow by approximately 25 percent
115
-------
which in turn reduces BODs and TSS by about 35 percent for the 2 Yr-2 Hr
storm. Likewise, eliminating overflows from drainage areas nos. 8, 9 and
25 in addition to those from areas nos. 7, 16 and 22 would theoretically
decrease the River overflow by nearly 70 percent with an attendant 83
percent BOD5 and 77 percent TSS reduction. However, as illustrated in
Table 33 and Figure 41 eliminating overflows from areas nos. 7, 16 and 22 and
increasing regulator capacities to handle peak flows reduces the total
River overflow by only 15 percent. Similarly, eliminating overflows
from areas nos. 7, 16, 22, 8, 9 and 25 results in approximately a 50
percent reduction in total overflow. Table 32 contains information
helpful in explaining these discrepancies. When drainage area overflows
are blocked, the downstream areas are surcharged even when the St. Paul
Interceptor constraints are removed. The drainage areas downstream on
the interceptor system therefore experience a greater overflow than when
not surcharged. For example, when drainage areas nos. 7, 16 and 22 are
blocked, the overflow volume increases in areas nos. 8, 9, 21, 25, and
Front Street while decreasing at the Siphon. As a result the overflow
reduction is only 14.9 mil gal instead of a theoretical 23.4 mil gal.
Blocking selected overflows appears to create the greatest surcharge
problem for area no. 21. From Table 33 it is obvious that drainage area no.
21 is where nearly all system relief occurs when other drainage area over-
flows are blocked. This table helps to explain the attenuated flow in the
St. Paul Interceptor.
As seen in Figure 40 the proposed tunnel interceptor system would
generate a peak flow through the St. Paul Interceptor of 260 cfs for the
2 Yr-2 Hr storm. The duration of the peak lasts for approximately 40
min. Increasing river regulator capacities threefold greatly increases
the peak flows from the individual drainage areas, yet the St. Paul
Interceptor peak flow is not significantly increased. In fact, the total
overflow reduction is only about 7 percent when the average peak flows
from individual areas are more than doubled. Increasing all regulator
capacities by 150 percent increases overflows at areas nos. 21 and 25,
even when the St. Paul Interceptor constraints are removed. This backflow
in areas nos. 21 and 25 attenuates the effect of the increased 7 mil gal
of overflow to the St. Paul Interceptor. Upon increasing the regulator
capacity by 150 percent and removal of the interceptor constraints, the peak
flow is only 265 cfs distributed over a period of 6 hr. Moreover, increasing
both the regulator and interceptor capacity results in reduced overflows
and essentially no change to peak flows to the Van Lare STP.
Blocking overflows from drainage areas nos. 7, 16, 22 and at
the same time increasing the respective regulator capacities to the
expected peak flow at the regulator (from the 2 Yr-2 Hr storm), results
in a reduction in the total River overflow of about 15 percent. The
additional 15 mil gal which reaches the Van Lare STP is spread out over a
period of slightly more than 7 hr at a peak rate of 270 cfs. By increasing
the blocked overflow regulator capacities to the predicted peak flow, no
flooding occurs at the regulators. Blocking overflows and increasing
116
-------
TABLE 33. SELECTIVE BLOCKAGE OF OVERFLOWS PROPOSED TUNNEL INTERCEPTOR SYSTEM
2YR-2HR DESIGN STORM
Areas No. 7, 16, 22, 8 & 9, and
25 blocked, regulator capacity
unaltered, regulator capacity
8 7, 16, 22, 8 & 9, and 25 in-
creased to peak, weirs unaltered,
St. Paul Interceptor Increased
Areas No. 7, 16, & 22 blocked.reg- [Overflows not blocked
ulators unaltered, weirs unalter- [regulators unaltered, weirs
ed, St. Paul Interceptor Unalterediunaltered, St. Paul Interceptor
iImproved
Capacity
Peak Flow
Overflow/ @
Volume Regulator
(mil gal)
_
_
_
28.6
_
10.5
14.9
10.0
20.6
10.6
(cfs)
330
328
292
765
72
328
_
-
-
-
-
Flooding
@
Regulator
N
N
N
Y
N
N
M
N
N
N
N
Overf 1 ow/
Volume
(mil gal)
_
.
42.4
16.4
9.3
7.2
11.0
9.9
20.9
10.5
Peak Flow
&
Regulator
(cfs)
23
79
48
91
159
21
_
-
.
-
-
Flooding
@
Regulator
Y
Y
Y
Y
N
N
N
N
Y
Y
I
i
Overflow/
Volume
(mil qal)
9.1
8.4
5.8
39.3
11.9
8.6
5.2
8.5
9.8
20.9
10.5
Peak Flow
?
Regulator
(cfs)
20
76
38
153
86
21
.
-
-
-
-
Flooding
@
Regulator
Y
M
N
N
N
N
N
N
N
N
N
Drainage
Area
No.
7
16
22
8,9
21
25
Front
Siphon
28
29
31
Note: Y = Yes, N = No, M = Maybe
-------
regulator capacities to peak flows therefore appears to be more effective
than just increasing regulator and interceptor capacities.
Figure 42 illustrates flow patterns through a typical overflow regulator
for drainage area No. 22, for the 2 Yr-2 Hr storm. It is seen that increas-
ing the regulator capacity 150 percent and removing the St. Paul Interceptor
constraints results in a reduction in overflow of 2.3 mil gal (40 percent
reduction) from this area. The remaining 60 percent is removed by blocking
the overflow and increasing the regulator capacity to the expected maximum
flow. The peak flow has a very short duration of only 20 min. Increasing
regulator capacities and eliminating interceptor constraints decreases
overflows and results in no flooding at the regulator. However, merely
plugging the overflow without any improvements to the regulators or the
St. Paul Interceptor results in heavy flooding in the lower portion of the
drainage area. This situation is evident in Table 33. As seen, the
reduction in total River overflow is the same when blocking drainage area
nos. 7, 16 and 22 overflows regardless of whether the regulators or
St. Paul Interceptor capacities are increased; however, when the capacities
are not increased, flooding occurs. The results would therefore indicate
that increasing both the capacity of the regulators and the St. Paul
Interceptor are required to effectively reduce potential flooding.
Selective Overflow Heir Evaluation
Raising the elevation of overflow weirs, although not always as effective,
is an alternative to completely blocking the overflows. In fact, increasing
the weir depth to the depth of the interceptor is analogous to blocking the
overflow. For those storms that do not create full depth flows in the trunk
sewers tributary to a regulator, weir elevation increases are not as
effective as blockage for surcharge conditions. This latter condition is
illustrated by Figure 42 and Table 34 which is similar to those used in the
discussions of selective overflow blockage for the 2 Yr-2 Hr storm.
Figure 43 shows that raising weir elevations to interceptor crowns as
well as increasing regulator capacities to peak flows for all River over-
flows except Front St. and the River Siphon is as effective as blocking
overflows from areas nos. 7, 16 and 22. The total River overflow
for increasing the weirs is 83.4 mil gal as opposed to 86.3 for blocking
overflows. Raising all River overflow weirs, Siphon excluded, and increas-
ing their regulator capacities 150 percent results in an equivalent overflow
(83.6 mil gal); however, this condition causes flooding at certain regulators
(See Table 32). Raising weir elevations only at drainage areas nos. 7, 16,
22 and 25 (increasing regulator capacities to peak flows) results
in a total River overflow of 95.9 mil gal or about one third the overflow
reduction experienced when blocking the overflows in areas nos. 7, 16,
and 33. Thus, this is not as effective as increasing all river
regulator capacities 150 percent which results in a total River overflow
of 94.6 mil gal. The feasibility of raising weir elevations appears to be
limited in light of the above.
118
-------
FLOW vs TIME
300 r
250
200
150
100
50
< Proposed Tunnel Interceptor System
Regulator capacity increased 150%,
St. Paul Interceptor constraints removed,
Overflow blocked, regulators increased
to peak flow capacity of St. Paul Blvd.
Interceptor
Overflow blocked, regulator unaltered,
St. Paul Blvd Interceptor unimproved
- 3.5 mil gal
Conduit 34 - Drainage Area 22 Regulator
2yr-2hr Storm
Proposed Tunnel Interceptor System
2.3 mil gal
^SSSSi^:S:::::::::$S:::::::::xr """""" .^ Pkwlinn ("Vinrllti/tn
i wjuuiuy v/unuiiion
I 234567
Time (hr of day)
FIGURE 42. Minimal Structural Alternative Modeling Results
Overflow Hydrograph Area No. 22 Regulator
8
119
-------
ro
CD
TABLE 34. SELECTIVE WEIR ANALYSIS PROPOSED TUNNEL INTERCEPTOR SYSTEM
2 YR-2 HR STORM
Regulator capacities increased
150» weirs all increased to
depth of interceptor St. Paul
Interceptor constraints un-
altered
Regulator capacities increased
150% weirs all increased to depth
of interceptor St. Paul Inter-
ceptor constraints removed
Weirs 9 drainage areas 7,
16, 22 and 25 raised reg-
ulators @ these drainage
areas increased to peak
flow St. Paul Interceptor
Constraints removed
Weirs & Drainage areas 7,16,22,
25,3&9 and 21 increased to depth
of interceptor regulators & these
drainage areas increased to peak
flow St. Paul Interceptor con-
straints removed
Drainage
Area No.
7
16
22
8&9
21
25
Front St.
Siphon
Overflow/
Overflow Volume
Conduit No. (mil qal )
6
25
35
31
214
40
250
251
6.4
4.2
3.7
34.7
4.1
10.7
7.8
15.0
Peak Flow
0
Regulator
(cfs)
65
223
110
208
244
61
-
-
Floodina
@
Regulator
Y
N
N
H
N
N
N
N
Overflow/
Volume
(mil qal)
6.4
4.1
3.5
33.7
9.2
9.8
6.3
10.7
Peak Flow
@
Regulator
(cfs)
65
224
111
232
235
61
-
-
Flooding Overflow/
@ Volume
Regulator (mil gal)
Y
N
N
H
N
N
N
N
1.3
1.4
0.3
40.5
16.8
7.9
20.9
6.7
Peak Flow
(? Flooding Overflow/
Regulator @ Volume
(cfs) Regulator (mil gal)
319
391
308
125
83
323
-
N
N
N
M
N
N
N
N
1.3
1.4
0.6
30.8
11.8
8.2
21.7
7.6
Peak Flow
@
Regulator
(cfs)
323
391
283
935
217
323
-
Flooding
0
Regulator
N
N
H
N
N
N
N
N
Note: Y = Yes, N = No, M = Maybe
-------
300
250
200
150
JO
Lu
100
50
Total River Overflow 83.6 mil gal
Total River Overflow
\
Total River
Overflow
95.9 mil gal
Total River i
Overflow
83.4 mil gal
\
\
i
raised regs increased SPI
Conduit I, Van Lore STP
2yr-2hr Design Storm
Proposed Tunnel Interceptor System
Weirs @ area nos. 7,16,22,25,
constraints removed
All river weirs raised, regs increased
All river weirs raised, regs increased
constraints removed
Weirs @ areas 7,16,22,8,9,21,25 raised, regs increased,
SPI constraints removed
I I I I I I
150%,
150%,
SPI
SPI
unaltered
I
34567
Time (hr of day)
8
10
Note: Regs = Regulators, SPI = St. Paul Blvd. Interceptor
FIGURE 43. Minimal Structural Alternative Modeling Results
Flow Projections at the Van tare STP
121
-------
Raising overflow weir elevations requires increasing regulator capacities
in order to significantly reduce the total system overflow to the Genesee
River. Under similar regulator and interceptor conditions, overflow blocking
results in a better reduction of the 2 Yr storm River overflow than raising
weir elevations, and therefore, would decrease River overflows at least as
well as raising weirs for any storm less than the 2 Yr storm. In summary,
raising weir elevation, as opposed to overflow blockage, does not appear to
be a feasible alternative to reducing combined sewer overflows.
Regulator Capacity Evaluation
Increasing regulator capacity has been discussed under minimal structural
alternatives. It has been shown that increasing regulator capacities is
more effective than raising weirs and that increasing regulator capacities
helps alleviate flooding when overflows are blocked. In this subsection
the effect of just increasing regulator capacities will be discussed.
Figure 44 shows the relationship between total River overflow and
percent increase in regulator capacity using the proposed tunnel interceptor
system with the St. Paul Interceptor constraints removed. Table 35
shows the overflow from individual River regulators which is associated
with the annual analysis.
TABLE 35. TOTAL RIVER OVERFLOW VS. REGULATOR CAPACITY BASED ON
SIMPLIFIED STORMWATER MODEL
REGULATOR CAPACITY INCREASE
Drainage Area No. 0% 50% 100% 150% 200%
7
8 & 9
16
21
22
25
Front St.
Totals (mil gal)
22.4
392.1
0.0
0.0
9.5
3.9
0.2
428.1
5.0
116.8
0.0
0.0
0.0
1.3
1.3
124.4
0.8
37.4
1.0
0.0
0.0
6.0
2.9
48.1
0.0
3.0
14.8
0.0
4.9
14.1
3.1
39.9
0.0
2.7
14.8
0.0
5.2
14.1
3.1
39.9
Previous work has shown that increasing regulator capacities on the
present interceptor system results in an asymptotic relationship whereby
increasing regulator capacities by more than threefold results in no
appreciable change in overflow reduction. Again, this phenomenon is apparent
on the proposed tunnel interceptor system. The analysis using the 2 Yr
storm shows only minimal overflow response to regulator increases, but the
yearly analysis dramatically expresses the response in overflow to the River
to changes in regulator capacities. Beyond a 150 percent regulator capacity
increase, no appreciable overflow reduction occurs. Figure 44 shows that
annual total River overflow can be reduced by about 95 percent by increasing
regulator capacities 150 percent and removing the St. Paul Interceptor
constraints.
122
-------
500
Annual Analysis
(SSM)
2yr-2hr Storm
(SWMM)
Proposed Tunnel Interceptor System
St. Paul Constraints Removed
80 120 160 200
Per Cent Increase of Regulator Capacity
240
FIGURE 44. Minimal Structural Alternative Modeling Results
Total River Overflow vs Regulator Capacity
-------
Table 35 presents the response of individual drainage areas to regulator
capacity increases and reveals an interesting point. As capacities increase
the total River overflow decreases, but the River overflows upstream of
drainage areas nos. 8 and 9 (specifically nos. 16, 22, 25 and Front St.)
increase. This is due to the influence of large flows leaving the regulator
from drainage areas nos. 8 and 9 on the upstream hydraulics.
STRUCTURALLY INTENSIVE ALTERNATIVES
Alternate 1
This alternate considers capturing the first-flush from all the River
overflow sites and treating all post first-flush flows with primary swirl
devices. The temporarily stored first-flush volume would be released back
into the existing interceptor system on termination of system overflows.
Treatment for this first-flush volume would be at wet-weather treatment
facilities located at the Van Lare STP. Also entering this plant for
wet-weather treatment would be flows from the St. Paul Interceptor and
the Cross-Irondequoit Tunnel. This alternate is shown schematically in
Figure 45.
The cost/benefit analysis for this alternate is presented in the follow-
ing section. First-flush capture and post first-flush swirl performance
quantities are presented in Section 7.
Alternate 2
This alternate involves evaluating various storage and treatment
options for capturing the total overflow to the Genesee River. The treatment
plant under consideration would be located on the Genesee River in the
vicinity of the lower falls. The influent to the plant would be the St. Paul
Boulevard overflows conveyed by individual conduits from each drainage
area regulator. By controlling the rate of treatment, storage within these
conduits would be affected. This alternate is shown schematically in Figure
46.
Two different approaches to analyzing this alternate were made utilizing
the SWMM. The 2 Yr-2 Hr design storm provided the required rainfall
input to the system. Computer simulation with the SWMM involved 3600 inte-
gration steps of 10 sec each. It is believed that such a short time
interval provides an accurate representation of the flowrates within the
sewer system. Table 36 shows the runoff and overflow volumes associated
with each drainage area under the application of the design storm.
The first approach involved the balance of storage and treatment assuming
the St. Paul Interceptor has continuous flow to the Van Lare STP throughout
the rainfall-runoff event. The initial analysis also did not incorporate
improvements to the interceptor system. The low regulator capacity from each
drainage area to the St. Paul Interceptor was increased by 150 percent.
Under this initial analysis, the overflows resulting from the east side of the
124
-------
N
STORAGE (F-F)
PRIMARY SWIRL CONC. ( POST F-F)
FIGURE 45. Schematic of Structurally Intensive
Alternate No. 1 - Storage First-Flush/
Treatment Post First-Flush
125
-------
N
OVERFLOW INTERCEPTOR
0 OVERFLOW TREATMENT PLANT
FIGURE 46. Schematic of Structurally Intensive
Alternate No. 2 - Overflow Treatment
Along Genesee River
126
-------
TABLE 36. RUNOFF AND OVERFLOW VOLUMES BY DRAINAGE AREA
AS PROJECTED BY THE SWMM FOR THE
2 YR - 2 HR DESIGN STORM
Drainage Area
No.
Runoff
(mil gal)
Overflow
(mil gal)
7
8 & 9
16
21
22
25
Front St. (from nos. 17,
18, 26 and part of
no. 31)
Siphon
11.25
54.22
14.11
8.42
8.99
8.78
28.77
Totals 134.54
9.00
41.41
8.51
8.43
5.80
8.90
6.46
12.95
101.55
Total Rainfall 2 yr - 2 hr storm = 1.15 in.
Total Rainfall over the drainage areas tributary to River = 286.40 mil gal
Gross Runoff Coefficient = 0.47
city that discharge to the Cross-Irondequoit Tunnel were stored and treated
at a maximum rate of 45 mgd at the Van Lare STP. The last condition which
separates this approach from the second, is that the river overflow inter-
ceptor conveying all overflow volume to the treatment facility is allowed
to fill completely before pumping and treatment begin. Figure 47 shows the
various combinations of storage capacity and treatment rates necessary to
limit the various overflow volumes. It should be noted that the overflow
volumes represent wastewater discharges to the various receiving waters
contributed by system relief and treatment plant by-pass induced by insuf-
ficient storage or treatment capacities.
Figure 47 shows that as the treatment rate increases, the required
storage capacity decreases for a given overflow volume. The necessary
storage requirements for a given treatment rate can be translated into a
specific diameter tunnel by using Figure 48. This figure shows the necessary
diameter tunnel for any given storage capacity based on a tunnel length of
18,800 ft. This represents the approximate length of the individual conduits
leading to a River treatment facility from the various drainage area regulator
overflow points.
From a quality consideration the total pollutant load reaching the treat-
ment facility from the river overflow conduits for the design storm is 70,930
Ib and 173,450 Ib of BODs and TSS, respectively. Total overflow volume is
101.55 mil gal.
The second approach taken to analyze this alternative was to balance
storage and treatment under the same conditions as the first approach with
the exception that flows from the individual river overflow conduits would
127
-------
no
CO
120
100
90
80
^5 70
o>
1 60
|, 50
o? 40
1 30
r^°
20
10
0
Assumptions: (1)
(2)
(3)
(4)
(5)
St. Paul Interceptor
has continuous flow with
existing constraints
Storage fills before
treating
Overflow regulator
capacities increased by
150%
East Side overflows and
additional 25 MGD treated
at minimum rate of 45 MGD
2 yr - 2 hr design storm
( ) - Overflow in mil gal
I
I
I
I
100 200 300 400 500 600 700 800 900 1000 1100 120013001400 1500
Treatment Rate (mgd)
FIGURE 47. Storage vs Treatment Requirements for Varying Overflow
Volumes for Centrally Located River STP
-------
r>o
to
130
120
110
S 90
5 80
I 70
1 60
8. 50
S
o
30
20
10
Assumptions
(I) Total Length of 18800 feet for all
Individual Overflow Interceptors
Leading to Central STP.
i i i i i i i i i
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
Diameter (ft)
FIGURE 48. Storage Volume vs Tunnel Diameter for Structurally Intensive Alternate No. 2
-------
be pumped and treated as they reach the treatment facility. That is, treat-
ment would begin on initiation of the overflow event. The second approach to
the analysis of Alternate 2 is based on the mass diagram of the inflow to
the river treatment facility. This method is similar to the Rippl procedure
(19) for the determination of storage required in impounding reservoirs.
Figure 49 presents the cumulative mass inflow curve for the river treatment
facility. Superimposed on the mass curve are straight lines representing
various treatment rates. The maximum vertical distance between the mass
curve and a particular treatment rate line is the maximum amount of storage
capacity necessary to prevent an overflow or a plant bypass. For varying
overflow quantities, Figure 50 shows the relationship between storage and
treatment for several river overflow conduits conveying overflows to a wet-
weather facility located on the river.
Given the amount of storage capacity required for a specific treatment
rate, Figure 48 shows the necessary tunnel diameter for this storage volume.
As previously indicated this figure is based on an assumed total length of
river overflow conduits of 18,800 ft.
Comparing Figures 47 and 50 it is seen that less storage capacity is
required for a given treatment rate for the second approach than for the
first. Since treatment begins with the collection of the first overflow
volumes instead of proceeding after the filling of the interceptor, it is
obvious that the second approach is a more cost-effective method for this
alternative.
The cost/benefit analysis for this alternate is shown in the following
section. Storage and treatment relationships for the analysis are based
on curves shown on Figure 50.
Alternate 3
An analysis similar to that presented for Alternate 2 was made on the
proposed tunnel sewer system incorporating Alternate 3. The difference
between Alternate 2 and 3 is the location of the treatment facility. Under
Alternate 3, the treatment facilities are located at the present Van Lare
STP. This alternate is shown schematically in Figure 51. As with Alternate
2, two approaches were taken for analysis. The first approach involved
the River overflow interceptor leading to the treatment plant filling
completely before pumping and treatment. Figure 52 shows the relationship
between storage capacity and treatment rate for various overflow volumes. The
second approach involved treating the overflows as they occurred. Figure 50
of Alternate 2 shows the relationship between storage and treatment for this
condition. These two figures indicate that the second approach is a more
cost-effective method. For a given treatment rate less storage capacity is
required when treating from the start of the overflow event than when
allowing the overflow interceptor to fill before treating. In both cases,
Figure 53 shows the size of storage and conveyance tunnel necessary for a
given amount of storage. This figure is based on an interceptor length of
36,000 ft which represents the approximate length of the river overflow
interceptor leading to the Van Lare STP along a route adjacent to the St.
Paul Interceptor originating from the Front Street overflow site.
130
-------
130
120
110
90
I
3 80
I 70
o:
0 60
o
« 50
~ 40
-------
CO
no
100
90
80
9 70
0>
1 60
I 50
w 40
30
20
10
0
Storage vs Treatment
2 yr - 2 hr Design Storm
Treatment at Van Lore
Treatment at River Plant
Bypass in mil gal
100
200
600
300 400 500
Treatment Rate (mgd)
FIGURE 50. Storage vs Treatment Rate for Varying Overflow Quantities
700 800
-------
N
Figure 51.
RIVER OVERFLOW INTERCEPTOR
OVERFLOW TREATMENT PLANT
Schematic of Structurally Intensive
Alternate No. 3 - Overflow Treatment
at Van Lare STP.
133
-------
oo
-pi
130
120
§» 110
JlOO
90
80
0)
T3
"
fc 60
5 50
40
30
20
10
0
Assumptions: (1)
(2)
(3)
(4)
St. Paul Interceptor has
continuous flow with existing
constraints
Storage fills before treating
Regulator overflow capacities
increased by 150%
East Side overflows and
additional 25 mgd treated @
minimum rate of 45 mgd
2 yr - 2 yr design storm
( ) Overflow in mil gal
.350 400
50 100 150 200 250 300
Treotment Rate (mgd)
FIGURE 52. Storage vs Treatment for Varying Overflow Volumes for
Treatment Plant Located at F.E. Van Lare
-------
u>
01
130
120
110
o>
= 90
1 80
| 70
o> 60
O>
I 50
-------
The cost-effective analysis for this alternate is presented in Section 9.
Storage and treatment relationships for the cost-effective analysis are
based on Figure 50.
Alternate 4
This alternate is similar to Alternate 1 with the exception that the
post first-flush is not treated but is directly discharged to the River.
This alternate is shown schematically on Figure 54.
The cost-benefit analysis is presented in the following section. First
flush capture and post first-flush discharge quantities are based on
calculations developed in Section 7.
Alternate 5
This alternate considers the use of primary swirl concentrators on each
of the River overflow locations for treatment of the entire overflow volume.
The swirl effluents are discharged directly to the River. The Van Lare STP
would treat flows entering from the St. Paul Interceptor and the Cross-
Irondequoit Tunnel. This alternate is shown schematically in Figure 55.
Based on calculations presented in Section 7, providing a swirl unit
at each of the overflow sites reduces total river suspended solids loading
by 33 percent - from 173,450 to 116,220 Ib under influence of the design
storm. Table 37 shows performance of the swirl units in reducing the
pollutant load to the River.
TABLE 37. PRIMARY SWIRL SOLIDS REDUCTION TREATING ENTIRE OVERFLOW
2 YR - 2 HR DESIGN STORM
Drainage Overflow
No.
7
3,9
16
21
22
25
Front
Siphon
Primary Swirl
Influent Load
Ib
14540
58900
23750
5220
24000
12770
4790
29480
173450
Load to
River
Ib
8590
39660
10410
5170
21590
11010
4790
15000
116220
Reduction
%
41
33
56
1
10
14
0
49
33
The size of each swirl concentrator was based on the analysis presented in
Alternate 1. These sizes are shown in Table 25 of Section 7. Since a swirl
unit at Front Street yielded no reduction in suspended solids, further
analysis of this alternate involved primary swirl units of the specified
diameter at all overflow locations except Front Street.
136
-------
N
STORAGE (F-F)
FIGURE 54. Schematic of Structurally Intensive
Alternate No. 4 - Storage of
First-Flush
137
-------
N
0 PRIMARY SWIRL CONCENTRATOR
( TOTAL FLOW)
FIGURE 55. Schematic of Structurally Intensive
Alternate No. 5 - Local Overflow
Treatment
133
-------
The cost/benefit analysis of Alternate 5 is presented in Section 9.
Alternate 6
This alternate involves the evaluation of a storage and treatment
balance for intercepting the river overflows and conveying them to the
Cross-Irondequoit Tunnel for treatment at the Van Lare STP. This alternate
attempts to take advantage of the storage capacity available on the east
side of the system to attenuate west side flows by providing a connecting
interceptor tunnel. The feasibility of this alternative will be shown using
the SSM for determining the relationship between storage treatment, and
overflow. This alternative is shown schematically in Figure 56.
The model analysis was made on a drainage area of 12,073 acres with a
gross runoff coefficient of 0.45. Available storage amounting to 84.8 mil
gal representing the Cross-Irondequoit and the Culver-Goodman Tunnels was
evaluated using varying treatment rates to predict overflow frequency and
volumes. Table 38 shows the storage-treatment-overflow relationship for this
alternative on an annual basis.
TABLE 38. WEST-EAST STORAGE-TREATMENT BALANCE SIMPLIFIED STORMHATER MODEL
Total Catchment Area 12073 Acres
Gross Runoff Coefficient 0.45
Storage 84.8 mil gal
Average Annual Overflow
Treatment Rate
(mgd) Days Volume (mil gal)
10
50
100
150
200
35.30
6.00
1.90
0.70
0.30
1494.90
315.76
104.34
35.90
8.90
Note: Storage of 84.8 mil gal represents available storage in the Cross-
Irondequoit and Culver-Goodman Tunnel Complex.
The cost/benefit analysis is presented in Section 9.
139
-------
N
EAST-WEST INTERCEPTOR
* OVERFLOW TREATMENT FACILITIES
FIGURE 56. Schematic of Structurally Intensive
Alternate No. 6 - Conveyance of Overflow
by East-West Interceptor
140
-------
SECTION 9
COST-EFFECTIVE ANALYSIS OF
STRUCTURALLY INTENSIVE ALTERNATIVES
ALTERNATIVES EVALUATED
This section presents the relative ranking of a number of alternatives
in terms of capital cost versus performance. Cost/performance is represented
based on costs developed for all storage, transmission, and treatment
facilities and performance is represented as the total impact of all facili-
ties on the receiving waters for the design 2 Yr-2 Mr storm.
Two categories of treatment trains were considered: (a) primary
treatment and (b) two-level treatment employing filtration. Treatment trains
evaluated are as shown on Table 39. All treatment trains included disin-
fection.
TABLE 39. PROCESS TRAIN CONFIGURATIONS
A. Primary Treatment
1. *G/S + DIS
2. #P/S + DIS
3. **F/S @ 800 gpd/ft2 + DIS
4. F/S @ 1500 gpd/ft2 + DIS
5. F/S @ 2000 gpd/ft2 + DIS
B. Two-Level Treatment
i. G/S +<£DMF + DIS
2. P/S + DMF + DIS
3. F/S @ 800 gpd/ft2 + DMF + DIS
4. F/S @ 1500 gpd/ft2 + DMF + DIS
5. F/S @ 2000 gpd/ft2 + DMF + DIS
* swirl degritter
-------
AH. 1 West side first-flush overflows are stored in 34.88 mil gal
facilities and routed to 148 mgd treatment facilities
(2000 gpd/ft2 F/S with alum and polymer) at the Van Lare
STP which also handles east side overflows. West side post-
first-flush overflows are treated by primary swirl concen-
trators designed to handle peak flow.
Alt. 4. Same as Alt. 1, except west side post-first-flush overflows
are discharged untreated.
Alt. 5. Same as Alt. 1, but all west side overflows to be treated
by the local swirl devices. East side flows to be sent to
the Van Lare STP.
Alternate 2 (treatment of west side overflows at a central plant on
the Genesee River) and Alternate 3 (treatment of all area overflows at a
central plant at the Van Lare STP)involved evaluation of many subalternatives.
The cost/benefit analysis included evaluation of several process trains
(one-and two-level treatments), storage vs treatment optimizations, and
allowance for various amounts of untreated overflow bypass. Computer
modeling of costs and performance allowed evaluation of 576 subalternatives
within these two alternatives.
DEVELOPMENT OF COST RELATIONSHIPS
Capital cost versus treatment capacity relationships were compiled
from several literature sources (8) and adjusted to a common basis to
allow comparison of the various sources. From these sources relationships
were selected incorporating the effect of design loadings. These curves
are based on November 1976 projected costs.
The total capital cost requirements for each alternative include the
costs of the assumed treatment facilities, storage, pumping, and sludge
treatment. The relationships used in the cost development are shown on
Table 40.
COST OPTIMIZATION
A tradeoff exists between costs of storage and treatment facilities;
i.e., larger storage facilities permit smaller capacity treatment plants.
Generally, an optimum combination of sizing treatment and storage facilities
will permit the development of least cost facilities. In the evaluation of
Alternates 2 and 3, each subalternative was evaluated for four different
combinations of storage-treatment sizes. The determination of storage versus
treatment requirements was discussed in Section 8. The overflow mass
diagram (Figure 49} was used to compile the combinations of storage vs
treatment in Figure 50 for different bypass volumes.
When storage and treatment costs were determined from the combinations
derived from Figure 50,relationships similar to Figure 57 were obtained.
In general, the trends of these curves showed that the more capital intensive
142
-------
systems (e.g., F/S at 800 gpd per ft^ and the two-level systems) showed
optimization which maximized the amount of storage and minimized the size of
treatment facilities. The less capital intensive systems such as swirl con-
centrators showed minimum cost with minimum storage. Other systems showed
minimum cost at some compromise between the maximized extremes of storage
and treatment capacities.
TABLE 40. PROCESS COST MODELS
A. Primary Treatment:
1. G/S
2. P/S Cost functions presented in detail in Volume II
3. F/S
B. Pumping from storage to Interceptor
$ mil = 0.185 QD°-484
C. Disinfection System
$ mil = 0.011 QD0-851
D. Sludge Pumping Cost
$ mil = 0.0274 Qo0'5
E. Sludge Treatment
$ mil = 4.73 (assumed independent of QD - based on average storm
conditions for volumes of CSO treated)
F. Filtration System
$ mil = 0.23 QD0-753
G. Tunnel Storage
Tunnel
$/ft = 35.80 (dia in
Drop Shaft
$ mil = 2.23 each
H. Force Main Cost
1. Van Lare STP Alt. $ mil = 0.625 Q|>
2. River Alt. $ mil = 0.337 Qp
Note:QD = design flowrate in mgd
COST-EFFECTIVE RELATIONSHIPS
The weighted performance for each system was developed by analyzing the
time-variable removal of TSS associated with the application of the 2 Yr-2 Hr
design storm. The projected TSS removal for each system was developed
for each half-hour time step and an overall weighted TSS removal was
calculated for the design storm. The performance models developed in the
pilot plant study were used for this purpose (8). The weighted performances
assuming constant flow and variable influent TSS were as follows:
143
-------
I40T
130--
120--
no--
en
O
u
I- IOO--
LU
DC
90--
cr
o
^ 80--
70--
60
F/S 800 gpd/ft'
F/S 1500 gpd/ft
F/S 2000 gpd/ft2
G/S
Alternate 3 (with 0 bypass)
All 'River Overflow Conveyed
to Wet-Weather Treatment Facility
at Van Lare STP
200 400 600
TREATMENT RATE (mgd)
800
FIGURE 57. Typical Storage-Treatment Cost Optimization
144
-------
% TSS Removal
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
F/S 0 800 gpd/ft2 with chem. trtmt.
F/S 01500 gpd/ft2 with chem. trtmt.
F/S 02000 gpd/ft2 with chem. trtmt.
P/S (D2/Di = 12, Nf = 0.167)
G/S (D2/D1 = 6, Nf = 0.167)
System
System
System
System
System
with
with
with
with
with
filtration
filtration
filtration
filtration
filtration
10
10
10
10
10
gpm/ft2
gpm/ft2
gpm/ft2
gpm/ft^
gpm/ft2
70,
66.
63.
39.
24.
94.
93.
92.
87.
84.9
Alternates 1 and 5 (local swirl concentrators with no storage) assumed TSS
removals as a function of both variable flow and influent TSS, again using
the pilot plant performance model.
The cost-effective relationships for the subalternatives of Alternate 2
are shown on Figure 58. Performance is represented by the total area wet-
weather TSS discharge (pounds per design storm). The total area wet-weather
TSS discharge is the sum of the amounts of solids discharged to the Genesee
River (in both bypass and treated effluents) and the solids in the effluent
of wet-weather treatment facilities located at the Van Lare STP. The TSS
discharge figures take into account that sludges from local treatment facili-
ties are routed back to the interceptor and must be treated again at wet-
weather facilities at the Van Lare STP.
The cost-effective relationships for Alternate 3 are shown on Figure 59 .
Figures 58 and 59 indicate that system performance becomes limited principally
by type of treatment employed and amount of bypass allowed. As the discharge
requirement becomes more stringent, the number of feasible available alter-
natives becomes limited with the subsequent increase in capital cost.
Cost-effective comparisons of all the alternatives are summarized on
Figure 60. Minimum cost lines are traced for Alternates 2 and 3 for conditions
of zero-bypass and variable permitted bypass. The minimum cost lines per-
mitting bypass are about the same for Alternates 2 and 3.
Central treatment on the Genesee River (Alt. 2) is slightly more cost-
effective than Alternate 3 for the zero-bypass condition. However, Alternate
2 results in the discharge of primary treated wastewater to the dissolved
oxygen assimilation capacity limited Genesee River whereas Alternate 3
results in the discharge of treated wastewater to the Rochester Embayment of
Lake Ontario.
The local treatment alternatives (I, 4 and 5) appear to offer no
cost performance advantages over the central treatment alternatives (2 and 3).
145
-------
150
200
West Side Treatment - River
XiaSP Treatment - Van Lore
_NO .STORAGE _OR TREATMENT 2yr-2hr Storm
W/O FILTERS
W/ FILTERS
\
mil gal BYPASS
-eo mil gal BYPASS
\ - -040 mil gal BYPASS
\ \ 28
\ \
\
U7
_ ,c_ mil gal BYPASS
IUU
^o mil gal BYPASS
20 mil gal BYPASS
- 50
9"
o mil gal BYPASS
Note: XI = Cross Irondequoit
SP = St. Paul Blvd Interceptor
H 1 f
50 100 150
STORAGE a TREATMENT COST {f mil )
FIGURE 58. Cost-Effectiveness Relationships for Subalternatives
of Alternate No. 2 - River Overflow Interceptors with
Overflow Treatment Facility Centrally Located
146
-------
250-
200--
i
o
SO
o
tr I5O-
aj
S
UJ
IU
<
UJ
o:
** IOO--
50--
All Overflow Treated at VanLare
2yr.-2hr. Storm
NO STORAGE OR TREATMENT
1
J20
i 111 r
I *f \
\ \ 1
\ \ \
\ \ \
v \ \
X \ V
4 \ \
35^, \ \
- W/O FILTERS
_ W/FILTERS
\
^^-ec mil gal \
\
40 mil gal BYRASS
eo mil .gal ,6
BYPASS ~~ \ 2° mil go! SYRftSS
p *9
\
mil gal BYBVSS
20 mil gal
BYPASS
o mil gal
BYRASS
50 100 150
STORAGE 8 TREATMENT COST { $ mil )
FIGURE 59. Cost-Effectiveness Relationships for Subalternatives
of Alternate No. 3 - River Overflow Interceptor with
Overflow Treatment Facility at Van Lare STP
147
-------
2yr-2hr. Storm
250T
NO STORAGE OR TREATMENT
(ZOO-
u
to
0 ISO-
IU
100+
i
50--
Art. 3 , 0 Bypass
50 IOO
STORAGE 8 TREATMENT COST (8 mil)
ISO
FIGURE 60. Cost-Effectiveness Summary for the
Structurally Intensive Alternatives
148
-------
EFFECTIVE ANALYSIS RELATIVE TO THE GENESEE RIVER
Study area loadings of TSS and oxygen demanding constituents are most
critical to the water quality of the Genesee River. Water quality modeling
under MA 7CD10 conditions indicates that the Genesee River can accept only
very small quantities of wet-weather overflows and still maintain the
dissolved oxygen standard of 4 mg/1.
The only alternative capable of completely eliminating wet-weather
loadings to the Genesee River under design storm conditions is Alternate 3
for the condition of zero bypass. Figure 61 shows the cost-benefit
relationship of the alternatives relative to the water quality of the
Genesee River. Effectiveness is indicated as the percent reduction of
solids loadings to the Genesee River. Additional loadings to the Rochester
Embayment are not included in this analysis. The best reduction attainable
with Alternate 2, assuming zero bypass, is approximately 93 percent.
The local treatment alternatives (1, 4 and 5) result in only 32-65
percent reductions in TSS loadings to the Genesee River under application
of the 2 Yr-2 Hr design storm.
149
-------
100 T
80-
tr
LJ
UJ
UJ
to
uj
UJ
o
(ft
60-
O 40+
z
O
Q
UJ
cr
*
2O- -
[Alt. 3-0 Bypass)
VIt. 2-0 Bypass
2 Level
A it. 3-20 mil gal Bypass
kit 2 -20 mil gal Bypass
Ait.3 - 40 mil gal Bypass
It. 2-0 Bypass
Level
Cost - Benefit Relation
Relative to Genesee River Impact
Mt2-60 mil gal
Bypass
50
100
150
TOTAL STORAGE » TREATMENT
COST ( * mil )
FIGURE 61. Cost-Benefit Summary for the Structurally Intensive
Alternatives
150
-------
SECTION 10
DESCRIPTION OF MASTER PLAN CONFIGURATION AND ABATEMENT SCHEDULE
The result of the CSO Abatement Program is a Master Plan and abatement
schedule derived from the analysis of the various alternatives. Such a
schedule has been developed based on the phasing of the most applicable
nonstructural, minimal structural, and structurally intensive alternatives as
presented and discussed in Sections 8 and 9.
Those measures found to be most applicable to the reduction in discharge
of combined sewer overflow to the Genesee River, Irondequoit Bay, and the
Rochester Embayment of Lake Ontario are presented in Figure 62 . The
nonstructural alternatives are presented in the form of "Preparation and
Implementation of Source Control Regulations" which include planning
concepts to minimize percent imperviousness associated with future growth,
the application of porous pavement in selected drainage areas, and the better
utilization of surface storage and overland flow attenuation.
In addition to the application of source control measures, it is also
recommended that a significant effort be extended in the area of system
maintenance. It has been shown that the level of conveyance system maintenance
is significant in reducing the impact of combined sewer overflows. Conveyance
system maintenance not only alleviates flooding problems but also helps to
eliminate or at least minimize the first-flush phenomenon.
It is also recommended that the extent (frequency and areal coverage)
and effectiveness of street sweeping be increased. Optimization of street
sweeping frequency and efficiency is an effective method in reducing the
surface buildup of pollutants. Control over street cleaning frequency and
efficiency can be an effective method for reducing the solids, toxicants,
and oxygen demanding constituents contributed to the receiving waters via
both combined sewer overflow and stormwater discharges.
Of the minimal structural alternatives evaluated, it was found that
modest improvements to the St. Paul Boulevard Interceptor, selective regulator
modifications, and the partial blockage of high impacting overflows could
have a very significant effect on reducing the discharge of combined sewer
overflows from the conveyance system relief points. It is projected that
the removal of the three throttling constraints in the St. Paul Boulevard
Interceptor will reduce the annual discharge of combined sewer overflow
by 73 percent. Regulator modifications would have to be conducted in
conjunction with the interceptor modifications in order to take advantage
of the enhanced conveyance capacity. It is interesting to note that due
to the attenuation of flow within the interceptor, the peak flow measured
151
-------
en
ro
Program
Facilities Plans
Blockage of Three Impacting Overflows
and Weir Elevation Changes
Interceptor Improvements ond
Regulator Modifications
%%%%3 Preparation and Implementation of Source
Control Regulations
Construction of Wet Weather
Facilities at Van Lore
Installation of Control System
//////////////////^^^^^
t t t
Correction of Localized System Flooding and
CSO Rtfcf Using Small Capacity Stormwater Retention
_ Tanks and Regulators
xi Existing
Augmentation wi»h
Additional ln~ System
Storage Capacity
Evaluation of System
Performance
Milestone Evaluations
-I 1 1
1977
1978
1979
1980
DATE
1981
1982
1983
FIGURE 62. Rochester Combined Sewer Overflow Abatement Program
Preliminary Implementation Plan
-------
at the existing Van Lare STP dry-weather facilities would not be expected
to increase beyond 270 cfs based on the 2 Yr-2 Hr design storm. The
duration of the peak flow would, however, be expected to be tripled as a
result of the modifications.
Other minimal structural alternatives evaluated and subsequently
recommended for application include the addition of control structures and a
control system to take optimum advantage of existing and proposed potential
in-system storage. It is proposed that control structures be placed at the
outlet of the Genesee Valley Interceptor, the proposed Genesee River
Southeast Interceptor, and the Culver Goodman Tunnel complex. There is a
potential storage capacity of approximately 80 mil gal which would not be
utilized without the placement of the these control structures. Furthermore,
the addition of control structures at selected locations within the surface
conveyance system can also possibly reduce the frequency and quantity of
overflow by utilizing the available storage in many of the smaller sewer
conduits that are tributary to the main interceptor.
Upon implementation of control structures, an operating management tool
is required to make the most intelligent use of the available storage during
a wet-weather event. For this purpose, it is recommended that a control
system be developed prior to or at the time of designing the required
control structures. It is proposed that the concepts involved in the
application of the SSM be incorporated as part of the predictive logic in
the control system.
Of the structurally intensive alternatives, it is proposed that the
high-rate flocculation and sedimentation facilities be constructed at the
site of the present Van Lare STP. These facilities with a hydraulic
capacity of 275 mgd and a design surface loading rate of 2000 gpd/ft^ would
have the capability of achieving a 63 percent average removal of total
suspended solids for the design storm.
It is of most importance that the wet-weather treatment facilities
be placed on line prior to the acceptance of combined sewer overflow dis-
charges to the Cross Irondequoit Tunnel. The present dry-weather facilities
at the Van Lare STP do not have sufficient hydraulic or process capacity to
handle the total wet-weather flows anticipated at the Irondequoit Pumping
Station.
It is also proposed that a number of the localized system flooding
and CSO problems be corrected using relatively small capacity stormwater
retention tanks. There are several areas identified in this report
where this type of remedial solution may have application. The retention
tanks must be applied to the retention of stormwater only and therefore
only connected to the catchbasins with attenuated discharge to the
trunk sewer. The rate of discharge from a stormwater detention tank
would be determined by the head presented by the trunk sewer.
As shown in Figure 62 the implementation of the Master Plan should
be well reviewed on a step-by-step basis to allow feed back as to the
effectiveness of each abatement measure following application. Of
153
-------
particular importance is the evaluation of the need for additional in-
system storage capacity which is proposed to be the last item to be
implemented. The initial network modeling indicates that 62 mil gal of
in-system storage capacity would be necessary to prevent discharge of
system overflow resulting from the design storm. Many of the measures
recommended prior to storage augmentation may further reduce the necessary
storage requirements. It is for this reason that an extensive evaluation
of system performance be required prior to the application of additional
storage.
Figures 63 through 66 show the expected response of the quality of
the Genesee River and Rochester Embayment of Lake Ontario to the implementation
of the most applicable nonstructural, minimal structural, and structurally
intensive alternatives as presented in Figure 61 The figures indicate that
the quality of the River and of the Embayment can be substantially enhanced
by the implementation of several nonstructural and minimal structural
alternatives. These alternatives represent small capital expenditures
relative to the benefits derived from their implementation.
154
-------
S2I
Percent Reduction of System Overflows
PO 01
o o o
m
co
a- o
0) <-h
C+ OJ
CO CD
~S O
a> -s
CO -h
ro
a.
c
o
fD
CO
T3
O
3
CO
ro
TD
(D
(D
3
O
3
O
-------
on
Oi
10,000,000
1,000,000
100,000
o
10,000
1,000
100
10
R a D Program
[^Facilities Plans
J^sS^s^^ Interceptor Improvements
Design Storm
2 yr-2 hr
gulator Modifications
Ave.
Annual
0--0
n-n
A--A
Parameter
TKN
TSS
BOD
High Impacting Overflows
Implementation of Source Control Regulations
if Wet Weather Facilities at Van Lore
Control Structures Qt
Control System
Correction of Localized System Flooding a
:SO Relief Using Small Capacity Stormwater
Detention Tanks a Regulators
Activity on Existing Tunnel System
n of System Performance
I
Milestone Evaluations
1 + t
Augmentation with
Additional In-System
Storage Capacity
1977 1978
1979 1980
DATE
1981
1982 1983
FIGURE 64. Total System Pollutant Loading Reduction to the Genesee River
Response to Implementation of Abatement Measures
-------
ZST
Average Annual Potential
Days of Dissolved Oxygen Concentration
row
CD
3D
rn
CTl
CJ1
^D CD
rc ro
CO 3
"a ro
o to
3 ro
O <
(D
i -S
3
"a xi
-J fD
CD Q-
B c
03 O
3 r+
r+ -1-
Cu O
O -1-
O
<-i-
m
t-l- -'
fD CU
05
3 O
<-+ -1-
LO
3; i/1
n> o
CU i
>J-\ <
c ro
-5 Q-
05
1/1 O
X
ro
3
o
13
-S
ro
-------
5-
4--
.2
I
c O
i- en o
g S S
2 m
g
o
RQO Program
Facilities Plan
Regulator Modifications
Blockage of High Impacting Overflows
Preparation and Implementation of Source Control Regulations
MILESTONE EVALUAT
I
Interceptor Improvements
Construction of Wet Weather Facilities at Van Lore
Addition of Control Structures and
Installation of Control System
Correction of Localized System Flooding
and CSO Relief
Construction Activity on Existing Tunnel System
Evaluation of System Performance
1977 1978 1979 I960 1981 1982 1983 1984 1985
DATE
Augmentation w/
Additional In-
ystem Storage
Capacity
FIGURE 66. Reduction in Potential Ontario Beach Closing Days
Response to Implementation of Abatement Measures
-------
REFERENCES
1. Lager, J. A., W. G. Smith, W. G. Lynard, R. M. Finn, and E. J. Finnemore,
Urban Stormwater Management and Technology: Update and User's Guide.
EPA-600/8-77-014, U.S. Environmental Protection Agency, Cincinnati,
Ohio, 1977.
2. Wyeoff, R. J., J. E. Scholl, and S. Kissoon. 1978 Needs Survey: Cost
Methodology of Control of Combined Sewer Overflows and Stormwater
Discharge. EPA-430/9-79-003, U.S. Environmental Protection Agency,
Washington, D.C., 1979.
3. Erdman Anthony Associates, Lozier Engineers, Inc., and Seelye Stevenson
Value and Knecht, Inc., (A Joint Venture). Wastewater Facilities Plan-
Interim Report. Rochester Pure Waters District, Rochester, New York,
1975.
4. Black & Veatch. Report on Comprehensive Sewerage Study for City of
Rochester, New York. New York State Department of Health, 1969.
5. Lozier Engineers, Inc. East Side Trunk Sewer Study. Monroe County
Division of Pure Waters, Rochester, New York, 1973.
6. Monroe County Division of Pure Waters. Facilities Report on Inter-
ceptor Improvements, C-36-837, for the Rochester Pure Waters District,
Rochester, New York, 1974.
7. Lager, J. A., T. Didriksson, and G. B. Otte. Development and Appli-
cation of a Simplified Stormwater Management Model. EPA-600/2-76-218,
U.S. Environmental Protection Agency, Cincinnati, Ohio, 1976.
8. Drehwing, F. J., C. B. Murphy, Jr., S. R. Garver, D. F. Geisser, and
D. Bhargava. Combined Sewer Overflow Abatement Program, Rochester,
New York, Volume II: Pilot Plant Evaluations. EPA-600/2-79-031b,
U.S. Environmental Protection Agency, Cincinnati, Ohio, 1979.
9. Canale, R. P., and S. Nachiappan. Steady State Modeling Program. Sea
Grant Technical Report No. 27, University of Michigan, 1972.
10. Brandstetter, A. Assessment of Mathematical Models for Storm and
Combined Sewer Management. EPA-600/2-76-175a, U.S. Environmental
Protection Agency, Cincinnati, Ohio, 1976.
11. Huber, W. C., J. P. Heaney, M. A. Medina, W. A. Peltz, H. Sheikh, and
G. F. Smith. Storm Water Management Model User's Manual, Version II.
EPA-670/2-75-017, Cincinnati, Ohio, 1975.
159
-------
12, Water Resources Engineers, Inc. Modifications to the EPA Stormwater
Management Model Documentation (Version of March, 1975), 1975.
13. O'Brien & Gere Engineers, Inc. Genesee River Water Quality Investiga-
tions; Part II Report submitted to Joint Venture, Rochester, New York,
1976.
14. American Public Works Association. Water Pollution Aspects of Urban
Runoff. Report No. 1103DNS01/69, U.S. Environmental Protection Agency,
Cincinnati, Ohio, 1969.
15. Colston, N. V., Jr. Characterization and Treatment of Urban Land Runoff.
EPA-670/2-74-096, U.S. Environmental Protection Agency, Cincinnati, Ohio,
1974.
16. Dorsch Consult, Consulting Engineers. Report on a New Method to Evaluate
Urban Runoff Pollution and Its Effects on Receiving Waters. Munich,
West Germany, 1975.
17. Keifer, C. J., and H. H. Chu. Synthetic Storm Pattern for Drainage
Design. In: Proceedings of the American Society of Civil Engineers,
Hydraulic Division, Paper 1332: 1-25, 1957.
18. Thelen, E., W. C. Grover, A. J. Hoiberg, and T. I. Haigh. Investigation
of Porous Pavements for Urban Runoff Control. Report No. 11034DUY03/72,
U.S. Environmental Protection Agency, Cincinnati, Ohio, 1972.
19. Fair, G. M., and J. C. Geyer. Water Supply and Wastewater Disposal.
John Wiley and Sons, Inc., New York, New York, 1965.
160
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-031a
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Combined Sewer Overflow Abatement Program, Rochester,
N. Y. Volume I: Abatement Analysis
5. REPORT DATE
July 1979(issuing date1)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Frank S. Drehwing, Cornelius B. Murphy, Jr.,
David J. Carleo and Thomas A. Jordan
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
O'Brien & Gere Engineers, Inc.
1304 Buckley Road
Syracuse, N. Y. 13221
10. PROGRAM ELEMENT NO.
A42B2A
11. CONTRACT/GRANT NO.
Y005141
12. SPONSORING AGENCY NAME AND ADDRESS
Great Lakes National Program Office
U. S. Environmental Protection Agency
536 South Clark St., Room 932
Chicago, Illinois 60605
13. TYPE OF REPORT AND PERIOD COVERED
Final Mav 1974 to Sept. 1977
14. SPONSORING AGENCY CODE
U.S. EPA
15. SUPPLEMENTARY NOTES Technical assistance provided by Municipal Environmental Research
Laboratory, Edison, N. J. - Richard Field, Tony Tafuri of SCSS-MERL and Larry
Moriarty of Region II, Rochester Field Office served as project officers.
16. ABSTRACTsr section ±(Jo(a) Demonstration Project.
Pollution abatement analyses, conducted in conjunction with system network modeling
studies and supported by combined sewer overflow (CSO) monitoring and sampling,
were initiated with the ultimate goal of formulating a cohesive and workable Master
Plan for CSO reduction and control. The Master Plan was developed in light of fiscal
constraints, sewer system complexities, necessity for optimized benefits from
minimal capital and operating expenditures, and best use policies of the affected
receiving waters. The presented methodology is considered applicable to other urban
areas.
Preliminary analysis of BMP and minimal structural alternatives indicated that by
addressing the major sources of pollution and by eliminating throttling constraints
within the existing sewerage system, a substantial decrease in the total annual load
of contaminants to the receiving waters from rainfall induced CSO can be achieved for
relatively small capital expenditures. These measures can be initiated within a
short period of time, thereby immediately reducing pollution to the receiving waters
while long term design and construction of more structurally intensive alternatives
are undertaken.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Combined sewers, overflows, water pollution
waste treatment, flow regulation,
abatement plan, cost benefit analyses, BMP,
minimal structural alternatives, structural
intensive alternatives, non-structural
alternatives, monitoring and sampling.
18. DISTRIBUTION STATEMENT
Available through National Technical
Information Service-Springfield, VA 22151
19. SECURITY CLASS (ThisReport}
Unclassified
21. NO. OF PAGES
161
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
& U. S. GOVERNMENT OFFICE 1981 - 751-632
161
------- |