EPA-600/2-77-065
Marcl) 1977 Environmental Protection Technology Series
Short Course Proceedings
APPLICATIONS OF STORMWATER
MANAGEMENT MODELS •1976
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-065
March 1977
SHORT COURSE PROCEEDINGS
APPLICATIONS OF STORMWATER MANAGEMENT MOLELS
1976
Edited By
Francis A. DIGlano
Donald D. Adrian
Peter A. Mangarella
University of Massachusetts
Amberst, Massachusetts 01002
Grant Number 803069
Project Officer
Anthony N. Tafuri
Storm and Combined Sewer Section
Wastewater Research Division
Municipal Environmental Research Laboratory (Cincinnati)
Edison, New Jersey 08817
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENT&SfliPAECTION AGENCY
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, 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 commercial products con-
stitute endorsement or recommendation for use.
ii
-------
FOREWORD
The Environmental Protection Agency was created because of increasing public
and government 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 problem solution
and it involves defining the problems, 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 preservation 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 communications link'between the
researcher and the user community.
The Short Course Proceedings contained herein include discussions of various.
computer assisted models applied to stormwater management, criteria for
selecting models either as planning or design tools and methods for col-
lecting f;Leld data for model verification. A case study designed for
instructional use highlights application of the EPA Stormwater Management
Model (SWMM).
Francis T, Mayo
Director
Municipal Environmental
Research Laboratory
iii
-------
PREFACE
This Short Course was first held at the University of Massachusetts
(August 19-23, 1974) and then revised and offered at the University of
Massachusetts (July 28-August 1, 1975), Asilomar Conference Grounds,
Pacific Grove, California (January 5-9, 1976) and the University of Chicago
(July 18-23, 1976). Registration totaled close to 300 with representation
by consultants; Federal, State and Municipal agencies, including those of
the Canadian government; and university researchers.
Based upon responses to a questionnaire completed by almost all
attendees, the following profile of professional backgrounds and interests
emerged:
• professional training areas represented were sanitary engineers (48%):
hydraulic engineers (24%); planners (6%) and other disciplines (22%).
• highest degrees earned were BS (35%); MS (52%); PhD (11%) and other
(2%).
• interests in stormwater management models were Section 208 imple-
mentation (32%); sewer system design (23%); research (9%); environ-
mental impact assessment (8%); and various combinations of the above
(28%).
• knowledge of stormwater management models prior to attending the
Short Course was indicated by 55% of the registrants.
• as a result of attending this Short Course, 87% of the registrants
indicated that use of models will be encouraged in their firms and
agencies.
• of the models presented, 68% of the registrants indicated that applying
a combination of EPA SWMM with other, more simple models would best
serve their needs.
iv
-------
ABSTRACT
This Short Course on applications of stormwater management models
is a follow-up to a course sponsored by the U.S. EPA and now available
as EPA Report 670/2-75-065, The proceedings contained herein represent
an entirely new set of contributions from participating speakers. The
objective of this Short Course is to provide practitioners with the cap-
ability to apply specific models directly. Toward this goal, a discussion
of the common components of stormwater management models first gives an
overview of modeling needs. The U.S. EPA Stormwater Management Model (SWMM)
is then described in detail and an illustrative case study presented.
The methodology for data preparation is outlined and sample input and output
data given for the Rainfall-Runoff, Transport, Storage/Treatment and
Receiving Water Blocks of the EPA SWMM, A discussion of criteria for
selecting models for application as either planning or design tools is
then presented along with illustrations of the use of two simplified models.
Finally, the techniques for collecting field data for model calibration are
presented and the performance of commercially available sampling equipment
assessed.
This report was submitted in partial fulfillment of Grant Number
803069 by the Department of Civil Engineering at the University of
Massachusetts, under the sponsorship, of the Environmental Protection Agency.
This report covers the period June 3, 1975 to August 31, 1976 and work
completed as of August 31, 1976.
-------
CONTENTS
Page
Foreword Ill
Preface iv
Abstract v
Acknowledgments . . . Ix
Short Course Faculty x
Introduction 1
Richard Field and Anthony N. Tafuri
Role of Models in Urban Stormwater Management Planning ... 4
James P. Heaney
Methodology for Evaluating the Cost of Urban Stormwater
Quality Management 15
James P. Heaney and Sheikh M, Hasan
Introduction to the EPA Stormwater Management Model
(SWMM) 34
Wayne C. Huber and James P. Heaney
Hydrology Review 74
Donald Dean Adrian
Hydraulics Review 77
Donald Dean Adrian
SWMM Application Study Guide . , 93
Thomas K. Jewell, Peter A. Mangarella,
Francis A. DiGiano and Donald D. Adrian
Criteria for Selection "of "S'to'rmwate'r 'Management
Models 239
John.A. Laeer
A Simplified Stormwater Management Model for
Planning ..."......> . ,259
John A. Lager
vii
-------
A Storage, Treatment, Overflow and Runoff Model for
Metropolitan Masterplanning 287
Larry A. Roesner
Collection of Field Data for Storrawater Model
Calibration 327
Philip E. Shelley
viii
-------
ACKNOWLEDGMENTS
I would especially like to thank Mr. Richard Field, Chief and Mr.
Anthony N. Tafuri, Project Officer of the Storm and Combined Sewer Section
and Mr. Harry Torno, Staff Engineer of the Municipal Pollution Control
Division of the EPA for their cooperation in planning and executing this
Short Course.
I would also like to thank the participating speakers. Special recog-
nition is to be given to Mr. Thomas K. Jewell, graduate research assistant
and Mr. David Gaboury, undergraduate research assistant, for their tireless
effort in applying the EPA SWMM to the UMASS computer system and in develop-
ing a case study for the workshop sessions.
I am also grateful to the staff of Conference Services at the University
of Massachusetts for handling all of the Short Course details. Finally, I
acknowledge the help of our Environmental Engineering secretary, Miss
Dorothy A. Blasko, for typing the workshop materials.
Francis A. DiGiano, Short Course Chairman
Associate Professor of Civil Engineering
University of Massachusetts/Amherst
ix
-------
SHORT COURSE FACULTY
Donald D. Adrian, PhD
Professor of Civil Engineering
University of Massachusetts/Amherst
Francis A. DiGiano, PhD
Short Course Chairman
Associate Professor of Civil Engineering
University of Massachusetts/Amherst
Richard Field, Chief
Storm & Combined Sewer Section
Municipal Environmental Research Lab (Cinn.)
USEPA/Edison, New Jersey
James P. Heaney, PhD
Associate Professor of
Environmental Engineering Sciences
University of Florida/Gainesville
Wayne C. Huber, PhD
Associate Professor of
Environmental Engineering Sciences
University of Florida/Gainesville
Thomas K. Jewell, Research Associate
Department of Civil Engineering
University of Massachusetts/Amherst
John A. Lager, Vice-President
Metcalf & Eddy, Incorporated
Palo Alto, California
Peter A. Mangarella, PhD
Assistant Professor of Civil Engineering
University of Massachusetts/Amherst
Larry A. Roesner, PhD, Principal Engineer
Water Resources Engineers, Incorporated
Walnut Creek, California
-------
Philip E. Shelley, PhD., Director of
Energy & Environmental Systems
Washington Analytical Services Center, Incorporated
Rockville, Maryland
Anthony N. Tafuri, Storm & Combined Sewer Section
Municipal Environmental Research Lab (Cinn.)
USEPA/Edison, New Jersey
xi
-------
INTRODUCTION
By
Richard Field* and Anthony Tafuri
Welcome to this fourth short course on the application of stormwater
management models. As many of you know, the University of Massachusetts
and the U.S. Environmental Protection Agency have jointly sponsored three
similar courses in the past. Each proved to be very successful and we're
looking forward to another success this week.
Before getting into the course, I would like to briefly introduce
you to EPA's Stormwater Program and the importance of mathematical
modeling to the Program. The Program started in 1965 under the
U. S. Public Health Service and is presently assigned to EPA's Municipal
Environmental Research Laboratory, Cincinnati; however, it is physically
located in Edison, New Jersey. A primary objective of the Program
has been to advance the technology for urban stormwater and combined
sewer overflow treatment, control and management. Part of this
objective involves the development and refinement of mathematical models
for the total management of sewer systems considering the full range
of wet-and dry-weather flow pollution control.
Traditional analytical methods are inadequate to evaluate the
dynamic reaction of the urban drainage system to storm events. The
Program recognized that a mathematical computer simulation program to
model urban runoff quantity and quality during the storm process would
provide an invaluable tool for engineers. Such a simulation system
was completed in 1971 and is known as the EPA Stormwater Management
Model (SWMM).
SWMM is capable of being used in detailed master planning as it
enables the decision-maker to learn the consequences of alternative
courses of action. For example, if storage facilities are being
considered, the planner can calculate the hydraulic and the pollutant
load that the facility will receive. Similarly, the pollutional
load imposed on streams by storms can be calculated. SWMM is also
useful in predicting and planning required changes in the hydraulic
conveyance capacities of the collection system brought about by local
overloading. For design purposes, the model is useful in determining
*
Chief and Staff Engineer, Storm & Combined Sewer Section, Wa'stewater
Research Division, Municipal Environmental Research Laboratory, U.S.
Environmental Protection Agency, Edison, New Jersey 08817, respectively.
-------
the anticipated flowrates that a specified sewer needs to convey.
The size and slope of the sewer can then be specified. Although the
model's operational capability has not yet been applied, there
is a likelihood it, or other versions, will be used in San Francisco
in the near future, to demonstrate real-time automatic control of
combined sewer systems. The model's operational capability lies
in its ability to provide information on what to expect at various
points in a stream in terms of flow from a rainfall event.
Since its release, SWMM has been refined and expanded. Program
"bugs" have been eliminated. We are now able to demonstrate the ability
of computer-assisted mathematical models to enhance urban stormwater
management by analyzing a major combined sewer system, by selecting
and designing control and treatment approaches based on cost effective-
ness, and by designing a computerized means of overall management
of the system during storm flows. It is the eventual goal to
handle all wet-and dry-weather flows in this manner.
A simplified mathematical model approach to the urban stormwater
management plan is presently being perfected. This approach determines
overflow occurrences on a probabilistic basis according to historical
rainfall records. Subsequently, the overflow problems within a
drainage system can be. pinpointed with the simplified model and then
defined in greater detail with SWMM for specific locations and storms.
Detailed discussion on SWMM versus the simplified model will be
presented later on this week along with information on other stormwater
management models that are presently available.
EPA FILM "STORMWATER POLLUTION CONTROL: A NEW TECHNOLOGY"
The film portrays a complete overview of the U.S. EPA's involve-
ments in developing countermeasures for pollution from combined sewer
overflows and stormwater discharges.
It starts with a background of sewer construction leading into
the pollution problems caused by urban runoff. Early drainage plans
made no provisions for storm flow pollutional impacts. Various
countermeasures are shown that are ready for implementation by
municipalities. The majority are full-scale operations and include
(a) control techniques, e.g. infiltration prevention, street cleaning,
porous pavement (for runoff attenuation), in-sewer routing and storage,
improved overflow regulators, and off-line storage; (b) physical-
chemical disinfection and biological treatment as shown adjuncts
to the sanitary plant or as remote "satellite" facilities at the outfall;
(c) and various schemes which reclaim stormwater for beneficial purposes
including enhancement of visual aesthetics, recreation and water supply.
-------
A couple of years ago, we felt that the time was right to make
an evaluation of previous program efforts, and we implemented a
contract to develop the state-of-the-art and assess techniques
available to manage and treat combined sewer overflow and stormwater.
A text on this subject has been completed and is available.
Expanding on the cliche "a picture is worth a thousand words", we
felt a film would be worth one hundred thousand—so a film was
included as part of the work.
The specific reasons for this film were to alert the environmental
engineering community to the stormwater problem and the immediate need
to apply available technology to counteract it.
-------
ROLE OF MODELS IN URBAN STORMWATER MANAGEMENT PLANNING
By
James P. Heaney*
INTRODUCTION
Section 208 of the Federal Water Pollution Control Act Amendments of 1972
requires that plans be developed to abate and control point and non-point
waste sources in urban areas throughout the country. This section has
brought about an unprecedented amount of activity in devising integrated
urban environmental management plans. In addition to multi-purpose water
related programs, attention is being given to air pollution and solid waste
as they relate to the water pollution control program.
Model developments during the past 10-15 years have produced a wide variety
of tools which are available to assist the engineers and planners in con-
ducting these studies. Many of these models simulate the relevant physi-
cal, chemical, and biologic processes. Various optimization techniques
are available to assist in making decisions regarding the "best" solution.
Literally hundreds of papers have been written describing these simula-
tion and optimization models. The balance of this paper discusses how
I perceive that some of these models could fit into the 208 planning pro-
cess as it is currently envisaged. Because 208 planning is still in the
formative stages of development, it is reasonable to expect that others
might view the "problem" quite differently. Brandstetter (1975) and
Brown et al. (1974) have developed assessments of urban water management
models.
At a recent workshop in stormwater management (Wanielista, 1975), Bowling
presented a preliminary work plan for 208 planning in the Orlando, Florida,
area. This outline will be used as a format for discussing the relevance of
various types of models to the various phases of the study.
Associate Professor, Department of Environmental Engineering Sciences,
University of Florida/Gainesville.
-------
MODELS IN 208 PLANNING
1.0 PUBLIC PARTICIPATION AND LOCAL GOVERNMENT INVOLVEMENT
1.1 PUBLIC CONTACT SECTIONS
1.2 RESPONSIVE COMMITTEES
1.3 PUBLIC PARTICIPATION
1.4 REGIONAL CLEARINGHOUSE REVIEWS
No significant modeling efforts are envisaged for this phase of the
study.
2.0 ORGANIZATION
2.1 ADMINISTRATION
2.2 RELATIONSHIPS TO OTHER PROGRAMS
Existing models would not have much relevance to this phase of the
work plan.
3.0 DATA COLLECTION
3.1 EXISTING DATA SOURCES
3.2 PROJECTED DATA COLLECTION EFFORTS
3.3 AGENCIES WITH POTENTIAL DATA COLLECTION FUNCTIONS
3.4 ASSEMBLY AND REVIEW OF EXISTING DATA
3.5 DETERMINE ADDITIONAL DATA NEEDS AND APPROPRIATE AGENCIES
3.6 ADDITIONAL DATA COLLECTION
If one knows the type of models needed to conduct the study, then a
relatively good idea of data needs is available. Of course, the con-
verse is true, i.e., the availability of data and resources to collect
additional data strongly influences the selection of models. Decision
theoretic models can be used to assess the value of data if the,
decision problem is well specified. Unfortunately, 208 planning
activities tend to have fuzzy decision criteria since it is not clear,
a priori, what the "problem" is. In particular, we have yet to concur
on performance criteria for wet weather pollution. Thus, one still
needs to rely heavily on "engineering judgment" in the data gathering
phase of the study.
4.0 NATURAL AND PHYSICAL SYSTEMS ANALYSIS
4.1 DESCRIPTION OF THE PLANNING AREA
4.2 IDENTIFICATION AND USE CLASSIFICATION OF RECEIVING WATERS
-------
4.3 IDENTIFICATION OF HYDROLOGICAL, GEOLOGICAL AND
CLIMATOLOGICAL FACTORS
4.4 IDENTIFICATION OF TOPOGRAPHICAL FACTORS
4.5 IDENTIFICATION OF AQUATIC, TERRESTRIAL AND CULTURAL FACTORS
4.6 IDENTIFICATION AND EVALUATION OF ENVIRONMENTALLY SENSITIVE
AREAS
This phase of the study deals with an inventory and characterization
of the above factors. I would not envisage any significant modeling
effort during this phase of the study.
5.0 SOCIAL. ECONOMIC AND LAND USE ANALYSIS
5.1 EXISTING POPULATION CHARACTERISTICS
5.2 EXISTING ECONOMIC BASE
5.3 EXISTING LAND USE
5.4 FUTURE POPULATION, ECONOMIC BASE AND LAND USE
5 5 IMPACT
There are several land use based urban planning models which could be
useful in this phase of the study. Kilbridge et al. (1970) have
reviewed many of these models. A key question to be addressed in such
modeling efforts is to what extent might environmental considerations
affect future land use. Similar types of Impact analyses have been
done in urban transportation studies for a number of years. Relatively
well documented versions of these models are available from the US
Department of Transportation. However, it would be reasonable to
expect that the city already has a model available from their transpor-
tation studies.
6.0 WASTEWATER SOURCE ANALYSIS
6.1 DOMESTIC WASTE DISCHARGES
6.2 INDUSTRIAL WASTE DISCHARGES
6.3 JOINT TREATMENT SYSTEMS
6.4 STORM SEWER DISCHARGE LOCATIONS
6.5 TOTAL POINT SOURCE WASTE DISCHARGES
6.6 IDENTIFICATION OF NON-POINT WASTE SOURCES
6.7 TOTAL NON-POINT WASTE LOADS
6.8 TOTAL WASTE LOADS
6.9 FUTURE WASTE LOADS-SOURCES AND CHARACTERISTICS
Discharges of domestic wastewater and industrial wastes are relatively
easy to estimate since they have been monitored for a number of years
and exhibit steady state behavior in most cases. FILTH in the SWMM
model can be used to estimate domestic wastewater Ipads. The wet
weather aspects of the wastewater source analysis are the main problem.
-------
One could expect to have a relatively large number of storm sewer
discharge locations. A problem arises in defining receiving waters
for cases where the original creek or river system has been improved
to accommodate stormwater runoff. Is it a receiving water or is it
part of the drainage network?
Wet weather sources can be estimated using a model such as STORM
(HEC, 1975) which provides an hourly accounting of runoff and pollutant
discharges. Local data are needed to calibrate the pollutant loading
factors. The use of literature values could lead to very misleading
results. The Universal Soil Loss Equation (USLE) is gaining growing
acceptance as a tool for doing nonpoint source analysis. This equation
was incorporated in SWMM (Heaney, Huber et al., 1975) a few years ago
and has recently been added to STORM. Using the USLE, other pollutants
are expressed as a function of the soil loss.
7.0 WASTEWATER TREATMENT SYSTEMS
7.1 DESCRIPTION OF EXISTING FACILITIES
7.2 IDENTIFICATION OF PROPOSED FACILITIES
7.3 EXISTING AND PROPOSED SERVICE AREAS
7.4 IDENTIFICATION OF EXISTING PLANS OF FEDERAL, STATE AND
LOCAL GOVERNMENTAL AGENCIES AND PRIVATE ORGANIZATIONS
7.5 SLUDGE GENERATION FORECASTS
7.6 INFILTRATION/INFLOW AND OVERFLOW ANALYSIS
This phase of the study does not appear to rely heavily on models.
8.0 NON-POINT SOURCES
8.1 ASSEMBLY OF AVAILABLE NON-POINT SOURCE DATA
8.2 EVALUATION OF NON-POINT POLLUTION SOURCES
See Section 6.0 for a discussion of non-point source analysis.
9.0 WATER QUALITY ANALYSIS
9.1 PRESENT WATER QUALITY STANDARDS
9.2 DETERMINATION OF EXISTING WATER QUALITY
9.3 VERIFICATION OF PRESENT STREAM STANDARDS
9.4 EVALUATION OF ACHIEVABLE WATER QUALITY
9.5 WATER QUALITY PROBLEM AREAS
9.6 RELATIONSHIP OF EXISTING WASTE SOURCES TO PRESENT WATER
QUALITY
9.7 PROJECTED WATER QUALITY
-------
A wide variety of receiving water models are available. The
receiving water model in SWMM is very popular. The vexing ques-
tion in this phase of the study is how to relate receiving water
quality to wet weather flows. One needs to extend the traditional
event analysis to a continuous simulation of an extended period
of time (year or more).
10.0 GROUNDWATER PROTECTION AND ENHANCEMENT
10.1 GROUNDWATER QUALITY
10.2 SOURCES OF POTABLE WATER
10.3 SOURCES OF RECHARGE
10.4 SOURCES OF DEGRADATION
As with receiving water models, there are several groundwater models
available. Perez et al. (1974) have developed a conjunctive surface
groundwater model which examines both quantity and quality. Unfor-
tunately, little data exist to support a very sophisticated ground-
water model especially with regard to water quality.
11.0 DEVELOPMENT OF NON-POINT SOURCE CONTROL STRATEGIES
11.1 DEVELOP ALTERNATIVES
11.2 RANKING AND PRIORITY OF NON-POINT SOURCE CONTROL STRATEGIES
A recent report by Metcalf and Eddy, Inc. (1974) provides a nice
summary of the available wet weather controls and their effectiveness.
Murphy (1975) presents a methodology for doing this analysis using the
appropriate cost data and storage/treatment isoquants generated by
STORM. One critical assumption in STORM and SWMM is the assumed rate
at which pollutants are washed off the catchment. At present, it is
assumed that a uniform runoff of one-half inch per hour would wash
away 90 percent of the pollutants in one hour. By varying this
assumption, one changes the significance of the "first flush" effect.
If the first flush is significant one would concentrate on controlling
the initial part of each storm event. This would imply a reservoir
operating policy of storing the first part of the storm and bypassing
the rest of it. Here again, local data are needed.
12.0 LAND USE/WATER QUALITY RELATIONSHIP
12.1 EVALUATION OF EXISTING LAND USE PLANS
12.2 ASSESSMENT OF LAND USE CONTROLS
12.3 LAND USE/WATER QUALITY RELATIONSHIP
12.4 DEVELOPMENT OF RECOMMENDED LAND USE GUIDELINES
12.5 RECOMMENDED LAND USE PLAN MODIFICATIONS
-------
SWMM and STORM can be used for this phase of the study since the
models generate runoff and pollutants as a function of land use.
13.0 PUBLIC WORKS/WATER QUALITY RELATIONSHIP
13.1 DRAINAGE
13.2 STORMWATER COLLECTION AND DISPOSAL
13.3 SOLID WASTE DISPOSAL
13.4 LAKE RESTORATION
Our experience to date indicates that it is vital to integrate wet
weather quality control with quantity control. Specifically, joint
utilization of storage facilities is a particularly promising alter-
native. SWMM provides output regarding the hydrograph attenuation
and treatment occurring in storage facilities.' Also, on-site storage
can reduce stormwater collection costs considerably. It does not
appear that an adequate data base exists to relate solid waste
incinerator discharges to wet weather quality. The groundwater models
discussed in Section 10 can be used to examine the impact of land
disposal of solid wastes.
14.0 INSTITUTIONAL ANALYSIS
14.1 PRESENT FEDERAL, STATE, REGIONAL, LOCAL AND PRIVATE
PROGRAMS AND ORGANIZATIONS
14.2 EVALUATION OF INSTITUTIONAL ARRANGEMENTS
14.3 RECOMMENDATIONS OF INSTITUTIONAL ALTERNATIVES
The overall problem needs to be viewed as a multi-purpose,multi-group
decision situation wherein one is attempting to find an equitable
and efficient solution. A disproportionate amount of the effort
to date has addressed only the efficiency question. Burke and
Heaney (1975) present a broad overview of the models which are
available at this time. For example, it is possible to assess the
relative voting power of the various interest groups to evaluate the
fairness of various voting rules. Also, simulation models exist to
aid in assessing the bargaining arena, interest groups, salient issues,
etc. (Bulkley and McLaughlin, 1966). This important phase of the study
is often overlooked entirely.
-------
15.0 WASTE TREATMENT MANAGEMENT TECHNIQUES
15.1 POINT SOURCE SUBPLANS
15.2 NON-POINT SOURCE SUBPLANS
15.3 INSTITUTIONAL (FINANCE AND MANAGEMENT) SUBPLANS
Given a multi-purpose, multi-interest group situation one needs to
know, for each plan, the incidence of benefits and costs to each
purpose and each group (James and Lee, 1971). Coordinated plans
reap savings by taking advantage of economies of scale, inter-
dependencies in production functions, and more efficient use of
assimilative capacity. Deininger and Loucks (1972) present a nice
summary of optimization procedures for point source analysis. Heaney,
Huber et al. (1975) describe a procedure whereby equitable financial
arrangements can be determined.
16.0 EVALUATION AND COMPARISON OF ALTERNATIVE PLANS
16.1 ACHIEVEMENT OF TREATMENT STANDARDS (BAT, BPWTT, ETC.)
AND WATER QUALITY OBJECTIVES
16.2 MONETARY COSTS AND ADMINISTRATIVE FEASIBILITY
16.3 TECHNICAL RELIABILITY
16.4 LAND USE IMPACT
16.5 ENVIRONMENTAL FACTORS
16.6 IMPLEMENTATION FEASIBILITY AND RELIABILITY
16.7 PUBLIC ACCEPTABILITY
Models from Section 5 (land use), 14 (institutional analysis) and
15 (waste treatment management techniques) can be integrated to
assist in the overall evaluation. Standards and environmental
factors can be included as additional constraints.
17.0 SELECTION OF WASTE TREATMENT MANAGEMENT PROGRAM
17.1 POINT SOURCE TREATMENT WORKS
17.2 NON-POINT SOURCE TREATMENT WORKS
17.3 NON-STRUCTURAL PROCESSES
17.4 MANAGEMENT
17.5 IMPLEMENTATION RATIONALE AND PRIORITY CRITERIA
17.6 PRESENTATION OF COMPREHENSIVE PROGRAM
This selection would be based heavily on Section 16 investigations.
No significant additional modeling appears necessary.
10
-------
18.0 ENVIRONMENTAL, SOCIAL AND ECONOMIC IMPACT ASSESSMENT
18.1 ECOLOGICAL INVENTORY
18.2 EVALUATION OF PROPOSALS, STRATEGIES, ALTERNATIVES AND
PROGRAM
18.3 PREDICTED EFFECTIVENESS - IMPACT
18.4 CONSISTENCY WITH, IMPACT ON OR ADJUSTMENT REQUIRED FOR
RELATED PLANS
A wide variety of procedures are now being used for environmental
impact assessment, e.g., Dee et al., 1973. Techniquesof multiple
objective analysis can be used to evaluate incommensurables such as
dollars and environmental quality. Cohon (1973) presents an excel-
lent survey of available techniques. We have applied this approach
to the Upper St. Johns River Basin in Florida using natural energy
as a surrogate for environmental quality (Butkovich and Heaney,
1975). The results are presented in Figure 1. In multiple objective
analysis one replaces notion of optimality with determining a set of
undominated or non-inferior solutions. This set is displayed to the
decision makers so that they might select a reasonable compromise
solution.
19.0 PLAN PREPARATION AND COMPLETION
20.0 GRAPHICS AND PRINTING
21.0 COORDINATION ACTIVITIES - COMPLETED AND CERTIFIED 208 PLAN
Little modeling seems relevant to these last three phases of the
study.
CONCLUSIONS
The purpose of this discussion has been to present some very
preliminary ideas regarding how models can be used in the 208
planning process. Following the format of a preliminary work
plan for Orlando, Florida, various types of models were described
with a heavy bias towards models we have worked with or are familiar
to us. No attempt was made to survey all of the available simulation
and optimization models; rather, the purpose was to provide some
preliminary suggestions on how models might fit into the 208 planning
process. An important consideration in successful modeling is to
11
-------
a
a
to
Vj
ra
VO
O
C
CZ
a.
100J
90
80
70
60
50
40
30-
20
10
2020
COMPOSITE
TRANSFORMATION
CURVE
N/
\
ENERGY VALUE DF.
NATURAL SYSTEM
\
\1980
\
2020(l)__.f 2
\
\
- X-a 2000
MAX[MUM ENERGY
1980
1965
O
1965
162 163 164 165 166 167 168 169 170 171 172 173 174
ENERGY (1012 kcal/year)
FIGURE 1. TRANSFORMATION CURVES FOR VARIOUS LEVELS OF DEVELOPMENT, UPPER ST. JOHNS RIVER BASIN.
-------
interact frequently with the decision makers. In view of the
existing uncertainty in wet weather pollution, the problem defi-
nition phase of the analysis is of vital importance.
REFERENCES
Brandstetter, A., An Assessment of Mathematical Models for Storm
and Combined Sewer Management, Battelle Northwest, Richland,
Washington, 1975.
Brown, J. W., et al., Models and Methods Applicable to Corps of
Engineers Urban Studies, Misc. Paper H-74-8, US Army Engineers
Waterways Experiment Station, Vicksburg, Mississippi, 1974.
Bulkley, J. W. and R. T. McLaughlin, Simulation of Political
Interactions in Multiple-Purpose River Basin Development, Report
No. 100, Hydrodynamics Laboratory, Department of Civil Engineering,
Massachusetts Institute of Technology, Cambridge, Massachusetts,
1966.
Burke, R. B. and J. P. Heaney, Behavioral Approach to Water Resources
Management, Lexington Books, Lexington, Massachusetts, 1975.
Butkovich, G. B. and J. P. Heaney, Multi-Objective Analysis of the
Upper St. Johns Water Resources Project, ORSA/TIMS Meeting,
Chicago, Illinois, 1975.
Cohon, J. L., An Assessment of Multi-Objective Solution Techniques
for River Basin Planning Problems, R73-49. Department of Civil
Engineering, Massachusetts Institute of Technology, Cambridge,
Massachusetts, 1973.
Dee, N., et al., An Environmental Evaluation System for Water
Resource Planning, Water Resources Research. Vol. 9, No. 3, 1973.
Deininger, R. A. and D. P. Loucks, "Systems Approaches to Problems
of Water Pollution Control," in Chacko, G. K. (editor) Systems
Approach to Enyironmental Pollution, ORSA Health Applications
Section, Arlington, Virginia, 1972.
Heaney, J. P., W. C. Huber, et_al_., Urban Stormwater Management
Modeling and Decision-Making, US EPA Report EPA-670/2-75-022, 1975.
Hydrologic Engineering Center, Urban Storm Water Runoff, STORM,
US Army Corps of Engineers, Davis, California, 1975.
13
-------
James, L. D. and R. R. Lee, Economics of Water Resources Planning,
McGraw-Hill Book Company, New York, New York, 1971.
Kilbridge, M. D., O'Block, R. P. and P. V. Teplitz, Urban Analysis.
Harvard University Press, 1970.
Metcalf and Eddy, University of Florida, Water Resources Engineers,
Storm Water Management Model
•
a. Volume 1, Final Report, US EPA Report No. 11024DOC07/71,
b. Volume 2, Verification and Testing, US EPA Report No.
11024DOC08/71,
c. Volume 3, User's Manual, US EPA Report No. 11024DOC09/71,
d. Volume 4, Program Listing, US EPA Report No. 11024DOC10/71.
Metcalf and Eddy, Urban Stormwater Management and Technology: An
Assessment, US EPA Report No. EPA-660/2-74-040, 1974.
Murphy, M. P., Economics of Urban Storm Water Quality Management,
ME Thesis, Department of Environmental Engineering Sciences,
University of Florida, Gainesville, 1975.
Perez, A. I., Huber, W. C., Heaney, J. P. and E. E. Pyatt, "A
Water Quality Model for a Conjunctive Surface-Groundwater System,"
US EPA Report No. EPA-600/5-74-013, 1974.
Wanielista, M. P. (editor), Proceedings: Storm-Water Management
Workshop, held at the University of Florida, Gainesville,
February 26-27, 1975.
14
-------
METHODOLOGY FOR EVALUATING THE COST OF
URBAN STORMWATER QUALITY MANAGEMENT
By
James P. Heaney and Sheikh M. Hasan*
INTRODUCTION
Traditionally, cities have utilized a single sewer network (called combine
sewers) to carry away domestic wastes and urban stormwater. Treatment facilities
were installed with capacity sufficient to treat two to five times the average
rate of dry weather sewage flow. If the rainfall was larger than that treat-
ment capacity, the excess flow containing a mixture of untreated sewage and
storm water was bypassed directly to the receiving water.
Following World War II, most cities installed separate sewer systems
in newly developing areas. Also, programs were initiated to separate
existing combined sewer systems. However, preliminary cost estimates
indicated that this separation program would be very expensive. In
addition, stormwater quality sampling programs indicated that this water
contains significant quantities of pollutants which it picks up as it
moves through the urban area. Thus, the current thinking indicates that
remedial measures to control combined sewer overflows are more cost-
effective than sewer separation.
During the same period, cities installed improved sewage treatment
facilities capable of removing about 85 to 90 percent of the bio-
chemical oxygen demand (BOD). During the late sixties and early
seventies, numerous cities were investigating tertiary treatment
of dry weather sewage. However, tertiary treatment is quite expensive.
Some people argued that it may be more cost-effective to treat combined
sewer overflows and/or urban stormwater. This question remains unresolved.
*
Associate Professor and Research Assistant, respectively, Department of
Environmental Engineering Sciences, University of Florida, Gainesville.
15
-------
Dry weather sewage flows are continuous and relatively small in
volume and the technology for their control is fairly well estab-
lished. On the other hand, the wet weather problem involves control
of massive and dynamically variable "wet weather" periods of short
duration. The quantity and quality of these flows depend upon many
factors such as the degree of urbanization, amount of rainfall, size
of the area, type of land use, soil types, etc. Thus, control of
wet weather flows involves both quantity and quality. Therefore,
the three facets of urban water management to be examined are con-
trol of:
1. dry weather sewage flow,
2. stormwater quality, and
3. stormwater quantity.
Since the mid-sixties, the federal government has played an in-
creasing role in the management of urban water quality. The
states have been required to establish water quality standards
and federal subsidies have been provided for the construction of
publically owned waste treatment works. The 1972 Amendments to
the Federal Water Pollution Control Act establish the year 1983 as
the goal for zero discharge. Towards achieving this goal, compre-
hensive guidelines have been issued which require planning on an
area wide basis, consideration of dry weather as well as wet weather
pollution sources and the installation of the best practicable
wastewater treatment technology for publically owned treatment
works.
Approximately two years ago, the American Public Works Association
and our group at the University of Florida initiated an EPA spon-
sored study to estimate the nationwide cost of treating combined
sewer overflows and stormwater runoff. Results from this effort
will be available later this year. This lecture describes the pro-
cedures used to conduct this study. Emphasis is placed on the
interrelationships among stormwater quality control and the more
established programs in wastewater control and storm drainage con-
trol.
The next section describes procedures for estimating the costs of
stormwater quality control. Then, the net costs for stormwater
quality management are determined by recognizing the complementari-
ties among the various purposes. N-person game theoretic concepts
are used for cost allocation purposes. Minneapolis, Minn, is used
as the test city.
16
-------
CONTROL TECHNOLOGY AND ASSOCIATED COSTS
A wide variety of control alternatives are available for improving
the quality of wet-weather flows (Field and Struzeski, 1972; Lager
and Smith, 1974; Becker elt a±., 1973). Rooftop and parking lot
storage, surface and underground tanks and storage in treatment units
are the flow attenuation control alternatives. Wet-weather quality
control alternatives can be subdivided into two categories: primary
devices and secondary devices. Primary devices take advantage of
physical processes such as screening, settling and flotation. Sec-
ondary devices take advantage of biological processes and physical-
chemical processes. These control devices are suitable for treating
stormwater runoff as well as combined sewer overflows. However, the
contact stabilization process is feasible only if the domestic waste-
water facility is of an activated sludge type. The quantities of
wet-weather flows that can be treated by this process are limited by
the amount of excess activated sludge available from the dry-weather
plant. At the present time, there are several installations through-
out the country designed to evaluate the effectiveness of various
primary and secondary devices. A summary of the design criteria and
performance of these devices is presented in Table 1. Based on these
data, the representative performance of primary devices is assumed
to be 40 percent BOD5 removal efficiency and that of secondary devices
to be 85 percent BOD5 removal efficiency. No treatment is assumed
to occur in storage. Hasan (1976) has synthesized available informa-
tion regarding stormwater pollution control costs. The results are
shown in Table 2.
COSTS OF URBAN STORMWATER QUALITY MANAGEMENT
The evaluation procedure for the nationwide assessment consisted of
relatively detailed studies of five cities: Atlanta, Denver, Minne-
apolis, San Francisco, and Washington, D.C. For each city, a single
storm event for a selected catchment was simulated using the EPA
Storm Water Management Model - SWMM (EPA, 1971a, b, c, d; 1975).
Also, one year of hourly precipitation, runoff, and discharage rates
were estimated using the HEC STORM model (HEC, 1974). STORM provides
estimates of the total volume of storm water which is treated for
a specified size of storage unit and treatment rate. Numerous com-
binations were tested for each of the five cities to derive storage-
treatment isoquants as shown in Figure 1 for Minneapolis. Given the
storage-treatment isoquants and knowning the relative costs of storage
and treatment, one can determine an optimal expansion path in terms
of control costs versus percent BOD removal. The optimal expansion
17
-------
TABLE 1. WET-WEATHER TREATMENT PLANT PERFORMANCE DATA
Device
Control Alternatives
Design Criteria
Reported
BOD5 Removal
Efficiency, n
Primary
Secondary
Swirl Concentrator8'
Micros trainer
Dissolved Air Flotation
w/ Chemical Addition*1
Sedimentation
Representative Performance
Contact Stabilization
Physical-Chemical8
Representative Performance
60.0 gpm/sq ft
20.0 gpm/sq ft
2.5 gpm/sq ft
0.5 gpm/sq ft
Cont 0.25 hrs
Stab 3.0 hrs
3.0 hrs
0.25 - 0.50
0.40 - 0.60
0.50 - 0.60
0.25 - 0.40
0.40
0.75 - 0.88
0.85 - 0.95
0.85
SField and Moffa, 1975
bAPWA, 1974
Wier, 1974
Lager and Smith, 1974
Performance data based on domestic wastewater treatment
Agnew, R. et al., 1975
^Estimate baaed on performance of these .units for domestic wastewater
18
-------
TABLE 2. COST FUNCTIONS FOR WET-WEATHER CONTROL DEVICES**b>1
Annual Cost: $/yr
Device
Primary
Control Alternative
C. u ft
Swirl Concentrator ' '
Microstrainer6* *
Dissolved Air Flotation
Sedimentation6
Amortized Capital Operation and Maintenance
CA - lTm
or lSm OM - pTq
Imp q
'l, 971.0
7,343.8
8,161.4
32,634.7
0.
0.
0.
0.
70
76
84
70
Representative Primary Device Total Annual Cost =
Secondary
Contact Stabilization8
Physical-Chemical6
19.585.7
32,634.7
Representative Secondary Device Total
Storage
High Density (15 per/ac)
Low Density (5 per/ac)
Parking Loth
Rooftop11
Representative Annual Storage
T = Wet-Weather Treatment Rate in mgd;
51,000.0
10,200.0
10,200.0
5,100.0
Costj ($
0.85
0.85
Annual Cost
1.
1.
1.
1.
00
00
00
00
per ac-in.
S = Storage Volume
aENR - 2200. Includes land costs, chlorination,
sludge
584.0 0.70
1,836.0 0.76
2,036.7 0.84
8,157.8 0.70
2
9
10
40
w
,555
,179
,198
,792
Total
TC - wTZ
or wS
.0
.8
.1
.5
0
0
0
0
z
.70
.76
.84
.70
$3,000 per mgd
4,894
8,157
.7 0.85
.8 0.85
24
40
,480
,792
.4
.5
0
0
.85
.85
= $15,000 per mgd
) = $122 e°*
in mg
°Agnew et
16 (PD)
al., 1975.
51
10
10
5
,000
,200
,200
,000
.0
.0
.0
.0
1
1
1
1
.00
.00
.00
.00
handling, engineering and contingencies.
Sludge handling costs based on data from Battelle North-
west, 1974.
cField and Moffa, 1975.
d
Senjes, et al.. 1975.
eLager and Smith, 1974.
fMaher, 1974.
j
and Robbins, 1975.
For T <^ 100 mgd. No economies of scale beyond
100 mgd.
PD = gross population density, persons/acre.
-------
.00
.90
T, cm/hr
.01 .02
T, cm/hr
.80-
.70-
.60-
.50-
.000 ,004 .008 .012 .016 .020
i
.15-
,000 .002 ,004 .006
T, in/hr
.000
.010
TREATMENT,!, in/hr
.020
008
ANNUAL RUNOFF - 10.50 in
.00
.030
Figure 1. Storage-Treatment Isoquants for Various BOD
Control Levels - Region III - Minneapolis
20
-------
path is determined by comparing the costs of the various alterna-
tives or
where CT = unit cogst of treatment ($/mgd) t
c = unit cost of storage*($/mg), and
MRS__ = marginal rate of substitution of storage for treat-
o 1 ._
ment.
The above problem can be expressed in the more compact mathematical
form shown below:
minimize
Z = cs(S) + cT(T)
subject to
.T) - 0 (2)
R1»S»T L °
where Z - total control costs,
Cg(S) = storage costs,
c_(T) = treatment costs,
S = storage volume,
T = treatment rate,
Ri= percent pollutant control, and
21
-------
f(R-;S,T) = production function relating the level of
pollution control attainable with specified
availabilities of storage (S) and treatment
(T).
Mathematical Representation of Isoquants
The storage/treatment isoquants are of the form:
(T2 - T1)e"KS (3)
where T = wet-weather treatment rate, inches per hour,
T. = treatment rate at which isoquant becomes asymptotic to
the ordinate, inches per hour,
T- = treatment rate at which isoquant intersects the abscissa,
inches per hour,
S ™ storage volume, inches, and
K = constant, inch
The value of T- occurs at a relatively high storage capacity in com-
bination with a low treatment rate such that the treatment plant
operates continuously. T- can be found as follows:
Ti • stto • •*
where AR = annual runoff, inches per year, and
R « percent runoff control.
By relating the parameters T-, T--T- and K to the level of control R,
one equation was developed for Minneapolis. The T2-T- and K terms ver-
sus R were found to be:
T2 - T! - .00193e-100894R (5)
K = 492.0e-'0319576R (6)
22
-------
Based on this analysis, the following general equation for the isoquants
is obtained:
.100894R-(492.0e
-.0319576R
)S
T = .0000345R + .00193e"""""n~ v"~'w" ' (7)
Adjusting Isoquants for Treatment Efficiencies
The above isoquants equation (7) characterizes the percentage of total
runoff that passes through "treatment," If the concentration of
pollutants is constant and "treatment" efficiency, ri, is 1.0 then per-
cent runoff control is synonymous with percent pollutant control. Ob-
viously, this is not the case. Thus, these results need to be refined
to account for treatment efficiency.
Recall that R is the percent runoff control. Let r\ equal treatment
plant efficiency. If RI denotes the percent pollutant control, then
to realize RI, one needs to process R]_/n of the runoff. Note that RI
may be percent BOD removal, percent SS removal, etc. In Table 1,
representative treatment efficiencies, in terms of BODs removal, were
derived for primary and secondary devices. Thus, if one desires 25
percent BODs removal with a primary device, then 62.5 percent of the
runoff volume must be processed whereas only 29.4 percent of the run-
off needs to be processed if a secondary device is selected. Thus,
to convert percent runoff control isoquants to percent pollutant
control isoquants, one simply uses
Rl
R =
in the isoquant equation (7).
Wet-Weather Quality Control Optimization
The wet weather optimization problem, assuming linear costs, may be
stated as follows:
minimize Z - cgS +c_T (9)
subject to
23
-------
(T2-T1)e
T,S > 0
This constrained optimization problem can be solved by the method of
Lagrange multipliers to yield
S* - max
s
(10)
I 0
and
T* . TI + (T2-T1)e"KS* (11)
Note that T* is expressed as a function of S*, so it is necessary to
find S* first. Knowing S* and T*, the optimal solution is
Z* - cgS* + cTT*. (12)
Data needed to estimate T., T. and K have already been presented.
For a primary device, CT - $3,000/mgd - $l>960/acr^ln
and n " 0.40.
flcrG—in
For a secondary device, CT - $15,000/mgd - $9,800/ ——
For storage cost,
cq($/acre-in) - 122 e0'16(PD) (13)
D
where FD - population density in persons per acre.
24
-------
The above optimization procedure was programmed to generate curves
(Figure 2) showing percent pollutant removed versus total annual
costs for primary and secondary treatment in conjunction with storage.
Note that, for wet-weather control, marginal costs are increasing
because of the disproportionately large sized control units needed
to capture the less frequent larger runoff volumes. The curves shown in
Figure 2 can be approximated by functions of the form:
Z = ke (14)
where Z « total annual cost, dollars per year
k, n = parameters
R, « percent pollutant removal. 0 < R. < R. and
j, —. J. ~~ J.
R. • maximum percent pollutant removal,
The results for primary and secondary levels are given below.
Unit Range k^ ri
Primary, C 0 < R, < 40 2.39 0.118
p — i —
Secondary, C 0 <_ R. <_ 85 4.98 0.049
S . X """""
The level of control at which one is indifferent between primary and
secondary levels of control in conjunction with storage i.e. C - C
is 10.6 percent. P s
POTENTIAL SAVINGS DUE TO MULTIPURPOSE PLANNING
The cost of wet weather quality control can be reduced by inte-
grating this purpose with dry weather treatment plants and/or storage
reservoirs for stormwater quantity control. Dry weather sewage treatment
25
-------
ro
i l70'
o
-I
56.775
0.800 p-
.
t
9 P 3 I
i
P
P
*$P 3
-
!
i
i
.1
p 1
M
1
: P
5
if
S S
!
p
p
,
p
1
p
1
:
i
••
s
s
'
;
8
: s
•
3
3
2-.
:
,
1
S
[ S
i
r !
'
i
9
8
1
;
rs
'
i
1
|
> |
i
0*0 JO.'OOO ..?0.*000 10.000 00,000 ^0.000 60,000 70,000 80.000 90.000 100,000
PERCENT POLLUTANT CONTROL, R,
FIGURE 2. CONTROL COSTS FOR PRIMARY AND SECONDARY UNITS AS A FUNCTION
OF PERCENT BOD REMOVAL.
-------
plants are designed to handle the peak flow anticipated 10 to 15
years hence. The full capacity of these plants is seldom utilized
because peak flows occur infrequently and also because additional
capacity is frequently added before the actual flow approaches the
design capacity of the plant. Utilization of this excess capacity
can reduce the treatment capacity needed for wet weather quality
control. Similarly, utilization of storage available for wet weather
quantity control can result in reducing the storage and treatment
requirements for wet; weather quality control. A rough estimate of
what the potential savings by integrating might be can be made as
follows:
Let Z = cost of dry weather control using a secondary
treatment plant (Hasan, 1976), dollars per year
= amortized capital costs + annual 0 & M costs,
118,300(D+E)'7715 + 54.800D'8345 (15)
D = actual sewage flow, mgd, and
E » excess capacity, mgd.
Assuming D = 10 mgd, E = 10 mgd, and a per capita sewage flow of
100 gallons,
Z1 - 15.67 PD (16)
where PD = population density, persons per acre.
Also,
Z. = annual cost of wet weather quality control ($/acre)
cgS* + cT[T1+(T2-T1)e ] (17)
where all terms are as defined earlier;
27
-------
and
Z, = annual cost of wet weather quantity control ,($/acre)
= cgV (18)
where V = storage volume required for wet weather quantity control
(inches)
= CR'(i) (19)
CR' = runoff coefficent in developed state - runoff coefficient
in undeveloped state; and
i = 24 hour rainfall for design frequency (inches).
If dry weather quality control and wet weather quality control are
integrated, then
Z _ = joint cost of two purposes, dollars per year
= 15.67 PD + cEE + cgS* + ^[(^ - E) + (T2-T1)e~KS*]
(20)
where E = excess capacity available at dry weather plant, assumed
to be 10 mgd or (1.474 x 10~4) PD inches per hour
c = assumed annual cost of treating E at dry weather plant
= $l,960/inch/hr
If wet weather quality control is integrated with wet weather quantity
control, then
28
-------
Z . = joint cost of two purposes, $/yr
-KS*
rV
|- In [^ K (Tg-T^] (22)
LO
where S* = max
It is assumed that dry weather control cannot be integrated with wet
weather quantity control. Therefore,
Z = joint cost of two purposes, $/yr
Z3 (23)
If all three purposes are integrated, then
Z.__ = joint cost of three purposes, $/yr
15.67 PD + c£E + cgS*
(24)
To determine the potential savings due to integrating the three pur-
poses, it is first necessary to apportion the joint costs among
various purposes. For cost allocation, the Use of Facility method can-
not be applied in this case because there is no single facility-
utilization parameter that is common to all three purposes. There-
fore, some other method must be devised. One may use the game-
theoretic procedure (Heaney et al., 1975). The method presented
below (called Separable Cost Remaining Benefits Method) yields equiva-
lent results.
Let <)> » cost apportioned to storm water quality control
7 _
Then * = (Z - Z13) +
(25)
29
-------
then the potential savings are Z ~
The above procedure was utilized to develop curves for various
frequency storms as shown in Figure 3 for Minneapolis.
Cost allocation is seen to have a very significant impact on the costs
attributable to wet weather quality control. In the lower levels of
control, joint utilization with the dry weather facility yields sig-
nificant savings. Sharing storage permits significant savings up
to much higher levels of pollutant control.
CONCLUSIONS
This study describes the methodology used to assess the cost of con-
trolling stormwater pollution. Cost data are combined with storage-
treatment isoquants to determine an efficient expansion path for con-
trolling varying degrees of wet weather control. It is possible to
reduce the costs of stormwater pollution control by integrating this
purpose with existing sewage treatment plants and storage facilities
for storm drainage systems. Sample results were presented for
Minneapolis, Minnesota.
30
-------
2 YEAR (E=V = 0)
u>
2 YEAR (E ONLY)
I YEAR(E AND V)
2 YEAR(V ONLY)
100 YEAR(E+V)
2 YEAR (E 4V)
E= AVAILABLE EXCESS CAPACITY IN DRY WEATHER PLANT
V= AVAILABLE STORAGE VOLUME IN STORM WATER QUANTITY CONTROL FACILITY
50 60
% POLLUTANT CONTROL, R,
100
FIGURE 3. EFFECT OF DESIGN STORM AND NUMBER OF PURPOSES ON COST ALLOCATION
FACTOR FOR VARIOUS LEVELS OF CONTROL.
-------
REFERENCES
American Public Works Assn,, The Swirl Concentrator as a Grit Separator
Device, 11023 GSC, USEPA, 1974,
Agnew, R. W., jat al., "Biological Treatment of Combined Sewer Overflow
at Kenosha, Wisconsin, USEPA Report EPA-670-275-019, NTIS-PE
242 120/AS, 1975.
Becker, B. C., et al,, Approaches to Stormwater Management, Hittman and
Assoc., U.S.D.I. Contract 14-31-001-9025, 1973.
Benjes, H,, et al., "Estimating Initial Investment Costs and Operation
and Maintenance Requirements of Stormwater Treatment Process,
USEPA Cont. EPA-68-03-2186, (unpublished), 1975.
Environmental Protection Agency, "Storm Water Management Model,"
Water Pollution Control Research Series, Washington, D.C. 1971
a. Volume I, "Final Report" No. 11024DOC07/71
b. Volume II, "Verification and Testing," No. 11024DOC08/71
c. Volume III, "User's Manual," No. 11024DOC09/71
d. Volume IV, "Program Listing," No. 11024DOC10/71
Environmental Protection Agency, "Urban Storm Water Management
Modelling and Decision Making," EPA-670/2-75-022, May 1975.
Field, R. A. and Struzeski, E. J, Jr., "Management and Control of
Combined Sewer Overflows," J, W, P. C. F., Vol. 44, No, 7,
pp. 1393-1415, 1973.
Field, R. I., and Moffa, P. E., Treatability Determinations for a
Prototype Swirl Combined Sewer Overflow Regulator/Solids-
Separator, IAWPR Workshop on Design-Operator Interactions at
Large Wastewater Treatment Plants, Vienna, Aus.,
September 1975.
Hasan, S., Integrated Strategies for Urban Water Quality Management,
Ph.D. Dissertation, University of Florida, Gainesville,
Florida, 1976.
•
Heaney, J. P., £jt a!L., Assessment of Combined Sewer Overflows,
Stormwater Discharges, and Non-Sewered Urban Runoff, USEPA
(unpublished), 1976.
Hydrologic Engineering Center, "Urban Stormwater Runoff: STORM,"
U.S. Army Corps of Engineers, Generalized Computer Program
723-58-L2520, 1975.
Lager, J. and Smith W., "Urban Stormwater Management and Technology:
An Assessment," USEPA Report EPA-670/2-74-040, NTIS-PB
240 697/AS, 1974.
32
-------
Maher, M. B., Microstraining and Disinfection of Combined Sewer
Overflows - Phase III, USEPA Report EPA-670/2-74-049, 1974.
Wiswall, K. C., and Robbins, J, C. Implications of On-Site Detention
in Urban Watersheds, ASCE Hyd, Div. Conf., Seattle, Washington.
33
-------
INTRODUCTION TO THE EPA STORM WATER MANAGEMENT MODEL (SWMM)
* A
By Wayne C. Huber and James P. Heaney
PROBLEMS OF URBAN RUNOFF
An enormous pollution load is placed on stream and other receiving
waters by combined and separate storm sewer overflows (Figure 1).
It has been estimated that the total pounds of pollutants (BOD and
suspended solids) contributed yearly to receiving waters by such
overflows is of the same order of magnitude as that released by all
secondary sewage treatment facilities (6,7). The Environmental
Protection Agency (EPA) has recognized this problem and lead and
coordinated efforts to develop and demonstrate pollution abatement
procedures. As shown in Figure 2, these procedures include not only
improved treatment and storage facilities, but also possibilities
for upstream abatement alternatives such as rooftop and parking lot
retention, increased infiltration, improved street sweeping, reten-
tion basins and catchbasin cleaning or removal (5,6). The complexi-
ties and costs of proposed abatement procedures require much time and
effort to be expended by municipalities and others charged with
decision making for the solution of these problems.
It was recognized that an invaluable tool for decision makers would
be a comprehensive mathematical computer simulation program that
would accurately model quantity (flows) and quality (concentrations)
during the total urban rainfall-runoff process. This model would not
only provide an accurate representation of the physical system, but
also provide an opportunity to determine the effect of proposed
pollution abatement procedures. Alternatives could then be tested
on the model, and least cost solutions could be developed.
The resulting EPA Storm Water Management Model is introduced below.
However, since its initial release in 1970, there has been an
insurgence of urban runoff modeling, and it is worthwhile to review
briefly objectives and options pertinent to management of urban
stormwater runoff.
*Associate Professors, Department of Environmental Engineering Sciences,
University of Florida/Gainesville.
34
-------
OWP TMATKCNT MCILITY
Note; DWP sources actually overlay the subcatchments. They are
separated in the figure only to simplify the representation,
-Figure 1. SCHEMATIC SYSTEM DRAWING RAINFALL THROUGH OVERFLOW.
35
-------
EXCESS FLOW
TREATMENT
FLOW INFILTRATION
DIVERTER
INTERNAL
IN-SYSTEM
STORAGE
-TREATED
EXCESS FLOW
DISCHARGE
Figure 2. TYPICAL STORAGE-TREATMENT APPLICATIONS TO LIMIT
UNTREATED OVERFLOWS.
36
-------
URBAN RUNOFF MODELS
Objectives
The overall objective of urban runoff modeling is to aid in.
decision making for control of quantity and quality problems.
Within this broad objective, three sub-objectives may be identi-
fied: planning, design and operation. Each objective typically
produces models with somewhat different characteristics and the
different models overlap to some degree. A more complete descrip-
tion of objectives, along with a description of several current
models is given by Huber (10).
.Planning Models
Planning models are used for an overall assessment of the urban
runoff problem as well as estimates of the effectiveness and costs
of abatement procedures. They may be used for "first cut" analyses
of the rainfall-runoff process and illustrate trade-offs among
various control options, e.g., treatment versus storage. They are
typified by relatively large time steps (hours) and long simulation
times (months and years). Data requirements are kept to a minimum
and their mathematical complexity is low.
A current example of such a model is the STORM Model (12,21) developed
by the Corps of Engineers Hydrologic Engineering Center (HEC) and
Water Resources Engineers, Inc. (WRE) for the City of San Francisco.
It utilizes hourly time steps and precipitation inputs and has simple
quantity and quality prediction procedures based on such parameters
as percent imperviousness and land use. Included are the effects
of snow melt and soil erosion as well as treatment and storage options.
The output may be used to illustrate, for example, the frequency and/or
volumes of discharges to receiving waters of untreated urban runoff
for a given treatment-storage combination. STORM has been run for
simulation periods of 25 years and longer, depending upon the desired
definition of return periods.
A planning model such as STORM may also be run to identify hydrologic
events that may be of special interest for design or other purposes.
These storm events may then be analyzed in detail using a more
sophisticated design model. Planning or long-term models may also be
used to generate initial conditions (i.e., antecedent conditions) for
input to design models. Although not originally developed as such,
as a result of recent modifications (23) the SWMM may also be used as
a planning model. More details are presented later.
37
-------
Design Models
Design models are oriented toward the detailed simulation of a
single storm event. They provide a complete description of flow
and pollutant routing from the point of rainfall through the
entire urban runoff system and often into the receiving waters as
well. Such models may be used for accurate predictions of flows
and concentrations anywhere in the rainfall-runoff system and can
illustrate the detailed and exact manner in which abatement proce-
dures or design options affect them. As such, these models are a
highly useful tool for determining least-cost abatement procedures
for both quantity and quality problems in urban areas. Design
models are generally used for simulation of a single storm event and
are typified by short time steps (minutes) and short simulation
times (hours). Data requirements may be moderate to very extensive
depending upon the particular model employed.
The EPA Storm Water Management Model (16,17,18,19), frequently
abbreviated "SWMM", is an example of a model developed specifically
for simulation of urban quantity and quality processes and useful
for the purposes mentioned above. It is also versatile enough to
be used for certain planning studies or adapted to uses other than
were originally intended. For instance, the surface runoff portion
may be used to simulate natural drainage systems, and the receiving
water portion may be applied to a variety of natural configurations
independent of the urban runoff context.
Many other urban runoff models have been described in the literature (10)
and are too numerous to enumerate here. Examples range from relatively
simple models, e.g., RKL (25), ILLUDAS (26), Chicago (14), to highly
complex models that utilize the complete dynamic equations of motion
to simulate every aspect of the drainage systems, e.g., the WRE version
of the SWMM (22), Hydrograph Volume Method (13), and Sogreah (24).
These latter three models, for instance, are probably the best
available for analysis of surcharging and backwater effects. However,
many of these other models lack quality calculations; of the afore-
mentioned ones, quality routing is included only in the WRE version
of the SWMM. Furthermore, many are either proprietary or ill-documented.
The EPA SWMM is well documented, widely tested and of a fairly high
level of sophistication. In addition, through its broad use improvements
and updating have been continuous. It is a widely accepted, detailed
simulation model.
Operational Models
Operational models are used to produce actual control decisions during
a storm event. Rainfall is entered from telemetered stations and the
model is used to predict system responses a short time Into the future.
Various control options may then be employed, e.g., in-system storage,
diversions, regulator settings.
38
-------
These models are frequently developed from sophisticated design
models and applied to a particular system. Examples are operational
models designed for Minneapolis-St. Paul (3) and Seattle (15).
DEVELOPMENT OF THE STORM WATER MANAGEMENT MODEL
Under the sponsorship of the Environmental Protection Agency, a con-
sortium of contractors — Metcalf and Eddy, Inc., the University of
Florida, and Water Resources Engineers, Inc. — developed in 1969-70
a comprehensive mathematical model capable of representing urban storm-
water runoff and combined sewer overflow phenomena. The SWMM portrays
correctional devices in the form of user-selected options for storage
and/or treatment with associated estimates of cost. Effectiveness is
portrayed by computed treatment efficiencies and modeled changes in
receiving water quality.
T-he project report is divided into four volumes. Volume I, the "Final
Report" (16), contains the background, justifications, judgments, and
assumptions used in the model development. It further includes
descriptions of unsuccessful modeling techniques that were attempted
and recommendations for forms of user teams to implement systems
analysis techniques most effectively. Although many modifications and
improvements have since been added to the SWMM, the material in Volume
I still accurately describes most of the theory behind updated versions.
Volume II, "Verification and Testing" (17), describes the methods and
results of the application of the original model to four urban
catchments.
Volume III, the "User's Manual" (18), contains program descriptions,
flow charts, instructions on data preparation and program usage, and
test examples. This volume has recently been revised (11).
Volume IV, "Program Listing" (19), lists "the entire original program
and JCL as used in the demonstration runs. Since many routines in the
updated version are similar or identical to the original, it is still
a useful reference.
All three original contractors have continued to modify and improve the
SWMM, as have numerous other users since its release. As a result, an
official "Release 2" of the SWMM has been made in August 1974. Although
it has been prepared for EPA by the University of Florida, it also
relies heavily upon contributions by Water Resources Engineers and
Metcalf and Eddy. The revised User's Manual (11) was written to accom-
pany Release 2. Several of the modifications incorporated in Release 2
also described in detail by Heaney and Huber et al. (8).
39
-------
OVERALL SWMM DESCRIPTION
The comprehensive Storm Water Management Model uses a high speed
digital computer to simulate real storm events on the basis of
rainfall (hyetograph) inputs and system (catchment, conveyance,
storage/treatment, and receiving water) characterization to predict
outcomes in the form of quantity and quality values. Single event
or continuous (e.g., yearly or longer) simulations may be performed,
the former for design and the latter for planning purposes.
The simulation technique — that is, the representation of the
physical systems identifiable within the Model — was selected
because it permits relatively easy interpretation, and because it
permits the location of remedial devices (such as a storage tank
or relief lines) and/or denotes localized problems (such as flooding)
at a great number of points in the physical system.
Since the program objectives are particularly directed toward complete
time and spatial effects, as opposed to simple maxima (such as the
rational formula approach) or only gross effects (such as total
pounds of pollutant discharged in given storm), it is considered
essential to work with continuous curves (magnitude versus time),
referred to as hydrographs and "pollutographs". The units selected
for quality representation, pounds per minute, identify the mass
releases in a single term. Concentrations are also printed out within
the program for comparisons with measured data.
An overview of the Model structure is shown in Figure 3. In simplest
terms the program is built up as follows:
1) The input sources:
RUNOFF generates surface runoff using a
kinematic wave approach based on arbitrary
rainfall hyetographs, antecedent condi-
tions, land use, and topography.
FILTH generates dry weather sanitary flow
based on land use, population density, and
other factors.
INFIL generates infiltration into the
sewer system based on available groundwater
and sewer conditions.
-------
RUNOFF
(RUNOFF)
DECAY
(QUAD
INFILTRATION
(INFIL)
DRY WEATHER
FLOW
(FILTH)
TRANSPORT
(TRANS)
INTERNAL
STORAGE
(TSTROT)
COST
(TSTCST)
EXTERNAL
STORAGE
(8TOMAO)
TREATMENT
(TRCAT)
COST
(TRC08T)
RECEIVING WATER
(RECEIV)
Not*i Subroutine MUMS are •hown in p*r«nth*Mi<
FIGURE 3. OVERVIEW OF MODEL STRUCTURE.
INPUT
> SOURCES
CENTRAL
> CORE
CORRECTIONAL
> DEVICES
EFFECT
41
-------
2) The central core:
TRANS carries and combines the inputs
through the sewer system using a modified
kinematic wave approach in accordance
with Manning's equation and continuity;
it assumes complete mixing at various inlet
points.
3) The correctional devices:
TSTRDT. TSTCST. STORAG. TREAT, and TRCOST
modify hydrographs and pollutographs at
selected points in the sewer system,
accounting for retentipn time, treatment
efficiency, and other parameters; associ-
ated costs are computed also.
4) The effect (receiving waters):
RECEIV routes hydrographs and pollutographs
through the receiving waters, which may con-
sist of a stream, river, lake, estuary, or
bay.
The quality constituents considered for simulation within the total
program are 5-day BOD, total suspended solids, and total coliforms
(represented as a conservative pollutant). These constituents were
selected on the basis of available supporting data and importance
in treatment effectiveness evaluation. In addition, the Runoff
Block also models COD, settleable solids, total nitrogen, phosphate
and grease. However, routing of these parameters through subsequent
blocks presently involves special programming efforts. The contri-
bution of suspended solids by urban erosion processes is also
simulated by the program..
PROGRAM BLOCKS
Executive Block
The adopted programming arrangement consists of a main control and
service block, the Executive Block, a service block (Combine), and
four computational blocks: (1) Runoff Block, (2) Transport Block,
(3) Storage Block, and (4) Receiving Water Block.
42
-------
The Executive Block assigns logical units (disk/tape/drum), deter-
mines the block or sequence of blocks to be executed, and, on call,
produces graphs of selected results on the line printer. Thus,
this Block does no computation as such, while each of the other
four blocks are set up to carry through a major step in the quantity
and quality computations. All access to the computational blocks
and transfers between them must pass through subroutine MAIN of the
Executive Block, Transfers are accomplished on off-line devices
(disk/tape/drum) which may be saved for multiple trials or permanent
record.
Combine Block
This block allows the manipulation of data sets (files stored on
off-linedevices) in order to aggregate results of previous runs
for input into subsequent blocks. In this manner large, complex
drainage systems may be partitioned for simulation in smaller
segments.
Runoff Block
Quantity Techniques —
Runoff hyetographs are generated in two phases: overland flow and
small gutter/pipes. Both phases utilize a kinematic wave formulation
through the use of Manning's equation and continuity. Gutter/pipes
may.be arranged in a tree-like network with inflows from upstream
gutters and/or surface subcatchments. Outflows are typically to inlet
manholes of the Transport Block, although output may be transferred
directly to any "downstream" block.
Infiltration from pervious areas is accomplished through the inte-
grated form of Hprton's equation. For continuous simulation (23),
infiltration capacity is recovered during periods of no runoff,
as is detention storage through evaporation.
Data inputs include physical parameters for each subcatchment and/or
gutter/pipe (e.g., area, imperviousness, slope, roughness, etc.) and
rainfall hyetographs. The latter may be input at short time intervals
(e.g., five minutes) for single event simulation for design purposes
or at long time intervals (e.g., one hour from US Weather Service
tapes) for long term, continuous simulation.
A3
-------
Quality Techniques —
Surface "washoff" quality is generated by assuming a linear build-up
of dust and dirt as a function of land use, dry days and street
sweeping practices. Available pollutant loads (i.e., BOD, suspended
solids, total coliforms, settleable solids, COD, N, PO, and grease)
are determined as fractions of the dust and dirt load. Values for
these parameters (dust and dirt loads and pollutant fractions) are
generally based on data from Chicago (1) but should be altered to
reflect local conditions if known. For continuous simulation,
surface loads are recharged during periods of no runoff.
Surface pollutographs are generated assuming an exponential erosion
of available pollutant loads. The decay coefficient is a function
of the runoff rate. Pollutographs are routed through gutter/pipes,
if used, assuming complete mixing within each conduit. Catchbasins
contribute BOD and COD in a similar exponential fashion.
Urban erosion is simulated using the Universal Soil Loss Equation (8),
Data required for quality computations include land use definition
for each subcatchment (i.e., single family residential, multiple
family residential, commercial, industrial and open or park), catch-
basin information, curb lengths and soil loss parameters.
Other Information —
Runoff may be run independently of other SWMM blocks to provide
"first cut" predictions of quantity and quality parameters for urban
catchments. However, more useful results may be generated by linking
Runoff directly to Storage/Treatment. The model may then be used in
a planning mode with output similar to that of STORM (12,21). This
will be discussed later.
Discretization of the urban catchment may be performed in a detailed
manner (i.e., many subcatchments and conduits) if necessary. However,
recent results indicate that a very coarse discretization (e.g., one
subcatchment, no conduits) may produce hydrographs and pollutographs
at the outlet that are entirely comparable to those of a detailed
schematization (23). Obviously, considerable work may be avoided under
the latter option.
Transport Block
Quantity Techniques —
The primary flow and quality routing through the drainage system is
accomplished in this block. In addition, dry weather flow (DWF) and
infiltration are generated, and internal storage units may be included.
44
-------
To categorize a sewer system conveniently prior to flow routing,
each component of the system is classified as a certain type of
"element." All elements in combination form a conceptual represen-
tation of the system in a manner similar to that of links and nodes.
Elements may be conduits, manholes, lift stations, overflow struc-
tures, or any other component of a real system. Conduits themselves
may be of different element types depending upon their geometrical
cross-section (e.g., circular, rectangular, horseshoe, trapezoidal).
A sequencing is first performed to order the numbered elements for
computations. Flow routing then proceeds downstream through all
elements during each increment in time until the storm hydrographs
have been passed through t*he system.
The solution procedure follows a modified kinematic wave approach in
which disturbances are allowed to propagate only in the downstream
direction. As a consequence, backwater effects are not modeled
beyond the realm of a single conduit, and downstream conditions
(e.g., tide gates, diversion structures) will not affect upstream
computations. Systems that branch in the downstream direction can
be modeled using "flow divider" elements to the extent that overflows,
etc., are not affected by backwater conditions. Surcharging is modeled
simply by storing excess flows (over and above the full-flow conduit
capacity) at the upstream manhole until capacity exists to accept the
stored volume. Pressure-flow conditions are not explicitly modeled and
no attempt is made to determine if ground surface flooding exists.
However, a message is printed at each time step for each location at
which surcharging occurs. The Transport routine has proven its ability
to model accurately flows in most sewer systems, within the limitations
discussed above, and as such it should be more than adequate for most
applications. However, it will not accurately simulate systems with
extensive interconnections or loops, systems that exhibit flow reversals
or significant backwater effects, or systems in which surcharging must
be treated as a pressure-flow phenomenon. These problems should be
overcome with the inclusion in 1975 of the WRE transport routine (22)
as an option in the Transport Block. This routine utilizes the complete,
coupled equations of motion and simulates surcharge and pressure-flow
conditions as well.
A present option in the program is the use of the internal storage model
which acts as a transport element. The model provides the possibility
of storage-routing of the storm at one or two separate points within the
sewer system (restricted by computer core capacity). The program routes
the flow through the storage unit for each time-step based on the
continuity equation in a manner analogous to flood routing through a
reservoir. Extensive backwater conditions may thus be modeled by treating
portions of the sewer system as a storage unit with a horizontal water
surface.
45
-------
Another program option is a modest hydraulic design capacity (8).
If requested, the program will size conduits such that surcharging
is avoided. Although the entire system will then have adequate
hydraulic capacity to pass the input hydrograph, this is sometimes
accomplished through the loss of considerable in-system storage and
loss of consequent attenuation of the downstream peak flow.
Data input consists of physical parameters of the drainage system
(e.g., shapes, dimensions, slopes, roughnesses), linkage information,
and data descriptive of special elements (e.g., flow dividers, storage
units). The program time step must be the same as that of the
Runoff Block (if Runoff is used to generate input).
Quality Techniques —
Pollutants are routed through the transport system by means of complete
mixing within each element. They may be introduced into the system by
four means:
1) Storm-generated pollutographs (and hydro-
graphs) may be computed by the Runoff Block
and transferred on tape/disk devices to enter
the system at designated inlet manholes.
2) Pollutographs (and hydrographs) may be input
from cards or generated by previous runs of the
Transport or Storage/Treatment Blocks.
3) Residual bottom sediment in the pipes may be
resuspended due to the flushing action of the
storm flows. In this manner, a "first flush"
effect is simulated.
4) For combined systems, DWF pollutographs
are also entered at designated inlet manholes.
The routing of the pollutants is then done for each time step. The
maximum number of contaminants that can be routed is four, although
suspended solids, BOD and coliforms are the only ones commonly input
from Runoff.
Dry Weather Flow and Infiltration •*•*
Average DWF quantity and quality are generated on the basis of land
use, population density and other demographic parameters. Measured
flows (e.g., at a DWF treatment plant) are used to calibrate the
prediction. Hourly and daily correction factors are used to modify
predicted average values.
46
-------
Total infiltration at the system outlet must be input by the user.
The program then apportions this total throughout the system on the
basis of conduit lengths and joint distances. If measurements are
unavailable, advice is given in the User's Manual (11) on infil-
tration prediction.
Other Information —
Obviously, the Transport Block may be used for separate systems by
not including DWF or infiltration. Overbank flow in a natural
channel may be simulated using certain types of flow dividers. Dis-
cretization may be extensive or simple as required. However, for
some SWMM runs, especially those for planning purposes, Transport
may be conveniently omitted entirely.
Storage/Treatment Block
Storage Routing Techniques —
Quantity — Flow routing through storage units is accomplished using
simple level-pool, storage routing techniques (9, p. 356). Geometries
may be regular or irregular with alternative inlet and outlet controls
such as by weir, orifice or pumping. Only the storage unit itself
possesses storage characteristics. Storage effects of treatment devices
such as sedimentation, swirl concentrator, dissolved air flotation, etc.,
are scheduled to be added to SWMM during 1975.
Quality — Removal of pollutants by sedimentation in storage is
computed as a function of detention time as shown in Figure 4. Deten-
tion time in turn is computed on the basis of either plug flow or com-
plete mixing (user specified), the latter probably being more appropriate
for short detention times and the former for long ones. Resuspension of
solids is not modelled.
Treatment Techniques —
Many treatment options are available in this block, as shown in Figure
5. Each option computes removal of BOD, suspended solids and total
coliforms as a function of flow rate, based on actual field data as
much as possible. Certain options require other design and operational
parameters as input data (e.g., particle size distribution, optional
addition of chemicals). In addition, DWF treatment may be simulated
prior to storage or treatment using constant removal functions, as
indicated in the overall storage/treatment schematization shown in
Figure 6.
47
-------
SUSPENDED SOLIDS
3 4
TIME , hr
FIGURE 4, POLLUTANT REMOVAL IN A STORAGE UNIT AS A FUNCTION OF
DETENTION TIME.
Source; Fair, G. M., J. C, Geyer and D. A, Okun, Water and
Wastewater Engineering, John Wiley and Sons, Inc.t
1968, p. 25-25.
48
-------
INFLOW
I
(BYPASS)
(01)
*
STORAGE
MODEL
PRECEDING
«
(02)
(BYPASS
Cl
-
OVERFLOW
DESIGN FLOW
.
(BYPASS) fill 1 RAF1 RLCWZ 1 "^ . l*^Wini
(12)
) ' ' • •- - -
* OVERFLOW
*"» ~*|OESIGN FLOW {
HIGH-RATE ravoAcel on 1 INLET O?\
DISINFECTION IBYRASS) (Zl) | puMp|N6 W
t *
1*— CHEM
ir
fauuA^^i jjui IK i i (FINE SCRE^?I
DISSOLVED AIR (34)
FLOTATION |
4 • ,,• • • - -
|* CHEM
t t
(BYPASS) (4.) STMRA«?E"RS «« "FOERS^ «3)
A *
1.
(BYPASS) (5,) | EFTlgW (6a
' i
.1 '
*
(BYPASS) (6.) | «jgg (62)
*
CHEM
.., '• ' • \
(BYPASS, (7.) CONTACT {?2) JJjjjHj
"4
. i
CONCENTRATORKI3) LEVEL I
LEVEL Z
(35)
I SEDIMENTATIQNl LEVEL 3
13
1(44)
BIOLOGICAL I LEV-L
TREATMENT I LEVEL
LEVEL 5
LEVEL 6
(73) LEVEL 7
FIGURE 5. OPTIONS AVAILABLE IN THE WET WEATHER TREATMENT STRING IN THE SWMM.;
49
-------
RUNOFF
DWF PLANT
WET WEATHER PLANT
Figure 6^ OPTIONS AVAILABLE IN THE SWMM STORAGE/TREATMENT BLOCK.
50
-------
Cost Computations —
Cost of storage is based on land and excavation costs. Treatment
costs are calculated as a function of design flow rate. These unit
costs in turn are included in the model on the basis of the most
recent demonstration grant or other installations. Costs are
localized using ENR factors and amortized over a specified period.
Other Information —
The overall storage/treatment schematization shown in Figure 6 is well
suited to both planning and design ojriented simulations. The former
is a relatively new application of the SWMM (23), but is advantageous
because of the more exact representation of the effect of storage in
a long terra simulation. In particular, SWMM accounts for treatment
that occurs in storage due to sedimentation, arid does not treat it
simply as a holding tank. This is highly significant when assessing'
the relative effect of treatment versus storage as control options.
For instance, if flow is allowed through a full storage unit, some
treatment by sedimentation may still occur if sufficient detention
time is planned (e.g., a minimum of 15 minutes). Thus, when combined
with Runoff, a powerful planning tool is available. Any of the
several pathways shown in Figure 6 may be utilized to simulate better
the actual effects of control options.
Receiving Water Block
.Quantity Techniques —
The Receiving Water Model simulates the behavior of estuaries, bays,
reservoirs, lakes and rivers. The program has two distinct phases
which may be simulated together or separately. In Phase A, the
time history of stage, velocity, and flow is generated for various
points in the system. In Phase B, the computed hydrodynamics are
utilized to model the behavior of conservative and nonconservative
quality constituents.
The receiving water is schematized by cutting the continuous system
into a series of discrete one- and two-dimensional elements which
connect node points. For the purpose of this analysis, the velocity
of flow is assumed constant with depth, one-dimensional elements '
represent rivers and specific channels, and two-dimensional elements
represent areas of continuous water surface. For each time-step,
the complete equations of motion and continuity are applied to all
nodal points to derive the hydrodynamics for the system. As a result,
the input and output of the model represent completely transient
51
-------
phenomena. Of course, the spatial distribution is determined by
observation of computed flows and velocities along channels and
stages at nodes (junctions).
Input data consist primarily of a description of bathymetry and
hydraulic characteristics of junctions and channels (e.g., surface
areas and widths, initial stages and depths, roughnesses and
geometry). Tidal or other boundary condition data may be input for
up to 10 junctions.
Quality Techniques —
Given channel flows and junction volumes, quality parameters are
routed by advection, with fluxes flow-weighted if more than one
channel flows into a junction. Diffusive fluxes are not modelled
explicitly although some "numerical diffusion" results from the
computations.
Non-conservative parameters (e.g., BOD, DO) are simulated assuming
first-order decay and reaeration. Under these restrictions, any
water quality parameters may be modeled, and pollutographs (and
hydrographs) may be conveniently entered from cards as well as (or
in place of) from tape/disk files generated by other SWMM blocks.
Quality input data include information on sources, sinks and decay
rates and a tidal exchange factor. The qualityportion may be run
several times after just one run of the quantity portion.
Other Information —
Receiving is a useful "stand-alone" model. Its many options and easy
discretization allow its use in a variety of receiving waters including
swamps and marshes, in applications independent of the rest of the SWMM«
Its primary limitations are its psuedo representation of two-dimensional
systems and lack of diffusive fluxes in quality calculations. The
Receiving Block has been considerably expanded with regard to available
water quality parameters and hydrodynamic options by recent work of the
Raytheon Corporation (20).
Total Simulation
In principle, the capability exists to run all blocks together in
a given computer execution, although from a practical and sometimes
necessary viewpoint (due to computer core limitations), typical runs
52
-------
usually involve only one or two computational blocks, together
with the Executive Block. Using this approach avoids overlay and,
moreover, allows for examination of intermediate results before
continuing the computations. Further, it permits the use of inter-
mediate results as start-up data in subsequent execution runs,
thereby avoiding the waste of repeating the computations already
performed.
USER REQUIREMENTS
Computer Facilities
A large, high-speed computer is required for operation of the SWMM
such as an IBM 360, UNIVAC 1108 or CDC 6600. The largest of the
blocks requires on the order of 90,000 words of storage. Through
considerable efforts, users have been able to adapt portions of the
program to small-core machines such as the IBM 1130, but only with
extensive use of off-line storage and a large increase in execution
time.
Data Requirements
As can be surmised, the data requirements for the SWMM are
extensive. Collection of the data from various municipal and other
offices within a city is possible to accomplish within a few days.
However, reduction of the data for input to the Model is time
consuming and may take up to two man-weeks for a large area (e.g.,
greater than 2000 acres). On an optimistic note, however, most of
the data reduction is straight forward (e.g., tabulation of slopes,
lengths, diameters, etc., of the sewer system). The SWMM is flexible
enough to allow different .modeling approaches to the same area, and
a specific, individual modeling decision upstream in the catchment
will have little effect on the predicted results at the outfall.
Moreover, a highly detailed schematization of the urban catchment
(requiring extensive data reduction) is often unnecessary, and
reasonably accurate results (compared to the detailed schematization)
may be obtained with a relatively "coarse" analysis, as mentioned
previously while discussing the Runoff Block.
Verification and Calibration
The SWMM is designed as a "deterministic" model, in that if all input
parameters are accurate, the physics of the processes are simulated
sufficiently well to produce accurate results without calibration.
This concept fails in practice because neither the input data nor
53
-------
the numerical methods are accurate enough for most real applications.
Furthermore, many computational procedures within the Model are based
upon limited data themselves. For instance, surface quality pre-
dictions are based almost totally on data from Chicago, and are unlikely
to be of universal applicability.
As a result it is essential that some local verification/calibration
data be available at specific application sites to lend credibility
to the predictions of any urban runoff model. These data are usually
in the form of measured flows and concentrations at outfalls or
combined sewer overflow locations. Note that quality measurements
without accompanying flows are of little value. The SWMM has sufficient
parameters that may be "adjusted", particularly in the Runoff Block,
such that calibrating the model against measured data is usually
readily accomplished.
EXAMPLE: RUNOFF, TRANSPORT AND STORAGE/TREATMENT
Introduction
Details of SWMM methodology and operation can only be learned through
a more lengthy study of the documentation, coupled with, experience.
As an alternative, two examples are presented that illustrate various
SWMM capabilities and features. The first will present an application
of the surface runoff and quality routines, as tested in Lancaster,
Pennsylvania.. Since the receiving water application at that location
is relatively uninformative, a different receiving example is given
later. Further details of the example presented below are given by
Heaney and Huber et al. (8).
The Study Area
The City of Lancaster, Pennsylvania, population 79,500, is situated
in a drainage area of about 8.24 square miles (21.34 km ). The
receiving stream in the Lancaster area is the Conestoga River which
drains an area of approximately 473 square miles (1225 km^) into the
Susquehanna River. The average flow is 387 ft^/sec (11 m^/sec)
with a maximum recorded flow of 22,800 ft3/sec (638 m3/sec).
There are two sewage treatment plants within the City, both of which
discharge into the Conestoga River. The North Plant with a capacity
of 10 mgd (3.80 x 10^ m3/day) serves a population of 36,000 people,
and the South Plant, recently expanded from 6 mgd (2.28 x 10^ m-Vday)
to 12 mgd (4.56 x 10^ m3/day), is designed to serve 69,000 people.
Both plants provide, secondary treatment. About one third of the flow
to the North Plant is derived from areas with separate sewers outside
the City serving an estimated population of 17,500 people and some
industries. The remaining two thirds of the sewage flow to the North
Plant is derived from the combined sewers serving the north part of
the City plus about 250 suburban acres estimated to have 18,500 people
and many water-using industries. In addition, most of the year the
54
-------
water table is high resulting in considerable infiltration. An over-
flow line diverts excess flow to the Conestoga during wet weather.
The North Plant drainage area is estimated at 3.72 square miles
(9.63 km2).
The South Plant is designed to handle a population of 34,500 served
by combined sewers and, in addition, up to an approximately equal
amount from separated sewers throughout the surrounding area. The
South Plant drainage area encompasses 4.52 square miles (11.71 km2)
and is comprised of four districts. Stevens Avenue district which is
the subject of an EPA demonstration grant is one of the four districts
connected to the South Plant. Three of the districts, including
Stevens Avenue, pump the sewage from a receiving station within the
district to the South Plant. All locations have overflow arrange-
ments that discharge into the Conestoga River when the capacity of the
system is exceeded.
The total drainage area of the Stevens Avenue district is 227 acres
(92 ha) which, while only about 4.3 percent of the total Lancaster
drainage area served by North and South treatment plants, is 17 per-
cent of the drainage area designed to flow into the South Plant from
combined sewers. The population within the Stevens Avenue district
is estimated at 3,900. Figure 7 illustrates various drainage districts
within the City.
Demonstration Grant Description
In order to remedy the situation resulting from combined sewer over-
flows, the City of Lancaster decided to explore means other than sewer
separation.
Construction of several underground silos at various locations within
the City is contemplated for retention of overflow during wet periods
and subsequent pumping to the treatment plants during low flow periods.
Stevens Avenue district was selected as the demonstration site for
evaluation of the effectiveness of a silo in combating combined sewer
overflows. The sewer layout for Stevens Avenue district is shown in
Figure 8. During normal dry weather periods, the dry weather flow
is pumped to the South Treatment Plant. During wet periods, when
the incoming flow to the pump station exceeds the capacity of the
station, the overflow discharges directly into the Conestoga River
through a 60-inch (1520 mm) sewer located at point 6 on Figure 7.
The City of Lancaster also authorized APWA to develop design parameters
for a full-scale swirl concentrator for removal of solids prior to
the retention of flow in the underground silo (2). Location of the
demonstration site is shown in Figure 8. A flow diagram of the pro-
55
-------
JUNCTION
' = PLANT TREATMENT - I..M.T*
6 = STEVENS AVE
DEMONSTRATION SITE
12 = SOUTH TREATMENT f
PLANT
- NORTH DRAINAGE DISTRICT
SOUTH DRAINAGE
DISTRICT
I h
SCALE IN FEET
2000 4QQO
1000 3000 5000
STEVENS AVE.
DRAINAGE DISTRICT
FIGURE 7. DRAINAGE DISTRICTS OF LANCASTER, PENNSYLVANIA AND NUMBERING
SYSTEM FOR RECEIVING JUNCTIONS.
-------
E. MIFfLIN 57
2) ©
TEVENS TRADE
SCHOOL
HAND J.H.S.
(RAIN
GAGE
CONDUIT USED IN STUDY
SUBCATCHMCNT MANHOI
I) MANHOLE
—— — CONDUIT NOT USED IN STUDY
DEMONSTRATION
PROJECT
STEVENS AVENUE DRAINAGE AREA
LANCASTER, PA.
SCALE IN FEET
0 500 1000
I . I I
Figure 8. STEVENS AVENUE DRAINAGE AREA WITK.
RUNOFF-TRANSPORT NUMBERING SYSTEM.
57
-------
posed swirl concentrator-silo treatment is presented in Figure 9.
In order to fully evaluate this treatment, the city decided to
include chlorination and microstraining as a part of this demon-
stration project. The capacity of the silo is expected to be
160,000 ft3 (4480 m3).
The tasks assigned to the University of Florida were as follows:
1) conduct further verification and testing of
the Storm Water Management Model based on
active overflow measurements on selected
storm events and to make refinements to the
Model;
2) provide results of simulations to the APWA
in order for it to develop design criteria
and sizing of the swirl concentrator;
3) simulate the effect of the swirl concentrator-
underground silo treatment; and
4) simulate the effect of combined sewer overflow
from the entire city to the Conestoga River,
Description of the Stevens Avenue Runs
A total of three studies comprising eight storms were simulated, two
of which are shown here. The City and its engineers provided input
data as well as one overall set of measurements. A description of
each study'and its results are given below.
Study No. 1.—The first study was based on a series of storms between
July 29 and August 3, 1971. The six-day period deposited a record
amount of precipitation throughout the Lancaster area (variously
measured between 7.3 and 9.46 inches, or 18.54 cm and 24.03 cm
respectively). During four of the six days, the storms were very
intense over short periods; in one case, being the second heaviest
on record. For purposes of simulation, Study No. 1 was divided into
six storms. Due to the unavailability of field data for the study,
verification of the results of computer simulations for each of
these storms was not possible. These runs do indicate that an over-
flow as high as 400 cfs (11.20 m^/sec) may be expected for a storm
event similar to Storm No. 6.
Results of this study were used by APWA in sizing the swirl concen-
trator.- A design flow of this device was established at 165 cfs
(4.62 in /sec).
58
-------
Ln
I EXIST. 60
SEWER
BLOCK
CONTROL GATE
EXIST JO^_SAN-SEWER
EXIST. 6" SAN-SEWER
INTERCEPTOR
I—t~"l
WET
WELL
1
TO SOUTH
I--HPUMP I—»TPX&MENT
I
T3EW TOTTST". 1972-, NOT
PART OF DEMO. PROJECT
1 OVERFLOW
/
DISINFECT
MIXING
APR A -
TION
DEVICE
1
SBJO TANK
EXIST. OUTFALL 60-IN.
o-
MAIN
PUMP
T
MICRO -
STRAINER
BACKWASH
•TO RECEIVING WATER
CONESTOGA RIVER
PRELIMINARY LANCASTER FLOW DIAGRAM
FIGURE 9. FLOW OF TREATMENT-STORAGE OPTIONS AT DEMONSTRATION SITE.
-------
Computer simulation studies were also conducted for all six storms
to evaluate the effect of the swirl concentrator-underground silo
facilities on the combined overflow quality. The results of Storm
No. 6 are shown on Figure 10. As illustrated, the quality of all
the overflow is significantly improved through the installation of
the swirl concentrator and a chlorine contact tank. Figure 10 also
shows that as soon as the swirl concentrator operates at full capacity
(165 cfs), excess stormwater simply overflows into the silo. The
shaded portion of the flow versus time graph indicates the volume of
silo storage, 160,000 ft3 (4530 m3). Computed flow from transport
reflects the stormwater inflow to the storage facility (which is the
same as the computed flow to the river without the silo simulation).
The silo dewatering rate is limited by existing pumping capacity
of the Stevens Avenue station (3.57 cfs or 0.10 m3/sec). For simul-
ation purposes, pumping was set 'to begin at a silo depth of 3 feet
(0.91 m).
Study No. 3.—This study is based on a relatively minor rainfall event
of March 22, 1972. This study is of special importance, however,
because it is one of the types-most frequently experienced in terms of
intensity of rainfall (return period of less than one year). It is
also one for which relatively complete verification data such, as
rainfall, flow readings and samples were collected. The rainfall
is shown in Figure 11 along with results of the computer simulation
showing overflow quantity and quality. Shown in the same illustration
are the actual quantity and quality measurements (points labeled x)
of the overflow. It can be seen that agreement between the. computer
simulation and the actual measurements of flow; is only fair. However,
considerable doubt exists as to the validity of the flow data, since
they were obtained from depths measured with a yardstick inserted into
the mouth of the outfall. The agreement between the computed and
measured quality parameters is only fair also. It should be noted that
no calibration of the SWMM was attempted due to the lack of good
verification data.
Computer simulations were also conducted on this study to determine the
effect of the swirl concentrator-underground silo system. These results
are also shown in Figure 11. With the silo system, the model indicates
no overflow into the Conestoga River.
Summary
Testing of the SWMM in Lancaster, Pennsylvania, has revealed the
importance of having sufficient and accurate measured data for model
calibration. It is recommended that the SWMM user begin his simula-
tion from a coarse analysis of the study area. Subsequent refinement
60
-------
I
>
z
.1.
-1
3:00
.COMPUTED FLOW FROM TRANSPORT
.OVERFLOW FROM SWIRL TO SILO
FLOW INTO SWIRL CONCENTRATOR
.COMPUTED FLOW TO RIVER, WITH SWIRL
a CONTACT TANK ONLY
COMPUTED FLOW OUT OF SILO TO
SOUTH PLANT
D.W. F = 0.90cf*
5:00
TIME, HOUR OF DAY
COMPUTED BOD FROM TRANSPORT
COMPUTED BOD TO RIVER, WITH,
S WIRL 8 C ON TAC T TANK ONLY
COMPUTED BOD OUT OF SILO
/ TO SOUTH PLANT
D.W.F = 363mg/l
4=00
TIME, HOUR OF DAY
SOO
CO
tn
300
COMPUTED SS FROM TRANSPORT
.COMPUTED SS TO RIVER, WITH SWIRL
CONTACT TANK ONLY
^COMPUTED SS OUT OF SILO TO
SOUTH PLANT
D.SV.F =
4iOO
TIME,HOUR OF DAY
5=00
Figure 10. SIMULATION FOR STEVENS AVENUE, STUDY 1, STORM 6.
HIGH CONCENTRATIONS OUT OF SILO REFLECT INITIAL
SLUSH OF DWF IN SEWER SYSTEM INTO THE SILO.
61
-------
c
~". "H
>• O
53
z
UJ
i-
Z Qj
_l
N
op.
U
COMPUTED FLOW
FROM TRANSPORT
COMPUTED FLOW OUT
OF SILO TO
SOUTH PLANT
COMPUTED FLOW TO
RIVER, WITH SWIRL S
CONTA'CT TANK ONLY
1120
12:20 l'20 2-20
TIME, HOUR OF DAY
3'20
A 20
Q
O
o
O
5
O
o
s
COMPUTED BOD FROM TRANSPORT
..-<—COMPUTED BOD OUTOFSILO TO SOUTHPLANT
COMPUTED BOD TO RIVER,WITH SWIRL
a CONTACT TANK ONLY
11:20
12:20 1.20 2:20
TIME,HOUR OF DAY
3 20
4 20
o
8
•
CO
"-MEASURED SS
-COMPUTED SS FROM TRANSPORT
COMPUTED SS TO RIVER,WITH SWIRL
CONTACT TANK ONLY
v_
\
••-... __ ,xs >< "
„ •—-——. N./••'•>< • *•"*
' '-T ^— —-T - ....
1*20 12:20 120 2:20
TIME,HOUR OF DAY
3:20
4:20
FIGURE 11.
SIMULATION FOR STEVENS AVENUE, STUDY 3. HIGH CONCENTRATIONS
OUT OF SILO REFLECT INTIAL FLUS OF DWF IN SEWER SYSTEM INTO THE
SILO. FOR THIS STORM, SILO WOULD STORE ALL RUNOFF, BUT RESULTS
WITHOUT IT ARE ALSO SHOWN.
62
-------
can be tailored with his knowledge of the more sensitive parameters
(impervious area, detention storage, catchbasin BOD content, infil-
tration and others) and available field measurements. The model is not
fully calibrated until satisfactory results have been obtained with
more than one storm within the study area. The application to Lan-
caster suffered from lack of good measured data for several of the
storms studied, but it can be generally stated that for the Study No. 3
storm:
1) The SWMM was able to predict only fairly the
quantity, of the combined overflow for the
Stevens Avenue district; however, the only
set of measured flows are questionable.
2) Computed suspended solids were universally
lower than measured values at the overflow
and in the receiving water (Conestoga River)
probably because erosion was not modeled and
appropriate upstream concentrations were
unknown.
3) The installation of the swirl concentrator
and silo complex will result in substantial
improvement in the quality of the overflow
at Stevens Avenue, provided the full-scale
performance of the swirl concentrator is
comparable to the results obtained in labora-
tory studies by APWA. This fact should be
true for any storm comparable to those des-
cribed earlier in this report.
EXAMPLE: RECEIVING BLOCK
The Receiving Block of SWMM was applied to the lower St. Johns River
for further developmental work on the model, and as part of a study
of the waste assimilation capacity of the river. Since the Receiving
Model is completely transient in both the hydrodynamic and water
quality portions of the program, it is especially well suited for
estuarine applications. Further details of the simulation discussed
below are given by Huber and Heaney et al. (11).
The St. Johns River, from Palatka, Florida to its mouth at Mayport,
a distance of about 75 miles (120 km) is presented as an example
application of the Receiving Water Block. Figure 12 illustrates the
layout of the 37 junctions and 40 channels used in the simulation.
In addition, Figure 13 is a detailed diagram of the area of junctions
13, 14 and 15. All physical data were reduced from US Coast and
Geodetic Survey nautical charts 685. and 636-SC and US Geological Survey
63
-------
FIGURE 12. SCHEMATIZATION OF THE ST. JOHNS RIVER FOR
RECEIVING SIMULATION. JUNCTIONS 19-24 PROGRESS
UPSTREAM TO JUNCTION 24 AT PALATKA. JUNCTIONS
34-37 ARE USED TO SIMULATE STORAGE AVAILABLE IN
TIDAL MARSHES. THE INTERACTION OF JUNCTIONS
33 AND 37 WITH THE INTRACOASTAL WATERWAY, ON WHICH
THEY ARE LOCATED, IS NOT SIMULATED.
64
-------
FIGURE 13. SCHEMATIZATION OF PORTION OF ST. JOHNS RIVER AT
JACKSONVILLE, FLORIDA.
65
-------
*> I—
HHW
ov
MIDNIGHT
10 15
NOON
TIME (T), hours
25
MIDNIGHT
Figure 14. REPRESENTATIVE SEMI-DIURNAL TIDE AT MOUTH OF ST. JOHNS RIVER AT
MAYPORT, FLORIDA, AUGUST 1, 1970. DATUM IS MEAN LOW WATER.
-------
quadrangal maps for Mayport and Eastport, Florida. Note that
channels represent meandering river segments and need not be
straight lines. Lengths and areas are scaled off the maps.
Measured tides at junction 1 were input from the National Oceano-
graphic and Atmospheric Administration tide gage at Mayport.
Figure 14 illustrates the type of semi-diurnal tide experienced
there. Measured transient river flows are entered at Palatka,
junction 24.
Quality input data reflect the base dry weather loads entering the
system at several junctions. The purpose of the run was to determine
the effect of these loads rather than to simulate a specific storm-
water runoff event. Thus, the only "stormwater" input is the river
inflow at Palatka and is read in from cards. The BOD loading at
Palatka is obtained through a known concentration measured there. This
is then multiplied by the flow and converted to pounds per day.
The principal output of the quantity portion of the Receiving Model is
stages at junctions and flows and velocities along channels. Samples
are shown in Tables 1 and 2. The tidal influence may be clearly seen.
Representative quality output is shown in Table 3. In addition,
maximum, minimum, and average concentrations are printed out after each
simulated day. In this example, the predicted chloride concentrations
compared favorably with measured values along the river, and are con-
sidered a useful verification parameter. Verification of predicted
BOD and DO values is difficult on the St. Johns because of extensive
benthal deposits and other unknown sources and sinks. Identification
and quantification of sources, sinks and decay rates of water quality
parameters is typically the most difficult single problem in modeling
of receiving water quality.
SUMMARY
The EPA Storm Water Management Model CSWMM) ia a useful tool for
analysis of problems related to urban stonnwater runoff. As such,,
it may be used for both the planning and design phases of investi-
gations. The former utilizes time steps on the order of one hour,
and simulation times on the order of years. The latter is oriented
toward single event simulation and utilizes time steps on the order
of minutes and simulation times on the order of hours.
Although only one of several urban runoff models currently available
the SWMM is perhaps the most accessible and widely disseminated of any.
It has found considerable acceptance in the engineering community and
continues to be at the state-of-the-art by virtue of extensive use
and continuous updating and maintenance.
67
-------
TABLE .1- SAMPLE JUNCTION OUTPUT, DAY 2.
O\
00
-M
.?o
«.(
iii
9.00
!:?8
JULY 19, 1972 TIDE AT HAYPORT.
DAY TS 2
HOUR
1:1*
3.00
o.OO
5.00
2.5570
:jj{[
:\l\
.9 5*
.9404
• 9586
10036
*-?fi*
2.4791
DATA.
tNC WATER HYDRODYNAMICS
Use MEASURED FLOW AT PALATKA» JULY 17-20* 1973.
T I H E HISTORY OF S T A 6 E
2.3955
\im
!
.?
1:
:f
hi
;1
i
ii
.96 4
iii
lit
;.;:
59
160
448
09
r •»
?4
I'l*
i
I. 85
2^597
i
8 Ji*?
4
-------
TABLE 2. SAMPLE CHANNEL OUTPUT, DAY 2.
IN DATA.
fitEtttx
NG WATER HYDRODYNAMICS
JULY 19, I»72 TIDE AT NAYMRT.
OAY T8 Z
USE NEA8UPED FLOW AT PALATKA, JULY 17-20, 1973.
o.o
:i!
N8
^
I
J;s!
9.00
^:S§
ii:§S
!!:§§
TIME HI
CHANNEL
?S>
,1111:
mi:
i
:il;
tip
92776.
4950l!
•30173.
•90876. -
-108949! -
0.44
•I
:4
•0.47
:K
.5.OM
0.3|
0.4
J:P
fc-i
CHANNEL 17 18
FLO*, VEL.
ft 9 -B
66004. 5!30 -1062
97772. 0.44 • 16
»«.!
57<
S T 0 R
16 17
%
$:82
Y OF FLOW AND VELOCITY
CHANN
EL
•0.
lizi
jjj:
jj|
IP
•3335.
s:!8
?•{
g«j
J}*i
2*
8:1
'•ill
-68|50!
108359!
Ikl:
"Ml:
-IP
-«!??
85.
m
•0*06
1:11
g-i
0.(
m
-123407: -o!$4
:«»{: :8:32
-7?225! -5:?i
2:o'
5!«3
i»§I!l« 8-S2
472]i4( 0.20
:i fli Hit!
CHANNEL 14 15 CHANNEL
it
JB»
-55147.
•3446$.
ittlt'
tiiii:
»320.
.66
-0.1.
•0!4<
m
_- s7. -c
•36803. -C
13516! t
$6565. C
53806.
.28
-460
>66^:
')!^
*
15 .
-0.
s- •
*• •
sin
!!4i
11. 8
I
-Si
:?»,..
:|WB:
mi: 8:1
ti^I* i-'
8zil: S:
43632. 0.
8:
-S1IS:
:MSl: :..
:!^IS: :S:
I
CHANNEL
FLOW
CCFS)
-50755
6i
13_ 14
:!!
59
?i
if:?
7728
•40_
-610
-6t
-64 "
-53
n
•1
*1
:l
0.41
0.0]
•oloj
-------
TABLE 3. QUALITY OUTPUT DURING DAY 2.
THIS IS A SAMPLE RUN OF QUANTITY AND QUALITY EPA STORMWATER MODEL
LOWER ST. JOHNS RIVER. JULY t972 VERIFICATION DATA.
JULY 19. 1972 TIDE AT MAYPOAT. USE MEASURED FLOW AT PALATKA, JULY 17-20, 1973.
JUNCTION CONCENTRATIONS, M6L (OR MPN/D* DURING TIME CYCLE (DAY) 2, QUALITY CYCLE 2* TINE (SEC) 7200.
CONSTITUENT NUMBER 1 BOD
JUNCTION 1 2 3 4 5 6 7 8 9 10
Jflft S:?I ?:§! 8:31 S:H 8:8 i:ft 1:11 5:18 8:51 j:»
21 TO SO 0.64 0.64 0.64 6.79 0.85 1.29 0.79 2.46 0.78 (.63
31 TO 37 0.56 0.62 0.48 0.64 0.60 0:?9 0^9
CONSTITUENT NUMBER 2 CHLORIDES (MOD
JUNCTION 1 2 3 4 5 6 7 8 9 10
JUNCJTQN
UNCTION --„...,,,
I TO TO 17732.35 14450.17 9961.09 8538.69 6707.11 7892.46 5075.64
1 ?81{ ml:h ! if :J ^l:ll 1JH i \M\ M\tt MM
1 TO Sf 8*80.00 6850.09 116U.25 lW\.ll H>t>i.l\ iSS?!?? \\lil.ll
!!!!! iil! !;!! !i!l !i!! III! 111! III! « a
ii § ! • -
CONSTITUENT NUMBER 3 BOD (DO)
JUNCTION l 2 5 * 5 * 7 8 « ' 10
COMPUTED OXYGEN SATURATION VALUES(MCL) AT THIS TIME STEP
123456789 10
2 $•!? $•*! !•?! !•» M* Li! Z»H !•« Z««« !.««
COMPUTED REAERATION COEFFICIENTS (I/DAY) AT THIS TIME STEP
ill iii! HR il ii !i!i ii i is a» Si
-------
REFERENCES
1. American Public Works Association, "Water Pollution Aspects
of Urban Runoff," Federal Water Pollution Control Admini-
stration, Report WP-20-15, January 1969.
2. American Public Works Association, Research Foundation, "The
Swirl Concentrator as a Combined Sewer Overflow Regulator
Facility," Office of Research and Monitoring, US EPA,
Report EPA-R2-72-008, September 1962.
3. Bowers, C. E., Harris, G. S., and A. F. Pabst, "The Real-
Time Computation of Runoff and Storm Flow in the
Minneapolis-St. Paul Interceptor Sewers," St. Anthony
Falls Hydraulic Laboratory, Memo No. M-118, University of
Minnesota, Minneapolis, MN, December 1968.
4. Fair, G. M., Geyer, J. C. and D. A. Okum, Water and Waste-
water Engineering. Wiley, 1968.
5. Field, R. and J. A. Lager, "Urban Runoff Pollution Control —
State of the Art," J. Env. Eng. Div.. ASCE, Vol. 101, No. EE1,
February 1975, pp. 107-125.
6. Field, R. and E. J. Struzeski, Jr., "Management and Control
of Combined Sewer Overflows," J. Water Pollution Control
Federation. Vol. 44, No. 7, 1962.
7. Gameson, A. L. and R. N. Davidson, "Storm Water Investi-
gations at Northhampton," Institute of Sewage Purification
Conference Paper No. 5, Annual Conference, Leandudno,
England, 1962.
8. Heaney, J. P., W. C. Huber .et al.. "Urban Stormwater
Management Modeling and Decision-Making," Office of Research
and Development, US EPA, Report EPA-670/2-75-022, Cincinnati,
Ohio, May 1975.
9. Henderson, F. M., Open Channel Flow. MacMillan, 1966.
10. Huber, W. C., "Modeling for Storm Water Strategies," APWA
Reporter. Vol. 42, No. 5, May 1975, pp. 10-14.
11. Huber, W. C., J. P. Heaney et al.. "Storm Water Management
User's Manual Version II," Office of Research and Development,
US EPA, Report EPA-670/2-75-017, Cincinnati, Ohio, March 1975.
12. Hydrologic Engineering Center, Corps of Engineers, "Urban
Storm Water Runoff: STORM," Generalized Computer Program
723-58-L2520, Davis, California, January 1975.
71
-------
13. Klym, H. Koniger, W., Mevius, F. and G. Vogel, "Urban Hydrological
Processes, Computer Simulation," Dorsch Consult, Munich,
Toronto, 1972.
14. Lanyon, R. F. and J. P. Jackson, "A Streamflow Model for
Metropolitan Planning and Design," ASCE Urban Water Resources
Program, Technical Memo No. 20, ASCE, 345 E 47 St, New York,
New York 10017, January 1974.
15. Leiser, C. P., "Computer Management of a Combined Sewer System,"
Office of Research and Development, US EPA, Report EPA-670/2-74-022,
July 1974.
16. Metcalf and Eddy, Inc., University of Florida, and Water
Resources Engineers, Inc., "Storm Water Management Model,
Volume I - Final Report," Environmental Protection Agency,
Water Quality Office, Report No. 11024DOC07/71, September 1971.
17. Metcalf and Eddy, Inc., University of Florida, and Water
Resources Engineers, Inc., "Storm Water Management Model,
Volume II - Verification and Testing," Environmental Protection
Agency, Water Quality Office, Report No. 11024DOC08/71,
September 1971.
18. Metcalf and Eddy, Inc., University of Florida, and Water
Resources Engineers, "Storm Water Management Model,
Volume III - User's Manual," Environmental Protection Agency,
Water Quality Office, Report No. 11024DOC09/71, September 1971.
19. Metcalf and Eddy, Inc., University of Florida, and Water
Resources Engineers, Inc., "Storm Water Management Model,
Volume IV - Program Listing," Environmental Protection Agency,
Report No. 11024DOC10/71, September 1971.
20. Raytheon Company, "RECEIV-II, Water Quantity and Quality
Model, New England River Basins Modeling Project, Documentation
Report," Vol. I, Final Report to US EPA, Raytheon Company,
Portsmouth, Rhode Island, December 1974.
21. Roesner, L. A. et al.. "A Model for Evaluating Runoff-Quality
in Metropolitan Master Planning," ASCE Urban Water Resources
Research Program, Technical Memo No. 23, ASCE, 345 E 47 St,
New York, New York 10017, 72 pp, April 1974.
22. Shubinski, R. P. and L. A. Roesner, "Linked Process Routing
Models," Spring Meeting, American Geophysical Union,
Washington, DC, April 1973.
72
-------
23. Smith, G. F., "Adaptation of the EPA Storm Water Management
Model for Use In Preliminary Planning for Control of Urban
Storm Runoff," ME Thesis, Department of Environmental
Engineering Sciences, University of Florida, Gainesville,
May 1975.
24. Sogreah, "Mathematical Flow Simulation Model for Urban
Sewerage Systems," Caredas Program, Partial Draft Report,
Sogreah, Grenoble, France, April 1973.
25. Stall, J. B. and M. L. Terstriep, "Storm Sewer Design —
An Evaluation of the RRL Method," Environmental Protection
Agency, Office of Research and Monitoring, EPA-R2-72-068
October 1973.
26. Terstriep, M. L. and J. B. Stall, "The Illinois Urban
Drainage Area Simulator, ILLUDAS," Bulletin 58, Illinois
State Water Survey, Urbana, Illinois, 1974.
73
-------
HYDROLOGY REVIEW
By
Donald Dean Adrian*
INTRODUCTION
A brief discussion of a few hydrologic principles will be presented.
The review will consist of a discussion of precipitation, its measurement
and description in graph form, its distribution in time and space;
infiltration and infiltration capacity; and runoff.
PRECIPITATION
Precipitation is measured by raingages which may be recording or
nonrecording. The nonrecording type measures the total amount of
precipitation between readings. It is being replaced by the recording
raingage which record the rate of precipitation and the time of its
occurrence.
Ordinary raingages may be modified with shields to record snowfall.
However, due to the snow being carried by the wind inaccuracies in
measurement may be introduced. Because snowtnelt is not converted immed-
iately into runoff, snowfall has received little attention in the SWMM.
Rainfall records are usually presented in the form of bar graphs
called hyetographs. The rainfall intensity serves as the ordinate and
time is the abscissa.
Rainfall records are frequently analyzed for intensity-duration
relationships. The average rainfall intensity in inches per hour decreases
as the duration increases. For example, in Los Angeles, at an average
of once each five years the rainfall intensity will be 3.2 in/hr for a
five minute duration, 1.6 in/hr for a 20 minute duration, and 0.85 in/hr
for a 60 min duration.
If multiple raingages are available their records for any storm will
be different, with each raingage being most representative of the area
in its immediate vicinity. Rainfall is allocated to the surrounding area
by the Theissen Polygon Method. Refer to Figure 1 for the construction
of the Theissen Polygon for the area shown. One has three raingages 1,
2, and 3. Straight lines are drawn connecting the raingages, then the
perpendicular bisector of each line is drawn. The perpendicular bisectors
serve as the boundaries for the area of influence of each raingage.
Thus raingage 1 serves area A., raingage 2 serves area A_, and raingage 3
serves area A-. The method may be extended to define the area served by
more raingages.
Professor of Civil Engineering, University of Massachusetts/Amherst.
74
-------
REPRESENT
RAINGAGES
REPRESENTS
PERPENDICULAR
BISECTOR
Figure 1. Construction of a Theissen Polygon
75
-------
INFILTRATION
Infiltration is the rate of movement of water into the ground. In
hydrology "infiltration" is used to express the notion that some precipi-
tation does not run off, but soaks into the ground. In environmental
engineering "infiltration" is also used to describe the movement of
water out of the ground and into sewers through cracks, poorly sealed
joints, and other openings. The reader is cautioned to note the dis-
tinction between these two usages.
Infiltration rate (in the hydrologic sense) is important in studying
runoff because if the rainfall rate is less than the infiltration rate
no runoff will occur. The maximum infiltration rate is called- the
infiltration capacity. It is a function of time and has been expressed
in equation form by Morton's equation
f = f + (f - f )e~ktR
p co c
where f is the infiltration capacity,
P
f is a constant rate of infiltration observed at the end of a
£
long storm
f is the maximum rate of infiltration at the beginning of a storm,
k is a positive constant, and
t is the duration of rainfall.
jx
OVERLAND FLOW AND SURFACE RUNOFF
Runoff starts when the precipitation rate exceeds the infiltration
capacity and depression storage has been filled. Runoff first flows in
trickles and rivulets until it joins larger streams and makes its way
to a gutter or channel. Water which is enroute to a channel is known as
overland flow. It is known as surface runoff after it enters a flow channel.
Water is stored temporarily while it is engaged in overland flow.
The temporary storage results because a certain depth of water builds up
as overland flow takes place. Such temporary storage is known as surface
depression storage. Depression storage represents water which does not
engaged in overland flow, such as water filling a pothole in a street.
It has been difficult to describe in equation form the behavior of
overland flow. Difficulties arise because of the complex flow paths which
the water follows as it travels to become surface runoff.
76
-------
HYDRAULICS REVIEW
By
Donald Dean Adrian*
BASIC CONCEPTS
Open channel flow is flow in a conveyance structure which allows
the upper surface of the water to be in contact with the atmosphere.
An example is flow in an irrigation canal. Another example is flow
in a stream. A third example is flow in a partially filled sewer.
Gravity is the driving force inducing flow in an open channel.
Pipe flow, or sometimes pressure flow, is flow in a conveyance struc-
ture, such as a pipe, in which the water completely fills the pipe.
An example of pipe flow, or pressure flow, is flow in a water main.
Another example of pipe flow or pressure flow is flow in a surcharged
sewer, i.e., a sewer which is completely filled during a storm event.
Unsteady flow is flow which changes with time. An example is the
flow in a storm sewer during the initial period of a rainfall event.
jtteady flow is flow which does not change with time. An example
might be the flow in a force main. (A force main is a sewer which
is designed to always operate under pressure, such as a sewer which
leads from a pumping station located in a low part of town and
discharges to a treatment plant located at a higher elevation.)
Uniform flow is flow in which the velocity stays the same in magni-
tude and direction throughout the whole of the fluid. Normally one
can think of uniform flow as being open channel flow in which the
depth remains constant along the channel length of interest.
Non-uniform flow is flow in which the velocity changes in magnitude
and direction over the channel length of interest. Again, it is eas-
ier to visualize non-unifqrm flow as open channel flow which changes
in depth along the channel length of interest. Sometimes it is advan-
tageous for computational purposes to treat a non-uniform flow as a
uniform flow. For example, a long combined sewer may have a signifi-
cant depth of flow change along its length, thus making the flow non
uniform, but if the length of the combined sewer is visualized as being
made up of several shorter segments, say 10, the flow in each segment
can be approximated as being uniform flow.
Discharge rate or sometimes the discharge, is the volumetric flowrate,
usually measured in units such as million gallons per day, cubic feet
per second, liters per second, and cubic meters per second.
K
Professor of Civil Engineering, University of Massachusetts/Amherst.
77
-------
CONTINUITY
The principle of conservation of mass, which states that the rate of
mass inflow minus the rate of mass outflow is equal to the change in
storage in a control volume, leads to a useful concept in hydraulics, the
equation of continuity.
The continuity equation is most readily visualized for steady flow.
It states that the discharge, Q, is equal to the cross sectional area
through which flow takes place, A, multiplied by the flow velocity, V.
Thus,
AV (1)
where A, V, and Q may be measured anywhere along a channel. At the junction
of two channels which combine to form a single channel the equation of
continuity states
Ql + Q2 " Q3
where Q- - discharge of the first upstream channel
Q9 « discharge of the second upstream channel
Q~ » discharge of the third channel, the downstream channel.
Along a single channel the equation of continuity may be written
A1V1 - A2V2 - A3V3
where the subscripts now refer to different locations along the channel.
Figure 1 summarizes the continuity relationships for steady flow.
The depth of flow, h, and the cross sectional flow area, A, are
functions of X for steady nonuniform flow. Thus h - h(X), A - A(X), and
of necessity V - V(X) since V « Q/A(X). Any differential changes in h, A or
V will then be expressed as ordinary derivatives. As the equation of
continuity states Q - AV one can take the derivative to obtain a differential
form of the equation of continuity for steady non-uniform flow
f .^(AV)orO.vf +Af (4)
78
-------
FOR STEADY FLOW : Q,.«Q2-«Q3 AND A | V, « A 2V 2 •
A3V3
FOR STEADY UNIFORM FLOW: y= CONSTANT AND A,»A2SA3
.-. V, -V2j V3
FOR STEADY NONUNIFORM FLOW! A, # Ag* A3
BUT A,V, s A2V2- A3V3
Figure 1. Equation of Continuity
79
-------
For "unsteady" flow the continuity principle is not so readily apparent
as for steady flow. If one considers a reach of length AX the discharge Q
may be different at either end of the reach. Thus Q can be changing with
distance. Also, since the flow is unsteady Q may be a function of time.
Thus Q - Q(x,t) and any differential changes in Q have to be expressed
as partial derivatives.
The flowrates Q1 and Q_ may be expressed in terms of Q(X,t), which is
shown in Figure 2 at the center of the control volume. By use of a Taylor's
Series, truncated after the second term,
Q(X,
(5)
Qo -
(6)
JNt,
The change in storage inside the control volume is AX • B -r— where B is
the water surface width. From the conservation of mass principle that the
rate of inflow minus the rate of outflow is equal to the rate of change of
storage, one can write
- Q2 - B AX
(7)
or, writing Q for Q(X,t),
f) - B -^ AX
<«
which becomes
and this expression simplifies to
U0>
Sometimes it Is useful to replace Q by its equivalent AV to obtain
80
-------
Figure 2. Definition Sketch'for Equation of Continuity
for Unsteady Nonuniform Flow
DATUM
3A , . 3V
(ii)
as the equation of continuity for unsteady non-uniform flow.
THE ENERGY PRINCIPLE
The energy content of a fluid is expressed in terms of the "head" in
hydraulics usage. The energy content is measured in ft-lb units, and by
expressing it on a per Ib basis one obtains the energy content per Ib with
the units of ft. Thus, the head available is expressed in terms of ft.
The Bernoulli Equation
The Bernoulli equation
+ Z- + •—
1 2g
(12)
expresses the head relationship between two points , 1 and 2 , along a pipe or
open channel. In the above expression each of the terms p/f, Z, V /2g and
have dimensions of feet. The terms represent:
p
Y
Z
pressure
specific weight of the fluid
elevation sbove a reference datum
head loss between points 1 and 2
(h. represents frlctional dissipation)
81
-------
For steady nonuniform flow one can examine the change in the total head,
H, as a function of distance. Let
H = p/Y + Z + V /2g
(13)
and note that
dH
dX
dh
S where S - frictional slope
(14)
Then (see Figure 3)
SLOPE
sf ••
dH
dX
BOTTOM SLOPE
S0 =" dfx
DATUM
Figure 3. Definition Sketch for Steady Nonuniform Flow
82
-------
f -3z « + * + >•-»f <15>
or
Letting V = Q/BY and differentiating yields
dX gB Y dX
which becomes, in terms of the Froude number
- 8f
_v
\here F - ; , or F_
r /BY ' r
Byi/iy~
The above relationship is widely used in developing water surface
profiles for non-uniform steady flow.
Critical Depth, Subcritical and Supercritical Flow
A gravity wave will travel through water in an open channel with a
velocity given by the expression
C - v'gy" (19)
where c - wave velocity
g » acceleration of gravity
y » depth of the water.
The above expression for wave velocity divides open channel flow regime
into two parts. By substituting for the wave velocity c - Q/A, and express-
ing for a rectangular open channel A -BY, one obtains
83
-------
(20)
as the expression for the critical depth of the channel.
If the depth of flow in the channel is greater than Y£ the flow is
called subcritical. If the depth is less than Y£ the flow is called
supercritical. In subcritical flow waves can travel faster than the
stream velocity, and are transmitted both upstream and downstream. In
supercritical flow waves travel slower than the stream velocity, and
are transmitted only downstream. The above property of waves is important
in devising numerical flow routing schemes.
A useful indicator of whether the flow regime is subcritical, critical,
or supercritical is the Froude number, F . It is defined as
F = v/i/g7 or F = ^ - <21)
where V is the actual stream velocity. If F < 1 the flow is subcritical.
If F = 1 the flow is at critical depth, and if F > 1 the flow is super-
critical.
THE MOMENTUM PRINCIPLE
The momentum principle is widely used in hydraulics. In applications
one applies the principle by summing all the forces acting on a control
volume of water. These forces usually are: hydrostatic pressure forces,
momentum flux forces (usually these are expressed as QpV where Q is the
flowrate, p is the specific density, and V is the stream velocity),
frictional resistance forces, and a component of the gravity force in
the direction of flow.
An application of the momentum principle is in the hydraulic jump
which is a violent transition from supercritical to subcritical flow. If
one lets y, and y_ represent the depth of flow before and after a hydraulic
jump, and F - and F „ represent the before jump and after jump Froude number,
the momentum principle yields
~ - T (/I + 8F* - 1) (22)
y;L 2 v'* ' rl
84
-------
which is useful when upstream conditions are known and the depth y2 is to
be solved for. An alternative form which will give the upstream depth
if the downstream conditions are known is
"2~2
(23)
FLOW RESISTANCE
The resistance of a channel or a pipe to flow results in a relationship
between the flow velocity and channel properties. Over the years a number
of equations describing the relationship have been developed. One of
the most widely used is the Manning equation. It is written
v . i-486 R23 s12 (24)
n
where V = the flow velocity
n = a coefficient dependent upon the channel roughness
S = slope of the energy grade line
R = hydraulic radius.
Values of n are tabulated for different materials, for example, n = 0.014
for unfinished concrete and n - 0.025 to 0.030 for clean straight river
channels. S is the slope of the energy grade line (the energy grade line
V2
is given from the Bernoulli equation as the sum of p/y + Z + •=— ) for the
channel or pipe. S may be calculated from the Bernoulli equation as
where h.. is the head loss between two points a distance L apart. R is
defined as the ratio of the cross sectional area through which flow takes
place, A, to the wetted perimeter, Pw. The wetted perimeter is the
length of the contact surface between the flowing liquid and the solid
boundary. For a pipe flowing full the wetted perimeter is the pipe circum-
ference 2irr, the cross sectional area is irr,2 resulting in a hydraulic
radius of R = r/2. If the pipe were flowing half full Pu> • TOT,
1 2
A » -r irr and R - r/2 again. For other values of flow depth R » r/2,
but must be calculated or looked up in a table.
The Manning equation is applicable to steady uniform flow, but is
frequently used for unsteady and non-uniform flows. For steady uniform flow
the value of S may be taken as the slope of the energy grade line or the
slope of the channel bottom. For non-uniform flow the slope S for use
in the Manning equation is the slope of the energy grade line. Sometimes
subscripts are used to differentiate between the two slopes, S representing
the slope of the channel bottom and Sf representing the frictional slope,
or the slope of the energy grade line.
85
-------
CHANNEL CONTROLS
Certain structures in an open channel serve to establish a unique
relationship between the head available and the flowrate. Such structures are
called controls. They are often encountered in calculations because they serve
as boundary conditions at upstream or downstream reaches of a channel. Two
types of control will be discussed here, namely the sharp crested weir
and the free overfall. See Figure 4 for sketches of these structures.
The flowrate over a sharp crested weir is given by the relation
2 r- 2/3
Q = ~ CDB /2g H (25)
where C is a discharge coefficient and all other terms have been previously
defined. The discharge coefficient varies with the ratio H/W and has a
value of 0.611 when W is large. A relationship for other values of the
ratio is
C = 0.611 + 0.08 H/W (26)
The free overflow is another type of control which may occur in
practice when a sewer discharges into a river or a drop structure.
Critical depth, Y , occurs slightly upstream from the brink of the free
overfall, establishing a unique relationship between the depth and the
flowrate. From the critical depth relation
(27)
the depth Y can be solved for if Q is known (if B is a function of depth
a trial and error solution may be necessary), or if y at the overfall is
known Q can be solved for. For some calculational purposes, especially
when A and B are functions of depth, another form of the critical depth
relation may be easier to use. It is
- 1 (28)
where the subscripts indicate that the relation is applicable only at
critical depth.
86
-------
\ \ \ \ \\A\\\\\\\\\\
THE SHARP CRESTED WEIR
/ / / /
- 3to4yc
THE FREE OVERFALL
Figure 4. Two Channel Controls
87
-------
UNSTEADY NON-UNIFORM FLOW
Unsteady non-uniform flow occurs frequently in a st orm or combined
sewer system as the storm event continuously changes the flowrate the
sewer has to carry. Two basic relations are necessary to develop equations
to describe unsteady non-uniform flow: a dynamic equation of motion
(that is one which considers the acceleration of the fluid) and an equation
of continuity.
The equation of continuity for unsteady non-uniform flow was derived as
00)
or
Refer to Figure 5 in the derivation of the equation of motion for
unsteady nonuniform flow. The slope of the total energy line is
JTJ jiu ji w
-rrr = S _ and —• = — (h + —) . Recall that for steady
uX f OA oA Zg *
nonuniform flow — = - Sf. The frictional slope S,. is expressed in terms of
OA £ I
a flow equation, such as the Manning equation
v . 1^8_6 R2/3 1/2 (24)
n f
which can be changed around to yield
v2 2
S , I p (29)
f 4/3 '
1 2.208 R /J
An alternate form of expression for the friction slope is obtained by
balancing the boundary shear with the gravity force leading to T • yRSf*
The energy available to the flow expressed by the slope of the total
energy line, dH/dX, can be used to overcome frictional resistance, expressed
by S,. , and the excess energy will be available for accelerating the flow.
88
-------
DATUM
Figure 5. Definition Sketch for Unsteady Nonuniform Flow
One can express the above concept in equation form by stating
e
(30)
where S is the slope of the total energy line and S is the acceleration
slope. Evaluating the forces acting on a control volume of liquid of length
AX in Figure 5, i.e., evaluating the pressure, resistance, and gravity
forces; results in an unbalanced force of
- yAAh - TQPAX
(31)
which is available to accelerate the mass of fluid in the control volume
where the mass is pAAX. TQ is the boundary shear resistance and P is the
wetted perimeter. Newton's law of motion F - ma can be applied to obtain
fAAh - TQPAX - pA AX a
(32)
dV
where a is the acceleration. The acceleration a - ~ and by noting that
V - V(x,t) and using the chain rule x dt
dV _ 8V dX , 3V
dt " ax dt at
89
(33)
-------
J-y
then noting that -r—
at
= V the acceleration is
dV 3V 3V
dt ax 3t
(34)
Then Newton's equation becomes
T
o _ ,.3h
vR ~ ~ 3X
- — + -—^
g ax g at'
(35)
after using the relation y-pg, and R = A/P. Now —
oX
3H a
j- (h + ) so that
M + I
3X g
(36)
which may be written in terms of slopes as
S + S + S. = 0
ear
(37)
The bed slope is introduced into the dynamic flow equation by
noting h = z + y, then - = - +
and recalling — « - SQ so that
— = - S + -££ . The dynamic flow equation then becomes
dX O 9X
f ~ o ax g ax g at
steady uniform flow
steady nonuniform flow
unsteady nohuniforn flow
(38J)
The dynamic flow equationand the equation of continuity provide the basis
for developing the solution to unsteady nonuniform flow problems. Due to
the difficulty in solving these simultaneous equations a number of
simplifications have been developed.
90
-------
SWMM APPLICATION STUDY GUIDE
By
Thomas K. Jewell, Peter A. Mangarella, Francis A. DiGiano
Donald D. Adrian*
INTRODUCTION TO WORKSHOP STUDY GUIDE
This study guide has been prepared to allow the reader to utilize the
U.S. Environmental Protection Agency Storm Water Management Model (SWMM)
in a typical case study. Information is presented concerning data gathering
and preparation for each of the four SWMM functional blocks. The attendee
is assigned the task of completing a case study that illustrates typical
conditions encountered in practice, for each block. All necessary data for
completion of the case study are provided along with sample solutions.
The Runoff and Transport Block case studies have been developed using
an urban drainage basin of approximately 300 acres. Figure la is a
topographic map of this basin on a scale of 1 inch = 660 feet. Figures Ib-le
show the four quadrants of the basin with a scale of 1 inch » 360 feet.
A schematic diagram of the basin sewer system (combined sewers) is given
in Figure 2. This sytem was designed about 20 years ago when the projected
land uses included industry, single and multi-family housing and open space.
At that time the interstate highway, large shopping center and drive-in movie
were not contemplated. Localized flooding during moderate to heavy rain-
falls has been reported.
The objectives of the Runoff and Transport Block case studies are
to model the existing sewer system and to identify areas where the system
may be inadequate. A ten year return period, two hour duration design
storm will be utilized. Derivation of this design storm is discussed in the
Study Guide Section entitled Case Study Materials directly following a
description of the Runoff Block.
The Runoff/Transport basin drains into a larger 3000 acre basin,
the output of which will be utilized in the Storage/Treatment and Receiving
Water Block case studies, Evaluation of the effects of treated and
untreated combined sewer overflows on receiving waters is the objective
of the Storage/Treatment and Receiving Water Block case studies.
^Research Assistant, Assistant Professor, Associate Professor, Professor,
respectively, Department of Civil Engineering, University of Massachusetts/
Amherst.
91
-------
c
INSERT 1: TOPOGRAPHIC
Scale: 1 inch
FIGURE IA, TOPOGRAPHIC MAP OF CASE STUDY BASIN,
-------
c
-
ELECTRONICS ASSEMBLY PLANT
-
Scale: 1 inch = 360 feet /
FIGURE IB, TOPOGRAPHIC Map OF NORTHWEST QUADRANT OF CASE STUDY BASIN,
-------
-
INSERT!: TOPOGRAPHIC
uw/a
FIGURE Ic, TOPOGRAPHIC M\p OF NORTHEAST QJADRANT OF CASE STUDY
BASIN,
-------
-•
o
B.SEBAU Scale: 1 inch = 360 feet D
FIELD «°
FIGURE ID, TOPOGRAPHIC MAP OF SOUTHEAST QUADRANT OF CASE STUDY BASIN,
-------
-c
Scale: inch = 360 feet
FIGURE IE, TOPOGRAPHIC MAP OF SourmcsT QUADRANT OF CASE STUDY BASIN,
-------
SEWER MAP
730'
200
Cl.O
1.00%
VO
-J
SCALE: 1 inch = 400 feet
E = EGG SHAPED
C = CIRCULAR
FIGURE 2, SEWER MAP FOR CASE STUDY BASIN,
-------
RUNOFF BLOCK STUDY GUIDE
PREPARATION OF RUNOFF BLOCK DATA
Considerable time and effort are required to gather and prepare the
necessary data to run the Storm Water Management Model. The following may be
of some assistance in identifying sources of these data. Municipal Department
of Public Works (DPW) or Engineering Office files will provide the bulk of
the data that is readily available. Most subcatchment data can be derived
from up-to-date, large scale (1:200 preferrable) topographic maps. If these
maps are not available or are inaccurate, appropriate aerial photos may be
used; particularly for the determination of land use and percent impervious
area. In very flat areas, sewer maps may be helpful in delineating subcatch-
ments. Land use and zoning maps are also useful.
During data preparation for the Runoff Block, reference must be made to
the existing sewer system. Most sewer data can be derived from sewer plates
available in the DPW files. Some municipalities may have a comprehensive
sewer system map that will assist greatly in coordination between Runoff and
Transport Block data preparation.
If available data is insufficient, on-site inspections to check recent
changes in land use and local topography, establish surface elevation, or
inspect the condition and configuration of the existing sewer system may be
necessary.
Preliminary Data Preparation
Preliminary data preparation for the Runoff Block include constructing a
composite topographic map for the basin to be modeled, overlaying the major
portion of the existing sewer system on the composite map and delineating the
outer basin boundary. Topographic maps used should be large scale (1:200 is re-
commended) and up to date. Individual map sheets can be assembled to yield
one topographic map of the basin. Data can be extracted from the sewer system
map to overlay the major sewer trunk lines on the topographic map. Special
attention should be given to including all of the sewer lines near the
periphery of the basin. Finally, the outer basin boundary can be delineated
from the composite topographic map, taking into account the sewer system lay-
out. Knowledge of basic map reading techniques is necessary to establish
basin boundaries. Additional information on topographic map interpretation
can be found in Reference 1. Figure 3 illustrates the result of preliminary
data preparation; i.e., the arrangement of the sewer system within the basin
on the composite topographic map.
Discretization of Basin Into Subcatchments
Discretization can be defined as the division of the basin into overland
flow areas and the sewer system into discrete elements. A decision must be
98
-------
Composited Topographic Maps
Outer Basin
Boundary Vv
FIGURE 3. PRELIMINARY DATA PREPARATION,
made as to how many subcatchments should be included In the basin. Finer
discretization grids (more subcatchments in a given basin) more accurately
represent the actual rainfall-runoff process. Because computers have a
limited memory capacity, a practical limit has to be placed on the number
°; subcatchments or sewer elements that can be modeled. With SWMM, a maximum
of 200 subcatchments can be modeled in Runoff and 160 sewer elements in
Transport. Factors to donsider when determining the approximate number of
subcatchments include the size of the basin, variations in land use within
the basin, the degree of accuracy required (planning study vs. design study)
and the method of calibration to be used. Funds available for the study must
also be taken into account because data gathering and program execution costs
will increase with the number of subcatchments modeled.
The choice of a discretization scheme must be based on engineering judg-
ment. In order to select an appropriately fine discretization, it may be de-
sirable to first apply two or more discretization grids to a small, represent-
ative basin. The small basin can then be used to calibrate the model; i.e.
the most efficient discretization scheme should be selected and values chosen
for the required input parameters to be used in modeling the remainder of the
urban area. Model calibration will be discussed further in the Transport Block
section.
Subcatchment Data Preparation
Once the outer basin boundary has been delineated and the approximate
number of subcatchments fixed, the next step is to divide the basin into
subcatchments. Topographic divides will again be used; however, it is now
increasingly important to consider carefully the sewer system layout. For
example, flow in the sewer system may be in a different direction from that
of overland flow. Thus, water will be conducted in the opposite direction
99
-------
from that indicated by the land surface slope. Consideration must also be
given to the location of the entry point of flow from the modeled subcatch-
ment to the modeled sewer system; i.e., all of the flow from a large sub-
catchment should not be routed through a small sewer if this does not actually
occur in the field. Placement of entry points may also have some effect on
the position of subcatchment boundaries.
Subcatchments should be drawn to represent as closely as possible a
homogeneous land use. Irregular subcatchment shapes should be avoided because
the SWMM depicts overland flow in subcatchments as if the subcatchment was a
rectangular plane of uniform slope and roughness. Unrealistic overland flow
conditions can result if subcatchment geometry differs substantially from
that assumed by the model.
Subcatchments should be arranged to reflect the actual field conditions
as closely as possible. The total runoff from the basin, however, is much
more sensitive to accurate outer basin boundary delineation than to the place-
ment of subcatchments within the basin. This is especially true for large
basins with a large number of subcatchments. Inaccuracies in subcatchment
placement tend to cancel each other out, while errors in outer boundary
delineation directly affect the total rainfall volume to be modeled.
Data that must be obtained for each subcatchraent include area, slope,
characteristic width, percent impervious area and land use. Preparation
of each of the elements of data will be discussed. Other data required by
the model, but not usually available, are overland flow resistance factors,
retention depths and infiltration rates. To compensate for these deficiencies
in data, the model has pre-programmed default values that are assumed if no
data are supplied. The default values are recommended unless data is avail-
able or engineering judgment warrants otherwise.
Area of Subcatchment - Area of subcatchments can be measured with a planimeter.
The planimeter readings can be scaled to give area in acres by first measur-
ing a known area drawn to the same scale as the map to be used. The
difference in planimeter readings at the start and finish divided by the
known area will give a scaling factor of planimeter units/acre, designated
as B. All subsequent planimeter results can be divided by this factor to
give area in acres. Equations 1 and 2 show the computation and use of the
scaling factor, respectively.
B = (R.-RJ/A. (1)
where R, = planimeter reading at start of trace
R» = planimeter reading at end of trace
A, = known area scaled to map (acres)
B = scaling factor (planimeter units/acre)
100
-------
A = (R1-R2)/B
where A = area of subcatchment (acres)
(2)
Subcatchment Slope - The subcatchment slope used for model input should be
the best estimate of the average slope of the subcatchment. Accuracy of
this estimate will be increased if subcatchments do not contain complex
slopes; i.e., subcatchments should be constructed so that flow-is over ground
of fairly uniform slope. In many instances, this may not be practical. A
method of calculating an area-weighted-average slope, is illustrated in
Figure 4. This method, if used properly, will give a good estimate of the
average subcatchment slope for complex cases. The steps involved in
Subcatchment
Boundary
FIGURE 4, DETERMINATION OF WEIGHTED SLOPE, Sw.
101
-------
calculating the weighted average slope are as follows;
1. Determine the line of maximum depression through the subcatchment
(see Figure 2). This line is analogous to the position of a free
flowing stream if one existed in the subcatchment,
2. Divide the line of maximum depression into equal increments of
length b. The number of increments to be used is based on engineer-
ing judgment and depends on the complexity of the subcatchment,
3. Draw perpendicular bisectors through each of the increments chosen.
Extend these bisectors to the subcatchment boundary.
4. Measure the distance, X., along, each bisector from the line of
maximum depression to tne subcatchment boundary. Measure the
difference in elevation, AZ., between the line of maximum depression
and the subcatchment boundary*
5. Compute the weighted slope, S , as shown by Equation 3;
Sw ' I AV I Xi (3)
W 1-1 ± i-1 1
where n - total number of areas formed.
The derivation of Equation 3 is given below:
b X S b X S b X S
S
w n n n
b I xi b I X b I X.
* *
X«S_ X S
-L-2 + ... + _n_n
n n
I x± I X
i-1 i l-l 1
AZi - XiSi (5)
102
-------
Characteristic Width of Subcatchment - The characteristic width of a
subcatchment is the distance over which overland flow leaves the subcatch-
ment surface and enters the main drainage conduit passing through the
subcatchment. This conduit may be a sewer element, a gutter or a natural
channel. Figure 5 illustrates the meaning of characteristic width.
The SWMM User's Manual recommends that twice the length of the main
drainage conduit be used as an approximation of the characteristic width.
This works well if subcatchments are symmetrical about the main drainage
conduit as is shown in Figure 5. Some amount of error would be introduced,
however, if this approximation is used to model subcatchments in which the
drainage system is located to either side of the subcatchment, as shown in
Figure 6(a) .
Analysis of an idealized 100 percent impervious subcatchment has indi-
cated that the p^ak runoff from a modeled subcatchment with W - 2L as
shown in Figure 6(b), can be as much as 20 percent higher than from a sub-
catchment modeled with W = L, as shown in Figure 6(c). Thus, it is recom-
mended that the characteristic width be taken as L when the drainage conduit
is skewed entirely to one side of the subcatchment .
For configurations that are between the two extremes, a relatively
simple method has been developed that accurately accounts for the position of
the main drainage conduit within the subcatchment (see page 56, Reference 2).
The method involves calculating a skewness factor S which is an area-weighted
measure of the position of the principle drainage conduit or canal with
respect to the centerline axis of the subcatchment. The equation for the
skewness factor, S, is
S = (A-B)/(A+B) (6)
where A and B represent the areas to either side of the drainage conduit as
shown in Figure 7 . The equation for calculating the corrected characteristic
width is given by
W = (2-S)L (7)
where L is the length of the drainage conduit. The skewness factor as a
function of W/L is shown in Figure 8.
Percent Impervious Area - The Runoff Block output is highly sensitive to the
percent impervious area modeled and this factor is difficult to estimate
accurately because some runoff from impervious areas may flow onto, pervious
areas, infiltrate and thus not contribute to the storm runoff. Therefore, it
is recommended that an initial study of topographic maps and aerial photographs
be made to estimate the actual percent impervious area with the understanding
that this parameter may need adjustment during the calibration stage. Table 1
103
-------
Direction of
Overland Flow
Main Drainage
.Conduit Through
Subcatchment
L = Total Length of Main Drainage Conduit
W = Characteristic Width = 2L = Total Width of Overland Flow.
FIGURES. CHARACTERISTIC WIDTH,
(a)
Actual Subcatchment
(b) (c)
Subcatchment as Modeled Subcatchment as
Using W = 2L Modeled using W
Overland
Flow
A/2
A/2
Drainage Conduit
A = Area.
FIGURE 6. SUBCATCWENT WITH SKEWED DRAINAGE CHANNEL,
104
-------
Overland Flow
A & B are the areas
on either aide of the
drainage conduit.
Drainage
Conduit
FIGURE 7, CALCULATED CHARACTERISTIC WIDTH,
105
-------
2.0
1.8 .
1.6 ~
W/L
1.4
1.2 -
1.0
0 0.2 0.4 0.6 0.8 1.0
Skew Factor
FIGURE 8, W/L Vs, SKEW FACTOR,
106
-------
Table 1. PERCENT IMPERVIOUSNESS AS FOR VARIOUS LAND USES IN THE SAN FRANCISCO
AREA (Reference 3)
Type of development
Density
in units
per acre
Percent impervious
Santa Clara
County
San Francisco
Bay Region
Residential:
Hill areas
Low urbanization
Medium urbanization
Heavy urbanization
(apartments)
Industrial:
Nonmanufacturing
Manufacturing
Reserve
Commercial
Transportation
Public buildings
Public parks
Agricultural
Natural watersheds
0.5- 2
3-6
7 -10
11 -20
6
10
20
32
8
15
25
40
50
40
20
50
70
40
12
4
2
60
50
25
60
75
50
12
4
2
-------
taken from Reference 3 may be used as a guide in estimating impervious area.
Land Use Designation - The model predicts runoff water quality based on one
of five land use designations: single family residential, multi-family
residential, commercial, industrial or undeveloped land. As much as possible,
subcatchments should be arranged so that each represents a single land use.
When this is not possible, the prevailing land use should be assigned to the
subcatchment. Zoning maps and land use maps will be helpful in determining
land use; however, up-to-date checks by area reconnaissance and/or discussions
with local officials may be necessary.
Gutters and Pipes - Up to 200 gutters or pipes can be modeled in the Runoff
Block. These are in addition to the 160 sewer elements that can be modeled
in the Transport Block. Use of gutters and pipes should be considered supple-
mental to the transport system, either to extend the number of elements
modeled beyond 160 or to model conduits such as street gutters or natural
channels that cannot be modeled in the Transport Block. The Transport Block
should be used in situations where: 1) backwater effects are significant;
2) hydraulic elements other than pipes and gutters (such as pumps or flow
dividers) are used, and 3) solids deposition or suspension is important.
To decide which elements to model as gutters and as pipes, the extent of
the transport system modeled in the Transport Block must be estimated.
Generally, all trunk sewers and most branch sewers three feet in diameter or
larger should be modeled in the Transport Block. Smaller branch lines,
gutters and natural channels can be modeled in the Runoff Block. Some adjust-
ment of the number of gutters and pipes modeled in the Runoff Block may be
desirable after the transport system has been completely descritized. When
a choice exists, an element should be modeled in the Transport Block because
of its more accurate flow routing procedure.
All of the flow through a modeled gutter or pipe enters at the head end;
i.e., there is no differential entry along the length of the element. The
same is true in the Transport Block but is less critical because of the
larger elements and flows which are modeled.
Data required for gutters and pipes include length, width or diameter,
invert slope and Manning's roughness factor. Special data required only
for gutters are side slopes and depth when full. Most data will be available
on sewer plates for pipes. Data for gutters or natural channels, on the other
hand, will have to be gathered in the field. Some surveying may be necessary
to ascertain slopes. Typical values of Manning's roughness coefficients
are found in Table 5-
Placement of Subcatchment Flow Outlet Point - All of the flow from a subcatch-
ment must enter either a single manhole or a single gutter/pipe. It is
important that this point of entry in the modeled system be located approxi-
mately where most of the flow would enter the actual system. In other words,
even though flows in the actual subcatchment may be entering the sewer system
at several manholes, the model must route all of the flow through a single
point. Inaccurate location of the entry point will result in errors in the
108
-------
modeled routing time. Placement of the outlet point will also influence
which gutters or pipes should be modeled.
Network Arrangement and Numbering - Completion of Runoff Block data prepara-
tion provides a system of subcatchments, gutter/pipes and outlet manholes.
These outlet manholes are the points through which flows will later be
transferred to the sewer system modeled in the Transport Block. A typical
runoff block arrangement of numbered manholes and gutter/pipes is shown in
Figure 9 .
The numbers assigned to subcatchments, gutter/pipes and inlet manholes
are the only way the computer has of joining the whole system together;
therefore, care must be taken to avoid errors in assigning numbers to the
various elements. The following numbering procedure is recommended. Sub-
catchments should be numbered from one to the total number to be modeled.
Because a maximum of 70 manholes can be used to transfer flows to the
Transport Block, the numbers 1-70 should be used to identify outlet manholes.
Subcatchments and outlet manholes may be assigned the same numbers; however,
numbers should not be duplicated between manholes and gutter/pipes. Therefore
it is recommended that gutters and pipes.be assigned numbers between 201 and
400. This numbering system is illustrated in Figure 9.
As shown in Figure 9, gutter/pipes may be arranged in tree structures;
however, all such trees must eventually terminate at an outlet manhole.
For example, gutter/pipes 203 and 204 flow into gutter/pipe 202 which in turn
flows into manhole 2. The overland flow from subcatchment 2 could either
enter the head of gutter/pipe 202 or manhole 2, whichever more accurately
represents the physical situation.
Quality Data - Data required for surface quality computations are included in
card groups 9-11 of Table 5-7, pp. 77-78, User's Manual (reference 6). The
number of dry days prior to a storm can be determined from U.S. Weather Bureau
rainfall records or local records if available. Street cleaning frequency,
number of street sweeper passes and catchbasin storage volume can be obtained
from municipal Department of Public Works records. In the absence of measured
BOD concentrations in catchbasins, the recommended value of 100 mg/1 should
be used.
The SWMM provides a choice of two methods for calculating suspended
solids generation from subcatchments. ISS = 0 is an exponential decay
function and ISS = 1 is a regression equation. No recommendation is given
in the User's Manual as to which method is better. Probably the best approach
would be to compare both methods to measured SS data during the model cali-
bration stage and choose the closer approximation.
The erosion portion of the quality subroutine is discussed thoroughly
•in the User's Manual and will not be included here (Reference 6). Other
data required are land use classification (as discussed previously), the
total number of catchbasins in each subcatchment (estimated from sewer maps
or field surveys) and total length of all gutters within each subcatchment
(scaled from the topographic map).
109
-------
Basin Boundary
I*/
Subcatchment
Boundary
1 \= Subcatchment
1
o
or)-!
——- Gutter/Pipe
Numbers
1-200
1-70
201-400
FIGURE 9, TYPICAL RUNOFF NETWORK ARRANGEMENT,
PREPARATION OF CODING FORMS
It is recommended that all Subcatchment and gutter/pipe data be
tabulated on forms similar to those supplied with these case study materials,
This will make transfer of data to the computer coding forms much easier.
Standard. FORTRAN coding forms should be used to prepare the data as shown
in Table 5-7, pp. 74-79 of the User's Manual,
CASE STUDY MATERIALS
General Description
The^objectives of the case study are to model an existing sewer system
in a typical urban drainage basin and to identify possible surcharging in
the system. The initial task is to complete the descritization and data
preparation for the Runoff Block. The basin (Figure 1) should be subdivided
into ten or less subcatchments for illustrative purposes. Although this is
a relatively small basin, some gutters and pipes should be included in the
Runoff Block. Data for pipes can be obtained from the sewer map supplied with
the Transport Block Case Study materials. Data for gutters can be scaled
from the topographic map. Zero side slope for gutters and 12" depth when full
are to be assumed. Streets are constructed of smooth asphalt.
110
-------
Table 2 may be used as a guide to organizing the various tasks. Tables
3 and 4 have been provided for data tabulation. The columns are arranged in
the order the data will be transcribed to computer coding forms.
Design Storm
The storm to be used in the case study is a ten year, two hour duration
storm. Data for this storm were taken from an actual point frequency
analysis of rainfall in Cleveland, Ohio (4), A ten year return period was
chosen because of the commercial and industrial land uses within the Case
Study basin. Figure 10, also taken from Reference 4, has been used to convert
the total volume of rainfall (2.10 inches) into a hyetograph. This hyetograph
will be used for program input and is shown in Figure 11,
Additional Data
The following additional data are provided.
1. Number of time steps (NSTEP): 50 should be sufficient.
2. Hour of start of storm (NHR): 12
3. Minutes of start of storm (NMN)• QO
4. Integration period (DELT); 5,0 minutes seems to be a good
compromise between accuracy and raingauge data gathering
capability.
5. Number of hyetographs (NRGAG): 1
6. Number of data points for each hyetograph (NHISTO); 24
7. Time interval between values (THISTO): 5.0 min.
8. Number of dry days (DRYDAY): 14 days
9. Street cleaning frequency (CLEFREQ): 21 days
10. Number of street sweeper passes (NPASS); 2
11. Catchbasin storage volume (CBVOL): 18 ft3
12. BOD concentration in catchbasins (CBFACT(4)): 100 mg/1
13. Number of catchbasins in subcatchment (BA): Assume 1
catchbasin/acre.
Ill
-------
Table 2, RUNOFF BLOCK DATA PREPARATION CHECK LIST
1. Composite Topographic Map
2. Sewer System Overlay
3. Outer Basin Boundary
4. Descritization of Basin into Subcatchments
5. Subcatchment Data Preparation
a. Area
b. Slope
c. Characteristic Width
d. Percent Impervious Area
e. Overland Flow Resistance Factors, Retention Storage
Depths and Infiltration Rates, when available,
f. Land Use
g. Gutters and Pipes
h. Placement of Subcatchment Flow Outlet Point
i. Network Arrangement and Numbering
j. Quality Data in Addition to Land Use.
6. Preparation of'Coding Forms
112
-------
Table 3. SUBCATCHMENT DATA
Quantity
Subcatchment
Number
Gutter or Manhole
for Drainage
Width
(ft)
Area
(Acres)
Percent
Impervious
Ground
Slope
(ft/ft)
QUALITY
Land
Use
Number of
Catchbasins
Total Length
of Gutters
-------
Table 4, GUTTER DATA
Gutter/
Pipe
Number
Gutter or
Inlet
Number for
Drainage
1 = Gutter
2 = Pipe
Bottom Width
of Gutter or
Pipe (ft)
Diameter (ft)
Length
(ft)
Invert
Slope
(ft/ft)
Left
Side
Slope
(ft/ft)
Right
Side
Slope
(ft/ft)
Manning ' s
Coefficient
Gutter
Depth
when Full
(in.)
-------
Curve represents average of all rainfalls studied.
CO
c
•H
OA
o
H
100
90
80 .
70
60
50 -
I 4°
-------
2.5
I I I I I
1 1 I I I I I I I I I
2.16
2.0-
1.68
JC
~c
(0
c
0>
1.5
1.0-
0.96
CO
os
0.5'
0.
1.92
1.44
1.08
Lp.96
I °-84
0.48
1.36
0.24
IlllBlillllllllllllllll
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120
Time from Start of Storm, Minutes
FIGURE II, RAINFALL HYETOGRAPH,
-------
SAMPLE DATA INPUT AND PROGRAM OUTPUT: RUNOFF BLOCK
The suggested solution to the case study depicts the given basin as
ten subcatchments with seven gutters or pipes. All pertinent input data
and program output for the Runoff Block Case Study are provided on the
following pages. The specific items of information included are:
Basin Map Showing Descritizatlon
Grid Used (Figure 12)
Sample Coding Forms (given for
Runoff Block only)
Runoff Block Card Data
Runoff Block Program Output
117
-------
-••
FIGURE 12, BASIN MAP SHOWING DECRITIZATION GRID USED
-------
FORTRAN Ceding Fa
V7327-«U/M050'
Printed mUS A
"°0~ S^W/f M
MOCUMMCI I 0*11
*;NCt«iN&
«V>" Xl'O —
C .it*.(C
*_.^H
^ ^5
>:*•«•>- tttfm- «•
-------
FORTRAN Drift* F«m
GX28-7327-6U/MOSO-
Printed mU.S A.
g W A) M
I StAKMNI
9 NUMM*
FORTRAN! STATEMENT
, > . io~r~
!• I* 30 71 ?? ?1 ?* M H 77 ?» fg » 3' 3? M 3* li 3* 37 3e
J7.o^
.0*1
o.
0*3
Of
£ttx!L
S
±
***&
43.0
S2.J
.034
ff
39-1
•ot*
£tA
46
-OJ3
&.7
01 o
.10
13. "7
£&'
-------
O«8-7JS7-eU/MOSO
PrtrrtdtnU-SA
-------
NJ
0 10 10 9
1 2 3 % 7
RUNOFF
RUNOFF BLOCK WORKSHOP
CASE STUDY BASIN* 10
RUNOFF BLOCK CARD DATA
YEAR - Z HOUR DURATION DESIGN STORM
C
1
0
1
2%
.29
.92
• %8
1
1
1
1
1
1
1
1
1
1
1
50
f.O
0.60
1.92
g.%8
50
51
52
53
5%
55
56
1
2
3
k
5
ft
- 7
ft
9
10
1
1
2
3
%
5
6
7
8
9
10
1
1200 5.0
9.8% 0.8%
1.%% 1.%%
1.36 0.2%
13 1
7 1
1% 2
9 1
3 2
5 2
5 1
561856.
511900.
52%%0fl.
122606.
53170C.
83%OP.
61600.
5%5109.
952%30.
56135C.
1%.
1
1
1
2
2
3
3
1
%
3
1
0.96 1.32
1.08 0.96
36.0
78.0
2.5
30.0
3.0
s.o
70. a
%3.0 25.
13.1 20.
36.1 15.
26.9 75.
%2.1 %0.
38.1 95.
19.7 %e.
21.7 tO.
50.5 75.
13.7 %C.
21.
%3.0
13.1
36.1
26.9
%2.1
31.1
19.7
21.7
50.5
13.7
1.%% 1.68 2.16 2.16
0.8% 0.8% ¥.%• K.%8
11CO. .ull °.
18CC. .J*9 0.
123C. .003
1560. .'C7 0.
120t. .010
10AO. .002
12PO. .P12 0.
.023
.011
.015
.621
.02%
.01%
.013
.610
.013
.03%
2 18.
39.5
%7.5
55.0
32.0
200.0
5%. 8
25.5
150.0
65.0
tC.5
CAKD
GROUP
-2
•3
1-4
C.
C.
.013
.013
• R13
.013
.013
.117
12.
12.
12.
12.
-7
100.
-9
-11
-12
-13
-------
ENVIRONMENTAL PROTECTION AGENCY - STORH MATER NANAGEMENT MODEL »»» RELEASE II •»•
DEVELOPED BY
SEPTEMBER 1970
HETCALF * EDDY, INC.
UNIVERSITY OF FLORIDA
NATER RESOURCES ENGINEERS* INC.
UPDATED BY UNIVERSITY OF FLORIDA
FEBRUARY
THIS IS A HEM RELEASE OF THE SHHM. IF ANY PROBLEHS
OCCUR IN RUNNING THIS MODEL PLEASE CONTACT HAYNE
HU3FR OR ALAN OELTZ AT THE UNIVERSITY OF FLORIDA.
PHONE l-90i»-392-08«f6
TAPE 0? DISK ASSIGNMENTS
JIN(1)
0
JOJTdl
10
JINC2)
10
JOUT(2)
NSCRATdl
1
JINI3I
-0
JOUTC3)
-0
JINUI
JOUT«(»)
-0
NSCRAT<2»
2
JIN(5)
-5
JOUT<5>
JINC6)
-0
JOUTC6)
-C
NSCRATC3)
3
JINI7)
-J
JOUT«7>
-D
JINC8I
-0
JOUT(8)
-0
"SCRJTW
JINC9I
-0
JOUTC91
-0
JINdOl
-0
JOUTtlO)
-0
NSCRAT(5)
7
-------
RUNOFF BLOCK MORKSHOP
CASE STUDY BASIN, 10 YEAR - 2 HOUR DURATION DESIGN STORM
fo
BASIN NJMBER 1
NUMBER OF TIME STEPS 50
INTEGRATION TINE INTERVAL (MINUTES)* 5*00
25.0 PERCENT OF IMPERVIOUS AREA HAS ZERO DETENTION DEPTH
FOR Zk RAINFALL STEPSt THE TIME INTERVAL IS 5.00 MINUTES
FOR RAINGAGE NUMBER 1 RAINFALL HISTORY IN INCHES PER HOUR
.
1.
GUTTER
NUMBER
1
2
3 *
t*
5 »
6 •
7
29
92
53
51
52
5^
5<»
55
56
.60
1.92
.1*8
Gil T T p n
U T T E R
GUTTER
CONNECTION
13
7
1*»
9
3
5
5
• 8%
!.<»%
.36
Ak| M D T
N D PI
WIDTH
CFT)
30.0
70.0
2.5
30.0
3.0
5.0
70.0
• 6<»
!.*<»
.2k
Pt* n A
E DA
LENGTH
-------
SUBCAT(H-
MENT
1
2
3
<•
5
6
7
)
9
19
NO.
t
Z
3
*
5
6
7
9
9
13
S 'J B C
GUTTER
01 INLET
50
51
"?
12
53
a
6
5*
55
56
A T C H
MTOTH
JFT)
185C.O
1990.0
M.00.0
Z60C.O
1700.3
3I.OC.O
16it.J
510C.O
Z*00.u
1350.9
M E N T DATA
AREA PERCENT
(AC)
*3.
13.
36.
27.
<»2.
31.
ZC.
ZZ.
51.
!«..
IMPERV.
25.)
ZO.O
15.0
75.0
40.?
95.0
<•Ci.it
<>i.O
75.0
fcC.J
SLOPE
(FT/FT*
.C23
.011
.15
.'3Z1
.r-zt.
.:iu
.;i3
.QIC
.:i3
.031*
TOTAL NUMBER OF SUBCATCHNENTS, 10
TOTAL TRIBUTARY AREA (ACR?SI«
RESISTANCE FACTOR
IMPERV. PERV.
SURFACE ST09AGEIINI
IMPERW.
PERV.
INFILTRATION
PATECIN/HR)
MAXIMUM MINIMUM
DECAY RATT GAGE
(1/SEC) NO.
.913
.013
.013
.01?
.013
.013
.013
.01?
.113
.013
.350
.?5U
.Z50
"""" .351
.?50
.250
.250
.2*0
.250
.Z50
.062
.062
•9.62
.06Z
.J6?
.062
.162
.062
.062
.162
.181,
.18%
•18*
.181.
.181.
.1S<»
.led
.181,
.lad
.141,
3.03
3. Ob
3.00
1.00
3.0C
3.0C
3.03
3.00
3.0C
3.00
.52
.5?
.52
.52
.52
.52
.52
.52
.52
.52
.00115
.00115
.00115
' TOO 115
.DC 115
.00115
.00115
.00115
.86115
.0011?
1
1
1
1
1
1
1
1
1
1
Ul
QUALITY SIMULATION INCLUDED IN THIS RUN
INPUT PARAMETERS AS FOLLOWS
NUMBER OF CONSTITUENTS 4
NUMBER OF DRV DAYS 1"».0
STREET CLEANING FPEO 21.0 DAYS
PASSES PER CLEANING 2
STO CATCHBASIN VOLUME 14.00 FT3
CATCHBASIN CONTENTS BOO 100.0 MG/L
METHOD FOR CALCULATING SS.
SPECIAL TECHNIQUE.
SAME AS IN ORIGINAL
RELEASE I OF THE SMMM.
ISS - 1
-------
WATERSHED QUALITY DEFINITIONS
SUBAREA
NUHBER
LAND USF TOTAL GUTTER
CLASS. LENGTH*10»»2 FT
1
2
3
5
6
7
8
9
10
1
2
3
5
6
7
8
9
lu
1
1
1
2
2
3
3
1
39.50
J>7.5"
55.00
32.00
200.00
5<*.on
Z5.50
150.00
65.J"
<»0.50
NUMBER OF
CATCHBASINS
<»3.0G
13.10
36.10
26.90
38.10
19.70
21.70
50.50
13.70
HVOROGRAPHS MILL BE LISTED FOR THE FOLLOWING 1 GUTTERS OR INLETS
GUTTER
GUTTER
GUTTER
GUTTER
GUTTER
GUTTER
GUTTER
GUTTER
GUTTER
* SURCHARGED*
3 SURCHARGED,
3 SURCHARGED,
3 SURCHARGED,
? SURCHARGED,
3 SURCHARGED,
3 SURCHARGED,
3 SURCHARGED,
3 SURCHARGED,
SURCHARGE=
SURCHARGED
SURCHARGES
SURCHARGES
SURCHARGE=
SURCHARGES
SURCHARGED
SURCHARGES
SURCHARGES
1053.
221
-------
RAINFALL HTETOGMPM
•ASIN NO 1
2.911
Z.Mt
l.MI
«t INFILL
IN
IN / H*
l.Ofl
.SOI
I.Oil
12.9 13.0
1..
13.5
.. 1 —i-........i .....•••••......!.........i
14.9 IS.t 15.5 16.0 16.S IT.I
TINE IN HOURS
RAIN6AGE LEGEND
-------
INLET HVOROGR4PH
RUNOFF
K- »
00
CFS
530. JOO
*09. 000
303.0(0
201.000
10P.OCO
3.000
12
I
I
I
I
I
I
I
I
I
T
-
I
I
I
I
I
I
I
I
I
-
I
I
I
I
I
I
I
I
I
-
I
T
I
I
I
I
I
I
I
-
I
I
I
I
I
I
I
I
I
**'
.0
•
*
• •
• ••
» •
• »
* •
•
»•*•»•••
—I I I I I »•••••••••.••••••••••••• .x :
12.5 13.C 13.5 !<>.v It.5 15, C 15.5 16.0 16.5 17.0
IN HOURS
-------
RUNOFF BLOCK WORKSHOP
CASE STUDY BASIN, II TEAR - 2 HOUR DURATION DESIGN STORK
SUMMARY OF QUANTITY AND QUALITY RESULTS AT LOCATION
FLON IN CFS AND QUALITY IN NG/L (AND COLIF IN NPN/LI
VO
Tine
FLOW
MO
sus-s
COLIF
COO
SET-S
MIT
P0
It9.31
106.62
76. tS
55. tt
ti .69
341.33
22 .9t
17.61
13.71
10.62
8.65
7.00
$.73
t.7t
3.97
3.36
2.86
2.1,7
Z.lt
1.67
1.65
I.t6
1.30
.226E»07
.125E»07
.129E»07
.lt2E»07
.lt6E»07
.119E»07
.881E»06
.591E»06
.3t2E»06
.10SE»*6
.9S1£»05
.t«3E»OS
.ZZ5E»05
.U7E»«5
.650E»Ot
.399E»Ot
.2S3E»8t
.169E»Ot
.125E»Ot
.97tE+«3
.726E»03
.57tE»03
.t8tE*03
.t29E»03
.t09E*03
.397E»«3
.35*E»03
.3Z8E»03
.310E»I3
.298E»03
.20ftE»03
.ZS1E»03
.275E«03
.27JP»03
.267E»03
.263E»03
.260E*03
.Z58E+03
.2S6E»03
.25tE»03
• 252E*03
. Z51E+03
.250E»03
.2t9E»03
.2t8E»l>3
.Zt7E«03
.2t6E»03
.2t5E»03
.2ttE*03
.2t3E»03
2t2.60
1Z5.V9
90.78
t6.56
33.95
27.3?
2t.2tf
13.56
7.85
t.2t
2.18
l.«t
&
.27
.15
.09
.06
.Ot
.03
.02
.02
.01
.01
.11
.11
.01
.81
.01
.01
.01
.01
.01
.01
.*!
.01
.01
.01
."1
.01
.01
.01
.01
.01
.01
.01
.01
.01
.01
.01
.81
1.72
2.09
3.13
11.17
32. tt
57. 3t
76.30
51.23
29,66
16.00
8.2t
3.93
1.95
1.02
.56
.35
.22
.13
.08
.05
.03
.02
.02
.02
.01
.01
.00
.on
.00
.00
.00
.Ok
.00
• OJ
.00
.00
.00
.Ofi
.00
.00
.00
.00
.00
.00
.00
.Ofi
.00
.00
.00
.00
1.30
2.90
22.63
39.67
29.71
16.00
2.65
2.33
1.97
1.77
.71
.61
.56
.50
.t2
.to
.32
.63
.20
10.83
11.60
11. Ot
10.69
11.05
12.23
12.21
9.30
6.72
t.80
3.tt
2.t9
1.83
1.36
1.03
.79
.62
.t9
.39
.31
.26
.21
.18
.15
.13
.11
.10
.08
.07
.07
.*6
.18
.32
2.28
•».99
3.00
1.62
.28
.2t
.Zt
.18
.17
.16
.16
.15
.It
.It
.13
.66
.83
1.00
1.16
1.10
1.07
1.11
1.22
1.22
.93
.67
.to
.3t
.25
.1H
.It
.U
.08
.06
.05
.Ot
.03
.03
.4)2
.02
.02
.01
.01
.01
.01
.01
.01
.01
1.33
.73
.76
.83
.86
.70
.52
.35
.Zf
.11
.06
.03
.01
.01
.00
.00
.08
.00
.BO
.10
.00
.00
.00
.00
.an
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.no
.60
.CO
.00
.00
.«0
.00
.CO
.00
.00
.00
-------
TRANSPORT BLOCK STUDY GUIDE
PREPARATION OF TRANSPORT BLOCK DATA
Sewer System Discretization
The Transport Block models the actual sewer system as a series of links
(conduit elements) and nodes (non-conduit elements). Flow and area data for
13 standard conduit shapes are included in the program and the user has the
option of supplying data for up to three additional shapes. Non-conduit
elements include manholes, lift-stations, flow dividers, storage elements
and backwater elements.
A maximum of 160 sewer elements can be modeled in the Transport Block.
Because all conduit elements modeled must be joined by non-conduit elements,
the practical limit of the number of conduit elements that can be modeled
is 80. In most practical applications, not all of the actual sewer system
elements can be modeled. The user has the choice of aggregating sewer ele-
ments or using more than one simulation to model the transport system and
then aggregating the results using Subroutine COMBINE of SWMM. Unless the
system is large or complex, the best approach is to combine elements and
reduce the total number modeled to below the permissible maximum.
The approximate extent of the transport system has been established
previously by selecting the location of manholes through which flows will
be transferred from the Runoff Block to the Transport Block. Discretization
of the sewer system involves dividing the system into a series of elements
that will adequately protray all significant changes in conduit geometry,
roughness and slope, and allow for any flow control structures. Constrictions
in the sewer system, rough elements or elements with flat slopes should be
modeled as separate elements because these are the locations where surcharging
is likely to take place. Conduits having similar characteristics can be
modeled as one element, provided that the flow capacity of the aggregate
element is approximately equal to the largest capacity among the grouped
elements. Flowrate, Q, as given below by the Modified Manning's Equation
is most sensitive to changes in area (geometry), less sensitive to roughness
and least sensitive to slope.
130
-------
1.49 A5/3 _ 3y_ v 9v.l/2 (8)
So 8x ~ g 8x'
where Q = flowrate (cfs)
n = Manning's roughness coefficient
2
A - cross sectional area of flow (ft )
R, = hydraulic radius » A/PW (ft)
P » wetted perimeter (ft)
w
s = invert slope (ft/ft)
and — -r^: Terms from the St. Venant Equation added to
g Manning's Equation to correct for spacially varied
flow.
The sensitivity of discharge to these parameters should be taken into account
when deciding which elements to combine.
In order to account for all subcatchment runoff, all manholes through
which flows will be transferred from the Runoff Block have to be modeled.
Transport Block Numbering System
After selecting the number of model elements for the sewer system,
and marking the ends of these elements on the sewer map or sewer plates,
elements should be numbered in a manner similar to that used in the Runoff
Block. Inlet manholes through which surface runoff will enter the modeled
transport system must be assigned the same number as used in the Runoff Block.
Those non-conduit elements through which flows can be transferred from the
Transport Block to other blocks must be numbered 1-100. Therefore, it is
recommended that the numbers 1 to 100 be reserved for all non-conduit elements
to enable use of any as a transfer point. Conduit elements could be assigned
numbers ranging from 401 to 600 to avoid any confusion with gutters or pipes
modeled in the Runoff Block. Figure 13 shows this numbering scheme applied
to a typical transport system.
131
-------
Inlet Manhole
Nutiibors
1-70
Transport Manhole 1-100
Transport Element 40l.60()
(a) Transport System Alone
~~~ — — = Gutter/Pipe
(b) Transport System in Relation to
Runoff System.
FIGURE 13, TYPICAL TRANSPORT NETWORK ARRANGEMENT
132
-------
Compilation of Sewer Element Data
Once the sewer system has been divided into elements, the compilation
of the required data follows logically. Input data requirements for
conduit elements are detailed in Table 6-2 (p. 120) and card group 15 of
Table 6-6 (pp. 160-161), User's Manual (Reference 6). Generally these data in-
clude geometric dimensions, invert slope and Manning's roughness coefficient. A
current sewer master plan, if available, should contain most of the data
needed for conduit elements. If not available, individual sewer plates will
have to be assembled. In most cases, some data will have to be obtained
and/or verified by on-site inspection.
Invert slope is required in ft/100 ft which is the same as percent
slope. Manning's roughness coefficients are best estimated from in-system
photographs or on-site inspection. Table 5 is a partial listing of accepted
Manning's roughness coefficients for various materials. Reference 5 gives
a more complete listing on pages 110-113.
Input data required for non-conduit elements are outlined in Table 6-3
(p. 125) of the User's Manual. Additional data are required for storage
elements, as outlined in pages 162-166 of the User's Manual. Data sources
for non-conduit elements are generally the same as those for conduit
elements.
Table 7, provided with the Transport Block case study materials, will
be a useful aid in organizing sewer element data for coding form preparation.
It is desirable to arrange entries within this table in a reverse tree order
(from extremities of system to outfall) to assist in the correct assembly
of the sewer network.
Infiltration Model
Modeling of infiltration into the sewer system requires simultaneous
records of rainfall,, water table and sewer flow data for several
weeks. Some municipalities may already have compiled this data as part of
infiltration-inflow studies; where unavailable, field studies should be
undertaken at the outset of the storm water management study. Descriptions
of the infiltration model and the data preparation required are contained
on pages 133-142 of the User's Manual. Organization of. infiltration data
for coding form preparation is given in card groups 31-33, .Table 6-6
(pp. 167-168) of the User's Manual.
Dry Weather Flow Model
Modeling of a combined sewer system requires data for dry weather
(sewage) flow. Estimates of dry weather flow and pollutant concentration
parameters can be obtained from standard engineering handbooks, water usage
data or sewage flow data for the basin. Measured data can be taken at the
sewage treatment plant or preferrably gathered at several points within the
system. Sewage treatment plant records should also contain data on
industrial process flows originating within the basin. The U.S. Bureau
133
-------
Table 5. TYPICAL VALUES OF MANNING'S n
A. Closed Conduits Flowing Partly Full
1. Cast Iron
a. Coated 0.013
b. Uncoated 0.014
2. Corrugated Metal
a. Sub drain 0.019
b. Storm drain 0.024
3. Concrete
a. Culvert w/bends, connection's and some
debris 0.013
b. Unfinished concrete 0.014
c. Rubble Masonry 0.025
B. Lined or Built up Channels
1. Concrete
a. Trowel finish 0.013
b. Unfinished 0.017
2. Brick in cement mortar 0.015
3. Asphalt
a. Smooth 0.013
b. Rough 0.016
C. Excavated Channels
1. Clean earth (straight channel) 0.022
2. Earth with weeds (winding channel) 0.030
D. Natural Streams
1. Clean and straight 0.030
2. Weedy reaches, deep pools 0.100
134
-------
of Census tract information will contain data on population distribution,
family income and the number and age of dwelling units. Land use can be
determined as in the Runoff Block. Subareas modeled in Subroutine FILTH
do not have to coincide with runoff subcatchments.
Obviously, the best data available should be used for input into
Subroutine FILTH. The additional time and expense of gathering field data
within the system may be worthwhile if the first flush effects of combined
sewers are to be modeled.
The dry weather flow model is discussed in more detail on pages 142-
153 of the User's Manual. Card groups 34-45, Table 6-6 (pp. 168-175) of
the User's Manual detail the organization of dry weather flow data for
coding form preparation. Note that card groups 42-44 are only used when
measured sewage flow data are input. Table 8 will aid in the preparation of
subarea data.
PREPARATION OF CODING FORMS
Table 6-6, pp. 154-175 of the User's Manual details the proper organiza-
tion of input data for coding form preparation. (Coding forms are supplied
with the case study materials.) All I Format numbers 'rnuit b'e right justified
and all numbers must be contained within the field width assigned to them.
Careful coding form preparation will greatly reduce input data errors and
facilitiate card punching.
At this point, a connectivity diagram of the transport system should be
constructed to ensure a correctly numbered sewer system. Using the data
from the computer coding forms, the network should be reconstructed by
starting at the outfall. All elements should be checked to insure proper
connection. Elements at the extreme tree branches should all be inlet
manholes. If errors exist,the problem should be located by working backward
in the data preparation sequence.
CASE STUDY MATERIALS
General Discussion
A 25-year-old combined sewer system is to be modeled. Infiltration-
inflow studies have been conducted and will be included in the modeling.
A sewer system map for the case study basin has been provided (Figure 2).
This map gives lengths, dimensions and slopes of sewer elements. All conduits
are concrete and in good condition unless otherwise denoted on the Sewer map.
Minor pipes and house connections are not shown on,the sewer map; however,
it can be assumed that all buildings are connected to the sewer system. The
dimension given for egg-shaped conduits is the height.
Descritize the sewer system and prepare all necessary input data • For
illustrative purposes, the number of elements modeled should be limited to 30.
135
-------
Tables 6, 7, and 8 should be of some assistance in data preparation. Inlet
manholes to be modeled in Transport are available from the completed Runoff
Block case study. Additional data is supplied for infiltration and dry
weather flow. Some execution control parameters are also provided.
Infiltration Data
The following infiltration data are available concerning the case study
basin.
Base Dry Weather Infiltration: 80 gal/min
Groundwater Infiltration: 0.0
Rainwater Infiltration: 125 gal/min
Day of Year of Estimate: 280
Peak Residual Moisture: 100 gal/min
Average Distance Between Joints: 7.0 ft
Monthly Degree Days: Use data for Cleveland, Ohio from
Table A-l, User's Manual
Dry Weather Flow Data.
The Storm starts on Monday at 12:00 AM. Dry weather flow quantity and
quality have been measured at the point of exit from the case study basin.
The average flow measured was 0.89 MGD with average concentrations of
360 mg/1 BODc, 250 mg/1 suspended solids and 2.5 x 107 MPN/100 ml total
coliforms. Daily variations of the data recorded are shown in Figure 14 and
hourly variations in Table 9. In addition, industrial waste contributions
are indicated in Table 10.
Demographic and land use data for the remainder of the basin are as
follows:
1. Average number of people/household: 4.0
2. The two apartment complexes have a total of 480 units with an
average occupancy of 2.9 people/unit.
3. The motel has 144 units, each occupied by an average of 2.0 people
during the hours of 4 PM-8 AM daily.
4. The shopping center parking lot is designed for 1250 cars and a 50%
average usage rate. Hours of operation are 10 AM-10 PM daily.
Studies have determined that the average number of people in each
car using the lot is 2.
136
-------
Table 6. TRANSPORT BLOCK DATA PREPARATION CHECK LIST
1. Sewer System Descritization
2. Sewer Element Numbering
3. Compilation of Sewer Element Data
a. Type
b. Length
c. Dimensions
d. Invert Slope
e. Manning's Roughness
4. Internal Storage Data Preparation
5. Infiltration Model
a. Data Gathering
b. Data Preparation
6. Dry Weather Flow Model
a. Daily and Hourly Correction Factors
b. Average Dry Weather Flow Derived from
(1) Monitoring points within study area
(2) Sewage treatment plant data
(3) Water usage data
(4) Default values
c. Subarea Designations
d. Census Tract Data
e. Garbage Grinder Data
f. Subarea Data Preparation
7. Coding Form Preparation
8. Connectivity Diagram
137
-------
Table 7. SEWER ELEMENT DATA
Element
Number
Upstream
Element 1
Upstream
Element 2
Upstream
Element 3
Class
Length
(ft)
First
Dimension
(ft)
Invert
Slope
(ft/ 100ft:
Manning's
n
Second
Dimension
(ft)
Number
of
Barrels
Third
Dimension
(ft)
LO
oo
-------
Table 8. DRY WEATHER FLOW SUBAREA DATA
NOTE: To conserve space, variable names are used for headings. Refer to Card Group 45, Table 6-6, pp.
173-175 of the User's Manual for further explanation.
U>
VO
KNU
INPU
KLAND
METHOD
KUNIT
WATE
PRICE
SEWAGE
A-SUB
POPOEN
OWL INGS
FAMILY
VALUE
PCtiti
SAQPF
SABPF
SASPF
XINCOM
MSUBT
-------
o
iJ
c
CO
o
4-1
CO
Qd
<
i
1
1
1
1
1
0
0
.IT
.3-
.7
. 1-
.0
.9'
• &
.ft
• -^
1.25
0.70
1 1.15
r Ll/P-i ,,
0.95 |
0.80
i i i t i 1
Sun Mon Tues Wed Thurs
Day of the Week
(a) Flow Variation
Fri
Sat
O -i 0 -
.H 1.2
4-1
0 P
4-1 4-< i i -
c 1.1
C U
co a
s o 1 0-
o
14-1
f flj
•H >-i
4-1 01
CO >
as <:
1.10
1-05 | t.05
liiUU
U-VD J0.95
0.90 |
Sun Mon Tues
Thurs Fri
C
o
O CO
Day of the Week
(b) BOD5 Variation
c
CO
-------
Table 9. HOURLY FLOW AND. CONCENTRATION VARIATION
AM
PM
Hour
Period
M-l
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-N
N-l
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-M
Ratio of Hourly Mean to Daily
Average Flow or Concentration
Flow
0.74
0.67
0.63
0.59
0.54
0.56
0.67
0.96
1.42
1.19
.1.20
1.15
1.17
1.11
1.08
1.15
1.21
1.23
1.25
1.21
1.17
1.15
0.88
1.07
BOD5
0.85
0.71
0.60
0.41
0.46
0.49
0.72
0.87
0.77
1.57
1.02
0.87
0.91
0.94
1.07
1.07
1.14
0.99
1.45
1.16
1.55
1.29
0.99
1.60
SS
1.05
1.05
1.10
0.50
0.66
1.33
1.10
0.88
1.03
0.91
0.66
0.63
0.94
0.94
1.05
1.05
1.16
0.94
1.33
1.22
1.44
1.10
0.88
1.05
COLIF
1.10
0.64
0.45
0.87
0.54
0.48
1.29
1.18
1.37
1.49
1.30
1.12
0.89
0.58
0.45
0.67
0.96
1.18
0.84
1.01
2.82
1.77
0.84
0.71
141
-------
Table 10. INDUSTRIAL WASTE CONTRIBUTIONS TO DRY WEATHER FLOW
Process Waste Characteristics* Number of Number of
Industry
Soft Drink
Bottling
Plant
Average
Daily
Flow
cfs
.340
Average
Daily
BOD
mg/1
430
Employees Shifts
Average
Daily
SS
mg/1
360 200 2
Bakery .155 480 360 150
Electronics
Plant - 600
*
Sanitary wastes are not included.
142
-------
5. Family income, dwelling value and garbage grinder usage for
selected areas of the basin are:
a. School, West, Franklin and Wells Streets
Family Income: $20,000/yr
Average Dwelling Value: $50,000
Garbage Grinder Usage: 100%
b. Apartment Complexes
Family Income: $10,000/yr
Average Dwelling Value: $9,000
Garbage Grinder Usage: 50%
»
c. Area bounded by Oak, Davis and Elm Streets
Family Income: $8,000/yr
Average Dwelling Value: $20,000
Garbage Grinder Usage: 10%
d. Area bounded by High, East, Elm and Oak Streets
Family Income: $ll,000/yr
Average Dwelling Value: $25,000
Garbage Grinder Usage: 30%
Execution Control Parameters
Additional data required for job control are given in Table 11. In
cases where no data are given, the default value should be used.
CALIBRATION OF MODEL
Calibration of the model is best accomplished utilizing a small,
representative test basin within the urban area to be studied. A comparison
of predicted with measured results for the test basin (for several storms)
will indicate which model parameters can be adjusted so that model output
will more closely agree with measured results. Test basin results can also
be used to choose the best descritization scheme. Information gained from
the calibrated model can then be used in modeling the remainder of the
urban area. Verification of the model for the entire urban area should be
accomplished at some point (preferably an outfall) other than where the model
was calibrated.
Most of the parameters that can be changed for calibration'are contained
in the Runoff Block. However, to more accurately portray routing delay
through the sewer system, it is recommended that calibration data be
gathered at a point downstream from several subcatchments after routing flows
through the Transport Block.
The model should first be calibrated for flow. Accurate field data
concerning flow can be gathered at points within the system. If field data
on quality exist, an attempt can be made to calibrate the model for quality
also. The state of the art of quality modeling is such that an accurate
143
-------
TABLE 11, EXECUTION CONTROL PARAMETERS
Card Group
1
12
13
Variable
NKLASS
KPRINT
NOT
NPRINT
NPOLL
DT
DWDAYS
NCNTRL
NINFIL
NFILTH
JPRINT
JPLOT
NDESN
Value
0
0
50
0
3
300
14
0
1
1
1
0
1
144
-------
calibration is difficult to obtain. This discussion will concentrate on
flow calibration.
There are no set rules as to which program parameters should be
adjusted during flow calibration. A great deal will depend on the individual
situation. Generally, measurable data such as area and ground slope should
not be varied. Although measurable, a reliable estimate of percent imper-
vious area is difficult to obtain because of possible interactions between
pervious and impervious areas. Model output has been found to be highly
sensitive to the percent impervious area modeled (2). Therefore, percent
impervious area modeled should be considered as an adjustable parameter
within limits. Other subcatchment flow parameters that may be adjusted are:
1. Resistance factor for impervious areas
2. Resistance factor for pervious areas
3. Surface storage on impervious areas
4. Surface storage on pervious areas
5. Maximum rate of infiltration
6. Minimum rate of infiltration
7. Decay rate of infiltration
Changes in the pervious area resistance factor and surface storage
depth will have almost no effect on the outflow hydrograph. Likewise,
changes in the impervious area resistance factors and surface storage
depths have minimal effect on the volume and routing of the flow. Infiltra-
tion rates, when properly adjusted, can have significant impact on output
(6). Figures 15 (a-d)illustrate the response of the model to several of
these parameters. Figure 15(e) is an example of incorrect routing time which
may result from an inaccurately modeled sewer system.
If adjustment of input parameters does not result in a good calibra-
tion, the original descritization scheme should be re-evaluated.
145
-------
Predicted
Measured
(a) Percent Impervious Area Too High
(b) Initial Rate of Infiltration Too High
Q
Predicted
Measured
(c) Decay Rate of Infiltration Too High
FIGURE 15, PROGRAM RESPONSE TO INPUT PARAMETERS,
146
-------
Q
4 \
*—Predicted
Measured
(d) Minimum Rate of Infiltration Too Low
Q
/ >
Measured
Predicted
ModefeS
(e) Inadequate Routing Delay
FIGURE 15, CONTINUED,
147
-------
SAMPLE DATA INPUT AND PROGRAM OUTPUT: TRANSPORT BLOCK
A total of 27 elements were used to model the case study basin sewer system.
Pertinent input data and program output are provided on the following pages.
The specific items of information included are:
Schematic Showing Descritized
Sewer System (Figure 16)
Transport Block Card Data
Transport Block Program Output
Note that surcharging occurs at modeled elements 102 and 107. Sur-
charge conditions at these locations are consistent with the data given
in the case study materials and indicate areas of the sewer system that may
require modification.
148
-------
TRANSPORT
SEWER MAP
vo
108
400'
§18%
14
FIGURE 16, SCHEMATIC SHOWING DESCRITIZED SEVCR SYSTEM,
-------
g
TRANSPORT^ TRANSPORT BLOCK CARD DATA
TRANSPORT BLOCK HORKSHOP
27
30
0
1 101
101 2
2 102
102 3
3 103
103 4
4 10V
104 5
5 0
105 5
6 106
106 7
7 139
107 8
8 1D8
108 9
9 0
109 10
10 113
113 1%
14 £
110 11
11 111
111 12
12 112
112 IS
13 9
1
8
50
0.
1
0
a
105
0
0
0
9
0
0
0
a
a
107
0
a
a
a
a
110
0
0
a
0
0
0
0
fl
9
0.
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Q
1
0001
1
16
3
16
3
16
3
16
3
16
3
16
3
16
1
16
1
16
3
16
1
16
3
16
1
16
1
16
1 1
14.
0 0
32C.
230.
400.
540.
230.
83f.
350.
401.
39C.
510.
640.
435.
920.
U
9.5
6.5
6.5
6.5
8.5
8.0
3.5
3.0
6.0
2.5
6.0
4.0
3.5
3 4
.15
.12
.12
.10
.15
.15
.90
.93
.25
.90
.25
.40
.40
.013
.013
.013
.013
.013
.013
.013
.013
.013
.013
.013
.013
.013
11
12
H5
-------
28C 100.
* 9 60
0.9"
0.98
1.42
1.21
9.8=
1.77
1.14
1.35
1.03
1.16
1.10
1.37
0.96
1C 1 fl
1.377
3uV.9 43.0
1 13310 240.
2 1112
3 1412
4 1222
5 912
6 912
7 83?
8 632
9 542
10 532
GRAPH
913
GRAPH OF OUTPUT
TIHF
125.
7.0
311 636 995 llC
1.25 1.15
P. 95 l.tS
1.00 l.OC
J.67 C.63
1.19 1.20
1.23 1.25
0.71 O.M
1,57
0.99
1.05
0.91
j . 94
0.64
1.49
1.18
2 12
3*?.
55.9
32.
1.02
1.45
1.10
0.66
1.33
C.45
1.30
O.A4
CO
25C.
23.4
1C. 7 u.
4.i) IB.
23.4 5.
17.7 79.
19.4 ?1.
15.6 13.
3?. 4 0.
7.0 0.
43.0 - 0.
7.8 P.
1 977 846 510
0.95 1.10
1.10 l.;5
l.i* 1.01
0.59 C.54
1.15 1.17
i.21 1.17
u .87
1.16
u.50
0.63
1.22
0.87
1.12
1.01
2.5E+7
56.7
18. 4.u
28. 4.3
480. 2.9
10T1. 4.3
52. 4.0
L .'1
1.55
0.66
0.94
1.44
p.54
P. 89
2.82
2?.
3.3
20.
50.
9.
20.
25.
223 49
I.u5
i.sa
G.56
1.11
0.49
39.
0.
in.
100.
5u.
IP.
30.
0.
sj.
j.
0.
u.94
1.29
1.33
O.Q4
1.1P
0.44
C.58
1.77
2 125.9
.082 2CO.
.035 200.
.585 408.
.02? 200.
C.8"
0.95
0.99
u.67
1.08
?.86
C.72
1.47
o.99
i.ir
1.05
0.88
1.29
0.4?
C.84
2CO.
200.
336.
230.
FROM TRANSPORT
IN HOURS
r33
J-34
1.15
1.07
0.87
1.07
1.60
0.88
1.C5
1.05
1.18
3.67
0.71
1-37
2
-38
-39
-4C
"-41
>42
-43
8.
20.
10.
8.
11.
-45
FLOW IN
BOO LBS/HIN
SS LBS/NIN
COLXFOR4 MPN/MIN
ENOPR069AH
CFS
-------
TRANSPORT BLOCK WORKSHOP
• • « • • ELENENT LINKAGES AND COMPUTATION SEQUENCE * •
CLEMENT NO. ZERO IS GIVEN INTERNAL NO. « NO. ELEMENTS » 1
Ul
ro
EXTERNAL INTERNAL TYPE DESCRIPTION
ELEMENT ELEMENT
NUMBER NUMBER
t 1 16 MANHOLE
101 2 3 EGG-SHAPED
2 3 16 MANHOLE
1»2 t 3 EGG-SHAPED
3 5 16 MANHOLE
It3 6 3 EGG-SHAPED
% 7 16 MANHOLE
14% 8 3 EGG-SHAPED
5 3 16 MANHOLE
ItS IV 3 EGG-SHAPED
-6 11 16 MANHOLE
106 1Z 3 EGG-SHAPED
7 13 16 MANHOLE
107 1% 1 CIRCULAR
S IS 16 MANHOLE
116 16 1 CIRCULAR
9 17 16 MANHOLE
119 !• 3 EGG-SHAPED
It 19 16 MANHOLE
113 21 1 CIRCULAR
1% 21 16 MANHOLE
lit 22 3 EGG-SHAPED
11 23 16 MANHOLE
111 2% 1 CIRCULAR
12 25 1« MANHOLE
112 26 1 CIRCULAR
13 27 16 MANHOLE
UPSTREAM ELEMENTS
{EXTERNAL NOS.I
127
ORDER OF COMPUTATIONS AT EACH TIME STEP (PROCEEDING DOWNSTREAMI
COMPUTATION EXTERNAL INTERNAL INTERNAL UPSTREAM
SEQUENCE NUNBER NUMBER ELEMENT NUMBERS
111
Z
It2 IB
3
It3
%
It*
5
•
6
106
7
189 16
8
10 a
9
a
it
113- 11
14
t
11
111
12
112
13
a
o
a
i
z
3
fc
5
6
7
a
9
1C
11
12
13
14
15
16
17
ia
19
20
21
22
23
2%
25
26
27
103
3
112
9
198
a
107
Ik
113
13
112
12
111
11
110
10
189
7
m
6.
its
2
101
1
9
a
7
6
5
%
17
16
15
1*
21
20
27
26
25
2*
23
2Z
19
18
13
12
11
ta
3
2
1
28 28 28
9 28 28
8 28 28
r za za
6 28 28
5 28 28
28 28 28
17 28 28
16 28 2B
15 za za
28 28 28
21 2a 28
za za za
27 28 28
26 za za
25 za za
zt za za
23 28 28
20 22 28
19 28 28
18 Ik 28
13 28 28
12 28 28
11 28 28
k It 28
3 2B 28
2 28 28
-------
RUNOFF BLOCK WORKSHOP
NUMBER OF ELEMENTS-
NUMBER OF TINE INT*
TINE INTERVAL*
zr
so
30B.O SECONDS.
ELEMENT PARAMETERS
EXT. TYPE DESCRIPTION
ELE.
NUN.
1
101
2
102
3
103
10%
105
6
106
107
a
101
9
109
10
113
1%
110
11
111
12
112
13
SLOPE
CFT/FTI
DISTANCE MANNING GEOM1 GEOH2 GEOM3
(FT) ROUGHNESS (FT) (FT) (FT)
NUMBER AFULL QFULL
OF (SQ.FT) (CFSI
QHAX
ICFS)
BARRELS
IS
3
IS
3
16
3
16
3
16
3
16
3
16
1
16
1
16
3
IS
1
16
3
16
1
16
1.
16
MANHOLE
EGG-SHAPED
MANHOLE
EGG-SHAPED
MANHOLE
EGG-SHAPED
MANHOLE
EGG-SHAPED
MANHOLE
EG6-SHAPEO
MANHOLE
EGG-SHAPED
MANHOLE
CIRCULAR
MANHOLE
CIRCULAR
MANHOLE
EGG-SHAPED
MANHOLE
CIRCULAR
MANHOLE
EGG-SHAPED
MANHOLE
CIRCULAR
MANHOLE
CIRCULAR
MANHOLE
-0.0000
.0015
-0.0000
.0012
-ft .0000
.0012
-0.0000
.0010
-8.0000
.0015
-0.0000
.0015
-8.0000
.0090
-0.0000
.0090
-0.0000
.0025
-0.0009
.0090
-0.0000
.0025
-0.0000
.00*0
-0.0000
.00*0
-0.0000
-0.0
320.0
-0.0
230.0
-0.0
*03.0
-0.0
5*0.0
-O.ii
238.8
-0.0
030.0
-0.0
350.0
-0.0
*00.0
-0.0
390.0
-0.0
510.0
-O."
6*0.0
-0.0
*35.0
-0.0
920.0
-0.0
-0.0000
.0130
-0.0000
.0130
-8.0000
.0130
-0.0000
.0130
-0.0000
.0130
-0.0000
.0130
-0.0000
.0130
-0.0010
.0130
-0.0000
.0130
-o.oooo
.0130
-0.0000
.0130
-0.0000
.013d
-0.0000
.0130
-O.OuOC
c.o
9,S
0.0
6.5
0.0
6.5
O.ii
6.5
0.0
8.5
0.0
8.U
0.0
3.5
O.b
3.0
O.b
6.b
0.0
2.5
0.0
6.0
o.e
*.o
O.C
3.5
0.0
-9.0 -O.C
-fl.O -0.0
-0.0 -0.0
-0.0 -O.C
-0.0 -0.0
-0.0 -0.0
-0.0 -0.0
-0.0 -0.0
-0.0 -O.G
-0.0 -O.C
-O.J -0.0
-0.0 -O.C
-0.0 -0.0
-o.o -o.o
-0.0 -0.0
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.0
-0.0
-0.0
-0.0
-0.0
-0.0
-0.0
-O.C
-0.0
-0.0
-o.o
-i.o
1.0
1.0
l.o
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1*0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
l.li
1.0
1.0
1.0
1.0
1.0
0.000
1*6.073
O.OQO
21.569
0.000
21.569
0.000
21.569
0.000
36.88*
Q.OQO
32.672
0.000
9.621
o.oao
7.069
0.000
18.378
0.000
d.939
g.ooo
18.378
0.000
12.566
0.000
9.621
O.OC3
0.000
305.1*80
0.000
99.6**
0.000
99.6**
0.000
90.963
O.OBO
227.819
0.000
193.811
0.000
95.70*
0.000
63.M»6
0.000
116.181
3.000
39.017
0.000
116.181
0.000
91.993
O.OOJ
63.802
0.000
0.000
326. *01
0.000
106.121
O.OCO
106.121
a. ooo
96.875
O.OCO
2*2.627
0.000
206. *09
0.000
103.360
0.000
68.522
0.000
123.732
0.000
1.2.138
0.000
123.732
0.0 Ot
99. see
0.000
68.907
O.OCO
SUPER-CRITICAL
FLOW HHfN I
THAN 95
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
EPSILON * .oooiao
NO. OF ITERATIONS IN ROUTING ROUTINE
HVOROGRtPHS AND POLLUTOGRAPHS PROVIDED TO SUBSEQUENT PROGRAMS FOR THE FOLLOWING ELEMENTS
1
-------
TOTAL AREA INFILTRATION(IN GPM) DUE TO-
BASE FLOW GROUND WATER MELT RAIN
80.0000 0.0000 65.79<»0 125.0000
APPORTIONED INFILTRATION
ELEMENT NO.
103
102
106
107
113
112
111
110
109
106
105
101
QINFIL(CFS)
.059
.043
.025
.025
.025
.026
.067
.036
.064
.039
.111
.033
.051
PROP. TOT. INFIL.
.0971
.0719
.0413
.0412
.0420
.01*37
.1104
.0597
.1062
.06
-------
QUANriTY AND QUALITY OF D M F FOR EACH SUBAREA
KNUN INPUT
01
1
2
3
it
5
6
7
8
9
10
13
11
!«.
12
9
9
8
6
5
5
A1BOD
A1SS
A1COLI
AONF
OUF *
CFS
.01
.01
.02
.12
• C4
.02
.08
*G%
.59
• C2
= 2
= 1
= 3
=
INF It
CFS
.01
.00
.01
.05
.02
.01
.04
.02
0.00
.01
2525.17 LBS PER DAY/CFS
1753.59 LBS PER OAY/CFS
3.83E*10 NPN/OAY PER CAPITA
1.38 CFS
OQOHF
CFS
.02
.01
.02
.18
.P6
.03
.12
.05
.59
.03
KLANO DM BOD
L9S/MIN
3
1
1
2
1
1
3
3
k
3
.02
.01
.05
.25
.07
.0«.
.13
.06
.89
.03
OMSS
LBS/NIN
.01
.01
.0?
.17
.05
.03
.09
.0<»
,7k
.02
TOT POP aoocoNc
"ERSOMS MG/L
SSCOMC
MG/L
COLIFORMS
HPN/100ML
TOTALS
sssa
.16
1.1P
7.7«, L9S 5.95 LBS
219S.
SMTOHF =
»T EACH
377.
290,
3.13E*06
-------
DAILY AND HOURLY CORRECTION FACTORS
FOR SEHAGE DATA
PAY
1
2
3
-------
ELEMENT FLOWS. AREAS, AND CONCENTRATIONS ARE INITIALIZED TO DRY HEATHER FLOrf AND INFILTRATION VALUES,
Ln
ELE. NO.
10 <»
103
102
108
107
113
11Z
111
110
109
106
105
101
TYPE
3
3
3
1
1
1
1
1
3
3
3
3
3
FLOW
(CFS)
1.1*8
1.192
1.217
.135
.308
.056
.089
.3*7
. U24
.519
.938
1.033
2.301
AREA
CSQ. FT.I
.83%
.807
.817
.057
.118
.027
.051
.182
.392
.H21
.773
.851
1.327
INIT. VEL.
(FPSI
1.3766
l.<»77&
1.U893
2.3606
2*6161
2.0905
1.7I»«»1
1.9065
1.0803
l.?319
1.2138
1.21<»9
1.73V1
BOO
(LBS/CF)
.J200
.0193
.0189
.0207
.0195
.0200
,005«f
.0192
.0165
.0156
.0150
.015C
,u!67
S.S.
CL8S/CF)
.0125
.0120
.0118
.0110
.0103
.0106
.0629
.0102
.0087
.0083
.0080
.11079
.0098
E30LI.
(Mf»N/MLI
0.
0.
0.
t.05E+09
1.77E*09
1.86E+09
0.
3. 63 E 4-09
3.10E+09
2.73E+09
2. ICE *09
1.90E*09
8.5i»E*08
CPOLL NO.
-------
INITIAL BED OF SOLIDS IN SEHER DUE TO
1«».0 DAYS OF DRY HEATHER PRIOR TO STORM
ELEMENT SOLIDS IN
NUMBER BOTTOM
(LBS)
103 *»9.28-971
00 102 <»8.52239
108 1.060^6
107 .51%31
113 1.29961
112 1.13011
111 5.73588
110 6.21376
109 6.225^9
106 18.?6<»
-------
TINE=
RATIO
TIME*
RATIO
S800. SECONDS, TIME STEP*
OF REQUIRED FLOM TO PRESENT
3300. SECONDS, TIME STEP*
OF REQUIRED FLOM TO PRESENT
Ln •
vo
360B. SECONDS, TINE STEP*
RATIO OF REQUIRED FLON TO PRESENT
TIME* 1600. SECONDS, TIME STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TIME* 3910. SECONDS, TINE STEP*
RATIO OF REQUIRED FLON TO PRESENT
TINE* 3900. SECONDS, TIME STEP*
RATIO OF REQUIRED FLON TO PRESENT
TIME* tZOO. SECONDS, TINE STEP*
RATIO OP REQUIRED FLOM TO PRESENT
TIME- *200. SECONDS, TINE STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TINE* %500. SECONDS, TINE STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TIME* %500. SECONDS, TIME STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TINE* V800. SECONDS, TIME STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TINE* V800. SECONDS, TINE STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TINE* 5100. SECONDS, TIME STEP*
RATIO OF REQUIRED FLON TO PRESENT
TIME- 5100. SECONDS, TINE STEP*
RATIO OF REQUIRED FLON TO PRESENT
TIME* S*»00. SECONDS, TINE STEP*
RATIO OF REQUIRED FLOM TO PRESENT
TINE* 5730. SECONDS, TINE STEP*
RATIO Oc REQUIRED FLOM TO PRESENT
10.
QFULL
It.
QFULL
12.
QFULL
ELEMENT
* 1.16
ELEMENT
= 1.38
ELEMENT
* 1.25
1Z. ELEMENT
QFULL* 1.58
13.
QFULL
13.
QFULL
1%.
QFULL
1%.
QFULL
15.
QFULL
1$.
OFULL
16.
QFULL
16.
QFULL
ELEMENT
* 1.51
ELEMENT
* 1.72
ELEMENT
* 1.71
ELEMENT
* 1.7*
ELEMENT
* 1.83
ELEMENT
= 1.67
ELEMENT
= 1.83
ELEMENT
= i.ta
17. ELEMENT
QFULL* 1.72
17. ELEMENT
QFULL* 1.20
18.
QFULL
19.
QFULL
ELEMENT
* 1.50
ELEMENT
= 1.20
107
107
102
107
102
187
102
ior
102
107
102
107
102
107
102
102
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGING.
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
SURCHARGE
OF
OF
OP
OF
Or
OF
Or
OF
OF
Or
Or
OF
0s
OF
OF
Or
<.53T.6(.CU. FT. STORED AT UPSTREAM ELEMENT 6
.<»1CU. FT. ST09EO AT UPSTREAM ELEMENT 8
7«»12.*OCU. FT. STORED AT UPSTREAM ELEMENT 3
16758.26CU. FT. STORED AT UPSTREAM ELEMENT 6
1519*.23CU. FT. STORED AT UPSTREAM ELEMENT 3
205*7.65CU. FT. STORED AT UPSTREAM ELEMENT 8
21135.53CU. FT. STORED AT UPSTREAM ELEMENT 3
Z13Z0.7ZCU. FT. STORED AT UPSTREAM ELEMENT 8
2I.668.23CU. FT. STORED AT UPSTREAM ELEMENT 3
19236.10CU. FT. STORED AT UPSTREAM ELEMENT 6
2<»91>.6fcCU. FT. STORED AT UPSTREAM ELEMENT 3
13850.*»CU. FT. STORED AT UPSTREAM ELEMENT 8
21M8.85CU. FT. STORED AT UPSTREAM ELEMENT 3
5839.3«»CU. FT. STORED AT UPSTREAM ELEMENT 8
15091.38CU. FT. STORED AT UPSTREAM ELEMENT 3
599$.32CU. FT. STORED AT UPSTREAM ELEMENT 3
-------
SELECTED INLET HYDROGRAPHS - CFS
EXTERNAL
ELEMENT
NUMBER
8
TIME STEP
1
.108
77.254
20.161
1.373
.274
2
.989
74.061
18. 888
1.104
.244
3
5.894
66.739
17.172
.904
.218
*
15.573
58.365
14.275
.752
.196
5
24.017
51.235
10.073
.633
.177
6
33.704
43.091
6.311
.539
.161
7
43.742
37.413
4.269
.463
.146
8
52.794
33.878
3.049
.402
.134
9
65.850
28.602
2.266
.351
.122
10
77.043
22.777
1.742
.309
.112
EXTERNAL
ELEMENT
NUMBER
TIME STEP
1
SELECTED INLET POLLUTOGRAPHS
»•* BOD IN LBS/HIN »»»
10
.030
.688
.974
.014
.800
.148
.560
.868
.008
.000
1.113
.461
.764
.005
.000
3.193
.377
.657
.003
.OOG
•» SUSPENDED
.007
10.318
19.466
.285
.005
.219
9.615
17.364
.168
.004
10.933
8.51"
15.284
.103
.003
50.631
7.211
13.130
.065
.002
•»• COLIFORM
3
1
1
%
4
. 83EO9
.6lE»08
.23E»06
.90E»05
.29C»05
2.12E«-09
7.70E+07
9.75E*05
4.77E»05
4.26E»05
2.19E*09
3.82E»07
8.21E»b5
4.67E+C5
4.24E»05
2.41E»09
1.99E»07
7.28E+05
4.59E»fl5
4.22E*05
2
1
6
4
4
3.569
.310
.513
.002
.000
SOLIDS
56.036
6.046
10.252
.042
.002
2.948
.254
.321
.001
.000
IN LBS/NIN
42.461
4.997
6.416
.028
.001
1.209
.208
.165
.001
.000
»*
6.102
4.109
3.304
.319
.001
1.100
.937
.085
.001
.000
7.353
18.705
1.705
.013
.001
.964
.986
.045
.000
.000
8.696
19.705
.906
.009
.001
.835
1.027
.025
.000
.000
10.032
20.526
.499
.007
.001
IN MPN/NIN »»•
.47E»09
.10E+07
.95E»05
•53E+C5
.20E»05
2.D2E+09
6.77E*06
6.74E*'05
4.47E»05
4.19EM5
1.50E»09
4.29-:»06
6.01E»05
4.42E»C5
4.17E»05
l.OOE+09
2.87E»06
5.57E»05
4.38E»05
4.16E+05
5.81E»08
?.13E»06
5.27E»05
4.35E»05
4»15E»(i5
3.14E»08
1.65E+06
5.06E+05
4.31E+05
4«13E«-05
-------
SELECTED OUTFLOW HYOROGRAPHS - CFS
EXTERNAL
ELEMENT
NUMBER
1
TINE STEP
1
2.325
317.737
177.025
19.878
Z
2.666
321.795
151.283
15.3*3
3
5.799
321.008
135.729
12.610
*
20.691
323.12ft
119.535
10.568
5
53.*38
31V. 035
102.178
8.898
6
90.502
309. *35
80.31*2
7.605
7
131.999
302.295
61.303
6.617
*
176.596
290.372
*2.09*
5.871
q
230.3*8
•268.192
31.918
5.293
10
278.217
22*. 7*9
2*. 620
*.B35
*.*8o
*.21*
3.950
3.759
3.579
3.*37
3.313
3.20*
3.11*
3.058
SELECTED OUTFLOW POLLUTOGRAPHS
EXTERNAL
ELEMENT
NUMBER
1
TINE STEP
1
2.315
12.*75
7.172
2.*99
2.386
2
2.*81
9.06*
7.190
2.38*
2.*19
3
3.705
7.7*1
7.0*2
2.31*
2.*3*
*
»»» BOO IN
7.788
7.393
6.3*9
2.283
2.*59
5
6
7
t
8
9
10
LBS/NIN •»»
1*.929
6.878
5.927
2.239
2.*6H
•• SUSPENDED SOLIDS
1
6.592
1*2.253
98.076
12.006
2.520
13.59*
97.917
97.081
8.928
2.*53
18.873
82.1*1
93.067
6.593
2.376
50.237
75. *2*
81.**2
5.688
2.335
1*8.156
72.365
71.893
(t.36*
2.283
IN
20.028
6.675
*.969
2.20*
2.*8*
LBS/NIN
251.3*7
69.9*0
55.0*8
3.861
2.255
22.168
6.968
*. 332
2.192
2.*97
• »
307.375
75.710
10
3.92E*1J
1.60E»11
1.0*E*11
5.93E*10
*.65E«10
3.96£»10
1.52E»11
1.0*E»11
*.93E»10
*.56E+10
3.99E»10
l.*8E*ll
9.86E+10
*.99E»10
*.*OE»10
*.d*E»10
1.35EU1
9.03EHC
5.00E+10
*.18F*10
*.16E«10
-------
SELECTED OUTFLOW POLLUTOGRAPHS
to
EXTEPNA;.
ELEMENT
NUMBER
TINE STEP
1
2
3
>»
5
6
7
8
9
10
••• BOD IN MG/L ••••
1
366.3*0
10.503
1C. 837
35.0 CO
U2.I.U1
2%8.S72
7.535
12.713
«.1.5S3
153.561
170.91U
6.*51
13.878
•.9.086
lf.«..8*»2
100.686
6.120
It. 208
57.687
175.019
7<».730
5.859
15.516
67.326
18U.<»25
••» SUSPENDED SOLIDS
1
1
758,1.66
119.761
11.1.201
16B.33I»
150. *79
3.02EO6
2.29E»0<»
3.06E»OU
1.51E*05
5.?7E»05
1363.8<.l
81.396
171.660
155.656
155.731
2.91E*06
2.20£»0«»
3.61E*Ot
1.87E»05
5.50E+05
870.579
60.d<»9
183. 421
139.871
16C.879
1.95E»06
2.12E»fli*
3.96E*«U
2.28E*05
5.79E*i.5
6<»9.U8<.
62.WVO
182.25«»
1<»3.711
166.168
•••COLIFORM
9.07F*05 <»
1.9i»E*0<. 1
<».58E*U
3.6*.E*85
6.70E*05
7.10E»0<»
2.02E»0<»
5.68E»0«.
(».12E»05
7.01E»05
5.06E^O<*
2.1uE*Ol»
6.88E*0<»
«t.56E*05
7.31E*05
3.76E»0<»
2.16E+0<»
9.18E»PV
-------
GRAPH OF OUTPUT FROM TRANSPORT
a*
500.000 I
I
I
I
I
I
I
I
I
I
too.ooo -
I
I
I
I
I
I
I
I
I
300.000 -
I
I
FLO* I
II
CFS
ZOO.000
100.190
I
I
I
I
I
I
I «
O.flOO •»»»»-
1Z.O
*
*
*•**»***•»*•»
»«*T*•*•*••••T»*»*••••»T•«
13.0 13.5 lfc.0
TINE IN HOURS
1*1.5
15.0
«••»••••••••»•••»
15.5 16.0
16.5
— I
17.0
-------
GRAPH Of OUTPUT FROM TRANSPORT
Z5.000 I
I
ZI.I06
15.000
BOO
LBS/tZN
10.000
5.000
*
*
•
*
— J.
12. 5
1
13.il
1
13.5
lt.0
mm*T^^••^•^^•^•••••••••^••••••»••!•••••••••!
15.C 15.5 16.0 IS.5 17.0
TIHE IN HOWS
-------
GRAPH OF OUTPUT FROM TRANSPORT
in
soa.iti
tOI.OM
ss
LBS/tlN
.•••
lai.ioi
g.ooi •• —i—
1Z.B 12.5
13.9 13.5
TIHf IN HOWS
*•••••••
.{^........j...... .•••*•••••»«••»••••••*••••
li.O lt.5 15.*. 15.5 16.t
16.5 ir.l
-------
a*
5.t»€»ll
%.••£•!!
3.IIE+11
COLIFORN
NPN/tlN
2.ME+11
• »
-•
!•
!•
••*»•••••
0.
1Z.O If. 5
13.0 15.5
TIRE IN HOURS
1*.5 H.O
...j.
15.5
16.0
..-I 1
16.5 17.C
-------
STORAGE/TREATMENT BLOCK STUDY GUIDE
DISCUSSION OF OPTIONS AVAILABLE
Primary uses of the Storage/Treatment Block include modeling of:
1. existing sewage treatment plants to estimate treatment effective-
ness under storm conditions.
2. proposed changes to existing treatment plants that are designed to
upgrade treatment of storm flows.
3. proposed plants designed specifically for storm water treatment.
4. storage facilities used in conjunction with any of the above.
Output from the block can be used to compare treatment or cost effectiveness
between two or more given storage/treatment schemes. The program is not a
primary design tool because no provisions have been made to optimize
storage/treatment units. All storage and treatment parameters are either
pre-programmed or input by .the user, and remain constant throughout the
simulation.
Input data requirements for the Storage/Treatment Block are minimal.
Data for existing facilities can be extracted from facility plans or
operational records. Data for proposed facilities can be taken from design
drawings or specifications. In the case of existing treatment facilities,
the user may wish to replace programmed treatment parameters with site
specific data. Equations used in treatment calculations are discussed in
a later section .
Multiple treatment schemes can be simulated during the
same computer run. Only the results of the first simulation, however, can be
transferred to the Receiving Water or other blocks. Figure 17 provides an
example of multiple simulations.
Three types of storage facilities can be modeled by the SWMM. The first
is a natural reservoir of irregular geometry. For this type, surface area
versus water depth data must be input. The second and third types are
covered and uncovered reservoirs of regular geometry. For these, base area,
base circumference and side slope are required inputs. Additional data
requirements are: the type of basin flow (plug flow or completely mixed),
maximum (flooding) reservoir depth, type of basin outflow works (pump,
orifice, weir), and storage facility initial conditions (storage volume
utilized and outflow rate at time zero).
167
-------
TRANSPORT
STORAGE
BARRACKS
OUTPUT FILE CREATED FOR
TRANSFER TO OTHER BLOCKS
4
PRINT
RESULTS
BAR RACKS
DISSOLVED AIR
FLOTATION
I
FIICROSTRAINERS
1
PRINT
RESULTS
RUN
*.
STORAGE
_w
BARRACKS
SEDIfCNTATIOTJ
BIOLOGICAL
CHLORINATION
PRIflT
RESULTS.
FIGURE 17 . EXAMPLE OF MULTIPLE STORAGE/TREATMENT SIMULATIONS
168
-------
Single or dual rate outlet pumps can be simulated. For dual rate
pumps, the second pump starts if the first pump fails to lower the water
level in storage. Data required are the outflow pumping rate (QPUMP),
the depth of water in storage at pump start-up (DSTART) and the depth of
water in storage at pump shut-down (DSTOP). The volume of water contained
in storage at DSTOP must be equal to or greater than the volume pumped
during one time step. Figure 18 depicts the important datum for outflow
pumping of a storage reservoir.
For orifice outlets, the orifice centerline is assumed to be located
at zero storage tank depth. The product of the orifice outlet area and
the orifice discharge coefficient, C, must be input. Values for C can
be found in King's Handbook of Hydraulics, Reference 7 or equivalent.
Required inputs for a weir outlet are weir height and length.
SWMM II provides the options of bypassing storage, routing all flow
through storage, or routing flow in excess of the treatment capacity
through storage. If the Storage/Treatment Block is called, flows must be
routed through at least one treatment module. Treatment modules are
arranged in seven levels; preliminary treatment, inlet pumping, primary
treatment, secondary treatment, effluent screens, outlet pumping and
chlorination. Preliminary treatment can include bar racks or swirl concen-
trators. Primary treatment options available are sedimentation, fine
screens, dissolved air flotation or fine screens in conjunction with dis-
solved air flotation. Secondary treatment can be accomplished by micro-
strainers, high rate filters or biological treatment. Standard or high rate
disinfection can be applied to the treatment plant effluent and high rate
disinfection can be applied to treatment plant overflows. Any of the levels
of treatment can be bypassed.
The user can either specify the treatment .capacity desired or have the
program determine the treatment capacity based on a percentage of the peak
flow. Once the treatment capacity has been set, all flows in excess of this
capacity are bypassed and re-combined with the treatment facility effluent
before the output hydrograph and pollutographs are written on the output file.
Many of the treatment parameters are pre-programmed. However, additional
data are required for the following treatment modules: high-rate disinfec-
tion of overflows (flow in excess of design1flow); swirl concentrator;
dissolved air flotation; sedimentation; and high rate filtration.
The Storage/Treatment Block will also provide estimates of the capital
and operational costs for the storage and treatment facilities modeled.
Figure 7-2 on page 220 of the User's Manual (reference 6) provides a schematic
diagram of all the storage/treatment options available,
PREPARATION OF STORAGE TREATMENT FLOW DIAGRAM
The storage/treatment flow diagram is a schematic diagram of the modules
to be called, annotated with all of the required user supplied data. The
development of such a flow diagram is straightforward for existing storage or
169
-------
nSTART
DSTOP
QPUMP
For Rectanuglar Basin
DSTOP*BASEA>QPUMP*DT
BASKA = Base Area
DT = Time Step, seconds
FIGURE 18, PUMP OUTLET PARAMETERS,
170
-------
treatment works, In the case of proposed facilities, the model may be
useful in testing various alternatives at the preliminary planning stage.
Several computer runs may be desirable to compare various treatment con-
figurations, to test the effects of varying treatment parameters or to
determine the best combination of storage and treatment.
To simulate proposed facilities, realistic storage and treatment para-
meters must be input to the model. Required capacity of the storage facility
can be estimated from the volume of flow generated by the design storm and
the volume of flow treated during the simulation. A mass curve (Ripple
diagram) can be used to determine the volume of the storage facility required
to prevent any overflow during the simulation period. Figure 19 shows such
a diagram constructed for the case study hydrograph.
Treatment parameters can be taken from bench or pilot scale studies
already completed, from design specifications or from standard engineering
references. References 8 and 9 contain suggested parameters for the types
of treatment modeled by the Storage/Treatment Block.
The storage/treatment flow diagram can be used to complete the computer
coding forms for the Storage/Treatment Block. Instructions for coding form
preparation are contained in Table 7-2, pages 227-237 of the User's Manual.
CASE STUDY MATERIALS
A hypothetical 3000 acre basin was used to generate hydrographs and
pollutographs in the Runoff Block. These were then routed through the
Transport Block. For the purpose of developing these data, land use within
the basin was assumed to be equally divided among the five types modeled by
SWMM; i.e., single family residential, multi-family residential, commercial,
industrial and open land. Impervious area of subcatchments varied from 10%
to 60%, with the average being 35.4%. The design storm used in the Runoff
Block Case Study produced the overland runoff. Dry weather flow and infil-
tration were modeled in the Transport Block. Thirty days of antecedent
dry weather and a street cleaning frequency of 40 days were chosen as
critical design conditions. The total length of all conduits modeled in
Transport was 3 miles. The resulting hydrographs and pollutographs for
entry into the Storage/Treatment Block are given in Figure 20. A total
volume of 116.85 million gallons (for an average flowrate of 112.18 MGD for
25 hours) is discharged. The total storm runoff contains 84,628 pounds of
BODc and 295,655 pounds of suspended solids. (Note the large first flush of
pollutants indicated in Figure 20.)
Transfer of flow from the Transport to the Storage/Treatment Block
occurs at modeled Element 10 as shown in Figure 21. Flows in excess of the
design capacity of 50 MGD are bypassed and then re-combined with the treat-
ment plant effluent prior to discharge into the receiving waters.
A preliminary planning study is now underway to determine the
feasibility of replacing the existing treatment plant with a facility der-
designed to treat storm flows as well as dry weather flow. The Storage/
171
-------
17 1
16
15
14
13 -
12
11 -
10 -
9 .
vO
O
8 -
n
£ 7 -
I 6 •
CO
CO
5J
4
3 -
2 -
1 -
0
Required Storage = ASMAV; for 100 MGD Plant = 6.2 x 106 ft3
MAX:
Treatment Rates
5 10
Time From Start of Storm, hrs.
I I I I I I I I I II I
15 20 25
FIGURE 19, fkss STORAGE CURVE,
172
-------
4(10
I | | | ) I I | I I I I I I I I I I I I I I I '
r, 10 13 20 '2
from Start of Sturm, lirs.
(a) Hyclrograpli
10
E
1000 -
c
o
o
S
500
i i
5
I I I I
10
l\ I
15
I I
20
I I
25
Time fri)m Start of Storm, hrs.
(h) BOH PolJutograph (my/J)
FIGURE 20, HYDROGRAPH AND POLLUTOGRAPHS FOR 3000 ACRE BASIN,
173
-------
300 1
200 -1
c
c
fl
1_
•u
C
c
c
100
0
I I
5 10 15
Time From Starr of Storm, hrs
(e) SS Pollutograph for Entire Storm (mg/1)
20
25
FIGURE 20-> CONTINUED,
174
-------
400
300
c
•r!
B
3
i—I
b
§
po
a
C
in
en
200
100
0
0
1 4 ' ' ' ' V ' ' '
Tl
Time from Start of Storm, hrs
(c) BOD Pollutograph (Ib/min)
15000"!
g 10000
•H
5000
1 2 3
Time from Start of Storm, hrs
(d) SS Pollutograph Showing First Flush (mg/1)
FIGURE 20,..CONTINUED,
175
-------
TRANSPORT
ELEMENT
10
FLOW RATE: 1600 GPD/n2
: 8 FT
)
50MGD
BARRACKS
OVERFLOW
ON
SEDII€NTATION
HBIOUOGICAL
TREffllfHT
EFFLUENT
RECOMBINED
RELEASE
RIVER
CHLDRHWTION
FIGURE 21. EXISTING TREATMENT PLANT CONFIGURATION,
-------
Treatment Block of SWMM will be used to evaluate the treatment effective-
ness that will be realized by the installation of a physical-chemical
treatment plant, with or without storage. Later, the results of these
simulations will be input to the Receiving Water Block to determine the
response in receiving water quality to treatment of storm flows.
The assignments for the Storage/Treatment Block workshop are to
develop the storage/treatment flow diagram for the physical-chemical treat-
ment plant with storage facilities and to prepare the necessary computer
coding forms. Use the following guidelines for design:
1. Design the storage and/or treatment facilities so that no overflow
will occur during the storm event. Figure 19 can be used for this
purpose.
2. Include chlorination in the treatment string.
3. Use any reasonable combination of treatment modules.
4. Transfer data from the Transport Block to the Storage/Treatment
Block through external element number 10.
5. Assume the time of start of simulation to be 12:00 AM.
Typical treatment parameters for the modules contained in the
Storage/Treatment Block are presented in Table 12 and may be used if desired.
Default values for cost computations should be used with the exceptions
given in Table 13 for excavation and the ENR Cost Index.
PROGRAMMED TREATMENT PROCESS CONSTANTS AND COST GENERATION EQUATIONS
To utilize the Storage/Treatment Block effectively, the user must have
an understanding of the logic involved in treatment and cost calculations,
This will aid in interpretation of simulation results and will allow the
user to change treatment and cost parameters to reflect local conditions.
The following synopsis of treatment efficiency calculations is presented here
as Table 14. This has been taken from Sections 15 and 16 of Reference 11 and
from the SWMM II program listing. Similarly, cost estimation equations are
included in Table 15. These have been revised from those given in Reference 11
to reflect capital costs given in Reference 6. Tables 16 and 17 are also ex-
tracted from reference 11 and give irreducible maintenance and storm event
costs used in the model.
177
-------
Table 12, TYPICAL TREATMENT PARAMETERS
Dissolved Air Flotation
2
Design Overflow Rate = 5000 gpd/ft
% Flow Recirculation = 15%
Depth of DAF Tank = 8 ft
Primary Sedimentation
Design Overflow Rate «• 1600 gpd/ft2
Depth of Sedimentation Tank = 8 ft
High Rate Filter
2
Maximum Operating Rate = 20 gpm/ft
Maximum Design Head Loss = 10 ft
Maximum Solids Holding Capacity = 3 lb/ft2
Swirl Concentrator*
Number of Particle Sizes = 5
Concentrator Diameter = 36 ft
Specific Gravity of Particles - 2.65
Particle Size (cm) = 0.02 0.05 0.1 0.15 0.20
Fractional % by weight - 0.10 0.10 0.15 0.25 0.40
*
Data taken from Table 2, page 15, Reference 10.
178
-------
Table JL3. EXCAVATION COSTS AND ENR COST INDEX
Dollars/Cubic Yard for Storage Excavation (CPCUYD): $6.00
ENR Cost Index for Year and Location (IENR):
Year Index Value
1970 1363
1971 1545
1972 1712
1973 1886
1974 2035
1975 2180
1976 2312
1977 2431
1978 2537
1979 2630
1980 2710
179
-------
Table 14, Synopsis of Treatment Efficiency Calculations
Fine Screens
2
Hydraulic Loading Rate: 50 gpm/ft
SS Removal: 27%
BOD5 Removal: 22%
Backwash Flow: 0.75% of entering flow
Sedimentation
SS Removal = 0,656 + SS <°-°6> - 0.40
2
OVFRA - overflow rate in gpm/ft /day
(assumed to be at least 300 gpd/ft )
SS removal efficiency is limited to a range of 0.30 to 0.76
BOD5 Removal =0.55 (SS removal) w/o chlorination
=0.66 (SS removal) with chlorination
Sludge volume is based on a solids content of 5% and a minimum
sludge pumping rate of 45 gpm.
Dissolved Air Flotation
SSREMV = 0.656 + 0.06J8SINF) _ Q^
100,00
SSREMV = Suspended Solids removal efficiency
SSINF = Suspended Solids in inflow, mg/1
ICHEM = 0 if no chemicals are added
= 1 if chemicals are added
SS removal efficiency is restricted to not less than 0.20 and not
more than 0.82.
180
-------
Table 14., Continued
BODREM = 0,59 + °'05 (™ C°nc> - 0.36
+ ICHEM (0.02) + ICL2 (0.15)
ICL2 = 0 if no chlorine is added
= 1 if chlorine is added
BOD5 removal efficiency is restricted to not less than 0.18 and
not more than 0.60
2
Minimum Overflow Rate: 1000 gpd/ft
If chlorination option is chosen:
10 mg/1 chlorine will be added if the BOD,, of the influent
is 130 mg/1 or less
15 mg/1 will be added if the influent BOD_ is greater than
130 mg/1.
Dissolved Air Flotation with Fine Screens
SSREMV = 0.528 + ' C°nd) - 0.486
KB* (1.37)
BODREMV = 0.475 + ' C°nC> -'0.405
+ ICHEM (1.30) (0.02) + ICL2 (1.30) (0.15)
SS Removal efficiency is limited to not less than 0.15 and not
more than 0.75.
BOD5 removal efficiency is limited to not less than 0.15 and not
more than 0.48.
181
-------
Table 14, Continued
Microstrainer
2
Design Based on: 4.3 rpm, application rate of 40 gpm/ft of
submerged area, 69% submergence and removal rate
2
of 0.0135 Ib of solids/ft /revolution
SS Removal (mg/1) = 10>°00
2
gpm/ft submerged area
F = area correction factor
=1.0 for areas up to 400 acres
area in acres
for areas >400 acres
2
The model restricts flow to 40 gpm/ft of submerged area, regard-
less of solids concentration. If solids concentration is so
great that flow cannot be strained, part is bypassed and the
overall performance is computed by the model.
BOD,. Removed (ppm) » BOD in influent - 10.0, if BOD_ in is
^27 ppm
= 17(BOD5 in influent)
27~~5 > if BOD5 in is <27 PPm
Biological Treatment
SSRM - SSIN (0.8)
BDRM = BDIN (0.8)
SSRM = Suspended Solids removed, Ib/time step.
SSIN - Suspended Solids inflow, Ib/time step
BDRM - BOD_ removed, Ib/time step
182
-------
Table 14, continued
BDIN = BOD inflow, Ib/time step
High Rate Filters
The filter modeled is a tri-media filter consisting of:
36 inches of anthracite, 24 inches of garnet-sand mixture, 3 inches
of coarse garnet and 9 inches of gravel.
2
Design Capacity: not to exceed 20 gpm/ft
2
Maximum Size of Units: 1400 ft
Maximum Capacity of Each Unit; 40 MGD or 62 cfs
An even number of units will always be specified.
Maximum Head Loss: not to exceed 10 ft.
2
Maximum Solids Holding Capacity: 3.0 Ib/ft
SS removal of Ripened Filter: 80% w/o chemicals
95% with chemicals
Reduction in removal efficiency for fresh filters is modeled.
If chemicals are specified, dosages of 150 mg/1 alum and 4 mg/1
flocculent aid are used.
Calculationa for Head Loss:
Hc - (q/V* x (0.40)
(q/qj (S/Sqm) (0.60 Hm>
«L " HC + H,
CL
H = head loss when filter is clean
\j
Hp • head loss due to clogging
183
-------
Table 14, continued
IL = total head loss
a = maximum design flow rate
H = maximum allowable head loss
m
S = integrated sum of the solids removed in each time step
(lb/ft2)
o
Sqm = solids holding capacity of the filter (lb/ft )
Effluent Screens
Effluent screens are used for aesthetic reasons only and are
assumed not to remove any SS or BOD-,
Included in the treatment stream for cost computations.
Chlorination
Detention Time: 15 minutes at design flow.
If detention time is less than 15 minutes, chlorination
efficiency is adjusted based on the ratio of the detention time
to 15 minutes.
Chlorine Demand: estimated at 10% of BOD_ content measured just
ahead of the point of application. Chlorine demand is limited
to not less than 6 ppm and not more than 25 ppm. Chlorination
results in a BOD- reduction of 2 times the chlorine demand, but
not more than 50% of the contact tank influent BOD5 concentration.
Coliforms at „ Coliforms in SS at point of application
Point of Application * influent SS at influent to treatment
units
Chlorination Efficiency: 0.999
184
-------
Table 1 5, TREATMENT COST SUMMARY
Option
Derived Capital Cost
Equation, Dollars
Applicable
Capacity
Range, mgd
Bar Racks
Supply:
Install:
1000(11 +
where n = Number of screens
s = Screen capacity
(cfs)-120,with a
minimum or zero
ENR - ENR index for pre-
scribed year (see
Volume III)
F - Site factor (see
Volume III)
10,000 (f •625(||L-)F
All
1,780 <
where Q - Capacity (mgd)
<. 100
> 100
Inlet Pumping
25,000 Q°-58(ff£4)F < 20
16,000 Q'(£)F 20100
Dissolved Air
Flotation
All
Fine Screens
12,000 Q
All
Microstrainers
10,000 Q
All
High Rate Filters
85,000 Q
°*67
-
1058
All
185
-------
Table 15, Continued
Sedimentation
In New Tanks
In Storage
430 n (70,000 0.91 fENR .
^ v R ' V1000'
where R = overflow rate
(gpd/sq ft)
<100
>100
HI r
27 (1314)F
AU
where U = Construction cost
($/cy)
V = Maximum storage
during storm (cy)
FNR
Biological Treatment 55,000 Q(ff§Q-)F
All
Effluent Screens
Supply and
Install:
Channel works:
5,000
Q
7,000 Q
°-625
<100
^100
<100
1,246 Q (
>100
Outlet Pumping
(Same as Inlet Pumping)
Chlorine Contact
Tank & Equipment
18>350 Q-
High Rate
Disinfection
2000
FNR
Qd = Design flow (MGD)
186
-------
Table 16 , IRREDUCIBLE MAINTENANCE COSTS
Treatment Process
Assumed Percentage of Total Capital
Investment Per Year
Bar screens
Pumping station, influent
Dissolved air flotation
Fine screens
Sedimentation tanks
Microstrainers
High rate filters
Effluent screens
Pumping station, effluent
Chlorine contact tanks
1
2
2
2
1
2
2
1.5
2
2
187
-------
Table 17. STORM EVENT COSTS
Treatment Process
Assumed Variable
Cost
Assumed Fixed
Cost, $/storms
Bar racks
Pumping station
Dissolved air
flotation
Fine screens
Sedimentation tanks
Microstrainers
High rate filters
Effluent screens
Pumping station
Chlorine contact
tanks including
chlorinators
Based on the volume of 15
solids to be disposed
Based on 2c/kwh and 15
assumed efficiencies
0.40/1,000 gal 15
0.60-1,000 gal 15
0.40/1,000 gal. 15
0.60/1,000 gal. 15
1.00/1,000 gal. 15
Based on the volume 15
of solids to be
disposed
Based on 2o/kwh and 15
assumed efficiencies
Based on cost of chlorine 15
20c/lb
188
-------
SAMPLE DATA INPUT AND PROGRAM OUTPUT: STORAGE/TREATMENT BLOCK
The sample solution for the Storage/Treatment Block case study uses a
physical-chemical treatment scheme in conjunction with storage facilities to
completely contain and treat the stormwater flow resulting from the 10 year-
2 hour duration design storm. A treatment rate of 125 MGD was chosen for this
run. Based on this treatment rate, the mass flow diagram (Figure 19)
indicated a required storage capacity of 5.0 x 10° gallons. Applying a 50%
safety factor to allow for future development within the basin, a storage
volume of 7.5 x 10^ gallons was selected.
Descriptions of the treatment modules used and the storage facility are
contained in the sample program output. Figure 22 compares the treatment
effectiveness of the chosen scheme with an identical scheme without storage
facilities and with the conventional treatment plant shown in Figure 17, Sample
data input and sample program output are also provided,
189
-------
v£>
O
CARD
9 i
-------
STORAGE BLOCK CALLED
ENTRY MADE TO STORAGE/TREATMENT MODEL
STORAGE/TREATMENT MODEL UPDATED BY UNIVERSITY OF FLORIDA FEBRUARY 1975
TEST CASE, 3000 ACRE BASIN
OUTPUT FRON EXTERNAL STORAGE/TREATMENT HODELS
M INPUT DATA-SET OUTFALLS AT THE FOLLOWING ELEMENT NUMBERS!
M l°
INPUT TO STORAGE/TREATMENT MODEL SUPPLIED FROM EXTERNAL ELEMENT NU1BER 10
NUMBER OF RUNS = 1
TIME-STEP SIZE = 10.00 NIN,
NO. TIME-STEPS MODELED = 150
TRIBUTARY AREA = 3000.00 ACRES
NO. TRANSP. MOO. OUTFALLS = 1
NO. OF POLLUTANTS * 3
TIME ZERO = «ZOO.«J SEC
-------
RUM MO. 1
INPUT DATA FOR TREATMENT PACKAGE FOLLOWS
CHARACTERISTICS OF THE TREATMENT PACKAGE ARE
LEVEL MODE PROCESS
0 03 STORAGE ROUTED
1 12 BAR RACKS
2 21 (BYPASS)
3 35 SEDIMENTATION
d « HIGHRATE FILTERS
5 51 (BYPASS)
6 61 (BYPASS)
7 72 CONTACT TANK
IPRIMT = 2i ICOST » 1, IRANGE = 1, ITABLE *
SPECIFIED TREATMENT CAPACITY USED.
DESIGN FLOURATE = 193.Off CFS.
TREATMENT SYSTEM INCLUDES MODULE UNITS
DESIGN FLOW IS THEREFORE INCREASED TO NEXT LARGEST NODULE SIZE
ADJUSTED DESIGN FLOMRATE = 193.38 CFS.t * 125.00 1GO.
£ «KNOO * 111
~ CHARACTERISTICS OF STORAGE UNIT ARE
OUTLET TYPE = 6
STORAGE MODE = 1
STORAGE TYPE * 3
IPX = It PRINT CONTROL USPRINI = 1
MAN-MADE RESERVOIR, HITH MAX. DEPTH ' 38.00 FT., AND CHARACTERISTICS
BASE AREA * 250(00. SQ.FT., BASE CIRCUMF. - 2000. FT.. COTISIOESLOPE1 - C.00000
RESERVOIR OUTFLOU BY FIXED-RATE PUMPING
PUMPING RATE * 193.00 CFS, PUMPING START DEPTH * 1.20 rT, PUMPING STOP DEPTH = .50 FT
DEPTHIFTI STORICU.FTI DEPTH(FTI STOR(CU.FT) OEPTH(FT) STOR (CU.FTI OEPTHIFTt STOR(CU.FT)
0.0" (. 1.09 rSOOOO. 6.0« 1500000. 9.00 Z250000.
12.0G 3000100. 15.09 3750000. It.00 tSOOOOS. 71.00 5250000.
2fc.OO 6000(00. 27.00 6750000. 30.00 7500000.
STORAGE BETWEEN PUMP START AND STOP LEVELS * 1.51 TIMES (QPUNP'OTI
ASSUMED UHIf COST (EXCAVATION, LINING, ETC.) * 6.00 t/CJ.YO.
PRELIMINARY TREATMENT BY MECHANICALLY CLEANED BAR RACKS (LEVEL II
NUMBER OF SCREENS * 2
CAPACITY PER SCREEN « 96.69 CFS
SUBMERGED AREA * 32.23 SQ.FT. (PERPENDICULAR tt THE FLOW)
FACE AREA OF BARS - <>S.12 SO.FT.
INFLON BY CRAVITY (NO PUMPING) (LEVEL 21
TREATMENT BY SEDIMENTATION (LEVEL 31 - (NO ASSOCIATED STORAGE)
DESIGN OVERFLOW RATE • 1600.00 GPO/SQ.FT. (1600 SUGGESTED)
SEO TANK DEPTH « 8.00 FEET (8 FEET SUGGESTED)
NUMBER OF SEO TANKS * 8
SURFACE AREA * 9765.63 SQ.FT./TANr
NO CHLORINE ADDED
-------
TREATMENT BY HIGH RATE FILTERS
NUMBER OF UNITS
FILTER AREA PER UNIT
NAX. OPERATING RATE
SOLIDS REMOVAL EFFICIENCY
BOO REMOVAL EFFICIENCY
MAX. DESIGN HEAD LOSS
MAX. SOLIDS HOLDING CAP.
CHEMICALS MILL BE ADDED
(LEVEL
1085.32 SQ.FT
20.00 GPM/SQ.FT.
95.00 PERCENT
80.00 PERCENT
10.00 FT.
3.00 LB/SQ.FT. (AT MAX H AND Q>
CO
NO EFFLUENT SCREENS (LEVEL 5)
OUTFLOW BY GRAVITY (NO PUMPING) (LEVEL 6)
TREATMENT BY CHLORINE CONTACT TANK (LEVEL 7)
NUMBER OF DOSING UNITS = <»
DOSING RATE PER UNIT = 8000.00 IB/DAY
MAXIMUM DEMAND RATE = 26077.ID LB/OAY
VOLUME OF CONTACT TANK = 17<»038. CU.FT, AT 15 HIN. DETENTION TIME
PERFORMANCE PER TINE STEP
NOTE- NO BOO OR SS ARE REMOVED IN LEVELS 2, 5, & 6, REGARDLESS OF THE OPTIONS SELECTED
NO SS REMOVALS IN LEVEL 7 (CHLORINE CONTACT TANK)
LEVEL It5 REMOVALS (AT BAR RACKS AND EFFLUENT SCREENS) A*E REPORTED IN SUMMARY ONLY
»*»»* IN HIGH RATE FILTER REMOVAL INDICATES MATERIAL IN FILTER BED
NOTE I TIME HISTORY OF FLOW RATES AND CONCENTRATIONS ARE PRINTED «T FACH TIME STEP
-------
PERFORMANCE PER TIME STEP FOR BOD
TINE TOTAL INFLOWS
FLOW BOO
HRININ CFS MG/L
12110 58.57 199.
12*20
12130
12.*0
12150
131 0
13*1*
13120
13130
13t*0
13150
1*1 0
1*110
1*120
1*130
58.1*
58.26
51.33
55.91
53.27
52.50
5*.2B
59.75
69.22
82.86
99.7*
128.83
1**.*8
169.96
196.
190.
193.
215.
381.
888.
1*19.
1552.
1336.
918.
558.
363.
2*0.
150.
~ IMf
FLOW
CFS
0.0
0.0
0.0
0.0
0.0
0.0
t.t
0.0
0.0
8.0
193.0
191.0
193.0
193.0
0.0
BOD
MG/L
0. IN
OUT
REMOVED
0. IN
OUT
REMOVED
0. IN
OUT
REMOVED
C. IN
OUT
REMOVED
0. IN
OUT
REMOVED
0. IN
OUT
REMOVED
0. IN
OUT
REMOVED
0. IN
OUT
REMOVED
0. IN
OUT
REMOVED
B. IN
OUT
REMOVED
111. IN
OUT
REMOVED
282. IN
OUT
REMOVED
9*0. IN
OUT
REMOVED
*78. IN
OUT
REMOVED
P. IN
LEVEL-1
HG/L
0.0
0.0
0.00
0.0
0.0
0.00
0.0\
o.e
O.OB
0.0
8.0
0.00
o.o
o.ae
o.t
8.0
0.00
0.0
0.9
0.1*
8.0
8.0
0.80
8.0
8.0
0.00
8.0
0.0
0.00
LEVFL-3 LEVEL-*
NG/L NG/L
C.OO 0.00
0.00 0.00
0.0 0.0
0.00
0.08
0.0
0.00
1.00
8.0
0.00
0.08
8.0
0.00
0.00
0.0
0.00
0.00
0.0
0.00
0.00
8.0
0.00
0.08
8.8
0.00
0.00
8.0
0.08
0.00
0.8
111.0 118.56
110.6 82.1.9
.*3 315*5.*
201.9
201.*
.57
9*0.0
938.*
1.65
*77.0
*76.B
.97
0.0
201.37
151.99
*138.9
938.36
7*7.25
1731.2
*76.79
371.5*
2793.6
0.00
0.00
O.iO
o.e
0.00
0.00
8.1
0.00
0.00
8.0
0.00
o.ot
o.e
0.00
I. 00
0.0
0.00
O.OB
O.B
0.0?
a .00
0.0
0.00
0.00
8.0
o.ot
0.00
0.0
82.* 9
*9.50
238.3
151.99
7*. 57
552.8
7*7. 25
1*9. *5
*C*5.6
371.5*
7*. Jl
532.1
0.00
LEVEL-7
MG/L
0.00
0.00
0.0
0.00
0.00
0.0
0.00
0.00
0.0
0.00
0.00
0.0
0.00
0.00
0.0
0.00
0.00
0.0
0.00
o.tt
o.t
0.00
0.00
0.0
0.00
0.00
0.0
0.00
0.00
0.0
1.9.50
37.50
12.0
7*. 57
59.65
1*.9
1*9. *5
119.56
29.9
7*. 31
59.1.5
1*.9
0.00
- OUTFLOWS -
FLOW BOD
CFS MG A.
0.0 0.
0.0 0.
0.0 0.
0.0 0.
0.0 0.
0.0 0.
0.0 0.
o.o a.
O.U 0 .
0." 0.
192.8 37.
190.6 60.
180.7 119.
1*8.* 59.
0.0 0.
- INFLOWS -
FLOW BOO
CFS NG/L
58.6 199.
58.1 196.
58.3 190.
58.3 192.
55.9 215.
53.3 380.
52.5 878.
5*. 2 1*16.
59.8 15*9.
69.2 1333.
82.9 916.
99. 7 5*9.
120. 8 363.
i**.; 2*j.
170.0 150.
- STORAGE - OUTFLOWS -
STORAGE DEPTH FLOW BOO
CF FT CFS MG/L
.S3E»05 .2 0.0 0.
. 88E»05 .* 0.0 0.
.12E»06 .5 0.0 0.
. 16E»06 .6 0.0 0.
.19E»06 .ft 0.0 0.
. 22E»06 .9 0.0 0.
.25E»06 1.0 0.0 0.
.29E»06 1.1 0.0 0.
. 33E»06 1.3 0.0 0.
. 26E»D6 1.0 193.0 202.
.19E»06 .8 193.0 9*0.
. 15E»P6 .6 193.0 *78.
.11E»06 .* 193.0 63*.
.20E»06 .8 0.0 0.
-O¥f KFLOH
- BYPASS -
FLOW BOO
CFS HG/L
0.0 0.
0.0 0.
0.0 0.
o.e o.
o.e o.
O.C 0.
0.0 0.
0.0 0.
0.0 0.
0.0 0.
0.0 b.
0.0 0.
0.0 0.
0.0 0.
-------
14140
14150
15» 0
15110
15120
15130
151*0
£ 15 ISO
Ul
16 « 0
16110
16120
16130
16140
16150
171 0
197.97 9*. 0.
2?0.74 66. 193.
0
0
260.47 52. 193.0
291.38 46. 193.
320.45 43. 193.
347.35 39. 193.
371.32 36. 193.
392.35 3*. 193.
*10.26 33. 193.
425.25 32. 19S.
448.47 32. 193.
446.01 32. 193.
469.46 31. 193.
461.79 30. 193.
476.69 30. 193.
0
0
0
0
8
8
0
0
0
0
0
0
OUT
REMOVED
0. IN
OUT
REMOVED
SB. IN
OUT
REHOVCO
351. IN
OUT
RENO VCD
113. IN
OUT
REMOVED
70. IN
OUT
REMOVED
56. IN
OUT
REMOVED
45. IN
OUT
REMOVED
40. IN
OUT
REMOVED
37. IN
OUT
REMOVED
35. IN
OUT
REMOVED
32. IN
OUT
REMOVED
30. IN
OUT
REMOVED
26. IN
OUT
REMOVED
27. IN
OUT
REMOVED
2fi. IN
OUT
REMOVED
0.0
0.00
0.0
0.0
0.00
58.1
.36
350.7
.79
112.8
112.4
.**
70.2
69.8
.37
56.4
56.1
.35
*5.5
*5.1
.34
39.6
39.5
.33
37.3
37.0
.32
35.2
34.8
.32
32.2
31.9
.32
30.5
30.2
.31
28.1
27.8
.31
27.0
26.6
.31
25.6
25.3
.31
0.10
0.0
".00
0.00
0.0
58.07
43.54
2568.1
349.94
269.76
2691.6
112.37
84.44
3574.6
69.76
52.19
5499.8
56.09
41.88
91*6.5
45.12
33.66
15266.8
39.50
29.46
1936*. 7
36.96
27.56
16118.1
3*.83
25.98
17076.5
31.86
23.76
15617.7
30.16
22.69
14783.5
27.79
20.73
13626.1
26.6*
19.87
13060.8
25.36
18.87
12*0*. 0
0.80
0.0
0.00
0.00
0.0
*3.5*
13.06
59*.*
269.76
80.93
986.8
8*.**
25.33
1101.7
52.19
15.66
263.3
41.88
8.38
2*1.8
33.66
6.73
50.*
29.46
8.8*
9*.*
27.56
8.27
135.7
25.98
7.79
17*. 5
23.76
7.13
120.2
22.49
*.5»
130.0
20.73
*.15
119.8
19.87
3.97
114.9
18.67
3.77
109.1
0.00
0.0
e.oo
0.00
0.0
13.06
1.06
12.0
80.93
6*.7S
16.2
25.33
13.33
12.0
15.66
3.66
12.0
8.38
*.19
4.2
6.73
3.37
3.4
8.84
4.42
4.4
8.27
4.13
7.79
3.90
3.9
7.13
3.56
3.6
4.50
2.25
2.2
2.07
2.1
3.97
1.99
2.0
3.77
1.89
1.9
0.
155.
ISO.
155.
192.
192.
156.
156.
156.
156.
192.
192.
192.
192.
0
6
4
2
4
7
6
6
6
6
9
9
9
9
192.9
0. 196.8 9*. .31E»06 1.3 0.0 0. 0.0
1. 226.7 66. .33E»«6 1.3 193.0 351. 0.0
65. 260.5 52. .36E»I6 1.* 193.0 113. 0.0
13. 291.* *5. .*1E*06 1.6 193.0 78. O.fi
*. 320.* *3. .*7E«06 1.9 193.0 56. 0.0
*. 3*7.* 39. .56E»86 2.2 193.8 *5. 8.0 0.
3. 371.3 36. .66E»06 2.6 193.0 *0. 0.0
*. 392.* 3*. .77E*I6 3.1 193.0 37. 0.0
*. *10.3 33. .96E+06 3.6 193.0 35. 0.0
*. *25.2 32. .10E»07 *.l 193.0 32. 0.0
*. **8.5 32. ,12E*07 *.7 193.0 38. 0.0
2. **8.0 32. .13E+07 5.3 193.0 28. 0.0
2. *69.5 31. .15E+07 6.0 193.0 27. 0.0
2. *61.6 30. .17E»07 6.6 193.0 26. 0.0 0.
2. 1.76.9 30. .16E»07 7.3 193.0 25. 0.0
17110 469.52
31. 193.0 25. IN 24.7 24.36 ' 18.17 3.63 192.9 2. 469.5 31. .20E»07 6.0 193.0 2*. 0.0
-------
SUHHARY OF TREATMENT EFFECTIVENESS
VO
TOTALS
IN»UT
OVERFLON (BYPASS)
TREATED
REMOVED
RELEASED
REMOVALS
LEVEL 1
LEVEL 3 trOTALI
LEVEL 4
LEVEL 5
LEVEL 7
TRASH-
BAR RACKS
FLOW (M.G.I BOD (LBI
115.202 51178.1
9.000 0.0
115.202 51178.1
1.512 ".6355.3
113.699 4822.8
FLOW (N.G.I BOD (LBI
.005 259.1
.205 12941.3
1.332 29739.2
0.000 0.0
0. OCa 3415.7
SS (LBI
l%9615.Pi
0.0
149615.8
142041.7
7574.1
SS (LB)
5181.)
66742.9
70116.9
0.9
0.0
COLIF (NPNI
1.4%E*16
0.
1.44E+16
1.44E+16
3.70E+11
= BAR RACKS
= SEOINENTATION
= HIGHRATE FILTERS
= NO EFFL. SCREENS
CONTACT TANK
690.92% CU.FT (AT 50 LB/CU.FT.t
EFFLUENT SCREENS
REMOVAL PERCENTAGES
OF OVERALL
OF TREATED
CONSUMPTIONS
LEVEL 3
LEVEL %
LEVEL 7
TOTAL
INPUTS
0.000 CU.FT (AT 50 LB/CU.FT.)
FLOW (VOLI BOD (LB)
1.31 90.58
FRACTIONS 1.31 90.58
(LBI
REPRESENTATIVE VARIATION
TIME
HATER
AV. FLOW (CFSI
BOO
ARRIVING (NG/LI
RELEASED (HG/LI
REDUCTION (PCTI
S. SOLIDS
ARRIVING (NG/LI
RELEASED (NG/LI
REDUCTION (PCTI
COLIFORNS
ARR (NPH/1BOMLI 0
REL (MPM/lflflMLI 0
REDUCTION (PCTI
12/10
0.00
.00
.00
.00
.00
.00
.00
•
•
0.00
CHLORINE POLYMERS
0.0 0.0
0.0 3839.1
5784.? 0.0
57»%. 2 3839.1
SS (LBI
94.94
94.91.
COLIF (HPNI
100.00
100.00
= SEDIMENTATION
= HIGHRATE FILTERS
= CONTACT TANK
OF TREATMENT PERFORMANCE HITH TINE (OVERALLI .
14/31 16/50 19/10
8.00 192.95 192.95
t.OO 26.95 19.29
0.0» 1.98 1.41
180.00 92.63 92.66
0.00 15.71 8.16
0.00 .28 .07
188.00 98.23 99.89
0. 8.28EMS l.uOE+06
0. 1.46E»01 9. ICE +00
0.00 100.00 100.00
21/30
192.95
20.72
1.52
92.65
8.32
.08
99.06
1.48E+06
1.%OE»01
100.00
23/50 2/10
192.95 192.95
18.6% 33.26
1.37 2.45
92.66 92.62
6.89 11.82
.0% .17
99.42 98.5%
1.89E+06 1.66E+06
1. 09E»61 2.%2E*01
100.00 100.00
4/30 6/50 9/10
192.95 192.95 192.95
37.99 5%. 74 33.83
2.81 4.05 2.50
92.61 92.59 92.62
14.98 18.18 17.18
.26 .3% .32
98.28 98.11 98.16
3.55E»06 6.22E»06 Z.64£»06
6.10E+01 1.17E«02 %.87E»bl
100.00 100.00 100.00
11/30
192.95
75.38
5.59
92.58
34.73
.79
97.73
8.96E»06
2.83E»02
100.00
SUMMARY OF BYPASS FLOH AND POLLUTANTS
OVERFLOWS OCCURRED DURING 0 TIME STEPS
FLOW (CU. FTI
0.
BOO (LBI
0.0
SS (LBI COLIF (HPNI
0.0 0.
-------
SUMMARY OF FLOWS - MAXIMA, AVERAGES* AND MINIMA
ARRIVING
OVERFLOW
TO
TREATMENT
LEVEL 3
REMOVAL OUTFLOW
LEVEL k
REMOVAL OUTFLOW
LEVEL 7
REMOVAL OUTFLOW
RECOMBINED
RELEASE
FLOW RATES IM.G.D.I
MAXIMUM
AVERAGE
MINIMUM
12%.758
110.618
0.000
0.000
O.OOG
0.000
lZt.758
110.618
0.000
7.970
.197
.065
12*.667
110.M6
116.782
2J.b38
1.250
0.000
13*,. 68 7
109.166
99.906
I*. 000
0.003
a.QUO
12<».687
109.166
' 99.906
I2(i.687
1C9.166
0.000
\O
HAXIHUN
AVERAGE
MINIMUM
BOD CONCENTRATIONS IHG/L1
940.0
e!o
0.0
0.0
0.0
980.9
1126.6
3975.U
7I».6
.7
25558.3
335.7
8.0
.0
0.0
0.0
0.0
335.7
8.C
.0
335.9
8.0
0.0
HAXIHUH
AVERAGE
MINIMUM
COLIFORM CONCENTRATIONS INPN/IOONL)
Z.Z6E*07 0. Z.Z6E*07
2.9*E*06 0. ?.<
0. 0. 0.
2.15F»03
7.78E»01
0.
2.15EH3
7.78E*01
0.
-------
INLET HYDROGRftPH - CFS
00
EXTERNAL
ELEMENT
NUMBER
10
TIME STEP
1
58.571
82.858
31.7.353
V69.522
V50.5V9
317.758
215.15V
15V.090
117.518
9*. 992
79.506
67.362
"8.38V
V8.979
VV.738
Z
58.139
99.71.5
371.320
VB?.7?6
1.38.709
305.083
2A7.553
1V9.V7
11*. 969
93.201
78.181.
66.215
56.60*.
V8.3BC
fcV.370
3
58.255
120.827
392.35V
1.69.500
1.26.10V
293.160
200.259
1V5.B29
112. V7V
91.50V
76.762
65.101.
55. VI 3
•.7.357
VV.993
V
54.325
1VV.V80
V10.2S8
V80.V12
V12. 755
281.686
193.317
1 VI. 068
110.006
89.857
75.595
6V. 026
5V.V05
V7.19S
V3.870
5
55.911
169.962
V25.250
V7V.569
398.857
276.657
186.7VO
136.903
107. V22
88.2V2
7V.VOV
63.288
53.510
V6.960
V3.78V
6
53.267
197.966
VV8.V7V
V78.338
38V. 922
260.362
180. 5V3
133.191
105.220
86.685
73.198
62.157
52.719
V6.65S
V3.777
7
5Z.V99
228. 7V1
VV8.009
V78.853
371.053
250. V36
173.577
129.715
103.053
85.GV2
71.976
61.05V
53.299
V6.32C
V3. 762
8
5V. 202
260 .V68
V69. V60
V75.V8V
357.375
2V0.975
168.798
126. V33
100.953
87.636
70.812
60.033
51.677
V5.9V2
1.3.808
9
59.750
291.377
V61.79V
169.V29
3V3.528
231.973
163.802
122.169
98.9VV
82.237
69.723
59.075
50.5V5
V5.5V6
V3.85S
10
69.216
320. VV7
V76.888
V60.78V
333. V30
223.319
158.902
120.03V
96.98V
80.857
68.5V7
58.165
V9.559
VS. 139
V3.923
INLET POLLUTOGRAPHS
EXTERNAL
ELEMENT
NUHBER
10
CIHE STEP
1
V3.606
28V. 327
51.086
5V.692
52.V12
59.V5V
7V.2V3
37.708
26.65*
19.855
10.613
11.98G
23.1*32
1.9.109
56.761
2
1.2.686
205.108
50.303
59.8«.9
52.V37
60 . 56V
66.685
37.V97
27.175
17.763
10.599
12.037
29.267
<»8.991
5*. 621
3
VI. 38 8
16V.135
50.V2D
59.098
60.929
60.538
63.386
V6.V27
27.839
16.551
10.705
12.029
3*. 966
55.53%
5V. 286
V
VI. 97*
129.83%
50.138
60.009
75.535
60.V17
63.939
6V. 622
27.526
16.531
10.889
12.031
36.703
70.391
54.228
5
BOO IN
VS. Oil
95.360
51.276
SB. 929
79.602
65.88V
6V. 032
7V. VI 3
25.986
16.576
lu.90V
13.716
37.001
80.1V5
52.563
6
LB/NIN
75.797
69.756
53.593
59.926
77.705
75.621
63.966
7V. 13V
22.6V2
16.562
IP. 879
17.628
36.967
B2.VOO
V8.V01
7
17Z.662
56.68V
52. 968
56.781
78,u37
79.011
57.618
73.75?
20.751
15.222
10.882
20.587
37.6V1
82.617
V5.218
8
287.599
50.6*1
53.667
52.9V5
78.115
78.055
«.«..« 87
73.905
20.805
12.350
10.88V
21.223
tO. V9V
82. 660
*V.VV4
9
3V6.691
V9.658
52. VIS
52.005
72.701
78.115
36.563
63.592
20.866
10.639
11.107
21.189
V5.938
77.513
VV. 3V5
10
3V5.699
51.V08
53.779
S2.6V9
62.965
78.175
37.381.
V0.721
20.838
10.559
11.618
21.197
V8.V18
65.50V
VV.322
SUSPENDED SOLIDS IN LB/NIN
Iw
29.V86
3637.961
VV.952
V3.33*
39.1.93
50.066
53.691
26.575
2V.V08
22.109
9.312
22.920
23.306
V6.9V7
26.180
38.719
2857.935
%2.171
V8.368
39.512
50.<.60
V6.09G
26.1.50
2V. 295
21.998
9.283
23.V65
2V. 022
V6.979
25.392
132.515
2227.587
VI. 139
V7.911
VS. 207
50.V86
V2. 716
29.167
2%. 555
21.831
9.735
23.V32
25.967
V5.V50
25.236
335.895
1608.771
V0.118
V8.V67
55.06V
50.V19
V3.239
3V.9BU
2V.VV9
21.796
10.688
23.1.50
26.716
1*1. 10V
25.205
502.515
967.019
VC.65V
V7.6V7
57.870
52.655
V3.326
38.52V
23.982
21.808
11.191
23.176
26.892
36.VV3
2V. 900
77V. 605
V75.211
V2.173
V8.V30
56.562
56.599
V3.259
38.57V
22.901
21.806
11.219
22.808
26.8V9
3V. 880
2V.UVV
1V87.07S
226.881
VI. 567
VV.900
56. 78 *
57.928
39. 3VV
38. V15
22.1V7
19.053
11.211
23.099
29.125
3V. 536
23.315
2V 7V. 576
117.677
I.2.023
V&.133
56.8V1
5'.5VO
30.850
38.V71
22.116
13.096
11.216
23.305
35.395
3V.V6U
23.10V
3339.552
67.550
V0.996
38.98Z
5V. 932
57.579
25.938
35.776
22.1VV
9.V77
. 13.716
23.265
V2.820
32.831
23.0G5
3863. V62
51.0V5
V2.03V
39.762
51.V63
57.597
26.360
29.25V
22.132
9.220
19.273
23.263
V5.998
29.0V3
23.057
-------
TREATED OUTFLOH HYDROGRAPH - CFS
EXTERNAL
ELEMENT
NUMBER
10
TIME STEP
1
O.uOQ
192.920
192.691
192.891
192.891
192.891
192.891
192*891
192.891
192.891
192.891
192.691
192.891
192.891
192.891
2
0.000
190.606
156.589
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.691
3
0.000
180.662
156.634
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.882
4
P. 030
148.366
156.634
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.833
5
0.000
O.OOu
156.634
192.391
192.891
192.891
192.991
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.841
TREATED OUTFLOW
EXTERNAL
ELEMENT
NUMBER
10
TIME STEP
1
0.000
27.028
3.U17
1.310
1.072
1.112
1.101
.987
1.663
1.671
2.502
2.922
1.158
1.615
4.034
2
0.090
42.506
1.970
1.250
1.C58
1.105
1.070
.998
1.719
1.635
2.702
3.026
1.151
1.554
3.l»79
3
0.000
80.748
2.588
1.221
1.U21
1.095
1.037
1.&03
1.728
1.606
2.763
2.925
3.442
1.169
6.466
4
O.OOP
32.971
2.421
1.198
1.413
1.093
1.036
1.011
1.738
1.644
2.829
2.403
2. 003
1.162
12.574
5
600 IN
0.000
0.01.0
2.282
1.176
1.025
1.110
.968
1.028
1.772
1.704
2.916
1.971
2.136
1.240
26.004
6
O.OOC
0.000
192.891
192.091
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
0.000
7
J «ObO
155.625
192.891
192.891
192.891
193.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
c.ooc
8
0.000
150.360
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
192.891
C.OOO
9
o.oco
155.192
192.891
192.891
192. 891
192.891
192.891
192.891
192.891
192.891
192.891
192,891
192.891
192.891
0.000
10
o.oco
192.369
192.891
192.891
192.891
192.891
192.691
192.891
192.891
192.891
192.891
192.891
192.H91
192.891
0.000
POLLUTOGRAPHS
6
L9/MIN
O.OQP
0.000
2.570
1.173
1.025
1.110
.968
1.032
1.810
1.758
3.u2U
2.047
2.103
1.300
0.000
7
i .000
.617
1.622
1.153
1.024
1.098
.964
1.231
1.846
1.789
2.981
2.134
1.802
1.404
0.000
8
0.000
36.393
1.495
1.115
1.029
1.083
.960
1.231
1.881
1.843
2.817
2.514
1.8ns
1.579
0.000
9
O.PCO
7.735
1.433
1.095
1.025
1.081
.962
1.500
1.842
2.326
2.745
3.798
1.841
1.978
c.oco
10
0.000
2.630
1.361
1.084
i.ior
1.102
.987
1.577
1.819
2.216
2.811
.863
1.623
3.064
3.000
-------
SUMMARY OF TREATMENT COSTS
ASSUMED FUTURE ENGINEERING NENS RECORD INDICES
CONSTRUCTION - 20 CITY AVERAGE
YEA*
1971
1972
1973
197*
1975
1976
1977
197«
1979
I960
ENR INDEX
1363
15*5
1712
1886
2035
2188
2312
2431
2537
2630
2710
COST PARAMETERS . .
INTEREST RATE
AMORTIZATION PERIOD
CAP. RECOVERY FACTOR
YEAR OF SIMULATION
SITE LOCATION FACTOR
7.00 PERCENT
25 YEARS
.0*58
1975
i.oaoo
NJ
o
o
UNIT COSTS . .
LAND =
POWER *
CHLORINE *
POLYMERS »
ALUM »
20008.10 S/*CRE
.020 f/KMH
.200 S/LB
1.250 »/L8
.03 I/LB
TREATMENT
BAR RACKS
NO INLET PUHPIM6
SEDIMENTATION
HI CURATE FILTERS
NO EFFL. SCREENS
NO OUTLET PUMPS
CONTACT TANK
LEVEL
1
2
3
%
5
6
7
CAPITAL COSTS
INSTAt LAND
517061 2*89.
0
0
•»**9585
0
•
829759
0.
229*8*.
7968.
0.
0.
1*697.
ANNUAL COSTS
INSTAL LAND
**369. 169.
o. a.
0. 1606*.
381821. 558.
0. 0.
0. 0.
71202. 1029.
NIN NAINT
5171.
0.
0.
88992.
0.
0.
1659S.
STORM
CHLORINE
0.
0.
0.
0.
0.
0.
1157.
EVENT COSTS
CHEN OTHER
P **2.
0
0
9118
0
0
0
0.
576.
1165.
0.
0.
15.
SUBTOTAL S 5796*0*. S 25*559. S *97392. S 17819. S 110757.
TOTAL t 6050963. S 625969.
1157. S 9118. t 2198.
f 12*72.
TOTAL PER
TRIB ACRE
3117.
209.
TOTAL LAND REQUIREMENT
7.96 ACRES.
-------
600
I i I i i I i r
5 10 15 20
Time from Start of Storm, hrs.
(a) Hydrograph
70-i
Scheme 1 = Conventional
Treatment
Scheme 2 - Physical-
Chemical w/o
Storage
Scheme 3 » Physical-
Chemical with
Storage
0
0
FIGURE 22,
5 10 15 20
Time from Start of Storm, hrs
(b) BOD, Pollutograph
=LJOU llYDROGRAPH AND POLLUTOGRAPHS FROM STORAGE/TREATMENT
:K
201
-------
1000
900 -
800 -I
700 -
c
1-4
E
600
500 -
C
in
"S 400
T3
C
(1)
§ 300 "
200
100
I i
5 10 15
Time from Start of Storm, hrs.
(c) Suspended Solids Pollutograph
20
25
FIGURE 72, CONTINUED
202
-------
RECEIVING WATER STUDY GUIDE
PREPARATION OF RECEIVING WATER BLOCK DATA
Data Gathering
Prior to application of the Receiving Water model, data for physical and
meteorological characteristics as well as data to define boundary and
initial conditions are required. Most data are either available or measur-
able; however, the complete description of the system may require consider-
able field work.
Physical data include tidal variations (for estuarine systems), water
surface elevations and areas, water depths and roughness coefficients.
Tidal variation data can be acquired from the National Oceanographic and
Atmospheric Administration. Water surface elevations (referenced to datum)
may be available at some points in the field where stage recorders are in-
stalled. Otherwise, the stage can be computed by that portion of Receive
which models the hydrodynamics of the receiving water body. Surface areas
and water depths can be determined from U.S. Coast and Geodetic Survey
nautical charts or U.S. Geological Survey quadrangle maps. Manning's
roughness coefficients can be estimated using standard tables as given in
Reference 5. Meteorological data required include rainfall, evaporation,
and wind velocity and direction. These data, or information on additional
sources of data, should be available from the U. S. Weather Bureau.
A required boundary condition is the mode of downstream flow control;
options available are tidal exchange, dam (weir) or specified outflow. For
tidal conditions, a tidal exchange ratio i.e. the ratio of the returning
outflow from prior ebb tide to the total inflowing flood tide must be
supplied. The tidal exchange ratio can be estimated from dye tracer or
pollutant analysis. For weir boundary conditions, the weir factor, W-, power
law for the weir, X, and height of the weir, H, referenced to the datum must
be supplied. The SWMM then uses Equation 9 to calculate flow over a weir.
Q - W, HX (9)
f
where Q - Flow, cfs
Wf • Weir Discharge Coefficient (Cd) x Weir Width (ft)
X - Power Law for Weir
203
-------
For a rectangular suppressed weir, values of 3,33 for C, and 1.5 for X
are generally accepted. Hydrograph data must be supplied to the model for
specified outflow boundary conditions. Data to generate the hydrograph
can be taken from stage-discharge rating curves based on the type of down-
stream boundary.
Optional boundary conditions are constant inflows to, or outflows from,
any of the modeled nodes (junctions). Either quantity or quality consti-
tuents can be introduced or removed from the system in this manner. This
option is mainly used to introduce flows at the upper end of the water body
to be modeled; quantity and quality constituents must be measured in the
field.
At the beginning of simulation, the initial conditions required for in-
put include velocity of channel flow, pollutant concentrations and dissolved
oxygen saturation level (at the boundary nodes), reaeration coefficient,
and first order decay coefficient for non-conservative pollutants. Velocity
of channel flow can be derived from flow/area data for stream segments.
For estuaries or lakes, velocity data can either be taken from a previous
computer simulation or set equal to arbitrary values and allowed to equi-
librate during the initial day of simulation. Pollutant concentrations
at the boundary node must be measured. Saturation concentrations for dis-
solved oxygen can be found from tables based on water temperature and
chloride concentrations, (Page 480 of Reference 12 contains such a table).
Reaeration coefficients and first order decay coefficients can be estimated
from laboratory tests conducted on the receiving water or from values found
in the literature for waters of similar characteristics. In addition to
the concentrations at the boundary nodes, an optional initial condition is
the concentration of pollutants at nodes to be modeled. Because these
parameters would be impractical to measure in the field, they can either be
estimated from previous computer simulations or set arbitrarily.
Discretization of the Receiving. Water
The model simulates the receiving body of water as a system of links
(channels) and nodes (junctions) organized in a linear, triangular or
higher order polygon configuration. The SWMM is designed to accept up
to 100 junctions and 225 channels. Figure 23 shows a representative re-
ceiving water divided into linear, triangular, quadrilateral and pentagonal
elements; here, circled numbers refer to channels and uncircled to junctions.
Choice of the discretization scheme to be used is based on engineering
judgments and will generally depend on the size and complexity of the re-
ceiving water system. The maximum permissible time step depends on the
values of channel length and depth as given by Equation 10.
At 0.75 (L/v/gd) (10)
204
-------
10000 9000 8000 7000 6000 5000 4000
3000
NJ
O
Ul
Scale: 1 in: 1000 ft
FIGURE 23, REPRESENTATIVE RECEIVING WATER,
-------
where At = time step length, sec,, usually between 30 and 300 sec,
L = length of channel, ft
d = average depth of channel, ft
Channels must be of sufficient length to provide a realistic time step
size. The grid system should be arranged so that it portrays all signifi-
cant changes in receiving water geometry and profile.
Junction and Channel Data Preparation
Required data preparation involves describing the physical system and
boundary and initial conditions. Boundary and initial conditions have been
discussed under data gathering. Physical system data can be divided into
junction and channel data.
Junction data include surface elevation, surface area and depth. The
surface elevation at the junction at the start of the simulation is referenced
to the standard datum chosen. For estuarine systems the datum is usually
taken as mean low low water. For bodies of water where surface elevation
is governed by a downstream weir, the elevation of the top of the weir may
provide a convenient datum. Any datum may be used as long as all measure-
ments are referenced to that elevation.
Surface area attributable to each junction can be found by constructing
the Thiessen polygon surrounding the junction and planimetering the area.
gons are constructed by locating perpendicular bisectors through the lines
joining that junction and all adjacent junctions. Each of these perpendicular
bisectors represents the locus of points equidistant from two junctions. The
closed figure formed by these bisectors defines the area belonging to each
junction. A typical polygon thus formed is designated as a-b-c-d-e^f-a in
Figure 24. In Figure 24, perpendicular bisectors have to be drawn through
some lines that are not being modeled as channels; for example, line b-c
of Figure 24 is the perpendicular bisector of the line drawn between junctions
7 and 8.
Depth is the distance measured from the junction surface to the bottom
of the water body. According to this convention, depth measurements are
positive in the downward direction.
Figure 24 shows the Thiessen polygon network constructed for the
receiving water shown previously in Figure 23, Table 18 gives a tabulation
of the junction data for this sytem. Figure 24 and Table 18 will be used for
the case study.
Channel data include length, average width and average depth. These
are in addition to Manning's roughness coefficient and initial channel velo-
city which were discussed previously. The length of the channel is the
distance along the channel centerline between the junctions on either end.
The channel axis need not be a straight line. The average width is taken
206
-------
to
o
10000 9000 8000
I
7000 6000 5000 4000 3000 2000 1000
Station
Scale: 1 in: 1000 ft.
FIGURE 24, SAMPLE RECEIVING WATER SHOWING THIESSEN POLYGON CONSTRUCTION,
-------
Table 18. JUNCTION DATA
to
o
CO
Junction.
lumber
1
2
3
4
5
6
7
8
9
10
11
12
13
Surface
Elevation
(ft)
0
0
0
0
0
0
0
0
0
0
0
0
0
Surface
Area
(106ft2)
1.38
2.59
2.04
0.99
1.58
4.32
3.55
"3.46
3.39
1.85
2.09
2.01
1.25
Constant Junction
Flow In
(cfs)
1000
Constant Junction
Flow Out
(cfs)
Junction Depth
(ft)
15.0
13.0
13.0
14.0
13.0
11.5
11.5
9.5
10.4
7.0
10.2
10.1
10.0
-------
as the length of the perpendicular bisector through the channel. Thus, in
Figure 24 distance a-b would be the average width of channel 12 and distance
d-g the average width of channel 8, Assuming a uniform slope between
junctions, the average depth of a channel is the average of the junction
depths on either end. Table 19 gives the channel data derived for the system
shown in Figure 24 and will be used in the case study.
For triangular networks, the program has the capability of generating
junction areas and channel lengths and widths for the interior of acute
triangles. Required data are the coordinates of each junction used in the
computations. Since the triangulation subroutine computes data only for the
interior of triangles, the area assigned to junction 2 of Figure would
be area 2-k-l-g-h-2 and the width of channeT 1 would be distance 1-k. The
user must add area 2-h-i-j-k-2 to junction 2 in card group 15. Similarly,
an additional channel must be modeled between junctions 1 and 2 to account
for the channel defined by width k-j. Therefore, unless an extensive network
of triangular elements is to be modeled, it is recommended that geometric
data be determined using the Thiessen polygon method.
Having described the receivine water system, the model can be run using
the output of the Runoff, Transport or Storage/Treatment Block (or a
combination of two or more) as the input to the Receiving Water Block.
Several optional data inputs may be used to allow the user to more accurately
model the receiving water system provided that such data are available.
Optional Data Input
Options available in the quantity portion of the Receiving Water Block
include multiple downstream boundary conditions; spatially variable rainfall;
stonnwater input from cards; parallel channels between the same two junctions;
Manning's roughness coefficient computed at each time step; and plots of
stage versus time for selected junctions. Quality options include the
ability to re-start the quality simulation; variable oxygen saturation co-
efficients; and variable reaeration coefficients. Each of these options will
be described briefly. Table 20 has been prepared to aid the user in deter-
mining which card groups are to be completed for a standard run and for the
various options available. A more detailed set of data preparation
instructions are contained on pages 273-288 of the User's Manual (Reference 6).
Multiple downstream boundary conditions can also be simulated•for systems
influenced tidally, systems controlled by weirs, or a combination of these.
For example, multiple boundary conditions could be used to represent more
than one outlet to the sea in the case of a tidal system, or a dam with spill-
ways at two different elevations in the case of a weir controlled system.
An estuarine system with relief works to control the water level would re-
present a combination of a tidal and weir controlled system.
Another optional input is spatially variable rainfall where two or more
rainguages are used to gather rainfall data at different locations in the
system. However, the User's Manual states, that unless the receiving waterii
are shallow or are characterized by low flows, rainfall is likely to have
209
-------
TABLE 19. CHANNEL DATA
N>
I-1
o
Channel
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Lower
Junction
1
1
3
4
3
2
2
3
6
6
5
7
8
8
9
10
11
12
1
Upper
Junction
2
3
4
5
5
3
6
6
8
7
7
9
9
10
11
12
12
13
4
Length of
Channel
(ft)
2325
2000
1225
1025
950
1475
1675
1725
2100
1800
1525
2050
1550
2600
2550
1225
1125
1000
1250
Average
Width
(ft)
750
1250
1000
650
1150
1450
1400
950
1575
1350
1600
1550
1500
950
1075
1125
1250
2250
500
Average
Depth
(ft)
14.0
14.0
13.5
13.5
13.0
13.0
12.3
12.3
10.5
11.5
12.3
11.0
10.0
8.3
10.3
8.5
10.1
10.0
14.5
Manning ' s
n
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
Initial
Velocity
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
-------
Table 20 . RECEIVE DATA PREPARATION REQUIREMENTS
I. Standard Run
Card Groups
1,2,3
4
5,6,7,8,9
11,12
13
15,16,17,18,19,20,21,30,31,33,34,
35,36,37
II. Multiple Boundary Conditions
Card Groups
4
10
11,12
13
III. Spatially Variable Rainfall
Card Groups
4
23,24,25,27
IV. Stormwater Input from Cards
Card Group
5
26
28
33
Comments
ISWCH(l) » 0, 1 or 2
if modeling tidal system
downstream weir boundary
Comments
ISWCH(l) = 3
tidal system
wier boundary
Comments
ISWCH(3) - 1
Comments
NJSW
if spatially varied rainfall is input.
if spatially varied rainfall is not
input.
NJSW
211
-------
Table 20 , continued,
IV. Stormwater Input from Cards, Continued
Card Group Comments
38,39
V. Parallel Channels
Card Group Comments
4 ISWCH(5) = 1
5 NCGT
VI. Triangles used to Compute Geometric Data
Card Group Comments
4 ISWCH(4) = 1
ISWCH (6)
17 NTEMP(3) = third junction making up
triangle. Also: new values for
variables ALEN, WIDTH, RAD
VII. Manning's Roughness Coefficient Computed at Each Time Step
Card Group Comments
4 ISWCH(7) » 1
14
VIII. Quality Restart Option
Card Group Comments
31 For first run ISWCH(3) = 1
for subsequent restarts ISWCH(l) » 1
212
-------
Table 20, continued
IX. Variable Oxygen Saturation Coefficients
Card Group Comments
31 ISWCH(7) = 1 or 2
34 TEMP & ICL required if ISWCH(7) - 2
35 CSAT • boundary DO concentration
36 STT Required
X. Variable Oxygen Reaeration Coefficients
Card Group Comments
31 ISWCH(8) = 1 or 2
36 ATT required.
213
-------
little effect on the simulation. Thus, data on spatial variations in rain-
fall may not be needed.
Card input can be used for storm flows and other variable inflows to,or
outflows from, the receiving waters. Parallel channels can be uned to model
a deeper channel within a marshy area. Manning's roughness coefficients
can be calculated at each time step; however,the user must supply the n vs.
depth data from which the coefficient will be calculated. The stage versus
time relationship for up to 50 junctions can also be plotted.
The re-start option allows quality calculations to continue from the
point at which the program stopped on a prior run. Quantity calculations from
the previous run are used over again. Use of the re-start option permits a
longer quality simulation.
Finally, variable oxygen saturation coefficients and reaeration co-
efficients are optional inputs which would probably be used only for estuarine
systems. Oxygen saturation is computed based on water temperature and
chlorides concentration. Reaeration coefficients are calculated based on
the O'Connor-Dobbins formula as given in Reference 13.
CASE STUDY MATERIALS AND PREPARATION OF CODING FORMS
The receiving water shown in Figure 23 accepts the effluent from the
sewage treatment plant modeled in the Storage/Treatment Bloclc. The outfall is
at junction 10 of the receiving water which corresponds to Element 10 of the
Transport Block.
Total area of the receiving water to be modeled is 700 acres (1.09 sq.
This is a tidal system with a constant inflow of 1000 cfs at junction 13.
Pollutant concentrations of the inflow include 1.0 mg/1 BOD5, 10.0 mg/1 SS
and 100 MPN/100 ml total coliforms. Depth and area data for each of the
junctions are given in Table 18. Channel data are contained in Table 19.
A tidal stage relationship is specified for junction 1. Tidal stage is
given for four times during the initial day of simulation. The SWMM will
develop tidal elevations for the remainder of the simulation period. Tidal
stages.as referenced .to mean low low water for day one are:
Time (hrs) Stage (ft)
0.3 -1.1
5.8 .7
11.6 -2.0
18.7 2.4
214
-------
The assignment for the Receiving Water Block Workshop is to collate
the data and prepare computer coding forms as outlined in Table 8-1, pages
289-305, User's Manual. Table 20 may be of assistance.
The following additional data can be used where required:
Initial Time for Start (TZERO): 12
Evaporation (EVAP): 3.0 in/month
Wind Velocity: 0 mpK
Wind Direction: 0
Quality Constituents Modeled: BOD5, SS, Coliforms
Water Temperature: 14 C
Oxygen Saturation Concentration: 9.0 mg/1
Reaeration Coefficient: 0.1 day"
First Order Decay Exponent: 0.2 day
DO Concentration of Inflow at Junction 13: 8.0 mg/1
Datum Chosen: Mean low low water
Tidal Exchange Ratio: 0.80.
SAMPLE DATA INPUT AND PROGRAM OUTPUT: RECEIVING WATER BLOCK
Data contained in Tables 18 and 19 and on pages 220 and 221 were used as
input for the Receiving Water Block. An additional assumption made was that
the initial concentrations of constituents at all nodes would be the same
as the inflow to Node 13. The hydrograph and pollutographs shown in Figure 20
(Input to Storage/Treatment Block) were used as dynamic input into the
Receiving Water Block to show the effects of raw combined sewage on the
receiving water. A standard run, as defined in Table 20, was performed.
The input data and sample program output are provided. Figure 25 shows
the variation of dissolved oxygen concentration with time, for selected points
within the receiving water. Dispersion of the storm water pollutants away
from the simulated outfall can be seen. In this case, pollutants have not been
completely dispersed and the system has not returned to its pre-storm state
during the five days simulated. Therefore, it may be desirable to use the
restart option and increase the length of quality simulation. The significant
quality oscillations shown in Figure 25 are the result of the tidal variation
and the high ocean exchange ratio for the system.
215
-------
9 10
1 2 *
RECEIVIM
QUANTITYQUALITY
DECEIVING HATER
CASE ST'jnv
OESIGN SfORN
1C YEAR, 2 HOUR
1
5 24. 1.
* 45
2.16
. 44
1
9
1 2
6 8
11 12
1
1 4 50
.3 -1.
I 0.
2 3.
3 0.
<* C.
S U.
6 0.
7 0.
8 0.
9 0.
1C 0.
11 0.
12 0.
13 u.
99999
4 7
RECEIVING
BLOCK WORKSHOP
DURATION STORM
40. 12. 13
145G. .84
1490. 1.92
153C. .48
2 ?
10 11
13 14
87 57
12 13
2 1
13
1 5.1
1.31
2.59
2. "4
P. 99
1.5S
4.32
3.55
3.46
3.39
1.65
2. 89
2«*1
1.251CQO.
CARD
GROUP
•N
WATER BLOCK CARD DATA
-
, TOTAL RAINFftLL=2.05 INCHES
1ft 10 3. c. J. i "• 12 1 ~
1460. 1.14 1470. 1.56 1<*9?.
15u:. l.!»4 1510. I.u2 1520.
154C . .48 156L. .34. 157C..
45671
12 13
45 35 23 2f 36~
79 « 9 « 10 9 11 1C 12
4 5 6 7 • 11 ~
.7 11.6 -2.P 1S.7 2.4
15.
13.
13.
14.
13.
11.5
11.5
9.5
10.4
7.
1C. 15
If* .05
10.
-
1
' 2
' 3
• 4
• 5
• 6
L 7
• 8
• 9
11
12
• 15
' 16
-------
ro
1
2
3
U
5
fc
7
8
9
10
11
12
IT
1<*
15
16
17
18
19
99999
1
1
•*
k
3
2
2
3
6
6
r
7
S
8
9
10
11
12
1
2
3
U
5
5
3
6
6
a
7
7
9
9
lo
11
12
12
13
it
232?.
2000.
1225.
1025.
95C.
1U75.
1575.
1725.
2100.
18 jQ.
1525.
2950.
1550.
260C.
2550.
1225.
1125.
nor.
1250.
7=0.
1251.
10CO.
650.
1151.
1«*50.
1UCO.
95?.
1575.
1350.
1600.
1550.
1500.
950.
1075.
1125.
1250.
2250.
500.
Ik.
1U.
13.5
13.5
13.
13.
12.3
12.3
10.5
11.5
12.3
11.0
10.0
6.25
10.25
9.53
10.1
in.c
1«».5
.025
.(125
.025
."25
.025
.P25
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
• 17
18
-------
WORKSHOP CASE STUDY, RECEIVING
TIME FROM START OF STORM, HPS
WATER LEVEL AT OCEAN
.2
WATER <*LOCK
1
5 3
1
1
2
3
it
5
6
7
8
9
IB
11
12
13
99999
1
1
2
3
k
5
6
7
8
9
10
11
12
13
99999
1
1
2
3
<*
5
6
7
8
9
10
11
12
13
99999
ENDPROGRAN
6
1
3.
1.
1.
1*
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
P.
10.
Iff.
10.
10.
10.
10.
10.
10.
10.
10.
10.
1C.
10.
0.
100.
luO.
100.
IfiO.
100.
100.
100.
100.
lilO.
100.
100.
ion.
100.
1 1
3D
.80 1<».
9. CO .1
10766.
9.i>0 .1
107660.
9.00 .1
3.39E+9
BOD
19
30
31
33
34
35
8.
9.
8.
8.
8.
0.
8.
8.
n.
e.
6.
8.
8.
8.
8.
8.
8.
«.
8.
8.
8.
8.
8.
8.
SUSPENDED SOLIDS
36
37
35
36
COLIFORMS
37
35
36
37
218
-------
RECEIVING WATER BLOCK WORKSHOP
CASE STJDY
E.P.I.
RECEIVING HATER HYDRODYNAMICS
DAYS SMULATEO 5
WATER QUALITY CYCLES PER 0»Y 21,
INTEGRATION CYCLES PER MATER QUALITY CYCLE 9D
LENGTH 9F INTEGRATION STEP IS J.8. SECONDS
INITIAL TIME 13.80 HOURS
EVAPORATION RATE. 3.0 INCHES PER HONTH
WINO VELOCITY, 0. NPH KINO DIRECTION, 0. DEGREES FROM NORTH
ESTURIAL SYSTEM
WRITE CfCLE STARTS AT THE 1 TINE CYCLE
VO
R»IM IN INCHES PER HOUR, AND TINE IN MINUTES, HEASUREO FRON START OF STORM
IN./HR. MINUTES
IN./HR. MINUTES
1 TO 5
6 TO 10
11 TO 15
16 TO 2D
14SO.DOt
1500.001
1560.aOi
0.000
.8VO
1.V.O
.300
0.000
1.920
s!o06
PRINTED OUTPUT AT THE FOLLOWING 13 JUNCTIONS
12 3 <• 5 6
1*60.oao
1510.000
1578.400
0.000
IN./HR. MINUTES
l.ldO
i.oza
0.006
0.000
1470.06)
15ZO.OCO
O.OC)
0.000
10
11
IN./HR. NINUTES
.8*0
0.000
0.000
12
13
ItSO.OOO
1530.000
0.000
0.000
IN./HR. HINUTES
Z.16G 1'•90.000
.MO 15*0.000
0.000 0.060
0.000 0.000
PRINTED OUTPUT FOK THE FOLLOWING 18 CHANNELS
1102
6008
1101Z
1003
6007
1Z013
HO*
5307
1.005
7009
3005
8809
ZOO 3
8010
Z006
9011
70(6
1001Z
KO IS 1 NUMBER OF TERMS IS
MAXIMUM NUMBER OF ITERATIONS IS 50 TIDE CHECK SWITCH IS-0
COEFFICIENTS FOR TIDAL INPUT WAVE AT JUNCTION 1
Al AZ A3 A* A5 A6 A7 PERIOOIHRS1
.018 -.808 -.109 .019 .531 -1.5G7 -.076 2*. 00
WHERE THE WAVEFORM IS GIVEN BY
H(JI = Al * A2.SINIHTI » A3.SINI2WTI * A*.SINC3WTI » AS.COSCWT) * A6.COS(2WT) » A7.COSC3MTI
-------
CHANNEL DAT*
CHANNEL
NUMBER
1
2
3
%
5
6
7
8
9
10
It
12
13
1*
15
16
17
IS
19
LENGTH
IFTJ
2325.
zaon.
1225.
10 25.
950.
1*75.
1675.
1725.
2100.
1800.
1S25.
2050.
1550.
2600.
2550.
1225.
1125.
1000.
1250.
WIDTH
IFTI
750 .00
1250.00
1000.00
650.00
1150.00
1*50.00
1*00.00
950.00
1575.00
1350.00
1600.00
1550.00
1500.00
950.00
1075.00
1125.00
1250.00
2250.00
500.00
AREA
ISO. FTI
1050(1.
17500.
13500*
8775.
1*950.
18850.
17220.
11605.
16538.
15525.
19680 .
17050.
15000.
7838.
11019.
9596.
12625.
22500.
7250.
MANNING
COEF.
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
.025
VELOCITY
IFPS»
-0.00
-O.OG
-0.00
"3.00
-0.00
-J.OO
-0.00
-o.ao
-o.on
-0.00
-0.00
-0.00
-0.00
-0.00
-0.00
-0.00
-0.00
-O.UO
-0.00
HYO RADIUS
tFTI
14.0
11.. 0
13.5
13.5
13.0
13.0
12.3
12.3
10.5
11.5
12.3
11.0
10.0
8.3
10.3
a. 5
10.1
10.0
1*.5
JUNCTIONS AT ENDS
1
1
3
l»
3
2
2
3
6
6
5
7
a
8
9
10
11
12
1
2
3
k
5
5
3
6
6
8
7
7
9
9
10
11
12
12
13
<»
MAX TIME STEP
(SEC)
IDS.
93.
58.
71.
83.
85.
11Z.
92.
75.
107.
8%.
155.
137.
72.
61.
5*.
57.
EXCEEDED BY
STEP OF <»0
to
ro
O
JUNCTION
NUMBER
1
2
3
*
5
6
7
8
9
10
11
12
13
J U N C T I 0
INITIAL HEAD
IFTI
U. 00
0.00
0.00
0.00
• .BO
0.00
O.Od
0.0"
0.00
0.00
0.00
0.00
0.00
N 0 A
DEPTH
IFTI
15.0
13.0
13.4
1*.0
13.0
11.5
11.5
9.5
10.*
7.0
10.2
10.1
10.0
T A * • » ' •
SURFACE AREA
I10«6 SO FTI
1.38
2.59
a.o«.
.99
1.58
*.3?
3.55
3.*6
3.39
1.85
2.09
2.01
1.25
INPUT
tCFS>
-0.
-u .
-0.
-0.
-8.
-0.
-U.
-0.
-0.
-p.
-0.
-8.
1000.
TOTAL A*EA FOR THE SYSTEH
3C.50 * 19»»6 SO. FT
OR
OUTPUT
ICFS)
-0.
-0.
-0.
-a.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-C.
1.09
CHANNELS ENTEFING JUNCTION
1
6
3
*
11
9
12
13
15
16
17
18
18
SQ
2
7
5
3
-------
RECEIVING HATER BLOCK WORKSHOP
C«SE STUDY
E.P.A.
RECEIVING HATER HYDRODYNAMICS
to
DESIGN STORH
DAY IS 2
HOUR
13.00
13.90
1>I. 00
15.90
16.00
17.30
18.00
19. CO
20.00
31.00
Z2.30
23.00
2*.00
25.30
26.00
27. OC
28.9?
29.00
30.00
31. Of
32.00
33.09
3*. 00
35.00
36.00
JUNCTION 1
HEAD (FEET)
-1.9537
-1.61*7
-.91*3
.0282
1.0257
1.8616
2. 3*21
?.3527
1.8973
1.10*5
.19**
-.5868
-1.3*29
-1.C870
-.76*1
-.2290
.3093
.6560
.6879
.3823
-.1856
-.«7C3
-t.*960
-1.8973
-1.9537
• T I H E HI
JUNCTION 2
HEAOIFEET1
-1.95*7
-1.7157
-.9123
.0983
l.P3*9
1.8390
2.31*8
2.33P*
1.88*5
1.1027
.1968
-.590*
-1.0*85
-1.0851
-.762*
-.2351
.31*0
.6655
.6917
.37ft 8
-.1930
-.8720
-l.*990
-1.9007
-1.95*3
10 YEAR, 2 HO
STORY OF :
JUNCTION 3
HEAD(FEET)
-1.95it5
-1.7067
-.9119
.0923
1.03*5
1.8*15
2.3175
2.33?3
1.8853
1.1023
.1961
-.59C*
-1.0*81
-1.0852
-.7622
-.23*2
.3138
.66*8
.6912
.3788
-.1699
-.8722
-l.*990
-1.9005
-1.95*1
10 YEAR, 2 HOUR DURATION STORM, TOTAL RAINFALLS.05 INCHES
JUNCTION *
HEAO(FEET)
-1.95*3
-1.69**
-.9115
.08*6
1.0339
1.8**7
2.3210
2.33*9
1.8865
1.1020
.1953
-.590*
-1.0*76
-1.0853
-.7621
-.2330
.3116
.6638
.6907
.3789
-.1898
-.872*
-l.*990
-1.9003
-1.95*0
JUNCTION 5
HEAOCFEETt •
-1.95*9
-1.7307
-.9125
.1065
1.035*
1.835*
2.3108
2.3275
1.8833'
1.1031
.1979
-.5903
-1.0*90
-1,0850
-.7625
-.236*
.31*2
.6666
.6923
.3786
-.1901
-.8716
-l.*989
-1.9009
-1.95**
JUNCTION 6
HEAD (FEET)
-1.9555
-1.7660
-.9128
.1277
1.0371
1.8263
2.3010
2.3203
1.8801
1.10*0
.2002
-.5903
-1.050*
-1.08*5
-.7626
-.2396
.31*7
.6693
.6937
.3782
-.190*
-.8710
-l.*989
-1.9015
-1.95*8
-------
RECEIVING HATER
-------
DESIGN STORM 10 YEAR, 2 HOUR DURATION STORM, TOTAL RAINFALL=2.05 INCHPS
RECEIVING MATER BLOCK WORKSHOP E.P.A.
CASE STUDY DYNAMIC STORM MATER QUALITY
MAXIMUM JUNCTION NUMBER 13
MAXIMUM CHANNEL NUMBER 19
NUMBER OF QUALITY CYCLES PER DAY 2*
NUMBER OF DAYS 5
NUMBER OF CONSTITUENTS 3
•» COLIFORMS MUST BE CONSTITUENT NUMBER 3
to
£5 LENGTH OF QUALITY INTEGRATION STEP ISECONDS! 3600.
PRINT INTERVAL, 1 DAYS
EXCHANGE REQUIREMENT AT OCEAN = RATIO OF ESTUARINE FLOW RETURNED TO AMOUNT THAT FLOMS OUT, PER TIDAL CYCLE .80
AVERAGE MATER TEMPERATURE (DEGREES CENTIGRADE.) lt.00
THERE ARE -0 STORMMATER INPUT JUNCTIONS
QUALITY CYCLE CONCENTRATIONS, PRINTOUT STARTS IN TIME CYCLE 1 ,
PRINTED EVERY 6 MOUR(S), FOR-A TOTAL OF 180 HOURS.
» • • * • SMITCH SETTINGS
SWITCH UUHBER 123*56789 10
SWITCH SETTING -0 -0-0 11 -0 -0 -b -0 -0
-------
CONSTITUENT NUMBER 1 BOD
SINK CONCENTRATION, MCL I OR NPN/L1 0.00
OXYGEN SATURATION IN6LI 9.00
REAERATION COEFFICIENT (I/DAT) .100
DECAY COEFFICIENT (1/OAYt .ZO
DISSOLVED OXYGEN FOR THIS CONSTITUENT IS CONSTITUENT *
INITIA1 CONCENTRATIONS.«6L (OR NPN/L) , BY JUNCTION
1 Z3<.5b78918
JUNCTION
1 TO H .1BOE»81 .108E»81 .100E»01 .10«E»01 .108E»(il .18i)E*81 .100E+01 .1IOE»01 .100E+01 .186E+01
11 TO 13 .10BE+I1 .180E»01 .100E»01
MASS LOADINGS. THOUSANDS OF IBS/DAY (OR OF MPN/NIN). BY JUNCTION
1Z3V5&78S
JUNCTION
1 TO 10 0. •. 0. P. 0. 8. 0. 0. 8.
11 TO 13 0. 0. .1»8E»8Z
INITIAL DISSOLVED OXYGEN CONCENTRATIONS (HGLI. BY JUNCTION
N> !Z3k56789
ro JUNCTION
*• 1 TO It .808E»81 .88&E*81 .SOOE»01 .»03E»01 .»OOE»01 .800E»81 .80CE»01 .t09E»ei .800E»«1 ,
11 TO 13 .88«E»01 .aOCC»01 .800E»01
DISSOLVED OXYGEN CONCENTPATION OF INFLON INGL1. JUNCTI3N
1Z3»5678S10
JUNCTION
1 TO 18 .88aC»81 ,8tOE»ll .800E»01 .88-E»01 .8DOE»01 .800E»01 .8BOE»01 .800EO1 .800E»01 .88dE»81
11 TO 13 .80aE*81 .8B8E»01 .880£»01
CONSTITUENT NUNBER 2 SUSPENOfD SOLIDS
SINK CONCENTRATION, NGL IOR HPN/LI 0.00
INITIAL CONSENT RATIONS,MSL (OR KPN/LI, BY JUNCTION
1 2 3 *<5 6 T 8 9 10
1 TO IB .l«CE*02 .10CE»62 .106£*D? .100E»3Z ,100E»tZ ,100E*92 ,108E»OZ .100EOZ .100E»0? ,100E*OZ
11 TO ii .iOuE»oz .iecE«ez .HOE»OZ
NASS LOIOINCS. THOUSANDS OF LBS/DAY (OR OF HPN/NINI, BY JUNCTION
1ZI»56T89
JUNCTION
1 TO 18 0. t. 0. 0. 0. 8. C. 0. 0.
11 TO 1» 0. B. .108E«v3
-------
TEST CASE, 3000 ACRE B»SIN
FOR STORAGE/TREATMENT BLOCK INPUT
DATA TRANSMITTED FROM INPUT FILE
ro
NJ
Ul
MUHBER OF STEPS
NUHBER OF INPUT POINTS
HUNBER OF CONSTITUENTS
TINE INCREMENT
INITIAL TINE
TOTAL AREA
LOADINGS FRON DATA
TINE IHR.I
12.17
12.13
LOADINGS FROH DATA
TINE (MR.I
12.50
12.67
12.83
13.11
19.17
11.33
13.51
13.67
13.83
14.80
14.17
14.33
14.58
14.67
14.63
15.81
15.17
15.33
15.59
15.67
FILE
JUNCTION
18
10
FILE
JUNCTION
IB
18
18
18
11
10
18
10
18
18
18
18
18
18
18
18
18
18
ID
10
ISO
1
3
600. SECS
12.08 MRS
3800.08 ACREA
BOO fLBS/NIN)
1.3.606
1.3.606
BOO (LBS/NINI
41.3*8
1.1.971.
•.5.011
T5.797
172.662
287.599
346.691
3*5.699
284.327
205.188
164.135
129.834
95.369
69.756
56.684
50.641
49.658
51.408
51.086
58.303
SS (L9S/HIN1
29.486
29.486
SS (LBS/MIMI
132.415
335.695
502.515
774.695
1487.076
Z 474 .5 78
3339.552
3863.462
3637.960
2857.935
2227.587
1608.771
967.019
475.211
2Z6.881
117.677
67.550
51.945
44.952
42.171
CLF IHPN/MINI
.131E»14
.131E«14
CLF IHPN/NINI
.449E»13
.12GE»13
.72SE»12
.158E+13
. 766E*13
.287EU4
.313E»14
.294E»14
.281E»14
.141E»14
.117E»14
.eroE*i3
,689£»13
.705E»13
.703E+13
. 685E+13
,636E»13
.524E»13
.".98E»13
.517E«13
-------
RECEIVING HATER BLOCK WORKSHOP
CASE
EPA STORMHATER MODEL
RECEIVING HATER QUALITY
DESIGN STORM lu YEAR, 2 HOUR DURATION STORM, TOTAL RAINFALL=2.05 INCHES
JUNCTION CONCENTRATIONS, HGL CO* HPN/L), DURING TIME CYCLE 10AY) 2, QUALITY CYCLE 6, TIME (SECI 6*800.
JUNCTI3N
1 TO 10
11 TO 13
CONSTITUENT NUMBER 1
123
.55
1.7%
.66
2.02
BOD
.62
1.9d
I
.55
.72
6
.81
7
.93
a
1.53
9
1.25
10
26.98
JUNCTION
1 TO IB
11 TO 13
CONSTITUENT NUMBER 2
6.35
19.70
8.20
21.72
SUSPENDED SOLIDS
d 5
7.61
19.99
6.52
9.07
6
10.19
7
11.72
8
17.11
9
15.2k
10
211.83
ro
N>
CT>
CONSTITUENT NUMBER 3 COLIFORMS
1 2 3 d
JUNCTION
1 TO 10 635.27 819.69 760.83 652.12
11 TO 13 12595.S3 %2DB8.59 1996.3d
JUNCTION
1 TO II
11 TO 13
56 7 8 9 10
997.17 1018.88 1171.86 10ZU7d.27 1680.5Z60d%500.d9
CONSTITUENT NUMBER d
12 3
BOO
8.33
7.8%
8.1%
7.87
8.20
7.96
LOADINGS FROM DATA FILE
TIME IHR.)
18.33
18.$0
18.67
IS.81
19.01
19.17
19.33
19.50
19.67
19.83
20.10
JUNCTION 600 (LBS/HINI
10
10
10
10
10
10
10
10
19
10
ia
52.9%5
52.005
52.6%9
52.dl2
52.%37
60.929
75.535
79.602
77.705
78.037
78.115
(00)
d
8.31
5
8.06
6
7.98
SS (LBS/MINI
.1.0.133
38.982
39.762
39.d90
39.512
d5.207
55.06d
57.870
56.562
56.786
56.8dl
CLF INPN/MIN)
.ld9E»ld
,152E«ld
.ld9E»ld
.156E»ld
.150E»ld
.136E*ld
.112E»ld
.105E«ld
.108E»ld
.108E+ld
7
7.91
8
7.83
9
T.85
10
d.86
-------
DESIGN ST3RM 10 TEAR, ? HOUP DURATION STORM, TOTAL RAINFALLS. 05 INCHES
RECEIVING MATER SLOCK HOUR SHOP E.P.A.
CASE STUDY OYNAHTC STORN MATE* QUALITY
AVERAGE JUNCTION CONCENTRATIONS DURING TIDAL OR TINE CYCLF Z , CONSTITUENT NUMBER 1 BOO
1 23h 56 789 10
JUNCTION
1 TO 10 .770E*88 .l99C»tl .1Z8E»81 .758E»08 .9I.IJE*00 .*ZOE»«1 .116E*01 . 9*8E»01 .15CE»01
11 TO 13 .179E»«1 .191C»I1 .194E»81
NAXINUNS
JUNCTION
1 TO ID .1T«E»«1 .578E»01 .Z96E+01 .111E»81 .120E*01 .970E»01 .lt3E*tl . 151E»02 .167E»B1
11 TO 13 .1«6E»01 .ZaSC«ll .19SE»I1
NININUNS
JUNCTION
1 TO 18 .52«e»tO .S*2E»«C .622E»BO .$49E»tO .723E»0« .»07e»00 .933f»00 .106E»D1 .1Z5E»01 . 160E+01
11 TO 13 .169EH1 .1
ro
M DESIGN STORH 10 YEAR, i HOUR DURATION STORM, TOTAL RAINFALL =2. 85 INCHES
RECEIVING MATER BLOCK NORKSHOP E.P.A. ,
CASE STUDY DYNAMIC STORN MATER QUALITY
AVERAGE JUNCTION CONCENTRATIONS DURING TIOAL OR TIME CYCLE Z , CONSTITUENT NUMBER <» BOD (001
123I»567«910
JUNCTION •
1 TO ID .817E*I1 .771E»I1 ,7S1E»01 .808E»01 .79ZE»B1 .715E»01 ,78fcE»01 .61ZE»01 .781Et01 .50LE»01
11 TO 13 .7««E*«1 .789E+01 .796E»01
NAXIHUHS
JUNCTION
1 TO 10 ,837E»tl .814E«I1 .8Z3E»01 .831E»31 .836E*Q1 .79«E»01 .791E*01 .T88E»01 .786E»01 .77GE*01
11 TO 13 ,787E»11 .791E»01 .796E»01
HININUMS
JUNCTION
1 TO IB ,771E»81 .650E*81 .731E»01 .786EO1 .78t£*01 .5I.7E+01 .776E»01 .«.5ZE*01 .777E*01 . 38JE*01
11 TO 13 ,782E»fll .787E»01 .795E»P1
LOADINGS FRON DATA FILE
TIME IHR.I JUNCTION BOO ILBS/NIN! SS (L8S/NINI CLP (HPN/NINI
-------
DESIGN STORM
10 YEAR, Z HOUR DURATION STORM, TOTAL RAINFALL=Z.05 INCHES
RECEIVING HATER BLOCK WORKSHOP
CASE STUDY
E.P.A.
DYNAMIC STORM HATER QUALITY
AVERAGE JUNCTION CONCENTRATIONS DURING TIDAL OR TIME CYCLE 2 , CONSTITUENT NUMBER 2
1231.5678
JUNCTION
1 TO 10
11 TO 13
JUNCTION
1 TO 10
11 TO 13
JUNCTION
1 TO 10
11 TO 13
SUSPENDED SOLIDS
9 10
.88<»E*01 .181E»OZ .133E+OZ .953E+01 .1ZOE*OZ .316E+OZ .lt6E+OZ .560E+OZ .179E»OZ .114E+03
.Z01E»«2 .Zfi6E»OZ .ZOOE+OZ
.173E+02
.6BOE»01
.190E»B2 .197E»12
MAXINUMS
.158E*02
.ZOOE»02
MINIHUHS
.761E»01 .65ZE«01 .907E«-01
.182E«-OZ .8%OE»D2 .199E*02 .27CE»03
. 117E«-02 .130E»02 . 151E«-02 .
to
00
DESIGN STOR1
RECEIVING HATER BLOCK HORKSHOP
CASE ST'JDY
10 YEAR, 2 HOUR DURATION STORM, TOTAL RAINFALLS.05 INCHES
E.P.A.
DYNAMIC STORM HATER QUALITY
AVERAGE JUNCTION CONCENTRATIONS DURING TIDAL OR TIME CYCLE
JUNCTION
1 TO 10
11 TO 13
JUNCTION
1 TO 10
11 TO 13
.559C»05
.13%E»05 .164E»05
3 "»
.183E»06 .168E»05
MAXINUtIS
.252E»05 .
.839E»06 .775E»B5
.200E»0<»
CONSTITUENT NUMBER 3
6 7
.119E*07
,390E+Of . 1«.JE»05
COLIFORHS
• 9 10
3%OE»07 ,105E»05 .826E»07
.175E»05 .12ZE*08
NININUMS
JUNCTION
1 TO 10 .600E»03 .820E*03 .761E»03 .65JE+03 .9u7E»03 .102E«-Ol> .117C»Oo .130E»0<> .151E»0% .179E»0<>
11 TO 13 .189€»«l» .197E»0
-------
to
ro
vo
oc
e
c
o
C
O
u
c
0)
60
O
T3
OJ
O
cn
tn
10
I I I
I I
Junction 7
8 _
Junction 1
4 -
2 _
0
Junction 10
I F
12 18
I
54
24 30 36 42 48
Time from Start of Storm, hrs .
I
60
I
66
72 78 84 90 96
FIGURE 25 . DISSOLVED OXYGEN VARIATION WITH TIME FOR SELECTED f
-------
REFERENCES
1. Map Reading, Field Manual 21-26, U.S. Department of the Army, Washington,
D.C (1956).
2. Jewell, Thomas K., Application and Testing of the U.S. Environmental
Protection Agency Storm Water Management Model to Greenfield,
Massachusetts, unpublished Manuscript, Department of Civil Engineering,
University of Massachusetts, Amherst (1974).
3. Roesner, Larry A., "Basic Data Requirements and Transferability of Data",
from Management of Urban Storm Runoff - Quantity and Quality, Proceedings,
Hydrologic Engineering Training Course, Water Resources Engineers, Inc.,
Walnut Creek, California, Table 1, pg.. 6-3 (1973).
4. Vittands, Jekabs P., "Three Case Studies on the Application of the Storm
Water Management Model", Proceedings, Short Course held at University of
Massachusetts, Amherst, August 19-23, pp. 6-10 (1974).
5. Chow, Ven Te, Open-Channel Hydraulics, McGraw-Hill Book Company,
New York (1959).
6. Huber, Wavne C., et al., Storm Water Management Model, User's Manual.
Version II, Environmental Protection Technology Series, U.S. Environmental
Protection Agency, Cincinnati, Ohio (1975).
7. King, Horace W. and Brates, Ernest F., Handbook of Hydraulics, McGraw-
Hill Book Company, New York (1963).
8. Lager, John A., and Smith, William G., Urban Stormwater Management and
Technology; An Assessment, Environmental Protection Technology Series,
U.S. Environmental Protection Agency, Cincinnati, Ohio (1974).
9. Metcalf & Eddy, Inc., Wastewater Engineering, McGraw-Hill Book Company,
New York (1972)
10. Sullivan, Richard H., The Swirl Concentrator as a Combined Sewer Overflow
Regulator Facility, Environmental Protection Technology Series, Office of
Research and Monitoring, U.S. Environmental Protection Agency,
Washington, D. C. (1972).
11. Metcalf & Eddy, Inc., et al., Storm Water Management Model, Volume 1, Final
Report, Water Pollution Control Research Series, U.S. Environmental
Protection Agency, Washington, D, C. (1971).
230
-------
12. Standard Methods for the Examination of Water and Wastewater, 13th Ed.,
American Public Health Association, Washington, D. C. (1971).
13. O'Connor, D. J., and Dobbins, W. E., "Mechanisms of Reaeration in
Natural Streams," Transactions, ASCE, 123:641 (1956).
Acknowledgements
The authors wish to express their sincere appreciation to
Mr. David Gaboury for assisting in preparing the Study Guide,
Mrs. Pamela Jewell for proofreading the manuscript and
Miss Dorothy Blasko for typing the manuscript.
231
-------
Part I. CRITERIA FOR SELECTION OF
STORMWATER MANAGEMENT MODELS - PLANNING VS. DESIGN
by John A. Lager, P.E.*
Selection of stormwater management models for specific
tasks should not be left to chance. Today there are both
choices and opportunities for those who will look beyond
the problems and focus upon the results.
• Planning without commitment is next to valueless.
• Commitment without objectives is simply foolish.
• Excellence without communication is wasted.
These situations, unfortunately, are not rare in the past
and present application of stormwater management technology.
In this lecture we will build up a selection process which,
I believe, will increase the visibility of the action
decisions necessary to successful performance. The process
is intended to be helpful to the model purveyor-developer
as well as the responsible manager-solicitor. The evalua-
tion procedure provides a vehicle for both project initia-
tion and project updating and reassessment.
The presentation, Part I, starts with a discussion on
defining the study objectives. With rare exception this is
the most difficult task of all. Next, the given data base
*Vlce President, Metcalf & Eddy, Inc., Western Regional Office, Palo Alto,
California.
232
-------
will be identified. Then the array of model capabilities
and requirements will be presented, leading to an identifi-
cation of necessary assumptions for application. Finally,
the selection process itself is concluded with emphasis
on results and fit into the continuing program.
In Part II, a simplified modeling approach will be
described with emphasis on the application strategy and
potentials for management decision making.
DEFINING THE STUDY. OBJECTIVES
Study objectives should be specific. Why? To facilitate
communication, focus of effort, and commitment. For
example, if the study objective is simply to
comply with the requirements of PL 92-500
it is unlikely that any two people in the chain of author-
ity will have the same idea in mind except by default. On
the other hand, if the objective were amplified to
assess the relative impact of the non-point discharges
of urban runoff from within the city limits of
Boomtown to other controllable sources on Segment V of
the High Flow River with respect to organic loadings,
given that loadings from Segment VI to V will not
change from today's level. It is intended th'at load-
ings be assessed for two seasonal occurrences...
the chance of common understanding would be greatly
improved, and the specifics of model requirements and
applicability would be well underway to identification.
The following, then, are the types of information that
require definition in the model selection phase.
233
-------
Improve What, How Much, and Where
Suppose a city is investigating the use of polymers to
reduce pipeline friction and thereby increase interception
of flows for treatment at the dry-weather plant, as opposed
to constructing a satellite overflow treatment facility.
Models operating off the same input data (hydrology, catch-
ment characteristics, pipe networks, etc.) would offer an
_..jell~.it tec"1 for "-" 0Mri£ the tvo alternatives. SJ'Tnp1:;-
fied models could be used to estimate the expected number
of occurrences annually before and after improvements, and
detailed dynamic models could identify design criteria for
specific critical events.
Other representative questions that should be considered in
setting the study objectives include the following:
1. Is flooding the major consideration? If so, then
the use of a dynamic backwater model is essential.
Why? Because flooding is the result of a system's
inability to react.to a specific storm event in
the required time interval (i.e., bottlenecks).
2. Is loading on the receiving stream the control?
If so, is the nature of the impact short term
(indicating the need for dynamic modeling) or long
term (simple models--i.e., Streeter-Phelps--
may suffice).
3. Is there a repetitive, defined occurrence that is
to be abated? For example, are overflow occur-
rences to be reduced to keep bathing beaches open
for the maximum number of days in season? Note
234
-------
that in bacterial contamination the occurrence or
nonoccurrence of an overflow is more important
than its duration.
4. What are the keys to the system's performance
today? A major study objective may be to deter
mine what makes the system respond the way it does.
That is, what and where are the controls and how
do they react?
5. Are there available facilities that are not being
put to maximum use or to effective use? These
may dictate some trial criteria for abatement
schemes.
6. Are the funds available on the basis of the
ability to sell the program? This again is a
question of communications and timing.
7. What is the balance between the known (given) data
and the assumptions? How will this change over
time? An exquisite model running on a series of
assumption's may yield no better results than a
simple model run on the same base data.
8. What are the potential impacts of new technology?
Recent work for the National Commission on Water
Quality [1] has shown the optimal storage-treat-
ment balance to be quite cost sensitive within the
relative range of costs experienced in EPA's R$D
program (Figure 1) . Thus, technological changes
which decrease the cost of one with respect to the
other may have significant impact on program
selection.
235
-------
tt
t-
I
J5-,
10-
I 5-
8-
TREATMENT COST = $50,000/MGD
TREATMENT COST = $40.000/MGD
TREATMENT COST = $30.000/MGD
100 200
TREATMENT RATE. MGD
300
AJSUMPTIONS
1.000 ACRES - REGION I
STORAGE : $1.0/CAL.
TREATMENT: i. JSO.OOO/MGO
b. $40.000/MGD
C. J30.000/MGD
OBJECTIVE: TO TREAT 2 YEAR - i HOUR STORM WITHOUT OVERFLOW
UTILIZING ANY COMBINATION OF STORAGE AND TREATMENT.
'FIGURE 1. COST SENSITIVITY
Required Sensitivity of Results
If the ranges of given variables are known, then the model
selected must be sufficiently sensitive to display output
differences within such ranges. Also, the modeling
requirements might be quite different if the model's
purpose is to locate points on the system where monitoring
information would be most beneficial, as opposed to addi-
tionally projecting expected flow and concentration ranges
at such locations.
236
-------
Who is the User or Reviewer
In order to be implemented, the results or operations of
the model must be made known to, and approved by, some
person or persons. What are his requirements and needs?
Arriving at the correct solution is not enough if you lose
the client in the process. The output must be recogniz-
able and usable in terms that the client can easily relate
to and work with, for after all, he_ has to sell it.
Boundaries of Investigation
What are the boundaries of the study area(s)? Who is
responsible for defining boundary transfers? Which sources
will be used? Which time spans are to be investigated--
historical, present, future? And which time increments--
annual, daily, hourly, etc.?
Timing and Budget Constraints
What are the milestone dates of the project, and what
information must be deliverable by these dates? Are the
results to take precedent over the effort? Where are the
risk areas—methods of approach, contingency plans? Will
the budget resources support the study objectives?
Personnel and Hardware Constraints
Who is going to operate the program and on which machine?
Are they compatible with the model requirements? Is the
program to be used or further developed following the
initial project? By whom and where? Are direct interfaces
with other current modeling/data management operations
essential or desirable? Are "canned" subroutines or
machine-specific notations used in the programming? Is the
documentation current, complete,and available? Is the
model available?
237
-------
IDENTIFY GIVEN DATA BASE
Models require input data upon which to operate, whether
the data are real or fabricated. Also, the quantity and
quality of the data base is likely to change over the
course of the work. To select model programs effectively
requires that the available data base strengths, weaknesses,
and dynamics be identified beforehand.
Rather than selecting a model and then seeing if you can
fill its data requirements, it is preferable to analyze
your available data and then choose the model that can use
these data most effectively to achieve the study objectives.
Statics
As used in this lecture, static data are the fixed data
bases available in the initial stages of the project. Such
data might include the plans and specifications of the
system, historical rainfall records, census tract enumer-
ations, planning and zoning documents, aerial photographs,
treatment plant records, files on trouble calls and system
maintenance, prior studies and reports, monitoring records,
and state-of-the-art literature.
Generally, the most prized data represent cause-effect
relationships. Examples might include changes in plant
inflow characteristics between dry- and wet-weather
periods; measured rainfall-runoff-quality relationships;
intensive stream surveys correlated with development,
treatment performance, and ideally,.storm occurences; and
documentation as to the time, location, and duration
of overflows.
238
-------
Personally, I believe that the most effective nationwide
pollution abatement step that can be taken by regulatory
authorities today is to require the monitoring and report-
ing of each and every overflow occurrence. The reason is
that the real water quality problem solution does not lie
in the culmination pf a grand cure-all program, but rather
in the early implementation of steps moving toward that
goal.
However, very important data describe the system itself.
Are as-built drawings up-to-date? Is the data available
in such a form that it can be taken off readily? Are
drainage and census boundaries shown? Are the topography
and structures superimposed on the maps? How does the
system function? Controls?
With respect to rain gages, what is the quality of the
network? Where are the records kept? In what form? What
periods are covered? Are they discontinued in the winter?
Have any data been computerized?
Dynamics
These are the data that will become available during and
subsequent to the study. To what extent can this effort
be modified or redirected to suit the needs of a specific
model? At what milestone points can "midpoint" corrections
be made both in the modeling and data gathering programs?
As a note of caution in setting a value on the dynamic
data: one storm does not represent a season, and one
season does not represent a historical record; however,
the ability or inability of a quantity of flow to get from
point A to point B may be both significant and repetitive.
239
-------
A major advantage of dynamic modeling is that it may
identify the key nodes that control or directly reflect
system efficiency. Concentrating dynamic data gathering at
these nodes may be far more cost effective than, say, a
shotgun or intuitive approach.
As a final thought, the data requirements for a specific
dynamic model may be so exhaustive as to require several
months to collect, check, and apply. Can a simplified
model be used to fill this time gap and improve the pros-
pects of success of the more comprehensive program?
IDENTIFY THE ARRAY OF MODEL CAPABILITIES
Mathematical model usage in stormwater management is basic-
ally a creation of the last decade, and application experi-
ence is even more limited (SWMM was first marketed in 1971).
Significantly, the more sophisticated models are in a
continuous state of refinement and modification as new uses
are attempted or requirements are identified. Documenta-
tion, availability, consultation advice, and record of use
are equally variable and changing.
Why Model at All
Models follow operating rules without exception or inter-
pretation and can process mountains of data with relative
ease. The developer sets the rules (and in most cases
options), and the user furnishes the data and is left with
the results. That is the package, plain and simple.
In deciding whether or not to model, the potential user
should assess this process closely. What are the data he
is prepared to give or invest in? What are the operating
rules that he is willing to live by? Where are the mountain*
240
-------
he wishes to overcome? What will he do with the results?
The faults are as obvious as the potential benefits--weak
or incorrect rule(s), erroneous or inadequate data, poorly
selected or defined mountains, unusable or indeterminant
results (i.e., what is the message?).
Where to Find Information
A comprehensive evaluation of stormwater management models
is contained in the study, "Evaluation of Mathematical
Models for Engineering Assessment, Control, Planning, and
Design of Storm and Combined Sewerage Systems," prepared
by Battelle Memorial Institute for EPA [1]. The objectives
and initial findings of this study were presented by the
project director at last year's seminar [2]. The complete
report covers model reviews, model selection criteria,
detailed descriptions of selected models, test data and
results, cost comparisons, and a source listing.
Selected summaries from this and other recent comparative
evaluations [3, 4] are shown in Tables 1, 2, and 3 along
with the identified references. The models are separable
by function (planning, design, operation) and by degree
of precision (simple, intermediate, and complex). The
purpose in presenting the three similar listings is to
illustrate the perspectives of these experienced users.
What (Engineering-Hydraulic) Processes are Being Simulated
The user should screen candidate model documentation to
identify what processes or devices are being simulated and
linked and with what degree of sophistication. Does the
model focus suit the need focus? For example, the Hydrocomp
Model is a multigeneration descendent of the Stanford Water-
shed Model. The original model dealt largely with
241
-------
too* ITDIUUILICS
nSTNkTEl OUAI4TT
nscEUJunotn
tl
i§
N>
UUffEMZfT Of
Table 1, MODEL CHARACTERISTICS BY BRANDSTETTER [2]
-------
COMPARISON OF URBAN RUNOFF MODELS, MARCH 1974
Degree of Degree of Accu- Flex- Re- Degree of
Sophist!- Sophisti. ale ibility of Explicit ceir- Cali
Sui- Quil- cation of cation of Model- Modeling Modeling Treat- ing brttion/
10
->
01
Model
Rational
Method
Chicago
Unit
Hydro-
, |
grapb
Unit
Pube
STORM
RRL
MIT
BalteDe
EPA-SWMM
WRE-SWMM
Qacin-
nati (UCUR)
Dotich
(HVM)
SOGREAH
Hydrocomp
mnofe
OSS)
face ity Surface Sewer
Rout- Sewer Rout- now Flow
ing Routinr, ing Routing Routing
Peak Peak No
Flows Flows
Only *'*-•--
Yes
Yet
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
vnuy
No No
In con- No
biut]on
with
surface
In com- No
bination
with
surface
In com- Yes
bination
with
surface
Yes No
No No
Yei Yes
Yes Yes
Yes Yes
Yes Yes
Y«s No
Ye*. ?
Yes Yes.
Yes No
Low Low
Moderate NA
Low Low
Low- Low-
Moderate Moderate
Low Low
Moderate Low-
Moderate
High NA
Low Moderate
High Moderate
High High
High Low
High High
High High
Moderate Moderate
Moderate High
ing of of Sewer
Stir- Compo-
charging nents
No
NA
No
No
No
No
NA
No
No
Yes1
No
Yes
Yes
No
No
Low
NA
Low
Low
Low
Low
NA
Moderate
High
High
Low
High
High
Low
Low
of In- men! Model Verifi-
Syitem Model- Avail- cation
Storage ing able Required
No
NA
No
No
No
No
NA
Yes
Yes
No
No
1
1
No
No
NA
NA
NA
NA
Yes
NA
NA
No
Yes
Yes
No
NA
T
No
NA
No
No
No
No
No
No
No
No
Yei
Yes
No
Yei
Yes
Yes
No
Docu- Data
Simulation Avail- men- Require-
Period ability tation ments
Usually not Individual Non- Good
verified stormi proprietary
Moderate
High
High
Low
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
High
Moderate
Individual Non- Fair
atorms proprietary
Individual Non- Fair
storms proprietary
Individual Non- Fair
storms proprietary
Long term Non- Good
proprietary
Individual Non- Good
storms proprietary
Individual Proprietary Fair
storms
Individual Non- Poor
storms proprietary
Individual Non- Good
storms proprietary
Individual Proprietary Poor
stormi
Individual Non- Fair
Individual proprietary
storms Proprietary Poor
Individual
storms or
?*?.""]' . Proprietary Poor
Individual
Individual Proprietary Fair
storms or
long term
Individual Non- Good
storms proprietary
Low-
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Extensive
Extensive
Extensive
Extensive
Extensive
Extensive
Extensive
TABLE 2. MODEL CHARACTERISTICS BY HUBER [3].
-------
UTILITY OF URBAN RUNOFF MODELS FOR VARIOUS STAGES OF
STORMWATER MANAGEMENT PLANNING
STORM
MITCAT HYDROCOMP EPA SWMM SF-WRE
Large-Scale Planning
(Alternative Screening)
Intermediate Scale
Planning
Excel lent
Poor
Detailed Planning/Analysis pnnr
(No Significant Backwater)
Detailed Planning/Analysis poor
(Complex Drainage Networks)
Good
Poor
Good
Poor
Poor
Excellent Excellent Excellent* Excellent*
Excellent Fair
Poor
Excellent Excellent
Fair
Excel lent
Runoff Quality Simulation
Flow Computation
Yes
Rational
Formula.
No
Yes
Yes
Yes
Kinematic Kinematic Kinematic Kinematic
Wave Wave Wave + Wave +
Manning Complete
Equ.with Dynamic Flow
Dynamic Equations
Continuity
*The Runoff Model is best suited for these applications
TABLE 3. MODEL CHARACTERISTICS BY ROESNER [4].
-------
unimproved watersheds to determine yields, streamflows, and
flood characteristics. Thus, its strengths may be presumed
to lie in representation of the natural hydrologic cycle.
On the other hand, the WRE-SWMM transport model modifica-
tions were developed specifically for complex urban pipe
networks where flow' reversals, backwaters, and special
diversions are not uncommon. The differentiation is obvious.
Getting from rainfall to runoff quantities may be as simple
a process as applying a direct percentage (perhaps.not a
bad assumption when the percent imperviousness is high), or
it may be a sophisticated accounting of depression storage,
infiltration rates, and surface and channel flows. Or it
may be an even more sophisticated accounting of soil, types
and moisture content, vegetation, ground slopes, air tempera-
ture, wind direction, etc.
In developing the treatment process simulation for SWMM, a
major consideration was not only to develop the operating
rules for efficiencies when operating within prescribed
limits, but more importantly, to account for the bypassed
flows or process inoperation when the unit was subjected to
flows outside the prescribed limits. Such unnormalized
conditions should not be overlooked in assessing capabilities,
Examples
Metcalf § Eddy has applied stormwater management modeling
techniques to suit many varied client requirements. In
the following examples, selection between planning and
design are emphasized.
Case 1 - Planning - In 1970 the City of Chicago wanted a
modeling capability to support its studies in developing a
245
-------
master plan to abate pollution from combined sewer over-
flows and flooding. A primary consideration was that under
the largest storms of record there must be no backflow into
Lake Michigan. Thus, the requirement was event specific.
The great magnitude of the problem--650 overflows to 75
miles of waterways--made it economically imperative that an
in-house capability of performing the modeling runs be
developed. Finally, abatement schemes were quality sensi-
tive. SWMM was selected and successfully applied. Sur-
charging was not a handicap.
Case 2 - Planning - In 1972 the District of Columbia was
committed to a massive upgrading of treatment at its dry-
weather facility to tertiary levels never approached in
such a scale. Concern was raised that the improvements at
the Blue Plains facility would be negated by raw discharges
through combined sewer overflows every time it rained. An
answer was needed in 90 days as to where the cost-benefits
divided between plant upgrading and wet-weather control
from the basis of in-stream quality. A combination of
simplified models (to determine the approximate frequency,
location, and magnitudes of loadings) and dynamic models
(SWMM - Receiving Water Block) was used.
Case 3 - Design - In 1972-1973 the City of Saginaw was
proceeding into initial implementation of its combined
sewer overflow abatement program under a State of Michigan
mandate to
Complete construction of facilities necessary to
abate pollution caused by said overflows and place
same in continuous operation on or before June 1,
1977.
246
-------
The 1972-1973 task is summarized in the following sentence
taken directly from the report: "The basic plan, being
well established and approved, leaves remaining the problem
of determining the final location, size, and type of facili-
ties." The Runoff, Transport, and Storage Blocks of SWMM
were used to test several alternatives incorporating various
sizes of storage and treatment facilities for the design
storm conditions. A quasi-steady state model was used to
relate impacts on the river system. The solution accounted
for both in-system and off-line storage as well as dual
treatment functions. The first facility is under
construction.
Case 4 - Design - The City of Hartford was interested in
developing a modeling capabilityVithin its staff and to
improve upon the Rational method for design of necessary
additions to the urban drainage system. City staff members
were trained in the application of the Runoff and Transport
Blocks of SWMM, and then they proceeded to check completed,
off-the-shelf designs of major drains scheduled for near-
term construction. By accounting for the dynamics of flow
routing, it was frequently found that one, two, or more pipe
diameter increments could be cut from the traditionally
designed systems while still handling the design storm.
IDENTIFY NECESSARY ASSUMPTIONS
Each model has a specific set of data requirements. Com-
paring these with the previously identified given data base
will identify the proportions of the assumed data base or
the dependency of the results on default values. If a large
array of default values is to be used in every case, can
they be sidestepped altogether by going to a simpler model?
That is, what is the chance of your answer being locked in
before you start?
247
-------
Statics
As with the given data, some assumptions will necessarily
remain fixed over the course of the project, while others
may change significantly through contributions from other
programs, sensitivity analyses, or verification runs.
Setting down and differentiating between the two may iden-
tify key areas for supplemental R§D work or literature
searches.
Typical static assumptions may include friction factors,
infiltration ratios, retention storage, daily flow cycles,
pre-storm conditions, diffusion coefficients, decay and
replenishment rates, boundary exchanges, accounting for
nonmodeling discharges, etc.
Dynamics
Dynamic assumptions may include the computational time
steps, the "design" event (may be approached through trial
and error executions), treatment-cost efficiencies, im-
perviousness and directly contributing impervious areas,
as well as any of static assumptions where special effort
is to be taken. The point is to make a realistic, pre-
application appraisal of what types of sensitivity analyses
are likely to be performed and whether or not they are
likely to be converging.
SELECT FOR RESULTS
The success or failure of a modeling attempt can be judged
only by the buyer. Did he get the information he needed
when he needed it? For the price he expected to pay for it?
If an initial attempt was a failure, where did it go wrong--
objectives, data base, model capabilities, assumptions? If
it was successful, where doe.s the credit lie?
248
-------
Decision Output Data
As introduced earlier, the display, listing, or packaging
of the output data may be as important to the decision
maker as the data itself. Are the output values to be
accepted because they display eight significant figures or
because the cause and effect relationships are visible,
reasonable, and capable of verification? If the output is
voluminous, where are the checks and balances? How is it
displayed? If part of the computation was made outside the
game rules, how is it signaled? Corrected? If a trace
back into the simulation is required, can one obtain inter-
mediate results at a specific location in the system? At a
specific point in the computations? How about restarts to
avoid repetitious computations?
A point, perhaps understressed heretofore, is the value of
accuracy in the results. A dynamic model may give the
impression of a very true result for a clean, circular
conduit where, in fact, 50 percent of the line is in a state
of collapse, subsided, or filled with debris. How valuable
is representing the response of a new system when you are
stuck with the one you have? Further, if the transfer of
quality data from Transport or Storage to Receive is to be
accomplished in hourly blocks, is it necessary that it be
computed at 5-minute intervals? The answer, of course,
goes back to the study objectives as to what is to be im-
proved, how much, and where.
Building Block in Continuing Program
One of the advantages of modeling is that almost as much
value might be obtained from a technically failed run as
from a successful one. Why? Because the effort of system
analysis that goes into preparing for and executing a run
249
-------
forces the issue into areas that have traditionally fallen
into neglect as lost items within a city budget. In
selecting a model, the user should assess its impact on his
broader continuing program. If the present objectives are
met, what is the next step? Will models again be a require-
ment? Should the experience and data base be built up now
in preparation?
Does It Fit
In summing up all of the above selection criteria, it is
seen that there are both choices and opportunities in
modeling urban stormwater runoff. Finding the model of best
fit is best approached by a systematic appraisal, such as
the one outlined herein, with emphasis on the preselection
definition of study objectives.
Planning Vs. Design
In the author's experience there are basically two levels
of approach available: the simplified and the detailed.
The advantage is that, when understood, they can and should
be entirely complementary. The advantage of the simplified
models is their ability to process long periods of record
and broad areal coverage at low cost. The advantage of the
detailed models is their ability to make a comprehensive
analysis of singular events and systems with a corresponding
increase in accuracy when supported by a viable data base.
In planning studies the simplified models offer a flexible
screening device to identify consequential storm events and
potentially attractive alternatives. The detailed models
permit the necessary follow-up and technical evaluation
between competing plans.
250
-------
In design, the roles may be somewhat reversed with the
detailed models fixing the component dimensions and the
simplified models testing the decision against the histori-
cal record. These capabilities and interfaces will be
described further in Part II: A SIMPLIFIED STORMWATER
MANAGEMENT MODEL FOR PLANNING.
251
-------
PART II. A SIMPLIFIED STORMWATER MANAGEMENT
MODEL FOR PLANNING
By
John A. Lager, P.E.*
What are simplified stormwater management models? Why are
they needed? How are they applied? What is their value?
These questions will be addressed in this lecture through
the context of a modeling strategy previously applied and
now undergoing critical refinement and documentation.
In the author's opinion, the simplest model that will get
the job done is usually the best. The intent here is not to
sell a particular model but rather to set forth the approach
methodology. If a simplified model has to be disseminated
in a preassembled package, then it has already lost a signi-
ficant fraction of its utility. The user should have the
flexibility to assemble model components and detail to build
on his data strengths and towards his study objectives.
PURPOSE AND AVAILABILITY
As brought out earlier, the target of the simplified models
is to break down data into a form meaningful to the user.
In so doing, precision is sacrificed for breadth of cover-
age. Further, because of the low cost of simplified models,
both for setup and execution, multiple assumptions can be
tested with relative ease and over a short period of time.
»T
Vice-President, Metcalf § Eddy, Inc., Western Regional Office,
Palo Alto, California.
252
-------
Functional Location in Master Program
The key sequential steps in a typical urban stormwater
management master program are shown in Figure 2. While
simplified models are principally an early planning screening
device, they can be used effectively throughout a program's
development, design, and implementation, including--
optimistically — real time control.
u
D
D
D
Q
FORMULATE APPROACH
• OBJECTIVES
• PROBLEM IDENTIFICATION
• SYSTEM EVALUATION
• DATA BASE
SELECT MANAGEMENT ALTERNATIVE
• ESTABLISH CONTROL
• HOD I Tt) RIHC
• STHUC
• AtlTONATION
• TREAT TO OBJECTIVE LEVELS
• S008CE
• IYtTU
• orr-iiNE
9 DUAL USE
IMPLEMENT AND ASSESS PERFORMANCE
• PILOT/PROTOTYPE
• OPERATION AMD MAINTENANCE
• MONITORING AND FEEDBACK
OF INFORMATION
FIGURE 2. URBAN STORMWATER MANAGEMENT STRATEGY [5]
253
-------
Subsystems
The subsystems to be described in this presentation are
(1) rainfall characterization, (2) storage-treatment balance,
and (3) discharge-receiving water response.
Rainfall Characterization - Rainfall characterization is
basically a sorting and ranking of historical rainfall data
from local gages. It requires a simple but precise storm
event definition to permit disassociation of chronological
sequence for statistical analysis. Official Weather Bureau
recordings of hourly data are a practical data-source.
Storage-Treatment Balance - Storage-treatment balance
models take historic rainfall data in chronological sequence
as input and log quantities, times, and durations of over-
flows for selected storage-treatment combinations.
Discharge-Receiving Water Response - Discharge-receiving
water response models apply significant quality character-
istics to the discharges and predict their behavior in the
receiving stream. The latter model can be very complex,
particularly in an estuarine environment.
Primary Goals
Simplified math models or their equivalent analyses are
needed for three reasons:
1. To introduce time and probability to
stormwater analyses
2. To promote total system consciousness on
the part of the user or reviewer
3. To establish size-effectiveness relationships
254
-------
Just as time and probability analyses are important in
sizing water supply impoundments and safe yields, they are--
or should be--equally important in determining the effective
use of stormwater facilities. Since total capture is not a
necessary goal, as opposed to flood control works, there is
greater latitude in facility sizing and staged implementation,
The trick is determining the relative merits of alternatives,
a task for which modeling is ideally suited.
Alternatives should not be limited to stormwater abatement
measures alone and may readily encompass dry-weather plant
upgrading and separate as well as combined discharges.
Source
Documentation and description of the modeling techniques
described herein is being prepared by Metcalf § Eddy under
a demonstration project sponsored by the Environmental
Protection Agency through a grant to the Rochester Pure
Waters District [6]. It is noteworthy that the work is
being performed concurrently and as an integrated part of
a much larger program involving extensive overflow moni-
toring and characterization, dynamic simulations, pilot
plant construction and operation, and outline of a total
program for collection, transmission, and treatment of
combined sewage under a central control and management
system. The simplified modeling report is expected to be
published under EPA's Environmental Protection Technology
Series in June 1976.
A similar, directly available model, STORM [7], will be
described in Dr. Roesner's lecture this afternoon. For
comparison, the present program has approximately 300
statements versus 4,000 statements in STORM and 14,000
in SWMM.
255
-------
APPLICATION STRATEGY
The successful application strategy of simplified models is
one of forced recognition of, and adjustment to, system
realities. In this section the component subsystems are
described, and in the following section they are assembled
into a working program.
Data Base
The first activities must be directed at answering the
question: What do we have and how does it work? The process
should inlcude not only the identification and compilation
of data, but also the development of a functional repre-
sentation of the system. A first attempt of such a drawing
for the Rochester system is shown in Figure 3. An important
requirement is that the points of dynamic data generation be
shown as well as the relationships between what may prove
to be test elements, including maximum capacities when
available.
Rainfall Characterization
Data for rainfall characterization are most readily obtained
from the U.S. Weather Bureau records either on tape files or
through published daily and hourly summaries. Tapes are
issued through:
U.S. Department of Commerce
National Climatic Center
NOAA Environmental Data Service
Federal Building
Ashville, N.C. 28801
Tel. (704) 258-2850
Data are available on two record files: Deck 488-USWB HOURLY
PRECIPITATION and Deck 345-WBAN SUMMARY OF DAY. Most first
order stations are covered. They range in number from 1 in
256
-------
LAKE ONTARIO
CHARLOTTE
PUMPING
ITATION
LEGEND
O MONITORING LOCATION
( ) CAPACITT IN NED
DURANO-EASTMAN
TREATMENT PLANT
MAPLEIOOD PARK
PUMPING STATION
G>
0-
(275)
( 4*7 )
(4)
(6)
(15)
(17)
(32}
(350)
SENESEE VALLET
CANAL SEWER —
17
^>
<~>
18
O
(200)
INTERCEPTOR
SEIER
\
X
\
(190)
N 0 R T 0 N - D E M S M 0 R E
SCREENING I CHLORINATION STATII
( 1 3 5 )
EAST SIDE TRUNK SEVER
(279)
(S2)
(81 )
^ 1EST SIDE TRUNK SEIER
(40)
(139
(27)
(25)
L
•O-i
I
TTRON PARK
PUM PING
STATI ON
(45)
(28)
(236)
(2)
ELMIOOD AVE
PUMPING STATION
FIGURE 3. FUNCTIONAL ELEMENTS OF THE ROCHESTER SEWERAGE SYSTEM
257
-------
Delaware to 19 in Texas. The period of record is generally
from August 1949 to the current date with some gaps. Tapes
are furnished 9 track-800 BPI, unless otherwise specified,
and are forwarded air parcel post. Recent experience with
these tapes has been excellent. The two tape files for
Rochester were ordered by telephone on March 11, shipped
on March 20, and received on March 26, for a combined cost
of $140.
Sorting the data requires a precise storm event definition.
The one Metcalf § Eddy has standardized upon for the present
defines a storm as starting with the first measurable rain-
fall after a minimum of 6 hours with no rainfall and ending
when a gap in measured rainfall (precipitation) of at least
6 hours is first encountered. Trace rainfall amounts are
disregarded. The 6-hour gap was selected to ensure relative
independence between events. For each event in the histori-
cal record, the following are noted and punched on data cards
or filed on disk: date, starting hour, duration, total
rainfall, maximum hourly rainfall and the hour in which it
occurred, elapsed days since the previous storm, and occur-
rences of excessive precipitation and snow. The program
routine is shown in Figure 4.
The data are then sorted and ranked in arrays according to
items of interest, such as total rainfall, duration, maxi-
mum hourly rainfall, and summer storms. Selected results
are reflected in Figures 5, 6, and 7. Other advantages of
the arrayed data are that current storms can be directly
classified in historical perspective, and probabilities of
multiple rare occurrences within a given time span can be
determined by inspection .or simple calculation. For
Rochester, sorting and ranking 3,000 storms by one parameter
*
took 0.65 minute of execution time and 0.12 minute of CPU
time on an IBM 360.
258
-------
SUNMART of on HOUILT PRECIPITATION
urcs
fCITKtl IURC»» RAINFUl RECORD
PROCESS
ONLINE
STORAGE
(DISK)
PROCESS
OHllHf
STORAGE
PROCESS
CARD
C345
OKI I HE
STORAGE
(DISK)
rtocR.
RAINFAll-
JTOIN EVENT
(ITORK I
EVENT I I
PROCESS
ONLINE
STORAGE
(OISI)
PNOCR.
SNOI HCt.
II STORM EVENT
(HORI /
CHIT 1 I
TRUUPfllED 10 tltl STORtCE
»\Sl JTOI1CE
CHlNCt "-" TO " " 11 Alt.
ELEVEN DOLE REMOVED.
RECott in EDITED FORM on oisi rot DIRECT «CCEJS.
HOURLY PRECINTHTION PROCESSED TO CUITE I NEI FILE
IESCRKIHC SIORK EVENTS VHICU »(IE DEFINED IS FOL101S:
EICH STORM EVENT SHUTS KITH THE FUST UEISURltlE
RAINFIIL AFTER » VIXUUU OF B COKSECUTIVE HOURS OF NO
lEtSURED miKFHL. THE STORM EVENT ENDS »T THE LIST
IEASURULE RAINFklL HAVING THE NEXT 6 CONSECUTIVE HOURS
OF 10 MEASURED RAIN.
tTORU {VENT FILE CONTAINING VEAN, MONTN AND OAV OF
DCCURANCE. DURATION OF STORM IN NOURS. TOTAL RAINFALL
OF STORM IN INCHES. MAXIMUM HOURLY RAIVFIII IN INCHES
•AND ITS OCCURRENCE AFTER START OF STORM, NUNIER OF
OATS SINCE LAST STORM EVENT, AND HOUR AT IHICH THE
STORK STARTED.
THE FILE CONTAININO THE SUMMARY OF OAV IS CHECKED TO
SEE IF SNOHFHl VAS INCLUDED IK A STORN EVENT. TNIS
INFORMATION IS ADDED TO 1HE STORM EVENT TILE.
STORM EVENT FILE INCLUDING SNOVFALL INFORMATION
MEFARED-FOR LISTING ANO RANKING PROCESS.
FIGURE 4. TAPED RAINFALL TRANSFER AND EDIT ROUTINE
259
-------
I
.AUG.
60 ON
1 SE
. ANNUAL
IASIS
IASIC DATA
• 21.5 YEARS OF RECORD
JULt 1.1950-010.31,19)1
• IASHINGION NATIONAL
AIRPORT CASE
0.1
.5 J S 10
OCCURRENCES PER TEAR
90 100
FIGURE 5. STORM MAGNITUDE VS. FREQUENCY OF OCCURRENCE
so
Itau
9 10 IS
HOUR AFTER START OF STORM
.
FIGURE 6. HOUR DURING STORM HAVING MAXIMUM RAINFALL
i
20
i
1C
-
1
.
_.
•
--
•
II
II
LENITK Of Pfllll. DATS
FIGURE 7. DRY PERIOD BETWEEN STORM EVENTS
260
-------
Storage-Treatment Balance
The storage-treatment balance concept is shown in Figure 8.
In the simulation, historical rainfall is read in chrono-
logical order, converted by a factor (coefficient of runoff)
into runoff, and stored in a specified volume which is
emptying at a specified rate. When the runoff exceeds the
combined storage-treatment rate, an overflow occurs. The
program flow-chart is shown in Figure 9. Initial runs are
executed on a daily basis for the full period of record and
over the range of the viable alternatives. Simulation of
20 years for one (alternative takes less than one minute at
a cost of approximately $20.
For periods of greatest interest, those producing overflows,
repeat runs may be cycled on an hourly basis to develop a
hydrograph and refine the storage-treatment simulation.
Input variables at the present time are simply the land
area, runoff coefficient, storage capacity, and treatment
rate. Sample output is shown in Table 4. A significant
operating rule under reevaluation controls the starting of
the treatment facility. As programmed, the treatment
deduction from storage does not start until the cycle
following the actual runoff, unless it happens to be
running from a previous storm. When run on a daily cycle,
this is intended to cover the uncertainty as to when during
the day the storm actually started. When run on an hourly
cycle, this delay represents a logical startup requirement.
The program is used by varying the storage capacity and
treatment rates, and noting the changes in overflow occur-
rences and durations. An example is shown in Figure 10
261
-------
R
U
N
0
F
T
u
0
V
R
F
0
W
&
J
erect i
MAX.
STORAGE
TREAT. RATE
ItTMT
FIGURE 8. STORAGE-TREATMENT BALANCE CONCEPT.
262
-------
READ
fENERAL OATA: HE*. DUNOfF COEFFICIENT, MIX I HUH STORAGE.
TREATMENT RATE. MINIMUM STORAGE. NUNIER OF
UIRS TO u Ammo.
00 I = 1
N TEAR
START CALCULATION: FOR A TEAR.
READ
YEAR TO RE ANALYZED, NUMBER OF IATA CARD:.
READ
res
•RITE
RAINFALL DATA: MONTH. DAY, QUANTITY OF RAINFALL. (10 DAYS
RECORDED ON EACH DATA CARD).
DATA CHECK: NUIDER OF DAYS IN MONTH. NUMBED IF MONTHS IN YEAR.
ERROR MESSAGE.
TERMINATE HOtRAN.
START CALCVLATIIMS ON DAILY RASH.
fROCEER Tl NEXT (AY'S DATA.
RAINFALL CNECI: III RAIRFALL OCCUR ON THIS OAYf
STORACE CNECI: II AIAIIAIU STIRACE EMfTYT
IETERMIIE IIAHTITY IF RAINFALL AND RUNOFF. NT RVIOFF INTO
ITIIAIE.
FI6URE 9. STORAGE-TREATMENT ROUTINE
263
-------
li TREATMENT OPERATIONAL FOI TNIt WtET
REDUCE HOUSE IT KITE OF TIEATIIEIIT.
1TAIT TREATMENT FOI NEXT CICIE.
II III I MUM ITOIACE EXCEEDED?
CALCULATE OUANTITI OF IATEI IN EUEIf OF NAXINUN STOIACC INI
IECURE IT TO IE OVEIFIOI.
CltCILITE: TOTH I1INFUL FOI MOUTH. OUUTITT OF IUN1FF,
OUIFIOI TREITED. STORED UD 0»I» OF HIM INO
KEIFIOI TO DATE FOI TEU.
HIT: D»TE (MONTH 1ND DM), IIINFUL. IUMOFF. STOIKE
HEATHEN!. IMO OKERFLOI FOI Oil OF OCCUMEUCE INO |g|
(F tUINTITIEt TO DATE OF KUNOFF. OUIFLOt. ITOIAKE
TKEATIEIIT, OATS Of IAIN UNO OVERFLO*.
lAt DATE FOI EACH (AT OF TEAI IEEN XCAO?
Hit EACH TEAR OF RECORO IEEN AIAITZEIT
TEININATE MOCIAI.
FIGURE 9. (CONTINUED)
264
-------
OCCURRING ON THE DATE
ACCUMULATED FROM START OF THE YEAR
in
I t-tv\ i ;
MONTH DAY
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
71 1
RAIN
IN
0.07
0.00
1.62
0.67
0.48
0.00
0.00
0.00
0.00
0.00
0.32
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.17
0.00
0.00
0.00
0.00
0.00
0.00
0.00
3.79
0.00
0.00
0.00
0.00
RAIN
DAYS
66
66
67
68
69
69
69
69
69
69
70
70
70
70
70
70
70
70
71
71
71
71
71
71
71
72
73
73
73
73
73
RUNOFF
MG
19.59
0.00
453.30
187.48
134.31
0.00
0.00
0.00
0.00
0.00
89.50
0.00
0.00
0.00
0.00
0.00
0.00
0.00
47.57
0.00
0.00
0.00
0.00
0.00
0.00
16.79
1060.50
0.00
0.00
0.00
0.00
STORAGE
MG
19.59
0.00
453.30
187.48
134.31
0.00
0.00
0.00
d.oo
0.00
89.54
0.00
0.00
0.00
0.00
0.00
0.00
0.00
47.57
0.00
0.00
0.00
0.00
0.00
0.00
16.79
590.00
340.00
90.00
0.00
O.-QO
OVERFLOW
MG
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
237.29
.0.00
0.00
0.00
0.00
TREATED
MG
0.00
19.59
0.00
250.00
250.00
250.00
25.09
0.00
0.00
0.00
0.00
89.54
0.00
0.00
0.00
0.00
0.00
0.00
0.00
47.57
0.00
0.00
0.00
0.00
0.00
0.00
250.00
250.00
250.00
90.00
0.00
T. RUNOFF
MG
6964.54
6964.54
7417.84
7605.31
7739.62
7739.62
7739.62
7739.62
7739.69
7739.62
7829.16
7829.16
6829.16
7829.16
7829.16
7829.16
7829.16
7829.16
7876.73
7876.73
7876.73
7876.73
7876.73
7876.73
7876.73
7893.51
8954.01
8954.01
8954.01
8954.01
8954.01
T. OVERFLOW OVERFL
MG
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
237.29
237.29
237.29
237.29
237.29
DAYS
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
T. TREAT
MG
7096.08
7115.66
7115.66
7365.66
7615.66
7865.66
7890.75
7890.75
7890.75
7890.75
7890.75
7980.29
7980.29
7980.29
7980.29
7980.29
7980.29
7980.29
7980.29
8027.86
8027.86
8027.86
8027.86
8027.86
8027.86
8027.86
8277.86
8527.86
8777.86
8867.85
8867.85
TREAT MAX STORAGE
DAYS
28.38
28.46
28.46
29.46
30.46
31.46
31.56
31.56
31.56
31.56
31.56
31.92
31.92
31.92
31.92
31.92
31.92
31.92
31.92
32.11
32.11
32.11
32.11
32.11
32.11
32.11
33.11
34.11
35.11
35.47
35.47
MG
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
495.27
590.00
590.00
590.00
590.00
590.00
TOTAL RAIN 7.18
TABLE 4. SAMPLE OUTPUT (DAILY CYCLE)
-------
11,000
1.000
y 4.000
I.OOO
3
• J.ooo
1.000
V
Mi
" 500
Cm
M
_, 100
100
lit
k-
5fi
i
•
v, ^^
x,^
fEIIOD ONlf-^"^
I.I O.J 0_J 0.4 0.5
V.
i
t
1.0 ) 141
mtm Huntit OF o»E«nou
.HHNUIL
^
10 10
^v
It 40
line
t me
«*INf»LL IICO.D HJ7-I9,,
UTi
. '«E»riiE»r sums
FILlOIINt (10*1
• coiimto SEIIMO HE* . i}.4n Ulll
FIGURE 10. ANNUAL OVERFLOW OCCURRENCES VERSUS STORAGE
CAPACITY.
266
-------
comparing varied storage volumes to a constant treatment rate
and noting the change in overflow occurrences.
Overflows, Quantity and Quality
The quantity and frequency of overflows is determined by the
storage-treatment balance computation. Assigning quality
characteristics can follow one of two approaches: (1) the
approach can follow a dust and dirt buildup and removal
formulation as in SWMM and STORM, or (2) it can directly
compute concentrations through some simple relationships or
regression techniques. The latter, more direct approach is
preferred when supported by at least a minimum of measured
real data.
In the District of Columbia study introduced in Part I,
significant overflows were computed on an hourly basis.
Quality characteristics were added using the following
assumed relationships:
Suspended solids (SS) concentrations in milligrams
per liter were computed from the following expression
for both combined and separate discharges:
SS = 400 x fi x f2 x £3
where f^ is a function of days since the last
storm and the time from the start of
overflow (matrix range 0.3 to 2.6)
f2 is a function of rainfall intensity
(0.5 to 1.5)
£3 is a function of catchment population
density (0.5 to 1.0) ;
A new concentration was computed for each time
increment (0.5 to 1.0 hours)
267
-------
BOD5 was computed from the suspended solids
concentrations as follows:
For separate storm drains
BODs (storm) = .10 x SS [SS <300]
BOD5 (storm) = 30 + (SS-300) x .08 [SS >300]
For combined overflows
BODs (comb) = <* D + (1 - <*) x BODs (storm)
where « is the proportion of combined flow attributed
to average dry-weather flow
D is the average BODs of dry-weather flow
Where needed, total nitrogen and total phosphorus were
approximated by the following expressions:
N = .10 BOD5
P = .033 BOD5
Typical utilization of these data are shown in Table 5 where
loadings from various discharges are compared for an intense
summer storm. The cost-benefit priorities, based on this
one parameter, are also shown. Similar comparisons were
made for a major long storm, average summer quarter, and
annual bases.
Table 5. COMPARATIVE DISCHARGES
FROM INTENSE SUMMER STORM
BOD5,
1,000 Ib Priority
Runoff from existing
separate sewer areas
Untreated 125 • S
Combined overflows
Untreated 174
If separated 164
With storage-treatment 9 2
STP discharged (daily)
Existing 13*
Upgraded 13 *
Note: Two-thirds of the existing area has
separate sewers.
268
-------
In Rochester, since field data are being collected concur-
rently, new relationships are being tested with emphasis on
overflow traits at specific locations. Early comparisons are
shown in Figure 11, which represents schematic plans of the
system with superimposed ranking of land use and population
densities, runoff coefficients, and constituent constraints.
The intent is not only to identify trends but also to screen
put potentially faulty data. While the present scatter may
appear-rather meaningless, the utility of the scheme is
expected to increase as the data base grows.
The purpose of the analysis is, of course, to rank the
potential pollution sources so as to guide emphasis in a
staged corrective program or operational strategy.
Receiving Water Response
Receiving water response is perhaps the highest ranked study
objective, yet it is also the least defined in terms of
present day real time data. In the absence of these data,
it is our practice to use the best available simulation
model that has some record of use in the area. This permits
testing wet-weather impacts on a technical level compatible
*
with previous analyses for treatment plant discharges, salt
water intrusions, etc.
While there are some misgivings with this limited approach
(we frankly have not come upon a simplifying breakthrough),
credibility requires and benefits from the similarities
in techniques.
ROCHESTER AND B.C. DEMONSTRATION PROJECTS
The final test of a model or approach to modeling is using it.
The application of simplified models to solve engineering
269
-------
I .
6
(
-
12
' . t-_.
3
10
i
9
i
,,j
M
*j
X-
^.
•
'
.
'i
|
1
Of •«
UJ C.
±s
•c >
i-
>
1
11
-
A
*t
21
8
21
21
11
21,29
7
it
r
g
2
8
5
,
12
11
i
i
8
*
1
•'
1
1
26
n
U
2 1
6
22
7
25
•
3
26 .2(
4
it
LAND USE - POP. DEN.
COD - TSS
(AREA AVERAGES)
RUNOFF COEF (ESTIMATED)
•»
i n
10
I
9
6
i
i
i
i
-
_.,
i ,•
1
3
21
7
21
8
22
7
25
1
21,21
4
ii,
t
10
5
4
8
i
«
6
', .,
•
e
3
21
'21
7
22
1
It
9
LLJL!
5
11
TIP
(AVE POSITION IN RANKING)
LEGEND: LOWEST RANKING NUMBER GIVEN TO HIGHEST CONCENTRATION.
POPULATION. OR PERCENTAGE OF COMMERCIAL INDUSTRIAL
AREA. RANKS 1-5 SHADED FOR EMPHASIS.
FIGURE 11. OVERFLOW RANKING TRENDS
270
-------
problems is not new; indeed, several cities, such as Chicago,
have rather distinguished programs. It is unfortunate, how-
ever, that their use is so little publicized.
Application of simplified models is an art. In fact, the
simpler the model, the greater the reliance on the judgment
of the user, and perhaps conversely, the greater the danger
of misapplication.
Tying It All Together
A simplified PERT worksheet is shown in Figure 12. This
worksheet, developed for the Rochester project, illustrates
the tying of the various programs together while focusing
on the result. Noteworthy are transfers of information
between tasks, shown by dashed lines.
Again, contributions of the individual tasks are dependent
on timing and level of effort, forcing decisions at the
designated transfer points.
Interfaces with SWMM
SWMM is a logical backup model for attempts at simplified
approaches. That is, it answer the question--If I had more
data or a more sophisticated approach, how would my results
be affected? Similarly, if the record of a large number of
SWMM runs exists for a particular area, the output may be
used as a basis for a simplified approach to broaden the
areal and time coverage.
Rochester - In Rochester, SWMM Runoff and Transport are being
calibrated for each of the 13 combined sewer overflow-points.
Once this is accomplished, the 13 overflows are to act as
nodes of an entire system analysis for specific events
271
-------
to
teiut i.s. Q ttrtLOf. crcKf. imsfn
m sun/a WBU i rim \MOCUT
fiFwrf
WHIU CKMOt,J!tU UtX tUI
^fitty b*t^ fci»« ^
\ icfir.'fr rff«5i V mit
w;i;i«
U
r.f 01 CK n
nisr ;«t
ff.HU.1
:«< s
flUI
P>KII gi»ff nwaa <
Knictfti n lofirffr eajfcm'fs < JLcttLtrt ro iirtK-fl \ wir n UM eniw
\ fltiiKK OF sxxiis war srsiai
\ ciYiicr swung, rxxs/u I MF;H i rect'W
X. isisurt cmtcju. .-tzu"'- ro Kic;i!a;ig snsir:»inJuxiri>T rust 3 ctmT.~
X muxiius Y 70 "s* s~/'«r ^-
ILIUM fxaua - Kstssica i
Sf7 ir TS1H K1IS
iiurier otn uwiri'ins
i Ofnms
ciccirc mn r«i i o;r?
\ UK.\
FIGURE 12. SIMPLIFIED APPROACH ACTIVITY DIAGRAM
-------
centering around the dry-weather treatment plant and the
interceptor system.
The simplified model application is directed at the same
objective but in less detail and in the context of the total
historical record. SWMM results are to be used to refine
the runoff coefficient estimates in the simplified approach.
Also, in the Rochester program, an extension in simplified
modeling capabilities is proposed, wherein each of the 13
areas will be run in a cascading loop with the intercon-
necting interceptor capacities determining the treatment
rates. This will give a spacial aspect to overflows and
storage devices not previously included in this level
of analysis.
District of Columbia - In the District of Columbia simu-
lation, the simplified model results for a storm occurring
once in 2 years were applied to a dynamic (SWMM) receiving
water model of the Potomac estuary. The receiving water
model was fed data believed characteristic of summer
quarter tidal cycles and headwater and other background
inflows. Figure 13 represents the calculated effect of
upgrading the regional dry-weather flow treatment efficiency
from 90 to 98 percent in terms of BODs- Figure 14 repre-
sents the impact of the 2-year storm on the upgraded system
3 and 5 days after the storm occurrence. Surprisingly,
reducing the storm BODs loading by one-third before dis-
charge for this major event raised the minimum dissolved
oxygen in the river by only 0.5 mg/1, from 2.6 to 3.1 mg/1.
From these runs, the time it took the river system to
recover from the shock impact of a storm was also computed.
In the selected example, dissolved oxygen recovery was
essentially complete in 10 days.
273
-------
(V) CHAIN BRIDGE
CD HOUTH Of ROCK CR.
Qj Mlh SI. BRIDGE
(F) HOUIH OF THE ANACOS1IA RIVER
(T) BLUE PLAINS PLANT
CASE:
YEAR
FLO*
UKP. 27°C
SUNOARO ............
ACTUAL DATA
FT. WASHINGTON MODEL DATA
c) (). (
NILE: FROM CHAIN BRIDGE
FIGURE 13. IMPACT OF UPGRADING REGIONAL PLANT
L&J CD) CO
*-*,
N
V
*i ^
rN
-
S--,
7v,
V
MAIN PIVER
./
...
i
;
,
-KO STORK
1
^
__•
• *
- _
•>
- •
.1
X
**-
^, f1
'. ,_>
JiC
^
4
b^-h
NO
0 i 10 1
! 1
in
TR
i
>
, ,
(«
ii
.
•
'
»»J
1 - ,
^'
'
1
^ .
,
r*n
-STORM
-
— -
"^J^"
^
X
(DAY S)
210 2^ i
•ILES FROM CHAIN BRIDGE
FIGURE 14. IMPACT OF 2-YEAR STORM WITH UPGRADED PLANT
274
-------
Present Status
As this paper is being prepared, the Rochester project is
approximately 25 percent complete. Overflow data from 12
storms and 12 locations (64 discrete data sets because few
storms are sampled at all locations) have been collected and
analyzed. The his-torical rainfall has been received, cate-
gorized, and ranked. The storage-treatment model is
operational on the client's system, and the client's
receiving water model has been tested on our system. Snow-
melt and sequential routines are under development.
The District of Columbia reconnaissance-level study was
concluded with the development of a feasibility study plan,
estimated to be capable of capturing and treating approxi-
98 percent of the average annual combined sewer overflows
and limiting the average number of overflow occurrences
to less than one per year. Implementation of the plan is
estimated to be capable of reducing the total long-term
pollution loadings on the river system from all District
sources by 25 to 35 percent, as opposed to an estimated
7 percent reduction for a program of total sewer separation.
Potentials for Management Decisions
\
These demonstration projects have shown simplified model
approaches to be a valuable tool to decision makers.
Benefit potentials lie in the following areas:
• Data reduction and classification
• Relating cause to effect
• Preliminary screening
• Testing alternative approaches to a
common baseline
• Sensitivity analyses
275
-------
• Assessment of data worth, tracking ability,
or tendency (i.e., Does it fit the concept?),
and basis for mid-course correction
• Setting project priorities
• Predicting outcomes of specific actions in
short response times (i.e., real time controls)
• Communication and credibility
• Simplicity and low cost
As an example, suppose a manager is weighing two alterna-
tives, one of which is treatment dominant and the other is
storage dominant. He knows that, once functioning, the
treatment system will consistently improve performance as
the storm progresses and as its operating run lengthens.
He also knows that the first hour's operation for each
storm is likely to be spotty. Having characterization data,
such as shown in Figure 6, would indicate the relative
significance of the first hour. In this case, the first
hour is by far the most significant; thus, the storage
dominant option would be favored. Another interpretation
might be that the minimum required storage should at least
equal the maximum first hour runoff (adjusted for in-system
travel time and equalization).
Distinguishing Between Model Weaknesses and Frills
Model weaknesses are limitations in logic or simulation that
lead to inadequate or misrepresented results. Frills are
the accessories to the basic logic and simulation that have
little basis for use or consequence in results. The danger
in frills is not so much that they create inefficiency in
data gathering and application but rather that they might
mask the true weaknesses by symbolizing a precision that
does not exist or an operation which is unnecessary.
276
-------
SUMMARY
Part I
Selection of stormwater management models should follow a
systematic preassessment process:
• Define the study objectives
• Identify the given data base
• Identify the array of model capabilities
• Identify necessary assumptions
• Select for results
Of these tasks, defining the study objectives is both the
most difficult and potentially rewarding. The given data
base and assumptions include both static criteria, fixed in
the initial stages of the project, and dynamic criteria,
which change during and following the course of the work.
Differentiating between these categories will assist the
model selection.
Models follow operating rules without exception or inter-
pretation and can process mountains of data with relative
ease. The developer sets the rules and the user furnishes
the data and is left with the results. There are several
models to choose from. These models are separable by
function (planning, design, operation) and by degree of
precision (simple, intermediate, and complex).
Simplified models are capable of processing long periods
of record and broad areal coverage at low cost. Detailed
models reflect the dynamics of a system, and emphasis is
placed on the comprehensive analysis of singular events and
component systems. The advantage of these two levels of
approach is that, when understood, they can and should be
entirely complementary.
277
-------
Only the buyer can make final judgment as to the success or
failure of a modeling attempt.
Part II
Simplified modeling approaches should be used by decision
makers, the basic reason being that the technique-permits
a direct assessment of the Ipgic of an approach and the
weighted credibility of each of the data inputs without
unnecessary masking by detail.
Thfe. potentials are not limited to the planning stage, but
extend through the design,and into the implementation (con-
trol) and feedback stage. Subsystem capabilities include:
• Rainfall characterization
• Storage-treatment balance
• Discharge-receiving water response
Interfaces with the more sophisticated dynamic models have
been successfully demonstrated. However, a principal
advantage lies in keeping the models simple, flexible, and
unencumbered by unnecessary frills. Because of their
simplicity and broad perspective, they have a strong potential
in communicating ideas and projecting long-term outcomes of
rather explicit actions.
Acknowledgment
A portion of the work upon which this presentation is based
was performed pursuant to Grant No. Y005141 with the
Environmental Protection Agency.
278
-------
REFERENCES
Battelle Memorial Institute. Evaluation of Mathematical
Models for Engineering Assessment, Control, Planning,
and Design of Storm and Combined Sewerage Systems.
U.S. Environmental Protection Agency, Project No. CI-
73-0070.
Brandstetter, A. Comparative Analysis of Urban Storm-
Water Models. Pacific'Northwest Laboratories Division
of Battelle Memorial Institute. August 1974.
Huber, W.C. Modeling for Storm Water Strategies.
APWA Reporter. May 1975.
Roesner, L.A. Personal communication to author.
June 1975.
Lager, J.A. and Smith, W.G. Urban Stormwater Manage-
ment and Technology - An Assessment. Metcalf § Eddy, Inc
U.S. Environmental Protection Agency Report EPA-670/2-
74-040. December 1974.
O'Brien § Gere Engineers, Inc. Combined Sewer Overflow
Abatement Program, Rochester, N.Y. EPA Grant Y005141
to Rochester Pure Waters District, June 1974.
U.S. Army Corps of Engineers. Urban Runoff: Storage,
Treatment and Overflow Model "STORM." Hydrologic
Engineering Center Publication 723-S8-L2520.
September 1973.
279
-------
A STORAGE, TREATMENT, OVERFLOW AND
RUNOFF MODEL FOR METROPOLITAN MASTERPLANNING
By
Larry A. Roesner, Ph.D.*
CONCEPT OF "STORM"
The quantity or urban runoff has traditionally been estimated by using a
design storm through frequency-duration-intensity curves or some other
statistical means based on rainfall records4 Such approaches normally
neglect the spacing between storms and the capacity of the urban system
to deal with some types of storms better than others.
Often, through natural and artificial storage mechanisms, intense short-
duration storms may be completely contained within storage so that no
untreated stormwater overflows to receiving waters. Alternately, a series
of closely spaced, moderately sized storms may tax the system to the point
that excess water must be released untreated. Consider, for example,
Figure 1 which shows the response of two different systems to the same
rainfall trace. System A, which has a relatively high treatment rate and
a small storage capacity, will overflow during the high intensity, short
duration storm. However, it will completely contain the second storm of
moderate intensity and longer duration. System B, on the other hand,
which has a low treatment rate and a large storage capacity, completely
contains the first storm. Notice that it would also contain the second
storm if the system were analyzed independently of the antecedent storm.
However, in this case the spacing of the storms is such that the system
analysis must include both rainstorms as a single event to accurately
describe the system's response to the rainfall trace illustrated in the
figure.
A storm cannot be defined by itself, but must be defined taking into
account the response characteristics of the urban stormwater system. It
is for this reason that an approach was developed that would not only
recognize the properties of rainfall duration and intensity, but would
also consider storm spacing and the capacity of the urban stormwater
system.
Principal Engineer, Water Resources Engineers, Incorporated,
Walnut Creek, California.
280
-------
'
'
RAINFALL TRACE
TIME—>
.
SYSTEM A ' TREAT • 4 UNITS/HOUR
STORAGE • 4 UNITS
OVERFLOW
4 UNITS OF
T
TIME
SYSTEM B : TREAT • 2 UNITS/HOUR
STORAGE • 24 UNITS
VERFLOW
STO«UGl CUKVC F0« SECOND
STORM IT FMST STORM IS IGNORED
TIME
Figure 1. System Response Examples
PRECIPITATION
(rolnfoll/tnow)
POLLUTANT
W*SHOFF AKO
111
Figure 2. Conceptualized View of Urban System Used in
"STORM".
281
-------
Figure 2 shows, pictorially, the interrelationship of the eight storm-
water elements considered in this approach for estimating stormwater
runoff quality and quantity. In this approach, rainfall washes dust and
dirt and the associated pollutants off the watershed to the storage-treat-
ment facilities so that as much stormwater runoff as possible can be
treated prior to its release. Runoff exceeding the capacity of the treat-
ment plant is stored for treatment later. When the storage facilities
become inadequate to contain the runoff the untreated excess is wasted
through overflow directly into the receiving waters.
For a given precipitation record, the quantity, quality, and number of
overflows will vary as the treatment rate, storage capacity, and land use
is changed. Land surface erosion is a function of land use, soil types,
ground slope, rainfall/snowmelt energy and erosion control practices. A
typical method of investigation is to alter the treatment, storage, and
land use and note the resulting response of the system. A group of alter-
natives can then be selected from among those meeting the overflow quan-
tity and quality objectives.
COMPUTATION OF THE QUANTITY OF RUNOFF
*
Runoff is calculated on an hourly basis as a function of rainfall plus
snowmelt using the following expression:
R - C(P - f) (1)
where
R « urban area runoff in inches per hour;
C = composite runoff coefficient dependent on urban
land use;
P = rainfall plus snowmelt in inches per hour over
the urban area; and
f - available urban depression storage in inches per hour.
For simplicity we will omit'the snowmelt computation in our discussion
here. The interested reader is referred to the User's Manual [ij for
details of that computation.
The runoff coefficient represents losses due to infiltration. It is com-
puted from land use data as follows:
L
C - C + (Cz - C ) I X± Ft (2)
where
C = runoff coefficient for pervious surfaces;
P
C- runoff coefficient for impervious surfaces;
X - area in land use i as a fraction of total watershed area;
F = fraction of land use i that is impervious; and
L » total number of urban land uses.
282
-------
Before the runoff coefficient is applied, depression storage losses must
be satisfied. Depression storage represents the capacity of the water-
shed to retain water in ditches, depressions and on foliage. The amount
of depression storage at any particular time is a function of past rain-
fall plus snowmelt and evapotranspiration rates. The function is computed
continuously using the following expression, where f is in inches:
f - f + N_ k, for f < D (3)
o u ~~~
where
f = available depression storage, in inches, after
previous rainfall;
Nn - number of dry days since previous rainfall;
k = recession factor, in inches /day, representing the
recovery (evapotranspiration) of depression storage
in inches ; and
D = maximum available depression storage in inches.
Figures 3a and 3b show graphically the hourly precipitation (P) , depres-
sion storage (f) , precipitation excess (P-f) and the resulting runoff (R) .
Figures 3b and 3c show how the runoff is distributed between treatment
storage and overflow for a system with a treatment rate of 0.02 inches/
hour and a storage capacity of 0.16 inches.
ESTIMATION OF THE RATE OF POLLUTANT BUILDUP ON URBAN WATERSHEDS
The estimates of pollutant accumulation rates on the urban watershed,
and the rate of washoff during a storm event follow closely the methods
used in SWMM(2) . The rate of dust and dirt accumulation, DD for a given
land use L is expressed as:
DDL - d^ x (GL/100) x ^ (A)
where
DD - rate of dust and dirt accumulation on a watershed
of land use L in Ibs/day;
dcL = rate of dust and dirt accumulation on watershed L
in Ibs /day/100 feet of gutter;
G = feet of gutter per acre in watershed L; and
A B area of watershed L in acres.
The rate factor ddL should be supplied by the user for his area. Default
values, (APWA factors; see Reference 2) are incorporated into STORM and
can be used if no better data are available.
The initial quality of a pollutant p on watershed L at the beginning of
a storm is then computed as:
Pp -
-------
or
v
2
2 •
Ul
RAINFALL RJ
,
.20 -
1 '
z
i"
<
CC
U.
U.
o
g
0
i/
"1
(_ JP
1 S 10
b/
i i i [~
t ' 's ' 10
RAINFALL
E23 EXCESS RAINFALL (fM)
AVAILABLE DEPRESSION STORAGE
(moximum = .06 inches)
rJ"P
15 20 25 30 35
TIME IN HOURS
— RUNOFF (C(P4));C = .9
TREATED RATE - .02 in/hour
^ UNTREATED RUNOFF
——
Wi
ma ^ ^
T 1 1 f n r~i
15 20 25 90 JS
TIME IN HOURS
STORAGE LEVEL
— STORAGE CAPACITY
ESS OVERFLOW
IS 20
TIME IN HOURS
Figure 3. Time Histories of Rainfall, Runoff, and Storage.
284
-------
F = pounds of pollutant p per pound of dust and dirt;
N_ = number of dry days since the last storm, and
P = total pounds of pollutant remaining on watershed L
p° at the end of the last storm.
In practice, pp is usually limited to the amount that would be accumulated
in a 90-day dry period. The reason for this is that the efficacy of
extrapolating daily buildup rates beyond this point (which was arbitrarily
selected) is uncertain. Moreover, if equation (5) is used repetitively
over long periods of time, positive errors could tend to accumulate in
P resulting in overly large values of P .
DETERMINATION OF URBAN RUNOFF POLLUTION LOADS
To compute the amount of pollutant washed off the watershed during a
storm, it is assumed that the amount of pollutant removed at any time t
is proportional to the amount remaining:
dP
We stated earlier that the runoff rate Q also affects rate of pollutant
removal, therefore K must be functionally dependent upon Q. However,
given two identical watersheds except for their area size, for the same
rainfall rate r on both watersheds a higher runoff rate would occur from
the larger watershed. This area effect can be eliminated by dividing the
runoff Q by the impervious area of the watershed. The impervious area is
used because only a negligible amount of the runoff comes from the per-
vious area. Since cfs per acre are equivalent to inches per hour, we can
say that K is functionally dependent on the runoff rate R from the imper-
vious area, where R is in inches per hour. Finally, assuming that K is
directly proportional to R and that a uniform rainfall of % inch per hour
would wash away 90 percent of the pollutant in one hour (a somewhat
arbitrary assumption), we can say that K = 4.6R. Making this substitution
into equation (6) and integrating over a time interval At (during which
R is held constant) gives:
Pp (t + At) - Pp (t)e-4'6RAt (7)
Equation (7) is the basic form of the overland flow quality model de-
veloped by Metcalf & Eddy, Inc., as part of the EPA Stormwater Management
Model [3]. Although it is simplistic and contains many assumptions, it
is the best overland flow water quality predictor or simulation model
that presently exists. Moreover, experience with that model (See Refer-
ence 4 and Volume II of Reference 3) has shown it to give fairly good
results. The rate of removal of mass from the watershed M is simply
[P(t)-P(t+At)]/At, which can be expressed as: P
Mp - P(t) x (1 -e"4'6RAt)/At (8)
285
-------
The variation o'f M with time for the associated hydrograph is plotted in
the lower graph ofpFigure 4. A plot of Mp versus t is termed a polluto-
graph . one of the most informative methods for expressing the pollutant
load carried by urban runoff. To determine the concentration of a
pollutant in the runoff as a function of time, one simply divides the
pollutograph value M by Q (with appropriate conversion factors).
Equation (8) must be modified, however, because not all of the dust and
dirt on the watershed is available for inclusion in the runoff at a given
time t. Thus pollutants which are tied to the dust and dirt are not all
available either. The Storm Water Management Model study [3] found that
for suspended solids the available fraction at any time was :
A = 0.057 + 1.4R1'1 (9)
sus
For settleable solids it has been assumed that the availability factor is
A - 0.028 + l.OR1'8 (10)
S 6 1
With regard to BOD, nitrogen and phosphate, recall that the APWA data [2]
described the dissolved fraction, which is independent of the amount of
solids available for runoff. In the Storm Water Management Model study
it was found that the BOD associated with the suspended solids was about
10 percent of the suspended solids load. We have further assumed that
the BOD tied to the settleable solids is two percent of the settleable
solids. For nitrogen and phosphate, we have assumed that BOD, N, and
P04 are associated with suspended and settleable solids in the same
proportion as they are in the dissolved state.
Thus, correcting equation (8) for available suspended and settleable solids
and adding the BOD, N and P04 found in the solids, we get the following
set of equations which are used in STORM:
Suspended Solids
M (t) - A P (t) x EXPT (11)
sus sus sus N '
where - ..
A - 0.057 + 1.4XT'1
sus
EXPT = (l-e"4'6RAt)/At, with At - 1 hour
Settleable Solids
Mset(t) - Aset Pset (t> * EXPT
where , R
A - 0.028 + IT*0
set
BOD
Pbod(t) X EXPT + °'10 A8US + °'02 Aset
286
-------
^J
^J
g p !
O ' i+At
EMAINING
>Pt
or Q
UJ
h- x
ZC/)
-------
Nitrogen
M , (t) = P , (t) x EXPT + .045 A + .01 A (14)
nit nitv ' sus set
P04
MP03Ct) = PP04(t) X EXPT + -°045 Asus + '001 Aset
COMPUTATION OF TREATMENT, STORAGE AND OVERFLOW
Computations of treatment, storage and overflow proceed in an hourly step-
by-step method throughout a period of rainfall/snowmelt record. For every
hour in which runoff occurs the treatment facilities are utilized to treat
as much runoff as possible. When the runoff rate exceeds the treatment
rate, storage is utilized to contain the runoff. When runoff is less than
the treatment rate, the excess treatment rate is utilized to diminish the
storage level. If the storage capacity is exceeded, all excess runoff
overflows into the receiving waters and does not pass through the storage
facility. This overflow is lost from the system and cannot be treated
later. While the storm runoff is in storage its age is increasing.
Various methods of aging are used including average, first-in: last-out,
first-in: first-out, or others depending on the physical conditions
encountered.
The computation of storage and the interplays among rainfall/snowmelt,
storage and treatment represent a simplistic approach for dividing a rain-
fall record into unique events such that the event is defined in terms of
the urban system. For example, whether two "storms" are considered as
two isolated occurrences or as one large storm is entirely dependent upon
how the system will react to them. If the system has not recovered from
the first when the second arrives, the two definitely will interact and
hence must be considered together. "Events" are defined as beginning
when storage is required and continues until the storage reservoir is
emptied. All the rainfall occurring within this period is regarded as
part of the same event. If precipitation produces runoff that does not
exceed the treatment rate, the runoff will pass through the treatment
process but will not register as an event. From the standpoint of the
urban stormwater system, such precipitation is inconsequential and hence
is not part of an "event" even if it should occur immediately preceding
an obvious event.
The runoff coming into the storage/treatment system is given by equation
1(1). The quantity of system overflows are computed using:
Q0 - R - QT * Qs (16a)
Q - minimum of (R + Q , T) (16b)
T Vl
Q » minimum of (R - Q_ , S) (16c)
S JL
where
Q » watershed runoff overflow, in inches;
Q_ - watershed runoff treated, in inches;
288
-------
Q = watershed runoff stored, in inches;
s
Q = watershed storage remaining in previous hour, inches;
St-l
R = watershed runoff as calculated using equation (1), inches;
T = treatment rate in watershed inches/hour; and
S = storage capacity in basin, inches.
The quality of system overflows are computed as follows for each pollutant
for each hour:
M = M (Q /R) (18)
po p xo
where
M _y = M - M (19)
pT/s p po
M = total pounds of pollutant overflowing from system;
M - total pounds of pollutant p coming into the system; and
M , = total pounds of pollutant p going to storage/treatment.
The program does not model the treatment process but it does compute the
quantity of water treated. It is assumed that the pollutants will be
reduced to an acceptable level before the storm water is released. The
age of pollutant in storage is computed as previously mentioned.
INPUT AND OUTPUT FOR "STORM"
The basic input data required by STORK and the basic output data generated
by the program are illustrated in Figure 5. This section defines the
input data requirements more specifically and contains details on STORM
output. :
INPUT DATA REQUIREMENTS
Hydrogeometrie Data
The first step in setting up data for the simulation model is to define
the boundaries of the basin which is to be investigated* specifically
that area which drains to some specific point of Interest such as a
receiving water. The size of the area is a computation variable but it
should be limited to less than 10 square miles so that travel time in the
system can be neglected.
Once the drainage basin boundaries are set the following information is
required:
1. Size of the total area of the basin
2. Percent of the total area in each of the following land
use groups:
289
-------
N)
VO
O
INPUT
I. LAND USE
INFORMATION
Residential
Multiple Family
Commercial
Industrial
Park
A. AREA
B. % IMPERV.
C. RUNOFF COEFF.
D. FT. OF GUTTER/ACRE
E. HYDROLOGIC RECORD
OF HOURLY RAINFALL
ANALYSIS
I. RUNOFF
where:
TL QUALITY
—KR
Mp=P(l-e )
A. SUSP SOLIDS
& SETT. SOLIDS
C. BOD
D. NITROGEN
E. PHOSPHORUS
IE. STORAGE
OUTPUT
I. STATISTICS BY
EVENT
A. RAINFALL
B. STORAGE
C. OVERFLOW
D. TREATMENT
E. QUALITY
II AVERAGE
STATISTICS FOR
ALL EVENTS
A. EVENTS/YR.
B. AVERAGES OF A-E
ABOVE FOR ALL
EVENTS
C. OVERFLOWS/YR.
AVERAGES OF A- E
ABOVE FOR ALL
OVERFLOW EVENTS
Figure 5. Input-Output Elements.
-------
a. Single Family Residential
b. Multiple Family Residential
c. Commercial
d. Industrial
e. Open or Park
3. Average percent imperviousness of each land use group
4. Feet of gutter per acre for each land use group
5. A runoff coefficient for impervious areas (the usual
range is 0.8 to 0.9)
6. A runoff coefficient for pervious areas (the usual
range is 0.1 to 0.3)
7. The depression storage available on the impervious
areas (usually 0.05 to 0.1 inches).
Determination of the percent of area under the various land uses can be a
tedious task. However, most jurisdictions have this information already
available in one form or another, or they have maps of sufficient scale
that the various land uses can be identified and their areas calculated.
Determination of the average percent imperviousness of each land use group
must be done carefully, for this is the most sensitive parameter affecting
the amount of storm runoff which comes off a watershed. In particular,
impervious surface areas which drain to pervious areas should be excluded
from the impervious fraction because runoff from these surfaces will
probably be held on the pervious surfaces to which they drain. In this
regard, special attention should be given to whether the house gutter
drains to pervious land or to the sewer system or gutter. Also, in
residential areas where there are parking strips, the sidewalks will most
likely drain to the pervious area on either side of the walk.
Estimation of the number of feet of gutter per acre is best done from a
plat. But if none is available a reasonable estimate can be based on the
average size of the block by taking the perimeter of the block times the
number of blocks per acre from a street map.
Hydrologic Data
A record of hourly rainfall is required. The rainfall record may be as
long or as short as desired but should be of sufficient length to assure
that all storms of interest are included in the record. Ten to thirty
years of record is desirable. A long raingage record exists for most
cities. Where such Information is lacking, however, standard hydrologic
procedures for areal translation of rainfall records will have to be
applied.
Quality Data
The quality data required for the simulation model consists of:
291
-------
1. The dally rate of dust and dirt accumulation in pounds
per- 100 feet of gutter for each of the land use areas:
a. Single Family Residential
b. Multiple Family Residential
c. Commercial
d. Industrial
e. Open or Park
2. The pounds of each of the following pollutants per 100
pounds of dust and dirt for each land use category:
a. Suspended solids
b. Settleable solids
c. Soluble BOD
d. Soluble N
e. Soluble P04
3. The interval in days between street sweepings for each
land use category
4. Street sweeping efficiency (usual range is .6 to .9).
Because this data is difficult to obtain, default values are provided in
the computer program as follows:
1. Daily Rate of Dust and Dirt Accumulation*
Land Use Amount of D/D by Land Use,
Ib./day/100 ft. of Gutter
Single Family Residential 0.7
Multiple Family Residential 2.3
Commercial 3.3
Industrial 4.6
Open or Park** 1.5
2. Pounds of Pollutant in Dust and Dirt*
Land Use Lbs. of Pollutant/100 Ibs^pf D/D
Sus. Sett.
Solids** Solids** BOD N PO4
Single Family Residential 11.1 |.J 0.5 0.048 0.005
Multiple Family Residential 8.0 0.8 0.36 0.061 0.005
Commercial 17.0 1.7 0.77 0.041 0.007
Industrial** 6.7 0.7 0.3 0.043 0.003
Open or Park** 11.1 1.1 0.5 0.048 0.005
3. Street Sweeping Interval 90 days
4. Street Sweeping Efficiency 0.7
*Data is taken from APWA Chicago Study (See Reference 2).
**Estimated values from other sources.
292
-------
OUTPUT FROM "STORM"
The computer program produces four output reports:
1. Quantity Analysis,
2. Quality Analysis,
3. Pollutograph Analysis, and
4. Land Surface Erosion Analysis.
For the quantity and quality analyses, STORM generates statistics by event
plus the average statistics for all events. A complete list of the output
statistics from the quantity and quality analyses are contained in Table 1.
Tables 2 through 7 are examples of STORM output from the Quantity Analysis,
Quality Analysis and Pollutograph .Analysis. For details on the output
from the Land Surface Erosion Analysis the reader is referred to the pro-
gram manual [1J.
PLANNING APPLICATIONS
City of San Francisco
The City of San Francisco funded the original development of the Quantity
Analysis portion of STORM. The purpose of the program was to create a
tool that would enable the City to evaluate the effectiveness of various
combinations of treatment rate and storage capacity with respect to their
ability to reduce combined sewer system overflows. Results from STORM
were used as a guide in the initial sizing of facilities for the City's
Master Plan [5]. The following paragraphs are abstracted from the Master
Plan Report. Additional details are described in References 12 and 13.
The rainfall record used for the analysis was a 62 year U.S. Weather
Service record of hourly values of rainfall measured at the Federal Office
Building in the City. A runoff coefficient of 0.65 was assumed for the
analysis. Table 7 shows the various combinations of treatment rate and
storage capacity that were examined, the resulting events per year, over-
flows per year and the average quantity of overflow per year. These data
are displayed graphically in Figures 6 and 7, which show the relationship
between given combinations of storage and treatment with overflow fre-
quencies and with overflow volumes, respectively.
Application of STORM to the Federal Office Building record with 0 storage
and a treatment rate of 0.02 inches per hour provided the baseline or
existing condition data. From this computation it was determined that
approximately one-third of the runoff is presently treated and discharged
by the three water pollution control plants and that the other two-thirds,
or about 6.0 billion gallons of runoff per year, overflows without treat-
ment. This volume of overflow occurs during an aggregate average of 206
hours per year. On the average, there are 46 days in the year during
which 82 overflows occur.
The storage needed to contain all overflows from the greatest recorded
storm utilizing the existing treatment rates would be 240 million cubic
feet. This storage volume Is then the upper limit of an all-storage
293
-------
TABLE 1
"STORM" Output
I. STATISTICS BY EVENTS
A. RAINFALL
/ DURATION OF RAINFALL EVENT
2. HOURS OF RAIN
3. TOTAL RAINFALL
B. STORAGE
/. TIME SINCE LAST EVENT
2. DURATION OF STORAGE
3. TIME TO EMPTY
4, MAXIMUM STORAGE USED
C. OVERFLOW
/. TIME OVERFLOW STARTS
2. DURATION OF OVERFLOW
3. QUANTITY OF OVERFLOW
4. OVERFLOW IN FIRST THREE HOURS
D. TREATMENT
/ DURATION OF TREATMENT
2. QUANTITY TREATED
E. QUALITY (susp. solids, sett, solids, BOD, nitrogen, phosphorous)
/ MASS EMISSION IN RUNOFF
2. MASS EMISSION OF OVERFLOW
3. MASS EMISSION DURING
FIRST THREE HOURS OF OVERFLOW
I. AVERAGE STATISTICS (A-E ABOVE)
A. FOR ALL EVENTS
B. FOR ALL OVERFLOW EVENTS
C. EVENTS / YR
D. OVERFLOWS / YR
294
-------
TABLE 2
Example of Quantity Statistics by Event
SCL8Y STREET WATERSHED SAN FRANCISCO, CA.
QUANTITY ANALYSIS
to
VO
Ul
Tai
ST:
EVI
1
2
3
4
S
6
7
7A
III
11
12
U
14
13
16
17
18
2U
21
22
E»T"ENI RATE • .ntMfl IN/HI, 37. e CKS, 24.400 HGO FED BLOC, SAN FRANCISCO, CA.
9RAGE CAPACITY" .1HJ3 INCHES, " 31.2 AC-FT. IB. 168 MC SEL8Y STREET
TEAK MQ or MR STORAC. DU4TN HHS OUANTY INCHES EMPTY 9URTN MAX NO ST DUR XASTE INITL MRS OtNTY AGEt
1 64 IB 28 16 3422 33 17 1.73 .84 4 37 .18 1 3 9 .47 .07 37 .37 4.9
2 64 1 t 1 2 *5 3 2 ' 1& 08 6 9 05 3 NO OVFRFLOM 9 .08 2.1
3 64 11 2 3 16 0 0 .49 .23 IB 16 .10 2 2 S ,H6 .86 16 .16 6.1
4 64 H S U 1J6 26 16 .86 .42 8 34 .10 3 2 S .08 .US 34 .33 5. B
3 64 U 9 24 3 2H HI .97 .46 4 24 .10 4 2 4 .22 ,2B 24 .23 4.3
6 64 tl U 6 8 1 1 .03 .02 1 2 .00 t NO OVERFLO* 8 .04 .5
7 64 U U 12 2116 .36 ,1B 7 18 .06 3 NO OVERFLO* 19 .18 3.1
8 «4 U 12 7 1 3 J .tl .«« 2 5 .01 2 NO OVE.HFLON S .84 1.3
9 64 U 12 18 1*8 5 3 .H ««7 S 7 .113 1 NO OVERFLOW 8 .07 1.3
10 64 U ?1 20 19 34 .24 .lii 6 11 .06 3 NO OVERFLOM 11 .10 3.4
DEFINITIONS OF QUANTITY COLUMN HEADINGS
EVENT • SEQUENCING NUMHEK.
DATE • OATC THIS E»EM etc**.
H« • NUMBER OF HOUHS PAST MIDNIGHT THIS EVENT BEGAN.
STORAC • NUMBER OF HOUHS SINCE END OF LAST EVENT, EXCLUDING SUMMER (MORE THAN, 960 HOURS).
DU«TN • DURATION OF STORM MOM FIHSI HOUR OF RAIN TO LAST HOUR OF RAIN.
HHS • NUMHE-i OF HOURS IN «HIC*4 RAINFALL OCCURRED DURING EVENT.
OUANTV • AMOUNT OF RAINFALL DURING IhE EVENT IN INCHES.
RUNOFF
INCHES • RUNOFF DURING EVENT IN INCHES.
EMPTY • NUMBER OF HUUKS FRO* LAST HAINFALL T0 END OF EVENT.
DURTN • TOTAL NUMaE" OF HOOKS STORAGE "AS UTILIZED. IE, LFNGTM OF THE EVENT.
HAX » MAXIMUM AMOUNT OF STOHAuE UTILIZED. IN INCHES.
ST I 2u"fIl0SFEHUjlsSELA«EOl'8f>0»F!:oVEHFLOt. STARTED. OR, IF NO OVERFLOW, HOUR OF HAXIMUM STORAGE.
OU* • NU'ott* OF MUUKS IN "MIC" OveMfLOK OCCUNF.O.
• ASTE • QUANTITY OF riATEH RELEASED UNTKt.ATEO, IN INCHES.
INITL • QUANTITY OF «ATER RELEASED UNTREATED DURING THE FIRST 3 HOURS OF OVERFLOH,
H43 • Ntli-iJEK OF HOUHS »AT£i» -AS TREATED DURING THE PaESfNT EVENT AND SINCE THE PREVIOUS EVENT.
OANTT « QUANTITY OF *ATEH I-»EAlt:D DURINr. THE EVENT AND SINCE THE PREVIOUS EVENT.
AGEt - »v£
-------
10
\o
TABLE 3
Example of Average Quantity Statistics for all Events
I 3EL8Y STREET WATERSHED 9AN FRANCISCO* CA,
QUANTITY ANALYSIS
TREATMENT MATE • .1101 IN/MR, 37.8 crs. 24.4«a «ct> FED BLOC. s»». FHANCISCO, CA,
STORAGE CAPACITY* .1808 INCHES, 31.2 AC-FT. 11.168 MC SELBt STHftT
EVENT — -o A T E— HRS NO -R A i H F A i L- RUNOFF MRS TO --STORAGE— — — o v E R r L o w— - — IREAT-'ENI- — . — AGC OF
YEAR MO OY HR STORAC OURTM HRS OUANTY INCHES EMPTY OURTN MAX NO ST OUR WASTE INHL HRS OANIY AGt 1 *G£2 ACE3 ACE4 AUE9
• •••} •••••••••2 O •••««4 ••••S ***6 *****7 ««««7A ••••*B ••••» •««18 •!! *I2 «1J •••14 •••!& •••16 •••17 ••Id **I9 ••?!) ••?! "22
AVE OF 188 EVENTS !Bt.7**10.9 7.8 .47 .22 3.8 14,8 .06 2.7* 15. S .lb 3.1 7.7 12.6 3,9 4.3
AVE OF «0 OVRFLH EVENTS 19.8 13.1 .92 .43 9.7 29,9 .03* 5.3 5.4 .20 .10 26. li .25 t>.2 |3.4 23. b A.I 7.7
• NON-OVERFLOW EVENTS ONLY.
••EXCLUDING H DRY PERIODS
AVERAGE ANNUAL STATISTICS FOR S YEARS OF RECORD FOR THE PERIOD BEGINNING 641P27 AfcO ENDING 661228
NUMBER OF EVENTS • 33.6
NUMBER OF OVERFLOWS • 13.e-
INCHES
TOTAL PRECIPITATION ON WATERSHED 17.95
TOTAL RUNOFF FROM WATERSHED 7.68 FRACTION OP RAINFALL • .42
OVERFLOW TO RECEIVING MATER 2.78 FRACTION OF HAINFALL • .19, OF RUNOFF • .36
INITIAL OVERFLOW TO RECIEYING HATER 1.42 FRACTION OF RAINFALL • .BB, OF KUNOFF • .19
-------
TABLE 4
Example of Quality Statistics by Events
PACE
StLSY STHEET hAT£«SHEO SAN FKANC13CO, CA.
QUALITY ANALYSIS
VO
T»EAT«3 Bbft JHM7 474 47 4 .22 4598 458 490 237
• 64 11 It 8 ,HJ ,t!2 104 16 15 61 NO OYEKfLO*
7 64 11 It 12 ,3ft .18 1666 ?II4 206 89 9 NO OVEHFLOU
> 64 1| |2 7 ,11 .04 310 43 46 |7 2 NO OVEHFLOH
9 64 t| 12 1A .14 .07 bs7 7J 81 31 3 NO OVEHKLOH
1* 64 It 21 2H .24 .in 236U 155 6A3 162 16 NO OVERFLOW
TITLE DESCRIPTIONS
1 . EVENT = Sequence number
2. DATE = Date this event began
3. HR '= Number of hours past midnight this event began
4. RAINFALL = Total quantity of rainfall for this event in inches
5. QANTY = Total quantity of runoff for this event in inches
6. SUSP = Suspended solids ^v
7. SETL = Settleable solids 1
8. BOD = BOD ) Tctal tbs in ^onoff for this event
9. N = Total nitropen j
10. P04 = Orthophosphate
1 1 . NO = Overflow event sequencing number
12. QANTY = Total Quantity of overflow for this event in inches
13. SUS =
14. SET =• 1
15. BOD = \Total Ibs. in overflow for this event
16. NIT J
17. P04 = '
18. 0 = Quantity of overflow during first three hours in inches
19. SUSP = >.
20. SETL = 1
21. BOD = /Total Ibs. in overflow during first three hours
22. N =1
23. P04 - '
FED BLDG, SAN FHANCI3CO, CA.
SEL8Y STHEET
INCH TOTAL POUNDS
P04 * SUSP SETL BOO N P04
144 .07 4248 129 922 264 26
7 ,B6 1260 94 157 67 7
9 .05 1073 79 154 59 6
24 .29 4362 429 463 225 22
-------
TABLE 5
Example of Average Quality Statistics for all Events
PAGE
vo
00
RATE •
STORAGE CAPACITY*
SELBV STREET WATERSHED SAN FRANCISCO,
QUALITY ANALYSIS
.ottin IN/HH. . 37.0 CFS, 24.4ii« HCD
INCHES, 3t.2 AC-FI, ifl.tea HG
CA.
FED BLOC, SAN FRANCISCO, CA.
SEL8Y STREET
EVENT DATE RAIN • S TOHH RL'NOF F—— — -S TORA6E 0 V E R F L 0 N—' FIRST 3 HOURS OVERFLOW——
EVENT BEGAN FHL INCHS TOTAL POUNDS SEQ INCHS TOTAL POUNDS INCH TOTAL POUNDS
YR HO 0» HR IMCHS UANTY SUSP SE1L BOO N P04 NO OANTY SUSP SETL BOO N P04 • SUSP SETL BOO N P0<
••••I •••••••2 «3 ****4 •••«5 ••••»6 ••««7 ««**B ••••9 **t0 Ml «*M2 «««*13 •••H *««15 •••t6 ««J7 «*|8 *«*19 ««2H *«21 **22 ««23
AVE OF 168 EVMS .47 ' .22
-------
TABLE 6
Example of Hourly Quality Statistics
to
vo
vo
,8188 IN/MR,
INCHES."
TREATMENT HATE
•Toiucc
EVENT Y» HO OV MR T(0i RAINFALL *\lnQf?
153
153
153
153
153
153
153
153
151
t53
153
IS J
153
153
153
153
153
68
68
68
68
68
68
68
68
68
68
68
68
68
68
68
68
68
11
11
1 1
11
11
It
It
It
It
11
It
It
It
It
tt
It
U
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
6
7
6
9
10
17
IB
20
21
?2
23
20
1.
2
3
•
3
1
2
3
a
5
12
11
15
16
17
18
I*
20
21
22
23
28
,13
.08
.05
«"
.00
.02
,na
.«!
.02
.01
.02
!o8
.08
.03
.07
*07
.0"
^01
.01
.01
.01
.02
.01
.01
.01
.01
.02
.00
,00
.02
.oa
er».
POLLUT06RARW Ft)* CVtNT
2«,«eo MCA
ia.t»8 MO
CA.
SUSP 3ETL
7177,1
•3,3
200,0
050,0
202,3
i083l6
2800.6
lsa.6
677,1
POLLUTANT LOAD. IV IBS/MR**
3l«a,Z 103.0 529.i
58,6 256,8
uz
BOO
861 TO 231^3
1251,5 «1«,5
S2t6
82.7
70,0
88,6
S6,2
58,5
109,5
3*.6,0
2<0?0 110?*
P0«
•1,3
'1*2
2*7
5<*
2.5
FED HOG. SAN FRANCISCO. CA.'
SELBV STRECt
»* AVE CDNCINT»ATION'; IN KO/L •»••••
SUSP SETL 810 N P0«
11.1
127,8
63,0
07,1
ll'l
52*1
62.7
52,2
71.0
2,6
|\*
2*6
22,3
15,J
11,9
Bil
10tO
2J7
6.69
7.31
5.25
3.20
3.0»
J.51
2.69
2.90
2.63
2.*»
«!s2
2>l
3.66
.70
.52
!?»
,32
^33
.27
?26
!3«
!«i
.27
.37
NOTE:
T(0)
RAINFALL
RUNOFF
CFS-OFF
Hours since start of storm
Inches of rainfall during indicated hour
Inches of runoff during indicated hour
Average runoff rate in cfs during indicated hour
-------
TABLE 7
San Francisco Hyetograph Storage/Treatment Analysis
Treat,
merit
.02
.04
.06
.03
•10
• 1 w
2
Storaoe
Capacity
.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
.25
.50
.75
1.00
1.50
2.00
2.50
3.00
.25
.50
.75
1.00
1.50
2.00
.10
.25
.50
.75
1.00
1.50
2.00.
.10
.25
.50
.75
1.00
1.50
USWB
Event/
yr.
38.806
36.661
35.726
35.435
35.194
35.113
35.048
35.048
42.355
40.226
39.435
39.113
38.903
38.871
38.790
38.790
37.774
36.855
36.468
36.339
36.290
36.274
33.016
31.952
31.419
31.274
31.194
31.145
31.145
27.484
26.855
26.484
26.435
26.403
26.387
62 Yr.
Ovrflw/
yr.
7.710
2.823
1.242
.597
.387
.242
.097
.032
11.726
5.435
2.823
1.419
.403
.177
.113
.016
9.323
3.645
1.500
.661
.177
.113
15.242
7.194
2.371
.903
.371
.113
.016
12.484
5.403
1.597
.581
.177
.032
Rec.
Quart/
yr.
4.254
1.907
.926
.504
.265
.107
.033
.003
4.115
2.101
1.125
.624
.231
.107
.031
.002
2.611
1.134
.524
.270
.095
.012
3.247
1.694
.633
.264
.121
.025
.001
2.363
1.121
.369
.128
.050
.007
treatment Rate In Inches per hour.
'Storage Capacity 1n inches.
300
-------
4.0
I I
62 YEAR USWB RECORD
CITY OF SAN FRANCISCO
6E YEAR USWB RECORD
CITY OF SAN FRANCISCO
.03 .04 .06 .08 .10
TREATMENT RATE - INCHES/HOUR
Figure 6. Overflow Frequencies.
TREATMENT RATE - INCHES/HOUR
Figure 7. Overflow Volumes.
62 YEAR USWB RECORD
CITY OF SAN FRANCISCO
I.O 20 10
STORAGE CAPACITY - INCHES
FIGURE 8. OVERFLOW OCCURRENCE RELATIONSHIPS.
301
-------
scheme and exceeds by a factor of 2 the volume requirement of an all-
treatment scheme.
The data presented above was necessarily based on two assumptions: that
the runoff loss is 35 percent and that rainfall occurrence is uniform
over the City. Each is a significant parameter in determining the total
volumes of runoff. At the time the Master Plan was being developed no
verified data existed on the losses experienced in the rainfall-runoff
process, although some measurements have been made in more recent charac-
terization studies.
The initial sizing of the Master Plan system was based upon the records
available from the Weather Service gage. This data represented the best
information available at the tine and all other data indicated that any
design based upon this gage would likely to conservative with regard to
size and costs. Refinement of the design will take place as data accum-
ulates from the City's-extensive field information collection system.
Figure 8 shows a composite of the effects of various combinations of
storage and treatment with regard to the frequency of uncontrolled over-
flow occurrences. It is apparent that, given a desired frequency of over-
flow occurrence, increasing the treatment rate decreases the storage
requirements and that, for any given storage volume, increasing treatment
capacity results in a lower occurrence frequency. Further inspection of
the figure also indicates that the law of diminishing returns results in
increasingly greater storage requirements for any treatment rate to
attain a lower frequency of overflow occurrence. Through the application
of cost factors for storage and treatment facilities, optimum design
points for minimum cost for various levels of control can be derived.
East Bay Municipal Utility District
The East Bay Municipal Utility District, Special District No. 1 (EBMUD,
SD1), provides wastewater disposal services to seven cities on the east
side of San Francisco Bay, California. Figure 9 shows the location of
the service area. The EBMUD systems is comprised of a 22 mile inter-
ceptor system plus a central treatment plant.
Wet weather inflow/infiltration is a major problem for EBMUD. The source
of the problem is the wastewater sewer systems that drain into the inter-
ceptor system. Most of these systems are over 30 years old. A number of
them have been converted from combined sewers to "sanitary" sewers, and
a few combined sewers are still connected to the interceptor system. In
addition, there are many illegal connections (especially to roof and yard
drains). Present estimates are that 11 percent of the rainfall on the
service area appears as inflow/infiltration in the interceptor system
causing the system to overflow to San Francisco Bay about 11 times a year.
Reduction of the overflows could be accomplished by reducing extraneous
inflows, by providing additional treatment and storage, or through some
combination of both. The question was, which particular combination
would give the control required and which combination would be the most
cost-effective.
302
-------
CERRITOsw,
s^.ja
ALBANY
RKEL_
PIEDMONT
EBMUD SPECIAL
DISTRICT No.
Figure 9. Vicinity Map of EBMUD Service Area.
303
-------
It was estimated by the District staff that conversion of the remaining
combined sewers to sanitary sewers would reduce the gross inflow/infil-
tration ratio (i.e., runoff coefficient) from 11 percent to 8 percent. If
80 percent of the direct connections to the sewer system could be elimi-
nated (roof and yard drains, parking lots, catch basins connected in error,
etc.) it was estimated that the inflow/infiltration ratio could be reduced
another 3 percent, i.e., from 8 percent to 5 percent. Finally, by the
additional removal of 50 percent of the percolation infiltration, it was
estimated that the inflow/infiltration ratio could be reduced another 2
percent, i.e., from 5 percent to 3 percent.
For purposes of determining what combinations of treatment rate and stor-
age capacity would meet the system requirements, the STORM Quantity Anal-
ysis portion was used to process 22% years of hourly rainfall data recorded
at the U. S, Weather Service station at Oakland International Airport.
An initial run was made with the STORM model set at existing Special Dis-
trict No. 1 treatment and storage capacities, assumed to be 0.0068 inches/
hour and 0.017 inches, respectively, in excess of average dry weather flow
requirements. For this run, an existing District-wide gross infiltration
rate of 11.1 percent of rain, developed in an earlier study [6], was used.
This infiltration rate includes the effect of combined sewers which drain
approximately 4 percent of the total area. The run showed an average
incidence of 10.9 overflows per year, which is in good agreement with
historical data.
All subsequent runs of STORM were based on the assumption that all combined
sewers were separated from the sanitary system. Treatment-storage combin-
ations were examined for three alternative infiltration rates (expressed
in percent of Oakland Airport rainfall):
1. 8% - Gross Infiltration Ratio without combined sewers
2. 5% - Removal of combined sewers plus 80% of "direct
connections"
3. 3% - Additional reduction by removal of 50% of
"percolation infiltration".
In order to provide enough points for curve plotting, a total of 49 infil-
tration-treatment-storage combinations were run. Treatment rates ranging
from 0.003 to 0.03 inches per hour (107.5 to 1075 MGD) and storage capac-
ities ranging from 0.001 inches to 0.08 inches (1.5 to 120 million gallons)
were analyzed.
For each of the three infiltration rates, the average number of overflows
per year are shown graphically in Figure 10. As expected, the number of
overflows decreases as the storage capacity or treatment rate is increased
or as Infiltration is reduced by upstream extraneous control measures, as
represented by the reduction in infiltration rates. Storage required to
totally contain the inflow from the 22 year period of rainfall record is
shown in Figure 11.
304
-------
tOO 4OO COO .100 OOO I20O
o
Ul
OM
OM
EBUUD SO Ha I
OVERFIO*
FRBOXNCr CURVES
iurt'
TKCATUCHT tare - Ufa
e aoo 400 coo aoo woo BOO
•O 004
i
»i
\ s
! ftfla
EBUUO SO Ha I
WEKFLO0
FREQUENCY OJM£S
10*1
- uta
no «eo wo MO noo
ttOJ
i
fc
EBUUD SO fit 1
FRBOVENCf CURVES
imflLTIUTlOH lUTC-3.0*1
on outa
moaueur *at - jMM0XMt
Figure 10. Overflows at Various Infiltration Rates.
-------
TREATMENT RATE-MGD
200 400 600 800
-J-T 1 1 K-
EBMUD SD No. 1
STORAGE REQUIRED
FOR NO WET
WEATHER OVERFLOWS
jExisfing Treatment Capacity
in Excess of Dry Weather How
1000
—\ 1 600
;
400
Existing Storage
Capacity in Excess
of Dry Weather Flow
300 I
k.
I
3
200 &
• 100
1
0.01 0.02
TREATMENT RATE - INCHES/HR.
0.03
Figure 11. Zero Overflows Requirements.
306
-------
APPLICATION OF "STORM" - EXAMPLES
The purpose of this section is to present some examples of other possible
applictions of STORM in addition to those already described. Four appli-
cations will be shown:
1. Computation of the quantity of storm runoff by month
and for single storm events.
2^ Computation of pollutographs for single storm events.
Use of STORM to find the most economical treatment-
storage combinations that meet system overflow
constraints.
4. Analysis of changes in the quantity and quality of
urban runoff due to alternative land use management
schemes.
A prototype area was selected for these applications of STORM, i.e., the
Castro Valley watershed near Oakland, California. Figure 12 shows the
USGS map of the area with the watershed boundary and the location of the
watershed relative to the San Francisco Bay area. Some land use inform-
ation can be obtained from the map but an aerial photograph plus ground
reconnaissance were necessary to obtain a better understanding of land
uses in the basin. Table 8 gives a summary of the estimated hydrogeo-
metric characteristics of the watershed. Runoff statistics for the water-
shed were generated by processing hourly rainfall data for the 17-month
period from November 1971 through March 1973.
COMPUTATION OF THE QUANTITY OF STORM RUNOFF
In the initial application of STORM to Castro Valley, the program was
calibrated to give the best comparison between computed and observed
values of:
1. Average annual precipitation; ;
2. Average annual runoff;
3. Monthly runoff volumes; and
4. Individual storm event volumes.
It is not appropriate to make comparisons with the instantaneous measure-
ments of discharge because the program computes runoff as hourly volumes
only. The hourly volumes should reflect the general shape of the observed
hydrograph and its volume, although the observedhydrograph will tend to
lag behind the computed hydrograph in larger basins where the time of
concentration is greater than one hour. The program can be applied to
larger basins where this lag problem exists as long as one realizes the
impact on the analysis. In storage analyses, however, this problem is
generally not critical.
307
-------
FIGURE 12. CASTO VALLEY WATERSHED,
308
-------
TABLE 8
Hydrogeometrlc Data
Land Use
Single Family Residential
Multiple Family Residential
Commercial
Open or Park
Percent
of Area
Percent
Impervious
70
3
7
20
40
50
80
2
Length of
Street Gutters
275(£t/ac)
430
400
20
Area of the watershed = 3136 acres
Degression Storage = . 10 inch
Depression Storage Recovery, inches/day
January
February
March
April
May
June
0.05
0.07
0. 12
0.17
0.23
0.26
July
August
September
October
November
December
0.28
0.25
0. 20
0. 13
0.07
0.05
Runoff Coefficient for Pervious areas = 0.45
Runoff Coefficient for Impervious area = 0.90
Area weighted basin average runoff coefficient = 0. 61
The program parameters which were "tuned" in the calibration process were:
2.
Rainfall factor relating basin average rainfall to the
gage rainfall;
Depression storage and the rate of recovery of
depression storage;
3. Imperviousness of the land uses; and
4. Pervious and impervious area runoff coefficients.
The recording rain gage for Castro Valley is centrally located in the
watershed and was assumed to reflect basin average precipitation. Cali-
brated values of the other parameters are those listed in Table 9.
Table 9 shows the computed and observed data for monthly runoff and for
total runoff over the 17-month period. Computed average annual rainfall
and runoff information is obtained directly from the program's average
annual summary. Monthly volumes were computed from the EVENT or POLLUTO-
GRAPH output. Figures 13 and 14 show the computed and observed hydrograph
fit for two events in the 13-month record.
309
-------
S50
CASTRO VALLEY
SCO
450
u
-
=
Figure 13. Comparison of Observed Hourly Values of
Values of Runoff With Values Computed
by "STORM" November 11, 1972.
Figure 14. Comparison of Observed Hourly
Values of Runoff With Values
Computed by "STORM"
February 6, 1973.
-------
TABLE 9
Comparison of Monthly Runoff Volumes
From Castro Valley Computed by Runoff Model
Year
1971
1972
1973
Month
11
12
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
Actual Runoff
(Basin-In)*
.35
1.50
.51
.60
. 17***
.34
. 10***
.17
. 11***
. 08***
.21
1.01
2.85
.91
5.78
4.34
2.01
=21.00
STORM**. In
.73
2.29
.58
.58
0
.44
0
.16
0
0
.48
1.51
2.82
1.03
4.13
2.45
1.71
18.90
•1-1 . 70 baseflow
20. 60
*From USGS Records.
**Does not include baseflow.
0.10 inches per month.
***No rainfall recorded.
Baseflow is approximately
COMPUTATION OF THE QUALITY OF RUNOFF
The zero storage, zero treatment combination was also used for an initial
calibration of the quality portion of STORM. The quality calibration was
made difficult because of the lack of adequate data. It is usually not
economically feasible to monitor every runoff event, thus monthly or
average annual data generally does not exist. In most instances, as was
the case for tlxe Castro Valley data, the quality calibration must be made
on data from several individual runoff events. However, unless care is
taken to obtain a good sampling over the duration of the entire event,
the comparison of total amounts of pollutant washoff will not -be possible.
This was also the case in Castro Valley; consequently, most of the cali-
bration had to rely on a few data points per event. It would be highly
desirable to have enough measurements during a runoff event to be able to
trace the hourly performance of the pollutant waahoff function.
The quality calibration for Castro Valley was made on the basis of com-
parisons of computed and observed concentrations for individual events.
The parameters calibrated were:
1. Dust and dirt accumulation rates;
2. Pollutant composition of the dust and dirt; and
3. The exponent in the pollutant washoff equation.
311
-------
The Initial pollutant loading rates used in the calibration were those
from the Chicago APWA study [3]. These data are programmed as default
values to be used if other values for these parameters are not specified
as input data. Table 10 shows the calibrated values of the pollutant
loading and washoff parameters. A comparison of these values with the
default values listed earlier reveals that the dust and dirt accumulation
rates had to be increased by a factor of two except for the open or park
area which was increased by a factor of about six. Pollutant composition
of the dust and dirt was also increased by a factor of four for all
parameters except suspended and settleable solids.
TABLE 10
Pollutant Loading and Washoff Parameters
Pounds Pollutant
Dust and Dirt
Accumulation
Land Use lbs/day/100 ft. gutter
Single Family 1 .4
Multiple Family 4.6
Commercial 6. 6
Open or Park 9. 2
Street Sweeping Efficiency
Washoff exponent = 2. 0
per 100
Susp.
Solids
11.1
8.0
17.0
11.1
= 70%
Ibs Dust
Sett.
Solids
1. 1
.8
1.7
1. 1
and Dirt
BOD
Z.O
1.44 .
3.08 .
2.0
N
19
24
16
19
P04
.02
.02
.03
.02
Comparisons of computed and measured BOD concentrations are shown in Figure
15. Agreement in the first case is very good, however, computed values
tor the second storm are consistently low. Notice that the first storm
occurred in November of 1972 which is the beginning of the winter rainy
season. The good agreement in this case indicates that the accumulation
of pollutant loads over the dry summer period is being computed well by
STORM.
The second comparison shown was for a storm in February which is near the
end of the winter. Since the computed values are consistently low, the
implication is that the pollutant load on the watershed at the beginning
of the storm was too small. This is quite likely the case because during
the period between the November storm and the February storm, the computer
program program has been performing a constant accounting of the pollutant
load on the watershed, i.e., how much was there at the beginning of each
storm, how much was left at the end of each storm, and how much accumu-
lated between storms. If we assume, on the basis of Figure 15 (the
November storm), that the pollutant load at the beginning of the rainy
season Is correct and that the computation of washoff rate is correct,
then the rate of pollutant accumulation (dust and dirt accumulation) on
the watershed must be larger for the prototype than the rate being used
In the model. As a result, the watershed is being washed overly clean
by STORM during the rainy season.
The next step in the calibration procedure, therefore, would be to
Increase the dally rate of dust and dirt accumulation during the winter
period which would result In larger pollutant loads at the end of the
rainy season.
312
-------
CASTRO VALLEY
.
r a 9
MOtfff Of DAY
10 II
12
13
CASTRO VALLEY
60
r
' 20
0
•
*•••».
^OBSi
^^-COMP
•O» — »
RVED
OTEO
•— »s
^>D
— .0- —
— -KI
20 21 • f 10 II 12 13 14
HOUR Of DAY
tl/5/73) (f/6/731
Figure 15. Comparison of Observed Hourly Values of BOD
Concentration With Values Computed by "STORM"
November 11, 1972 and February 6, 1973.
313
-------
STORAGE-TREATMENT ANALYSIS EXAMPLE
The primary purpose of the STORM program is to analyze the effectiveness
of storage and treatment facilities for use in controlling the quantity
and quality of urban storm runoff. The criteria for control of the storm
runoff may be in terms of maximum allowable:
1. Overflow events per year;
2. Volume of overflow per year;
3. Volume of overflow during some design storm event;
4. Pounds of BOD (or other pollutant) overflow per
year; and/or
5. Pounds of BOD (or other pollutant) overflow during
some design storm event.
It may be possible to achieve the desired control of runoff through a
number of different combinations of storage and treatment. A cost analy-
sis is then necessary to determine which storage-treatment alternative
achieves the desired control at least cost.
The decision criteria for the purposes of this hypothetical example are
as follows:
1. No more than five overflows per year; and
2. No more than 10,000 pounds of BOD overflow
per year.
In order to determine what storage-treatment rate combinations will meet
these objectives, the STORM program was run for several different storages
for each of several treatment rates. For each storage-treatment combi-
nation, information on the average annual number of overflows, inches of
overflow, and pounds of BOD (or other pollutant) was obtained directly
from the output (see Tables 8 and 10 for example).
This information can be plotted, as was done in Figure 6, to enhance the
analysis. For this problem, however, it would be better to first plot
Number of Overflows/year vs. Storage Capacity and Pounds of BOD Overflow/
Year vs. Storage Capacity, both plots showing lines of equal treatment
rates. The two plots are shown in Figure 16. Notice on the BOD overflow
curves that the overflow frequency is also shown. It can be seen from
this figure that overflow frequency is a more severe requirement for the
system if its storage capacity is less than 0.3 inches. For greater
storage capacities, however, the quality constraint is more limiting.
Thus all treatment-storage combinations shown below the shaded line in
Figure 17 are acceptable from the standpoint of meeting or exceeding the
system performance criteria. However, it can be seen that for a given
treatment rate, the smallest allowable storage will be that identified
on the shaded performance line in Figure 17. Thus, in our example the
most economical control system that meets the performance criteria will
be one of the following three:
1. Treatment = 0.01 in/hr Storage - 0.46 in
2. Treatment - 0.03 in/hr Storage - 0.28 in
3. Treatment - 0.05 in/hr Storage - 0.20 in
314
-------
40
8 1.2
STORAGE CAPACtrr
* !
TREATMENT RATE:
.01 in/hr
.03 in/hr
.05 in/hr
Dojried lines ^ow number
of overflow* per year
Mos1 economicol treatment-storage
combination!
.8 1.2
STORAGE CAPACITY
fine***}
2.0
Figure 17. Most Economical Treatment-Storage
Combinations To Meet System
Overflow Constraints.
STORAGE CAPACITY
(incfut)
Figure 16. Relationship of Overflow Frequency and
Pollutant Overflows to Treatment-Storage
Alternatives Castro Valley.
-------
Economic Analysis of these three alternatives would identify the most
economical system.
LAND USE MANAGEMENT ANALYSIS
Another possible application of STORM is to estimate the impact of proposed
land use changes within a watershed on the quantity and quality of urban
runoff. To illustrate this use, it was assumed that the upper arm of the
Castro Valley watershed (see Figure 12) would be developed. This 470 acre
area is presently open space. Under the assumed development plan, single
family residential dwellings would be constructed on 67 percent of the
area. The remaining 33 percent of the area would be developed as multiple
family residences.
Table 11 shows the results of applying STORM to the 470 acre subarea in
both its present undeveloped state and in the proposed developed state.
In the developed state, the annual quantity of runoff can be expected to
increase by 40 percent. The annual BOD, on the other hand will increase
more than 400 percent.
The impact of the proposed development on the watershed as a whole is
shown in Table 12. As expected, this impact is significantly less. As
a result of the proposed development, annual storm runoff from the water-
shed can be expected to increase by approximately 5 percent. The annual
BOD load washed off the watershed will increase approximately 14 percent.
TABLE 11
Effect of Changing Land Use on
Storm Runoff from 470 Acre Subarea
Annual Runoff
Inches
8.34
11.72
Land Use
Existing
100% open
Proposed
67% single family residential
33% multiple family residential
Annual BOD Load
Pounds
5,700
23,400
316
-------
TABLE 12
Effect of Changing Land Use of 470 Acre Subarea on Storm Runoff
From Entire Caatro Valley Watershed (3140 Acres)
Annual Runoff Annual BOD Load
Land Use Inches Pounds
Existing 11.08 130,500
70% single family residential
3% multiple family residential
7% commercial
20% open
Proposed 11.59 148,300
80% single family residential
8% multiple family residential
7% commercial
5% open
317
-------
REFERENCES
1. "Urban Storm Water Runoff 'STORM'." Generalized Computer Program,
723-S8-L2520, Hydrologic Engineering Center, U.S.Army Corps of
Engineers (May 1974).
2. "Water Pollution Aspects of Urban Runoff." American Public Works
Association, Water Pollution Control Research Series, Federal Water
Pollution Control Administration, Report No. WP-20-15 (January 1969).
3. "Stormwater Management Model, Vol, I, Final Report," 11024 DOC 07/71,
Metcalf and Eddy, Inc., University of Florida, Gainesville, and Water
Resources Engineers, Inc., Report for EPA (July 1971),
4. Roesner, L. A., D. F. Kibler, and J, R. Monser, "Use of Storm Drain-
age Models in Urban Planning." Presented at the AWRA Symposium on
Watersheds in Transition, Colorado State University, Fort Collins,
Colorado, June 1972, Proceedings, AWRA, Urbana, Illinois, pp. 400-405
(1973).
5. "San Francisco Master Plan for Waste Water Management," Department of
Public Works, City and County of San Francisco (September 15, 1971).
6. Metcalf & Eddy, Inc., "Storm Water Problems and Control in Sanitary
Sewers, Oakland and Berkeley, California." 11024EQ603/71, Report
for EPA (March 1971).
318
-------
COLLECTION OF FIELD DATA FOR STORMWATER MODEL CALIBRATION
By
Philip E. Shelley, PhD*
INTRODUCTION
This lecture on the collection of field data for storrawater management
model calibration and verification is intended to stress the importance
of data gathering procedures and techniques for data products that
are to be used in conjunction with such models and to summarize the
current state of the art with respect of wastewater sampling and flow
measurement. It is assumed that the reader has a baccalaureate degree
in one of the basic engineering disciplines (or the equivalent), an
overall appreciation for urban hydrology and general stormwater char-
acteristics, and at least a limited amount of field experience.
BACKGROUND
Hydrology today is in a period of transition due to the advent of the
digital computer. This tool has made possible the development and use
of complex mathematical models during the 1960s. Where once the
"design storm" concept was the basic approach to design, near con-
tinuous simulation leading to estimates of flow probabilities has now
become possible. The literature probably contains references to over
200 stormwater-related models. Very few of these, however, have been
specially designed for urban areas. Most of these models can be clas-
sified into a few basic types, and many differ little from conventional
hydrologic techniques of the past few decades.
Director of Energy and Environmental Systems, EG & G Washington
Analytical Services Center, Incorporated, Rockville, Maryland.
319
-------
Precipitation Models
Although precipitation modeling and measurement lie outside the bounds
of the present lecture, a few comments are appropriate for the sake
of completeness since the more sophisticated urban stormwater models
need a detailed knowledge of time and space variability of precipita-
tion as inputs. For example, during the progress of a storm, the
spatial distribution of precipitation for consecutive, specified time
intervals (e.g., five minutes precipitation accumulation) might be
required continuously on a real time basis. The procedure for com-
puting the spatial distribution of a variable over some region from
data at a limited number of sites is often referred to as objective
analysis. Bergthorssen and Doos (1955) present the development of a
typical objective analysis technique. The accuracy of any such analy-
sis depends upon the variability of the analysis parameter and upon
the density of observations.
Many researchers have used dense raingage networks to Inquire directly
of nature for particular regions. The time distribution character-
istics of rainfall rates were investigated by Huff (197 Ob) who used
data from a 60-storm sample on two dense networks in Illinois. He
found variability parameters to fit closely a log normal distribution.
Both absolute and relative variability showed a wide range within and
between storms and between areas of different size. Little difference
in variability properties was noted between rain and synoptic weather
types associated most frequently with warm-season storms. In a study
of the spatial distribution of rainfall rates, Huff (1970a) found it
to be so great within and between convective storms that the equipment
and operational demands for accurate rate measurements may be prohibi-
tive for areas over 16 km square. In related work, Huff and Shipp
(1969) found that a gage spacing of 0.3 minutes is needed for
1-minute rain rates compared with 7.5 minutes for total storm rainfall,
a 625 to 1 increase.
320
-------
Stormwater Models
Given the foregoing, it is no surprise that a number of different meth-
ods or models for the calculation of stormwater runoff from rainfall
have arisen. However, there seem to be only a few basically different
models that appear in the literature. The many so-called methods that
are presented usually consist only of different procedures for evaluat-
ing the key parameters of one of the basic models, different computa-
tional procedures fo* solving similar equations, or sometimes only a
name. Linsley (1971) has suggested that there are four basic techniques
for calculating or estimating flows in a watershed; empirical formulae,
statistical correlation with watershed characteristics (regression
models), frequency analysis of streamflow, or computation from precipi-
tation data (hydrolic synthesis). He admits, however, that the bound-
aries of the four can become quite blurred. A brief discussion of
each technique follows.
Empirical techniques are in common use today, the most popular of which
is the so-called rational method first put forth by Mulvaney (1851) and
introduced in the United States by Kuichling (1889). Chow (1962) has
given a good bibliography, and the method was reviewed recently by
McPherson (1969). The real issue of how long must the rainfall con-
tinue for equilibrium to be reached so that the assumption holds that
peak outflow will be equal to inflow rate (rainfall intensity) is not
answered by the rational method. The best approach is to make use of
information that is now available on overland flow. The technique of
Izzard (1946) appears fairly satisfactory, although several other
methods have been suggested in the more recent literature, e.g., Grace
and Eagleson (1966) and Woolhiser and Liggett (1967).
Regression techniques seek to relate a causal factor (precipitation
and/or watershed characteristics) with an effect (peak flow, storm
*
runoff volume, etc.) by statistical correlation. One could view them
merely as extensions of empirical techniques based upon the existence
321
-------
of more data and more sophisticated methods of analysis. Potter (1961)
presents a typical example. Care must be exercised in such techniques,
and Freeze (1972) notes, "This evidence for a wide range of watershed
response functions leads to the development of a healthy skepticism
toward black-box rainfall-runoff correlations, the concept of basin
linearity, and the rationality of hydrograph separation."
Frequency analysis techniques can be used to generate probabilities
of peak flow statistically if enough prior data are available and there
have been no significant changes in characteristics over the data per-
iod. Methods of transposition to broaden the use of such techniques
have been advanced, see Anderson (1970) for example. Dalrymple (1960)
has presented an adaptation of regional frequency analysis.
Hydrologic synthesis techniques employ two basically separable rela-
tions; a rainfall-runoff relationship to estimate the volume of storm-
water runoff (rainfall excess), and a procedure for determining the
shape of the flow hydrograph having the estimated runoff volume.
Rainfall-runoff relations can be broken down into three basic types -
regression relations, infiltration indices, and water balance models.
l
Runoff that reaches the receiving water has, in general, three compo-
nents that follow different paths; surface (sheet or overland) flow,
interface (through the upper soil layer) flow, and base (groundwater)
flow. For urban drainage basins only surface flow is relevant in most
instances. Regression techniques range from a simple plot of rainfall
vs runoff to fairly complicated multivariable procedures. Kohler and
Linsley (1951) proposed a coaxial correlation technique that has been
used by the National Weather Service in flood forecasting. Cook (1946)
introduced the use of infiltration indices which can be viewed as
representing an average loss rate. These can be determined from se-
lected storm hydrographs and rainfall data. The water balance method
maintains a running account of the water in soil moisture storage by
adding each new rainfall, less direct runoff and accretion to ground-
water, and subtracting evapotranspiration. Thus, the amount of runoff
322
-------
and groundwater accretion are made functions of the prevailing soil
moisture storage. This technique has been used in the Stanford
Watershed Model as reported by Crawford and Linsley (1966).
Linear hydrologic synthesis has been popular since the unit hydrograph
was first proposed by Sherman (1932). The black-box technique of
deriving unit hydrographs from input and output data while ignoring
physical basin processes is not uncommon and has been approached in
many ways. Snyder (1938) used a simple least squares fit, O'Donnell
(1960) used a Fourier expansion, and Dooge (1959) employed Laguerre
functions. All such approaches seek to find the optimal shape of the
response to a unit impulse of rainfall; the difference merely lies in
the method of doing this.
As was noted earlier, Freeze (1972) has questioned the assumption of
basin linearity, and there is no real dearth of nonlinear approaches.
For example, Amorocho and Brandsetter (1971) investigated the determina-
tion of nonlinear response functions in the rainfall-runoff process,
and Bidwell (1971) performed a regression analysis of nonlinear catch-
ment systems.
TYPES OF FIELD DATA
Regardless of whether the model of interest is stochastic versus deter-
ministic, on the one hand, or analytic versus synthetic, on the other,
there are two uses that are generally made of field data, and to drive
home the importance of distinguishing between the two, they will be
referred to here as data types. Because of the complexity of the proc-
esses being modeled, most of the models that are popular today require
field data both for estimating empirical parameters in their structure
and for fitting other application-specific parameters (calibration).
For example, one version of the Stanford Watershed Model has twenty
parameters, two based on meteorological data, four on hydrograph separa-
tion, five are computed from physical measures, three are estimated
323
-------
from empirical tables, and six are fitted. All field data that are
used within the model structure, i.e., to ready the model for specific
application, will be referred to here as calibration data.
The chief concern of the model user is how well the model outputs
(which are_its sole reason for being) compare to reality, this compari-
son forming a measure of the predictive capability of the model. Here,
also, input and output field data are required, but their fundamental
use is quite different from that of calibration data. They will be
referred to here as verification data, since they are used to verify the
results of a particular model exercise.
This distinction between field data types is not made to suggest that
different gathering techniques are required for calibration versus
verification data; in point of fact they are the same. The reason for
making the distinction is simply that calibration data must never
(repeat never) be used for model verification. The importance of this
simplistic statement cannot be overestimated. Having made this point,
the matter will be dropped, with the proviso that everything said about
the collection of calibration data applies equally to verification data.
FACTORS INVOLVED IN WASTEWATER CHARACTERIZATION
The various stonnwater management models that are available today
require data on the catchment, precipitation, and runoff quantity and
quality. They vary widely, however, in the detail of temporal and spa-
tial distribution of data required. For example, some models require
time steps of less than one minute to satisfy numerical stability condi-
tions, while others can be run with hourly, daily, and even up to semi-
monthly data.
Considerations other than the model structure are also involved. Both
the quantity and quality of stormwater runoff are highly variable and
transient in nature, being dependent upon meteorological and climato-
324
-------
logical factors, topography, hydraulic characteristics of the surface
and subsurface conduits, the nature of the antecedent period, and the
land use activities and housekeeping practices employed. It is this
highly variable and transient nature of stormwater flows that makes
their characterization so difficult. In addition to tremendous dynamic
ranges, the poor quality of stormwater draining from the urban environ-
ment has a significant effect on the choice of suitable sampling and
flow measurement equipment and methods as well as an impact in the
analytical laboratory.
The three factors involved in stormwater characterization efforts and
treated in this lecture are flow measurement, sampling and laboratory
analyses of the samples taken. Sections II,. Ill, and IV are devoted to
a discussion of each of these factors in turn, and we will limit our-
selves here to noting that although they represent three different
efforts requiring different types of skills and training, they are also
interdependent tasks and must be considered together in the establish-
ment of a successful stormwater characterization program.
THE NEED FOR TIME-SYNCHRONOUS DATA
Measurements of stormwater characteristics are useful only with respect
to their relationship with each other and with other phenomena. An
assignment of time of occurrence to a data product makes possible a
determination of its relationship to other parameters whose times of
occurrence are known. The other parameters of interest may or may not
be synchronous with the particular data. In some cases, definition of
the time interval between events is sufficient but, generally, the true
clock time of the value of the concerned parameter is required.
The required accuracy of the time element in stormwater characteriza-
tion data is very different from requirement to requirement. An ex-
ample is the use of peak flows for each year to determine their
frequency. On the other hand, flow and quality data at intervals as
325
-------
short as one minute may be needed to define the discharge hydrograph
from small urban areas.
A particular need for attention to the time element occurs in the meas-
urement of flows from small urban storm sewers in order to define the
hydrograph and to provide data for the development and verification of
rainfall-runoff-quality models. Accurate definition of both the time
and discharge elements of the hydrograph makes possible the computation
of total volume of runoff during the storm by computing the area under
the hydrograph, exclusive of base flow. By selecting a number of we.ll
defined hydrographs resulting from storms of similar rainfall character-
istics, a typical hydrograph for the basin may be defined. In combina-
tion with time-synchronous water quality data, pollutographs can be
generated. Shape of the unit hydrograph is determined by accurate
timing as well as by discharge, although it is independent of clock
time. The hydrograph as defined by clock time and discharge is often
used to route flows along a stream channel or through a reservoir.
Peak flows, storm runoff volumes, daily flows, or other flow parameters
are often correlated with similar flows at other points on a storm
sewer or stream, or with flows of other storm sewers or streams, to
provide a means for flow estimation. Also, correlations may be made
with various physical characteristics of a basin, such as area, slope,
population density, etc. Correlations with temperature, soil moisture,
or antecedent precipitation may be made at times. In most cases, it is
essential that the correlated variables be synchronous, so accurate
timing of the data is often required.
Timing of measured flows and collection of quality samples can be useful
in determining sources of pollution. For example, they can be related
to time of release of pollutants from industrial plants, or to the time
of accidental spills of pollutants. The time of travel of pollutants
along a stream or storm sewer can be estimated from the time of travel
326
-------
of small rises or other flow changes in the channel.
DATA QUALITY
We are not yet overwhelmed with existing data on stormwater character-
istics. As stated by Torno (1975), "One of the serious problems that
faces either a new model developer or one who must evaluate several
models is the incredible lack of good data...". To be effectively
utilized, we need more than simple values as data products. We need to
know something about the'data quality, i.e., about the "goodness" or
truthfulness of the data. Two terms that relate to the data gathering
process are conventionally used to describe data quality, accuracy
and precision. Accuracy refers to the agreement between the measure-
ment and the true value of the measurand, with the discrepancy normally
referred to as error; while precision refers to the reproducibility
(repeatability) of a measurement when repeated on a homogenous time-
stationary measurand, regardless of the displacement of the observed
values from the true value, and thus indicates the number of significant
digits in the result. We are therefore interested in establishing the
best estimate of a measured quantity and the degree of precision of
this estimate from a series of repeated measurements. Calibration,
whether it be of a piece of flow measurement equipment, a chemical
method for wastewater analysis, a stormwater management model, or what-
ever, is simply the process of determining estimates of accuracy and
precision.
Discrepancies between the results of repeated observations, or errors,
are inherent in any measurement process since it is recognized that
the true value of an object of measurement can never be exactly estab-
lished. These errors are customarily classified in two main groups,
systematic and random (or accidental) errors. Systematic errors usually
enter into records with the same sign and frequently with either the
327
-------
same magnitude (e.g., a zero offset) or with an establishable relation-
ship between the magnitude of the measurement and the error. The meth-
ods of symmetry and substitution are frequently used to detect and
quantify systematic errors. In the method of symmetry, the test is
repeated in a symmetrical or reversed manner with respect to the par-
ticular condition that is suspect. In the method of substitution, the
object of measurement is replaced by one of known magnitude (a calibra-
tion standard), an instrument with a known calibration curve is substi-
tuted for the measuring instrument in question, and so on. Thus, sys-
systematic errors bear heavily on the accuracy of the measurement.
Random errors, on the other hand, are due to irregular causes, too many
in number and too complex in nature to allow their origin to be deter-
mined. One of their chief characteristics is that they are normally
as likely to be positive as negative and, therefore, are not likely to
have a great effect on the mean of a set of measurements. The chief
aim of a data quality assurance effort is to account for systematic
errors and thereby reduce errors to the random class, which can be
treated by simple probability theory in order to determine the most
probable value of the object of observation and a measure of the confi-
dence placed in this determination.
The statisticalmeasures of location or central tendency (e.g. , the var-
ious averages, mean, median, mode, etc.) are related to accuracy. The sta
statistical measures of dispersion or variability (e.g., variance and
standard deviation, coefficient of variation, and other measures derived
from central moments of the probability density function) are related
to precision.
Even lacking enough data for statistical treatment^here are some anno-
tations that the data gatherer can make to increase the usefulness of
the data. For example, inspection of equipment and records may indicate
periods of instrument malfunction or failure (e.g., power interruptions).
328
-------
These facts are important and should form a part of the total record.
There may be circumstances discovered during site visits that would
have an effect on preceding data that can not be readily determined,
e.g., a partially blocked sampler intake, a rag caught in the notch of
a weir, etc. These facts should also be noted and, where at all fea-
sible, some qualitative notation as to expected data quality (e.g.,
poor, very good, etc.) should be made.
The importance of notations of data quality arises from the ultimate
use of the data. At the risk of seeming ridiculous, ±50% data
should not be used to calibrate a model whose outputs are desired
within ±20%, nor should strong model verification judgements be made
based upon a very small sample of data with a high variability, for
example. The levels of data quality desired vary with the intended
use of model outputs. The needs for overall basin planning, treatment
plant design, plant operation, and research are all quite different and
this must be kept in mind in designing the data gathering program (or
system).
329
-------
FLOW MEASUREMENT
The subject of flow measurement in general, and sewer flow measurement
In particular, has received considerable attention in the past and con-
tinues to as reflected in the current literature and the introduction
of newHequipment into the commercial marketplace. Only a very sketchy
treatment can be given here; for more detail, reference should be made
to the recent monograph by Shelley and Kirkpatrick (1975), from which
the bulk of the material, including all Figures and Tables, in this
section was taken.
SITE SELECTION
The success or failure of selected flow measurement equipment or methods,
with respect to accuracy and completeness of data collected as well as
reasonableness of cost, depends very much on the care and effort exer-
cised in selecting the gaging site. Except for a few basic require-
ments which are applicable to all types of equipment and methods and
which will be discussed at this point, there are significant differ-
ences in site needs for various flow measurement devices.
A requirement which appears to be obvious, but which is frequently not
sufficiently considered, is that the site selected be located to give
the desired flow measurement. Does flow at the site provide information
actually needed to fulfill project needs? Sometimes influent flows,
diversions, or storage upstream or downstream from the selected site
would bias the data in a manner not understood without a thorough study
of the proposed site. Such study would include reference to surface
330
-------
maps and to sewerage maps and plans. Sometimes groundwater infiltra-
tion or unrecorded connections may exist. For these reasons, a thorough
field investigation should be made before establishing a flow measure-
ment site.
There are some situations where there is no choice of sites. Only a
single site may be available where the desired flow measurement can be
made. In this case, the problem is one of selecting the most suitable
flow measurement equipment and methods for the available site.
A basic consideration in site selection is the possible availability of
flow measurements or records collected by others. At times, data being
collected by the U.S. Geological Survey, by the State, or by other pub-
lic agencies can be used. There are locations where useful data, al-
though not currently being collected, may have been collected in prior
years. Additional data to supplement those earlier records may be more
useful than new data collected at a different site. Other general site
considerations include any history of surcharging, entry and backwater
conditions, and intrusion from receiving waters.
Requirements that apply to all flow measurement sites are accessibility,
personnel and equipment safety, and freedom from vandalism. If a car
or other vehicle can be driven directly to the site at.all times, the
cost in time required for installation, operation, and maintenance of
the equipment will be less, and it is possible that less expensive
equipment can be selected. Consideration should be given to access
during the periods of adverse weather conditions and during periods of
flood stage. Sites on bridges or at manholes where heavy traffic oc-
curs should be avoided unless suitable protection for men and equipment
331
-------
is provided. If entry to sewers is required, the more shallow loca-
tions should be selected where possible. Manhole steps and other facil-
ities for sewer access must be carefully inspected, and any needed
repairs made. Possible danger from harmful gases, chemicals, or explo-
sion should be investigated. With respect to sites at or near streams,
historical flood marks should be determined and used for placement of
access facilities and measurement equipment above flood level where
this is possible. Areas of known frequent vandalism should be avoided.
Selection of sites in open, rather than secluded, areas may help to re-
duce vandalism. Often, the only solution to prevent destruction of fa-
cilities is to place them in solid concrete or steel shelters, and to
surround them with heavy fencing. Erection of warning signs is futile,
as they often serve only to provide targets.
In development of a system or network of flow measurement stations,
primary consideration must be given to cost if the maximum benefit is
to result from available funds. Therefore, cost must be considered in
the selection of each gaging site. Cost reduction can result from
selection of sites where the less expensive types of equipment, which
will fulfill project requirements of accuracy and completeness, can be
installed. For example, if a site is selected where conditions are
such that satisfactory records can be obtained with a weir installation,
this would be preferable to selecting a site where the head loss re-
quired by a weir would not be available, and the expense of installing
a Parshall flume must be met.
METHODS OF FLOW DETERMINATION
This brief discussion is intended to provide an overview of the physical
principles that have been utilized in the design of equipment for the
quantitative measurement of flows. For further reading see Shelley and
Kirkpatrick (1975), ASME (1971), Replogle (1970), McMahon (1964),
332
-------
USDI Bureau of Reclamation (1967), Leupold and Stevens (1974), and any
of the many standard texts on hydraulics and fluid mechanics.
Any flow measurement system can be considered to consist of two dis-
tinct parts, each of which has a separate function to perform. The
first, or primary element, is that part of the system which is in con-
•
tact with the fluid, resulting in some type of interaction. The second-
ary element is that part of the system which translates this inter-
action into the desired readout or recording. .While there is almost
an endless variety of secondary elements, primary elements are related
to a more limited number of physical principles, being dependent upon
some property of the fluid other than, or in addition to, its volume
or mass such as kinetic energy, inertia, specific heat, or the like.
Thus, the primary elements, or rather their physical principles, form
a natural classification system for flow measuring devices and are so
used in this discussion.
Flow measurement systems may be thought of as belonging to one of two
rather broad divisions, quantity and rate. In quantity meters, the
primary element measures isolated (i.e., separately counted) quantities
of fluid either in terms of mass or volume. Usually a container or
cavity of known capacity is alternately filled and emptied, permitting
an essentially continuous flow of the metared supply. The secondary
element counts the number of these quantities and indicates or records
them, often against time. In rate meters, by contrast, the fluid passes
in a continuous, uninterrupted stream, which interacts with the primary
element in a certain way, the interaction being dependent upon one or
more physical properties of the fluid. In the secondary element, the
quantity of flow per unit time is derived from this interaction by
known physical laws supplemented by empirical relations. A general
categorization of flow meters by division, classification, type, and
sub-type is presented in Table 1.
333
-------
TABLE 1. FLOW METER CATEGORIZATION
DIVISION CLASSIFICATION
QUANTITY GRAVIMETRIC
QUANTITY GRAVIMETRIC
QUANTITY GRAVIMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
QUANTITY VOLUMETRIC
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE DIFFERENTIAL PRESSURE
RATE VARIABLE AREA
RATE VARIABLE AREA
RATE VARIABLE AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE HEAD-AREA
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FLOW VELOCITY
RATE FORCE-DISPLACEMENT
RATE FORCE-DISPLACEMENT
RATE FORCE-DISPLACEMENT
RATE FORCE-DISPLACEMENT
RATE FORCE-DISPLACEMENT
RATE FORCE-MOMENTUM
RATE FORCE-MOMENTUM
RATE FORCE-MOMENTUM
RATE FORCE-MOMENTUM
RATE THERMAL
RATE THERMAL
RATE THERMAL
RATE OTHER
RATE OTHER
RATE OTHER
RATE OTHER
RATE OTHER
RATE OTHER
RATE OTHER
TYPE
WEIGHER
TILTING TRAP
WEIGH DUMP
METERING TANK
RECIPROCATING PISTON
OSCILLATING OR RING PISTON
NUTATING DISC
SLIDING VANE
ROTATING VANE
GEAR OR LOBED IMPELLER
DETHRIDGE WHEEL
VENTURI
DALL TUBE
FLOW NOZZLE
ROUNDED EDGE ORIFICE
SQUARE EDGE ORIFICE
SQUARE EDGE ORIFICE
SQUARE EDGE ORIFICE
SQUARE EDGE ORIFICE
CENTRIFUGAL
CENTRIFUGAL
CENTRIFUGAL
IMPACT TUBE
IMPACT TUBE
LINEAR RESISTANCE
LINEAR RESISTANCE
LINEAR RESISTANCE
GATE
CONE AND FLOAT
SLOTTED CYLINDER AND PISTON
WEIR
WEIR
FLUME
FLUME
FLUME
FLUME
FLUME
FLUME
FLUME
FLUME
OPEN FLOW NOZZLE
FLOAT
FLOAT
TRACER
VORTEX
VORTEX
TURBINE
ROTATING ELEMENT
ROTATING ELEMENT
VANE
HYDROMETRIC PENDULUM
TARGET
JET DEFLECTION
BALL AND TUBE
AXIAL FLOW MASS
RADIAL MASS
GYROSCOPIC
MAN6U5 EFFECT
HOT TIP
COLD TIP
BOUNDARY LAYER
ELECTROMAGNETIC
ACOUSTIC
DOPPLER
OPTICAL
DILUTION
ELECTROSTATIC
NUCLEAR RESONANCE
SUBTYPE
CONCENTRIC
ECCENTRIC
SEGMENTED
GATE OR VARIABLE AREA
ELBOW OR LONG RADIUS BEND
TURBINE SCROLL CASE
GUIDE VANE SPEED RING
PITOT-STATIC
PITOT VENTURI
PIPE SECTION
CAPILLARY TUBE
POROUS PLUG
SHARP CRESTED
BROAD CRESTED
VENTURI
PARSHALL
PALMER-BOWLUS
DISKIN DEVICE
CUTTHROAT
SAN DIMAS
TRAPEZOIDAL
TYPE HS. H, AND HL
SIMPLE
INTEGRATING
VORTEX-VELOCITY
EDDY-SHEDDING
HORIZONTAL AXIS
VERTICAL AXIS
334
-------
DESIRABLE EQUIPMENT CHARACTERISTICS
It is obvious from Table 1 that not all types of flow meters are suit-
able for measuring stormwater flows. In fact, the severe conditions
and vagaries of these flows place a number of very stringent design
requirements on flow measurement equipment if it is to function satis-
factorily in such an application. It should also be apparent that no
single design can be considered ideal for all flow measurement activ-
ities in all storm flows of interest. Characteristics of the available
sites, as well as the particular flows in question, make a device that
might be acceptable for one location totally unsuitable for another.
Despite this, one can set forth some equipment "requirements" in the
form of primary design goals and some desirable equipment features in
the form of secondary design goals.
Primary Design Goals
The following are considered to be primary design goals for equipment
that is to be used to measure storm and combined sewer flows:
Range - Since flow velocities may range from 0.03 to 9 m/s (0.1 to
30 fps), it is desirable that the unit have either a very wide range of
operation; be able to automatically shift scales; or otherwise cover
at least a 100 to 1 range.
<
Accuracy - For most purposes, an accuracy of ±10% of the reading at the
readout point is necessary, and there will be applications where an
accuracy of ±5% is highly desirable. Repeatability of better than
±2% is desired in almost all instances.
Flow Effects on Accuracy - The unit should be capable of maintaining its
accuracy when exposed to rapid changes in flow; e.g., depth and velocity
changes in an open channel flow situation. There are instances where
335
-------
the flows of interest may accelerate from minimum to maximum in as short
a time period as five minutes.
Gravity and Pressurized Flow Operation - Because of the conditions that
exist at many measuring sites, it is very desirable that the unit have
the capability (within a closed conduit) of measuring over the full
range of open channel flow as well as with the conduit flowing full and
under pressure.
Sensitivity to Submergence or Backwater Effects - Because of the possi-
bility of changes in flow resistance downstream of the measuring site
due to blockages, rising river stages including possible reverse flow,
etc., it is highly advantageous that the unit be able to continue to
function under such conditions or,, at a minimum, be able to sense the
existence of such conditions which would lead to erroneous readings.
Effects of Solids Movement - The unit should not be seriously affected
by the movement of solids such as sand, gravel, debris, etc. within
the fluid flow.
Flow Obstruction - The unit should be as non-intrusive as possible to
avoid obstruction or other interference with the flow which could lead
to flow blockage or physical damage to some portion of the device.
Head Loss - To be-usable at a maximum number of measurement sites, the
unit should induce as little head loss as possible.
Manhole Operation - To allow maximum flexibility in utilization, the
unit should have the capability of being installed in confined and
moisture-laden spaces such as sewer manholes.
336
-------
Power Requirements - The unit should require minimum power at the meas-
uring site to operate; the ability to operate on batteries is a definite
asset for many installations.
Secondary Design Goals
The following are desirable features for flow measuring equipment,
especially for use in a storm or combined sewer application:
Site Requirements - Unit design should be such as to minimize site re-
quirements, such as the need for a fresh water supply, a vertical drop,
excessive physical space, etc.
Installation Restrictions or Limitations - The unit should impose a
minimum of restrictions or limitations on its installation and be
capable of use on or within sewers of varying size.
Simplicity and Reliability - To maximize reliability of results and
operation, the design of the unit should be as simple as possible, with
a minimum of moving parts, etc.
Unattended Operation - For the majority of applications, it is highly
desirable that the equipment be capable of unattended operation.
Maintenance Requirements - The design of the equipment should be such
that routine maintenance is minimal and troubleshooting and repair can
be effected with relative ease, even in the field.
Adverse Ambient Effects - The unit should be unaffected by adverse
ambient conditions such as high humidity, freezing temperatures, hydro-
gen sulphide or corrosive gases, etc.
337
-------
Submersion Proof - The unit should be capable of withstanding total
immersion without significant damage.
Ruggedness - The unit should be of rugged construction and as vandal
and theft proof as possible.
Self Contained - The unit should be self contained insofar as possible
in view of the physical principles involved.
Precalibration - In order to maximize the flexibility of using the
equipment in different settings* it is desirable that it be capable of
precalibration; i.e., it should not be necessary to calibrate the sys-
tem at each location and for each application.
Ease of Calibration - Calibration of the unit should be a simple,
straightforward process requiring a minimum amount of time and
ancillary equipment.
Maintenance of Calibration - The unit should operate accurately for
extended periods of time without requiring recalibration.
Adaptability - The system should be capable of: indicating and re-
cording instantaneous flow rates and totalized flows; providing flow
signals to associated equipment (e.g., an automatic sampler); imple-
mentation of remote sensing techniques or incorporation into a com-
puterized urban data system, including a multi-sensor single readout
capability.
Cost - The unit should be affordable both in terms of acquisition and
installation costs as well as operating costs, including repair and
maintenance.
338
-------
Evaluation Parameters
It is, of course, not necessary that all of the primary and secondary
design goals be achieved for all flow measurement requirements. For
example, "spot" measurements of all flow rather than continuous records
are sufficient at times. Flow measurement devices used to calibrate
others need not necess'arily be self contained, nor would unattended
operations be required. Furthermore, meeting all of the listed design
goals for all installations and settings would be difficult, if not
impossible, to achieve in a single design.
Nonetheless, the primary and secondary design goals can be used to
formulate a set of evaluation parameters against which a given design
or piece of equipment can be judged. Since application details may
make certain parameters more or less important in one instance or
another, no attempt has been made to apply weighting factors or assign
numerical rank. It is hoped that the evaluation factors will prove
useful, as a check list among other things, for the potential user
who has a flow measurement requirement and who may require assistance
in the selection of his equipment.
The evaluation parameters together with qualitative scales, are pre-
sented in the form of a flow measurement equipment checklist in Table 1.
GENERIC EVALUATIONS OF SOME PROMISING DEVICES
A slightly modified form of the flow measurement equipment checklist
given in Table 2 has been used to evaluate the various flow measuring
devices and techniques of Table 2, and a matrix summary is given as
Table 3. It must be emphasized that these evaluations are made with a
storm or combined sewer flow measurement application in mind and will
not necessarily be applicable for other types of flows. They are nec-
essarily somewhat subjective, and the writer apologizes in advance
339
-------
TABLE 2. FLOW MEASUREMENT EQUIPMENT CHECKLIST
Designation:
Evaluation Parameter
Scale
Weight and Score
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Range
Accuracy
Flow Effects on Accuracy
Gravity & Pressurized Flow
Operati on
Submergence or Backwater
Effects
Effect of Solids Movement
Flow Obstruction
Head Loss
Manhole Operation
Power Requirements
Site Requirements
Installation Restrict*! ons
or Limitations
Simplicity and Reliability
Unattended Operation
Maintenance Requirements
Adverse Ambient Effects
Submersion Proof
Ruggedness
Self Contained
Precali brati on
Ease of Calibration
Maintenance of Calibration
Adaptability
Cost
Portability.
D Poor D Fair DGood
D Poor D Fair D Good
DHigh D Moderate D Slight
D No DYes
DHigh D Moderate D Low
D High
D High
DHigh
D Poor
D High
D High
D High
D Poor
D No
D High
D High
D No
D Poor
D No
D No
D Poor
D Poor
D Poor
D High
D No
D Moderate D Slight
D Moderate D Slight
D Medium D Low
D Fair D Good
D Medium D Low
D Moderate D Slight
D Moderate D Slight
D Fair D Good
DYes
D Medium D Low
D Moderate D Slight
D Yes
D Fair DGood
D Yes
DYes
D Good
D Good
D Good
D Fair
D Fair
D Fair
D Medium
D Low
D Yes
Comments:
340
-------
TABLE 3. FLOWMETER EVALUATION SUMMARY
GrdvImetrlc-al] types
Vo1umetric- a 11 types
Venturl Tube
Dall Tube
Flow Nozzle
Or1flee Plate
Elbow Meter
Slope Area
Sharp-Crested Weir
Broad-Crested Weir
Subc ri t i ca1 Flume
Pdrshall Flume
Pa 1mer-Bowlus Flume
Dlskin Device
Cutthroat Flume
San Dnnas Flume
Trapezoidal rIume
Type HS, H 4 HL Flume
Open F1ow Nozzle
Float Velocity
Tracer Veloclty
Vortex Velocity
Eddy-Stieddl ng
Turbine Meter
Rot at 1ng-E1ement Meter
Vane Meter
Hydrometrlc Pendulum
Target Meter
Force-Momen turn
Hot-Tip Meter
Boundary Layer Meter
Electromagnetic Meter
Acoustic Meter
Doppler Meter
Optical Meter
01 lutton
*
Cn
>
u
S
E
a
L
L
L
L
L
L
i.
H
M
H
L
N
M
M
L
L
L
H
H
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
K
H
S
M
S
H
S
5
H
H
S
S
S
M
S
S
S
M
M
S
S
H
M
H
H
M
M
M
M
H
S
S
H
II
S
S
tructi on
o
X
0
H
H
S
S
S
H
S
S
H
M
S
S
S
H
S
S
S
S
S
S
S
H
M
H
M
H
M
H
M
M
S
S
S
S
S
S
l/>
_J
T3
M
H
M
H
H
H
H
M
M
M
M
M
S
S
S
S
S
M
M
S
S
H
S
H
S
S
S
S
H
M
H
M
M
M
S
S
c
o
in
*j
"e
o
C
O
4J
~-
I/I
OJ
a;
0
ra
i/>
c
H
H
H
M
M
S
S
s
M
M
S
M
S
S
s
s
s
M
M
S
S
H
S
M
S
M
S
M
H
M
M
M
M
M
S
S
as
.a
-
a
T>
c
>i
'£
E
P
F
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
6
F
F
F
F
G
G
G
F
P
F
F
F
F
F
S
F
o
19
L.
01
a.
•a
•o
a>
+j
n
c
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
H
Y
N
Y
Y
Y
Y
Y
Y
Y
N
Y
c
a>
at
3
cr
at
a:
QJ
U
c
aj
=
nO
H
H
M
H
L
H
L
L
H
L
L
L
L
H
L
L
L
M
M
L
M
H
M
H
H
M
L
H
H
H
M
M
N
M
L
M
M
M
M
M
H
M
M
M
M
H
M
M
M
H
M
M
H
M
M
H
S
s
s
5
H
M
H
S
S
M
S
s
s
s
H
S
o
0
k.
•f- a)
U TJ
E en
.a en
3 ^
- F
- F
- G
- G
- G
- F
- G
- G
- G
- G
- G
- G
- G
- F
- G
- G
- G
- G
- G
- G
- F
p
- F
- F
- G
- G
- G
- P
- P
- F
- G
- F
- F
- F
- G
- F
T3
0)
c
•a
c
o
L_>
£
Y
Y
r
Y
Y
Y
K
N
Y
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
.
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
c
0
>o
.a
o
ai
>a
G
3
G
G
G
C
F
F
G
F
G
G
G
f
G
G
G
G
G
.
G
F
G
G
G
F
F
G
G
G
G
G
G
G
G
G
>.
1
'• S"
F -
F -
G -
F -
G
P -
G -
G -
P -
F -
G -
G -
G -
F -
G -
G -
G -
F -
F -
_
G -
F -
F -
F -
G -
F •
F -
F -
G -
F -
G -
G -
G -
G *-
G -
G -
i/i
H
H
H
H
M
L
L
H
L
L
M
M
L
L
L
L
L
L
L
L
H
H
M
H
L
L
L
H
H
H
H
H
H
H
L
H
>>
£1
L.
N
It
N
N
N
Y
N
N
Y
N
N
Y
Y
Y
N
N
H
Y
Y
Y
Y
N
Y
N
Y
N
Y
N
N
N
N
N
N
N
Y
Y
Legend:
F - Fair N - No
G - Goal P - Poor
H - High s - Slight
L - .Low y - Yes
M - Medium or Moderate
341
-------
to the clever reader who has made a particular device work satisfacto-
rily in such an application and, hence, feels that it has been treated
unfairly.
Only a few of the evaluation parameters normally have numbers associated
with them. To assist the reader in interpreting the ratings, the fol-
lowing general guidelines were used. If the normal range of a partic-
ular device was considered to be less than about 10:1, it was termed
poor; if it was considered to be greater than around 100:1, it was
termed good. The intermediate ranges were termed fair. The accuracy
that might reasonably be anticipated in measuring storm or combined
sewer flows was considered rather than the best accuracy achievable by
a particular device. For example, although a sharp-crested weir may be
capable of achieving accuracies of ±1.5% or better in clear irrigation
water flows, accuracies of much better than ±4-7% should not necessarily
be anticipated for a sharp-crested weir measuring stormwater or com-
bined sewer discharges. If the accuracy of a particular flow measuring
device or method was considered to be better than around ±1-2%, it was
termed good; if it was considered to be worse than around ±10%, it was
termed poor. The intermediate accuracies were termed fair.
The flow measuring devices and techniques were not rated on two evalua-
tion parameters, submersion proof and adaptability, because these fac-
tors are so dependent upon the design details of the secondary element
selected by the user.
Discussion
In comparison with Table 3, Table 4 offers a different (and even more
subjective) comparison of the primary devices or techniques just de-
scribed. Each method is numerically evaluated in terms of its percent
of achievement of several desirable characteristics. Dilution tech-
niques as a class appear to be the most promising of all. In view of
342
-------
TABLE 4. COMPARISON OF MOST POPULAR PRIMARY DEVICES OR TECHNIQUES
Primary Device
or Technique
Dilution
Acoustic (Open Channel)
Parshall Flume
Palmer Bowlus Flume
Current Meter
Electromagnetic
Acoustic (Pressure Flow)
Open Flow Nozzle
Sharp-Crested Weir
Flow Tube
Vehturl Tube
Trajectory Coordinate
Slope Area
Range
100
100
90
80
90
50
100
60
60
50
20
80
80
Desirable Characteristic (% of Achievement)
Uncal ib.
Accuracy
100
100
95
90
95
100
100
95
95
100
100
70
50
Head
Loss
100
100
80
85
100
100
100
70
70
95
90
50
100
Free From
Upstream
Effects
100
60
90
90
100
100
60
80
80
40
70
100
20
Free From
Downstream
Effects
100
90
80
85
100
100
90
75
80
100
100
70
100
Solids
Bearing
Liquids
100
95
90
90
90
100
95
80
50
95
90
100
100
Portabil 1 ty
100
80
70
90
100
0
0
80
80
0
0
100
100
Unattended
Operation
80
100
•100
100
0
100
100
.95
90
100
100
0
0
Comments
Especially useful as a calibra-
tion tool .
Good In large flows but
expensive.
Requires drop in floor.
Good overal 1 .
Results are very operator
dependent.
Generally requires pressure flow
Wetted transducers recommended.
Good if head drop is available.
Will require frequent cleaning.
Pressurized flow only.
Pressurized flow only.
Requires free discharge.
Use as last resort.
to
*>
CO
-------
the current state of the art, however, their usefulness is probably
greatest as a tool for in-place calibration of other primary devices.
They have also been extremely useful for general survey purposes and
have found some application as an adjunct to other primary devices
during periods of extreme flow such as pressurized flow in a conduit
that is normally open channel.
Acoustic open channel devices are also quite promising; but, because of
their dependency upon the velocity profile and the frequently resulting
requirement for several sets of transducers, they are presently only
justifiable for very large flows in view of the expense involved. The
usefulness of the Parshall flume is evidenced by its extreme popularity.
The requirement for a drop in the flow is a disadvantage, and submerged
operation may present problems at some sites. Known uncertainties in
the head-discharge relations (possibly up to 5%) together with possible
geometric deviations make calibration in place a vital necessity if
high accuracy is required. Palmer-Bowlus type flumes are very popular
overall. They can be used as portable as well as fixed devices in many
instances, are relatively inexpensive, and can handle solids in the flow
without great difficulty.
All point velocity measuring devices have been lumped together in the
current meter category. In the hands of a highly experienced operator,
good results can be obtained (the converse is also true, unfortunately),
and they are often used to calibrate primary devices in place or for
general survey work. They are generally not suited for unattended op-
eration in storm and combined sewer flows, however.
Electromagnetic flowmeters show considerable promise where pressurized
flow is assured as do closed pipe acoustic devices. Neither can be
considered portable if one requires that the acoustic sensors be
wetted, a recommended practice for most wastewater applications.
344
-------
Open flow nozzles and sharp-crested weirs are often used where the re-
quired head drop is available. Weirs will require frequent cleaning
and are best used as temporary installations for calibration purposes.
Flow tubes and Venturis are only suitable for pressurized flow sites
such as might be encountered, for example, at the entrance to a treat-
ment plant.
Trajectory coordinate techniques, such as the California pipe or Purdue
methods, require a pipe discharging freely into the atmosphere with
sufficient drop to allow a reasonably accurate vertical measurement to
be made, a situation not often encountered in storm or combined sewers.
Slope area methods, as explained earlier, must generally be considered
as producing estimates only, and consequently should be considered as
the choice of last resort (despite their apparent popularity).
REVIEW OF COMMERCIALLY AVAILABLE EQUIPMENT
The number of commercial firms that offer flow measuring equipment in
the marketplace today is astoundingly large. Even when attention is
limited to devices intended for liquid flowmetering, the number is still
extremely large, probably well in excess of two hundred. Many manufac-
turers offer more than one type of primary device (and these typically
in numerous models), and when combined with secondary device choices,
the number is virtually overwhelming. Thus, no attempt to cover all
available equipment can be made here, and this explains the great
amount of space devoted to the generic descriptions in the preceding
subsection. We will content ourselves here to note that two or more
firms offer all devices that were described except for sharp-crested
weirs, which are usually fabricated directly by (or for) the user In
accordance with specifications for the particular measuring site.
The firms offering flow measuring equipment as at least a part of their
product line range from very large, well known manufacturers that have
offered a wide range of flow measuring equipment for over a century
345
-------
to relatively small organizations with a limited product line which has
only recently been introduced. This latter category should not be
excluded from consideration solely because of their seemingly novitiate
status. The principals involved frequently have many years of experi-
ence, and their designs often reflect the most up-to-date expressions
of the state of the art.
The revolution in the electronics industry, especially as regards
solid-state designs and integrated circuitry, has not gone unnoticed
by most flowmeter manufacturers; and as a result, many new, sophisti-
cated secondary devices have recently appeared, and older equipment is
frequently being upgraded in design to reflect the more modern tech-
nologies. Furthermore, many of these new secondary devices are of
digital (rather than analog) design and are frequently computer com-
patible as supplied, offering tremendous possibilities for system
structure.
A listing, by no means complete, of some manufacturers who offer flow
measuring equipment in the categories listed in Table 4 is presented
in Table 5. Under the heading "Company", the name, address, and tele-
phone number has been provided. Under the heading "Products", only
those products bearing on the flow measurement categories of Table 4
have been listed, even though the particular company may have a much
more extensive flow measurement product line. The product emphasis
was placed on primary devices, with secondary devices (in the form of
level gages) indicated only where they are offered as "flowmeters". It
can be generally assumed that each manufacturer offers a complete line
of secondary elements for use with his primary devices.
Table 5 can be used to obtain direct, up-to-date information on all of
the types of equipment discussed from at least two suppliers. Refer-
ence can be made to Shelley and Kirkpatrick (1975) for descriptions of
the offerings of these and a number of other manufacturers.
346
-------
TABLE 5. SOME FLOW MEASUREMENT EQUIPMENT MANUFACTURERS
COMPANY
AMERICAN CHAIN AND CABLE COMPANY,
ACCO BRISTOL DIVISION
WATERBURY. CONNECTICUT 06720
TELEPHONE (203) 756-4451
BADGES METER, INC.
INSTRUMENT DIVISION
45*5 WEST BROWN DEER ROAD
MILWAUKEE, WISCONSIN 53223
TELEPHONE (414) 355-0400
BADGER METER, INC.
PRECISION PRODUCTS DIVISION
6116 EAST 15TH STREET
TULSA, OKLAHOMA 74115
TELEPHONE (918) 836-4631
BIF - A UNIT OF GENERAL SIGNAL
1600. DIVISION SOAD
WEST WARWICK. R.I. 02893
TELEPHONE (401) 885-1000
INC.
BROOKS
DIVISION
EMERSON ELECTRIC COMPANY
407 WEST VINE STREET
HATFIELD, PENNSYLVANIA 19440
TELEPHONE (215) 247-2366
COKTRDLOTRON CORPORATION
176 CONTROL AVENUE
FAJWWCDALE, L.I,, MM YORK 11735
TELEPHONE (516) 249-4400
CUSHIHG ENGINEERING INC.
3364 COMMERCIAL AVENUE '
NORTHBROOK, ILLINOIS 60062
TELEPHONE (312) 564-0500
C.W. STEVENS, INC.
P. 0. BOX 619
KENNETT SQUARE, PENNSYLVANIA 19348
TELEPHONE (215) 444-0616
DKEXELBROOK ENGINEERING COMPANY
205 KEITH VALLEY ROAD
BORSHAM, PENNSYLVANIA 19044
TELEPHONE (215) 674-1234
ENVIRONMENTAL MEASUREMENT SYSTEMS
A DIVISION OF WESMAR
905 DEXTER AVENUE NORTH
SEATTLE, WASHINGTON 98109
TELEPHONE (206) 285-1621
EPIC INC.
150 NASSAU STREET
SUITE 1430
NEW YORK, NEW YORK 10038
TELEPHONE: (212) 349-2470
FISCHER & PORTER CO.
WASMINSTER, PENNSYLVANIA 18974
TELEPHONE (215) 675-6000
Combination depth and
velocity measuring device
in a single unit
Flow tubes, open flow
nozzles, Parshall flumes
Acoustic (open channel)
Flow tubes, open flow
nozzles, Parshall flumes,
"universal" venturi tubes
Electromagnetic
Acoustic (pressure flow)
Electroaagnetic
Acoustic level gage
Electronic level gage
Acoustic (open channel)
Current meters, level
gages
ElectroBagnetic, flow
tubes, open flow nozzles,
Parshall fluaes, level
gages
COMPANY
CARL FISHER AND COMPANY
DIVISION OF FORMULABS, INC.
529 WEST FOURTH AVENUE
P. O. BOX 1056
ESCONDIDO, CALIFORNIA 92025
TELEPHONE (714) 745-6423
FLUMET CO.
P. O. BOX 575
WESTFIELD, NEW JERSEY
N.Y. OFFICE: TELEPHONE (212) 227-6668
THE FOXBORO COMPANY
FOXBORO, MASSACHUSETTS 02035
TELEPHONE (617) 543-8750
HINDE ENGINEERING COMPANY OF CALIFORNIA
P. O. BOX 56
SARATOGA, CALIFORNIA 95070
TELEPHONE (408) 378-4112
INTEROCEAN SYSTEMS, ISC.
3510 KURTZ STREET
SAN DIEGO, CALIFORNIA 92110
TELEPHONE (714) 299-4500
KAHL SCIENTIFIC INSTRUMENT CORPORATION
P. 0. BOX 1166
2L CAJOH, CALIFORNIA 92022
TELEPHONE: (714) 444-2158
F. B. LEOPOLD COMPANY
DIVISION OF SYBRON CORPORATION
227 S. DIVISION ST.
ZELIENOPLZ, PENNSYLVANIA 16063
TELEPHONE (412) 452-6300
LEUPOLD & STEVENS, INC.
P. O. BOX 588
600 N.«. MEADOW DRIVE
EEAVERTON, OREGON 97005
TELEPHONE (503) 646-9171
MANNING ENVIRONMENTAL CORP.
120 DU BOIS STREET
P. 0. BOX 1356
SANTA CRUZ, CALIFORNIA 95061
TELEPHONE (408) 427-0230
MASTIC BUB-L-AIR
2116 LAKEHOOR DRIVE
OLYMPIA, WASHINGTON 98502
TELEPHONE (206) 943-2390
METRITAPE, INC.
77 COMMONWEALTH AVENUE
WEST CONCORD, MASSACHUSETTS 01742
TELEPHONE (617) 369-7500
NB PRODUCTS, INC.
35 BEDLAH ROAD
NEW BRITAIN, PENNSYLVANIA 18901
TELEPHONE (215) 345-1879
PRODUCTS
Fluorescent dyes
Palmer-Bowlus flumes
Electromagnetic, level
gages
Palmer-Bowlus flumes
Current meters, level
gages
Current meters,
fluorescent dyes
Open flow nozzles,
Palmer-Bowlus flumes,
Parshall flumes
Float level gages
Acoustic and "dipper"
level gages
Bubbler level gage
Electronic level gage
Portable V-notch weirs,
level gages
-------
TABLE 5. SOME FLOW MEASUREMENT EQUIPMENT MANUFACTURERS Continued)
oo
CUWANY
H-CON SYSTEMS COMPANY
308 MAIN STREET
NEW ROCHELLE, NEW YORK 10801
TELEPHONE (914) 235-1020
NDSONICS, INC
9 KEYSTONE PLACE
PARAMOS, NEW JERSEY 07652
TELEPHONE (201) 265-2400
OCEAN RESEARCH EQUIPMENT, INC.
FALMOUTH, MASSACHUSETTS 02541
TELEPHONE (617) 548-5800
THE PERMUTIT COMPANY
DIVISION OF SYBRON CORPORATION
£49 MIDLAND AVENUE
PAKAMUS, NEW JERSEY 07652
TELEPHONE (201) 262-8900
PLASTI-FAB, INC.
11650 S.W. RIDCEVIEW TERRACE
BEAVERTON, OREGON 97005
TELEPHONE (502) 644-1428
POLCON, INC.
AS AFFILIATE OF CARL F. BUETTNER
& ASSOCIATES, INC.
5106 HAMPTON AVENUE
ST. LODIS, MISSOURI 63109
TELEPHONE (314) 353-5993
PORTAC
MIN-ELL COMPANY, INC.
1689 BLUE JAY LANE
CHERRY HILL, NEW JERSEY 08003
TELEPHONE (609) 429-0421
ROBERTSHAW CONTROLS COMPANY
P. 0. BOX 3523
KNOXVILLE, TENNESSEE 37917
TELEPHONE (615) 546-0524
SARATOGA SYSTEMS, INC.
10601 SOUTH SARATOGA-SUNNYVALE ROAD
CUPERTINO. CALIFORNIA 95014
TELEPHONE (408) 247-7120
SCARPA LABORATORIES, INC.
46 LIBERTY STREET, BRAINY BORO STATION
METUCHIN, NEW JERSEY 08840
TELEPHONE (201) 549-4260
PRODUCTS
Float and "dipper"
level gages
Acoustic (pressure flow)
Acoustic (open channel)
Flow tubes, open flow
nozzles, Parshall flumes,
venturi tubes
Palmer-Bowlus flumes,
Parshall flumes, V-notch
weir boxes
Open channel flow tube
Current meter flow tube
Parshall flumes,
level gages
Acoustic (pressure flow)
Acoustic (pressure flow)
COMPANY
SIGMAMOTOR, INC.
14 ELIZABETH STREET
MIDDLEPORT, NEW YORK 14105
TELEPHONE (716) 735-3616
SINGER-AMERICAN METER DIVISION
13500 PH1LMONT AVENUE
PHILADELPHIA, PENNSYLVANIA 19116
TELEPHONE (215) 673-2100
SIRCO CONTROLS COMPANY
8815 SELKIRK STREET
VANCOUVER 14, BRITISH COLUMBIA, CANADA
TELEPHONE (604) 261-9321
TAYLOR
SYBRON CORPORATION
TAYLOR INSTRUMENT PROCESS CONTROL DIVISION
TELEPHONE (716) 235-5000
TRI-AID SCIENCES, INC.
161 NORR.IS DRIVE
ROCHESTER, NEW YORK 14610
TELEPHONE (716) 461-1660
UNIVERSAL ENGINEERED SYSTEMS, INC.
7071 COMMERCE CIRCLE
PLEASANTON, CALIFORNIA 94566
TELEPHONE (415) 462-1543
VICKERY-SIMMS, INC.
P. 0. BOX 459
ARLINGTON, TEXAS 76010
TELEPHONE (817) 261-4446
WALLACE-MURRAY CORPORATION
CAROLINA FIBERGLASS PLANT
P. 0. BOX 580
510 EAST JONES STREET
WILSON, NORTH CAROLINA 27893
TELEPHONE (919) 237-5371
WESMAR INDUSTRIAL SYSTEMS DIVISION
905 DEXTER AVENUE NORTH
SEATTLE, WASHINGTON 98109
TELEPHONE (206) 285-2420
WESTINGHOUSE ELECTRIC CORPORATION
OCEANIC DIVISION
P. O. BOX 1488, MAIL STOP 9R30
ANNAPOLIS, MARYLAND 21404
TELEPHONE (301) 765-5658
PRODUCTS
Bubbler level gage
Palmer-Bowlus flumes,
Parshall flumes, level
gages
Acoustic level gage
Electromagnetic
Acoustic level gage
Palmer-Bowlus flumes
Parshall flumes, venturi
Parshall flumes
Acoustic level gages
Acoustic (open channel)
-------
In these days of inflation, little can be said about equipment costs
except in a very cursory fashion. Recording level gages generally
start around $1K, and a full system for measuring flows at a difficult
site could well run $100K. Construction, installation, and (impor-
tantly) projected maintenance and repair costs must be considered in
addition to equipment acquisition costs to arrive at true cost of
ownership, which is the only real basis for comparison.
In closing this discussion of commercially available equipment, it must
be noted that none of the devices completely measures up to the desirable
equipment characteristics listed earlier, and much remains to be done
before such an "ideal" device is available.
REVIEW OF RECENT FIELD EXPERIENCE
A brief review of flow measurement experiences from recent projects in
the storm and combined sewer area will be given to allow a better appre-
ciation of the application of some of the flow measuring devices and
techniques in an actual field setting. The projects are reviewed in a
generally chronological order, starting with the oldest (circa 1966)
and coming up to the present time. The project title will be used for
identification purposes. In some instances final reports have been
issued, while in others project files or interviews with project engi-
neers formed the source of the information presented.
Characterization and Treatment of Combined Sewer Overflows - was a
study whose general objectives were: (a) to develop workable systems
to manage overflows from the combined sewers of San Francisco, thereby
alleviating pollution and protecting beneficial uses of local receiving
waters, and (b) to provide the rationale and methodology for controlling
pollution from combined sewer overflows in other metropolitan areas of
the United States. Data collection for the project included measure-
ment of dry weather flows and storm overflows of the Selby Street and
Laguna Street trunk sewers. Monitoring included measurement of rain-
349
-------
fall and discharge as well as the quality characteristics of the over-
flows. The tracer dilution method was selected for use in measurement
of dry weather flows. Pontacyl Brilliant Pink B fluorescent dye was
used for the tracer, and quantitative analyses for the tracer were made
with a Turner Model 110 Fluorometer.
Although successful at the Laguna Street site, use of the dilution
method did not prove satisfactory for measurement of the Selby Street
overflow. Uneven distribution of the tracer when injected resulted
from exposed sludge banks, and there was insufficient turbulence for
adequate mixing of the dye. Because of the resulting lack of reliable
data, a Palmer-Bowlus flume with a 1.22m (4 ft) throat and 15 cm
(6 in.) invert slab, constructed from 16 gauge galvanized sheet metal
was installed. A continuous record of the upstream water level was
obtained by mounting a Stevens water level recorder, operated by float,
in a stilling well constructed of sandbags.
Because of generally unsatisfactory conditions, several methods were
used for the measurement of storm flow in the Selby Street outfall
structure. These were:
a. Velocity determination with current meters at a point 15m
(50 ft) above the outfall structure. Not considered to
yield reliable data as the meters were immediately fouled
with rags and other debris.
b. Differential head measurements over the broadcrested weirs
of the outfall structure. Because of expected interfer-
ence by tide gates, the theoretical head-discharge rela-
tionship for a broad-crested weir of similar shape was
used for comparison purposes only.
350
-------
c. Measurement of surface velocities in the outfall structure
by timing the traverse of styrofoam floats across a meas-
ured control section. A factor of 0.64 was applied to
surface velocities to convert to average velocity, thus
accounting for both horizontal and vertical velocity
distributions.
d. Measurement of vertical velocity profiles in the outfall
structure with an especially designed current meter having
low velocity sensitivity. This was to establish discharge
values under low head conditions and to check the results
of the surface velocity determinations.
Water levels in the outfall structure were continuously measured by
means of a "bubbler" gage. A rating curve was developed from the re-
sults of the surface velocity determinations and the current meter
measurements in the outfall structure. A theoretical curve computed
from the broadcrested weir formula was approximately 10% larger than
this developed curve. Flow determinations in the Laguna Street
overflow were made by measuring the depth of flow at the outfall sewer
and calculating the discharge by means of the Manning equation. Use of
the Manning equation was said to be justified because the slope of the
outfall sewer is known, and a uniform reach extends about 210m (700 ft)
upstream from the point at which depths of flow were determined.
Engineering Investigation of Sewer Overflow Problem - Roanoke,
Virginia - included investigations of approximately 25 percent of
Roanoke, Virginia's separate sanitary sewer system concerning the
amounts of infiltration for various storm intensities and durations and
the amounts of sewage overflow from the system. Flows in three sani-
tary sewer interceptors, and streams draining the corresponding basins,
were gaged during storm events to measure infiltration and runoff
351
-------
quantities to establish their relation to rainfall intensities and
durations. After significant variation in dry weather flows was ob-
served, continuous monitoring of flows in the interceptors was main-
tained. Overflows bypassing the water pollution control plant were
measured during rainfall events.
Sharp-crested weirs were used to measure flows in two of the streams.
In the third stream, a stage-discharge curve was developed from the
Manning formula using the measured hydraulic characteristics of the
stream. In two of the interceptor sewers, and in the water pollution
control plant overflow, a stage-discharge curve based on the Manning
formula and the measured hydraulic characteristics of the sewer were
used to convert depth measurements to flow estimates. In the third
interceptor sewer, dry weather flows were estimated using the Manning
formula, but during rainfall events the sewer became surcharged and
overflow was measured by means of a weir installed in the side of the
manhole wall.
No problems with use of the streamflow measurement devices and methods
were noted. However, accuracy of measurements with use of the Manning
formula in the natural stream channel photographed must be considered
very poor. A photograph of one of the weirs used for streamflow meas-
urement shows a significant accumulation of trash and debris on the
weir. Under this condition, accuracy must be considered poor. Hydro-
graphs indicated that two of the interceptor sewers were surcharged
during many of the storms, and a record of discharge was not obtained
during those periods.
Water levels at the gaging sites were recorded by means of six float-
actuated, continuous water-level recorders manufactured by the
Instruments Corporation (now a part of Belfort Instrument Company),
and one pressure type recorder manufactured by the Bristol Company.
After the float-actuated recorders were serviced and supplied with an
expanded time scale, they performed satisfactorily for the duration of
352
-------
the program. During dry-weather periods, the bubbler pipe in the
Bristol recorder collected debris and required frequent cleaning.
Because of this maintenance problem, it was replaced with float-
actuated, continuous water-stage recorder.
Combined Sewer Overflow Abatement Alternatives - Washington B.C. - was
a project whose objectives were to: (a) define the characteristics of
urban runoff in the subject area, (b) investigate the feasibility of
high-rate filtration for treatment of combined sewer overflow; and
(c) develop and evaluate alternative methods of solution. Investigative
activities included field monitoring of combined sewer overflows at
two sites, and of separated stormwater discharges at one site. The
monitoring program was conducted over a period of about six months,
April 1 to September 23, 1969.
Selection of a satisfactory technique for flow measurement presented a
problem. Weirs were not used because backwater elevations would have
caused surcharging and flooding at the expected high flow rates.
Depth of flow measurements with the use of a steady state empirical
equation such as the Manning equation for calculating flow were not
used because flow conditions were not steady state during periods of
storm runoff. The method selected was use of lithium chloride as a
tracer in a procedure similar to that of the salt dilution method
(continuous injection type). Use of a lithium salt is said to improve
the technique because the background of lithium in wastewater is usually
low and because lithium concentrations at fractional parts per million
levels can be accurately and conveniently determined by atomic absorp-
tion or flame emission spectroscopy. The slope-area method was used as
a check.
A number of difficulties experienced in use of the equipment resulted
in loss of flow record during several major storms. Greatest trouble
was in clogging or damage to the submersible pump used to collect sam-
ples of wastewater required for measuring lithium chloride concentra-
tion. Flooding of one of the lithium chloride release stations caused
353
-------
damage to the bubbler instruments used to measure water stage, to the
lithium chloride release system, and to other equipment. Flow rate
estimates based on depth-of-flow measurements and the Manning formula
were compared with results of the tracer method. Only a very general
correlation with significant spread resulted. Large differences in
flow were attributed to inaccurate measurement of depth of flow and
the assumption of steady-state conditions inherent in Manning's formula.
Urban Runoff Characteristics - concerned investigations of a combined
sewer watershed in Cincinnati, Ohio for the refinement of the US EPA
Storm Water Management Model. Flows at three sites were monitored.
At two of the sites, flow was actually measured in two tributary sewers
immediately upstream from their junction. Third site was at the
outlet of the watershed; thus, five sets of flow measuring equipment
were used. The flow measuring equipment essentially consisted of a
compressor, a manometer, and a pressure-type recorder. This recorder
operated by measuring the pressure due to the depth and velocity of
water flowing into a pipe by bubbling air through a long tube inserted
into the water. The gage released air is at a constant pressure, and
as the depth and the velocity of flow changed, the pressure differ-
ential was recorded on a circular chart in inches of water.
To obtain the discharge corresponding to the measured specific energy,
Manning's equation was used to express the depth and the velocity of
flow as functions of the discharge and the hydraulic radius of the sewer
cross sections. Thus, curves have been calculated relating the meas-
ured specific energy and the corresponding discharge at the five meas-
ured locations.
Apparently, the value of slope used in the Manning formula was that of
the sewer line at each of the five measuring sites. A photograph in
the report shows a heavy, metal, top-hanging gate at the outlet of the
sewer watershed. The outlet flow monitoring site is described as about
6m (20 ft) upstream from this gate. If this is the case, flow past the
354
-------
site probably would not be uniform, and the Manning formula would not
be applicable. Flows in the two pairs of sewers just upstream from
their junction with two sewers to be monitored would be controlled by
the slopes of each of the two monitored sewers rather than the slopes of
the four tributary sewers immediately upstream, which were the slopes
used in the Manning formula to compute flow. In any case, water surface
slope is more properly used with the Manning formula than is the sewer
slope. Plotting of storm hydrographs for the measured sites discloses
a number of serious inconsistencies in the data.
Storm and Combined Sewer Pollution Sources and Abatement - Atlanta,
Georgia - was a study of six urban drainage basins within the City of
Atlanta, Georgia, served by combined and separate sewers, to determine
the major pollution sources during storm events. Rainfall frequency
analysis and simulation techniques were utilized to obtain design cri-
teria for alternative pollution abatement schemes. Measurements of
flow were made of three major overflows, as well as of the interceptor
sewers originating at these points of overflow. Three branches of
South River were measured, as was the South River main stream at four
points. A bypass at the interceptor entering one of the wastewater
treatment plants was measured.
Data were collected from January 1969 until April 1970. Continuous
flow monitoring was maintained at the river and its tributary stations,
and in the interceptors where dry weather flow characteristics were of
interest. Event monitoring only was conducted of overflows and of the
bypass to the treatment plant. Rating curves for all gaging stations
were developed by stage-discharge measurements with current meters.
Either Price Type AA or Pigmy Type current meters were used. Some
discharge measurements were verified by alternate methods or formulas,
but results of these verifications were not given.
355
-------
Stevens Type F water level recorders were used throughout. Gaging sta-
tions were reported to be constructed in accordance with established
U.S. Geological Survey practice. For flow level recording at intercep-
tors, scow floats were installed at manholes a short distance downstream
from regulators. Although probable accuracy of the records collected
was not reported, an indication of their accuracy is available, based
on one of the gaging stations in the project which was operated by the
U.S. Geological Survey. This station, having a drainage area of
3.86 sq km (1.49 sq mi), had been operated since October 1963, or for
more than six years. The greatest flow measured by current meter during
the period was 5947 £/s (210 cfs), but the rating curve was extended
to 23,220 £/s (820 cfs) by computation of flow through a culvert.
Records at the station are stated by the USGS as poor, with no
quailification.
Stormwater Problems and Control in Sanitary Sewers - Oakland and
Berkeley, California - was an engineering investigation conducted on
Stormwater infiltration into sanitary sewers and associated problems in
the East Bay Municipal Utility District, Special District No. 1, with
assistance from the cities of Oakland and Berkeley, California. Rain-
fall and sewer flow data were obtained in selected study sub-areas that
characterized the land used patterns predominant in the study area.
Results obtained were extrapolated over large areas. Palmer-Bowlus
flumes were installed at three of the ten metering stations established
specifically for the study. These flumes were constructed of stainless
steel and were designed to fit the respective sizes and shapes of the
sewers. They were mounted in the outlet sewer from the manholes so
that head measurements could be made at the proper distances upstream
from the throats. Wooden channel extensions through the manholes were
installed to prevent water spreading out in the manhole as depth in the
sewer increased.
At seven of the new metering stations, 90-degree, V-notch weirs were
installed. These were constructed of marine plywood and covered for
356
-------
additional water resistance with two coats of polyurethane finish.
For ease of installation, a channel closure was provided so that it
could be easily slipped down into the flowing sewage and quickly bolted
in position after installing and sealing the actual weir plate.
Stevens type 2A35 water stage recorders were used at three weir loca-
tions where submergence of the weir was anticipated. These recorders
were selected for the ability to record two liquid levels simultaniously,
upstream and downs'tream from the weir. Taylor pressure recorders were
installed at the other seven new metering stations. With these re-
corders, liquid level sensing consists of measuring the back pressure
from a continuously purging nitrogen gas bubble system. In several
cases, equipment was installed in manholes near the centers of streets,
thus complicating the installation and use of equipment. Otherwise,
no problems were noted with the use of the equipment at the newly
established stations.
The flow rate at two pumping stations was determined by means of a
system-head curve for the station which gives the discharge rate for
each pump or combination of pumps. A recording ammeter was attached to
the pump electrical leads to indicate the total number of pumps running
at any given time. Relief overflow at one of these pumping stations
was measured by means of a broad-crested weir and a wet-well liquid
level recorder. A third pumping station was equipped with both a wet-
well liquid level recorder and a flow recorder. The primary device for
the flow recorder was a venturi tube mounted on the discharge manifold
of the pumps. The influent pumping station at the water pollution con-
trol plant was equipped with individual flowmeters on each pump dis-
charge which were connected to a combined flow recorder. The primary
devices for the flow recorder were discharge weirs in the grit chambers
which reflect the respective pump discharge rates.
A pumping station relief overflow structure was provided with a flow
measuring device for measuring the volume of water that overflowed.
The flow measuring arrangement consisted of measuring the liquid levels
357
-------
on both sides of a rectangular tide gate and extracting the flows from
the manufacturers rating curve. Because of difficulties in installa-
tion and operation, no usable flow measurements were obtained during
periods of overflow.
Dispatching System for Control of Combined Sewer Losses - concerned a
regulator control system which is said to demonstrate impressive reduc-
tions in combined sewer overflow pollution of the Mississippi River in
Minneapolis and Saint Paul. The project Includes a computer-based data
acquisition and cpntrol system that permits remote control of modified
combined sewer regulators. Data from rain gages, regulator control
devices, trunk sewers and interceptors, and river quality monitors
provide real-time operating information.
Water surface elevations in the system were monitored at about
48 points by the installation of bubbler gages with telemetry units
for transmitting data to a central location. This equipment provided
information on the frequency and duration of overflows. Because flow
rate was not measured, data on the volume of overflows were not thus
determined.
Flow in each of the three Minneapolis interceptors at the
Minniapolis-St. Paul city line was metered by dual venturi meters in
each interceptor. This equipment, which was in use prior to the sub-
ject study, provides information on the effectiveness of the program to
use the interceptors for temporary storage of combined sewage. Flow
data in the interceptors served to provide a check on the accuracy of
rainfall runoff modeling. Probable error of flow measurement by the
venturi-meters was not discussed.
Preconstruction Evaluation of Combined Sewage Detention Facilities -
was a lengthy study of combined sewer flows in Somersworth, New Hampshire
prior to construction of detention facilities. In order to get reason-
ably accurate and reliable flow measurements a section of the outfall
358
-------
was replaced with a weir pit large enough to provide a fair amount of
stilling behind the weir. This weir pit was 1.8 x 1.8 x 7.5m
(6x6x24.75 ft), and the weir was located 5.6m (18.33 ft) from the inlet.
Three different 0.63 cm (1/4 in.) thick steel plate rectangular weirs
with crest lengths of 0.3, 1.2, and 1.7m (1, 4, and 5.6 ft) were used,
the 1.2m (4 ft) one being employed for all but three months of the
year-long study. The design was such that the different weirs could be
changed easily in several minutes. The main difference in the 1.2m
(4 ft) weir was that its crest was elevated 1.2m (4 ft) above the floor
of the weir pit as opposed to 0.76m (2.5 ft) for the other two. This
weir was constructed after initial operation with the 0.3 and 1.7m
(1 and 5.6 ft) weirs, and observation of high flow rates using the
5.6' weir indicated that it was desirable to increase the stilling in
the weir pit by raising the crest height.
Head measurements over the 0.3 and 1.7m (1 and 5.6 ft) weirs were made
using an air operated Fischer and Porter recorder. A float operated
transmitter and recorder manufactured by the Penn/Measure-Rite
Division of Badger Meter Manufacturing Company was used for head
measurements over the 1.2m (4 ft) weir. Both recorders were set up to
use 24 hour, 30.5 cm (12 in.) diameter charts. The Fischer and Porter
recorder charts had a range of 0-51 cm (0-20 in.) of head, and the
Penn/Measure-Rite Recorder charts had a range of 0-76 cm (0-30 in.).
Flow into the weir pit was such that some turbulence was created.
This caused the flow recorders to print out head measurements in short
vertical strokes instead of a smooth line, a condition that was cor-
rected as much as possible by using the 1.2m (4 ft) weir. Sludge build-
up behind the weirs did not appear to upset their hydraulic
characteristics. However, before any sampling programs were under-
taken, the sludge build-up was totally removed to obtain accurate
chemical and biological characteristics of the combined sewage flow.
The actual sludge build-up would occur within a couple of days after
359
-------
the installation of a weir. The majority of this sludge consisted of
grit with a small percentage of organic matter.
Head measurements were converted to flows using "appropriate formulae
for the rectangular weirs used in the weir pit." By this it is assumed
that the Kindsvater-Carter equation was used rather than the Francis
formula, which would require corrections for both less than standard
contraction and velocity of approach much of the time. For example,
at the maximum head of 0.6m (2 ft) on the 1.2m (4 ft) weir (any greater
head would overflow the 1.8m deep weir pit) the Kindsvater-Carter
equation indicates a discharge of around 91 MLD (24 MGD), whereas the
uncorrected Francis formula yields approximately 83 MLD (22 MGD), a
10% difference. The maximum discharge for the 1.7m (5.6 ft) weir, as
computed assuming the recorder's full 51 cm (20 in.) of head was used,
is approximately 98 MLD (26 MGD). Maximum flow rates recorded with the
1.2m (4 ft) weir of from 99 to 148 MLD (26.2 to 39.0 MGD) are reported
but not explained. It is possible that the 148 MLD (39.0 MGD) dis-
charge was estimated but not so indicated. The point is that high
accuracies should not necessarily be expected with the given installa-
tion at the higher flow rates.
Urban Storm Runoff and Combined Sewer Overflow Pollution - Sacramento,
California - was a program to develop a general method for determining
the extent of pollution resulting from stormwater runoff and combined
sewer overflows occurring in an urban area, and the application of this
method to the City of Sacramento, California. Combined sewage and
stormwater in the system were characterized by collecting samples and
measuring flows at each of 19 sampling locations during six wet weather
periods. The intention was to collect (as nearly as practical) at the
commencement of rainfall, three hours thereafter, and approximately
12 to 18 hours after the commencement of sampling. However, comparison
with rainfall records indicated that the first data of each storm period
were not collected until the time of maximum rainfall intensity, or
later.
360
-------
The wastewater flows were established at manhole sampling stations by
use of the Manning formula. The coefficient of roughness was assumed
to be 0.013, a design value used by the City of Sacramento Engineering
Department. The value of S used was the measured slope/of either the
water surface or the invert. Because of difficulties in making the
»
required measurements for determination of slope, flow data at three of
the stormwater runoff sites are not considered to be reliable. None of
the computed wastewater flows at manhole sampling stations was checked
by means of flow measurement equipment. For design purposes, the peak
stormwater runoff flow in each of the individual pipes comprising the
Sacramento collection and conveyance system was estimated from rainfall
records by use of the rational method. These estimated flows for a
full pipe condition could be checked at three locations with computed
flows at the manhole sampling stations. They differ with the computed
flows by -6, +4, and -32 percent. This agreement is exceptional, par-
ticularly as these are the three locations where the computed flows
are not considered to be reliable.
Storage and Treatment of Combined Sewer Overflows - was a.project to
demonstrate the feasibility and economic effectiveness of a combined
wastewater overflow detention basin. Overflow from the combined sewer
area to the detention basin was measured by a Palmer-Bowlus flume
installed in a 198 cm (78 in.) reinforced concrete pipe. The flume
was designed as outlined by Ludwlg and Ludwig (1951). It was fabri-
cated of steel, and installed in the concrete pipe as it was laid.
Space between the flume and pipe was grouted. A Stevens Type A35
water-level recorder with a scow float was installed in the pipe to
measure head on the flume. The chart time scale was 73,2 cm (28.8 in.)
per day and the gage scale was 1:6. The volume discharged into the
basin was computed from the recorded head on the flume and a rating
curve for the flume, which was approximated in four linear sections.
Discharges indicated by rainfalls on eleven different occasions were
missed due to the recorder being out of order, amounting to eleven per-
cent of the runoff events that were missed.
361
-------
Overflow from the basin to the river was measured by a 6.7m (22 ft)
long sharp crested weir located in the overflow structure. The weir
head was measured by a Stevens Type A35 water-level recorder with a
cylindrical float. The chart time scale was 24.4 cm (9.6 in.) per day
and the gage scale was 1:6. The recorder was out of order during one
of the three overflow events which occurred during the period of project
operation.
Wastewater Flow Measurement in Sewers Using Ultrasound - concerns the
use of newly-developed ultrasonic velocity measurement equipment in
conjunction with ultrasonic level measurement equipment for the measure-
ment for sewage flow. Each of two existing combined sewers in the
Milwaukee, Wisconsin sewerage system, one 3.8m (12.5 ft) and the other
1.5m (5 ft) in diameter, were thus equipped initially. Subsequent
discovery of an excessive amount of entrained air at the larger diam-
eter sewer site necessitated the transferral of that equipment to a
more favorable location upstream in a 3.7m (12 ft) diameter sewer.
Performance of the ultrasonic meters was compared with other metering
devices at each of the locations. Relationships between average volume
flow, water level, and average velocity along selected horizontal
chords of the sewer cross sections were determined. The unit installed
in the samller sewer had operated for 18 months without failure and
required only routine maintenance. Similarily, the relocated unit in-
stalled at the 3.7m (12 ft) sewer had operated without failure since
its installation. No deterioration of the ultrasonic transducer probes
had been detected, indicating their suitability for use in the sewer
environment. The electronics of the ultrasonic velocity metering unit
were modified to include peak protection, automatic gain control, and
automatic trigger control to minimize the effects of variations in the
solids loading. Further observations concerning use of the ultrasonic
equipment are as follows:
a. Similar equipment has been used in Japan to successfully
measure full pipe flow of return sludge with high solids
362
-------
loadings. For practical line diameters, say from 0.2 to
5m (0.5 to 16 ft), no operational limitations due to sus-
pended solids would be expected.
b. Entrained air bubbles were found to cause operational
problems because of dissipation of the ultrasonic pulse
due to scatteririg by the bubbles. Therefore, it is
recommended that measurement sites be selected which are
«
reasonably free of severe upstream agitation that would
cause air entrainment.
c. Difficulties with level gage performance resulted from
standing ripples in the sewage surface which interfered
with echo returns. This was alleviated by moving the
level sensor a few feet to a point where the sewage sur-
face was more still.
d. Comparative figures of flow as measured by the ultrasonic
equipment with those measured by other metering devices
are not given for the demonstration sites described.
e. Total system cost for each site was about $15,000. Fu-
ture simplifications of the ultrasonic circuitry made
possible through more extensive use of integrated cir-
cuits have the promise of reducing this system cost by
a factor of two or three.
Biological Treatment of Combined Sewer Overflow - Kenosha, Wisconsin -
concerns the design, construction, operation and two year evaluation
of a biological process used for the treatment of potential combined
sewer overflow. During 1970, while design and construction of the
demonstration system facilities was occurring, a program to determine
the quality of the combined sewer overflows in Kenosha was carried out.
363
-------
This included measurement of rainfall, combined sewer overflow quantity
and quality, and influent quality to the water pollution control plant
during rainfalls.
Flow measurement equipment was installed in the outfall lines of three
overflows, known as the 57th Street, 59th Street, and 67th Street over-
flows. Bubbler-type depth recorders were used. Depth-discharge rela-
tionships were developed for the three overflow lines by means of dye
tests. However, no details of the test procedures were given. As a
result of an unsuccessful attempt to correlate rainfall volumes with
overflow volumes, it was disclosed that the overflow measuring devices
were of questionable accuracy. In some cases, the volume of overflow
measured exceeded the volume of rainfall over the combined sewer area.
Apparently, the depth-discharge relationships were not accurate and so
the overflow data were not used.
Measuring sites at the 57th Street and 59th Street overflows were aban-
doned in 1972. The depth recorder at the 67th Street site was moved
upstream above a weir diverting flow to the interceptor sewer. The
end of the bubble tube was placed just upstream from the overflow
mechanism, and a formula for broad-crested weirs was used to convert
level over the weir into flow rates and volumes. Flows computed in
this manner were used to estimate the total volume of overflow to the
demonstration treatment plant.
Surge Facility for Wet and Dry Weather Flow Control - concerns a 3-year
demonstration project which encompassed the design, construction, opera-
tion, testing, and evaluation of a surge facility designed to provide
flow equalization and some degree of treatment to all storm flows and to
provide rate control of all wet weather and dry weather wastewater flows
to interceptor sewers. The principle elements of the facility are a
sedimentation-equalization basin, a clarifier, a storage pond, a chlo-
rine contact basin, and a sludge digester. Flow and hydraulic measure-
ments include: (a) influent to the sedimentation-equalization basin;
364
-------
(b) underflow from the sedimentation-equalization basin to the clarifier
and the storage pond; (c) overflow from the sedimentation-equalization
basin directly to the storage pond; (d) water surface elevation of the
sedimentation-equalization basin; and (e) effluent to the chlorine
contact basin and the receiving stream.
The influent metering structure is a Parshall flume with a 0.30m (1 ft)
throat width having a maximum capacity of approximately 30 MLD (8 MGD).
The influent flow transmitter to the control building can be used to
control a flow proportional sampler of the influent. Underflow from
the sedimentation-equalization basin is measured with a 15.2 cm (6 in.)
magnetic flowmeter. The underflow can be set at any desired rate up
to 8.7 MLD (2.3 MGD). Overflows from the basin are measured by four
sharp-crested rectangular weirs, totaling 2.54m (100 in.). Head on the
weirs, and the water surface fluctuations in the sedimentation-
equalization basin, are monitored with a Stevens Type F water-stage
recorder. Time gears were selected to give an 8-day chart, and stage
gears were selected to give an indication of 0.06m/cm (0.5 ft/in.) of
chart or 0.012m/cm (0.1 ft/in.) of chart, depending on the depth
variation expected during any particular testing period. Effluent
from the facility is measured with a combination of 15.2 cm (6 in.)
Parshall flume and a 137 cm (54 in.) sharp-crested rectangular weir.
The two flow measuring,devices are set to give a combined capacity of
22.7 MLD (6 MGD). The effluent flow transmitter can be used to control
a flow proportional sampler of the effluent.
No discussion of problems with flow measurement equipment is given in
the report. Flow data presented in the report appear to be complete
and accurate.
Joint Construction Sediment Control Project - was a demonstration proj-
ect concerned with sediment control. As part of the project, a gaging
and sampling program was conducted to determine the effects of urbaniza-
tion on storm runoff and water quality of natural areas. Four automatic
365
-------
flow gaging and water sampling stations were installed on small streams
of the study area. Two of the stations were installed adjacent to each
other on streams just upstream from their junction. One of these
streams drains an experimental watershed and the other drains an adja-
cent reference watershed. Two other gaging and sampling stations were
established immediately upstream and immediately downstream from a
1.6 Hectare (4 acre) pond. At three of the gaging sites, precalibrated,
broad-crested, V-notch weirs developed and tested by the U.S. Department
of Agriculture were installed. At two sites, the concrete weir caps
have 2:1 side slopes; at the third site, side slopes of the weir cap
are at 3:1. At the gaging site downstream from the four-acre pond, a
sharp-crested, compound, 90°, V-notch and rectangular weir was
installed.
A Stevens Duplex Water-Level Recorder Type 2A35 was used to simultane-
ously record water levels at the two adjacent stream sites. A
Stevens Type A35 water level recorder was used upstream from the pond,
and a Belfort liquid level recorder was used at the downstream site.
A Gurley pygmy current meter was available to perform additional stream
gaging, but was used primarily to check the calibration of the perma-
nently installed weirs.
The weirs and level recording devices used are said to have proven to
be accurate, reliable, and easy to maintain. However, it was found
necessary to clean out sediment above the three U.S.D.A. weirs after
each storm and sometimes under base flow conditions to maintain the
calibration and accuracy of the weirs. No cleanout was required at
the compound weir downstream from the pond. Because of the reported
accumulation of sediment found after each storm, accuracy of runoff
records at the three sites during periods of storm runoff may be ques-
tioned. An unknown pattern of alternate sediment accumulation and
366
-------
flushing during the storm runoff period could occur. Calibration may
have been incorrect during the falling limb of the hydrograph, when
sediment was accumulating.
Combined Sewer Overflow Abatement Plan, Des Moines, Iowa - was a proj-
ect designed to provide engineering information regarding the volume,
character, and impact of combined sewer overflows and urban stormwater
discharges from a typical midwestern metropolitan area. Several dif-
ferent types of equipment and methods were used to measure flows and to
provide secondary stage measurements. Weirs of the 90° V-Notch type
and rectangular weirs, both suppressed and contracted, were used. At
sites with larger volumes of flow, a stage-discharge relationship was
determined using current meters. Natural controls of a permanent
nature were found at these sites. The Manning formula, with water-
surface slope and an "n" value of 0.018, was used to determine discharge
in one conduit. The main outfall of the wastewater treatment plant
was measured with a raw sewage flow totalizer of an undisclosed type.
Stick gages painted with water soluble paint were used to detect over-
flow occurrences. Both Stevens Type F Recorder, Model 68, and project-
designed and fabricated compressed-air bubbler recorders were used to
record water levels.
Dry-weather sanitary sewer flows were measured in eight sewers and in
the main outfall of the wastewater treatment plant. Weirs were used
in all sewers and float recorders were used in seven of them. A bub-
bler recorder was used in the eighth. No problems in measuring dry-
weather flows with this equipment were reported, although no check on
j
accuracy is available. Because of the extended period of high river
stage, flow measurements during wet weather were generally not obtained
in the sewers or overflow points. During this period, most of the flow
measuring sites were either submerged or at least intermittently af-
fected by high water. Overflows at one point were estimated by use of
the Manning formula, and at another point by use of a current meter.
367
-------
Stonnwater runoff was measured at four gaging sites on storm sewers.
At three sites, flow was measured by means of weirs and bubbler re-
corders. During periods of high river stage, one of these sites was
submerged or affected by backwater, when no record was obtained. Run-
off at the fourth site was determined from a stage-discharge relation-
ship established by current meter measurements. A concrete sewer line
crossing the channel provided a permanent type control. The rating
curve was extended an undisclosed amount above the highest current
meter measurement.
Computer Management of a Combined Sewer System - concerns a computer-
controlled "total systems management" complex, which affects much of the
combined sewer system of Seattle, Washington. Computer-augmented treat-
ment and disposal (CATAD) takes advantage of storage in the sewers to
limit overflows, and selects overflow points based on water quality
data. Development of the control system included the installation of
36 remote sensor stations. Work continues on a fully automatic, opti-
mizing model to program decisions so the system can maintain an 80%
overflow reduction.
Water levels at many locations are probably the most important single
category of information used to calculate flows. By incorporating
Manning's equations, various orifice and weir formulae, and pump effi-
ciency curves, it is possible to calculate flow at almost any location
in the system. Monitoring, control, and modeling of the entire system
depends upon flow information from many locations, some of which cannot
be obtained from water level measurements alone. In the CATAD system,
other on-line flow measuring techniques employed are;
a. Flow measuring weirs are installed at various locations.
b. A calibrated propeller type flowmeter is monitored at the
West Point Treatment plant site.
368
-------
c. Force main pressure calibration methods are used at pump
stations where there is sufficient friction head loss to
calibrate the force main at various flow ranges.
In addition, more than 100 Palmer-Bowlus flumes are installed in man-
hole locations. A computer program has been applied for rating these
measuring flumes, using data including sewer diameter, shelf height,
flume side slopes, and elevation of vertical side slopes.
Level sensors used in the system are generally pneumatic bubblers with
back pressure read by a differential pressure transmitter. The accu-
racy of measurements was checked by direct level measurements to bring
instrument calibration of the various stations into agreement with the
system datum. It was determined that the overall accuracy probably ap-
proaches about 2% of the full scale measurement. Project experience
demonstrated that manufacturer's pump unit performance curves may be
used to calculate reliable flows provided that critical analog sensors,
particularly pump speed sensors, are reliable. Because flows calcu-
lated using performance curves had been considered dubious, force main
pressure sensors were installed at many pump stations to provide alter-
native means of calculating station discharge. As a result of checking
the pressure gages using the salt velocity method, it was found that
flows thus calculated are not entirely reliable due to rapid fluctua-
tion of analog pressure values.
Deficiencies in the flow calculation procedures for regulator stations
were revealed. No allowance had been made for the effect of intercep-
tor backwater affecting the tailwater at a regulator ga'te and no tran-
sition had been provided between fully-submerged and free discharge
conditions. A backwater allowance was added, and a method of calculat-
ing the degree of gate submergence was, developed.
369
-------
Characterization and Treatment of Urban Land Runoff - was a project to
characterize the runoff from a 4.3 sp km (1.67 sq mi) urban watershed
in Durham, North Carolina with respect to annual pollutant yield. The
U.S. Geological Survey operates a continuous stage recorder and two
digital punch tape precipitation recorders within the basin. Stream-
flow control is provided by a shallow V-notch weir located on
Third Fork Creek some 21m (70 ft) downstream from a bridge culvert and
11.3 km (7 mi) upstream from the mouth of Third Fork Creek. Water
quality samples were taken from the center of the stream approximately
1.5m (5 ft) below the weir. Thus, rainfall, runoff, and water quality
data were gathered for the basin and analyzed. The USEPA Storm Water
Management Model (SWMM) was also evaluated with respect to actual condi-
tions as measured in the field and "was judged to predict peak hydro-
graph flows and total hydrograph volumes with reasonable accuracy;
however it was not judged effective for predicting pollutant
concentrations".
In order to assess the impact of varying types of land use within the
basin on urban runoff quality, five storms were manually sampled at
sub-basin discharge locations. A control section, usually a pipe or
box culvert, was utilized with Manning's equation to arrive at stage-
discharge relationships for each basin sampled. The stage was manually
read when a sample was taken. No accuracy estimates are available.
370
-------
SAMPLING
The collection of a wastewater sample that Is representative of the
source in all respects of interest is a frequently underrated task; a
situation that is unfortunate in view of the complexities involved,
especially if suspended solids are present, as will be the case with
most stormwater flows. This section, whose contents are mostly taken
from Shelley (1974, 1975a, b, c, d) and Shelley and Kirkpatrick (1973),
briefly treats the subject and automatic equipment for the purpose.
MECHANICS OF POLYDISPERSE SYSTEMS
The mechanics of polydisperse systems such as stormwater flows are among
the most complex and least thoroughly understood of all aspects of
science. This is not surprising when one considers that it covers
dynamic processes ranging from such sedimentology subjects as the movement
of soil to the thixotropic world of colloid chemistry. With complete
descriptions having to account for such topics as electrokinetics,
descriptive and structural rheology, sorption, flocculation, diffusion,
and Brownian motion as well as hydraulic system influences, it is little
wonder that empirical progress has outpaced analytical descriptive efforts.
Even under well controlled laboratory conditions, the study of suspended
solid laden flows remains very difficult, with considerable data scatter
the rule and statistical treatment of results being usually required.
The. point of all of this is not to suggest that any attempt to seriously
study the subject is doomed to failure but, rather, to point out that
one should not approach it as though it is a + 1 percent cosmos.
Table 6 has been prepared to give a better appreciation of suspended
solids orders of magnitude and characteristics. Eight decades of parti-
cle sizes are covered, and nominal dimensions are given in millimeters,
microns, and angstroms as some readers may have a better appreciation
371
-------
TABLE 6, PARTICLE ORDERS OF MAGNITUDE AND CHARACTERISTICS
Millimeters
10
10"
10
10
-3
10
-4
10
-5
10
-6
10
- 7
Microns
10,000
1,000
100
10
0.1
0.01
0.001
0.0001
Angstroms
10°
10
10C
10'
10s
10-
10
Relative
Weight of
a Particle
Relative Fall
Velocity
10J
4.8
10
-3
10
-6
10
-9
10
-12
10
-15
10
-18
5.6xlO~2 5.6X10""4 5.6xlO~6 5.6xlO"8 5.6x10 10 5.6xlO"12
10
-21
Relative Brownian
Displacement
10J
10
10-
10
10-
10
CO
•VJ
ro
Classification
s and
silt
clay
ultra clay solution
Phase
Appearance
Observed w/
Separated w/
Form after
evaporation
bed load
on bottom
naked eye
screen
granular
coarse suspension
very cloudy
naked eye
filter paper
loose powder
coll. susp.
turbid
microscope
clay filter
powder or
gel
colloidal solution
virtually clear
electron or ultra-
microscope
ultraf liter
gel
molecular
soln .
clear
-
-
crystal
Notes :
3.
mm in diaraeter essen-
1. The fall velocity of a 1 mm diameter quartz sphere is 16.0 cm/sec. Similar particles much below 10
tially do not settle, ^heir fall velocities being better stated in centimeters per century.
2. The time for average Brownian displacement of a 2 mm diameter sphere by one centimeter is around 7,300 yr. Brownian motion
-2
starts as a practical consideration for diameters smaller than 10 mm.
The resolution Unit of an ordinary microscope is around 2000A compared to 10A for electron or ultramicroscopes.
4. The size of particles passing the finest practical sieves {300 mesh) is around 0.05 nm. The limit of an ultrafilter is
approximately 10 nir.
-------
for size in one set of units rather than another. The changes in par-
ticle weight (or mass or volume) with size are indicated relative to a
1 mm particle. Changes in hydraulic size with mean particle diameter
are indicated relative to a quartz sphere (s.g. = 2.65) 1 mm in diame-
ter. Within the range of validity of Stokes' Law they vary with the
square of the diameter. The displacement due to Brownian motion rela-
tive to a 1 mm diameter particle is also indicated.
The major divisions of particle size classification set forth in Table 6
are indicated, as is the physical nature or phase of the mixture. Sev-
eral other characteristics of particle-water mixtures are also given,
including the visual appearance, methods of particle observations, sep-
aration techniques, and the form of the solids after evaporation.
There are characteristics of particles other than size which are of in-
terest to explain certain phenomena. For example, the charge on a par-
ticle attracts counterions and repels similiions. Some of the
counterions may be immobilized in the Stern layer, while the remainder
form a diffuse Gouy layer. Within this double layer, the decay of poten-
tial is measured by its thickness, which is very sensitive to the concen-
tration and valence of the counterions as is the surface change density.
Potentials at the particle surface, at the boundary between the Stern and
Guoy layers, and at the hydrodynamic plane of shear are of special
importance.
As an example, clay minerals usually have a negative charge. This may be
due to preferential adsorption of anions, especially hydroxyl ions;
cationic substitutions within the crystal lattice; and residual valences
(broken bonds) at particle edges (Postma; 1967). The double layer of
counterions (hydrated cations) described above balances the negative
charge. If the electrolytic potential (thickness of the double layer)
decreases below a critical value, coagulation occurs. In this process,
while the double layer is present, two clay particles approaching each
other by Brownian movement are repelled because their charges are equal.
373
-------
Other forces, however, may tend to cause the particles to approach each
other. The electrolytic content of the water is a strong factor here as
it affects the thickness (and hence electrolytic potential) of the double
layer. Increases in pollution may cause the flocculation process to be
completed, perhaps due to the presence of polyvalent cations in indus-
trial wastes and excessive amounts of particulate organic sewage, which
can act as a binding substance for fine-grained particles. On the other
hand, dissolved organic solids may inhibit flocculation, at least to
some degree.
The diameter of a floccule may be considerably larger than the diameters
of its constituent particles, with the result that the fall velocity of
flocculated clays is much greater than that in peptized form. Much water
is included in the floccules, however, with the result that fall velocity
increases are not as great as might otherwise be expected. For example,
as pointed out by Postma (1967), a unit particle with a specific weight
of 2.7 and a diameter of 5 microns has a fall velocity in seawater of
about 0.002 cm/s. A particle of the same density but a 500 micron diam-
eter would sink at a rate of approximately 20 cm/s. A floccule of clay
particles of the same size but containing 95 percent water, on the other
hand, would have a fall velocity of only 0.4 cm/s, the hydraulic size of
a 20 micron quartz sphere.
The transport of solid particles by a fluid stream is an exceedingly
complex phenomena and no complete theory which takes into account all of
the parameters has yet been formulated. It normally occurs as a combi-
nation of bed movement, saltation, and suspension. Although these are
interrelated, they are usually discussed separately because the phenom-
ena are not understood sufficiently to allow one to satisfactorily con-
sider them together.
The distribution of suspended solids or sediment in a transport stream Is
expressed in terms of concentration in one of two ways. Spatial concen-
tration is defined (FIASP; 1963) as "the quantity of sediment relative to
374
-------
the quantity of fluid in a fluid-sediment mixture". Thus, it could be
expressed as the dry weight of solids per unit volume of water-solids
mixture. There are three such concentrations that may be of interest:
spatial concentration at a point in the cross section, which is the con-
centration in a small volume at the point; spatial concentration in a
vertical, which is the concentration in a small water column extending
from the stream bed to the water surface; and spatial concentration in a
cross section, which is the concentration of the mixture contained in a
unit length of channel at the cross section. Turbidity, density, and
other fluid properties of the water-solids mixture are related to the
•
spatial concentration.
On the other hand, the discharge-weighted concentration is defined as the
quantity of suspended solids relative to the discharge of the fluid-solids
mixture. Thus, it could be expressed as the dry weight of solids in a
unit volume of discharge, or the ratio of the dry weight of solids dis-
charge to the weight of the water-solids discharge. Again there are three
such concentrations that may be of Interest: discharge-weighted concen-
tration at a point in the cross section which is the concentration in
the water-solids discharge through a small cross-sectional area at the
point; discharge-weighted concentration in a vertical, which is the con-
centration in the water-solids discharge through a unit width cross-
sectional area centered on the vertical; and discharge-weighted
concentration in a cross section, which is the concentration in the dis-
charge through the entire cross section. The discharge-weighted con-
centration may be multiplied by the overall stream discharge to obtain
suspended solids discharge.
Although a number of theories on the suspension of solids in flowing wa-
ter have been proposed, it is now generally recognized that it is di-
rectly related to turbulence of the flow as explained by Lane and
Kalinske (1939, 1941). In turbulent flow the instantaneous current vec-
tcr has vertical and lateral components as well as the horizontal one,
375
-------
and these are constantly changing in magnitude and direction with time in
a random fashion. As expressed in FIASP (1963) for sediment:
"Sediment carried in suspension is acted on in the vertical direction by
momentary currents which move upward or downward in the stream. Because
the water level in the stream remains unchanged, the quantity of upward
and downward flow must be equal. If the upward and downward currents
were the only forces affecting the vertical movement of sediment, com-
plete mixing would soon take place and the concentration of sediment
would become uniform throughout the depth. However, all particles of
specific gravity greater than that of water settle steadily downward.
Under the combined action of vertical currents and gravitational force,
a particle caught in a current moving upward at a rate greater than the
settling velocity of the particle should be transported upward, but if
it is suspended in water moving downward, or moving upward at a rate less
than its settling velocity, the particle should move downward. It might
seem that the downward currents would take down as much sediment as the
upward ones carry up, with the result that all the material finally
would settle to the bottom. However, as settling takes place the sedi-
ment concentration increases toward the bottom, and the upward currents
travel from a region of higher concentration to one of lower concentra-
tion, whereas for the downward currents the opposite relation prevails.
As the amounts of water moving upward and downward are equal and the
sediment concentration in the rising currents is potentially greater than
in the downward currents, more sediment must be acted upon by the rising
than by the falling currents. The settling action superimposed on the
fluctuating upward and downward currents tends to produce a balanced
suspension in which the rate of increase in sediment concentration toward
the bottom depends upon the degree of turbulence in the stream and the
settling velocity of the suspended particles."
Since coarse particles settle faster than fine particles of the same spe-
cific gravity and the vertical water motion due to turbulence will not
be appreciably different for particles of different sizes, equilibrium
376
-------
can only be obtained if the vertical concentration distribution varies,
with the concentration increasing more towards the bottom for coarse than
for fine particles. As particle size decreases, the vertical concentra-
tion gradient becomes less and less until it is essentially zero, the
case for most silts and clays. As an example, Figure 1 depicts typical
distribution curves for various sizes of sediment in the Missouri River
at Kansas City. Of course, for buoyant particles the reverse is true.
With all of the analytical difficulties discussed or alluded to in the
foregoing, it is obvious that there must be great reliance upon empir-
ical data in the study of the sampling of suspended solids. A wide va-
riety of sampling equipment designs is available (Shelley; 1974); but
none of them is universally acceptable for representatively sampling
all flows of interest (Shelley; 1975a), and differences in designs can
produce marked differences in results as reported by Harris and Keffer
(1974) and Shelley (1975c).
Let us consider the ability of a sampler intake probe to gather a rep-
resentative sample of dense suspended solids in the sediment range, say
up to 0.5 mm with specific gravity of 2.65. The results of a rather
thorough examination of relatively small diameter intake probes (0.63
and 0.32 cm) are given in FIASP (1941b). The argument is developed
that, for a nozzle pointing directly into the flow, the most represent-
ative sample of a fluid/suspended-solids mixture will be obtained when
the sampling velocity is equal to the flow velocity at the sampling
point. Using this as the reference criteria, investigations were con-
ducted to determine the effects of a) deviations from the normal sam-
pling rate, b) deviations from the straight-into-flow position of the
probe, c) deviations in size and shape of the probe, and d) disturbance
of sample by nozzle appurtenances. The effect of the sampling velocity
on the representativeness of the sample is indicated in Figure 2 which
presents the results for 0.45 mm and 0.06 mm sand. For the latter size,
377
-------
-.
:-
-,
I
h
r-
u
*>
i.
oz
i
SS
2
?
2 O b. f>
v
CONCENTRATION:- I SPACE - IOO P P.M BY WEIGHT
;
U
Figure 1, Vertical Distribution of Sediment in the Missouri River at Kansas City*
* Taken from FIASP (1948).
-------
CJ
^J
\0
l_>
z
o
«=£
00
10.0
0.4 0.5 0.6 0.7 0.8 0.9 1.0
2.0
3.0 40 5.0 6.0 7.0 80 9.0 IO.O
INTAKE VELOCITY
FLOW VELOCITY
Figure 2. Effect of Sampling Velocity on Representativeness of Suspended Solids*
* Data from FIASP (1941b).
-------
which falls within the Stokes' Law range, less than + 4 percent error in
concentration was observed over sampling velocities ranging from 0.4 to
4 times the stream velocity. For the 0.45 mm particles, the error at
a relative sampling rate of 0.4 was +45 percent, and at a relative
sampling rate of 4 the error was -25 percent.
For probe orientations up to 20 to either side of head-on, no appreci-
able errors in concentration were observed. Probe diameter and inlet
geometry (beveled inside, beveled outside or rounded edge) showed compara-
tively little effect on the representativeness of the sample. In
summary, it was found that for any sampler intake, facing into the stream,
the sampling rate is the primary factor to be controlled.
TYPES OF SAMPLES
In general, the selection of the type of sample to be collected depends
on a number of factors such as the. rates of change of flow and of the
character of the wastewater, the accuracy required, and the availability
of funds for conducting the sampling program. All samples collected,
either manually or with automatic equipment, are included in the following
types.
Descrete Samples
Discrete samples are those that are collected at selected intervals, and
each sample is retained separately for analysis.
Simple Composite Samples
Simple composite samples are those that are made up of a series of ali-
quots (smaller samples) of constant volume (Vc) collected at regular
time intervals (Tc) and combined in a single container. One could denote
such a sample by TcVc, meaning time interval between sucessive aliquots
constant and volume of each aliquot constant.
380
-------
Flow Proportional Composite Samples
Flow proportional composite samples are those collected in relation to
the flow volume during the period of compositing, thus indicating the
"average" waste condition during the period. One of the two ways of
accomplishing this is to collect samples of equal volume (Vc), but at
variable time intervals (Tv) which are inversely proportional to the
•
volume of the flow. That is, the time interval between samples is re-
duced as the volume of flow increases. Alternatively, flow proportion-
ing can be achieved by increasing the volume of each sample in
proportion to the flow (Vv), but keeping the time interval between
samples constant (Tc).
Sequential Composite Samples
Sequential composite samples are composed of a series of short-period
composites, each of which is held in an individual container. For ex-
ample, each of several samples collected during a 1-hour period may be
composited for the hour. The 24-hour sequential composite is made up
from the individual 1-hour composites.
Although the various types of composite samples are useful when an over-
all picture of "average" wastewater conditions is desired (e.g., moni-
toring industrial discharges, checking treatment plant efficiencies,
etc.), they do nothing to describe the pattern of changes that may occur
over the compositing period. Therefore, discrete sampling is generally
more useful for stormwater characterization, since it allows pollutant
parameters to be determined over a known time history and this provides
information about their variability with time.
PRACTICAL CONSIDERATIONS OF A SAMPLING PROGRAM
The adequacy of a sampling program depends largely on the optimum se-
lection of sampling sites. Both the program cost and its effectiveness
381
-------
in collecting samples representative of the character of sewer flovs
are seriously affected by the care exercised in site selection. Aware-
ness of the general character of sewer flows and of flow modes in storm
sewers and combined sewers, and knowledge of the variability of pollut-
ant concentration, leads to an understanding of how best to select sites
for sampling.. Some of the considerations in making such selections are:
a. Maximum accessibility and safety - Manholes on busy
streets should be avoided if possible; shallow depths
with manhole steps in good condition are desirable.
Sites with a history of surcharging or submergence by
surface water, or both, should be avoided if possible.
Avoid locations which may tend to invite vandalism.
b. Be sure that the site provides the information desired -
Familiarity with the sewer system is necessary. Know-
ledge of the existence of inflow or outflow between the
sampling point and point of data use is essential.
c. Make certain the site is far enough downstream from
tributary inflow to ensure mixing of the tributary
with the main sewer.
d. Locate in a straight length of sewer, at least six
sewer widths below bends.
e. Locate at a point of maximum turbulence, as found in
sewer sections of greater roughness and of probable
higher velocities. Locate just downstream from a drop
or hydraulic jump, if possible.
f. In all cases, consider the cost of installation, bal-
ancing cost against effectiveness in providing the
data needed.
382
-------
A decision as to whether manual or automatic sampling should be employed
involves more than just cost considerations. Experience has indicated
that operator training is necessary if manual sampling is to produce re-
producible results. Instances have been noted wherein two different op-
erators were asked to obtain a sample at a particular site with no other
guidance given. Analyses of samples taken at almost the same instants
in time have shown variances up to 50%. Other work conducted solely to
compare manual sampling methods has indicated such discrepancies in re-
sults that suspicion must be cast upon manual methods which involve dip-
ping of samples out of raw waste sources, and has raised questions
regarding the suitability of such manual grab sampling as a yardstick
against which to measure other techniques.
The preferred method of gathering manual samples from a raw waste
stream is to use a nonclogging submersible pump to actually extract the
fluid and tubing of appropriate size to transport it up to the sample
container. Pump and tubing sizes should be such that effective collec-
tion and transport of all suspended solids of interest is assured.
Both small, flexible impeller centrifugal pumps and progressive cavity
screw pumps have been successfully used with good repeatability of re-
sults. It should be noted, however, that the collection of flow pro-
portional composite or sequential samples can become quite tedious if
performed manually at the sampling site.
The decision to use automatic sampling equipment does not represent the
universal answer to wastewater characterization, however. For initial
characterization studies, proper manual sampling may represent the most
economical method of gathering the desired data. It is also prudent
from time to time to verify the results of an automatic sampler with
samples obtained manually. In this regard, because of the strict chain
of custody procedures that can be exercised with manually collected
samples, they can be used to support those data resulting from the use
of automatic sampling equipment. It should be noted, however, that
383
-------
more and more use is being made of automatic samplers as time goes by,
and this trend is expected to continue.
Presently available automatic sewage samplers have a great variety of
characteristics with respect to size of sample collected, lift capa-
bility, type of sample collected (discrete or composite), materials, of
construction, and numerous other both good and poor features. A number
of considerations in selection of a sampler are:
a. Rate of change of sewage conditions
b. Frequency of change of sewage conditions
c. Range of sewage conditions
d. Periodicity or randomness of change
e. Availability of recorded flow data
f. Need for determining instantaneous conditions, average
conditions, or both
g. Volume of sample required
h. Need for preservation of sample
i. Estimated size of suspended matter
j. Need for automatic controls for starting and stopping
k. Need for mobility or for a permanent installation
1. Operating head requirements.
Because of the variability in the character of storm and/or combined
sewage, and because of the many physical difficulties in collecting
samples to characterize the sewage, precise characterization is not
practicable, nor is it possible. In recognition of this fact, one
must guard against embarking on an excessively detailed sampling pro-
gram, thus increasing costs, both for sampling and for analyzing the
samples, beyond costs that can be considered sufficient for conducting
a program which is adequate for the intended purpose.
A careful study of costs should be made prior to commencing a program
of sampling, balancing cost against the number of samples and analyses
384
-------
required for adequate characterization of the wastewater. As the pro-
gram progresses, current study of the results being obtained may make
it reasonable to reduce or increase the number of samples collected.
ELEMENTS OF AN AUTOMATIC SAMPLER SYSTEM
In a system breakdown by functional attributes, an automatic liquid
sampler may be divided into five basic elements or subsystems. Each
of these will be discussed in turn.
Sample Intake Subsystem
The operational function of the sample intake is to reliably allow
gathering a representative sample from the flow stream in question.
Its reliability is measured in terms of freedom from plugging or clog-
ging to the degree that sampler operation is affected and invulnera-
bility to physical damage due to large objects in the flow. It is also
desirable, from the viewpoint of sewer operation, that the sample in-
take offer a minimum obstruction to the flow in order to reduce the
possibility of blockage of the entire pipe by lodged debris, etc.
The sample intake of many commercially available automatic liquid sam-
plers is often only the end of a plastic suction tube, and the user is
left to his own ingenuity and devices if he desires to do anything
other than simply dangle the tube in the stream to be sampled. Some
manufacturers provide a weighted, perforated plastic cylinder that
screens the hose inlet from the unwanted material that might cause
choking or blockage elsewhere within the sampler. Typical hole sizes
are around 1/3 cm (1/8 inch) in diameter and, if there are sufficient
holes to assure free flow, results have been satisfactory in some ap-
plications. Samplers that employ pneumatic ejection have their own in-
take chambers that must be used in order for the equipment to function
properly. There will be some sampling sites where the use of custom
designed intakes is indicated. However, their individual attributes
385
-------
will be a function of the flow characteristics; e.g., is the stream
trash or debris ladened, is there any flow stratification, are sus-
pended solids high, etc.?
Sample Gathering Subsystem
Three basic sample gathering methods or categories can be identified;
mechanical, forced flow, and suction lift. Several different commercial
samplers using each method are available today. The sample lift re-
quirements of the particular site often play a determining role in the
gathering method to be employed.
Mechanical Methods - There are many examples of mechanical gathering
methods used in both commercially available and one-of-a-kind samplers.
One of the more common designs is the cup on a chain driven by a
sprocket drive arrangement. In another design, a cup is lowered within
a guide pipe, via a small automatic winch and cable. Other examples in-
clude a self closing pipe-like device that extracts a vertical "core"
from the flow stream, a specially contoured box assembly with end clo-
sures that extracts a short length (plug) of the entire flow cross-
section, a revolving or oscillating scoop that traverses the entire
flow depth, etc.
Some of the latter units employ scoops that are characterized for use
with a particular primary flow measurement device such as a weir or
Parshall flume and extract an aliquot volume that is proportional to
the flow rate. Another design for mechanically gathering flow propor-
tional samples involves the use of a sort of Dethridge wheel with a
sample cup mounted on its periphery. Since the wheel rotation is pro-
portional to flow, the effect is that a fixed volume aliquot is taken
each time a certain discharge quantity has passed, and total discharge
can be estimated from the size of the resultant composite sample.
386
-------
The foregoing designs have primarily arisen from one of two basic con-
siderations. First, site conditions that require very high lifts have
dictated the use of mechanical gathering units due to the limitations
of suction lift pumps and space considerations. Some mechanical units
are capable of lifts of 61m (200 ft) or more. Second, the desire to
gather samples that are integrated across the flow depth has led to some
of the different mechanical approaches mentioned above. Unless vertical
velocity and pollutant gradients are quantified and accounted for, their
presence makes the results of such depth integrated samples question-
able, at least in a mass discharge sense.
One of the penalties that one must trade-off in selecting a mechanical
gathering unit is the necessity for some obstruction to the flow, at
least while the sample is being taken. The tendency for exposed mech-
anisms to foul, together with the added vulnerability of many moving
parts, means that successful operation will require periodic inspection,
cleaning, and maintenance.
Forced Flow Methods - All forced flow gathering methods require some
obstruction to the flow, but usually it is less than with mechanical
gathering methods. It may only be a small inlet chamber with a check
valve assembly of some sort or it may be an entire submersible pump.
The main advantage of submersible pumps is that their high discharge
pressures allow sampling at greater depths, thereby increasing the
flexibility of the unit somewhat insofar as site depth is concerned.
Pump malfunction and clogging, especially in the sizes often used for
samplers, is always a distinct possibility and, because of their loca-
tion in the flow stream itself, maintenance is much more difficult and
costly to perform than on above ground or more easily accessible units.
They also necessarily present an obstruction to the flow and are thus
in a vulnerable position as regards damage by debris in the flow.
387
-------
Pneumatic ejection is a forced flow gathering method used by a number
of commercial samplers. The gas source required by these units varies
from bottled refrigerant to motor driven air compressors. The units
that use bottled refrigerant must be of a fairly small scale to avoid
an enormous appetite for the gas and hence a relatively short operating
life before the gas supply is exhausted. Furthermore, concern has re-
cently been expressed about the quantities of freon that are being dis-
charged into the atmosphere. The ability of such units to backflush
or purge themselves is also necessarily limited. The advantages of few
moving parts, inherent explosion proof construction, and high lift ca-
pabilities must be weighed against low or variable line velocities, low
or variable sample intake velocities, and relatively small sample capac-
ities in some designs. Another disadvantage of most pnuematic ejection
units is that the sample chamber fills immediately upon discharge of
the previous sample. Thus, it may not be representative of flow condi-
tions at the time of the next triggering and, if paced by a flowmeter,
correlation of results may be quite difficult.
Suction Lift Methods - Suction lift units must be designed to operate
in the environment near the flow to be sampled or else their use is
limited to a little over 9m (30 ft) due to atmospheric pressure. Sev-
eral samplers that take their suction lift directly from an evacuated
sample bottle are available today. Vacuum leaks, the variability of
sample size with lift, the requirement for heavy glass sample bottles
to withstand the vacuum, difficulty of cleaning due to the requirement
for a separate line for each sample bottle, the necessity of placing
the sample bottles near the flow stream (and hence in a vulnerable
position), the varying velocities as the sample is being withdrawn,
etc., are among the many disadvantages of this technique
Other units are available that use a vacuum pump and some sort of meter-
ing chamber to measure the quantity of sample being extracted. These
units in some designs offer the advantages of fairly high sample intake
388
-------
and transport velocities, the fluid itself never comes in contact with
the pump, and the pump output can easily be reversed to purge the sam-
pling line and intake to help prevent cross-contamination and clogging.
Their chief disadvantage, shared with certain other suction lift de-
signs arises from the following consideration.
With all suction lift devices a physical phenomenon must be borne in
mind and accounted for if sample representativeness is to be maintained.
When the pressure on a liquid (such as sewage) which contains dissolved
gases is reduced, the gases will tend to pass out of solution. In so
doing they will rise to the surface and entrain suspended solids in
route. (In fact, this mechanism is used to treat water; even small
units for aquariums are commercially available.) The result of this is
that the surface layer of the liquid may be enhanced in suspended sol-
ids, and if this layer is a part of a small sample aliquot, the sample
may not be at all representative. In the absence of other mitigating
factors, the first flow of any suction lift sampler should therefore
be returned to waste.
A variety of positive displacement pumps have been used in the design
of suction lift samplers, including flexible impeller, progressive cav-
ity rotary screw, roller or vane, and peristaltic types. Generally
these pumps are self-priming (as opposed to many centrifugal pumps),
but some designs should not be operated dry because of internal wearing
of rubbing parts. The desirability of a low-cost pump that is rela-
tively free from clogging has led many designers to use peristaltic
pumps. A number of types have been employed including finger, nutating,
and two- and three-roller designs using either molded inserts or reg-
ular tubing. Most of these operate at such low flow rates, however,
that the representativeness of suspended solids is questionable. Newer
high-capacity peristaltic pumps are now available and should find appli-
cation in larger automatic samplers. The ability of some of these
pumps to operate equally well in either direction affords the
389
-------
capability to blow down lines and help remove blockages. Also, they
offer no obstruction to the flow since the transport tubing need not be
interrupted by the pump, and strings, rags, cigarette filters and the
like are passed with ease.
All in all, the suction lift gathering method appears to offer more ad-
vantages and flexibility than either of the others for a storm or com-
bined sewer application. The limitation on sample lift can be overcome
by designing the pumping portion of the unit so that it can be separated
from the rest of the sampler and thus positioned within 6m (20 ft) or so
of the flow to be sampled. For many sites, however, even this will not
be necessary.
Sample Transport Subsystem
The majority of the commercially available automatic samplers have
fairly small line sizes in the sample train. Such tubes, especially
at 1/3 cm (1/8 in.) inside diameter and smaller, are very vulnerable to
plugging, clogging due to the build-up of fats, etc. For many applica-
tions, a better minimum line size would be 1 to 1.3 cm (3/8 to 1/2 in.)
inside diameter.
For flows that are high in suspended solids, it is imperative that ade-
quate sample flow rate be maintained throughout the sampling train in
order to effectively transport them. In horizontal runs, the velocity
must exceed the scour velocity, while in vertical runs the settling or
fall velocity must be exceeded several times to assure adequate trans-
port of solids in the flow. Sharp bends and twists or kinks in the
sampling lines should be avoided if there is a possibility of trash or
debris in the lines that could become lodged and restrict or choke the
flow. The same is true of some valve designs. In summary, the sam-
pling train must be sized so that the smallest opening is large enough
to give assurance that plugging or clogging is unlikely in view of the
390
-------
material being sampled. However, it is not sufficient to simply
make all lines large, which also reduces friction losses, without
paying careful attention to the velocity of flow. For many appli-
cations, minimum velocities of 0.6 to 1 m/s (2 to 3 fps) would ap-
pear warranted, and even higher velocities are required for some
applications.
In few equipment designs the sample is delivered under pressure
all the way to the sampler container. It is more often the case,
however, that a break occurs at a distribution funnel or some sim-
ilar appurtenance, and flow from this point on into the sample
container is by gravity alone. Field experience with some such
designs has been less than satisfactory, with instances of clog-
gling, accumulation of fats, etc., being noted.
Sample Storage Subsystem
For storm and combined sewer applications, discrete sampling is
generally desired. This allows characterization of the sewage
throughout the time history of the storm event. If the samples
are sufficiently large, manual compositing can be performed based
on flow records or some other suitable weighting scheme. Although
the quantity of samples required will be a function of the subse-
quent analyses that are to be performed, in general at least
1 liter and preferably 2 liters will be desired. An additional
benefit arises because such relatively large samples are less vul-
nerable to errors arising from cross-contamination.
The sample container itself should either be easy to clean or dis-
posable. The cost of cleaning and sterilizing makes disposable
391
-------
containers attractive, especially if bacteriological analyses are
to be performed. Although some of today's better plastics are
much lighter than glass and can be autoclaved, they are not so
easy to clean or inspect for cleanliness. Also the plastics will
tend to scratch more easily than glass and, consequently, cleaning
a well-used container can become quite a chore. The food packag-
ing industry, especially dairy products, offers a wide assortment
of potential disposable sample containers in the larger sizes.
Both the 1.9A (1/2 gal) paper and plastic milk cartons can be con-
sidered viable candidates, and their cost in quantity is in the
pennies-each range.
The requirements for sample preservation are enumerated in MDQARL (1974) and
will not be repeated here except to note that refrigeration is
stated as the best single preservation and will, in all likelihood,
be required unless the sampling cycle is brief and samples are re-
trieved shortly after being taken. It should be mentioned, how-
ever, that if the samples are allowed to become too cold, they may
no longer be representative.
For example, destruction of the organisms necessary for the devel-
opment of BOD may occur, or freezing may cause serious changes in
the concentration of suspended solids. Light can also affect sam-
ples and either a dark storage area or opaque containers would seem
desirable. Unless disposable containers are used, however, it will
be difficult to inspect an opaque container for cleanliness.
Again the paper milk carton is attractive since not only is it rela-
tively opaque, but its top opens completely allowing visual inspec-
tion of its contents.
392
-------
Controls and Power Subsystem
The control aspects of some commercial automatic samplers have come
under particular criticism. It is no simple matter, however, to pro-
vide great flexibility in operation of a unit while at the same time
avoiding all complexities in its control system. The problem is not
only one of component selection but packaging as well. For instance,
•
even though the possibility of immersion may be extremely remote in a
particular installation, the corrosive highly-humid atmosphere which
will, in all likelihood, be present makes sealing of control elements
and electronics desirable in most instances.
The automatic sampler for storm and combined sewer application will, in
all likelihood, be used in an intermittent mode; i.e., it will be idle
for some period of time and activated to capture a particular meteoro-
ligical event. If field experience to date is any indication, the
greatest need for an improved control element is for an automatic
starter. While the sensor is not a part of the sampler proper, its
function is essential to successful sampler utilization. Although re-
mote rain gages, etc., can be used for sensing elements, one of the
most attractive techniques would be to use the liquid height (or its
rate of increase) to start a sampling cycle. This will avoid the dif-
ficulties associated with different run-off times due to local condi-
tions such as dryness of ground, etc.
The controls determine the flexibility of operation of the sampler, its
ability to be paced by various types of flow measuring devices, etc.
Built-in timers should be repeatable and time period should not be af-
fected by voltage variations. The ability to repeatedly gather the
required aliquot volume independent of flow depth or lift is very im-
portant if composite samples are to be collected. Provisions for man-
ual operation and testing are desirable as is a clearly laid out control
panel. Some means of determining the time when discrete samples were
393
-------
taken is necessary if synchronization with flow records is contem-
plated. An event marker could be desirable for a sampler that is to
be paced by an external flow recorder. Reliability of the control sys-
tem can dominate the total system reliability. At the same time, this
element will, in all likelihood, be the most difficult to repair and
calibrate. Furthermore, environmental effects will be the most pro-
nounced in the control system. The power switching function of the
control system may be required to deal with multiple switching of in-
ductive loads and must achieve the switching of these loads without
the possible damage associated with transfer of energy interruptions.
The above tasks can probably be best executed, in the light of the
current electronics state-of-the-art, by a solid state controller ele-
ment. In addition to higher inherent reliability, such an approach
will allow switching of high level loads in a manner that eliminates
RFI emissions and destructive results. In addition, the unit should
be of modular construction for ease of modification, performance moni-
toring, fault location, and replacement/repair. Such an approach also
lends itself to encapsulation which will minimize environmental effects.
Solid state switching eliminates the possibility of burned or welded
contacts either of which will cause complete sampler breakdown.
•
Solid state controllers can be easily designed with sufficient flexi-
bility to accept start commands from a variety of types of remote sen-
sors, telephone circuits, etc.
One of the attributes essential to the control system of an automatic
sampler to be used in a storm or combined sewer application is that it
be able to withstand power outages and continue its program. Such
power interruptions appear to be increasingly common as demand for
electricity continues to grow. Although desirable in some instances,
the provision of a random interrogate signal to be coupled with a se-
quence sample mode generates programming problems, especially when
coupled with power interrupt possibilities.
394
-------
The foregoing discussion as it relates to problems associated with
interruptions in electrical service is, of course, directed to samplers
that rely upon outside power for some aspect of their operation. The
need for high sample intake and transport velocities, larger sample
lines and capacities, together with the possible requirement for me-
chanical refrigeration make it unlikely that such a sampler can be
totally battery operated today. Although recent break-throughs have
resulted in 1 kw dry-cell batteries, their cost is prohibitive for
this sort of an application. Other approaches to self-contained power
such as custom designed wet-cell packs, diesel-generators, etc., while
within the current state-of-the-art, introduce other problems and com-
plexities that must be carefully weighed before serious consideration
can be given to their incorporation in an automatic sampler design.
REVIEW OF COMMERCIALLY AVAILABLE EQUIPMENT
The foregoing discussion of the elements of an automatic sampler can be
viewed as equipment requirements. In addition to these requirements,
there are also certain desirable features which will enhance the utility
and value of the equipment. For example, the design should be such that
maintenance and troubleshooting are relatively simple tasks. Spare
parts should be readily available and reasonably priced. The equipment
design should be such that the unit has maximum inherent reliability.
As a general rule, complexity in design should be avoided even at the
sacrifice of a certain degree of flexibility of operation. A reliable
unit that gathers a reasonably representative sample most of the time
is much more desirable than an extremely sophisticated complex unit
that gathers a very representative sample 10 percent of the time, the
other 90 percent of the time being spent undergoing some form of repair
due to a malfunction associated with its complexity.
It is also desirable that the cost of the equipment be as low as prac-
tical both in terms of acquisition as well as operational and mainte-
nance costs. For example, a piece of equipment that requires
395
-------
100 man-hours to clean after each 24 hours of operation is very undesir-
able. It is also desirable that the unit be capable of unattended opera-
tion and remaining in a standby condition for extended periods of time.
The sampler should be of sturdy construction with a minimum of parts ex-
posed to the sewage or to the highly humid, corrosive atmosphere asso-
ciated directly with the sewer. It should not be subject to corrosion
or the possibility of sample contamination due to its materials of con-
struction. The sample containers should be capable of being easily re-
moved and cleaned; preferably they should be disposable.
For portable automatic wastewater samplers, the list of desirable fea-
tures is even longer. Harris and Keffer (1974) give a number of features
of an "ideal" portable sampler which are based upon sampler comparison
studies and over 90,000 hours of field experience. Included were:
• Capability for AC/DC operation with adequate battery
energy storage for 120-hour operation at 1-hour sampling
intervals.
• Suitable for suspension in a standard manhole and still
provide access for inspection and sample removal.
• Total weight including batteries under 18 kilograms
(40 pounds).
• Sample collection interval adjustable from 10 minutes
to 4 hours.
• Capability for collecting both simple and flow-
proportional composite samples.
396
-------
• Capability for multiplexing repeated aliquots into dis-
crete bottles (i.e., sequential composite).
Intake hose liquid velocity adjustable from 0.61 to
3 m/sec (2.0 to 10 fps) with dial setting.
• Minimum lift of 6.1 meters (20 feet).
• Explosion proof.
• Watertight exterior case to protect components in the
event of rain or submersion.
• Exterior case capable of being locked and with lugs for
attaching steel cable to prevent tampering and provide
some security.
No metal parts in contact with waste source or samples.
• An Integral sample container compartment capable of main-
taining samples at 4 to 6°C for a period of 24 hours at
ambient temperatures up to 38°C.
• With the exception of the intake hose, capable of oper-
ating in a temperature range between -10 to 40°C.
• Purge cycle before and after each collection interval
and sensing mechanism to purge in event of plugging during
sample collection and then collect complete sample.
• Capable of being repaired in the field.
Although some types of automatic liquid sampling equipment have been
available commercially for some time, project engineers continue to
397
-------
design custom sampling units for their particular projects due to a lack
of commercial availability of suitable equipment. In the last few years,
however, there has been a proliferation of commercial sampling equipment
designed for various applications. New companies are being formed and
existing companies are adding automatic sampling equipment to their prod-
uct lines. In addition to their standard product lines, most manufac-
turers of automatic sampling equipment provide special adaptations of
their equipment or custom designs to meet unique requirements of certain
projects. Some designs which began in this way have become standard
products, and this can be expected to continue.
The products themselves are rapidly changing also. Not only are improve-
ments being made as field experience is gathered with new designs, but
attention is also being paid to certain areas that have heretofore been
largely ignored. For example, one company is introducing sampling probes
that allow gathering oil or various other liquids from the flow surface;
solid-state electronics are being used more and more in sampler control
subsystems; new-type batteries are offering extended life between charges
and less weight; and so on. Table 7 lists the names and addresses of
some 32 manufacturers who are known to offer standard lines of automatic
wastewater sampling equipment. In view of the burgeoning nature of this
product area, it is inevitable that some omissions have been made.
An overall matrix, which summarizes the equipment characteristics to fa-
cilitate comparisons, is presented in Table 8. There are several col-
umn headings for each sampler model (or class of models). "Gathering
Method" identifies the actual method used (mechanical, forced flow, suc-
tion lift) and type (peristaltic-, vacuum-, centrifugal-pump, etc.).
Depending upon the gathering method employed, the sample flow rate may
vary while a sample is being taken, vary with parameters such as lift,
etc. Therefore, the "Flow Rate" column typically lists the upper end of
the range for a particular piece of equipment and values significantly
less may be encountered in a field application. "Lift" indicates the
398
-------
TABLE 7. AUTOMATIC WASTEWATER SAMPLER MANUFACTURERS
Bestel-Dean Limited
92 Worsley Road North,
Worsley
Manchester, England M28 5QW
BIF Sanitrol
P.O. Box 41
Largo, Florida
33546
Brallsford and Company, Inc,
Milton Road
Rye, New York 10580
Brandywlne Valley Sales Co.
20 East Main Street
Honey Brook, PA 19344
Chicago Pump Division
FMC Corporation
622 Dlversey Parkway
Chicago, Illinois 60614
Collins Products Co.
P.O. Box 382
Livingston, Texas 77351
Environmental Marketing
Associates
3331 Northwest Elmwood Dr.
Corvallis, Oregon 97330
ETS Products
12161 Lackland Road
St. Louis, Missouri
63141
Fluid Kinetics, Inc.
3120 Production Drive
Fairfield, Ohio 45014
Horizon Ecology Company
7435 North Oak Park Drive
Chicago, Illinois 60648
Hydra-Numatic Sales Co.
65 Hudson Street
Hackensack, NJ 07602
Hydraguard Automatic
Samplers
850 Kees Street
Lebanon, Oregon 97355
Instrumentation Specialties
Company
Environmental Division
P.O. Box 5347
Lincoln, Nebraska 68505
Kent Cambridge Instrument
Company
73 Spring Street
Ossining, New York 10562
Lakeside Equipment Corp.
1022 East Devon Avenue
Bartlett, Illinois 60103
Manning Environmental Corp.
120 DuBois Street
P.O. Box 1356
Santa Cruz, California 98061
Markland Specialty Eng. Ltd.
Box 145
Etobicoke, Ontario (Canada)
Nalco Chemical Company
180 N. Michigan Avenue
Chicago, Illinois 60601
Nappe Corporation
Croton Falls Industrial Complex
Route 22
Croton Falls, New York 10519
N-Con Systems Company
308 Main Street
New Rochelle, New York 10801
Paul Noascono Company
805 Illinois Avenue
Collinsville, Illinois 62234
399
-------
TABLE 7. AUTOMATIC WASTEWATER SAMPLER MANUFACTURERS (Cont'd)
Peri Pump Company, Ltd.
180 Clark Drive
Kenmore, New York 14223
Phipps and Bird, Inc.
303 South 6th Street
Richmond, Virginia 23205
Protech, Inc.
Roberts Lane
Malvern, PA 19355
Quality Control Equipment
Company
P.O. Box 2706
Des Moines, Iowa 50315
Rice Barton Corporation
P.O. Box 1086
Worcester, MA 01601
Sigmamotor, Inc.
14 Elizabeth Street
Middleport, New York 14105
Sirco Controls Company
8815 Selkirk Street
Vancouver, B.C.
Sonford Products Corporation
100 East Broadway, Box B
St. Paul Park, MN 55071
Testing Machines, Inc.
400 Bayview Avenue
Amityville, New York 11701
Tri-Aid Services, Inc.
161 Norris Drive
Rochester, New York 14610
Williams Instrument Co., Inc.
P.O. Box 4365, North Annex
San Fernando, California 91342
400
-------
TABLE 8, SAMPLER CHARACTERISTIC SUMMARY MATRIX
Stapler
Bestel-Dean Hk 11
Beatel-Dean Crude
BIT 41
Brill, ford DC-F 1 EP
Brallaford EVS
Brailsford DV-2
BVS PP-100
BVS FE-400
BVS SE-800
BVS PPE-400
Chicago Fuap
Collin, 42
Collins 40
EDA 200
ITS rs-«
Borlion S7570
Horizon S7576
Borlion *7S78
Eydracuard Hf
Rydrae;uard A
Bjdri-Iuaatic
ISCO 1392
ISCO 1*80
ISCO 1580
Kent SSA
Kent SSI
Kent SSC
Lakeside 1-2
Manning s-4000
Marklaad 1301
Narklaod 101 4 102
Harkland 104T
lalco S-100
Rsppt Scries 46
Noavcono Shift
N-Con Surveyor II
K-Con Scout I I
N-Con Sentry 500
N-Con Trebler
N-Con Sentinel
S-sc reu type
M-cup on ch> in
S-pistoo type
S-plston type
F-pneuB.it ic
F-atubaerftlble pu»p
F-«ubaerslble puap
F-pnetia»t Ic
user supplied
user supplied
P-plston type
S-perlstsltic
S-peristsltlc
F-pncuBatic
F-pD«UB«t 1C
S-centrif ugal
S-perlataltic
S-perlst«lti~c
S-peristaltic
S-scrcv type
H-scoop
S~ vacua* pnap
P-pneuB«tlc
F-paeuss Stic
F-pncumat ic
F-*uba*r«ibl« puap
S-flcxlblt l»p«ll«r
S-fl«xlble lipallar
S-pirl3.78S
•v.5,000
Dnk.
%20
100
100
100
a
a
5.700
1,500
IA
1.400
150
200
33,000
«A
3,800
a
a
a
28,400
11,400
13.200
8
20.000
1)0
ISO
hA
63.000
Lift
(•)
6.1
6.1
4.9
<2
3.7
<2
85
9.8
9.8
85
HA
(A
HA
<1
8.8
9.1
9.1
9.1
>9
>9
4.6
7.9
7.9
7.9
4.9
4.0
5.0
0
6.7
18.3
18.3
18.3
7.6
1.8
4.6
9.1
1.8
5. 5
5.5
0
NA
Line
Sice
(»•)
6.4
19.1
25.4
4.S
4.8
4.8
3.2
12.7
12.7
3.2
25.4
2.4
2.4
9.5
6.4
0.8
0.8
0.8
6.4
6.4
6.4
6.4
6.4
6.4
6.4
6.4
25.4
12.7
9.5
6.4
6.4
6.4
12.7
6.4
9.5
4.8
6.4
6.4
6.4
12.7
25.4
Saapl«
Type
D, IcVc, IvVc
D. TcVc, TwVc
TcVc, TvVc
Continuous
TcVc, TvVc
TcVc, TvVc
TcVc, TvVc
D, TcVc, TvVc
TcVc, TvVc
TcVc, TvVc
TcVc, TvVc
Cont inuoua
Crab
TcVc
TcVv
TvVc
TcVc, TvVc
Discrete
D, TcVc, TvVc. S
TcVv
D, S
TcVc, TvVc
D, TcVc
TcVc, TvVc
TcVc
TcVc, TvVc
Cont inuoua
TcVc, TvVc
TcVc , TvVc
Sequential
TcVv
TcVc, TvVc
Instal-
lat ion
Portable
Portable
Fixed
Portable
Portable
Portable
Portable
Portable
Fl>ed
P or F
Fixed
P or F
P or F
Portabl*
Portable
Portable
Portable
Portable
Portable
Portable
Portable
Portable
Portable
Fixed
Fixed
Portable
Portable
Fixed
Fixed
Portable
Portable
Fixed
Portable
Portable
Portable
Portable
Fixed
Fixed
Unk.
link.
•n.ooo
296-373
520-672
373
853-1,525
1,500-2,510
5,650
1,450-3,350
2,600-3,200
985-2,478
835-2,328
199-456
1,095-up
•>-410
•\-220
595
246-541
286-668
1.800
1,095-1,498
645-1,020
750-1,130
1.240
2.35*
2,354
•v700-up
1.290
1,095-1.350
594-2,189
1,094-2,644
U«k.
225-285
1,100-1,800
Unk.
290-590
575-935
1,125-1,203
1,050-1,350
--2,600
AC/DC
AC
AC
DC
AC/DC
DC
AC/DC
AC /DC
AC
AC /DC
AC
AC
AC
AC /DC
AC
AC/DC
AC
DC
Air
Air 4 AC
AC
AC/DC
AC/DC
AC/DC
AC/DC
AC
AC
AC
DC
Air 4 DC
Air 4 DC
Air 4 AC
AC
AC /DC
AC
AC
AC
AC /DC
AC/DC
AC
AC
w
•
1
N
-------
TABLE 8. SAMPLER CHARACTERISTIC SUMMARY MATRIX (Cont'd)
o
to
Sampler
Perl 704
Phlppa and Bird
Prolech CG-110
ProTech CG-125
Prolech CG-125FP
ProTech CEG-200
ProTech CEL-300
Prolech DEL-4005
QCEC CVE
QCEC CVE II
QCEC E
SERCO HW-3
SERCO TC-2
Sigaaaotor WA-1
Slgaaaotor UAP-2
Sigaaaotor UM-3-24
SigaaBOtor HA-5
Slgaaaotor WAP-S
Sigaaaotor HH-5-24
Slrco B/SI-VS
Slrco B/IE-VS
Slrco B/DP-VS
Slrco MK-VS
Sonford HG-4
Streaagard DA-24S1
TMI Fluid Stream
IMI Mk 3B (Hants)
Trl-Ald
Ullliaas Oscillaaatic
*
F-pneuB.it ic
F-poeuBatlc
F-pneuvat Ic
F-pneua.it ic
F-submerslble pu»p
user auppl led
S-periataltlc
user supplied
M-dlpper
user supplied
F-pneuaa t ic
S-perlstalt Ic
S-diaphraga type
Flov
Rate
(al/ain)
160
NA
1,000
1,000
1,000
1.000
•>-«,ooo
3.000
3,000
NA
Unk.
42,000
60
60
6Q
80
80
80
12,000
NA
6,000
NA
NA
*
500
60
Lift
<•)
7.4
IB. 3
9.1
9.1
9.1
16.8
9.1
9.1
6.1
6.1
18.3
3.7
^3
NA
6.7
6.7
6.7
5.5
5.5
5.5
6.7
61
NA
6.7
0.5
NA
7.6
7.5
3.6
Line
Sire
(..)
6.4
DA
3.2
3.2
3.2
3.2
12.7
12.7
6..'*
6. A
HA
25.4
6.4
1,19
3.2
3.2
3.2
6.4
6.4
6.4
9.5
9.5
9.5
9.5
19.0
6.4
12.7
9.5
6.4
Saiple
Type
TcVc
TcVc , TvVc
Tc»c
TcVc
f
*
Discrete
IcVc , TvVc
TcVc, TvVc
TcVc
Discrete
TcVc , TvVc
TcVc
TcVc , TvVc
Discrete
TcVc
TcVc , TvVc
Discrete
TcVc , TvVc
TcVc, TvVc
TcVc , TvVc
D, TcVc, TvVc, S
TcVc, TvVc
Discrete
TcVc
TcVc , TvVc
TcVc
Instal-
lation
Portable
. Fixed
Portable
Portable
Portable
P or F
' t or F
Fixed
Portible
Portable
Fixed
Fixed
Portable
Fixed
Portable
Portable
Portable?
Portable
Portable
Portable
P or F
Fixed
P or F
Portable
Portable
Portable
Fixed
Portable
P or F
P or F
Unk.
•^1,000-up
485
695-1,205
925-1,610
1,345-2,445
1,495-2,750
3,995-4,765
570-1.030
1.1,000-up
•vl.OOO-up
Unk.
•x.1,000
•\.2,500
430-730
650-870
975-1,525
750-990
850-1.215
1,225-1,775
1,900-3,000
1,500-3,000
1,600-3,000
•H.SOO-up
325-495
775
1.800
•X.700-UP
650-985
438
DC
AC
-/AC
AC/DC
Alr/AC
AC
AC
AC/DC
AC/DC
AC
AC
Air & AC
AC/DC
AC/DC
AC/DC
AC/DC
AC/DC
AC/DC
AC/DC
AC
AC/DC
yic/oc
AC/BC
Air t AC
AC
H - Mechanical
F - Forced Flow
S - Suction Lift
* - Depends on pressure and lift
NA - Not Applicable
Unk - Unknown at tl»< of writing
-------
maximum vertical distance that is allowed between the sampler in-
take and the remainder of the unit (or at least its pump in the case
of suction lift devices).
"Line Size" indicates the minimum line diameter of the sampling train.
"Sample Type" indicates which type or types of sample the unit (or
series) is capable of gathering. Not all types can necessarily be
taken by all units in a given model class; e.g., an optional controller
may be required to enable taking a TvVc type sample, etc. The "Instal-
lation" column is used to indicate if the manufacturer considers the
unit to be portable or if it is primarily intended for a fixed instal-
lation. "Cost Range" indicates either the approximate cost for a typ-
ical unit or the lowest price for a basic model and a higher price
reflecting the addition of options (solid state controller, battery,
refrigerator, etc.) that might enhance the utility of the device. Fin-
ally, the "Power" column is used to indicate whether line current (AC),
battery (DC), or other forms of power (e.g., air pressure) are required
for the unit to operate.
In general, the commercially available automatic samplers have been de-
signed for a particular type of application. In the present work, how-
ever, they are being considered for application in a storm or combined
sewer setting. Because of the vagaries of such an application, it is
altogether possible that a particular unit may be quite well suited for
one particular application and totally unsuitable for use in another.
The following 12 points should be considered in the selection of a
piece of automatic sampling equipment:
1. Obstruction or clogging of sampling parts, tubes and
pumps.
2. Obstruction of flow.
403
-------
3. Operation under the full range of flow conditions pecu-
liar to storm and combined sewers.
4. Operation unimpeded by the movement of solids such as
sand, gravel and debris within the fluid flow; including
durability.
5. Operation automatic Cduring storm conditions), unat-
tended, self-cleaning.
6. Flexibility of operation allowed by control system.
7. Collection of samples of floatable materials, oils and
grease, as well as coarser bottom solids.
8. Storage, maintenance and protection of collected samples
from damage and deterioration as well as the sample
train and containers from precontamination.
9. Amenability to installation and operation in confined
and moisture laden places such as sewer manholes.
10. Ability to withstand total immersion or flooding during
adverse flow conditions.
11. Ability to withstand and operate under freezing ambient
conditions.
12. Ability to sample over a wide range of operating head
conditions.
404
-------
REVIEW OF RECENT FIELD EXPERIENCE
In order to assess the efficacy of both commercially available samplers
and custom engineered units in actual field usage, a survey of recent
USEPA projects, many of which were in the storm and combined sewer pol-
lution control area, was conducted. None of these projects was under-
taken solely to compare or evaluate samplers, but all required
determination of water quality. In^the following, difficulties en-
countered with various elements of the liquid samplers are described.
The small diameter, low intake velocity probes found in several com-
mercial units were felt to be unable to gather as representative a
sample of the flow as could be obtained manually. There were many in-
stances of inlet tube openings being blocked by rags, paper, disposable
diapers, and other such debris. Although less a fault of the equipment
than an installation practice, there were several instances of intake
tubes being flushed over emergency overflow weirs, up on to manhole
steps, etc., during periods of high flow and left high and dry and un-
able to gather any samples when the flow subsided.
There were numerous instances of pre-evacuated bottle samplers losing
their vacuum in 24 to 48 hours, resulting in little or no data. Fur-
thermore, personnel find these units with their 24 individual intake
tubes virtually impossible to clean in the field. The low suction
lifts on many commercial units render some sites inaccessible. In one
project, three sites required manual sampling because none of the sam-
plers on hand could meet the 5 to 6m lifts required at these sites.
There were several instances of sample quantity varying with sewage
level as well as with the lift required at the particular site. On at
least two occasions, submersible pumps were damaged or completely swept
away by heavy debris in the flow.
Within the sampling train itself, line freezing during winter operation
was a problem in several projects with instances of up to 60% data loss
405
-------
reported. In one project, the intake line was too large, which allowed
solids to settle out in it until it ultimately became clogged. There
were numerous instances of smaller suction tubes becoming plugged with
stringy and large sized material. A very frequent complaint, applied
especially to discrete samplers, was that they gathered inadequate sam-
ple volumes for the laboratory analyses required.
Although not directly the fault of the sampling equipment itself, on
one project data were lost for 14 storms due to improper sterilization
of non-disposable sample bottles.
The control subsystems of commercial units probably came in for more
criticism than any other. Comments on automatic starters ranged from
poor to unreliable to absolutely inadequate. There were instances
where dampness deteriorated electrical contacts and solenoids causing
failure of apparently well-insulated parts. The complexity of some
electrical systems made them difficult to maintain and repair by field
personnel. Inadequate fuses and failures of microswitches, relays, and
reed switches were commonly encountered. The minimum time between ex-
traction of samples for some commercial units was too long to adequately
characterize some rapidly changing flows.
Collected USEPA experience in one region, involving over 90,000 hours
use of some 50 commercial automatic liquid samplers of 15 makes and
models, has indicated that the mean sampler failure rate is approxi-
mately 16% with a range of 4% to 40% among types. They have found that
the ability of an experienced team to gather a complete 24-hour compos-
ite sample is approximately 80%. When one factors in the possibility
of mistakes in installation, variations in personnel expertise, exces-
sive changes in lift, surcharging, and winter operation, it is small
wonder that projects on which more than 50 or 60 percent of the desired
data were successfully gathered using automatic samplers were, until re-
cently, in the minority.
406
-------
In fairness to present day equipment, it must be pointed out that some
of the above cited complaints stem from equipment designs of up to six
years ago, and many commercial manufacturers, properly benefitting from
field experience, have modified or otherwise improved their products'
performance. The would-be purchaser of commercial automatic samplers
today, however, should keep in mind the design deficiencies that led
to the foregoing complaints when selecting a particular unit for his
application.
407
-------
LABORATORY ANALYSIS
As an integral part of any wastewater characterization effort, the role
of the analytical laboratory must not be underestimated. To be of value,
the data it provides must correctly describe the characteristics or con-
centrations of the constituents of interest in the sample submitted to it.
In many cases, an approximate answer or incorrect result is worse than no
answer at all, since it could lead to faulty interpretations and incor-
rect decisions. The Methods Development and Quality Assurance Research
Laboratory (MDQARL) (1972) has prepared a handbook for analytical qual-
ity control that provides sufficient information to allow inauguration
or reinforcement of a program that will emphasize early recognition,
prevention, and correction of factors leading to breakdowns in the va-
lidity of data products from water and wastewater laboratories.
It is especially important that the laboratory analyst be involved in
the initial design of any wastewater characterization program. The re-
quirements for analysis of constituents of interest vary widely and
must be accounted for in the program design if it is to be sucessful.
Nothing is more frustrating to the analyst, for example, than to be
given a 100 ml sample and asked to perform a large number of analyses,
any one of which would require about all of the available sample quan-
tity to run correctly.
For any program, 500 ml should be considered the bare minimum sample
quantity, and at least a liter will be required in many instances. It
is always safer to err on the side of collecting, too much sample rather
than too little. Furthermore, such larger samples are less sensitive
to slight pre- and cross-contamination effects that might result from
some sampling equipment designs, improper sample collection and han-
dling techniques, etc.
408
-------
SAMPLE PRESERVATION AND HANDLING
Having collected a representative sample of the fluid mixture in ques-
tion, there remains the problem of sample preservation and analysis.
It is a practical impossibility either to perform instant analyses of
the sample on the spot or to completely and unequivocally preserve it
for subsequent examination. Preservative techniques can only retard
the chemical and biological changes that inevitably continue following
extraction of the sample from its parent source. In the former case,
changes occur that are a function of the physical conditions - metal
cations may precipitate as hydroxides or form complexes with other
constituents; cations or anions may change valence states under certain
reducing or oxidizing conditions; constituents may dissolve or volatize
with time, and so on. In the latter case, biological changes taking
place may change the valence state of an element or radical; soluble
constituents may be converted to organically bound materials in cell
structures; cell lysis may result in release of cellular material into
solution, etc.
Preservation methods are relatively limited and are generally intended
to retard biological action, retard hydrolysis of chemical compounds
and complexes, and reduce volatility of constituents. They are gener-
ally limited to pH control, chemical addition, refrigeration, and
freezing. MDQARL (1974) has compiled a list of recommendations for
preservation of samples according to the measurement analysis to be
performed. Since it is frequently of interest to examine a sample for
a number of parameters, this list has been reproduced here as Table 9.
Standard methods for the examination of water and wastewater have been
set forth by the American Public Health Association (1971) and MDQARL
(1974). As regards the solids content in particular, termed residue in
these publications, methods are given for determining total residue
(upon evaporation), total volatile and fixed residue, total suspended
matter (nonfilterable residue), volatile and fixed suspended matter,
409
-------
TABLE 9. RECOMMENDATION PRESERVATION OF SAMPLES ACCORDING TO MEASUREMENT
(1)
— •
Acidity
alkalinity
arsenic
HO
Broejlde
COO
Chloride
CUorlne leg.
Color
Cyanide*
Dissolved Oxygon
Proo.
Hnaler
Fluoride
Berdneee
Iodide
KUS
Metals
DlMOlTed
•Mpended
• IMal
Dleeolved
Utroaen
• '-
rjaldaU
•Itrate
•Itrate
•nni
-'l. Tnbn from •
2. Tlaetlc or C]
3. If lenvla* cl
Vol
Baa, Container
(«*>
100 P.G«>
100 F.G
100 P,G
1000 F.G
100 F.C
50 F.G
50 F.G
50 F.G
50 F.G
500 F.G
300 C only
300 G only
300 F.G
100 F.G
100 F.C
250 F.G
200 F.E
100
100 F.C
400 F.G
500 7.G
100. F.C
,
58 F.C
NUL <1»74>.
lane.
moot bo returned to the labori
Preeer...!"
Cool. 4-C
Cool. 4*C
BM>3 to pi <2
Cool, 4-C
Cool, 4-C
BjSOt to pB <2
Hone laq
Cool, 4-C
Cool, 4'C .
Cool, 4"C
HeOB to pH 12
Dnt on .Ite
Fix on alta
Cool, 4-C
Cool, 4-C
Cool. 4'C
Cool. 4-C
Filter on sits
BK>} to pi <2
Filter on site
Ba»3 to pi <2
Filter
B»3 to pa <2
Cool, 4*C
BgSO^ to pB <2
Cool, 4-C
BjS04 to pi <2
Cool, 4-C
BjSOj to pi <2
Cool, 4'C
•torv la laa. tbna o be
Holding Tln.(i>
24 Bre
24 Bre
6 Nos
6 Br.(3)
24 Hre
7 Deye
7 Days
24 Bra
24 Bra
24 Brs
Ho BoUlnt
•o Boldln.
7 Day.
7 Days
24 Ire
24 Br.
i Hoe
6 Ho.
i Mo.
38 Deye (Claee)
13 Deye (Berii
Plastic)
24 Brs«>
24 Brs(*'
24 Brs'*'
24 „,<«>
mra and holdins tlm*
Heeeuree«nt
IRA
Oil end Grea.e
Orfanlc Carbon
PB
Pbenollce
Phoephorus
Orthopboaohet* ,
Dleeolvxl
Bydrolyseble
Total
Total,
Dleeolved
Filterable
Bonflltareble
Total
Volatile
Settleable Better
S-1"1-
SUlca
Specific
Conductance
Sulfete
Sulflde
Sulflte
Tenoereture
Thrvehold Odor
Turbidity
Vol
R*g
(ml)
5O
1000
23
25
500
50
50
SO
50
100
100
100
100
1000
SO
50
100
50
50
50
1000
200
100
Container
P,C
G only
P.G
P.C
G only
P.C
F.G
F.G
P.C
F.C
F.G
F.G
F.C
F.G
F.G
Ponlj
F.C
P.C
F.G
F.G
F.G
C only
F.C
Preeervetive
Cool, 4-C
Cool, 4'C
BjSO^ to pB <2
Cool, 4'C
BjSOj to pB <2
Cool, 4'C
Det on site
Cool, 4'C
B.PO to ph <4
Filter on site
Cool, 4*C
Cool, 4'C
H.SO, to pfl <2
Cool, 4-C
Filter on site
Cool, 4-C
Cool, 4"C
Cool, 4'C
Cool, 4-C
Cool, 4'C
Hone Beq
BUOj to pB 2
Cool, 4-C
Cool, 4'C
Cool, 4*C
2 at zinc
acetate
Cool, 4*c
Det on site
Cool, 4-C
Cool, 4*C
Holding Tj
24 Brs
24 Bre
24 Bre
6 firs'3'
24 Bre
24 Br.<«>
2* Br.">
24 Br.<"
24 Br.<*>
7 Dare
7 Day.
7 Deye
7 Deye
24 Bre
6 Hoe
7 Dsys
24 Br.5'
7 Deye
24 Brs
24 flrs
Ho Boldlng
24 Brs
7 Days
4. Hcrcurlc chloride my b* uMd •• an alt«rnate prM*rvatlvt at a conconcratlon of
40 **/l, MpecUHjr if • longer boldlns tlM t« r«
-------
dissolved matter (filterable residue), and settleable matter. These
tests do not determine specific chemical substances, but classes of
matter that have similar physical properties and similar responses to
ignition. The tests are basically empirical in nature and, as a result,
the constituents of each form of residue are defined to a large extent
by the procedures employed. For this reason, close adherence to the
established procedures is necessary to insure reproducibility and com-
•
parability of results. It is interesting to note that MDQAKL (1974)
states that precision and accuracy-data are not available for methods
used to determine total dissolved solids (total filterable residue),
total suspended solids (total nonfilterable residue), total solids
(residue), or settleable matter.
Sampling for certain basic wastewater parameters is essential. In gen-
eral these Include:
a. Biochemical oxygen demand (BOD) - Used to determine the
relative oxygen requirement of the wastewater. Data
from BOD tests are used for the development of engi-
neering criteria for the design of wastewater treat-
ment plants.
b. Chemical oxygen demand (COD) - Provides additional in-
formation concerning the oxygen requirement of waste-
water. It provides an independent measurement of
organic matter in the sample, rather than being a
substitute for the BOD test. For combined sewer over-
flows and stormwater, COD may be more representative
of oxygen demand in a receiving stream because of the
presence of metals and other toxicants which are rela-
tively nonbiodegradable.
c. Total oxygen demand (TOD) - A recently developed test
to measure the organic content of wastewater in which
411
-------
the organics are converted to stable end products in a
platinum-catalyzed combustion chamber. The test can
be performed quickly, and results have been correlated
with the COD in certain locations.
d. Total organic carbon (TOG) - Still another means of
measuring the organic matter present in water which
has found increasing use in recent times. The test
is especially applicable to small concentrations of
organic matter.
e. Chloride - One of the major anions in water and
sewage. The concentration in sewage may be in-
creased by some industrial wastes, by runoff from
streets and highways where salt is used to control
ice formation, salt water intrusion in tidal areas,
etc. A high chloride content is injurious to vehi-
cles and highway structures, and may contaminate
water supplies near the highway.
f. Nitrogen Series - A product of microbiologic activity,
is an indicator of sewage pollution, or pollution re-
sulting from fertilizers, automobile exhausts, or
other sources. Its presence may require additional
amounts of chlorine, or introduction of a nitrogen
fixation process, in order to produce a free chlorine
residual in control of bacteria.
g. pH - The logarithm of the reciprocal of hydrogen ion
activity. State regulations often prescribe pH limits
for effluents from industrial waste treatment plants.
Provides a control in chemical and biological treat-
ment processes for wastewater.
412
-------
h. Solids (Total, Suspended, Volatile, and Settleable) -
Usually represent a large fraction of the pollutional
load in combined sewage. Inorganic sediments, in a
physical sense, are major pollutants, but also serve as
the transporting or catalytic agents that may either
expand or reduce the severity of other forms of
pollution.
i. Oil and Grease - Commonly found in sanitary sewage, but
also appear in industrial wastes as a result of various
industrial processes. Present a serious problem of
removal in wastewater treatment facilities.
j. Bacterial Indicators (Total Coliform, Fecal Coliform,
Fecal Streptococcus) - Indicate the level of bacterial
contamination.
Where more exotic wastes are combined with stormwater and sanitary
sewage, additional treatment facilities may be required for the re-
moval of industrial byproducts and nutrients such as cyanide, fluoride,
metals, pesticides, nitrogen, phosphorus, sulfate and sulfide. For
planning and design of such treatment facilities, additional analyses
are required in accordance with the pollutant material expected in the
wastewater. This may, in turn, require significant expansion of the
analysis program.
SAMPLING EQUIPMENT CLEANING
The proper cleaning of all equipment used in the sampling of waste-
water is essential to assuring valid results from laboratory analyses.
Cleaning protocols should be developed for all sampling equipment early
in the design of the stormwater characterization program. Here also,
the laboratory analyst should be consulted, both to assure that the
413
-------
procedures and techniques are adequate as well as to avoid including
practices that are not warranted in view of the analyses to be performed.
As an example, Lair (1974) has set down the standard operating procedures
for the cleaning of sample bottles and field equipment used by USEPA
Region IV Surveillance and Analysis field personnel engaged in NPDES
compliance monitoring. They are reproduced below for a typical auto-
matic sampler and related sampling equipment.
2-1/2 Gallon Pyrex Glass Composite Bottles
1. Rinse twice with spectro grade acetone.
2, Rinse thoroughly with hot tap water using a bottle
brush to remove particulate matter and surface film.
3. Rinse thoroughly three times with tap water.
4. Acid wash with at least 20 percent hydrochloric acid.
5. Rinse thoroughly three times with tap water.
6. Rinse thoroughly three times with distilled water.
7. Rinse thoroughly with petroleum ether and dry by pulling
room air through bottle.
8. Dry in drying oven overnight.
9. Cap with Aluminum foil.
414
-------
ISCO Glass Sample Bottles
1. One spectre grade acetone rinse.
2. Dishwasher cycle (wash and tap water rinse, no
detergent).
3. Acid rinse with at least 20 percent hydrochloric acid.
4. Dishwasher cycle, tap and distilled water rinse cycles,
no detergent.
5. Replace in covered ISCO bases.
Sample Tubing (1/4, 3/8 or 1/8 pexcon or tygon)
1. Do not reuse sample tubing. No cleaning required.
New sample tubing is to be used for each new sampling
setup.
2. Use teflon tubing where samples for organics are to
be collected.
ISCO Pump Tubing
1. Rinse by pumping hot tap water through tubing for at
least 2 minutes.
2. Acid wash tubing by pumping at least a 20 percent
solution of hydrochloric acid through tubing for at
least 2 minutes.
3. Rinse by pumping hot tap water through tubing for at
least 2 minutes.
415
-------
4. Rinse by pumping distilled water through tubing for at
least 2 minutes.
Teflon Sample Tubing
1. Teflon sample tubing should be cleaned in the same manner
as the 2-1/2 gallon pyrex sample containers.
ISCO Rotary Funnel and Distributor
1. Clean with hexane to remove any grease deposits.
2. Rinse thoroughly with hot water and a bottle brush to
remove particulate matter and surface films.
3. Use a squeeze bottle of 20 percent hydrochloric acid
and rinse thoroughly, rinse funnel as well as funnel
and distributor depressions.
4. Rinse thoroughly with tap water.
5. Rinse thoroughly with distilled water.
6. Replace in sampler.
ISCO Sample Headers
1. Rinse entire header with hexane or petroleum ether.
2. Disassemble header and rinse thoroughly with hot tap
water, using a bottle brush to remove particulate matter
and surface films.
416
-------
3. Rinse the plastic portion of the header with at least a
20 percent solution of hydrochloric acid. Do not use
acid on the metal parts.
4. Rinse thoroughly with tap water.
5. Reassemble header.
6. Rinse all header parts thoroughly with distilled water.
One Gallon Plastic Sample Containers
1. Use only new bottles when sampling wastewater sources.
One Quart Wide Mouth Bottles for Organics. Pesticides and Oil and
Grease Samples
1. Use only new bottles with teflon liners.
2. Rinse twice with petroleum ether and allow to dry.
One Pint Narrow Mouth Bottles for Phenol Samples
1. Use new bottles only.
One Pint Narrow Mouth Mercury Sample Bottles
1. Use only new bottles.
2. Rinse with at least 20 percent nitric acid.
3. Rinse at least three times with distilled water.
417
-------
One Liter Plastic Storemore Cyanide Sample Bottles
1. Use only new bottles.
418
-------
SUMMARY
1. There are numerous stormwater management models available and
their data needs vary both as to time and spatial coverage.
Therefore a knowledge of model structure and requirements is
desirable to facilitate a data gathering program.
2. Regardless of the type of model involved, two types of data are
required; calibration, for fitting a model to a particular basin,
and verification, for model testing. There is no real difference
in the data gathering requirements, but the distinction of their
uses must be kept in mind. In particular, no data used for cali-
bration should ever be used for verification.
3. The three major factors in the field data considered here are
flow measurement, sampling, and laboratory analysis of the samples
taken. Although they represent three different efforts, requiring
different types of skills and training, they are also inter-
dependent tasks and must be considered together in the establish-
ment of a successful stormwater characterization program.
4. The required accuracy of the time element in stormwater charac-
terization data is application dependent. Time synchronization
of data is required, however, to allow parameter inter-
relationship to be determined.
5. The required quality of the field gathered data, expressed as
accuracy and precision, will depend upon the ultimate use.
Records should include all information that would bear on data
quality, as well as estimates of accuracy and precision, to
maximize data utility.
419
-------
6. A flowmeter is one tool of several that must be employed for
the characterization of a wastewater stream. Its selection must
be based upon consideration of the overall flow measurement pro-
gram to be undertaken, the nature of the flows to be measured,
the physical characteristics of the flow measurement sites, and
the degree of accuracy required, among other factors.
7. In view of the large number of highly variable parameters asso-
ciated with the storm and combined sewer application, no single
flowmeter can exist that is universally applicable with equal
efficacy. Some requirements are conflicting, e.g., an open drain-
age ditch versus a closed conduit deep underground, and a careful
series of trade-off studies is required in order to arrive at a
"best" selection for a particular program and site.
8. The proper selection of flow measurement sites can be as important
as the selection of methods and equipment. A clear understanding
of the data requirements and ultimate use is necessary, as is a
familiarity with the sewer system to be examined. There are
measurement sites where no presently available equipment can
operate unattended for long at a high degree of accuracy (better
than ±5% of full scale).
9. Where large flows are to be measured with fairly high accuracy,
considerable expense in terms of initial equipment cost, site
preparation and installation, and operator training and maintenance
is involved; fifty to one hundred thousand dollars should not be
considered atypical.
10. The most consistently reliable flow measurement data have been
taken at sites where the equipment has been calibrated in place
over the entire range of flows anticipated.
420
-------
11. In an extensive review of field experience in wastewater flow
measurement, it was found that in most instances errors of greater
than 10% seem to be the rule. It is not at all uncommon to find
readings that differ from spot field checks by from 50 to 200 per-
cent. Some wastewater discharge data are of such poor quality as
to be virtually useless.
12. An automatic liquid sampler is one tool of several that is often
employed for the characterization of a flow stream. Its selection
must be based upon consideration of the overall sampling program
to be undertaken, the characteristics of the flows to be sampled,
the physical characteristics of the sampling sites, and the sam-
ple analyses that are available and desired.
13. In view of the large number of highly variable parameters associ-
ated with the storm and combined sewer application, no single
automatic sampler can exist that is universally applicable with
equal efficacy. Some requirements are conflicting, and a careful
series of trade-off studies is required in order to arrive at a
"best" selection for a particular program. Such a selection may
not be well suited for a different program, and a systems approach
is required for either the selection or design of automatic sampl-
ing equipment for storm and combined sewer application.
14. The proper selection of sampling sites can be as Important as the
selection of sampling methods and equipment. A clear understanding
of the data requirements and ultimate use is necessary as is a
familiarity with the sewer system to be examined,
15. Field experience with automatic sampling equipment, with emphasis
on recent USEPA projects, revealed leaks in vacuum operated units;
faulty automatic starters; inlet blockage and line plugging; lim-
ited suction lift; low transport velocities; complicated electri-
cal systems; and failures of timers, micro-switches, relays and
421
-------
contacts, and reed switches among the difficulties frequently
encountered.
16. There is a plethora of sampling devices available in the market-
place today. These automatic samplers are of various designs and
capabilities and incorporate both good and poor features. There
are numerous claims (and counter-claims) made by the various manu-
facturers and their representatives, including limited data in
certain instances, as to the efficacy of one particular piece of
equipment (i.e., design approach) or another. The present state
of affairs can be summarized as follows:
(a) Comparisons of water quality data gathered using dif-
ferent commercially available samples demonstrate
without question that there can be marked differences
in results obtained with different types of equipment;
(b) Different wastewater flow characteristics call for dif-
ferent equipment requirements in order to assure repre-
sentative sampling; and
(c) The results of manual sampling are extremely methodo-
logically dependent, and data strongly indicate that
they may or may not be representative of the wastewater
flow in question,
17. The laboratory analyst should be involved early in the design of
the stormwater characterization program. It is especially im-
portant to determine the constituents that are to be examined
early .in the program design to assure that sufficient quantity of
sample for the analysis to be performed is delivered to the
laboratory.
422
-------
18. Complete and unequivocal sample preservation is a virtual impossi-
bility. However, sample preservation guidelines are given for a
number of parameters and should be closely adhered to if the re-
sults of laboratory analysis are to be meaningful.
19. It is important that cleaning protocols be developed for all
sampling equipment used in a stormwater characterization program.
Here also the laboratory analyst should be consulted, both to as-
sure that the procedures and techniques are adequate and to avoid
including unnecessary practices in view of the analysis to be
performed.
423
-------
REFERENCES
American Public Health Association (1971). Standard Methods for the
Examination of Water and Wastewater, 13th Ed., published jointly with
the American Water Works Association and the Water Pollution Control
Federation, Washington, D.C.
American Society of Mechanical Engineers (1971). Fluid Meters - Their
Theory and Application, Report of the ASME Committee on Fluid Meters,
Sixth Edition, New York, NY.
American Society of Testing Materials (1973). 1973 Annual Book of ASTM
Standards, Part 23 - Water and Atmospheric Analysis. 1916 Race Street,
Philadelphia, PA.
Amorocho, J. (1969). "Generalized Analysis of Small Watershed
Responses", California Water Resources Center Technical Completion
Report No. WRC-133.
Amorocho, J. and A. Brandsetter (1971). "Determination of Nonlinear
Functional Response Functions in Rainfall-Runoff Processes", Water
Resources Research, Vol. 7, No. 5.
Anderson, E. A. and N. H. Crawford (1964). "The Synthesis of Con-
tinuous Snowmelt Runoff Hydrographs on a Digital Computer", Technical
Report 36, Department of Civil Engineering, Stanford University.
Anderson, E. A. (1968). "Development and Testing of Snowpack Energy
Balance Equations", Water Resources Research, Vol. 4, No. 1.
Beard, L. R. (1966). "Basin Rainfall and Snowmelt", Program No. 23-
266, Hydrologic Engineering Center, Corps of Engineers, Sacramento, CA.
424
-------
Bergthorssen, P. and B. Doos (1955). "Numerical Weather Map Analysis,"
Tellus, Vol. 7, No. 3..
Bertie, F. A. (1966). "Effect of Snow Compaction or Runoff from Rain
or Snow", U.S. Bureau of Reclamation, Engineering Monograph No. 35.
Bidwell, V. J. (1971). "Regression Analysis of Nonlinear Catchment
Systems", Water Resources Research, Vol. 7, No. 5.
Buchanan, T. J. and W. P. Somers (1969). "Discharge Measurements at
Gaging Stations," SurfacA Water^echni^ues^ Book 3, Chapter 8, U.S.
Geological Survey, U.S.D.I., Reston, VA.
Carter, R. W. and I. E. Anderson (1963). "Accuracy of Current Meter
Measurement," J. of the Hydraulics Div.t ASCE, Vol. 89, No. HY 4.
Chen, C-L., C. G. Clyde, M-S. Chu and C-Y. Wei (1972). "Calibration of
Parshall Flumes With Non-Standard Entrance Conditions," Utah Water
Research Laboratory Report PRWG 102-1, College of Engineering, Utah
State University, Logan, UT.
Chow, V. T. (1962). "Hydrologic Determination of Waterway Areas for
the Design of Drainage Structures in Small Drainage Basins", University
of Illinois Engineering Experimental Station, Bulletin 462.
Cobb, E. D. and J. F. Bailey (1965). "Measurement of Discharge by
Dye-Dilutions Methods," Surface Water Techniques, Book 1, Chapter 14,
U.S. Geological Survey, U.S.D.I., Reston, VA.
Collinge, V. K. and D. G. Jamieson (1968). "The Spatial Distribution
of Storm Rainfall", Journal of Hydrology, Vol. 6, No. 1.
Cook, H. L. (1946). "The Infiltration Approach to the Calculation of
Surface Runoff", Transactions. American Geophysical Union, Vol. 27,
(October).
425
-------
Crawford, N. H. and R. K. Linsley (1966). "Digital Simulation in
Hydrology: the Stanford Watershed Model IV", Technical Report 39,
Department of Civil Engineering, Stanford University.
Dall, H. E. (1962). "Flow Tubes and Non-Standard Devices for Flow
Measurement With Some Coefficient Considerations," in Proceedings of
a Symposium on Flow Measurement in Closed Conduits held at the
National Engineering Laboratory, Gt. Brit., Sept 27-30, 1960, Vol. 2,
Section D, Her Majesty's Stationary Office, Edinburgh.
Dalrymple, T. (1960). "Flood Frequency Analysis", USGS Water Supply
Paper 1543-A.
Dooge, J. I. (1959). "A General Theory of the Unit Hydrograph",
Journal of Geophysical Research, Vol. 64, No. 2.
Duckstein, L., M. M. Fogel, and C. C. Kisiel (1972). "A Stochastic
Model of Runoff-Producing Rainfall for Summer Type Storms", Water
Resources Research, Vol. 8, No. 2.
Federal Inter-Agency Sedimentation Project (FIASP) (1941a), "Methods
of Analyzing Sediment Samples," Report No. 4, St. Paul U.S. Engineer
District Sub-Office Hydraulic Laboratory, University of Iowa,
Iowa City, 10.
(1941b), "Laboratory Investigation of Suspended Sediment Sam-
plers," Report No. 5, St. Paul U.S. Engineer District Sub-Office
Hydraulic Laboratory, University of Iowa, Iowa City, 10.
(1948), "Measurement of the Sediment Discharge of Streams,"'
Report No. 8, St. Paul U.S. Engineer District Sub-Office Hydraulic
Laboratory, University of Iowa, Iowa City, 10.
426
-------
(1957), "Some Fundamentals of Particle Size Analysis," Report
No. 12, St. Anthony Falls Hydraulic Laboratory, Minneapolis, MN.
(1962), "Investigation of a Pumping Sampler With Alternate
Suspended Sediment Handling Systems," Report No. Q, St. Anthony Falls
Hydraulic Laboratory, Minneapolis, MN.
(1963), "Determination of Fluvial Sediment Discharge," Report
No. 14, St. Anthony Falls Hydraulic Laboratory, Minneapolis, MN.
(1964), "Electronic Sensing of Sediment," Report No. R,
St. Anthony Falls Hydraulic Laboratory, Minneapolis, MN.
(1966), "Laboratory Investigations of Pumping Sampler Intakes,"
Report No. T, St. Anthony Falls Hydraulic Laboratory, Minneapolis, MN.
Freeze, R. A. (1972). "Role of Subsurface Flow in Generating Surface
Runoff", Water Resources Research, Vol. 8, No. 3.
Grace, R. A. and P. S. Eagleson (1966). "The Modeling of Overland
Flow", Water Resources Research, Vol. 2, No. 3.
Grace, R. A. and P. S. Eagleson (1967). "A Model.for Generating
Synthetic Sequences of Short-Time-Interval Rainfall Depths", Proc. of
International Hydrology Symposium (Colorado State University), Vol. 1,
Paper 35.
Harris, D. J. and W, J. Keffer (1974), "Wastewater Sampling Methodo-
logies and Flow Measurement Techniques," USEPA Region VII Surveillance
and Analysis Division Report No. EPA 907/9-74-005, Kansas City, KS.
Hershfield, D. M. (1961). "Rainfall Frequency Atlas of the United
States," National Weather Services Technical Paper No. 40, U.S.
Department of Commerce, Washington, D.C.
427
-------
Huff, F. A. and W. L. Shipp (1969). "Spatial Correlations of Storm,
Monthly, and Seasonal Precipitation," Journal of Applied Meteorology.
Vol. 8, No. 4.
Huff, F. A. (1970a). "Spatial Distribution of Rainfall Rates", Water
Resources Research, Vol. 6, No. 1.
Huff, F. A. (1970b). "Time Distribution Characteristics of Rainfall
Rates," Water Resources Research, Vol. 6, No. 2.
Ippen, A. T. (1966), "Sedimentation in Estuaries," in Estuary and
Coastline Hydrodynamics, McGraw-Hill, Inc., New York, NY.
Izzard, C. F. (1946). "Hydraulics of Runoff from Developed Surfaces",
Proceedings of the Highway Research Board, Vol. 26.
Kilpatrick, F. A. (1964). "Use of Flumes in Measuring Discharges at
Gaging Stations," Surface Water Techniques, Book 1, Chapter 16, U.S.
Geological Survey, U.S.D.I., Reston, VA.
Kilpatrick, F. A. (1967). "Dye-Dilution Measurements Made Under Total
Ice Cover on the Laramie, River at Laramie, Wyoming," Water Resources
Division Bulletin (July-December).
Kohler, M. A. and R. K. Linsley (1951). "Predicting the Runoff from
Storm Rainfall", U.S. Weather Bureau Research Paper 34.
Kuichling, E. (1889). "The Relation Between the Rainfall and the
Discharge of Sewers in Populous Districts", Transactions, ASCE, Vol. 20.
Lair, M. D. (1974). Private communication.
Lane, E. W., and Kallnske, A. A. (1939), "The Relation of Suspended to
Bed Materials in Rivers," Transactions. American Geophysical Union,
Vol. 20, Part 4.
428
-------
Lane, E. W., and Kallnske, A. A. (1941), "Engineering Calculations of
Suspended Sediment," Transact ions, American Geophysical Union, Vol. 22,
Part 3.
Leupold and Stevens, Inc. (1974). Water Resources Data Book, first ed.,
P.O. Box 688, Beaverton, OR.
Linsley, R. K. (1971). "A Critical Review of Currently Available
Hydrologic Models for Analysis of Urban Stormwater Runoff," Office of
Water Resources Research Contract No. 14-31-0001-3146 (Project XllOCAL)
Report (August).
Liptak, B. G. and R. K. Kaminski (1974). "Ultrasonic Instruments for
Level and Flow," Instrumentation Technology, Vol. 21, No. 9.
Ludwig, J. H. and R. C. Ludwig (1951). "Design of Palmer-Bowlus
Flumes," Sewage and Industrial Waste, Vol. 23, No. 9.
Methods Development and Quality Assurance Research Laboratory (MDQARL)
(1972). Analytical Quality Control in Water and Wastewater
Laboratories, a Technology Transfer Publication, U.S. Environmental
Protection Agency, Washington, D.C.
(1974). Methods for Chemical Analysis of Water and Waste, a
Technology Transfer Publication, U.S. Environmental Protection Agency,
Washington, D.C.
McMahon, J. P. (1964). "Flow of Fluids," Section 5-1 in Handbook of
Applied Instrumentation, D. M. Considine, ed-in-chief, McGraw-Hill
Book Co., New York, NY.
McPherson, M. B. (1969). "Some Notes on the Rational Method of Storm
Drain Design", ASCE, Urban Water Resources Program, Technical Memo
No. 6.
429
-------
Mkhitaryan, A. M. (1959), Gidravlika 1 Osnovy Gazodinamiki, (Hydraulics
and Fundamentals of Gas Dynamics), Gosudarstvennoe Izdatel'stvo
Tekhnicheskoi Literatury UkrSSR, Kiev.
Mulvaney, T. J. (1851). "On the Use of Self-registering Rain and Flood
Gauges in Making Observations of the Relation of Rainfall and Flood
Discharges in a Given Catchment", Transactions-Institute of Civil
Engineering (Ireland), Vol. 4, No. 2.
O'Donnell, T. (1960). "Instantaneous Unit Hydrograph Derivation by
Harmonic Analysis", Int. Assoc. Sci. Hydrology.
Palmer, H. K. and F. D. Bowlus (1936). "Adaptation of Venturi Flumes
to Flow Measurement in Conduits Transactions, ASCE, Vol. 101.
Postma, H. (1967), "Sediment Transport and Sedimentation in the
Estuarine Environment," in Estuaries, American Association for the
Advancement of Science Publication No. 83, Washington, D.C.
Potter, W. D. (1961). "Peak Rates of Runoff from Small Watersheds",
U.S. Bureau of Public Roads Hydraulic Design Series 2 (April).
Replogle, J. A., L. E. Meyers and K. J. Burst (1966). "Flow
Measurements With Fluorescent Tracers," J. of the Hydraulics Div., ASCE,
Vol. 92, No. HY 5.
Replogle, J. A. (1970). "Flow Meters for Water Resource Management,"
Water Resources Bulletins, Vol. 6, No. 3.
Robinson, A. R. (1957). "Parshall Measuring Flumes of Small Sizes,"
Colorado Agricultural and Mechanical College (Now Colorado State
University), Agricultural Experimental Station Bulletin No. 61,
Fort Collins, CO.
430
-------
Schulz, E. F., Wilde, R. H. and Albertson, M. L. (1954), "Influence of
Shape on the Fall Velocity of Sedimentation Particles," Missouri River
Division Sedimentation Series Report No. 5, U.S. Army Corps of
Engineers, Omaha, NB.
Shelley, P. E. (1974), "Sampling of Wastewater," a Technology Transfer
Publication of the Industrial Wastewater Monitoring Seminar Program,
U.S. Environmental Protection Agency, Washington, D.C.
(1975a), "Review of Automatic Liquid Samplers," in Urban
Runoff - Quantity and Quality, W. Whipple, Jr., ed., American Society
of Civil Engineers, New York, NY.
(1975b), "Implementation of New Technologies in the Design of
an Automatic Liquid Sampling System," in Proceedings of International
Exposition and Seminar on Water Resources Instrumentation, International
Water Resources Association, Milwaukee, WI (in press).
(1975c), "Design and Testing of a Prototype Automatic Sewer
Sampling System, USEPA Environmental Protection Technology Series
Report No. EPA-R2-75-XXX (in press).
(1975d). "Sediment Measurement in Estuarine and Coastal Areas,"
EG&G Washington Analytical Services Center Report No. TR-7115-001,
Rockville, MD.
Shelley, P. E. and 6. A. Kirkpatrick (1973). "An Assessment of
Automatic Sewer Flow Samplers," Environmental Protection Technology
Series Report No. EPA-R2-73-261, U.S. Environmental Protection Agency,
Washington, DC.
431
-------
(1975). "Sewer Flow Measurement - A State-of-the-art
Assessment," Environmental Protection Technology Series Report
No. EPA-R2-75-XXX (in press), U.S. Environmental Protection Agency,
Washington, DC.
Shercliff, J. A. (1962). The Theory of Electromagnetic Flow
Measurement, Cambridge University Press, New York, NY.
Sherman, L. K. (1932). "Streamflow from Rainfall by the Unit-Graph
Method", Engineering News-Record, (April 7).
Shuster, J. C. (1965). "Canal Discharge Measurements With
Radioisotopes," J. of the Hydraulics Div., ASCE, Vol. 91, No. HY 2.
Skogerboe, G. V., M. L. Hyatt and K. 0. Eggleston (1967a). "Design and
Calibration of Submerged Open Channel Flow Measurement Structures -
Part 1, Submerged Flow," Utah Water Research Laboratory Report WG 31-2,
College of Engineering, Utah State University, Logan, UT.
Skogerboe, G. V., M. L. Hyatt, J. D. England and J. R. Johnson (1967b).
"Design and Calibration of Submerged Open Channel Flow Measurement
Structures - Part 2, Parshall Flumes," Utah Water Research Laboratory
Report WG 31-3, College of Engineering, Utah State University,
Logan, UT.
Skogerboe, G. V., M. L. Hyatt and L. H. Austin (1967c). "Design and
Calibration of Submerged Open Channel Flow Measurement Structures -
Part 4, Weirs," Utah Water Research Laboratory Report WG 31-5, College
of Engineering, Utah State University, Logan, UT.
Smoot, G. F. and C. E. Novak (1968). "Calibration and Maintenance of
Vertical-Axis Type Current Meters," Techniques of Water Resources
Investigations. Book 8, Chapter B2, U.S. Geological Survey, U.S.D.I.,
Reston, VA.
432
-------
Smoot, G. F. (1974). "A Review of Velocity Measuring Devices," U.S.
Geological Survey Open File Report (April), U.S.D.I., Reston, VA.
Snyder, F. F. (1938). "Synthetic Unit Hydrograph", Transactions.
American Geophysical Union, Vol. 19, pt. 1.
Spencer, E, A. and J. S. Tudhope (1958). "A Literature Survey of the
Salt-Dilution Method of Flow Measurement," Institute of Water
Engineers J., Vol. 12, No. 2.
Torno, H, C. (1975). "Stonnwater Management Models," in Urban Runoff -
Quantity and Quality, W. Whipple, Jr., ed., American Society of Civil
Engineers, New York, NY.
Townsend, F. W. and F. A. Blust (1960). "A Comparison of Stream
Velocity Meters," J. of the Hydraulics Div.. ASCE, Vol. 86, No. HY 3.
U.S. Department of Interior, Bureau of Reclamation, (1967). Water
Measurement Manual, second ed., U.S. Government Printing Office,
Washington, DC.
Vanleer, B. R. (1924). "The California Pipe Method of Water
Measurement," Engineering News Record, August 3 and 21.
Wells, D. M., T. A. Austin, and B. C. Cook (1971). "Variation of
Urban Runoff with Duration and Intensity of Storms", Texas Tech
University Water Resources Center Project Completion Report
No. WRC-71-5.
Wells, E. A. and H. B. Gotaas (1956). "Design of Venturi Flumes in
Circular Conduits," J. of the Sanitary Engineering Div.. ASCE, Vol. 82,
No. SA2.
433
-------
Woolhiser, D. A. and J. A. Liggett (1967). "Unsteady, One-dimensional
Flow Over A Plane: The Rising Hydrograph", Water Resources Research.
Vol. 3, No. 3.
ACKNOWLEDGEMENTS
This lecture handout material has been primarily based upon work per-
formed for the U.S. Environmental Protection Agency under con-
tracts 68-03-0155, 68-03-OA09, 68-03-0426, and 68-01-2568, and the
support by the members of the Storm and Combined Sewer Section of the
Wastewater Research Division (especially R. Field, H. Masters, and
D. Cesareo); of the Municipal Environmental Research Laboratory,
Cincinnati, Ohio, and the Office of Technology Transfer (especially
N. Lailas and R. Crowe(, Cincinnati, Ohio is gratefully acknowledged.
434
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/2-77-065
3. RECIPIENT'S 4CCESSION«NO.
4. TITLE AND SUBTITLE
Short Course Proceedings:
MANAGEMENT MODELS
APPLICATIONS OF STORMWATER
5. REPORT DATE
March 1977
(Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOfl(S)
Editors: Francis A.DiGiano, Donald D. Adrian,
and Peter A. Hangarella
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Civil Engineering
University of Massachusetts
Amherst, Massachusetts 01002
10. PROGRAM ELEMENT NO.
1 BC 611
11. CONTRACT/GRANT NO.
R-803069
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Cin.,OH
Office of Research and Development
U.S. Environmental Protection AGency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
6/75-8/76 Proceedings
14. SPONSORING AGENCY CODE
EPA/600/14
16. SUPPLEMENTARY NOTES
Project Officer Anthony Tafuri 340-6679
16, ABSTRACT
This Short Course on applications of stormwater management models is a follow-up to a
course sponsored by the U.S. EPA and now available as EPA Report 670/2-75-065. The
proceedings contained herein represent an entirely new set of contributions from
participating speakers. The objective of this Short Course is to provide practitioneis
with the capability to apply specific'models directly. Toward this goal, a discussion
of the common components of stormwater management models first gives an overview of
modeling needs. The U.S. EPA Stormwater Management Model (SWMM) is then described in
detail and an illustrative case study presented. The methodology for data preparatior
is outlined and sample input and output data given for the Rainfall-Runoff, Transport,
Storage/Treatment and Receiving Water Blocks of the EPA SWMM. A discussion of ,
criteria for selecting models for application as either planning or design tools is
then presented along with illustrations of the use of two simplified models. Finally
the techniques for collecting field data for model calibration are presented and the
performance of commercially available sampling equipment assessed.
This report was submitted in partial fulfillment of Grant Number 803069 by the
Department of Civil Engineering at the University of Massachusetts, under the sponsor-
ship of the Environmental Protection Agency. This report covers the period
June 3, 1975 to August 31, 1976 and work completed as of August 31, 1976.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Computer programs,* Mathematical Models,
*Storm sewers, *Combined sewers,*Simulat-
ion,*Rainfall intensity,*Runoff Pollution,
*Stream pollution,*Waste treatment, Cost
analysis, Cost effectiveness,Optimization,
*Water Quality
b. IDENTIFIERS/OPEN ENDED TERMS
el
Infiltration,1
storm flow, Combined
.sewer overflows,Urban
runoff,Water quality con-
trol.
COSATI Field/Group
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19'. SECURITY CLASS (ThU Reportj
UNCLASSIFIED
21. NO. OF PAGES
447
20. SECURITY CLASS (TMlpagt)
UNCLASSIFIED
22. PRICE
"A Form 12X0-1 (t-7>)
435
------- |