United States
Environmental Protection
Agency
Office of Air
Land and Water Use
Washington DC 20460
EPA-600/9-78-035
October 1978
Research and Development
-------
EPA-600/ 9-78-035
October 1978
URBAN RUNOFF CONTROL PLANNING
M.B. McPherson
Director, ASCE Urban Water Resources
Research Program
Marblehead, Massachusetts
Project Officer
Harry C. Torno
Office of Air, Land, and Water Use
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
OFFICE OF AIR, LAND, AND V1ATER USE
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
DISCLAIMER
This report has been reviewed by the Office of Air, Land and Water Use,
Office of Research and Development, 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 coninercial products constitute
endorsement or recommendation for use.
-------
FOREWORD
A major function of the Research and Development programs of the
Environmental Protection Agency is to effectively and expeditiously transfer,
to the xoser community, technology developed by those programs. A corollary
function is to provide for the continuing exchange of information and ideas
between EPA and users, and between the users themselves. In this latter
spirit, the Office of Air, Land, and Water Use publishes work which, while
not originally supported by EPA, is of sufficient interest and merit to be
useful to engineers and planners working on EPA-supported programs.
This report, by the ASCE Urban Water Resources Research Council under the
sponsorship of the National Science Foundation, provides insight into the urban
runoff problem, and ashould prove invaluable to those engaged in the planning
for or design of urban runoff pollution abatement or control projects.
Courtney Riordan
Acting eputy Assistant Administrator
Office of Air, Land, and Water Use
111
-------
ABSTRACT
Section 208 of Public Law 92-500 (Federal Water Pollution Control Act
of 1972) encourages areawide planning for water pollution abatement manage-
ment, including urban runoff considerations where applicable. Areawide
studies are under way or planned in just about every metropolitan area.
Deadlines for initial areawide reports are not far off, and it is expected
that many of the agencies preparing reports are presently resolving their
projected activities beyond the current first planning phase. This report
has been prepared to assist agencies and their agents that are participants
in the preparation of areawide plans, from the standpoint of major urban
runoff technical issues in long-range planning. Emphasized is the importance
of conjunctive consideration of urban runoff quantity and quality and the
need to development a factual basis that will support expected reliability
of performance of proposed actions and programs. While not intended as a
handbook for urban runoff control planning, this report delves into sane
important technical issues that are often slighted or poorly handled, such
as the utilization of simulation. Recognizing that the ultimate test of
any plan lies in its implementation, topics are viewed from the perspective
and experience of the local government level where implementation takes place.
IV
-------
TABLE OF CONTENTS
SECTION 1 - INTRODUCTION AND SUMMARY 1
Synopsis 1
Overview 1
Problem 2
Purpose 2
Delivery 3
Intentions and Expectations . 3
Summary, Section 2 - Selected Planning Considerations ... 5
Summary, Section 3 - Incentives for Comprehensive Planning ...... 5
Summary, Section 4 - Some Flow Management Considerations . 6
Summary, Section 5 - Utilizing Simulation 7
Summary, Section 6 - Urban Runoff Models ... 8
Summary, Section 7 - Some Modeling Considerations 8
Summary, Section 8 - References 8
Acknowledgment 8
Principal Information Listings 9
SECTION 2 - SELECTED PLANNING CONSIDERATIONS 11
Definition 11
An Example of Local Government Planning Elements 11
An Illustration of Local Government Planning ...12
A Broader Perspective ............13
Figure 1 - Block Diagram of Urban Runoff Control
Master Planning Concept 15
Marshalling for a Plan ....»., 14
An Emphasis on Flooding and Drainage ....18
Some Metropolitan Examples ...19
Two-Direction Communication and Time Horizons 21
Table 1 - Summary Guidelines for Plan Components
of Continuing Planning Process 22
SECTION 3 - INCENTIVES FOR COMPREHENSIVE PLANNING 23
Synopsis 23
Land-Use Controls 23
Disparities in Economic Evaluations 25
Water Reuse 26
Figure 2 - Uses for Reclaimed Water 27
Extraneous Sewer Flows 28
Erosion and Sedimentation 29
A Metropolitan Issue 29
Conclusions 30
SECTION 4 - SOME FLOW MANAGEMENT CONSIDERATIONS 31
Synopsis ...... 31
Drainage Versus Flood Control 31
Flood Aspects 32
Conjuctive Planning 33
-------
TABLE OF CONTENTS (Continued)
Detention Storage 34
Other Measures . . . , 35
Miscellaneous .......... ..... 37
SECTION 5 - UTILIZING SIMULATION ^33
Synopsis 38
Why Simulation? 39
Background 40
Performance Reliability 42
Figure 3 - Subjective Estimate of Increased Reliability
and Associated Costs for More Thorough Analysis . . 43
Testing Simulation Models . . 44
Table 2 - Physical Characteristics of Representative
Catchments Being Gaged in Philadelphia 45
Long-Term Precipitation Record as a Reference 46
Figure 4 - Sample of Yarnell Charts 48
Figure 5 - Relation for Various Durations 48
Developing Frequency Rationales for Levels of Protection ...... 49
Simulating a Range of Storms and System Loadings 49
SECTION 6 - URBAN RUNOFF MODELS 51
Contents 51
Role of Simulation in Planning 51
Some Reservations ..... 52
Categories of Model Applications .... 53
Models for Planning Applications 54
Figure 6 - "STORM" Simplified Logic Diagram 55
Models for Analysis/Design Applications 57
Model Comparisons 57
Receiving Water Modeling 58
SECTION 7 - SOME MODELING CONSIDERATIONS 61
Pitfalls 61
Land-Use Data 61
Flood Plain Mapping 63
Storage Manipulation Via Automatic Control 64
Ralngage Networks 66
EPA Overview 68
SECTION 8 - REFERENCES 69
ADDENDUM 1 - METROPOLITAN INVENTORIES 88
ADDENDUM 2 - THE DESIGN STORM CONCEPT 100
ADDENDUM 3 - NOMOGRAPHS FOR TEN-MINUTE UNIT HYDROGRAPHS FOR
SMALL WATERSHEDS 119
ADDENDUM 4 - RESEARCH ON THE DESIGN STORM CONCEPT 153
-vi-
-------
SECTION 1
INTRODUCTION AND SUMJ&RY
Synopsis
Section 208 of Public Law 92-500 (Federal Water Pollution Control Act of
1972) encourages areawide planning for water pollution abatement management,
including urban runoff considerations where applicable. Areawide studies are under
way or planned in just about every metropolitan area. Deadlines for initial areawide
reports are not far off, and it is expected that many of the agencies preparing
reports are presently resolving their projected activities beyond the current first
planning phase. This report has been prepared to assist agencies and their agents
that are participants in the preparation of areawide plans, from the standpoint of
major urban runoff technical issues in long-range planning. Emphasized is the
importance of conjunctive consideration of urban runoff quantity and quality and the
need to develop a factual basis that will support expected reliability of performance
of proposed actions and programs. While not intended as a handbook for urban runoff
control planning, this report delves into some important technical issues that are
often slighted or poorly handled, such as the utilization of simulation. Recognizing
that the ultimate test of any plan lies in its implementation, topics are viewed from
the perspective and experience of the local government level where implementation
takes place. On-site elaboration of the principles presented in this report, and
their adaptations to local conditions, will be made in seminars, workshops and
conferences to be led by the author in a number of metropolitan areas over 1977-19-79.
During that period this report will be updated and supplemented in a series of
subsequent releases.
Overview
Perhaps never before in our history has there been as rapid a change in the
public's attitudes towards its surroundings as has occurred in the past few years.
This haa beeu fueled by unprecedented world-wide population growth, a resultant
intensified international communication of pollutants as well as information, and a
growing appreciation of the finite limitations of the resources of our planet.
Manifestations are concerns over environmental protection, more rational use of land,
better husbandry of energy and other resources and greater relief from natural disasters.
Because three out of four Americans live in urban settlements that occupy less than
one-fortieth of our land area, it is equitable to say that the winds of change are
predominantly urban in character, or at least that urban dwellers will be the larger
number affected.
The growth impact of projected areal enlargement of urban areas'1'^' on
present and planned urban water resource facilities is almost too stunning a reality
to be comprehended fully at this time. Even if expected urban growth is checked by a
renaissance of our central cities, the required reconstruction will still be monumental.
On the basis of past performance, it has been asserted that while the
construction phase of water resource project implementation takes somewhere around
1 to 5 years for smaller projects and around 5 to 15 years for larger projects, because
of protracted delays in conflict resolution the true lead-time necessary for
accomplishing the objectives sought has been on the order of 20 to 25 years,
references are listed in Section 8.
- 1 -
-------
particularly for larger projects, including certain large urban water supply
developments.(3) Anticipated geometric increases in the impacts on urban water
resources suggest that total project development time must be substantially compressed.
Thus, means must be found for accelerating conflict resolution, the element most likely
to cause protracted delays, despite the fact that new works will increase steadily in
hydrologic complexity because of a growing interdependence among project components.
That is, local governments are confronted with public demands for instant solutions
while simultaneously there is more direct involvement by the public in decision-making
as the problems and their solutions become increasingly more complex. On top of all
this is the shifting of targets as national and State laws and regulations are
constantly revised and augmented in an upward spiraling of stiffened requirements.
There are numerous instances where elaborate plans have had to be abandoned or
drastically altered because the "rules of the game" had suddenly become more demanding,
Problem
Considerable planning activity is being focused on the improved management
of all types of water pollution, including that from urban runoff. Section 208 of
Public Law 92-500 (Federal Water Pollution Control Act Amendments of 1972) encourages
areawide management planning in areas which, as a result of urban-industrial
concentrations or other factors, have substantial water quality control problems.
The 208 studies are resulting in a many-fold increase in planning activities on urban
runoff pollution management, with studies under way or planned for just about every
metropolis in the nation. While an opportunity is presented through these studies
"to plan and manage a comprehensive program based on integrated planning and control
over such activities as municipal and industrial wastewater, storm and combined sewer
runoff, nonpoint source pollutants, and land use as it relates to water quality,"^'
evidence is accumulating that, in order to meet initial reporting requirements, many
of these plans may favor short-term solutions over the meeting of long-range goals.
Because knowledge on quantity and quality of stormflows is limited, long-
range planning should be accompanied by programs for the acquisition of local field
data, and its analysis by such means as simulation techniques. The 208 study schedule
does not provide sufficient time to develop a suitable factual base where it did not
already exist, and as a result in many metropolitan areas problem identification,
and definitive planning are expected to continue well beyond current initial reporting
activities.t5'
Purpose
The major purpose of this report is to encourage long-range comprehensive
planning for urban runoff quality and quantity management, as a supplement to
applicable current Section 208 planning.
Another purpose is to stress the importance of planning not only for
control of urban runoff quality but also quantity, a major economic and environmental
consideration that might be inadvertently slighted in the new national emphasis on
quality management.
Only a very few instances have been identified where local government
planning that was started several years ago for urban runoff control has conjunctively
addressed both quality and quantity management aspects. These and other but narrower
leading-edge examples are cited whenever methods or procedures that have been used
might be adapted to advantage elsewhere.
- 2 -
-------
This report is not a handbook for urban runoff control planning, and it
is not a definitive exposition of the planning process. It does delve into some
important technical issues that are often slighted or poorly handled, such as the
utilization of simulation.
A synthesis of local government perspectives on planning and associated
problems is presented with the intention of facilitating metropolitan intergovernmental
cooperation and coordination. Because an intended audience is Section 208 planning
agencies, a central purpose is to bring to their attention some potential technical
impediments that might frustrate effectuation of their proposals. There are cautions
also, directed at local governments, that are intended to be reminders of some
technical pitfalls that should be avoided in their dealings with areawide plans.
In short, we hope that the uninvited intervention of an outside referee will be of
some assistance in improving the effectiveness of the participants in a very complex
process.
Delivery
This report is being distributed to area-wide planning agencies and
supportive entities in all metropolitan areas. On-site elaboration of the principles
presented herein, and their adaptation to local conditions, will be provided in
seminars, workshops and conferences conducted in a number of metropolitan areas over
1977-1979, As the delivery phase progresses, special topics not treated adequately
or sufficiently elsewhere that are found to need synthesis and summarization will be
reviewed in subsequent Addenda to this report. The two Addenda included with this
report are examples of what is intended. Only a fraction of the metropolitan areas
can be assisted via site visits made for the purpose of elaborating the report's
contents, but through the distribution of post-report Addenda to all metropolitan
planning agencies a wider audience will be reached. Also, the Addenda will serve as
a vehicle for updating the content of the present report.
Intentions and Expectations
Just about all broad-scale, leading-edge, local government plans for urban
runoff control we have encountered are called "master plans," and although this term
appears to be abhorred by many metropolitan planning agencies we do not propose a
wholesale semantic reform. This difference in viewpoint probably springs from the
realistic need for areawide planning agencies to preserve the concept of a "living
document" in principle and in fact, whereas local governments cannot afford this
luxury because any planning dealing with capital improvements must lead to acts of
implementation, and once construction starts the die has been cast.
The conventional local government master plan has generally dealt with only
one aspect of water resources such as water supply, or wastewater^ treatment, or
flood control, or pollution abatement. We favor and encourage the emerging type of
comprehensive planning that provides integrated consideration of flooding protection,
pollution abatement including erosion control, and water conservation. Even some of
the advanced examples cited in this report have integrative deficiencies. There are
simply not many local government plans under preparation that are truly comprehensive
because the impact of changing policies and objectives has not been factored into
many such plans. Moreover, because comprehensive local government urban water
resource planning is a relatively new concept, very few of the broader plans have
been carried to the threshold of Implementation.
- 3 -
-------
We have no illusions about the possibilities for wholesale change. Perhaps
the most we can hope for in the near term, in a majority of metropolitan areas, is a
conjunctive planning approach, at least with regard to the quantity and quality
aspects of urban runoff. Comprehensive planning faces a number of obstacles, including
fractionalized authority of administrative agencies and balkanized local governments
in metropolitan areas,(6) specialist approaches taken individually on each water
service component, and uneven advances in knowledge on the various components, The
term "water management," widely used only in recent years, has not yet acquired a
specific meaning. Although rational control of water is its objective, there is no
common understanding of needed policies and institutional arrangements for achieving
rational control. On the other hand, metropolitan area planning is being conducted
formally almost nationwide, although the planning function is normally advisory, only;
and in terms of comprehensive water resource development, the planning, implementation
and operation of works is usually fragmented in both the central cities and in their
metropolitan districts. Even water and wastewater services are commonly fragmented in
metropolitan districts (e.g., our largest metropolis is served by some 400 separately
managed water agencies). A major cause of administrative fragmentation is the
existence of often gross dissimilarities in geographic areas: viz., differences in
hydrologic, service and revenue areas, and in political jurisdictions. One of the
consequences is that data collection and its analysis and related research and
development is currently pursued independently by individual local government
departments and metropolitan special districts, with only limited collaboration on a
national or even metropolitan scale. Therefore, early coordination by local governments
of data collection and its analysis is strongly advised.
"Although metropolitan-wide planning and coordination is necessary for
successful urban water resource management, the key to the implementation of such
planning generally lies at the local government level and at the State level in terms
of what the State enables or requires local government to do."(7) (Emphasis in
original). '
While urban water resource implementation agencies may do some long-range
or strategic planning, their basic functions as mandated by authorized responsibilities
are management activities that emphasize specific local demands, a narrow range of
alternatives, detailed rules for operation and maintenance, and the inclusion of
measures such as service charges and service extension policies.^)
"There have been some attempts to define an integration of regional planning
with local guidance and management, but there have been few, if any, successful
attempts to apply it."^''
Mare comprehensive management approaches would integrate or give due
consideration to all aspects of water in a metropolis, together with related
environmental, energy and other relevant considerations. The need for a comprehensive
overview becomes more crucial as the complexity of urban areas increases, caused by
such things as population migration and growth, competition over available energy
supplies, rising expectations of urban dwellers in the face of inflation of local
government costs, and greater public awareness of environmental issues and enlarging
demands for public participation. A means for assuring a comprehensive overview is
the conduct of metropolitan water balances, or inventories, described in Addendum 1
of this report.
- 4 -
-------
Summary. Section 2 - Selected Planning Considerations
Based on experiences with a leading-edge master plan for combined sewer
overflow pollution abatement, a set of planning elements is presented that should
serve as a useful check list.
The same master planning effort is reviewed in terms of project financing
scenarios for meeting a variety of possible abatement goals. Illustrated by this
example plan is the urgency of incorporating sufficient flexibility in a plan to
satisfy future shifts in regulationsi_a_s_impleinentation proceeds.
Six discrete steps can be identified in the typical planning exercise:
problem definition, establishment of goals and identification of constraints; data
assembly and problem redefinition; formulation of alternatives; analysis of
alternatives; identification of more promising alternatives; and evaluation of the
trade-offs among selected alternatives. These fundamental steps are elaborated in a
detailed block diagram of the urban runoff control master planning concept that
incorporates all features of several less comprehensive depictions. Noted
parenthetically is that a local agency can become so obsessed with long-range planning
that more immediate, partially mitigative solutions can be overlooked or inordinately
deferred.
Next, an example is given of a master plan for combined sewer overflow
pollution abatement that was launched from the metropolitan level. The careful
procedures that were followed in marshalling for the plan are detailed because they
provide an example worth emulating.
Examples are then presented that include: a County master plan for flood
control, erosion control and drainage; the master plan for the largest combined sewer
overflow pollution and drainage control project in the nation; flood control master
planning for a metropolitan area; and inferences of the Urban Studies Program of the
Corps of Engineers.
The examples given in this Section deserve consideration by many readers
because they are prototypes in many respects of desirable practices and they suggest,
by what they do not include, possibilities for improvements elsewhere.
Lastly, the cyclic process of reviewing alternatives in a continuing
planning process is briefly considered. Noted is the need for local government
entities to be in close communication with regional planning agencies. That is. a
jinn, positive linkage should be forged between metropolitan planning and local
government implementation. Noting that the detailed local government master plans
alluded to earlier in this Section have required a minimum of three to five years for
their development, it is important to accept that the local government planning time-
frame must be measured in terms of a number of years.
Summary. Section 3 - Incentives for Comprehensive Planning
The thesis of this Section is that rational planning requires conjunctive
consideration of the quantity and quality aspects of urban runoff within a
comprehensive multiple-use framework. Some of the more obvious arguments are
explored, in connection with public viewpoints on land-use controls, disparities in
approaches to economic evaluation, potentials for water reuse, opportunities to reduce
extraneous sewer flows, control of erosion and sedimentation, and disposal of solids
from new joint or ad hoc treatment facilities.
- 5 -
-------
Acting against the conjunctive approach is the tendency of citizen groups
and Federal laws to dwell on land-use controls as a panacea while minimizing the
interrelations and interconnections inherent in water resources. This point surfaces
recognition of some of the challenges confronted by conjunctive planning and the
great importance of public information, and even public education, in promoting the
larger view.
Economic evaluation of water quality control projects at the local level
appears to be restricted to some form of minimum cost analysis, as opposed to flood
control economic justification via cost-benefit analysis at the river basin scale
Because of the trend towards the blending of river basin planning with metropolitan
planning, local agencies will increasingly face the need to reconcile this dichotomy
of project evaluation criteria. Further, these disparities in economic bases for
project evaluation would seem to suggest rather strongly that planning for management
of urban runoff quantity and quality should be conjunctive to avoid absolute
confusion.
After briefly reviewing various national developments and trends it is
concluded that the reuse of wastewater treatment plant effluents and reclamation of
surface runoff is more likely to be via artificial recharge of groundwater supplies
in general. This eventuality can be provided for only through conjunctive planning*.
A national Symposium concluded that, because the majority of publicly-owned
wastewater treatment works will be committed by the time Section 208 plans are ready
for review and implementation, the only questions that will still be open by that
time will be nonpoint sources of pollution and growth management and who pays for
growth. Thus, land-use control issues will continue to simmer for some time.
Erosion and sedimentation control is one of several water quality issues
that cannot be divorced from their causative driving force, urban runoff.
With regard to treatment of stormwater and combined sewer flows, jiandljng
water volumes is only part of the problem, for once solids conveyed by water are
removed they must be disposed of somewhere.
The point of resolution for inconsistencies, conflicts and duplications of
legal instruments imposed by higher levels is at the local level of government. It
is concluded that for this reason, among several, local governments are forced to
define their own problems.
Summary, Section 4 - Some Flow Management Considerations
Because urban drainage and flood plains are almost always the responsibility
of separate and different kinds of jurisdictions, their interconnected behavior is the
responsibility of neither type of organization. A similar splintering of authority
over water quality aspects plagues most metropolitan areas. Given this institutional
patchwork, about the only avenue for conjunctive consideration of these otherwise
disparate issues is via comprehensive planning and coordination at the metropolitan
level.
Asserted in this Section is that the guiding principle should be to reduce
the liabilities and increase the assets of urban runoff.
After reviewing a range of techniques applicable to urban runoff control
the reader is warned that although numerous schemes have been postulated for
- 6 -
-------
controlling the quantity and quality of urban runoff, very few of these concepts
have been tested in full-size system applications^, and the research on which many of
them are based is fragmented and the findings cannot be sufficiently generalized for
universal applications. By default, local governments are obliged to assure themselves
of the relevance and reliability of most such schemes before making a commitment to
their wholesale use. This leads to an extension to the last conclusion above for
Section 3: local governments are forced to define their own problems and to take the
initiative in finding solutions to those problems.
Summary, Section 5 - Utilizing Simulation
While simulation is undoubtedly an effective means for defining problems and
analyzing alternative urban runoff control strategies, its most important use can be
in the assessment of expected system performance. Performance reliability is
ultimately the fulcrum of local government political acceptability.
This Section commences with reasons for using simulation and a historical
reckoning of practical reservations against incautious acceptance of runoff models.
Emphasized is that simulation techniques adopted should not exceed the level of
mastery of such tools by the user and that tools should be selected on the basis of
their suitability for solving problems.
Performance reliability is defined in terms of four serially connected
principal considerations, which must be completed in the following order to be fully
effective: testing simulation models against local field data; using a long-term
precipitation record as the basic reference; developing frequency rationales for
levels of protection; and simulating a range of storms and system loadings. An
emphasis is placed on the testing of simulation models against local field data
because this _issyie_JLSthe crux of credibility as well as reliability. As things now
stand, local governments are substantially on their own for the acquisition of field
data.
The concept of spatial sampling for field instrumentation of catchments
having representative land uses is outlined. Stressed is the importance of rapid
application of data to models, to check data reliability but principally for
expeditious calibration of models, the primary use of such data in planning.
Control of flooding and water pollution must be based on probabilities of
occurrence because of the randomness of precipitation. Rationales are developed for
using long-term rainfall records as the reference for defining urban runoff quality
and quantity control objectives in probabilistic terms. Addendum 2 is an extension
of this presentation.
Underscored in this Section is that the only realistic defense for planning
in an atmosphere of ambiguous policy is to employ procedures the results of which have
an inherent flexibility for conversion to meet alternative policy goals.
Reliability in the employment of calibrated tools for the exploration of
alternatives planned for the future is a function of several things, but the most
important are probably the shrewdness of design of the field data network and__the
extent to which its results approach an ideal set of representative samples.
- 7 -
-------
Summary. Section 6 - Urban Runoff Models
Reiterated in this Section is the importance of calibrating any urban
runoff model against local field data.
Models are characterized in this Section in terms of their applications.
Specific models are identified, with an emphasis on sewered system planning
applications, although associable analysis/design applications are duly noted.
Receiving water modeling is handled separately because of the tendency towards
customized adaptation of available models for that purpose. Examples are given of
some regional receiving water simulation studies.
Summary. Section 7 - Some Moclelinfi Considerations
Models can be used ineffectively if they are selected and employed without
prior careful consideration of their data requirements and their place in overall
information processing.
While flood plain mapping of 100-year stream stages is a minimum requirement
the mapping of more frequent occurrences is required for analysis of the economics of
flood damage mitigation.
Permanent field data acquisition systems for operations, particularly for
automatic control, can be readily justified, whereas only temporary installations can
be justified for other purposes except for long-range monitoring.
The three themes described above are singled out in this Section because
they are particular pitfalls that have been encountered in past planning efforts.
Also, the latest developments in the manipulation of storage via automatic
control are described.
Summary. Section 8 - References
As a convenience to the reader, references are divided according to the
Section in which they are cited. Considerable discrimination was employed in
selecting these references from an initial listing with over three times as many
entries. Rather than cite, say, six related references, the writer chose to use
either the latest (which itself cites the ones omitted) or the one or two with the
most useful information. Some references are cited to acknowledge credit for ideas
or material while others are mentioned merely to document a point being made.
Knowing that every reader will have different interests and a unique
familiarity with the body of literature involved, it would be folly to attempt to
single out a short list as the "most important reading". The need to cite numerous
references where each dwells on.a narrow portion of the spectrum merely confirms a
contention in this report that research and development on urban runoff control has
travelled a disjointed, almost haphazard journey, with numerous efforts pursued
independently.
Acknowledgment
This report has been prepared on behalf of and with the assistance of the
ASCE Council on Urban Water Resources Research by its full-time operating arm, the
ASCE Program bearing the same name. The activities and products of the Program are
- 8 -
-------
well known.(°) Suffice it to say that the central objective of the Council is to
help advance the state-of-the-art by identifying and promoting needed research and
by facilitating the transfer of the findings from research to users. A Steering
Committee designated by the ASCE Council gives general direction to its Program:
Mr. S. W. Jens (Chairman); Dr. W. C. Ackermann; Dr. J. C. Geyer; Mr. C. F. Izzard;
Mr. D. E. Jones, Jr.; and Mr. L. S. Tucker, Mr. M. B. McPherson is Program Director
(23 Watson Street, Marblehead, Mass. 01945). Administrative support is provided by
ASCE Headquarters in New York City.
Financial support for the preparation and duplication of this report is
entirely from a grant to ASCE by the Research Applied to National Needs program of
the National Science Foundation. Technical liaison representative for NSF/RANN is
Dr. J. Eleonora Sabadell. Any opinions, findings, and conclusions or recommendations
expressed herein are those of the author and do not necessarily reflect the views of
the National Science Foundation. The ASCE Council is greatly indebted to NSF/RANN
for supporting this nine man-month project.
Formal review of the two drafts of this report was generously provided by
an ad hoc Advisory Panel assisting the ASCE Council:
Mr. Dennis Athayde, NPS Branch, Water Planning Division, EPA, Washington, D.C.
Mr. Harold C. Coffee, Jr., Department of Public Works, San Francisco
Dr. Charles N. Ehler, Office of Research and Development, EPA, Washington, D.C.
Dr. Neil S. Grigg, UNC-WRRI, North Carolina State University, Raleigh
Mr. Forrest C. Neil, Metropolitan Sanitary District of Greater Chicago
Mr. Harry C. Torno, Office of Research and Development, EPA, Washington, D.C.
Mr. L. Scott Tucker, Urban Drainage and Flood Control District, Denver
Dr. Stuart G. Walesh, Southeastern Wisconsin Regional Planning Commission, Waukesha
Dr. Donald H. Waller, Nova Scotia Technical College, Halifax
Information, assistance and review were also provided by scores of other
persons from among the hundreds of ASCE Program cooperators. In particular, vital
assistance was provided by officials responsible for some of the more innovative
master planning projects. However, none of the persons named or alluded to in this
subsection are to be held responsible for any of the contents of this report. They
have nevertheless helped insure a better professional consensus on the issues involved
and the ASCE Council and the writer are therefore greatly indebted to them for their
invaluable assistance and generous cooperation.
A significant portion of this report draws upon relevant parts of a prior
report that was prepared for the U.S. Geological Survey.(*'
Typing of this report and its drafts was by Mrs. Richard Symmes of
Marblehead, Mass.
Principal Information Listings
Some readers may be interested in the sources of principal information
listings in urban water resources research. Brief descriptions of on-going research
on specific subjects are available from the Smithsonian Science Information Exchange.
One service of the Exchange is the availability of "research information packages"
that are updated every 90 days. Of particular interest here are the packages on urban
water management, urban runoff, and effects of storm water runoff on water quality in
receiving streams,
- 9 -
-------
For completed research, reference should be made to the selected water
resources abstracts issued twice per month by the National Technical Information
Service that are compiled by the Office of Water Research and Technology^11/ Nearly
all of the reports cited in the abstracts that have been supported by public funds
are available for a cost-recovery charge from NTIS, in which cases the NTIS
identification numbers are included in the abstracts. Information searches on urban
water publications can be greatly facilitated by referring to the annual cumulated
indexes of the selected water resources abstracts.(12) In addition, NTIS has packages
of bibliographies with abstracts on some specific subjects, such as urban surface
runoff. Both NTIS and SSIE also make custom searches of completed and on-going
work, respectively.
- 10 -
-------
SECTION 2
SELECTED PIANNING CONSIDERATIONS
Definition
Comprehensive, areawide planning is conducted by metropolitan planning
agencies. Implementation of such plans or their variants is by local units of
government. Moreover, more detailed planning is done by local governments and their
plans are carried through construction into operations. The term "comprehensive
planning" thus means quite different things at the metropolitan and local levels.
Also, local government plans can follow, be concurrent with, or lead areawide plans.
It would be totally presumptuous and entirely inappropriate to attempt in this
report to instruct regional planning agencies on the nuances of their responsibilities.
Rather, a major purpose is to present a synthesis of local government perspectives
on planning and associated problems which hopefully will facilitate metropolitan
intergovernmental cooperation and coordination. Because an Intended audience of
this report is Section 208 metropolitan planning agencies, a central purpose is to
bring to their attention some potential impediments that might frustrate effectuation
of their proposals. There are cautions also, directed at local governments, that
are intended to remind them of technical pitfalls that should be avoided in their
dealings with areawide plans. In short, we hope that the uninvited intervention of
an outside referee will be of some assistance in improving the understanding of the
participants in a very complex process.
Regardless of the governmental level involved, when we use the term
"comprehensive planning" we are referring to a multipurpose scope, such as for
integrated management of flooding protection, pollution abatement including erosion
control, and water conservation. That is, despite the label of comprehensiveness,
we are really addressing only a subset of the overall metropolitan comprehensive
planning matrix. Just about all broad-scale, leading-edge, local government plans
we have encountered are called "master plans," although this term appears to be
abhorred by many regional planning agencies. .There are not many local government
plans under preparation that are truly comprehensive because the impact of newer
policies has not been factored into many such plans and, as a result, there are
simply not many broad urban runoff control plans under way and those that are
usually have been initiated to comply with a State directive against a particular
local government. Further, comprehensive urban water resource planning is a
relatively new concept to local governments and very few plans have been carried to
the threshold of implementation, even for the narrower objective of urban runoff
control.
An Example of Local Government Planning Elements
What is called for above is essentially a systems approach to facilities
and management planning. For example, elements needed in water quality control
combined sewer system planning are said to include the following:(")
. An analysis of implementing governmental entities.
. A public information system.
. A set of goals, consistent with existing and proposed regulations for the foreseeable
future.
. A realistic and attainable set of regulations.
. A definite understanding and description of the problems and their analysis.
. A complete description of the causes of the problems.
(Continued)
- 11 -
-------
. A finite description of the existing physical system and its inadequacies.
. A finite description of the receiving-water environment for biota.
. A finite description of solids disposal locations.
. A detailed rainfall history perhaps spanning 30 to 50 years.
. A detailed history of water-demand and water-use factors.
. A detailed pollutant-constituent history.
. A complete water quality inventory showing constituents in the water supply, in
wastewater prior to treatment, and in effluents.
. A means for evaluating treatment in terms of storage and overflow options.
. A series of viable alternatives based on real data.
. A recommended solution or set of solutions.
. A flexible facilities and operations staging program.
. A practical financing program.
. A source of funds.
Experience indicates that a steering committee comprised of outside experts can be
very helpful in avoiding oversight or slighting of the above elements. While the
above elements may be regarded as idealistic by some, the local government that ignores
them does so at its peril. Because of unique combinations of local conditions, it
should be evident that satisfaction of the needs listed above would necessarily vary
in detail, methodology and procedure from one local jurisdiction to another, even
in the same State. However, the institution of a public information program early in
a local government planning endeavor is very important.
An Illustration of Local Government Planning
One of the most outstanding comprehensive, imaginative and thoroughly
developed plans for municipal pollution control has been prepared by the Department
of Public Works, City and County of San Francisco, California. San Francisco is
served almost exclusively by a combined sewer system. The plan is founded on a
data bank of unusual scale, ranging from runoff-quality field measurements initiated
in 1965, through installation of a pilot wet-weather dissolved-air flotation treatment
plant on a field catchment, to the use of a unique automatic rainfall-runoff
monitoring system. A deliberate step-by-step approach was used in developing the
plan, which was carefully phased with field information acquisition. The quality of
the engineering work in the development of the plan was superb, and the plan was
supported by several DPW studies and 25 ancillary reports prepared by consultants.
Pioneering development of automatic control capability by the DPW is discussed in
Section 7.
Water quality standards for both dry-weather and wet-weather flows have
presented a moving target for pollution abatement compliance by local agencies, and
more changes in regulations may be anticipated in the future. Recognized in the
development of the "master plan" for wastewater and combined sewer overflow control
in San Francisco was that for various abatement plans the timing imposed by Federal
and State regulatory agencies and subsequently supported by their funding is the
critical factor controlling the final selection of such a plan. The flexible plan
adopted by the City is amenable to adjustment during its implementation to meet
future shifts in regulations, but the feature to be discussed here is concerned
instead with project financing scenarios for meeting a variety of possible abatement
goals.
The preliminary master plan proposed in 1971 provided four alternates that
would allow an overflow to occur from a maximum of eight times per year to a minimum
of once in five years. Because the amount of Federal funds and the eligibility
- 12 -
-------
requirements for grants had changed almost continuously, and wet-weather projects had
a lower priority than dry-weather projects (and were not scheduled for either
California or Federal grants), the six variables affecting the scheduling of a bond
issue were analyzed:
. Frequency of overflow occurrence (4 alternates).
Effluent disposal site for a wastewater treatment plant (3 sites).
. Level of dry-weather treatment (3 levels).
. Number of years for completion of all facilities (10 to 60 years).
. Per cent of total costs obtained from grants (07. to 807.).
. Amount of the City's financial resources that might be allocated (four values).
Results of the analysis of the 1,944 combinations for each of the four
alternates were presented in a single ingenious graph displaying all six variables.
Thus, for a given overflow control level, a given effluent disposal site, a given
dry-weather treatment level, a given number of years for facilities completion, and a
given per cent support by grants, the amount of the City's resources that would be
required can be determined from the correlation graph. Through this graph the question
of who would pay what share of the costs for meeting a particular water quality
objective under a particular set of circumstances becomes immediately clear. This is
an excellent illustration of the fulfillment of an engineer's responsibility in this
instance: to present the range of options available to officials with the authority
to make decisions and recommendations. In essence, the Department's engineers
provided the City's elected officials with the means for negotiation with regulatory
agencies having both enforcement powers and any funds for supportive grants.
While the above description is for abatement in terms of the average number
of combined sewer overflows permitted per year, this could be readily converted into
overflow volumes or pollutant loads or seasonal serial events instead, by reference
to the City's performance simulation results described elsewhere in this report.
Parenthetically, although the basic featuresC15) of the preliminary master
plan^ ' have not been altered, some modifications were made in 1973(1'' and again in
1976(18.) when implementation of the first phase commenced.
Of the 28,000 acres in San Francisco, 24,000 are drained by the sewer system
(plus 2,000 acres in adjoining San Mateo County). While the City is surrounded on
three sides by receiving waters, the Pacific Ocean and the Bay, strearaflow flooding
does not exist in the City and therefore was not a consideration in development of
the master plan. Also, the City is essentially completely developed and thus is
not involved in the conversion of low-density land into suburban expansion. We now
turn our attention to the more general situation where both sewer systems and streams
drain the land and where both existing and projected drainage systems are involved.
A Broader Perspective
A review of a number of local government planning projects has revealed a
commonality of six discrete steps, which might be regarded as the framework of the
standard planning process;(19)
. Define problems, establish goals and identify constraints.
. Assemble data and refine definitions of problems.
. Formulate alternatives.
. Analyze alternatives.
. Identify more promising alternatives.
. Evaluate the trade-offs among selected alternatives.
- 13 -
-------
Embracing the above six parts of the basic planning process, Figure 1
is a block diagram depicting the urban runoff control planning concept, which attempts
to accommodate situations with sewer systems and streams and land already developed
and projected for development. The term "master planning" appears in the title to
insure a local government application connotation. The sequence of steps shown is
not intended to specify the scheduling of the steps or the degree of their completion
needed, but the order in which decisions should be reached in order to assure an
efficient planning progression. Considerable feedback would be involved from any
given major step to various preceding steps, but no attempt is made to show the
raorass of possibilities on the diagram because it is complicated enough as it is.
Simulation and related hydrological and water quality field data acquisition aspects
are discussed in detail in Section 5.
In the preparation of Figure 1, all the significant features of three
selected representations were also taken into account, for water flow control(20)
and water quality controlv^l»22) local government planning.
For conjunctive flow control and quality control planning, water quality
considerations tend to dominate. This is primarily because flow control is concerned
solely with quantity while quality control is concerned with the complex relationships
between runoff and its pollutant burdens. A compelling reason for engaging in
conjunctive planning, however, is that the overall flow processes for quantity and
quality planning are shared in common, making it difficult to justify two independent
planning efforts which draw upon the same field information resources and often use
the same or similar or related tools of analysis. Other arguments favoring
conjunctive planning are presented throughout the remainder of this report.
Before leaving Figure 1, it is important to note that a local agency can
become so obsessed with long-range planning that more immediate, partially
mitigative solutions can be overlooked or inordinately deferred. Zoning ordinances,
flood insurance and incorporation of detention storage in new land development are
examples that come quickly to mind. Preparation of most local government plans
completed to date has required a minimum of three to five years. There are interim
or initial actions that can be taken early in this period that do not require
significant local government funding and hence are not likely to be as highly
controversial politically.
Marshalling for a Plan
Mentioned earlier was that the first phase of the San Francisco master plan
is being implemented, reflecting the effectiveness of its careful preparation. We
now turn to a planning effort recently completed for metropolitan Milwaukee. To be
discussed here are the deliberate, careful steps that should be taken to organize
and marshall forces for the conduct of plan preparation. While the San Francisco
effort originated with and was pursued by an operating department of a municipality,
that for the Milwaukee area was launched from a metropolitan level. The description
which follows is a digest of coverage in an earlier ASCE report.(9)
A preliminary engineering study leading to a plan for the abatement of
pollution from combined sewer overflows in Milwaukee County was initiated by the
Metropolitan Sewerage District of the County, a special-purpose metropolitan agency.
The combined sewers in the County are owned and operated by the City of Milwaukee and
one of its adjacent suburbs, the Village of Shorewood, both of which are entirely
within the County and the Sewerage District. Three rivers pass through the area
served by combined sewers: the Kinnickinnic, the Monoraonee and the Milwaukee Rivers,
- 14 -
-------
INVENTORY
EXISTING
DRAINAGE
SYSTEMS
CALIG
SIMUL
MOO
IRATE
ATION
ELS
ASSEMBLE
PRECIPITATION-
RUNOFF-
QUALITY
DATA
1
ACHIEVE
ACCEPTABLE
UlCTTIRirAI
PERFORMANCE
SIMULATION
SPECIFY
FORESEEABLE
FLOW AND
QUALITY
OBJECTIVES
ASSEMBLE
LONG-TERM
uiernpfCAl
RAINFALL
DATA
POSTULATE
ALTERNATF
SOLUTIONS
SPECIFY
EXPECTED
TIME, MONEY
AND POLICY
CONSTRAINTS
L
A
SIMULATE
PERFORMANCE
FOR
ALTERNATES
ACQUIRE
AUXILIARY
FLOW AND
QUALITY
DATA
1
DEVELOP
CRITERIA
FOR
ECONOMIC
ANALYSIS
*
SUPPLY
FLOODING
DISBENEFIT
PARAMETERS
EVALUATE
ATTAINABLE
TREATMENT
FACILITY
EFFICIENCIES
REVIEW
UPDATED
OBJECTIVES
AND
CONSTRAINTS
DETERMINE
OPERATION
AND
MAINTENANCE
LATITUDE
EVALUATE
EFFECTIVENESS,
PERFORMANCE
AND COSTS OF
ALTERNATES
SUPPLY
UNIT
COSTS
1,
SCREEN
Al TCBklATT
*»LI CLPfNAI C
APPROACHES
DETERMINE
BEST
LOCATIONS
FOR
TREATMENT
1.
i
SELECT
BASIC
APPROACH
L
1
ANALYZE
A
D AftJfSF
nMN\>t
OF
ALTERNATIVES
REVIEW
OTHER
WATER
RESOURCE
CONSIDERATIONS
DETERMINE
CAPITALIZATION
LATITUDE
1 DEVELOP PLANNING GUIDELINES
DETERMINE
POLITICAL
FEASIBILITY
OF LAND-USE
REGULATIONS
REPORT
FINDINGS
TO
ELECTED
OFFICIALS
NEGOTIATE
ALTERNATIVES
WITH
REGULATORY
AGENCIES
REVIEW
AND
UPDATE
PLAN
FIGURE 1 - BLOCK DIAGRAM OF URBAN RUNOFF
CONTROL MASTER PLANNING CONCEPT
- 15 -
-------
all in the Lake Michigan Basin. The three rivers merge in the downtown area of the
City of Milwaukee and enter Lake Michigan through a common channel. The mouth, the
confluence and reaches part way up each river are navigable and used for various types
of commercial shipping.
Alternative abatement methods for the master plan were investigated as a
part of a comprehensive study of water-related problems of the Milwaukee River
Watershed conducted by the Southeastern Wisconsin Regional Planning Commission (SEWRPC)
at the request of the Common Council of the City of Milwaukee and the Board of
Supervisors of Milwaukee County. After three years of intensive work by SEWRPC
retained consultants and a technical advisory committee comprised of Federal. State
and local citizens, public officials and technical experts, the report(23,24) was
published and later adopted as a regional guide for development and for solution to
existing problems. That part of the report that dealt with the combined sewer
problem investigated many different methods of abating pollution from combined sewer
overflows and recommended, based on the pre-feasibility investigation, the use of a
combination of methods entitled "Deep Tunnel/Mined Storage/Flow-Through Treatment".
The report also recommended that an existing sub-regional agency, the Metropolitan
Sewerage District of the County of Milwaukee, agree to conduct a more in-depth
engineering study leading to a master plan.
In April, 1973, the Metropolitan Sewerage District, through the agencies of
its constituent operating bodies, the Sewerage Commission of the City of Milwaukee, and
the Metropolitan Sewerage Commission of the County of Milwaukee, requested SEWRPC to
prepare a Prospectus for a preliminary engineering study for combined sewer overflow
abatement for use in selecting a consultant to conduct the study. The Prospectus was
prepared by SEWRPC with the assistance of a Technical Advisory Committee of outstanding
Federal, State and local specialists in the water pollution field and completed in
July, 1973.(25) ^he Prospectus outlined a proposed engineering study based on the work
contained in the Milwaukee River Watershed report and directed a more in-depth study of
possible alternatives as well as further advances in the field that had occurred since
the publication of the report. The Prospectus was used as an invitation for proposals
and was sent to many of the most prominent consulting engineering firms in the nation.
Twenty-one proposals were received, as evidence of the interest in such a project,
with six consultants invited for oral presentations.
The selection of the six consultants, as well as the recommendation to the
District of the consultant to be retained, was made by the same Technical Advisory
Committee; and the Committee was later transferred to the jurisdiction of the District.
The District then entered into a contract with a consultant for the preliminary
engineering study. The consultant worked under the direct supervision of the staff of
the Metropolitan Sewerage District assisted by the Technical Advisory Committee. The
primary jurisdiction of the Sewerage District is the planning, construction and
operation of the sanitary sewerage facilities for approximately 420 square miles,
comprising the Metropolitan Milwaukee area. The District operates two regional
wastewater treatment plants, 240 miles of interception sewers and 20 miles of improved
stream channels.
Three basic approaches for abatement and control were considered in the
Milwaukee River watershed study: storage of overflows with their subsequent slow
release for conventional treatment at existing wastewater treatment facilities; flow-
through and inflow treatment of overflows at special treatment facilities during
peak flow periods; and complete separation of combined sewer systems. Ten different
storage schemes and three different methods of separation were explored in the total
of fifteen different alternatives considered.^1*/ On the basis of careful economic,
engineering and legal analysis of the fifteen alternatives, it was recommended that a
- 16 -
-------
combination of deep-tunnel conveyance, mined storage and flow-through treatment be
utilized for combined sewer overflow abatement for the entire County combined sewer
service area of 27-square miles.(25) The purpose of the preliminary engineering
study was to verify the feasibility of the recommended plan or the best alternative
thereto, and to determine the needed system configuration with particular emphasis on
the balance to be struck between capacities for conveyance, storage and flow-through
treatment.
The study area not only encompassed the combined sewer service area of the
three rivers and the lakefront but was extended to include all areas that contribute
to the hydraulic flows in the combined and intercepting sewers, and to permit proper
analysis of related wastewater flows and stormwater discharges and related drainage
and flooding problems. Provided was all information necessary for the respective
political entities involved to make the public policy determinations required to
proceed with the pollution abatement program.
Also considered in the study was the potential application of various
combinations of collection, conveyance, storage and treatment facilities for the
additional handling of stormwater runoff from throughout the Milwaukee urbanized
area, should stormwater runoff treatment become necessary in the future to protect
surface water quality, particularly in Lake Michigan and the interconnected Milwaukee
Harbor estuary.
Tunnel systems have been considered in Boston, Chicago, Detroit, Milwaukee
and Washington, D.C.(26' A brief description of the Chicago plan is included as a
case study in an ASCE Program Technical Memorandum.(1°) The Chicago studies were far
more advanced than those for Milwaukee in 1973(2**' and, as noted later in this
Section, a major phase of that master plan is under construction. Because little has
been published on the Milwaukee master plan' » '' more description is given here than
would otherwise be justified.
Local government plans do not evolve overnight and carefully weighed plans
are usually founded on several years of deliberate planning and development, as
epitomized by the initiation of the Milwaukee master plan.
Work on the Milwaukee Combined Sewer Overflow Pollution Abatement Project
commenced late in 1974 and the first of three phases was completed less than a year
later/28) The organizational and scheduling plan for conducting the project, by
tasks/29) would be of interest to planning entities embarking on the development of
such plans. The last phase was completed in mid-1977, and the final report provided
the basic information required to start design for construction.
"Reliable local municipal stormwater drainage facilities cannot be properly
planned, designed, or constructed except as integral parts of an areawide system of
floodwater conveyance and storage facilities centered on major drainageways and
perennial waterways designed so that the hydraulic capacity of each waterway opening
and channel reach abets the common aim of providing for the storage, as well as the
movement, of floodwaters. Not only does the land pattern of the tributary drainage
area affect the required drainage capacity, but the effectiveness of the floodwater
conveyance and storage facilities affects the uses to which land within the tributary
watershed, and particularly within the riverine areas of the watershed, may properly
be put."^) The preceding is given as the guiding principle for water control
facility development for the Milwaukee River watershed, and hence for the subject
master plan. Flood mitigation aspects are discussed later in this report.
- 17 -
-------
It is important to note that in the 1970 inventory,specific model
information needs were not an issue. In f-".ct, most of the hydrologic models deployed
in the master planning were only partially developed, at most, in 1970. This
illustrates the recent phenomenon where information requirements shift in only a very
few years as modeling capabilities are improved.
An Emphasis on Flooding and Drainage
An outstanding example of a comprehensive local government plan for flood
control, erosion control and drainage is the program in Fairfax County, Virginia.
The County lies to the west and southwest of Washington, D.C. The County's population
leaped from 100,000 in 1950 to 560,000 residents in 1974. With this population spread
over an area of 405-square miles the County can be said to be "suburbanizing" because
it is being transformed from a predominantly rural area to a residential area for many
people employed in the District of Columbia, Such explosive growth has strained
various public facilities and services.
Problems associated with stormwater runoff were parried in November, 1971
when the voters of the County passed an $ll-million storm drainage bond referendum.
The objectives of the bond program are to solve existing flooding and drainage problems
to make plans for avoidance of such problems in the future, and to improve the storm
drainage system operation and maintenance capability of the County staff. In the
spring of 1972, a consulting firm was awarded a contract on the planning program,
which will be completed late in 1977. Although the program has been described
/Q o f\ OT j
elsewhere,V5»-'"»-'' some of its salient features deserve mention here.
Briefly stated, the specific goals of the program are to: develop
immediate solutions for existing flooding and channel erosion problems; improve methods
and procedures for developing budget projections for storm drainage capital facilities;
establish a standard baseline from which to measure environmental changes; rewrite
the drainage code and policy by incorporating a balance of on-site detention and
off-site pro rata costs; and make full utilization of stormwater runoff by regarding
it as a resource out of place.(32)
The thrust of the planning program is on flooding and erosion. Because it
commenced prior to the enactment of PL 92-500, water quality was not considered to be
a major issue at that time. However, included among the planning activities were
"environmental baseline studies"'-*3) and environmentally-oriented design solutions,
in all of which water quality was a consideration. Detailed study of water quality
management in the County commenced with the Occoquan basin.
The "environmental baselines"^^) established for each of the 29 watersheds
provided the environmental framework for developing basin plans and the necessary data
for analysis of projected environmental conditions. Rather than use a benefit-cost
approach for evaluating alternative plans, a "problem identification matrix" was
chosen to permit better inclusion of benefits and costs for such usually elusive elements
as visual impact, aesthetics, and creation or destruction of wildlife habitat and aquatic
systems. The matrix approach, which has also been used elsewhere, was a technique for
systematically identifying the general extent and importance of stormwater-related
problems in the County and evaluating the adequacy of proposed solutions to those
problems. Its use required three different elements: the objectives of the drainage
planning program; the relative importance of each objective; and the degree to which
each objective would be met under a given set of circumstances. Although program
objectives can and may be changed or regrouped as experience by the County in the
application of the method is gained, initial objectives, not necessarily in order of
- 18 -
-------
importance, were: freedom of residences from flooding damage; freedom of commercial
and industrial facilities from flooding damage; freedom of public and institutional
facilities and equipment from flooding damage; control of bank and channel erosion;
protection of aquatic ecosystems; protection of wildlife habitats; freedom of parks,
recreation and aesthetic areas from flooding damage; and prevention of traffic
interruptions. Although protection of health and safety is not specifically included
in this listing, it is an overall consideration and would take precedence over other
objectives as part of any flood control plan.
By the end of 1977 the consulting firm will make available a summary report
on the program, encapsulating procedures and findings embodied in the dozens of
program reports for the County.
Some Metropolitan Examples
Chicago Area. There are more than 5,000-railes of combined sewers within a
375-sq. mi. portion of the territory served by the Metropolitan Sanitary District of
Greater Chicago (MSDGC), which includes the 220-sq. mi. of the City of Chicago. The
District serves more than 5,500,000 people in an 860-sq. mi. area, exercising control
over 75-miles of open waterways with respect to drainage, pollution control and
navigation, much of which is monitored for water quality.() Lake Michigan is the
water supply source for the area, which is drained by three rivers away from the lake.
An intergovernmental committee resolved a "Tunnel and Reservoir Plan" for
combined sewer overflow pollution abatement^-*' that incorporates flood control
features. In 1972 the cost of implementing the plan was estimated at 3.3-billion
dollars.^ ' Currently, the first major phase is under construction. "A unique
feature of the plan is its inherent flexibility to take advantage of discoveries made
during its ten year construction period."(36) Principal features of the plan have
been outlined.(1»)
The primary function of the MSDGC is the collection and treatment of the
water-borne wastes from the City of Chicago and its suburbs. Because flooding is
related to such collection and treatment, the District has initiated long-range
planning and local flood plain control. The District has used its authority over the
granting of sewer permits to require local authorities to enact flood plain ordinances
that safeguard flood plain developments against high water and requires on-site
detention of excess runoff from new developments.(") A streamflow model developed
by the District^*' is used in the preparation of watershed master plans. One of the
completed plans, for a 52-square mile watershed, calls for measures consisting of:
five floodwater retarding structures; a multiple-purpose structure (flood protection
and recreation); 261-acres of flood plain preserve; about 1.8-miles of channel
modification; and a land treatment program for more than half of the watershed.'
Participants in this planning, in addition to the MSDGC, were the local Soil and Water
Conservation District, the Cook County Forest Preserve District, the State of Illinois,
and four villages and four park districts,(37) illustrating the multiple jurisdictions
commonly involved in flood control planning. (The MSDGC spearheaded the combined
sewer overflow pollution abatement master plan mentioned above, in cooperation with
the State of Illinois, Cook County and the City of Chicago^5').
Denver Area. A local government special district, the Urban Drainage and
Flood Control District was established in 1969 to solve growing drainage problems in
metropolitan Denver. The District covers about 1,200-square miles of all or part of
six counties. A joint project by the Denver Regional Council of Governments and the
District, partly supported by the HUD Urban Systems Engineering Demonstration Program,
developed procedural methodology for urban drainage and flood control master planning
in the region.(39)
- 19 -
-------
Major drainageways were targeted in the methodology development, defined
as those streams, creeks and gulches that have definable flood plains. Three types
of principal activities to be pursued over two decades are called for under the
Denver regional program: (1), preventive master planning for areas where flood
plain regulation, land use controls and other preventive actions can be utilized;
(2), design master planning for areas where problems already exist and facility
construction is known to be required; and (3), construction for developed areas where
preventive measures are not feasible and where channels, culverts, sewers and other
structures are needed to provide protection.^0' Activities included are: delineation
of flood plains on major drainage channels; regulation of all unoccupied and occupied
100-year flood plains; 100-year protection on occupied flood plains; National Flood
Insurance Program coverage on occupied flood plains where protection is not economical-
the provision for limitation of runoff from new real-estate development by ordinance
and State law; flood storage capacity and spillway protection for dams in the region*
integration of major drainage measures with the regional water resource management
scheme; and a flooding early warning system. (^0) The first master plan for a major
drainage area using the methodology developed was completed in
Motivation for preventive master planning was strong. The cost of protection
for the quarter of the total District flood plain area already occupied, in 1972
dollars, was estimated at $110-million.(39) Use of a preventive approach for the
remaining three- fourths is proving to be comparatively inexpensive.
Urban Studies Program. Corps of Engineers. Congress and the Office of
Management and Budget in 1970 expressed an interest in having the Corps of Engineers
conduct pilot studies of wastewater management for several major metropolitan areas,
to assist metropolitan agencies in their planning efforts. "The initial objective
for each study area was to produce a feasibility report that would identify alternative
means of reaching very high standards of wastewater quality on a regional basis. A
further requirement was that the Corps of Engineers consider land treatment as an
alternative to advanced plant treatment sys terns. "(^2) Initial feasibility studies
were completed in 1971 for five metropolitan areas: Merrimack Basin in New England,
Cleveland-Akron, Chicago, Detroit and San Francisco Bay. Detailed studies for these
five pilot areas were completed in 1973. By the end of 1974 a total of 37 such
planning studies had been authorized, and the number has since increased. Several
studies have been completed.
Provided will be a range of urban water resource plans that are compatible
with comprehensive urban development goals of the region under study, and these plans
will provide "an integrated approach to water resources management".'^3'
Features have been summarized of the Chicago'**2' and San Francisco(^4) area
studies, and the growth of interest in the land disposal of wastewater concept has
been briefly defined.(^^ A significant feature of a study of possibilities for land
disposal in the San Francisco Bay region was the exclusion of places where urban
development might be anticipated by the year 2000.t45' The important point here is
that not only are the effects of local pollution often felt well outside the metropolis
of origin, but with land disposal the abatement of pollution via this form of treatment
could also take place beyond the metropolis.
Corps responsibilities under the Urban Studies Program can include: urban
flood control and flood plain management; municipal and industrial water supply;
wastewater management; bank and channel stabilization; lake, ocean and estuarine
restoration and protection; recreation management and development at Civil Works
projects located in close proximity to urban areas; and regional harbor and waterway
development. Urban runoff quantity and quality is addressed in several of the Urban
- 20 -
-------
Studies. Where associated master plans have been adopted by local governments they
are treated mostly as givens in the Urban Studies.
Two-Direction Communication and Time Horizons
A 1970 analysis of regional land use and transportation planning drew
inferences -from case studies of thirteen metropolitan areas. At that time the
conventional planning concept was the selection of a metropolitan plan from several
complete and integrated development proposals. Proposed as a more viable substitute
was a style in common use today, a cyclic use of alternatives. In the cyclic style,
"alternatives become a means of exploring and understanding the effects and implications
of diverse objectives, assumptions, plans and policies, often in response to a specific
problem. * o . . . flexible and partial alternatives should be prepared and evaluated for
several types of plans at appropriate geographic scales, time horizons and levels of
detail.
Table l characterizes the cyclic employment of alternatives in a
continuing planning process. The reasons for introducing Table 1 here are to critique
the role of local government master plans therein and to remind the reader of the
long time horizons required for such plans.
Seemingly overlooked in Table 1 under "Sub-metropolitan Studies" is, in
addition to the impact of metropolitan plans, the identification of potential problems
at the local level, the input of such signals to the metropolitan phases, and the
acknowledgment of these problems in all metropolitan planning. Consequently, the
allocation of only ten per cent of the total effort in Table 1 to Sub-metropolitan
Studies appears to be too small. On the other hand, the more detailed planning for
local development generally fails to address the broader considerations involved in
metropolitan planning, with the consequence that there is often an appreciable
planning void between the two levels, with no organized or assured effort made to
accommodate these considerations. These comments illustrate the common occurrence
of a discontinuity between metropolitan and local planning efforts. Needed is not
only the interdisciplinary interaction that is becoming more widely practiced, but
also a firmer interinstitutional approach. The Consulting Panel on Water Resources
Planning of the National Water Commission indicated in 1973 that there was a long
way to go, claiming that: "Water-resource planning has rarely been integrated or
coordinated with overall urban planning". W) Section 208 planning has a great
potential for closing this gap.
As noted in Section 1, and as illustrated in Table 1, considerable time is
required for the resolution of plans. The detailed local government master plans
cited in this report have required a minimum of three to five years for their
development. Most will require at least ten additional years for their full
implementation. Those being implemented have faced some degree of delay for
resolution of conflicts, right up to the letting of construction contracts.
- 21 -
-------
TABLE 1
SUMMARY GUIDELINES FOR PLAN COMPONENTS OF CONTINUING PLANNING PROCESS(46)
Plan Type
Metropolitan
Development
Framework
Facility-
Service
System
Flans
Capital
Improvements
Program
Metropolitan
Studies and
Alternatives
Sub-
Metropolitan
Studies
_
Purpose Time
Horizon
general framework for urban 20
development and basis for years
metropolitan public facility
and service system plans and
review of local plans and
programs
metropolitan system plans 20
and basis for designing the years
location and characteristics
of services and facilities
and programming their imple-
mentation (transportation,
water-sewer, open space,
sub-regional centers,
education, health, housing)
coordination of the alloca- 10
tion of fiscal resources at years
all levels of government and
the timing of implementation
of projects, both within and
among facility systems
identification and recon- 20-50
ciliation of development years
objectives, policies and
combinations of systems
and examination of long-run
impact of technological,
social and economic change
on urban development, as
basis for the metropolitan
development framework and
facility-service system
plans
impact of plans on the various
development process and
quality of urban environ-
ment at the community to
project scale
Level of Detail
pattern of ur-
banization,
location and
density of ac-
tivities and
layout of fa-
cility systems
sufficiently
detailed for
preliminary
design
studies
large or
composite pro-
jects as basis
for programming
and budgeting
by operating
agencies
highly
generalized to
very specific
as appropriate
to particular
purpose
detailed to
generalized
Proportion
of Effort
257.
30%
15%
20%
10%
- 22 -
-------
SECTION 3
INCENTIVES FOR COMPREHENSIVE PLANNING
Synopsis
Emphasized in a 1975 national symposium^"' co-sponsored by the ASCE Urban
Water Resources Research Council was that urban water quality control plans should be
comprehensive, making provision for and fully taking into account future water needs
and withdrawals and recognizing and making provision for appropriate future usage of
waterways and adjacent areas as wildlife habitats and for recreation. Recognized was
that water supply, flood control, drainage and pollution control are interrelated in
urban planning and that independent development of drainage and flood control plans
might result in higher costs for achieving water quality objectives.
The thesis of this Section is that rational planning requires conjunctive
consideration of the quantity and quality aspects of urban runoff within a comprehensive.
multiple-use framework. Some of the more obvious arguments are explored, in connection
with viewpoints on land-use controls, disparities in approaches to economic evaluation,
potentials for water reuse, opportunities to reduce extraneous sewer flows, control
of erosion and sedimentation, and disposal of solids from new joint or ad hoc treatment
facilities. Admittedly, difficulties and problems are emphasized, but for the sole
purpose of indicating that the only rational way in which a reasonable degree of
order can replace possible chaos is probably through highly skilled comprehensive
planning and policy coordination at the metropolitan level.
Land-Use Controls
Citizen participation is urged or required in various national laws dealing
with urban runoff management. While the public is perceived by various observers as
demanding more holistic or overall solutions to problems from its elected officials,
there is a countervailing tendency for responsible public groups to single out
specific parts of the puzzle for detailed scrutiny and simultaneously to oversimplify
the interconnections among closely related issues.
Exemplifying this inconsistent view, the national Citizens' Advisory
Committee on Environmental Quality has defined planning as "the conscious selection
c choices n land use'C1**) weeas a more enerall accetable defntion
of policy choices in land use/'1** whereas a more generally acceptable definition
would be that planning is the process of deciding what resources to allocate, over
time and space, to achieve a set of specified objectives. Noted by the Committee is
that many communities begin the planning process with a comprehensive study that
takes into consideration interactions between all present and potential man-made
activities and various natural characteristics, resulting in a document called a
master plan or comprehensive plan for use in guiding community development.
Emphasized by the Committee is the importance of a land-use inventory as "a useful
data collection device with which to obtain a fresh view of an area's relevant
features. It is a device which will focus attention on the important problems and
proposals that are evolving around us often gradually, without shock to our daily
lives but with a tremendous long-range impact on our future.''^*9' Water aspects
seem to be regarded by the Committee as mostly locational issues in an inventory,
as for other natural resources such as flora and fauna and geologic features. By_
stressing the location of water aspects, the transitory nature of water resources
and the intermingling of waters in multiple uses is thereby completely ignored, as
opposed to the water orientation to supplement land-use inventories advocated in
Addendum 1 of this report. Independent of planning issues, it has becm alleged that
- 23 -
-------
management of urban water resources has been viewed for the most part independently
from land-use policy and management rather than as an integral element.($0) These
points are raised in an attempt to indicate some of the challenges confronted by
comprehensive planning and the great importance, of public information, and even
public education, in promoting the larger view.
Attempts by Congress to force a marriage between land and water management
are exemplified by 40 CFR Part 131.11 supporting Section 208, P.L. 92-500. Although
intended by Congress to exploit the police power of local governments over land use,
it is nevertheless expected that in a number of urban areas it may be politically
impossible to utilize land-use controls for pollution abatement.(51) For this and
other reasons, there are strong local biases towards capital-intensive solutions.
One of the other reasons is that nonstructural solutions, which seem to be favored
by EPA for nonpoint source control,(52,53) ten(j to requ£re additional investment in
operation and maintenance of affected urban services, and thus would inflate already
burgeoning local agency operating budgets. Also, the efficacy of most nonstructural
solutions has not been adequately demonstrated with the support of field data. The
intents of Congress in Section 208 are said to be the establishment of a land-use
regulatory mechanism to achieve water-quality objectives over the longer term, and to
provide an incentive and a mechanism for States and local governments to come to grips
with nonpoint source pollution.(54) Local governments face difficulties in acceptance
of land-use regulatory mechanisms in the absence of reliable demonstration of their
effectiveness.
EPA has released in three stages a two-volume manual covering procedures
available for water quality management, with particular emphasis on urban stormwater,
to assist the conduct of State and areawide planning under Section 208 of P.L.
92-500.(55) xhe manual notes that the establishment of an overall wastewater
management plan calls for examination by State and areawide planning agencies of the
wide variety of pollutant sources and corresponding receiving water impacts in terms
of the necessity and feasibility of their control.
Existing and future surface water quantity and quality problems are
inextricably tied to existing and future land use patterns. For example, because
land use significantly influences the volume and rate of runoff from the land surface
to stream systems, land use is a primary determinant of the location and severity of
stormwater inundation problems and of riverine area flooding problems. The general
locations of discharges of wastewater treatment facility effluents to surface water
systems are likely to be influenced by overall land-use patterns, and the nature and
density of land use will affect the quantity and quality of pollutants received by
and passed through such facilities. Similarly, the type and quantity of pollution
from diffuse sources and the points of entry of these pollutants into stream systems
are also largely determined by existing and future land-use patterns. Engineering
and planning studies for amelioration of existing surface water quality and quantity
problems and for the avoidance of future problems should be undertaken within the
framework of an agreed-upon areawide land-use plan. To do otherwise is to ignore the
interdependence of land and water resources.
Because the land-use control issue is so central to the implementation of
any urban runoff control scheme, this writer can find no realistic alternative to
comprehensive planning and policy coordination at the metropolitan level. Because we
have only a handful of areawide, general-purpose metropolitan agencies with operations
authority, for nearly all metropolitan areas the baton must go by default to their
regional planning agencies.
- 24 -
-------
Disparities in Economic Evaluations
Implementation of plans ultimately hinges on political acceptability,
which in turn depends on economic justification. Explored in this subsection are
disparities in measures of economic efficiency employed, and again the regional
planning agency appears to be the most likely agent for reconciliation.
A review(56), of j-^e report^?) from a symposium sponsored by EPA concluded
that, because of the emphasis in the current Federal water pollution control program
on waste control at the source, it appears that environmental decisions will be based
on tests of cost-effectiveness and noted that EPA is inclined in that direction.
Symptomatic is the agency's proposed mandatory requirement for the use of value-
engineering in certain wastewater projects.'")
From a conservationist point of view, there are evaluation hazards in
attempting to assign monetary values to what are usually regarded as non-resources,
such as endangered animals and plants,(59) which is one of the reasons why a cost-
benefit approach finds little popular support as a basis for environmental protection
justification.
Such terms as "public welfare," "social well-being" and "quality of life"
encountered energetic discussion at a recent international urban water workshop where
agreement on terminology appeared to be better for the concept of "the good life" and
fairly definitive on basic human and social needs requisite to survival.v^O)
Because aesthetically-oriented water uses can in many ways be incompatible
with utilitarian uses, conflicts have arisen in multiple-objective water resource
management. There has been a definite trend in recent years towards greater legal
recognition of aesthetic values associated with natural processes, especially with
regard to water, in essentially all areas of the law, including water quality protection,
allocation and planning, and authorization of projects.v"l) An outstanding example is
the environmental impact analysis requirement of the National Environmental Policy Act
of 1969.
Because both facts and value judgments are involved, environmental decision
analysis is both economic and political in character, conditions which are said to
lead to the conclusion that models for environmental planning can be no more definitive
than the political processes to which they provide information on the consequences of
alternative decisions.(*>2) Thus, cost-benefit analysis is truly helpful only in
handling the monetizable assets and liabilities, the economic effects but not the
environmental and social effects, although some of the latter can be assigned subjective
market values.
Federal flood control planning proceeds essentially on a river basin basis
and uses cost-benefit analysis for project justification. Thus, we find in urban
runoff management a tendency towards cost-benefit evaluation for water quantity and
a tendency towards least-cost evaluation for water quality. The traditional approach
for water project evaluation by local governments has been minimum costs, which should
facilitate the water quality side. However, because of the trend towards the blending
gf river basin planning with metropolitan planning, local agencies will increasingly
face the need to reconcile the dichotomy of project evaluation criteria. A problem
that local governments have always had with cost-benefit analysis is that, being a
test of economic efficiency, identification of neither beneficiaries nor bearers of
costs is included, whereas the issues of "who benefits" and "who pays" are central
feasibility questions in local government projects.
- 25 -
-------
Disparities in economic bases for project evaluation would seem to suggest
rather strongly that planning for management of urban runoff quantity and quality should
be conjunctive to avoid absolute confusion.Federal policy with regard to both flood
control and runoff pollution abatement in urban areas is biased towards non-structural
remedial measures, underlining the need for conjunctive planning. In sum, and to
repeat the allegation at the beginning of this subsection, the metropolitan planning
agency appears to be the roost likely agent for the reconciliation of disparities in
approaches to economic evaluation mandated by a tangle of Federal and State lavre.
Water Reuse
Water conservation has often been emphasized as one of the advantages of
wastewater reuse in urban areas.(63,64) whereas reuse of reclaimed wastewater for
domestic purposes is still controversial, recycling for industrial uses is not.
Furthermore, the requirements of P.L. 92-500 for wastewater treatment in the 1980's
are expected to result in such stringent controls on the disposal of effluents into
receiving waters that recycling via groundwater recharge and augmentation of
industrial supplies may become more cost-effective than one-time use.
As local shortages and conflicts over the enlargement of facilities continue
to grow, more serious attention to reuse in more places can be expected; but this
attention is now only beginning to emerge.
There is considerable internal recycling of process water by industry.
Irrigation and cooling water for industry are essentially the exclusive direct reuses
of treated municipal wastewater, a practice that is not widespread.(") Because of
its location in a semi-arid region where the bulk of its water roust be imported, the
City of Los Angeles has been increasingly leaning towards large-scale reclamation for
uses such as aquifer injection as a seawater barrier, surface spreading for water supply
augmentation, industrial cooling, and park irrigation. In 1973 it appeared that the
only instance of a carefully engineered system for water reuse in a major U.S. cit
was in the wastewater portion of the City's legally adopted master plan. Figure
shows the general types of reuse that have been considered for Los Angeles.
Los Angeles practices surface spreading of storm and imported water in an
area to the north, to replenish water pumped as part of the supply from aquifers
normally affected by river waters.(66' The hundreds of recharge basins on Long Island,
New York, are very effective in offsetting the effects of urbanization on the
groundwater supply, according to water budget studies made of the catchments for
three basins.(&/) Sooner or later, stormwater will more commonly be reclaimed as a
source of water supply, probably mostly through recharge of groundwater aquifers.
However, in prior-appropriation doctrine States, legal questions may arise when urban
runoff is captured and put to a beneficial use.(^' There are other formidable
management obstacles to water reuse and conservation,("°' and groundwater sources are
incompletely developed in a number of metropolitan areas.()
We can conclude from the observations above that reuse of wastewater
treatment plant effluents and reclamation of surface runoff is more likely to be via
artificial recharge of groundwater supplies, in general. Where this occurs, the water
supply-through-wastewater treatment cycle will interconnect with the stormwater
disposal cycle (an interconnection shown in Figure 1 of Addendum 1)( particularly for
combined sewer systems, through the long-term storage replenishment of groundwater.
This eventuality can be provided for only through conjunctive planning.
- 26 -
-------
RESIDENTIAL
USE
WASTEWATER
FRESH
WATER
SUPPLY
PRIMARY
TREATMENT
SECONDARY
TREATMENT
PARK
IRRIGATION
AND LAKES
TERTIARY
TREATMENT
GROUNDWATER
REPLENISHMENT
SEAWATER
INTRUSION
BARRIER
OCEAN
OUTFALL
FIGURE 2-USES FOR RECLAIMED WATER
- 27 -
(M)
-------
Considerable capabilities exist for analysis of aquifer performance^!)
and for monitoring groundwater pollution.(72) Also, knowledge on groundwater
pollution has greatly expanded in recent years,(73) partly because of national concern
for control of water pollution and protection of public water supplies,
Extraneous Sewer Flows
Whereas the preceding subsection dealt with beneficial aspects of
groundwater manipulation, the entrance of extraneous flows that include leakage of
groundwater into sewers is a liability.
Plows of stormwater and infiltrated groundwater into wastewater and
combined sewer systems may considerably exceed domestic flows where poor sewer
construction and illicit connections are prevalent.(74) Results from a survey of
communities in the U.S. and Canada(75) indicated that such infiltration and inflow
problems are widespread. More stringent Federal pollution abatement requirements
(P.L. 92-500) and related increased investment in wastewater treatment facilities
have necessitated consideration of ways to diminish these extraneous sewer flows in
order to reduce the capacities of new or enlarged treatment plants. Specifically,
applicants for treatment works grants must demonstrate that each sewer system
discharging into the treatment works is not subject to excessive infiltration and
inflow.(76) xt has been calculated that almost $5.3 billion will have to be spent
to correct inflow and infiltration problems nationally.(77) Diminishment of
extraneous flows would also increase the effective carrying capacity of wastewater
sewers, which in some cases would offset future growth in wastewater loads. Under
certain circumstances, auxiliary capacity could be added by installing flow-smoothing
basins at key locations in a system, as an alternative to the provision of relief
sewers.(78)
Considerations for the conduct of infiltration/inflow evaluations have
outlined.(79,80)
The connection between stormwater and wastewater is obvious for combined
sewer systems, in terms of burdens of infiltrated groundwater that must pass through
wastewater treatment plants. Perhaps less obvious is the fact that the promotion of
consolidation or regionalization of treatment facilities to achieve economies of
scale may prove to have been a poor policy when such facilities are later called upon
to handle attenuated stormwater flows from combined and separate sewer systems from
dispersed storages. Thus, schemes for optimizing the regionalization of wastewater
treatment plants(81»82) should be modified to accommodate wet-weather flow
considerations where their joint treatment at central facilities is projected.
Further, while regionalization of wastewater collection and treatment has been
encouraged by various Federal and State regulations, diseconomies of scale are
encountered with increase in collection system size, and minimum-cost system size is
highly sensitive to the density of the population in the area being served.(83)
At a national Symposium on Regional Solutions to Regional Problems held in
Portland, Oregon, in 1976, concluded was that because the majority of publicly-owned
wastewater treatment works will be committed by the time Section 208 plans are ready
for review and implementation, only two questions will still be open by that time
nonpoint sources of pollution and growth management and who pays for growth.(%<*)
Growth management complexities are compounded by the proliferation of restrictive
zoning, moratoria on sewer or water connections, building bans and dedication
ordinances, adopted as measures to slow growth rates and control the location of
development.(85) Although a variety of motivations have spurred these actions, local
- 28 -
-------
growth control is viewed as a unifying theme.(86' Some of the antigrowth mood can be
laid to the higher costs of unlimited urban sprawl, A study for the U.S. Council on
Environmental Quality explored the cost of urban sprawl.(87} On the basis of
quantified results from a supporting analysis of energy consumption related to
alternative metropolitan futures for the Washington, D.C., area, it was concluded
"that the present concern for land management or growth management can be widened to
total resource management, including consumption of energy".(^8) We can expect
increasing pressure for the integration of energy costs into any metropolitan study
of urban water resources versus land use.
Erosion and Sedimentation
Clearance of land for urban construction exposes it to rapid erosion.'""'
In 1965, Montgomery County, Maryland, adopted the first sediment control program in
the nation.(90) A Statewide Sediment Control Law was adopted by the Maryland
Legislature in 1970. Every project involving grading requires a permit from either
the local government or the State Department of Natural Resources.(91' Each County
and municipality serves as the primary unit of government for administration,
inspection and enforcement for sediment control, the statewide program utilizes the
U.S. Soil Conservation District in each County as technical advisor for erosion and
sediment control, and the program is fully integrated with the State's pollution
control activities.(92) The statewide sediment control program is gradually being
expanded to include storm water management. Preliminary evaluations of "individual
site" stormwater detention storage, such as mandated for erosion and sediment control
in Maryland, suggest that such dispersed storage might aggravate downstream flooding
and that a regional approach to stormwater management might be superior.(93)
Of particular significance here is the fact that erosion and sedimentation
control is one of several water quality issues that cannot be divorced from their
causative driving force, urban runoff. This is one more reason why conjunctive
quantity and quality planning for urban runoff control is needed.
A manual has been developed for use in connection with the master plan for
control of flooding and drainage in Fairfax County. Virginia.(94) Some of the
features in the manual have been summarized.(9-*>9 ^
For the person concerned with erosion and sediment control there are a
number of important references available.(2,97-102)
A Metropolitan Issue
Reference is made to Figure 2 in Addendum 1, which shows the annual
quantities of water passing through a hypothetical metropolitan area of one-million
inhabitants. Total wastewater solids for the million residents would be on the order
of 100-tons/day, dry weight, most of which would be removed in conventional wastewater
treatment. However, sludge is often transported to disposal sites as a slurry. Costs
of transportation and handling can be reduced by slurry volume-reduction through
concentration or "thickening" of sludge solids. Based upon the maximum concentration
that could be hydraulically transported for the hypothetical metropolis producing
100-tons/day, dry weight, the minimum slurry load would be about 400-tons/day.
However, normal practice, which reflects reliability of operation as well as cost-
attractiveness, would result in about a 1,000-tons/day slurry for the hypothetical
metropolis, or in the neighborhood of half the weight of solid wastes handled per
day. As opposed to solid waste disposal, the disposal of wastewater sludge encounters
two monumental liabilities: an extremely large water content .and a very high organic
- 29 -
-------
composition. As a result, the feasible avenues of disposal generally differ from,
and often are more restricted than, those for solid wastes. In sum, handling water
volumes is only part of the problem, for once solids conveyed by water are removed
they must be disposed of somewhere.(69)
A number of major cities, such as New York and Philadelphia, transport
wastewater sludge for disposal outside their metropolitan areas. With many of them
on a desist notice to comply with new environmental regulations, they are very hard
pressed to find acceptable solutions for alternative disposal. Wherever urban runoff
is captured for treatment in combined sewer and storm sewer systems in the future
the sludge disposal problem will be magnified. Even if the added annual amounts
captured were only about one-tenth to one-quarter more, their episodic and
unpredictable accumulation (because of the random nature of storm occurrences) would
tax the scheduling and operation of sludge disposal systems that otherwise handle
fairly uniform loadings throughout the year.
Conclusions
Explored have been some of the more cogent arguments for comprehensive
urban runoff control planning. Additional arguments will surface in succeeding
Sections, particularly in the next Section on conjunctive planning for flow
management. There appear to be very few local government plans extant that have
integrated water quantity management with water quality management. Perhaps the
most we can expect, given the institutional constraints on planning in metropolitan
areas, is an integrated, or comprehensive, or systems approach; but such an approach
is subtle in a documentary sense, and its existence would probably be difficult to
detect from the outside. Optimistically, much more rational conjunctive planning
may be going on than external indications seem to imply.
As an experienced observer of the local water resources operating agency
scene, the writer is impelled to note that any inconsistencies, conflicts and
duplications between and among Federal and State laws, policies, guidelines and
regulations must be and are, reconciled, arbitrated and resolved mostly at and by
the local level of government. For this reason, among several, local governments
are forced to define their own problems. Numerous examples abound of dire
consequences where this responsibility has been abrogated.
- 30 -
-------
SECTION 4
SOME FLOW MANAGEMENT CONSIDERATIONS
Synopsis
This Section is essentially a continuation of the previous Section on
"Incentives for Comprehensive Planning," and has been given separate billing because
the issues involved require a comparatively greater exploration.
Integrated consideration of convenience drainage systems and urban flood
plains has long been recognized as a logical and efficient basis for their
management, tore recently, the logic of integrated or conjunctive consideration of
runoff quantity and quality has become widely apparent. Impeding these ideals is a
fractionalized authority in nearly all metropolitan areas, with responsibilities
for drainage, flood control and their associated water quality aspects splintered
among numerous units of local government. Additionally, in many cases there are
State and Federal agencies exercising specialized authority. Given this institutional
patchwork, about the only avenue for conjunctive consideration of these otherwise
disparate issues is via comprehensive planning at the metropolitan level. While
institutional reform has had strident advocates for decades there have been only rare
indications of wholesale change.
Asserted in this Section is that the guiding principle should be to reduce
the liabilities and to increase the assets of urban runoff. Some of the more
promising ideas for accomplishing this are cited. However, the reader is cautioned
that very few of these concepts have been tested in full-scale system applications,
the research on which many of them are based is fragmented and the findings cannot be
sufficiently generalized for universal application. Here, as in later Sections of
this report, local governments are advised that they are mostly on their own in the
conduct of the research needed to rationalize fiscal commitments for improved resource
management.
Drainage Versus Flood Control
Geohydrologic processes have formed natural drainage channels that convey
storm waters to the seas. Because of an early dependence by commerce on water
transport, most large metropolitan areas originated as urban centers on or near
streams, lakes, estuaries or seacoasts. Intrusion of urban development on natural
flood plains has resulted in damages to occupying structures and sometimes loss of
life. Subterranean systems of conduits facilitate human occupancy by draining
sheet-flow runoff from the land surface. That is, fluvial drainage areas contributing
to urban flood-plain inundation are often of gigantic size compared with individual
underground conduit catchment areas which rarely exceed 10-square miles in extent.
Thus, underground drainage systems are more of a convenience or amenity than a
preserver of public safety. Structural means for mitigating flood-plain inundation
are designed to provide a much higher level of protection than that for storm
drainage systems because of the much greater threat to human life and more apparent
community-wide economic implications.
The function of underground drainage conduits is to remove storm water from
urban surfaces (except combined sewers, which in addition convey wastewater on a
perennial basis). The smallest catchment area (on the order of an acre in size) is
that tributary to a street inlet. Flow in storm and combined sewer systems is
principally by gravity. Like natural drainage basins, smaller sewer branches unite
- 31 -
-------
with larger branches, and so on, until a main sewer is reached. Thus, a main sewer
not only transmits upper reach flow to a receiving watercourse but serves as a
collector of surface runoff all along its route.
Whereas there is a continuum between the subsystems of water supply, water
use and wastewater reclamation (Figure 1, Addendum 1), stormwater has been historically
regarded as purely a negative good or nuisance and its subsystem (Figure 4, Addendum 1)
has seldom been deliberately connected to the other urban water subsystems.
Urban drainage facilities are generally owned, operated and maintained by
local governments, and designed and constructed by local governments and private
land developers. Human life is seldom threatened by the flooding of these facilities.
Because the principal local detrimental effects of flooding are damage to the below-
ground sections of buildings and hindrance of traffic, the consequences of flooding
range from clearly assessable property destruction to annoying inconvenience. It
follows that provision of complete protection from flooding can only rarely be
justified. Instead, facilities are designed which will be overtaxed infrequently.
However, because of the marginal level of protection afforded, storm drainage flooding
damages are also of considerable magnitude, probably equalling those in urban flood
plains. In addition, indirect damages from local drainage flooding are much more
extensive than for stream flooding and generally recur more often, and direct damages
are usually much more widely dispersed throughout a community.
Because urban drainage and flood plains usually are the responsibility of
separate and different jurisdictions, their interconnected behavior is the responsibility
of neither type of organization. Thus we find storm sewers designed for common storm
events without an investigation of their expected performance during rare events, and
flood plain improvements selected on the basis of rare events but with little
investigation of flood damages during more common events. This dichotomy is epitomized
by a reliance on local detention storage for common storm events and the unexplored
discounting of the possible effects of such storage on receiving waters during rare
events.
The safety of people and properties occupying flood-prone areas is the
concern of every level of government, and Federal, State, special districts and other
local government units are involved in the development of an ever-greater arsenal of
remedial or mitigative measures and policies. While storm drainage facilities and land
use regulation are provided by local governments, development of most major stream
flood-management works is undertaken by national agencies.
Flood control, drainage and the quality of receiving waters are all closely
related. (Refer to Section 3, "Incentives for Comprehensive Planning," for a review
of current trends). Frequently overlooked is the fact that precipitation cleanses
the land surface. However, because pollutants together with aesthetically objectionable
materials are washed off the land and transported to receiving waters in runoff, the
result is merely a transfer of land surface pollution to water pollution despite the
benefits accruing to the land. Considering that urbanization increases the rates and
volumes of runoff delivered locally to receiving waters, it is evident that the
conveniences of surface cleansing and land drainage are obtained at the expense of
higher stages and greater pollutant burdens in receiving waters.
Flood Aspects
One out of six acres of urbanized areas is in the 100-year natural flood
plain. There are about 20,000 flood-prone communities and some 16,500 square miles
- 32 -
-------
of urban flood plains in the nation, an area equivalent to the States of Maryland
and Hawaii combined or Massachusetts and New Jersey combined. Mare than one-half of
the nation's flood plains in urbanized areas have been developed, an area of 8,800
square miles or more than the entire State of Massachusetts. These estimates, based
on a 1973 evaluation,'*03' appear in the proceedings of a 1976 seminar on nonstructural
flood plain management measures.(104)
A supporting study(105) indicated that slightly more than half of all
federal investments in structural flood control works for flood damage prevention can
be credited towards flood protection or reduction of flood damage potential in urban
areas. While average annual urban area stream flooding damages are not known, the
U.S. Water Resources Council estimated the rural-urban total at more than $1.7-billion
for 1966, of which the urban portion probably would not be much more than $l-billion.
In contrast, the American Public Works Association estimated 1966 average annual
flooding damages in sewered systems as being in excess of $l-billion.'*^ ' The major
portions of urban areas are served by underground systems of storm drainage. Thus,
somewhere around half of all urban annual flooding damages occur on the large segment
of urban land outside of 100-year flood plains, a matter of considerable importance
to the National Flood Insurance Program.
Conjunctive Planning
There was a time when stream flood plains and sewered drainage systems could
be considered separately, and when urban runoff could be regarded as merely an adjunct
consideration in land-use planning. New public demands and policies require widespread
use of true comprehensive planning, which includes integrated land and water management.
The guiding principle should be to reduce the liabilities and increase the assets of
urban runoff«<.1(^0 The National Flood Insurance Program has been supported by
investigations and analyses concerned mostly with stream flood-plains. It is to be
expected that the scope will greatly enlarge as more communities become eligible for
insurance against storm sewer flood damages. There will be severe difficulties,
alone, in defining what is "improper" drainage or "inadequate" storm sewers.
Determination of criteria for defining "construction and land use practices that will
reduce flooding" will be quite difficult for several reasons: (1), techniques for
hydrologic analysis of rivers and urban streams are at a stage of verification far
beyond that for local drainage, mostly because of a much broader data base for the
former; (2), designs by municipalities, developers and consulting firms, and tools
of analysis developed by universities and other research organizations, have
predominantly emerged as an assortment of unconnected, independent products; (3), the
potential transferability of innovative plans and designs has been drastically impaired
because a range of alternatives was very rarely explored in any given case and a
bias of specialized site conditions was generally suspected; and (4), new requirements
for protecting the quality of water supplies and for abating the pollution of water
bodies add a complex dimension rarely considered in past development projects.
All the basic engineering methods for in-stream flood regulation are
applicable to flood mitigation works within local urban catchments, but generally
on a smaller scale: acceleration of flood flows by canalization through threatened
reaches with consequent reduction in stages; isolation of contiguous land in
threatened reaches from flood flows by means of embankments, such as levees and
flood walls; and the attenuation of flood peaks by means of storage, located either
upstream from threatened reaches or at lateral points fed by diverted flows.
Similarly, possibilities also exist for social-economic-administrative-legal controls
to preclude some degree of damages, such as flood-plain zoning, flood-proofing of
structures, flood insurance and related schemes.
- 33 -
-------
Increased volumes of direct runoff from underground drainage conduits
clearly can aggravate flooding of urban flood plains. On the other hand, increased
receiving-stream stages can cause or induce flooding of underground drainage systems
because of the intimate hydraulic linkage between them. Extensive programs of
integrated flood plain management are particularly crucial for metropolitan areas
that have lttle toorahic relef such as reater Chicao '^°' e
that have little topographic relief, such as greater Chicago. ' There are large
cities that early dedicated most of their flood plains over to parkways, such as
Philadelphia, or developed extensive systems of major drainage channels in step with
urban development, such as Los Angeles. (109, 110) Regardless, new suburbs have more
than occasionally aggravated flooding with subsequent induced damages in many major
cities located on large streams or bodies of water because their low topographical
elevation makes them hydro logically subservient.
It has been suggested that for a growing metropolitan area the thrust of
drainage and flood control solutions should be in two basic directions: preventive
activities in the form of flood plain management together with good planning; and
remedial actions where flood plains have been improperly occupied and developed and
where local drainage problems have not been adequately considered and handled. (mJ
Preventive activities and remedial actions can involve both structural and non-
structural considerations. Structural components include storm sewers, inlets, curbs
and gutters, culverts, and channelization and detention facilities; and non-structural
activities include flood plain management, flood plain warning and flood insurance.
Environmental corridors in southeastern Wisconsin encompass about 18 per
cent of the total area served by the regional planning commission, 486 out of 2,700
square miles. (H2) Asserted is that riverine areas should be viewed as environmental
corridors managed in such a way as to minimize flood problems and simultaneously
satisfy a variety of noneconomic human needs. (H3) In concept, an environmental
corridor is essentially a continuous linear pattern in the landscape consisting of a
composite of natural resource and natural resource-related elements that are important
for maintenance of overall environmental quality. This concept is partly founded on
the demonstrated unsuitability of riverine areas for urban development, and favors,
instead, multiple-purpose floodland management in terms of flood damage mitigation
and environmental corridor protection. From this perspective, it is impossible to
conduct planning separately for streamflow quantity and quality.
Considerations in water-oriented amenities in the urban environment have
been evaluated and tested in an analysis that included case studies of: Harrisburg,
Pennsylvania; Waterloo, Iowa; Houston and San Antonio, Texas; Sacramento, California;
Minneapolis-St. Paul, Minnesota; and the two new towns of Flower Mound, Texas, and
Columbia, Maryland/114'
Detention Storage
Discharges from conventional storm drainage sewer facilities and flood-
plain intrusion by structures both tend to aggravate flooding, and thereby jointly
tend to raise the potential for stream flooding damages. Revising storm sewering
criteria, such as by including much more in-system storage, can be an effective
adjunct in flood plain management. While there is universal agreement that the
planning and development of drainage systems and flood-plain management programs
should be coordinated and integrated, and there is evidence that such efforts are
increasing, prospects for accommodation may tend to diminish rather than improve
because of an increasing concern over water quality considerations in sewered systems
and a contemporary neutrality or indifference on water quality matters by some agencies
- 34 -
-------
dealing predominantly with flood plains. Often overlooked is that much of the
flood-plain flooding problem as well as the land runoff water quality problem could
possibly be more effectively countered on the land feeding urban watercourses.
While the advantages of local detention storage in lieu of the traditional
rapid removal of storm flows have long been recognized,^115) such storage has been
only occasionally employed as part of overall flood mitigation, as in Denver'11"'
and Fairfax County (Section 2). However, specific projects attempting to derive
general storage-requirement parameters have heretofore been necessarily site-specific,
insensitive to storm actualities, limited to one or very few storage options, and
usually have ignored water quality considerations. For example, while an excellent
study in its own right, the research by North American Rockwell^11?) typifies these
limitations.
Findings from a general study of stormwater detention usage have been
reported.(H8-120) Documented in another study were management techniques in the use
of detention storage.(121) While the principle of the use of local detention storage
is often championed in lieu of main channel improvements for metropolitan areas,
sharp differences of opinion have been known to arise, partly or perhaps principally
because implementation of the two methods commonly falls within different
jurisdictions of authority.
There is evidence which indicates that the effects of urbanization on the
magnitudes of streamflows may decrease as rarer floods are approached,'122) to the
extent of a difference that may be undetectable or at least relatively insignificant
at the 100-year level. That is, the effects of land use changes due to urban
development are more pronounced the more common the occurrence. Underground conduit
systems for urban drainage and adjunct detention storage are deliberately designed
to be overloaded rather often (such as once in 5 years, on the average), whereas
waterways flooding protection is usually much greater. Because storm sewers drain to
waterways, they are obviously connected and their performance is related. Mare
attention should be accorded the interrelations between land runoff and streamflow,
over a range of occurrence frequencies, in a conjunctive planning sense. For example,
plans for upstream local detention storage for flow attenuation and/or pollution
control might be found to aggravate unduly streamflow rates and/or pollutant burdens
in a given instance. A study in the Atlanta, Georgia, area indicated that small
detention storage basins are effective in holding runoff from newly developed areas
to their former peaks under natural conditions but become progressively less
effective with increasing watershed size,(123)
A computer-based procedure for application in comprehensive flood plain
information studies' ) is in advanced stages of development. Accommodated are
land use changes and interrelations between conduit systems, detention storages and
streamflows from the standpoint of both quantity and quality.
Through use of simulation, it is feasible to analyze the influence of
alternative urban development on stream hydraulics and to assess expected flood
damages. Elucidation of incremental increases in average annual flood damages are
an effective means for demonstrating flood damage effects of arbitrary or
indiscriminant urban development.C125) Also, there are ways to analyze via
simulation the influence of alternative urban development on storm water quality.^ '
Whereas cost information on local detention storage has not been
comprehensively assembled, a manual is available for estimating construction and
0 & M costs for combined sewer overflow storage and treatment.^12") Also,
- 35 -
-------
some useful data has been assembled on costs of traditional drainage(1^7) an(j
storage capacities.^ °'
Urban lakes have been a subject of concern to the U.S. Geological
Survey(129,130) and the American Geophysical Union. (131)
Other Measures
New national urban planning priorities include: enhancement of urban
environments; conservation of water resources; and reduction in water pollution.
These new priorities require use of stormwater management practices that are much
more comprehensive and complex than generally used at present. Research on desirable
practices has been fragmented and diffused, a severe handicap in making deliberate
efforts to encourage abandonment of historical practice, where urban settlements have
been drained by underground systems of sewers that were intentionally designed to
remove stormwater as rapidly as possible from occupied areas, with little regard for
resultant receiving-stream flooding and ignorance or oversight of lost environmental
enhancement opportunities.
There are at least four basic methods for reducing sewered system flood
peaks and controlling runoff pollution: (1), retardation of flows by induced
friction; (2), control of flow paths and gradients by grading; (3), induced
infiltration of stormwater into the ground; and (4), provision of detention or
retention storage. All of these are more effective the nearer they are to the
sources (individual properties) under the conditions for which they are designed.
However, as noted in subsections above, during rarer runoff events the benefits of
such means may decline substantially by the time flows reach downstream waterways and
other receiving bodies of water.
Alternatives have been conceptualized that would cause a reduction in
quantity or improvement in quality of urban runoff, yet could be implemented by
investment in other than the construction of major new facilities. Among the
possibilities are the following: (132)
Source Control Alternatives
Roof storage
Ponding
Porous pavements
Erosion control
Street cleaning
Deicing methods
Utilization of natural drainage features
Collection System Control Alternatives
Sewer flushing
Inflow/ infiltration
Sewer cleaning
Polymer injection
In-line storage
Remote monitoring and control.
These and other alternatives are discussed in an EPA report. (133) A. "desktop"
procedure is available for the comparison of selected alternative control technologies
for stormwater management.
- 36 -
-------
"Urban hydrology is one field which has failed to attract major Federal
support in spite of strenuous efforts of professional groups such as the Urban Water
Resources Research Council, ASCE. The development of acceptable ways to measure
water quality in urban streams and storm sewers has been even more neglected. The
Federal data-gathering program has concentrated on the larger streams, presumably
because of their traditional use in formulating basin plans. There has been a
tendency to leave urban data-gathering to the municipalities themselves. Even so,
Federal research funds have been available only to a very limited extent."(135)
Thus, the reader is warned that although numerous schemes have been
postulated for controlling the quantity and quality of urban runoff, very few of these
concepts have been tested in full-scale system applications, the research on which
many of them are based is fragmented and the findings cannot be sufficiently
generalized for universal application. By default, local governments are obliged to
assure themselves of the relevance and reliability of most such schemes before making
a commitment to their wholesale use. One way to investigate expected performance is
the employment of hydrological simulation methods founded on local field data, the
subjects of the remainder of this report.
Miscellaneous
A very useful compilation of 800 references has been prepared that emphasizes
nonstructural measures for flood damage abatement, flood insurance and
floodproofing.(136) Citations span 1928-1976.
A study of a 29-acre urban catchment in West Lafayette, Indiana, investigated
the economic and environmental impact of alternative drainage systems for varying land
use and population density.v"7) Drainage alternatives included different levels and
kinds of runoff treatment plants and either open channel or closed conduit drainage,
but consideration of detention storage was not included. However, the techniques for
cost evaluation of single-family versus multiple-dwelling versus commercial development
are potentially useful for applications elsewhere.
Several drainage ordinances, including those in Fairfax County and
metropolitan Chicago, have been, examined in terms of the effectiveness of their
specific provisions and the problems encountered in administering them, and a model
drainage ordinance based on the survey was promulgated.
- 37 -
-------
SECTION 5
UTILIZING SIMULATION
Synopsis
While simulation is undoubtedly an effective means for analyzing
alternative urban runoff control strategies, its most important use can be in the
assessment of expected system performance. Local government officials must be able
to answer pointed voter questions on the exactness of cost estimates versus the
degree of confidence that may be placed on proposed facilities and management
programs to meet specified abatement goals. Performance reliability is ultimately
the fulcrum of political acceptability.
This Section commences with reasons for using simulation and a historical
reckoning of practical reservations against incautious acceptance of runoff models.
Emphasized is that simulation techniques adopted should not exceed the level of
mastery of such tools by the user and that tools should be selected on the basis of
their suitability for solving defined problems.
Performance reliability is defined in terms of four serially connected
principal considerations, which must be completed in the following order to be fully
effective: testing simulation models against local field data; using a long-term
precipitation record as a reference; developing frequency rationales for levels of
protection; and simulating a range of storms and system loadings. The first two of
these four steps are more or less inconveniences that must be accepted to get to the
threshold of reliability required for undertaking the latter two steps where the
true payoff resides.
An emphasis is placed on the testing of simulation models against local
field data because this issue is the crux of credibility as well as reliability.
As things now stand, local governments are substantially on their own for the
acquisition of_field data. The concept of spatial sampling of catchments for field
instrumentation having representative land uses is outlined. Stressed is the
importance of rapid application of data to models, to check data reliability but
principally for expeditious calibration of models, the primary use of such data in
planning.
Control of flooding and water pollution must be based on probabilities of
occurrence because of the randomness of precipitation. Rationales are developed for
using long-term rainfall records as the reference for defining urban runoff quality
and quantity control objectives in probabilistic terms. (Reasons for using actual
records rather than synthetic storms are reviewed at length in Addendum 2).
Reference is made to indications in other Sections of a history of moving
targets for flood management and pollution abatement over the past decade and the
likelihood that policy targets will continue to shift over the next several years.
Underscored in this Section is that the only realistic defense for planning in an
atmosphere of ambiguous policy is to employ procedures the results of which have
an inherent flexibility for conversion to.meet alternative policy goals.
This Section closes with a discussion of what was termed "simulate
performance for alternates" in Figure 1, Section 2. In runoff control planning,
simulation is mostly for future or extrapolated conditions. Reliability in the
employment of calibrated tools for such extrapolations is a function of several
- 38 -
-------
things, but the most important are probably the shrewdness of design of the field
data network and the extent to which its results approach an ideal set of
representative samples.
Most of the written record on simulation has been authored by modelers
and data collectors and reflects their research orientation and interest. Details
over which these specialists fret and expostulate are of only passing interest to
persons responsible for developing comprehensive plans within limited budgets and
rigid schedules. Therefore, an attempt has been made in this Section to dwell only
on fundamental simulation issues in planning and to sidestep niceties that are the
grist of the researcher's mill. Thus, some readers concerned with applications may
not necessarily appreciate the full reasons for some of the trenchant assertiveness
of this Section.
Why Simulation?
The name of the game is system performance. Readers who are disinterested
in the reliability of performance of plans for projected drainage can skip the
remainder of this report.
All but a small fraction of storm sewers here and elsewhere in the world
have been sized by means of wholly empirical methods. Given a lack of evidence of
superior methods, these overly simplistic procedures proved adequate when the primary
purpose of storm sewers was to drain the land and express the accelerated convergence
of surface runoff to receiving waters. Out of sight, out of mind. Once restrainment
or containment of flows and their pollutant burdens become added primary objectives,
traditional procedures of analysis are no longer adequate because of added system
complexities for which conventional tools are unsuited.
Why not use observed discharge variations as a guide? There are several
compelling reasons precluding this possibility: (1), very few urban catchments,
particularly sewered ones, have been gaged; (2), a statistical approach requires a
period of record spanning at least ten years, substantial physical changes commonly
take place on most urban catchments over this long a time, and the mixed statistical
series that results is not interpretable; (3), while such a statistical series would
characterize the existing situation, there would be substantial uncertainty over its
extension to differing future situations; and (4), the clinching reason, in the usual
case where no field measurements have been made, is that it would be necessary to
postpone planning and analysis until new long-term field records were accumulated, an
unacceptable option under contemporary imperatives. An even less acceptable alternative
would be to rely solely on empirical tools and determine prototype system performance
after system changes had been instituted, a procedure that would indicate the overall
errors implicit in the tools used, but would be very expensive experimentation.
Thus, in order to anticipate future system performance under changed
conditions, because these changes can very rarely be simulated by manipulating
prototype systems, recourse must be made to performance simulation by calculation
or analogy. Once having accepted simulation as the means for analyzing future
performance, the central issue remaining is the degree of reliability of system
performance that is desirable, under what conditions and at what costs, the central
theme of this Section.
- 39 -
-------
Background
Hydrology may be defined as the science that is concerned with the waters
of the earth, their occurrence, circulation and distribution, chemical and physical
properties, and their reaction with the environment, including their relationship to
living things. Thus, hydrology embraces the full history of water on the earth.^39)
Because the complex interactions of human activity in concentrated
settlements with air, water and land must be taken into account, urban hydrology is
a distinctive branch of the broad field of hydrology. As opposed to conventional
hydrology, because urban development everywhere has been in continuous states of
expansion and flux, urban hydrology contends with the dimension of dynamic change.
Also, urban water resources management utilizes the social and biological sciences as
well as the physical sciences. Reflecting the lag in the recognition of the fact
that America became an urban nation over half a century ago, the term urban hydrology
gained currency less than two decades ago. Since then, the term has been tacitly
expanded to include all urban water resource matters in, or interfacing with, the
hydrologic cycle, including water quality considerations.
Whereas generic hydrology is termed a geoscience, engineering hydrology
is strictly an applied science. There are no truly fundamental principles exclusive
to engineering hydrology, and this is particularly true of urban hydrology.
Fundamental laws of soil-water movement are borrowed from soil physics and
fundamental equations of water motion are adapted from hydrodynamics. For example,
the principle of conservation of mass is applicable to all kinds of systems, not
merely the water cycle. It is important to recognize that urban hydrology is a
complex art, but an art all the same, because otherwise a more faithful replication
of system performance would be expected from simulation than warranted by the
attainable perfection of the art.
Without question, the advent and rapid evolution of electronic computation
has accelerated development of tools of analysis in urban hydrology as in every other
field. Advances are clearly evident in a recent review.(l^D) perhaps more obvious
advances have been made in water supply and receiving water hydrology simply because
these are essentially suprametropolitan. Another factor is undoubtedly the fact that
they have benefited from advances in traditional hydrology that came about because of
national and State imperatives. Federal legislation on water pollution, flood
insurance and environmental protection, has resulted in a great intensification of
urban water model development over the last few years, with sewered catchment models
perhaps very recently eclipsing urban receiving-water model development.(1^1)
Relatively few runoff-quality field gagings in sewered catchments have been
made, and these have been mostly at outfalls. Source quality has been investigated
principally as a function of street surface pollutants accumulated between rainfalls.
In order to accommodate cause-effect relationships required for modeling, it is
current practice to estimate potential street loadings, separately for individual
parameters, on the basis of the few documented solids-accumulation histories.
Arbitrary allowances are then added to account for off-street contaminant accumulations
expressed as multiples of the potential street loadings. Thus, no direct verification '
of the hypothesized buildup of pollutants and their transport to receiving waters is
presently available. It is reasoned that when "pollutegraphs" generated by models
reasonably approximate field observations for a catchment, that the overall
accumulation and transport hypothesis is validated. As a result, it might be
concluded that model development has already greatly outstripped the data base for
model validation, in the sense of bracketing probable reliability. However, if field
- 40 -
-------
research and model testing continue at anywhere near the level of activity of the
past decade, substantial advances in reliability appear to be an inevitable result.
Or, as summarized in a recent national conference on urban runoff: "When
we consider the length of time that studies have been made of treating sanitary
wastes and the principal industrial wastes, or of hydrology in general, it is
apparent that water quality analysis of urban runoff is a relative newcomer and a
neglected field".(1^2) Much more also has to be learned on the quantitative aspects
of urban runoff.
A recent appraisal of future modeling needs^^O) revealed that, from the
standpoint of runoff quality simulation, progress is stymied principally by a lack
of sufficient field data to evaluate thoroughly the quality segments of various
models, let alone to evolve improvements.
A plan for a national program for acquisition and analysis of field data
from urban sewered catchments was designed by ASCE in 1969. (^3) Despite the best
of intent ions, (144) otliy a token portion of the proposed, critical, field gaging
portion has been implementable. As things now stand, local governments are
substantially on their own for the acquisition of field data and no integrated or
national program exists or is in sight.
The fact that three-fourths of our people live in metropolitan America
continually escapes the attention of framers and implementers of national policy.
In the meanwhile, fractionalized, largely independent, fretful but perhaps impressive
progress is being made in urban hydrology research, and accelerating planning
activities nationwide imply even greater attention in the immediate future.
Acknowledging that significant advances have been made in many aspects of
urban water resources, research on drainage has long been neglected, particularly
with regard to field data acquisition, and information needs are growing faster than
new knowledge is accumulating. This neglect can be traced to the continuing
ambivalence of the Federal government on its role in metropolitan areas, an inheritance
from our colonial past. One view holds that a national program of drainage research
would require the least public expenditure, because the primary objective would be
transferability of findings. However, the prevailing view is that each of the more
than 10,000 local units of government in metropolitan areas involved with stormwater
management should "do their own thing" in such research. Seemingly overlooked is
the fact that national investment in urban facilities has already constituted seven-
tenths of total investments in all water resource facilities, with its share still
growing. While it is of small comfort, the U.S. is far from unique in this regard
because the most urbanized countries still exhibit the rural origins of their
institutions, and urban water resources research around the world commonly has
suffered from inadequate attention and support and from discontinuous and erratic
efforts.
Against this historical perspective is a viewpoint that deserves quoting:
"There does not seem to be a 'perfect* model for analysis of stormwater. The models
are either too complicated, do not allow for distributed inputs and parameters, do
not simulate continuous streamflow, or have not been tested extensively on hydrologic
data There remains much uncertainty in stormwater modeling. There appear
to be enough parametric models available which have been shown to be feasible
conceptualizations of the stormwater runoff process. What is needed now is a
continued and accelerated verification of the existing models and a follow-up
regionalization of the parameters,"(22) ^11 this will take some time. What can be
- 41 -
-------
done in individual metropolitan areas in the meanwhile is to be discussed in the
remainder of this Section.
Performance Reliability
Users of the newer simulation models are confronted with a seeming dilemma.
While the risk of inadequate representation of prototype systems decreases as more
flexible models are used, the complexity of the models necessarily increases at the
same time, as does the difficulty in obtaining an acceptable simulation of expected
performance. A recommended rule is "the premise that the simplest model that will
do the job is usually the best".(132' Further, the more complicated the model the
greater the demands on the user for a mastery of complex hydrodynamic and
biochemical relationships. Indeed, the technological sophistication of some urban
runoff models is as great as in any water resources application. This may be seen
by reviewing the process details of modern models.(22) There is evidence of simulation
failures that have been attributed to model inadequacies where the blame more properly
belonged to improper handling by the user because of insufficient comprehension of
the complex processes involved. Thus, the first cardinal rule is that simulation
techniques adopted should not exceed the level of mastery of such tools by the user.
Putting it in a more pertinent way, the range of simulation possibilities can be
constrained by the level of skills of the user. The finesse and mastery of the
technical personnel involved directly in the simulations can therefore be a major
contributing influence on the reliability of system performance that can be attained.
A second cardinal rule is that tools should be selected on the basis of
their suitability for solving defined problems. This is to say that what is wanted
from simulation should be defined first, and the selection of techniques should
follow, not lead, this decision. Models may be useful as tools in solving problems
but they have no inherent capability for defining them. Further, there being no
single or universal model, a stable of tools will be involved over the course of
comprehensive planning, starting with preliminary screening models through post-
implementation operations models. It should be evident that model needs should be
projected through the total planning epoch, to assure compatibility (at least for
data required) and to avoid extravagant costs and almost irreversible commitments to
particular data processing schemes. Also, the package of tools used should permit
the incorporation of expected improved versions because such changes have accumulated
in the past in the midst of comprehensive plan evolvemenus.
We now turn to other principal considerations affecting the reliability of
projected prototype system performance. Four of these principal elements are
itemized in Figure 3:
. Testing simulation models against local field data. Without some adjustment of
model factors to reflect local peculiarities, whatever they may be, any model is
necessarily under a cloud of suspicion. This item is included in Figure 1,
Section 2.
. Using long-term precipitation record as reference. Water quality and quantity
control objectives must be defined in probabilistic terms, for which this is usually
the only way open. This item is included in Figure 1, Section 2.
. Developing frequency rationales for levels of protection. Flood levels and
receiving water quality goals must be expressed in several ways, such as frequency
of events or degrees of protection, as implied in Figure 1, Section 2,
- 42 -
-------
THOROUGHNESS
OF
ANALYSIS
RELYING
SOLELY ON
EMPIRICAL
ESTIMATES
TESTING
SIMULATION
MODELS
AGAINST LOCAL
FIELD DATA
USING
LONG-TERM
PRECIPITATION
RECORD AS
REFERENCE
DEVELOPING
FREQUENCY
RATIONALES
FOR LEVELS OF
PROTECTION
SIMULATING
A RANGE OF
STORMS AND
SYSTEM
LOADINGS
/
// J*
ADDED
COST
ADDED PERFORMANCE RELIABILITY
FIGURE 3- SUBJECTIVE ESTIMATE OF INCREASED RELIABILITY AND ASSOCIATED COSTS
FOR MORE THOROUGH ANALYSIS
-------
. Simulating a range of storms and system loadings. This, too, is implied in Figure 1
Unless a system is operated (albeit via surrogate simulation) over a reasonable
range of expected conditions, how can we be assured it will perform as we expect
and under what circumstances may it be expected to be "overloaded"?
These elements are discussed at greater length below. However, before
leaving Figure 3 there are important points that should be made. First, it should
be noted that the four components are serially connected. That is, a "range of
loadings" depends on the "frequency rationales" developed, which depends on reference
to a "long-term precipitation record," all of which hinges for collective reliability
on prior "testing of the simulation models" that are used against local field data.
Second, as graphed in Figure 3, the first two elements probably provide a greater
payoff relative to their cost than the other two, although the latter are the really
important ones from the standpoint of comprehensive planning. That is, the first two
are more or less inconveniences that must be accepted to get to the threshold of
reliability required for undertaking the latter two steps. Figure 3 is a purely
subjective estimate by the writer, made in an attempt to hypothesize a nationally
applicable representation. It will be some time before supporting evidence
accumulates. There does not appear to be a suitable way to accommodate in Figure 3
the question of the use of tools with complexities compatible with master plan
objectives, discussed at the beginning of this subsection.
Testing Simulation Models
In the analysis of water distribution systems, the first major objective,
prior to the investigation of requirements for the future, is the attainment of
satisfactory simulation of existing system performance. System simulation is
considered validated during preliminary analysis for design when calculated pressures
are satisfactorily close to observed field-gage readings for given field water source
send-out and storage conditions. "The simulation capability for the existing system
having been verified, analysis of the system or district is pursued for conditions
expected within the next 5, 10, 25 and possibly 50 years."(l^) water distribution
system simulation is singled out to illustrate the fundamental planning principle of
regarding present conditions as prologues or points of reference for the future.
Good information on current actualities is needed to proceed confidently with
extrapolations, where the degree of reliability of projections diminishes as
estimates move into more remote time-frames.
Returning to urban catchments, the "systems" involved become considerably
more complicated. Even if there are parallel similarities between drainage conduits
and water distribution piping, stormwater catchments are characterized by their
large number and variegated land use characteristics, and to this is added the
complex performance interaction of underground conduits with receiving channels,
streams and other water bodies. Thus the calibration of simulation model variables
is much more involved. It is important that the field data used be for catchments
"representative" of prevailing land uses. In principle, identification of candidate
catchments would be on the basis of an inventory of existing catchments in the
metropolitan area, classifying them by predominant land use, size of area, extent of i
storm sewer drainage, and location of outfall relative to receiving waters. Detailed
guidelines for the selection, instrumentation and processing of data of catchments
have been presented recently.(1^6)
Table 2(147) lists the principal physical characteristics of the nine
catchments presently being gaged in the City of Philadelphia to obtain regional
model calibration data. Note that uniformity of land use is deliberately high for
- 44 -
-------
TABLE 2
PHYSICAL CHARACTERISTICS OF REPRESENTATIVE CATCHMENTS BEING
GAGED IN PHILADELPHIA^147'
PREDOMINANT
LAND USE
Residential
Single Houses
Semi-Detached
Row Houses
Commercial
Shopping Center
Industrial
Light Industry
Heavy Industry
CATCHMENT
NAME
Linden
Overbrook
Tustin
Whitaker
Large
Mifflin
Leo Mall
Industrial Park
Erie
UNIFORMITY
OF LAND USE
86
88
100
81
98
90
100
100
100
SEWERAGE
Separate
Comb ined
Separate
Combined
Comb ined
Comb ined
Separate
Separate
Comb ined
SIZE
(Acres)
145
32
82
120
74
80
9
43
100
IMPERVIOUSNESS
(Per Cent)
43
31
63
68
84
98
100
75
76
DWELLING UNITS
(No. /Acre)
1.4
3.4
10.9
8.4
16.0
24.4
-
-
I
-p-
-------
the catchments selected, to ensure a representativeness of land use types for later
application to the Philadelphia metropolitan area of the models to be calibrated.
There are about 250 separate storm and combined sewer catchments within the City of
Philadelphia alone.
For smaller metropolitan areas it is more feasible to make comparatively
more extensive gagings simply because of the few catchments involved. An example
is the portion of metropolitan Rochester, New York, served by combined sewer systems.
There, all thirteen combined sewer outfalls involved have been instrumented for
automatic flow-sensing and water quality sampling.(14°) The land-use features of
the catchments are listed elsewhere.(132) j^e catchments range in size between 230
and 2,600 acres, residential occupancy ranges between 37% and 94% and imperviousness
ranges between 35% and 80%. Ten of the thirteen combined sewer outfalls discharge
overflows to the Genesee River, which is also monitored for water quality. Considerable
modeling work has been undertaken.(132,148)
A number of considerations are involved in the deployment of gaging
stations,(146) such as in the Philadelphia case above, which may be termed a spatial
network design. The purpose of these stations is to serve planning and design
functions. (The use of such networks for real-time operation is discussed in
Section 7). Specifically, short-term records are necessary to calibrate models so
that long-term performance simulations can be made. That is, rainfall variations are
sampled over a season or two on catchments that are samples of predominant land uses
(The issue of storm variations over time in discussed in the next subsection).
The goal of spatial network design is to attempt an optimization of the
best set of "representative catchments sariples" that can be realistically funded for
gaging; and the maximum number of station points where measurements of flow-quality
can be made, for a given total budget, is then affected by how fast a point can be
abandoned and measurements at some other point can be initiated with the same
equipment. Speed of abandonment is a function of how rapidly the data are used to
calibrate a model or models, and the deliberate rush to calibrate may cause an extra
non-data cost for analysis which has to be taken into account in weighing budget
schedules. Because such field work can often be funded only on a year-by-year basis
the staging of investment is therefore affected by the staging strategy employed, in*
addition .to considerations of total project and yearly project budgets. That is,
the staging over the project should maximize results from year to year, with an
early payoff and sufficient payoff thereafter to discourage premature abandonment by
officials with fiscal veto power. All this is to say that considerable judgment,
wisdom and foresight are required. It is obvious that spatial network design is
presently more art than science. On top of all this is the restriction that some
catchments cannot be gaged at all (e.g., very complex backwater flooding of a site
at a wide flow range, such as by tides), and some can be gaged only at inordinate
expense (e.g., because of difficulties of access). ASCE had proposed a 3-year national
gaging program in 1969,(143) ancj it ^s therefore expected that any metropolis could get
nine-tenths of its work done within 3 years.
Long-Term Precipitation Record as a Reference
Control of flooding and water pollution must be based on probabilities of
occurrence because of the randomness of precipitation. Reasons for using actual
records rather than synthetic storms are reviewed at length in Addendum 2.
The primary long-term rainfall data base for sewered catchments in any
metropolis is the (usually) single U.S. Weather Service first-order station five-minute
- 46 -
-------
interval data, commonly spanning fifty years or more. It is necessary to assume that
short-term rainfall samples from field-gaging programs lasting a season or two
comprise a representative sample of the long-term record. To repeat, the single
rainfall station record is the primary (and sometimes sole) source of long-term data
available. So far, there has been no successful simulation of 5-minute single-point
data via stochastic methods, so there is no other way. On top of the assumption that
our precipitation field gaging is for a representative sample, we must further assume
that the long-term single-point record would be more or less replicated over a
similar period of time at any other point in the same metropolis. The latter is a
generous assumption because we have proof of areal persistence only for pieces of
storms: i.e., maximum intensities of a specified duration. Figure 4 shows examples
of the areal persistence of one-hour maximum average rainfalls of particular
frequencies, after Yarnell'l^'' (only six of the total of 56 charts he prepared).
The same persistence is implicit in later mappings of the U.S.W.S., including
Technical Report No. 40(1^0) aiuj ^n recent updates. In addition, the Corps of
Engineers studied the Yarnell maps and found that the depth for any duration was an
approximate function of the same frequency one-hour duration depth for all national
station data, regardless of recurrence interval, Figure 5.(151) it should be
carefully noted that the "isohyetals," such as in Figure 4, connect equal values of
precipitation for a given duration and frequency; and that these values are for
different storms at different years from one location to another. (Details on such
data are given elsewhere'"^). Thus, there is an implication of long-term persistency
from one point to another over long distances, and hence over relatively short
distances, tenuously based on the documented strong persistence of separate pieces
of storms regarded independently.
Theory of catchment raingage network design is relevant only for initial
placement; and optimum or critical placement can be assured only by modeling data as
it accumulates to see if reliability of the objective function (runoff-quality) is
adversely affected by the initial placement. The underlying assumption in using a
single point (Weather Service gage) long-term record for all points in a metropolis
is that variations are completely random and therefore over a sufficiently long
period (whether 10-years, 50-years, or more, nobody knows), essentially the same
objective function results would be obtained at any point in the metropolis as at
the single reference point. However, we have no adequate theory, based on objective
function response and not input response, on the statistical relation between more
than one point of input data and metropolitan-wide points. Based on our fundamental
assumption, it follows that if more than one raingage is required to characterize a
catchment we will not be able to extend the record as above-described realistically.
This leads us to the inescapable conclusion that all catchments with a flow-sampler
installation should need only a single raingage. At first, this would seetn to
restrict consideration to catchments no larger than perhaps 500 acres, but it does
not. Rather, what it tells us is that subcatchments should be no larger than some
magic limit (say, 300 acres), that each subcatchment of about that size should have
its own raingage, and that each subcatchment should have its own runoff-quality
station at its outlet, so that when long-term records are simulated the simulation
will take account of the separate subcatchments' calibration peculiarities. This is
still stretching our basic assumption, in the application of long-term single-point
rain data, because we will be assuming perfect joint correlation of objective
function results for the multi-point catchment (i.e., between each subcatchment) as
we have for between the reference Weather Service record and some other single point.
This, however, is the best we can possibly do I
An important conclusion that results from the above reasoning is that,
except where a catchment does not change in character over a long period of time
- 47 -
-------
On* Hour Rainfall, in Inch**, to
B* CHp*cl*d One* in 2 Yian
Ont-Hour Rainfall, in Inchti. lo
84 Cxpccte! One* in 10 Yon
On* Hoar Rainfall, m Inch**, lo
I* CipKttd One* m SOYtart
On*4tour Rtinfall, M Inch**. ID
t Eip*cM One* 'n 5 Ytara
One-Hour Rainfall, In Inert**, to
( E*p*ct*d One* in J5 r«ar»
FIGURE 4- SAMPLE OF YARNELL CHARTS
On* Hour Rainfall, m Inchn, to
* Exp*ct*d One* in 100 Y*ar>
(149)
< g
fr»qu*ncy
2-Yur
S-YMr
10-YMT
25-Y**r
50-YMf
lOOYMt
"0123,56789 10 12
R*inlall Intindty for Durations lndicat*d by Param*t«n, in Inc-it* p*r Hour
FIGURES- RELATION FOR VARIOUS DURATIONS
(151)
- 48 -
-------
(unchanged land use and hydrology over 10 years or more, but how can this be
realistically anticipated in advance?) the justifiable duration of data collection
for planning purposes is limited to that amount necessary to calibrate the catchment
response with whatever model or models will be employed to extend the record of that
catchment. Dallying could be justified on the basis that the "right" model or models
is either not known with assurance at the outset, or that improvements in all models
will be coming, sometime, but such temporizing must be weighed against the cost of
acquiring data only as a "holding action" and the more important fact that the
equipment tied down by equivocation cannot be moved to calibrate another catchment,
and the extent of "catchment sampling" (see the preceding subsection) will
accordingly be shorted. Of course, there is merit in the perpetuation of a few
stations as a means for sharpening reliability as planning proceeds. But this is
virtually identical to an extra-cost option on an insurance policy and should be so
regarded. The only major exceptions are for long-term monitoring and where automatic
control in planned operations is contemplated, discussed in Section 7.
Developing Frequency Rationales for Levels of Protection
Emphasized throughout this Section, and in Addendum 2, is the importance
of putting results of simulation on a probabilistic footing as a means for maximizing
the reliability of performance of facilities and programs planned for the future.
Noted several times in earlier Sections was the history of moving targets for flood
management and pollution abatement over the past decade, and the likelihood that
policy targets will continue to shift over the next several years. The only realistic
defense for planning in an atmosphere of ambiguous policy is to employ procedures
the results of which have an inherent flexibility for conversion to meet alternative
jjolicy goals,
In Section 2 we noted that the policy goal of the San Francisco master plan
for combined sewer overflow pollution abatement was to limit the average number of
overflows per year. We also rioted that because the City had founded its estimates of
attainable system performance on a long-term precipitation record (62-years), it
would be relatively simple to convert the original simulation results summaries into
their equivalent average annual or seasonal or serial expected occurrences, such as
in terms of overflow volumes or pollutant loads or seasonal number of overflows or
whatever. Such a conversion would be facilitated by the existence of field data that
has accumulated over the past six years in San Francisco in support of a total system
automatic control development program (Section 7). Although the first phase of the
master plan is now committed in terms of physical deployment and capacities of a
about half of the projected facilities, the ultimate incorporation of
automatic control and interbasin dispatching features will provide sufficient
flexibility in operating the total system so that reasonable shifts in policy targets
could be accommodated. Had the above elements of flexibility not been included, the
City would be rather helpless if the policy target was shifted,,. The cost of
add-on revisions to new facilities could quickly escalate abatement costs to
completely unrealistic levels,
While flexibility of analytical tools is the issue in this Section, it is
important to note that total project flexibility is much more important.
Simulating a Range of Storms and System Loadings
The preceding three steps are prologues to this step, called "simulate
performance for alternate;;" in Figure 1. To reiterate, it is necessary to acquire
short-term field dato records over one or two seasons to calibrate models for the
- 49 -
-------
subject simulation. That is, rainfall variations need to be sampled over a season
or two on catchments that are samples of predominant land uses. Because urban
catchments, whether fully storm sewered or only partially storm sewered, often change
in land use and drainage system characteristics from year to year, the river basin
analysis approach of deriving hydrologic indicators from a long series of field
observations (such as 20 years or more) is simply not feasible. If reliability of
the results of analysis is a genuine issue, the only recourse is a time-series
synthesis of runoff-quality events via simulation using hydrologic models.
Now for the moment of truth. What are we going to do with the resulting
catchment calibrations? We are going to extrapolate them. We are going to assume
that we have a fix on the characteristics of a "representative sample" of catchments
and that the model hydrologic constants, exponents and other factors that were found
to work on a given class-size-land use catchment will apply to all catchments of that
given character and (sending up a prayer) that we can extrapolate (and perhaps
interpolate) these findings for other class-size-land-use combinations where we did
not have enough money to sample for their characterization. Obviously, the level
of reliability of the tools we have calibrated is not merely a function of the
goodness of fit of hydrographs and pollutographs in the calibrations or in inherent
instrument or recording errors, but also on the "representativeness" of the gaging
network; and this latter, very important point must be brought home to the so-called
dec is ion-makers. That is, reliability of the employment of calibrated tools is a
function of several things, but the most important are probably the shrewdness of
design of the data network and the extent to which its results approach an ideal set
of representative samples. Let the modelers lament observed-calculated differences
and let the field engineers sermonize on instrument error, but always remember that the
degree of data network adequacy is a much more important measure of the reliability
of planning or design conclusions based on simulations founded on data from the
network. At this point we must qualify the preceding assessment of overall reliability
of results, downward, but by a necessarily qualitative amount, to allow for the
universal use of a single-point rainfall data record within the metropolis, described
earlier in this Section, as the representative historical rainfall for the entire
metropolitan area. There are a few cities and metropolitan areas that have raingage
networks that have been in operation for five years or longer. A few of these have
been able to derive empirical relationships between rainfalls in various sectors of
their jurisdictions on the basis of selected storms. While such indicators can
certainly be useful in operations, and perhaps also in design, planning applications
are necessarily referenced to the long-term record of single-point rainfall data.
- 50 -
-------
SECTION 6
URBAN RUNOFF M3DELS
Contents
Reiterated in this Section is the importance of calibrating any urban
runoff model against local field data. The national dearth of field data
(particularly for sewered catchments) available for validating models is cited
as one of the principal liabilities in the use of such models. However, there are
a number of potential advantages in the use of models that can be exploited when
certain liabilities, such as a national scarcity of validating field data, can be
overcome.
Models are characterized in this Section in terms of their applications.
Specific models are identified, with an emphasis on sewered system planning
applications, although associable analysis/design applications are duly noted. Of
the several attempts to compare the performance of various models, the most useful
are those defining extents of application flexibility. Receiving water modeling
is handled separately because of the tendency towards customized adaptation of
available models for that purpose. Examples are given of some regional receiving
water simulation studies.
Role of Simulation in Planning
A special session at an annual meeting of the American Geophysical
attempted to define appropriate rationales and incentives for the more extensive use
of urban runoff mathematical models, for planning, analysis/design and operations.
Among the advantages cited for the use of such models for planning were that: tests
can be made of alternative future levels of development and their impact on facilities
needed in the future; several models well-suited to master planning are in the public
domain and are regularly upgraded and made readily available by the Federal agencies
that supported their development; when detailed models are used in advanced stages
of planning the user is able to understand better the physical performance of a
system; the interrelation between land-use projections and planned raitigative
programs and their costs can be made more apparent; revisiting plan assumptions to
update projects can be done with consistency and relative ease; joint consideration
of quantity and quality of runoff in sewered catchments and in streams can be
accommodated; hydrologic-hydraulic effects of future urbanization can be explored;
and deficiencies in existing facilities and prevailing management programs can be
identified.
Of the liabilities, outstanding is the dearth of field data on rainfall-
runoff-quality, particularly for sewered catchments, for development of more
acceptable measures of reliability of all types of models. A recent workshop
conducted by the ASCE Urban Water Resources Research Council resolved guidelines for
the acquisition of such data by local governments.^^"' The spectrum of investigative
stages utilizing field data for the sewered areas and receiving waters of a
metropolitan area include the following:
. Identification and evaluation of quantity and quality problems.
. Exploration of alternatives for pollution and flooding abatement.
. Analysis of the most attractive alternatives.
. Preliminary design of adopted alternatives.
. Detailed design of adopted alternatives and their implementation.
. Post-implementation operation via a range of possibilities extending from simple
monitoring to automatic control.
- 51 -
-------
Consistent with the comprehensive planning scope outlined in Figure 1 (Section 2)
this Section will concentrate on the use of models up to preliminary design.
However, because models used in the implementation and operation stages that evolve
from a comprehensive plan rely on extensions of the same data base, some mention
must be made of their characteristics.
Before embarking on a discussion of model capabilities, it is necessary to
establish certain caveats on the state of the art.
Some Reservations
Because complex processes, such as in the hydrological response of a
sewered catchment to a precipitation occurrence, can never be fully replicated in a
computation due to incomplete technical understanding of the processes and the
infeasibility of detailing the literally myriad pieces involved, resort is made to
simulation of response of a conceptually equivalent system. The simulation package
is commonly called a "model". Reality dictates that a model should be selected on
the bases of the type of application involved, how it is to be used, how much can be
invested in its use, how often it would be used, what levels of precision are
required or desired, what kinds of outputs are wanted, how much time can be spent to
get the model to work, and how much can be committed to verify and calibrate the
model. Calibration is the process of varying model parameters to minimize the
difference between observed and simulated records.
We have been reminded that until each internal module of an overall
catchment model can be independently verified, the model remains strictly a hypothesl
with respect to its internal locations and transformations.(154) Because O£ t^e vg s
limited amount and kind of field data available, just about all sewer applications
model validation has been for total catchment response, at outfalls. That is, under
contemporary conditions a distributed system model deteriorates into a lumped system
model for all practical purposes. It should therefore be evident that validation
using transferred data by the model's developer is not nearly enough. Credibility
requires at least token calibration using some local rainfall-runoff-quality data
Unfortunately, the acquisition of such data is commonly regarded as the exclusive
problem of local governments, and too many planning and analysis exercises have
proceeded without benefit of local field data, using one model or another.
Calibration and validation is further confused by the fact that much more
field data are available for partially sewered catchments, where flow is measured in
receiving watercourses, than for totally sewered catchments. (That water quality
samples have been taken for only a fraction of these gaging sites does not help).
Adding streamflow hydraulics to sewer hydraulics hardly simplifies the lumped system
dilemma alluded to above, yet much of the data used to verify various models has been
from such mixed catchments., This should add additional incentive for calibration with
local data. "General EPA guidance to date (1975) has tended to minimize basic samplin
and analysis of urban runoff. Collection of new stormwater data is severely needed to
define the magnitude of its impact on water quality."(1-55)
Concluded in a comprehensive Canadian study was that sufficient information
is not available on relationships between street surface contaminants, their
pollutional characteristics, and the manner in which they are transported during storm
runoff periods. Also concluded was that basically only one type of model exists for
analysis of urban runoff quality, and that the accuracy of the water quality
computations using models extant has not been sufficiently established to be used
with confidence for prediction purposes, in particular the formulation relating water
quality with land use.'^6)
- 52 -
-------
A fundamental objective in water pollution abatement is public health
protection, yet little information is available on the level of pathogenic
microorganisms in stormwater runoff in urban areas.(^7)
Water resources are not an exclusive major consideration in environmental
protection, and interest is growing in comprehensive or integrated planning,
management and control of all types of pollution sources. For example, a model for
metropolitan application has been developed which considers environmental pollution
as a set of interrelated problems, using submodels for land use, residuals, and
dispersion-disposal of residuals.^"' While analogous completely comprehensive
modeling of urban water resources processes has been found to be technically
feasible,' 159) conciusions reached in a study of future needs in urban water models
included the opinion that emphasis in urban water modeling in the immediate future
should remain on the simulation of discrete processes (individual subsystems such as
storm runoff) before attempting simulation or optimization of connected subsysterns.
This viewpoint was predicated on the likelihood that the functional responsibilities
of local government water agencies will remain as splintered as they typically are
now. "It is likely that only where wastewater reclamation, including municipal
reuse, becomes a viable alternative will a single agency require comprehensive urban
water modeling approaches."(140)
Categories of Model Applications
Mathematical models used for the simulation of urban rainfall-runoff or
rainfall-runoff-quality can be divided into three different application categories:
planning, analysis/design and operations. Some particular models have been employed
in both planning and analysis/design, and a few models have been applied in analysis/
design and operations applications, making it difficult to allocate them to a single
category. Additionally, the reader is cautioned that on no account should most of
the models to be mentioned be regarded as typical tools. Rather, common practice
still favors rudimentary techniques and only a modest number of offices currently
employ the more advanced tools routinely.
Planning applications are at a macro-scale, such as in metropolitan or
city-wide master plans. Model requirements for planning are less rigorous and require
and permit less detail than for analysis/design because investigation of a range of
broad alternatives is at issue. What are sought for planning tools are general
parameters or indicators for large-scale evaluation of various alternative schemes.
Hence, the degree of model detail required in jurisdictional planning is much less
than in analysis/design. However, a certain amount of conjunctive intensive detailed
modeling nevertheless is needed to establish parameters and indicators and to provide
an underlying understanding of the governing hydrological processes, so that
simplified expedients are not inadvertently misused.
Total lengths of underground drainage conduits dwarf those of open
watercourses in major cities. For example, total lengths in the 97-square miles of
the City of Milwaukee as of the beginning of 1970 were as follows:(160)
Lakefront length - 8-miles
River lengths - 37-miles
Combined sewers - 550-miles
Storm sewers - 820-miles.
(In addition, there were 685-miles of wastewater sewers). These drainage conduits
are distributed over 465 drainage catchments having a maximum size of 1,820-acres and
- 53 -
-------
a median size of 25-acres.(I"-*-) When dealing with so many components the model used
must be as simple and as flexible as possible. That is, data processing for planning
applications becomes a much more important practical consideration than the level of
sophistication of hydrological process modeling.
Models used for analysis/design applications are more sophisticated and
thus are more detailed tools. They are used for analyzing individual catchments and
subcatchments where the simulation of detailed performance of discrete elements within
a subcatchment must be achieved. Whereas hourly rainfall data is an appropriate input
for planning models and for simulating flows in larger urban streams, 5-minute
interval rainfall data (the shortest duration reported by the U.S. Weather Service)
is the appropriate input for simulating flows in sewers and small urban streams for
design applications. That is, the level of sophistication of hydrological process
modeling for design becomes a much more important practical consideration than data
processing, just the opposite of the emphasis imposed by planning requirements.
Models used for operations applications are likely to be more use-specific
because of wide diversities in management practices, operating problems and
individual service-system configurations. However, the most potentially transferable
technology will be for automatic operational control of total community runoff a
capability that is currently receiving intensive development (Section 7). The
mathematical models required feature control algorithms that have to be painstakingly
derived from numerous indicator applications of both detailed analysis/design models
(for generalization of the performance of individual process components by simulation)
and planning models (for generalization of community-wide system performance by
simulation). Here also, analysis/design models are used as tactical tools and
planning models are used as tools of strategy.
Models for Planning Applications
STORM (Storage, Treatment and Overflow Model). Designed specifically for
urban runoff and quality evaluation for total jurisdiction master planning, this
computer model is eminently suited for that purpose and currently enjoys, in one
version or another, the most extensive use of any urban drainage planning model. Not
only is it non-proprietary but the computer program, model documentation^162) and a
users' manual(163) are all readily available. The original version of the model was
employed in part of the development of the Department of Public Works, City and Countv
of San Francisco, master plan for combined sewer overflow abatement, described in
Section 2.
A simplified logic diagram for STORM is presented in Figure 6.(28) Note
that this model focuses on structural means for flow and pollutant containment
(storage and treatment). It is designed for use with many years of continuous hourly
precipitation records (but can be used for individual storm events). For example,
a 62-year U.S. Weather Service hourly precipitation record was used for the San
Francisco master plan. Essentially, the model employs an accounting scheme that, for
each storm event, allocates runoff volumes to storage and treatment, noting volumes
exceeding storage or treatment capacities (overflows, in the case of combined sewer
systems) as these capacities are exercised from one event to the next. Water quality
is handled as a function of hourly runoff rates, with generated quantities of
constituents allocated to storage, treatment and non-capture as for runoff volumes.
Statistics are generated for each event and collectively for all events processed
including average annual values.
- 54 -
-------
INITIALIZE SYSTEM
READ PRECIPITATION DEPTH)
CALCULATE EFFECTIVE PRECIPITATION
CALCULATE RUNOFF FLOW a QUALITY
FLOW > TREATMENT
RATE?
DECREASE STORAGE
BY TREATMENT RATE
MINUS FLOW RATE
YES
VOLUME = (FLOW - TREATMENT RATE) A TIMEJ
VOLUME < RESERVE
STORAGE CAPACITY
ADD VOLUME TO
STORAGE
NO
OVERFLOW = VOLUME MINUS
RESERVE STORAGE CAPACITY
STORAGE - MAXIMUM
STORAGE CAPACITY
FIGURE 6- "STORM" SIMPLIFIED LOGIC DIAGRAM*"*
-------
STORM accommodates non-urban catchments and snowpack accumulation and
snowmelt, and land surface erosion for urban and non-urban areas can be computed in
addition to basic water quality parameters.
Until recently, hydraulic simulation or flow routing was not incorporated
in STORM. The latest Corps of Engineers' version includes a capability for routing
to the outlet of each sub-basin through the use of triangular unit hydrographs based
on the Soil Conservation Service procedure.(1°3) Planned for future versions are such
added capabilities as an expanded procedure for sizing detention reservoirs. Hence we
find a tendency towards enlargement of a model originally designed for broad scale
planning to include preliminary analysis/design capabilities. Simultaneously,
detailed models for analysis/design are being simplified to make them more amenable
for planning applications, and both types of models are being interfaced to
facilitate feedback of information from one to the other. All this is making it more
hazardous than ever to categorize a model in terms of its major applicability.
A modification of the STORM model has been linked with the receiving water
module of the SWMM analysis/design model (noted below) for continuous simulation of
receiving water quality,(164,165;
As part of a nationwide assessment of stormwater pollution control costs
a "desktop" procedure was developed, by streamlining the STORM model, that can be
used to estimate the quantity and quality of urban runoff in combined sewer and
storm sewer areas and unsewered portions of a jurisdiction. Combinations of storage
and treatment for pollution abatement and their costs can be estimated taking
advantage of generalized results from the nationwide assessment.(166)
A methodology has been suggested for screening urban development alternatives
and preliminary design of storage and treatment capacities for runoff control on the
basis of water quality impacts utilizing STORM.(167)
Other Planning Models. A simplified stormwater management planning model
has been developed that is an inexpensive, flexible tool for planning and preliminary
sizing of stormwater facilities.(^32,168) Time and probability considerations are
incorporated in the model. Joint usage of a complex model and a simplified planning
model, such as this one, is said to be not only compatible but also complementary.
A very simple methodology has been advanced for preliminary screening of
stormwater pollution abatement alternatives.(169,170) while the method was conceived
for combined sewer system applications, it could as easily be applied to stormwater
systems. Developed for use at the national or State decision-making level for early
identification of poor candidates for abatement project funding, it might as readily
be applied for rough, early assessment of the maximum pollutional impacts of storms
at the metropolitan level.
Reported verifications of the simple process planning models, including
STORM, have been limited, although hearsay indicates that the number of verifications
is growing. Because suggested magnitudes of model coefficients are based on the
sparse amount of field data available nationally, it is very important that local
rainfall-runoff-quality field data be used to calibrate such models for the sake of
enhanced reliability of results. Too many planning exercises have proceeded without
benefit of local field data, using one model or another.
The original MIT Catchment Model was predominantly an analysis/design type.
However, a modified version has been employed for estimating sewered catchment and
stream runoff in the Fairfax County master plan development,(30,31) section 2.
- 56 -
-------
Synthetic unit hydrographs , based on regional parameter general izat ions,
were employed in planning the drainage for the large "Woodlands" developmental^)
near Houston. However, unit hydrograph applications have bean mostly in analysis/
design. For example, a unit hydrograph is in general use for designing drainage and
flood protection works in metropolitan Denver. (173, 174)
Models for Analysis/Design Applications
There are a number of this type, including the two noted immediately above.
Although it is essentially an analysis/design type when applied to
smaller-scale catchments, the Hydrocomp Model can and has been used for broad- scale
receiving-water planning applications, described later. The Hydrocomp Model has also
been employed in several metropolitan sector studies. (175-178)
There have been some applications of versions of the British Road Research
Laboratory Model. (1'9-lal) Available to the public is the computer program for a
very simple streamflow and drainage system model used in the Chicago area(l°^) for
preliminary project planning and design.
The USGS has modified the MIT and ILLUDAS models for continuous simulation,
detention-storage accommodation and water quality simulation.
As far as drainage system simulation is concerned, the EPA
(Stormwater Management Model) and its variants have enjoyed more applications and been
subjected to more verification than any other model. For example, a version of SWMM
has been used to supplement and compare results obtained by the Chicago Department of
Public Works from its own models(185,186) as part Of tne raaster plan for combined
sewer overflow abatement in metropolitan Chicago, Section 2. Another version, used
in connection with the San Francisco master plan, (187*188) has the most versatile
transport module of those extant. This version is now included as an option in the
computer program for SWMM. The latter is continually updated. (189)
For the Milwaukee master planning project, described in Section 2, three
models were used: STORM, SWMM and Harper's Multipararaetric Water Quality Model. (28)
Model Comparisons
There have been a rash of projects comparing the merits of various models
on the basis of a variety of criteria. (156, 190-197) instances where tests had been
published of the performance of various types of models against field data were
reported in 1975. (1^1) Advances in modeling capability occur almost too rapidly to
keep track, and in 1975 it could be said that mathematical model development for
sewered system applications had already seemingly greatly outpaced the data base for
model *
Results of the various tests are mixed, mostly because there is no acceptable
basis for multiple-objective comparison. Peak flow is the major consideration in
sizing conduits, volume and hydrograph shape are critical for sizing storage, and
concentrations and loadings of pollutant emissions are essential for evaluation of
receiving water impact and helpful in sizing treatment facilities. Each model has its
strengths, weaknesses and outright faults for a given application. Over and above the
correlation definition problem is the inherent difficulty with any runoff model in the
necessarily subjective separation of abstractions (infiltration, depression storage,
etc.) from total rainfall to resolve rainfall excess (amount and pattern), which is
- 57 -
-------
the input from which an equal volume of .direct runoff is generated by models of one
kind or another. After analyzing the performance of a variety of models, it was
concluded that the weakest link is the proper estimation of rainfall excess.(198)
All this is to reiterate that urban hydrology is still, in the absence of an adequate
body of field data, more of an art than a science, and that under this circumstance
the choice of a model for a given application is largely a matter of taste.
Receiving Water Modeling
Receiving waters are the common repository of effluents from just about
every community and self-supplied industry in a metropolis, constituting perhaps
the most shared aspect of urban water resources. Impressive advances have been made
in receiving water modeling. Initial attention was on hydrology and hydraulics in
support of flood control objectives. Water quality modeling capability has evolved
more recently, with a tendency to use tailor-made models for discharge-quality
simulation in planning applications, and earlier development was focused on estuaries.
The choice of a model or models to be used in any given planning effort therefore
requires careful and discriminating study. Consequently, it is appropriate to cite
recent capability summaries and describe a few examples of applications. Reference
has been made earlier in this Section to SWMM capabilities.
A compendium, (19/) two companion reports, (199,200) a jjorth American
summary/201'' and a text,(202' survey features of large-scale water quality models-
and an annotated bibliography of models for tidal rivers, estuaries and coastal
waters is available.(20-») Tidal water models have been comprehensively
classified,(204-,205) an£j capabilities for modeling estuary and streamflow water
quality have been assessed.(206)
Aquatic ecosystem submodels have been delineated for process analysis.(207)
Aquatic ecosystem models were surveyed in 1974 for the National Commission on Water
Quality via a questionnaire,(208) and while several models reported are more generally
applicable, nearly all have been developed or tested on a specific water body and only
a fraction of the applications have urban implications. Computer graphics techniques
might be useful for displaying oceanographic data for coastal metropolises.(209)
Water quality modeling for systems containing rivers and reservoirs has
been advanced through the issuance of a description of a combination of models.(210)
The Hydrologic Engineering Center is having dynamic flow routing routines added to
the model and plans to upgrade the documentation as new developments occur.(211)
Dynamic or unsteady-flow water quality modeling is particularly important in the
case of significant pulse loadings from urban runoff or when man-made controls such
as dams are involved.
Effects of risk and uncertainty in the application of operations research
techniques, including hydrologic modeling, have been critically reviewed at length.
Although receiving waters represent only a part of the total urban water
resource, they commonly traverse entire metropolitan areas and are affected by the
actions of a multitude of local jurisdictions. Recent emphasis on regionalized
wastewater treatment and disposal has resulted in some receiving water simulation
studies on a grand scale, a few of which are noted here as examples.
San Francisco Bay. The study area extends upstream well beyond the San Francisco
metropolitan area. About two-fifths of all the surface runoff in California passes
beneath the Golden Gate bridge via the Bay estuary. About 800-mgd of treated
- 58 -
-------
wastewater effluents from industrial and municipal sources enter the Bay. Funded by
the California State Water Resources Control Board, the estuary water quality project
is modifying and calibrating an existing set of interrelated hydraulic, water quality
and ecologic models of the Bay.(213) xhe hydraulic models were designed to permit
rapid evaluation of alternative wastewater management plans at a reasonable cost and
to serve as a screening tool to reduce alternative plans to a reasonable number. Once
the screening of plans has been resolved, the ecologic models are to be used in the
simulation of the ecologic response of the Bay for the more attractive plans.
Descriptions are available of some of the water quality'21^' and flow'^lS)
characteristics of the group of models used.
Denver Area. Partly supported by an EPA planning grant, a metropolitan water
quality management planA216) ^s being resolved by the Denver Regional Council of
Governments that encompasses: a plan for an areawide system of wastewater treatment
facilities; financial arrangements and legislation and institutional guidelines for
implementation of the plan; a water quality surveillance system for regional water
quality management; and an environmental impact assessment of the proposed facilities.
Virtures of alternative treatment plant locations, sizes and groupings were analyzed
using the Hydrocomp model. Flow and water quality simulations were performed for the
mainstem and principal tributaries of a 750-square mile sector of the South Platte
River basin that essentially includes the metropolitan Denver area. Numerous flows
enter the study sector from upstream. Stream sections within the sector were divided
into 33 reaches for the simulations and hourly water quality levels and streamflows
were computed for each reach. Model calibration was complicated by a necessity to
account for upstream reservoir releases and irrigation diversions and shortcomings in
long-term water quality data. Simulations were for anticipated future land use. An
earlier DRCOG study(39; (Section 2) had developed some contemporary runoff-related
land use characteristics for 398 sub-basins, most of which are in the study area,
and this information was exploited in characterizing streamflow inputs from urban
land runoff and its associated pollution burden.
Seattle Area. In response to a State of Washington directive to develop water
pollution control and abatement plans for two basins, the Municipality of Metropolitan
Seattle and surrounding King County formed the River Basin Coordinating Committee
(RIBCO), which adopted a coordinated approach to insure development of not only an
integrated plan but an integrated planning process as well.^2!/) Study elements
included water resource management, water quality management, urban runoff and basin
drainage, solid waste management, and land use allocation. Because of an emphasis on
a continuing planning process, a systems approach was specified. This led to
extensive use of computer models and provisions for their future updating and use.
The water resource management study(21°' included use of the Hydrocomp
to develop a regional water supply plan for municipal and industrial
usage, navigation lockage, stream flushing and fish flows. Both streamflow quantity
and quality characteristics were included, together with firm yields and upper estuary
considerations. Seventeen chemical and biological parameters were simulated. Model
input/output is responsive for segments 5-square miles or larger in size.
A variant of the EPA SWMM model was used in connection with the urban runoff
evaluation portion of the RIBCO Study.(220)
Hydrocomp has since extended the usefulness of its model for non-point
pollution simulation/221'
Milwaukee Area. Responsible for finding solutions to areawide developmental
and environmental problems in a rapidly urbanizing seven county area, the Southeastern
- 59 -
-------
Wisconsin Regional Planning Commission was created in 1960. The Commission has
completed comprehensive studies for the Root River, Fox River, Milwaukee River and
the Menomonee River watersheds, and is conducting a similar study of the Kinnickinn"
River watershed. The Milwaukee master plan, outlined in Section 2, is an extension
of part of the Milwaukee River watershed comprehensive study.
Among the outputs of the comprehensive watershed studies are flood hazard
maps for computed 100-year recurrence interval flood stages and estimated annual
flood damages for floodprone reaches. Extensive simulations are made using a Corps
of Engineers backwater computer program(222' and a Soil Conservation Service flood-
routing computer program.(223) SEWRPC also uses the Hydrocomp model. There are
several other computer programs suitable for practical application that are readily
available.(224) j^g watershed studies have included an examination of both
structural and non-structural alternative plan elements for the resolution of
existing flood problems and for determination of the effect of changing land use in
an evaluation of the 1990 regional land use plan.(225) Also included have been the
development of guides for designation of floodland regulatory zones, analysis of
floodland encroachments, provision of bridge design data and criteria, provision
for subsequent updating and refinement of flood stage profiles and floodland mappine
preparation of Federal Flood Insurance reports and dissemination of flood hazard
information.'2"' Other modeling has been undertaken in connection with Section 201
and Section 208 planning.
More recently, the Hydrocomp water quality submodel has been incorporated
by its developers into flow simulation capabilities in a Hydrologic, Hydraulic Wate
Quality and Flood Economics Model.(153) ' r
San Diego Area. Under an Urban Systems Engineering Demonstration Program
comprehensive planning grant from the Department of Housing and Urban Development
the Comprehensive Planning Organization of the San Diego Region commissioned a joint
venture to formulate water resource planning methodologies and technical approaches
for estimating costs of alternative land use plans. Five interrelated computer
programs were developed: Data Management System; Water Supply Model; Wastewater
Model; Flood Control Model; and Economic Analysis Program.^22")
- 60 -
-------
SECTION 7
SOME MODELING CONSIDERATIONS
Pitfalls
Models can be misused or be used ineffectively if they are selected and
employed without prior careful consideration of their data requirements and their
place in overall information processing. Because the response formulation of
current models is keyed to land-use parameters, the extent and detail of land-use
characterization that can or should be assessed becomes a function of the intended
uses of model outputs.
While flood plain mapping of 100-year stream stages is a minimum requirement,
the mapping of more frequent occurrences is required for analysis of the economics
of flood damage mitigation.
Permanent field data acquisition systems for operations, particularly for
automatic control, can be readily justified. In Section 5, only temporary
installations could be justified for planning and analysis/design purposes. On the
other hand, a skeletal instrumentation network should be maintained throughout the
planning through operations cycle, but such long-term foresight has not been well
received in the past.
The themes described above are discussed in this Section. They have been
singled out because they are particular pitfalls that have been encountered in past
planning efforts.
A special feature of this Section is a brief review of the latest
developments in the manipulation of storage via automatic control.
Land-Use Data
Land-use information needs for planning and analysis/design models are
similar, differing mostly in the amount and detail required, the latter model
category being more sophisticated and demanding by far. For current models, the key
land-use variable for calculating runoff is the imperviousness for each land-use
category being considered. Runoff rates and volumes rise as the degree of
imperviousness increases. Because the basic hypothesis for source quality for current
models hinges on street-surface pollutants accumulated between rainfalls, the key
land-use variable for calculating pollutant accumulations is the extent of street
gutters, expressed in terms of length of street curbs. Indeed, these two parameters
may be the most sensitive of the parameters affecting the estimation of runoff and
pollutant loading amounts, respectively, according to the results from limited
sensitivity tests.
Intuitively, land-use type, degree of imperviousness and extent of curbs
should be correlatable, using residential population density as the dependent
variable. Such correlations have been found for the State of New Jersey,^"/)
sectors of metropolitan Washington, D.C.(228,229) ancj for a few other samples in the
U.S. and Canada.*166' Similar relationships can be developed for any metropolitan
area or major jurisdiction for use in planning, by characterizing representative
samples of various land-use population densities.
- 61 -
-------
However, because more detailed representation may be justified for later
preliminary design, recourse to total jurisdiction or metropolitan area assessment
of current (and perhaps prior) land use might be more consistent. While aerial
photographs can certainly be used for land-use surveys, new technology for computer-
aided interpretation of satellite data might be exploited for the explication of
imperviousness.(230) Compared with costs for interpretation of aerial maps the
new method appears to be considerably faster and less expensive.(231) Interpretive
resolution is of about the same precision as for manual reading of aerial photographs
but only for watersheds of a minimum of about one square mile in size.(231) This '
tentative conclusion is based on a case in which comparative model results were
obtained for both types of mapping.
At this juncture it is again important to remind the reader that planning
is a dynamic process concerned with alternative futures. Present conditions are
prologues or points of reference for the future and projections are best made from
the springboard of a reasonably reliable assessment of current conditions. Good
information on current actualities is needed to proceed confidently with
extrapolations, where the degree of reliability of projections diminishes as
estimates move into more remote time-frames. That is, one of the tests of
contemporary information usefulness relates to its amenability for the projection
of expected ranges of future conditions or requirements, a consideration over and
above the question of model response reliability for a given land-development
configuration (Section 5). This is true for all levels of planning but certainly
most visible at the implementation level. Land-use maps show current or recent
land use, but planning for specific facilities requires projection of land use
through the full period of implementation, which would be around two decades for
most comprehensive plans.
Because anticipated future conditions cannot be simulated by manipulating
actual land use, recourse is made to simulation by some form of calculation or
analogy. Again, as in the analysis/design of urban water resource facilities, the
first major requirement, prior to investigating requirements for the future, should
be attainment of a satisfactory simulation of existing conditions. This might be
called the tool calibration phase. Immediately the question emerges: to what detail
should basic land-use information for existing conditions be acquired? Because cost
of information acquisition rises with the degree of detail sought, it is prudent
first to derive collaterally an indication of simulation precision through
sensitivity testing. In the case of hydrological modeling, the subject of the
preceding Section, there is usually too little field data for conducting a realistic
sensitivity analysis. On the other hand, even a year of rainfall-runoff-quality
field data for one to a few catchments would lead to the use of much more reliable
model coefficients than the use of values transferred from elsewhere by default.(232)
All this is to say that if local field data is not available for some degree of
hydrological model calibration, what would be the logic in determining land use
imperviousness and curb lengths at a scale of, say, 10-acres, when a scale of perhaps
100-acres or 200-acres would be more consistent with the probable errors in assumed
or transferred hydrological model coefficients? Similarly, why determine model
information input for every bit of land involved, when a sampling of perhaps a tenth
of the planning area would be more consistent? Conversely, when land-use projection
of only specious quality are available, why insist on process elegance in the
hydrological tools employed?
Perhaps a suitable answer to the information scale and detail question
would be to suggest that: when information on a small scale and with good detail
is available it should simply be used to the maximum extent possible consistent with
- 62 -
-------
available budgets for simulation; and when only coarser raw information is available
the scale and detail used should be weighed against the presence or absence of field
rainfall-runoff-quality data for hydrological model calibration in tempering the
level of model sophistication adopted.
There are a number of computer models for land-use projection and land
allocation. An example of a projection model is one advocated for wastewater system
planning.(' A land-allocation model was employed in the Fairfax County master
planning (Section 2) to derive future runoff characteristics for the hydrological
models employed.(9»2J4) pOr both of these examples there are computer-graphics
subroutines for display of variables. Of course, land-use mapping is only one
component of composite mapping and associated environmental impact analysis.(235)
Prior to conducting inventories of land-use data and establishing procedures
or designing systems for the storage, retrieval, analysis, and display of such data,
it is important to consider the potential need, either in the modeling process or in
other aspects of master planning, for other types of areal data. These other types
of areal data, that is, natural resource and man-made features data having an areal
characteristic as opposed to point location, might include soil type, vegetal cover,
population density, land market values, and flood-prone status. All such data types
are similar to the land-use data discussed above in the sense that they have an
areal dimension and, therefore, may be readily inventoried and managed in a manner
that parallels that used for land-use data. The potential utility of a systematic
approach to the inventory, storage, retrieval, manipulation, and display of land-use
and other areal data has been demonstrated.(236)
Guidelines for acquisition of information on natural characteristics of
land for land-use planning are available.(237)
Flood Plain Mapping
It has been posited that economic criteria can and should be used in
planning nonstructural flood control measures generally, just as they have long been
used in planning structural measures.(238) gut little data is available at the local
level for making benefit analyses. (m> 239) Fundamentally, what is needed from a
hydrological standpoint is stage-frequency information, which would have to be
developed by simulation.
As noted earlier, there is considerable evidence(122^ which suggests that
the effects of urbanization on streamflows decreases as rarer floods are approached,
to the extent of a difference that may be undetectable or at least insignificant at
the 100-year level. That is, the effects of land use changes are more pronounced
the more common the occurrence, adding further substantiation to the need for some
sort of regional stage-frequency information over and above the mapping of the 100-
year event. Without a reasonable indication of stage-frequency relations for
existing conditions, how can one make flood plain planning projections with reasonable
confidence?
An attempt has been made to generalize the 2 1/3-year, 5-year, 10-year,
20-year and 50-year frequency of peak flows for urban waterways on the basis of
available data.(240) input variables of interest here are drainage area slope and
degree of imperviousness, akin to the needs for conduit flow simulation models.
- 63 -
-------
Over half of natural 100-year flood plains in urban areas are already
occupied. (103) Damages to buildings are a function of the height of flooding level
at individual buildings in a flood plain. That is, two identical buildings located
at different elevations of a waterway flood plain would suffer differing damages for
a given flood, assuming that both were not totally destroyed. Looking at it another
way, in any given instance some buildings are exposed to more frequent flooding
hazard than others. There is so much attention accorded the 100-year flood level (241)
that more frequent occurrences can be inadvertently overlooked. There is a convincin
indication that potential flooding losses might be significantly mitigated in numerou8
situations if flood protection was provided up to some relatively frequent recurrence
interval, meaning that conventional protection against a rarer flood level could be
much more costly.' '
In essence, needed are stage-frequency relations for as many streamflow
locations as is feasible if benefits are to be assessed realistically. Clearly
delineation of 100-year flood plains is not enough. The National Flood Insurance Act
of 1968 and the Flood Disaster Protection Act of 1973 call for measures implying the
necessity of using stage-frequency relations. Whether, say, 10-year, 25-year and
50-year flood plains should or could be mapped along with the 100-year levels is a
moot point.
Developers of the Fairfax County master plan (Section 2) were fortunate
that the USGS had mapped not only the 100-year County flood plains but also the
25-year and 50-year flood plains. The Southeastern Wisconsin Regional Planning
Commission (Section 2) not only mapped the 100-year flood plains of the Milwaukee
River watershed but also the 10-year highwater surface profiles using a simulation
model.
Storage Manipulation Via Automatic Control
While we may not go that far as a nation, the ultimate solution for the
problem of abating pollution from urban storm sewer discharges and combined sewer
overflows is the treatment of such flows prior to their release into receiving
waters. Outflow surface-water hydrographs of sewered catchments exhibit very large
peaks, on the order of two or more times those of equivalent non-sewered areas.
Capturing, transporting and treating all discharges/overflows, unattenuated, would
require gigantic collection sewers, pumping stations and treatment facilities all
of which would be used less than the equivalent of about an hour a day, on the
average, over a typical year. Furthermore, there will be periods running into several
weeks or even months where the discharge/overflow control works will not operate at
all, because of little or no precipitation. Therefore, schemes for system-wide
discharge/overflow collection and treatment incorporate some form of auxiliary storac
for the purpose of attenuating sudden inflows to collection and treatment facilities
with the objective of scaling down the size of such facilities to reasonable and '
manageable proportions. The complexity added by converting a facility with simple
gravity flow into a multiple-component interacting scheme, where practically
instantaneous response to flow incidences must be made, requires incorporation of
some degree of automatic operational control. In addition, automatic control permits
manipulation of dispersed storage capacities for the purpose of maximizing the
effective use of their collectively available volumes, and the diversion of
uncontainable releases to lower impact receiving water reaches, by taking advantage
of the areal non-uniformity of rainfall over large jurisdictions.
There are three distinct levels of complexity in automatic control for the
operation of urban water resource systems. In order of complexity, they are
- 64 -
-------
automated monitoring, remote supervisory control, and complete, or "hands off,"
automatic control.' ^^' So far, the use of automatic control for the manipulation
of storage has been considered seriously only for existing combined sewer systems,
although the same technology could be applied in the future to existing or new
separate stormwater sewer systems.
Two basic schemes have been considered: the exploitation of ambient
storage capacity in existing trunk and interceptor sewers in combined systems by
adding dams and diversion structures that can be remotely actuated; and the much
more expensive provision of altogether new auxiliary in-line or off-line storage
facilities that can be remotely controlled. The level of automatic control adopted
has been scaled to the investment in storage capacity added. Remote supervisory
control is the highest level of automation to which schemes for exploitation of
ambient storage capacity have aspired. This is exemplified by the metropolitan
Seattle remote supervisory control system.(244,245) &t least two other cases have
not progressed beyond automated monitoring.
Originally planned features have been outlined of the San Francisco master
plan for combined sewer overflow pollution abatement. (2^' Implementation of the
master plan (Section 2) represents the exclusive instance where complete automatic
control capability is being sought. There are fundamentally two distinctive modes
of automatic control, reactive and predictive. The reactive mode refers to the
manipulation of facilities in response to the actual occurrence of a rainstorm, as
the storm progresses over the affected jurisdiction. The predictive mode refers to
the anticipation of how facilities should be handled just prior to the beginning of
a rainstorm, and blends with the reactive mode once rainfall on the affected
jurisdiction commences, but the predictive feature is retained until the cessation
of incipient rainfall. Reactive strategies are apt to be site-specific whereas
predictive strategies are likely to be adaptable to any system.
San Francisco is currently engaged in automatic control reactive capability
development but plans to advance to a predictive capability soon. (I-') In cooperative
projects with the City, Colorado State University has advanced the automatic control
capability state of the art with respect to individual catchments and the total City
system.(' Mathematical models employed for developing control algorithms for
both control levels have had somewhat more specific information requirements than for
STORM but much less detailed requirements than for SWMM. That is, simpler models
than SWMM were used for developing control algorithms for individual catchments.
However, use of analysis/design models such as SWMM are needed to monitor and provide
a physical understanding of results from wholesale applications of far simpler models.
That is, a certain amount of intensive detailed modeling is needed to establish
parameters and indicators and to provide an underlying understanding of the governing
hydrological processes and system responses so that simplifying expedients are not
inadvertently misused.
With respect to predictive control capability development, a procedural
scenario can be hypothesized with regard to storm characterization. The data source
would be at least two years of rainfall data for a network of raingages surrounding
the jurisdiction, with their ideal density heavier towards the side from which storms
passing over the metropolis tended to prevail. Because of the importance of time of
occurrence, synchronization of raingage records would be mandatory. By means of
computer-mapping, the depth of rainfall over any point within the network at a given
time interval could be interpolated and the prevailing storm direction at a given
instant would be indicated by the slope of the maximum tilt of the fitted surface.
Storm path and traversing speed would be indicated by the progression of prevailing
- 65 -
-------
storm directions across the network from one time interval to the next. The statistical
means and a measure of variance would be derived for prevailing direction, path and
speed of storms and for total storm depths outside and inside the metropolitan area.
As an adjunct of total storm depth characterization, similar statistical characteristics
of the time-pattern of accumulation of total depths would be obtained. Next a theorv
of storm temporal and spatial variations would be postulated, the statistical
indicators would be applied to the resultant theoretical storm characterization model
and the validity of the model would be verified by testing it against at least a
season of data that had not been used in the prior analysis. It might take some
adjustment in approach or theory to reach an acceptable predictive capability. Of
some help would be the fact that advances continue to be made on the statistical
characterization of rainfall records for a single point for intervals of an hour or
more.(2^9'
Consider now the following capability, expected to become available in 1977
at each node of a metropolitan spatial grid (such as a 64-point by 64-point grid with*
nodes 3-miles apart) the digital read-out of rainfall intensity every few minutes
(such as at a 3-minute interval), the cumulative depth of rainfall at each time of
reporting (such as at a 3-minute interval), and rates and depths for each node
predicted as far in advance as 6 hours. This capability is emerging from the use of
special radar equipment in conjunction with the deployment of 200 raingages in a
recently concluded study in metropolitan St. Louis and a subsequently implemented
study in metropolitan Chicago involving a network of 300 raingages. Storm
characterization will be possible by analysis of the data from these extensive
networks; however, in system control applications later it will be necessary to have
only a radar facility and only a few telemetry-connected raingages, with the emerging
special software converting radar readings into rainfall intensities and depths
occurring and predicted, for a grid of points that could be as large as almost 200-
miles by 200-miles in size. These exciting developments are part of the "Chicago
Hydrometeorological Area Project," partially supported by an NSF/RANN grant to the
Illinois State Water Survey.t250'
Criticism of the use of synthetic storm patterns for planning and analysis/
design applications is elaborated in Addendum 2, with particular reference to cases
where substantial new storage facilities would be involved. The criticism centers
about the point that for these applications the issue is system output characterization
not rainfall characterization. This substantial reservation does not apply for '
real-time applications, as in computer-assisted supervisory control or in complete
automatic control, where storm prediction capability would be directly interfaced with
facilities operation simulation capability. The remaining element is then the
development of predictive mode tactical algorithms for operation of facilities, such
as rule curves or decision trees or the like, as is being done for the San Francisco
system.
Rajngage Networks
In 1968, fifteen of the largest metropolitan areas had a. network of
recording raingages of between about 5 and 192 instruments, with records spanning
periods of 2 to about 50 years.(251) There has since been a growth in the number
and size of networks, but the most advanced automatic network was installed in
San Francisco in 197.1, where each 0,01-inch of rainfall from 30 raingages are
transmitted at the specific second of occurrence to a central data acquisition and
recording station.(252) A. very similar, but smaller, telemetry and automatic data
logging facility has since been installed in Portland, Oregon.(253) (Both facilities
also automatically log signals from a number of field flow-indicating instruments).
- 66 -
-------
A computer-mapping program (SYMkP'"') operated off-line has been employed
by the San Francisco DPW to process a number of storm records obtained from their
network. For a given storm, the program interpolates between raingage network
readings at any specified clock time and calculates rainfall depths for an imaginary
fine grid covering the entire City. On the basis of the grid values, the program
prints a contour map showing the variation of depths over the City for the time
interval in question. Visual inspection of a series of such maps for an entire
storm, plotted for some specified time interval such as five minutes, provides
dramatic evidence of rainfall variability over time and space and the movement of
the most intense rainfall sectors across the City.
Study of storm mappings showed that, for more intensive rainfalls, only
portions of San Francisco are subjected to the higher rainfall rates at any given
time. Storms of low total depth are more uniformly distributed, but these would not
tax facilities as severely. The spatial and temporal differences observed in the
occurrence of rainfall has confirmed the presumption of the 1971 master plan that a
system of interconnected components would result in more efficient utilization of
facilities. That is, use of real-time computer-actuated control, based on sensing
the direction and likely volumes of rainfall, accompanied by a constant concurrent
updating of system status, would permit fully efficient use of all storage and flow
capacity throughout the system. The result would be a wider latitude in staging the
master plan implementation, potential improved system performance reliability, and
greater flexibility in meeting water quality standards as they are refined or revised.
In essence, system-use allocations will be balanced between storage vacancies available
and unencumbered treatment capacity, to meet sensed or predicted rainfall loadings.
Further, when overflow occurrences are unavoidable, the sites of their release can be
selected on the basis of minimal environmental impact.
Use of a raingage network for developing and using some form of remote
supervisory or complete automatic control can be rather readily justified. There is
considerable merit in installing networks for surveillance of much less exotic
systems, such as those in several other metropolitan areas. However, older networks
have had raingages with their own on-site recording charts. Synchronization of
charts was always a problem, with time correlations within 15-minutes about the best
that could be expected, but the much larger liability was the onerous and time-
consuming task of reading the inked traces and reducing the results to a digital
tabulation. The result was that usually only the data for selected storms of unusual
interest were reduced. Unless field data is reduced and analyzed, it is difficult to
justify indefinitely the maintenance of a raingage network.
The above considerations lead us to the inescapable conclusion that the
principal value in the use of raingage networks is in the operation of systems,
but that full exploitation of such networks requires automatic data logging and
reporting, an expensive commitment. However, when raingage networks are deployed
as part of an automatic control scheme their cost compared with their importance is
almost trivial. (These qualifications apply equally to flow monitoring stations,
which are an integral and indispensible part of any automatic control system).
Installation of raingage networks for planning and analysis/design is more
difficult to justify. There are some obvious advantages when employed simultaneously
with the deployment of temporary runoff and water quality field stations for the
calibration of simulation models. Also, they could prove invaluable in the checking
of post-improvement performance of systems, and there is much to be said for the
maintenance of monitoring stations on selected catchments and receiving water sites
from the beginning of planning through operation, but this type of reliability
- 67 -
-------
insurance has always been one of the most difficult to sell. Not to be overlooked
are possibilities for extending features of temporary data acquisition systems into
permanent urban flood warning systems, such as in metropolitan Melbourne.(254) jt
is not entirely inconceivable that the day when real-time bacteriological and
chemical warning systems might be added may not be too distant.
Parenthetically, mention should be made of recommendations that have been
offered for raingage densities in the operation of in-line storage or selective
discharge systems.(255) Also, there are new useful references on the effects of
storage in underground systems(256) aiuj ^n surface systems,(257-259) but mo^ Qf
these are marred by a reliance on hypothetical cases and a dependence on synthetic
storms. The latter is a practice deplored in Addendum 2 when used for important
projects.
EPA Overview
An important reference is the recent review of EPA's research and
development program for urban runoff pollution control.(133) Outstanding features
are: estimated national costs for stormwater pollution control; a summary of
references on simulation models; more numerous and more detailed urban stormwater
quality management alternatives than could be justified for inclusion in Section 4
of the present report; and identification of 174 EPA reports and 74 on-going EPA
projects, plus 15 additional references.
EPA personnel have provided annual literature reviews on urban runoff
and combined sewer overflow papers and reports.(260)
Difficulties with planning called for under P.L. 92-500. particularly in
Section 208, have been discussed by experts from outside of EPA.(261,262)
Lastly, a methodology has been developed for assessing secondary impacts
of wastewater treatment facilities on urban ecosystems in a project by The
of Ecology for EPA.(263)
- 68 -
-------
SECTION 8
REFERENCES
Section 1
1. Godfrey, Kneoland A8J Jr., "ASCE Tackles Land Use Planning," Civil Engineering.
Vol. 44, No. 7, pp. 66-69, July, 1974.
2. Guy, Harold P., and D. Earl Jones, Jr., "Urban Sedimentation in Perspective,"
J.Hyd.Div.. ASCE Proc., Vol. 98, No. HY12, pp. 2099-2116, December, 1972.
3. Hall, Warren A., "The Research Interrelationships of the National Water Commission,
The Water Resources Council and the Committee on Water Resources Research,"
Report on the Fifth Annual Water Resources Research Conference, February 3-4, 1970,
Washington, D.C., Sponsored by the Office of Water Resources Research, p. 11.
4. U.S. EPA, Draft Guidelines for Areawjde Waste Water Management, GPO, Washington,
B.C., 126 pp., May, 1974.
5. Whipple, W., Jr., and J. V. Hunter, "Nonpoint Sources and Planning for Pollution
Control," J.WPCF, Vol. 49, No. 1, pp. 15-23, January, 1977.
6. McPherson, M. B., Prospects for Metropolitan Water Management. ASCE, New York,
N.Y., 250 pp., December, 1970.
7. University of North Carolina, Chapel Hill, Promoting Environmental Quality
Through Urban Planning and Controls. U.S. EPA Socioeconomic Environmental
Studies Series EPA-600/5-73-015, GPO, Washington, D.C., 441 pp., February, 1974.
8. McPherson, M. B., and G. F. Mangan, Jr., "ASCE Urban Water Resources Research
Program," J.Hyd.Div., ASCE Proc., Vol. 101, No. HY7, pp. 847-855, July, 1975.
Authors' closure to the 27 discussions: Vol. 103, No. HY6, pp. 661-663,
May, 1977.
9. McPherson, M. B., Regional Earth Science Information in Local Water Management.
ASCE, New York, N.Y., 155 pp., July, 1975.
10. Smithsonian Science Information Exchange, Research Information Services for
Energy and the Environmental Sciences. Room 300, 1730 M Street, N.W., Washington,
D.C. 20036, 36 pp., 1977.
11. Selected Water Resources Abstracts, issued twice per month, available from Office
of Water Research and Technology, Water Resources Scientific Information Center,
Washington, D.C. 20240, and from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
12. Annual Cumulated Indexes. Selected Water Resources Abstracts. Water Resources
Scientific Information Center, Office of Water Research and Technology,
Washington, D.C. 20240. (In two volumes: Part 1, Author, Organization and
Accession Number; and Part 2, Subject. For 1976, there are 700 pp. in Part 1
and 1,201 pp. in Part 2).
- 69 -
-------
Section 2
13. (Communication from Harold C. Coffee, Jr., Department of Public Works, City
and County of San Francisco, California).
14. Department of Public Works, "Preliminary Summary Report," San Francisco Master
Plan for Waste Water Management, City and County of San Francisco, California
147 pp., September 15, 1971. '
15. Friedland, Alan 0., "City-Wide Master Planning," Urban Runoff. Quantity and
Quality, ASCE, New York, N.Y., pp. 8-12, 1975.
16. McPherson, M. B., "Innovation: A Case Study," ASCE Urban Water Resources
Research Program Technical Memorandum No. 21, ASCE, New York, N.Y., 59 pp.
February, 1974. (Available from NTIS as PB 232 166). "'
17. Department of Public Works, "Supplement I," San Francisco Master Plan for
Waste Water Management, City and County of San Francisco, California 88 pp
May 15, 1973. '
18. Roesner, L. A., "Real Time Automatic Control of San Francisco's Combined
Wastewater Collection System," a paper presented at the Storm Water Management
Model User's Group meeting at Toronto, Ontario, October 22, 1976, 26 pp.
19. (Communication from Neil S. Grigg, University of North Carolina Water Resources
Research Institute, North Carolina State University, Raleigh).
20. Golding, Bernard L., Discussion of "Master Planning Methodology for Urban
Drainage," by Harold F. Bishop, J.Hyd.Djv.. ASCE Proc., Vol. 100, No. HY9
pp. 1301-1303, September, 1974. '
21. Me tealf and Eddy, Inc., Palo Alto, California, Urban Stormwater Management anH
Technology; An Assessment, Environmental Protection Technology Series EPA-670/2-
74-040, GPO, Washington, D.C., 447 pp., December, 1974. (Available from NTIS as
PB 240 687).
22. Overton, Donald E., and Michael E. Meadows, Stormwater Modeling. Academic Press
New York, 358 pp., 1976. *
23. Southeastern Wisconsin Regional Planning Commission, "A Comprehensive Plan for
the Milwaukee River Watershed," Vol. One, "Inventory Findings and Forecasts "
Planning Report No. 13, Waukesha, 514 pp., December, 1970.
24. Southeastern Wisconsin Regional Planning Commission, "A Comprehensive Plan for
the Milwaukee River Watershed," Vol. Two, "Alternative Plans and Recommended
Plans," Planning Report No. 13, Waukesha, 625 pp., October, 1971.
25. Southeastern Wisconsin Regional Planning Commission, "Prospectus, Preliminary
Engineering Study for the Abatement of Pollution from Combined Sewer Overflow
in the Milwaukee Metropolitan Area," Waukesha, 36 pp., July, 1973.
26. Harza, Richard D., and Ramon S. LaRusso, "Deep Tunnel Technology," APWA
Reporter. Vol. 40, No. 9, pp. 11-13, September, 1973.
- 70 -
-------
27. Anonymous, "Tunnels May Store Storrawater for Later Treatment," The American
City. Vol. 88, No. 12, p. 75, December, 1973.
28. Metropolitan Sewerage District of the County of Milwaukee and Stevens, Thompson
& Runyan, Inc., "Phase One, Volume One, Technical Report, Combined Sewer Overflow
Pollution Abatement," Milwaukee, Wisconsin, 114 pp., 1975.
29. Metropolitan Sewerage District of the County of Milwaukee and Stevens, Thompson
6e Runyan, Inc., "Phase One, Volume Two, Project Design, Combined Sewer Overflow
Pollution Abatement," Milwaukee, Wisconsin, 114 pp., 1975.
30. Michel, Henry L., and William P. Henry, "Flood Control and Drainage Planning in
the Urbanizing Zone: Fairfax County, Virginia," Urban Runoff. Quantity and
Quality. ASCE, New York, N.Y., pp. 119-139, 1975.
31. Michel, H. L., and W. P. Henry, "Fairfax County Works to Solve Runoff Problems,"
Water & Wastes Engineering. Vol. 13, No. 9, pp. 38-42, September, 1976.
32. VanWeele, Brian, "Management of Flood Control and Drainage in the Urbanizing
Zone," Proceedings, National Symposium on Urban Rainfall and Runoff and Sediment
Control, Office of Research and Engineering Services Publication UKY B106,
University of Kentucky, Lexington, pp. 229-238, October, 1974.
33. Allee, Debra C., "Environmental Baseline Relationship Matrix," Urban Runoff,
Quantity and Quality. ASCE, New York, N.Y., pp. 150-159, 1975.
34. Lanyon, Richard, and Jerome Kurland, "Monitoring System Boosts Chicago Water
Data Collection System," Water & Wastes Engineering. Vol. 8, No. 1, pp. 20-23,
January, 1971.
35. Flood Control Coordinatinp Committee, "The Chicago Underflow Plan Development
of a Flood and Pollution Control Plan for the Chicagoland Area," State of
Illinois, County of Cook, MSDGC, City of Chicago, 29 pp., December, 1972.
36. MSD NEWS, Vol. 3, No. 3, 4 pp., Fall, 1972.
37. Lanyon, Richard, "Flood Plain Management in Metropolitan Chicago," Civil
Engineering. Vol. 44, No. 5, pp. 79-81, May, 1974.
38. Lanyon, Richard F., and James Jackson, "Flow Simulation System," J.Hyd.Div.,
ASCE Proc., Vol. 100, No. HY8, pp. 1089-1105, August, 1974.
39. Denver Regional Council of Governments and Urban Drainage and Flood Control
District, "Urban Storm Drainage and Flood Control in the Denver Region," Project
REUSE Final Report, 219 pp., August, 1972.
40. Denver Regional Council of Governments, "Storm Drainage and Flood Control for
Metropolitan Denver," Project REUSE Summary Report, 18 pp., May, 1973.
41. Bishop, Harold F., "Master Planning Methodology for Urban Drainage," J.Hyd.Div.,
ASCE Proc., Vol. 100, No. HY1, pp. 189-199, January, 1974. Author's closure to
discussion: Vol. 101, No. HY4, p. 412, April, 1975.
- 71 -
-------
42. Roper, Willard, "Wastewater Management Studies by the Corps of Engineers "
J.Envir.Engrg.Div., ASCE Proc., Vol. 99, No. EE5, pp. 653-669, October, 1973.
Author's closure to discussion: Vol. 101, No. EE1, p. 164, February, 1975.
43. Department of Defense, Corps of Engineers, U.S. Army, "Urban Studies Program
Proposed Policies and Procedures," Federal Register, Vol. 39, No. 130, Part III
pp. 24754-24771, Friday, July 5, 1974. '
44. Corps of Engineers, "San Francisco Bay and Sacramento-San Joaquin Delta Water
Quality and Waste Disposal Investigation, Public Brochure, Wastewater Management
Technical Alternatives," U.S. Array Engineer District, San Francisco, 30 pp.
August, 1972.
45. Michel, H. L, P. H. Gilbert and H. K. Creed, "Land Disposal, What's the
Realistic View?," Water & Wastes Engineering, Vol. 11, No. 6, pp. 30-35,
June, 1974.
46. Boyce, David E., Norman D. Day and Chris McDonald, Metropolitan Plan Making
"An Analysis of Experience with the Preparation and Evaluation of Alternative
Land Use and Transportation Plans," Regional Science Research Institute
Monograph Series Number Four, Philadelphia, Petina., 450 pp., 1970.
47. Banks, Harvey 0», and Jean 0. Williams, "Water-Resources Planning," J.AWWA.
Vol. 65, No. 11, pp. 665-669, November, 1973.
Section 3
48. American Water Resources Association, Urbanization and Water Quality Control
Edited by William Whipple, Jr., Minneapolis, Minnesota, 302 pp., 1975.
49. Citizens' Advisory Committee on Environmental Quality, How Will America Grow?
A Citizen Guide to Land-Use Planning, GPO Stock No. 040-000-00347-9, Washington
D.C., 39 pp., April, 1976. '
50. Walker, William R., and William E. Cox, "Water - An Element of Land-Use and
Urban Growth Policies," J.Urban Plan.&Dev.Div.. ASCE Proc., Vol. 102, No. UP1,
pp. 81-94a August, 1976.
51. Lienesch, William C., and Gerald A. Emison, "Planning for Implementation Under
Section 208," J.Wat.Res.Plan.&Mgt.Div.. ASCE Proc., Vol. 102, No. WR2, pp. 283-
295, November, 1976.
52. Pisano, Mark A., "Nonpoint Sources of Pollution: A Federal Perspective,"
J.Envir.Engrg.Div., ASCE Proc., Vol. 102, No. EE3, pp. 555-565, June, 1976.
I
53. Pisano, Mark, "208: A Process for Water Quality Management," Civil Engineering
Vol. 46, No. 11, pp. 55-57, November, 1976.
54. Billings, Leon G., "The Evolution of 208 Water-Quality Planning," Civil
Engineering. Vol. 46, No. 11, pp. 54-55, November, 1976.
55. "Areawide Assessment Procedures Manual," in three volumes, report EPA-600/9-76-014
Municipal Environmental Research Laboratory, Office of Research and Development '
U.S. EPA, Cincinnati, Ohio 45268, July, 1976, et seq. *
- 72 -
-------
56. Flannery, James J., "New Publications," EOS. Transactions of AGU, Vol. 57,
No. 7, pp. 510-515, July, 1976.
57. Peskin, Henry M., and Eugene P. Seskin, Cost Benefit Analysis and Water
Pollution Policy. Urban Institute, Washington, D.C., 370 pp., 1975.
58. Gulp, Gordon L., and Leonard S. Horner, "Coping with EPA's Value Engineering
Requirement," Water & Sewage Works. Vol. 123, No. 12, pp. 48-51, December, 1976.
59. Ehrenfield, David W., "The Conservation of Non-Resources," American Scientist,
Vol. 64, No. 6, pp. 648-656, November-December, 1976.
60. Lindh, Gunnar, Socjo-Economic Aspects of Urban Hydrology, The Unesco Press,
Paris, 1977 (in press).
61. Cox, William E., and William R. Walker, "Recognition of Aesthetic Values in
Water Law," J.AWWA. Vol. 69, No. 1, pp. 31-34, January, 1977.
62. Hill, Douglas, "A Resource Allocation Model for the Evaluation of Alternatives
in Section 208 Planning Considering Environmental, Social and Economic Effects,"
pp. 401-406 in Environmental Modeling and Simulation. Office of Research and
Development and Office of Planning and Management, U.S. EPA, report EPA-600/9-
76-016, 847 pp., July, 1976. (Available from NTIS as PB 257 142).
63. Water Resources Scientific Information Center, OWRT, Water Reuse. A Bibliography,
GPO, Washington, D.C., Vol. 3 (444 pp.) and Vol. 4 (448 pp.), March, 1975.
64. Dworkin, Daniel M., "Water Reuse: A Flexible and Efficient Management Alternative
for Municipal Water Supply," Water Resources Research. Vol. 11, No. 5, pp. 607-
615, October, 1975.
65. SCS Engineers, Long Beach, California, Demonstrated Technology and Research
Needs for Reuse of Municipal Wastewater. U.S. Environmental Protection Agency
report EPA-670/2-75-038, GPO, Washington, D.C., 338 pp., May, 1975. (Available
from NTIS as PB 249 151).
66. Bargman, Robert D., George W. Adrian and Donald C. Tillman, "Water Reclamation
in Los Angeles," J.Envir.Engrg.Div.. ASCE Proc., Vol. 99, No. EE6, pp. 939-955,
December, 1973.
67. Seaburn, G. E., and D. A. Aronson, Influence of Recharge Basins on the Hydrology
of Nassau and Suffolk Counties, Long Island, New York, U.S. Geological Survey
Water Supply Paper 2031, GPO, Washington, D.C., 66 pp., 1974.
68. Hulob, H., "Some Legal Problems in Urban Runoff," in Hydrology and Water
Resources in Arizona and the Southwest, pp. 169-179, Vol. 2, Proceedings of the
1972 meetings of the Arizona Section, American Water Resources Assn., and the
Hydrology Section, Arizona Academy of Science, Prescott, Arizona, 1972.
69. McPherson, M. B., "Management Problems in Metropolitan Water Resource Operations,"
ASCE Urban Water Resources Research Program Technical Memorandum No. 14, ASCE,
New York, N.Y., 76 pp., September, 1971. (Available from NTIS as PB 206 087).
70. Schneider, W. J., and Andrew M. Spieker, Water for the Cities the Outlook.
U.S. Geological Survey Circular 601-A, GPO, Washington, D.C., 6 pp., 1969.
- 73 -
-------
71. Appelj Charles A., and John D. Bredehoeft, Status of Ground-Water Madeline -in
the U.S. Geological Survey. U.S. Geological Survey Circular 737, GPO, Washington
D.C., 9 pp., 1976. '
72. Todd, D. K., R. M. Tinlin, K. D. Schmidt and L. G. Everett, "A Groundwater-
Quality Monitoring Methodology," J.AWWA. Vol. 68, No. 11, pp. 586-593, November
1976.
73. Todd, David K., and Daniel E. 0. McNulty, Polluted Groundwater. Water Information
Center, Port Washington, N.Y., 178 pp., 1976.
74. Geyer, J. C., and J. J. Lentz, "An Evaluation of the Problems of Sanitary Sewer
System Design," J.WPCF. Vol. 38, No. 7, pp. 1138-1148, July, 1966.
75. American Public Works Association, Control of Infiltration and Inflow into Sewer
Systems, Water Pollution Control Research Series 11022 EFF 12/70, GPO, Washington
D.C., 121 pp., December, 1970. (Available from NTIS as PB 200 827). *
76. Office of Water Program Operations, Guidance for Sewer System Evaluation. U.S.
EPA report EPA-520/9-74-018, GPO, Washington, D.C., 10 pp., March, 1974.
77. Cesareo, David J., and Richard Field, "Infiltration-Inflow Analysis," J.Envjr.
Engrg.Div.. ASCE Proc., Vol. 101, No. EE5, pp. 775-785, October, 1975. Authors'
closure: Vol. 102, No. EE5, pp. 1126-1127, October, 1976.
78. Click, C. N., and F. 0. Mixon, "Flow Smoothing in Sanitary Sewers," J.WPCF
Vol. 46, No. 3, pp. 522-531, March, 1974. '
79. Beckwith, R. A., "Infiltration/Inflow Program for EPA Grants," pp. R-16 - R-18-
P. A. Podolick, "Preparing an Infiltration/Inflow Analysis," pp. R-31 - R-34-
and D. J. Cesareo and R. Field, "How to Analyze Infiltration/Inflow," pp. R-84
R-90: Water & Sewage Works, Reference Number, 1975.
80. Office of Water Program Operations, Handbook for Sewer System Evaluation and
Rehabilitation, U.S. EPA Municipal Construction Division Technical Report MCD-19
EPA 430/9-75-021, Washington, D.C., 218 pp., December, 1975. '
81. Klemetson, Stanley L., and William J. Grenney, "Development of a Dynamic
Programming Model for the Regionalization and Staging of Wastewater Treatment
Plants," Utah Water Research Laboratory report PRWA20-2, Utah State University
Logan, 124 pp., June, 1975. (See J.WPCF. Vol. 48, No. 12, pp. 2690-97, Dec. 19*76)
82. Weeter, Dennis W., and James G. Belardi, "Analysis of Regional Waste Treatment
Systems," J.Envir.Engrg.Div.. ASCE Proc., Vol. 102, No. EE1, pp. 233-237,
February, 1976.
83. Dajani, Jarir S., and Robert S. Gemmell, "Economic Guidelines for Public
Utilities Planning," J.Urban Plan.&Devel.Div.. ASCE Proc., Vol. 99, No. UP2,
pp. 171-182, September, 1973.
84. Anonymous, "Symposium Studies Solutions to Regional Wastewater Problems,"
"STRIDES," Stevens, Thompson & Runyan, Inc., Portland, Oregon, 3rd Quarter,
8 pp., 1976.
- 74 -
-------
85. Anonymous, "Builders Assault the No-Growth Laws," Business Week, pp. 44 and 46,
September 8, 1973.
86. Rabe, Frank T., and James F. Hudson, "Highway and Sewer Impacts on Urban
Development," J.Urban Plan.&Devel.Div.. ASCE Proc., Vol. 101, No. UP2, pp. 217-
231, November, 1975.
87. Real Estate Research Corporation, The Costs of Sprawl. GPO, Washington, D.C.,
278 pp., April, 1974. (An "Executive Summary," 15 pp., is also available
from GPO).
88. Roberts, James S., "Energy and Land Use: Analysis of Alternative Development
Patterns," Environmental Comment, Urban Land Institute, Washington, D.C.,
pp. 2-11, September, 1975.
89. Wolman, M. Gordon, "The Physical Environment and Urban Planning,"
Geographical Perspectives and Urban Problems. National Academy of Sciences,
pp. 55-70, 1973.
90. Ports, Michael A., "Sediment and Erosion Control Design Criteria," APWA
Reporter. Vol. 42, No. 5, pp. 18-19, May, 1975.
91. Sachs, Herbert M., "The State Water Resources Planning Program in Maryland,"
Proceedings. Symposium on Water Resources Needs Facing the District of
Columbia, Washington Technical Institute, WRRC Report No. 3, Washington, D.C.,
pp. 17-28, June 20, 1974.
92. Water Resources Administration, "The Maryland Sediment Control Program,"
Department of Natural Resources, Annapolis, 8 pp., August, 1972.
93. McCuen, Richard H., "A Regional Approach to Urban Storm Water Detention,"
Geophysical Research Letters. AGU, Vol. 1, No. 7, pp. 321-322, November, 1974.
94. Parsons, Brinckerhoff, Quade & Douglas, New York City, "Erosion, Sedimentation
& Debris Control," Task Order 11.0, Master Plan for Flood Control and Drainage,
Fairfax County, Virginia, 183 pp., May, 1975.
95. Chen, Charng-Ning, "Evaluation and Control of Soil Erosion in Urbanizing
Watersheds," Proceedings, National Symposium on Urban Rainfall and Runoff and
Sediment Control, Office of Research and Engineering Services Publication
UKY B106, University of Kentucky, Lexington, pp. 161-173, October, 1974.
96. Chen, Charng-Ning, "Planning Tools for Erosion Control in Urbanizing
Watersheds," Urban Runoff. Quantity and Quality. ASCE, New York, N.Y., pp. 159-
165, 1975.
97. Powell, M. D., W. C. Winter and W. P. Bodwitch, "Community Action Guidebook for
Soil Erosion and Sediment Control," The American County. Vol. 35, No. 5, 64 p.
insert, May, 1970.
98. Guy, Harold P., Sediment Problems in Urban Areas. U.S. Geological Survey
Circular 601-E, GPO, Washington, D.C., 8 pp., 1970.
99. Hittman Associates, An Executive Summary of Three EPA Demonstration Programs
in Erosion and Sediment Control. Environmental Protection Technology Series,
Report EPA-660/2-74-073, GPO, Washington, D.C., 50 pp., June, 1974. (Available
from NTIS as PB 239 333).
- 75 -
-------
100. ULI, ASCE and NAHB, Residential Storm Water Management. ASCE, New York, N.Y.
64 pp., 1975. '
101. Task Committee, "Urban Sediment Problems: A Statement on Scope, Research,
Legislation, and Education," J.Hyd.Div.. ASCE Proc., No. HY4, pp. 329-340,
April, 1975. Committee's closure to discussion: Vol. 102, No. HY10, pp.'l598-
1599, October, 1976.
102. Dallaire, Gene, "Controlling Erosion and Sedimentation at Construction Sites "
Civil Engineering. Vol. 47, No. 10, pp. 73-77, October, 1976. '
Section 4
103. Schneider, William J., and James E. Goddard, Extent and Development o£ Urban
Flood Plains, U.S. Geological Survey Circular 601-J, GPO, Washington, D.C.,
14 pp., 1974.
104. "Proceedings of a Seminar on Nonstructural Flood Plain Management Measures,"
4-6 May 1976, The Hydrologic Engineering Center, U.S. Army Corps of Engineers
609 Second Street, Davis, Cal. 95616, and Institute for Water Resources, U.S.*
Army Corps of Engineers, Kingman Building, Fort Belvoir, Virginia 22060, 193 pp.
105. Goddard, James E., "An Evaluation of Urban Flood Plains," ASCE Urban Water
Resources Research Program Technical Memorandum No. 19, ASCE, New York, N.Y.
40 pp., December, 1973. (Available from NTIS as PB 227 337).
106. Poertner, H. G., Robert L. Anderson and Karl W. Wolf, Urban Drainage Practices
Procedures and Needs. APWA Research Foundation, Special Report No. 31, Chicago
Illinois, 54 pp., December, 1966.
107. Thomas, H. E., and W. J. Schneider, Water as an Urban Resource and Nuisance.
U.S. Geological Survey, Circular 601-D, GPO, Washington, D.C., 9 pp., 1970.
108. Sheaffer, J. R., D. W. Ellis and A. M. Spieker, Flood-Hazard Mapping in
Metropolitan Chicago. U.S. Geological Survey, Circular 601-C, GPO, Washington,
D.C., 14 pp., 1970.
109. Wood, Walter J., "Los Angeles County Flood Control System and the Early 1969
Storms," Civil Engineering. Vol. 40, No. 1, pp. 58-61, January, 1970.
110. Rantz, S. E., Urban Sprawl and Flooding in Southern California. U.S. Geological
Survey, Circular 601-B, GPO, Washington, D.C., 11 pp., 1970.
111. Grigg, N. S., "Evaluation and Implementation of Urban Drainage Projects,"
J.Urban Plan.&Devel.. ASCE Proc., Vol. 101, No. UP1, pp. 61-75, May, 1975.
112. Walesh, Stuart G., "Floodland Management: The Environmental Corridor Concept,"
Technical Record, Vol. 3, No. 6, Southeastern Wisconsin Regional Planning
Commission, Waukesha, pp. 1-13, April, 1976.
113. Walesh, Stuart G., "Floodland Management: The Environmental Corridor Concept,"
Hydraulic Engineering and the Environment. ASCE, New York, N.Y., pp. 105-111, 1973
114. Wendell, Mitchell, and Harvey 0. Banks, "Management of Water Resources for Urban
Environment Enhancement," Environments for Tomorrow, Inc., McLean, Virginia,
275 pp., July, 1974. (Available from NTIS as PB 239 301).
- 76 -
-------
115. Bauer, W. J«, "Economics of Urban Drainage Design," J.Hyd.Div.. ASCE Proc.,
Vol. 88, No. HY6, pp. 93-114, November, 1962. Author's closure to discussion:
Vol. 90, No. HY1, p. 257, January, 1964.
116. Rice, Leonard, "Reduction of Urban Peak Flows by Ponding," J.Irrig.&Drajn.Div.,
ASCE Proc., Vol. 97, No. IR3, pp. 469-482, September, 1971.
117. Kalvinskas, John J., Wallace A. Moser and Thomas E. Bullock, "Water Reservoir
Systems," Water Resources Bulletin. Vol. 7, No. 2, pp. 330-342, April, 1971.
118. Poertner, Herbert G., "Better Storm Drainage Facilities At Lower Cost,"
Civil Engineering. Vol. 43, No. 10, pp. 67-70, October, 1973.
119. Poertner, Herbert G., Practices in Detention of Urban Stormwater Runoff. APWA
Special Report No. 43, Chicago, Illinois, 231 pp., June, 1974. (Available from
NTIS as PB 234 554).
120. Poertner, Herbert G., "Urban Stormwater Detention and Flow Attenuation," Public
Works. Vol. 107, No. 8, pp. 83-85, August, 1976.
121. Becker, B. C., M. L. Clar and R. R. Kautzman, "Approaches to Stormwater
Management," Hittman Associates, Columbia, Maryland, 258 pp., November, 1973.
(Available from NTIS as PB 228 124).
122. Hollis, G. E., "The Effect of Urbanization on Floods of Different Recurrence
Interval," Water Resources Research. Vol. 11, No. 3, pp. 431-435, June, 1975.
123. Lumb, A. M., J. R. Wallace and L. D. James, "Analysis of Urban Land Treatment
Measures for Flood Peak Reduction," Georgia Institute of Technology, Georgia
Environmental Resources Center Report No. ERC-0574, Atlanta, 146 pp., June,
1974. (Available from NTIS as PB 236 418).
124. Hydrologic Engineering Center, Phase I Oconee Basin Pilot Study, Trail Creek
Test. An Investigation of Concepts and Methods for Broadened Scope Flood Plain
Information Studies, U.S. Army Corps of Engineers, Davis, California, 98 pp.,
September, 1975.
125. Walesh, S. G., and R. M. Videkovich, "Urbanization: Hydrologic, Hydraulics
and Flood Damage Effects," a paper presented at the ASCE Hydraulics Division
Conference, August 4-7, 1976, West Lafayette, Indiana, 29 pp.
126. Benjes, Henry H., Jr., Cost Estimating Manual Combined Sewer Overflow Storage
and Treatment, Environmental Protection Technology series EPA-600/2-76-286,
U.S. EPA Municipal Environmental Research Laboratory, Cincinnati, Ohio, 123 pp.>
December, 1976. (Available from NTIS as PB 266 359).
127. Rawls, Walter J., and John N. Knapp, "Methods for Predicting Urban Drainage
Costs," J.Hyd.Div.. ASCE Proc., Vol. 98, No. HY9, pp. 1575-1585, September, 1972,
128. Kohlhaas, Charles A., "The Optimization of Storm-Holding Tanks: A Problem of
Water Pollution Control," Stanford University, California, PhD dissertation,
321 pp., February, 1970.
129. Rickert, David A., and Andrew M. Spieker, Real-Estate Lakes. U.S. Geological
Survey Circular 601-G, GPO, Washington, D.C., 19 pp., 1971.
- 77 -
-------
130. Britton, L. J., R. C. Averett and R. F. Ferreira, An Introduction to the
Processes. Problems, and Management of Urban Lakes. U.S. Geological Survey
Circular 601-K, GPO, Washington, D.C., 22 pp., 1975.
131. Anonymous, Man-Made Lakes: Their Problems and Environmental Effects. Geophysical
Monograph Series Volume 17, American Geophysical Union, Washington, D.C.,
900 pp., 1973.
132. Metcalf & Eddy, Inc., Palo Alto, California, Development and Application of a
Simplified Stormwater Management Model. Environmental Protection Technology
Series EPA-600/2-76-218, U.S. EPA Municipal Environmental Research Laboratory,
Cincinnati, Ohio, 139 pp., August, 1976. (Available from NTIS as PB 258 074)!
133. Field, Richard, Anthony N. Tafuri and Hugh E. Masters, Urban Runoff Pollution
Control Technology Overview. Environmental Protection Technology Series
EPA-600/2-77-047, Municipal Environmental Research Laboratory, Cincinnati,
Ohio 45268, 91 pp., March, 1977.
134. Heaney, James P., and Stephan J. Nix, Storm Water Management Model: Level I
Comparative Evaluation of Storage-Treatment and Other Management Practices.
Environmental Protection Technology Series EPA-600/2-77-083, Municipal Environmental
Research Laboratory, U.S. EPA, Cincinnati, Ohio 45268, 88 pp., April, 1977.
135. Whipple, William, Jr., "Water Quality: Problems and Research Needs," Water &
Sewage Works. Vol. 123, No. 12, pp. 40-41, December, 1976. *
136. Weathers, John W., Editor, Flood Damages Prevention. An Indexed Bibliography
Tennessee Valley Authority, Knoxville, Tennessee 37902, and University of
Tennessee Water Resources Research Center, Knoxville, Tennessee 37916, 61 pp.
October, 1976.
137. Miller, William L., and Steven P. Erickson, "Systematic Development of
Methodologies in Planning Urban Water Resources for Medium Size Communities,
Economic and Environmental impacts of Surface Runoff Disposal Systems," Purdue
University, Water Resources Center Technical Report No. 39, West Lafayette,
Indiana, 74 pp., December, 1973. (Available from NTIS as PB 234 315).
138. Debo, T. N., "Survey and Analysis of Urban Drainage Ordinances and a Recommended
Model Ordinance," Georgia Institute of Technology, Environmental Research Center
Report No. ERC-0475, Atlanta, 294 pp., February, 1975. (Available from NTIS as
PB 240 817).
Section 5
139. Gray, Donald M., Editor, Handbook on the Principles of Hydrology. Water
Information Center, Port Washington, N.Y., 720 pp., 1973.
140. Water Resources Engineers, Walnut Creek, California, Future Direction of Urban
Water Models, Environmental Protection Technology Series EPA-600/2-76-058,
U.S. EPA Municipal Environmental Research Laboratory, Cincinnati, Ohio, 86 pp.
February, 1976. (Available from NTIS as PB 249 049). " *'
141. McPherson, M. B., "Urban Hydrological Modeling and Catchment Research in the
U.S.A.," ASCE Urban Water Resources Research Program Technical Memorandum No.
IHP-1, ASCE, New York, N.Y., 49 pp., November, 1975. (Available from NTIS as
PB 260 685).
- 78 -
-------
142. Whipple, William, Jr., "Conference Summary," Urban Runoff. Quantity and Quality.
ASCE, New York, N.Y., pp. 255-258, 1975.
143. American Society of Civil Engineers, Basic Information Needs in Urban
Hydrology. ASCE, New York, N.Y., 112 pp., April, 1969.
144. Schneider, William J., "The U.S. Geological Survey Urban Water Program," Effects
of Watershed Changes on Streamflow. University of Texas Press, Austin, pp. 165-
168, 1969.
145. Committee Report, "Water-Distribution Research and Applied Development Needs,"
J.AWWA. Vol. 66, No. 6, pp. 385-390, June, 1974.
146. Alley, William M., Guide for the Collection. Analysis and Use of Urban
Stormwater Data. ASCE, New York, N.Y., 115 pp., 1977.
147. Radziul, Joseph V., and Patrick R. Cairo, Philadelphia Water Department, "Case
Study of Philadelphia, Pennsylvania," paper presented at Engineering Foundation
Conference on Instrumentation and Analysis of Urban Stormwater Data, Quantity
and Quality, November 28-December 3, 1976, Easton, Maryland, 39 pp.
148. Richardson, Steven L., "A Solution to Combined Sewer Overflows," Clearwaters.
New York Water Pollution Control Association, Vol. 4, No. 4, pp. 14-15,
January, 1975.
149. Yarnell, D. L., Rainfall Frequency-Intensity Data. U.S. Department of Agriculture,
Miscellaneous Publication 204, 1935.
150. Hershfield, David M., Rainfall Frequency Atlas of the United States. Weather
Bureau, U.S. Department of Commerce, Technical Paper No. 40, May, 1961.
151. Hathaway, G. A., "Design of Drainage Facilities," Transactions. ASCE, pp. 697-
730, Vol. 110, 1945.
152. McPherson, M. B., "Some Notes on the Rational Method of Storm Drain Design," ASCE
Urban Water Resources Research Program Technical Memorandum No. 6, ASCE, New
York, N.Y., 74 pp., January 22, 1969. (Available from NTIS as PB 184 701).
Section 6
153. McPherson, M. B., (Editor), "Utility of Urban Runoff Modeling," ASCE Urban Water
Resources Research Program Technical Memorandum No. 31, ASCE, New York, N.Y.,
126 pp., July, 1976. (Available from NTIS as PB 261 460).
154. Gburek, William J., Discussion of "Hydrologic Consequences of Rainfall
Augmentation," J.Hyd.Div.» ASCE Proc., Vol. 97, No. 12, pp. 2114-2115,
December, 1971.
155. Black, Crow and Eidsness, Inc., and Jordan, Jones and Goulding, Inc., "Study
and Assessment of the Capabilities and Cost of Technology for Control of
Pollutant Discharges from Urban Runoff," a report for the National Commission
on Water Quality, 380 pp., October, 1975. (Available from NTIS as PB 247 391).
- 79 -
-------
156. Proctor and Redfern, Ltd., and James F. MacLaren, Ltd., Storm Water Management
Model Study. Volume II, Research Program for the Abatement of Municipal Pollution
under Provisions of the Canada-Ontario Agreement on Great Lakes Water Quality
Research Report No. 48, Environment Canada, Ottawa, 148 pp., September, 1976.*
157. Field, Richard, et al., Proceedings of Workshop on Microorganisms in Urban
Stormwater, Environmental Protection Technology Series EPA-600/2-76-244 U.S
EPA, Cincinnati, Ohio, 120 pp., November, 1976. (Available from NTIS as
PB 263 030).
158. Department of Economics, Georgetown University, The Integrated Multi-Media
Pollution Model, U.S. EPA Socioeconomic Studies Series EPA-600/5-74-020 GPO
Washington, D.C., 259 pp., February, 1974. ' *
159. American Society of Civil Engineers, Urban Water Resources Research. ASCE New
York, N.Y., 43 pp. + 588 pp. appendices, September, 1968.*
160. Prawdzik, Ted B., Milwaukee DPW, "Environmental and Technical Factors for Open
Drainage Channels in Milwaukee," ASCE Urban Water Resources Research Program
Technical Memorandum No. 12, ASCE, New York, N.Y., 37 pp., February, 1970
(Available from NTIS as PB 191 710).
161. Tucker, L. S., "Sewered Drainage Catchments in Major Cities," ASCE Urban Water
Resources Research Program Technical Memorandum No. 10, ASCE, New York N.Y.
71 pp., March 31, 1969. (Available from NTIS as PB 184 705). ' * *'
162. Roesner, L. A., et al., Water Resources Engineers, The Hydrologic Engineering
Center/Corps of Engineers, and Department of Public Works/City and County of
San Francisco, "A Model for Evaluating Runoff-Quality in Metropolitan Master
Planning," ASCE.Urban Water Resources Research Program Technical Memorandum
No. 23, New York, N.Y., 73 pp., April, 1974. (Available from NTIS as PB 234 312
or from the H.E.C., Davis, Cal. 95616).
163. Hydrologic Engineering Center, Corps of Engineers, "Storage, Treatment, Overflow
Runoff Model (STORM), Users' Manual," Computer Program 723-S8-L7520, 609 2nd '
Street, Davis, California 95616, 170.pp., July, 1976.
164. Winner, Jochen, and Michael Shapiro, Discussion of "Urban Runoff Pollution
Control ~ State-of-the-art," J.Envir.Engra.Div.. ASCiS Proc., Vol. 102, No. EB1
pp. 220-223, February, 1976. * '
165. Meta Systems, Inc., Land Use-Water Quality Relationships. U.S. EPA, Water Quality
Guidance report WPD 3-76-02, Water Planning Division, 233 pp., March, 1976.
166. Heaney, J. P., W. C. Huber and S. J. Nix, Storm Water Management Model; Levoi T _
Preliminary Screening Procedures., Environmental Protection Technology Series "^
EPA-600/2-76-275, U.S. EPA Municipal Environmental Research Laboratory,
Cincinnati, Ohio, 77 pp., October, 1976. (Available from NTIS as PB 259 916).
167. Jalal, Kazi F., "Water Quality Impacts of Urbanization A Methodology,"
J.Envir.EnRrg.Div.. ASCE Proc., Vol. 103, No. ESI, pp. 49-57, February,.1977.
168. Lager, John A., "A Simplified Stonnwater Management Model for Planning," pp. 259
279, Short Course Proceedings. Applications of Stormwater Management Models! 197!
Environmental Protection Technology Series EPA-600/2-77-065, U?S. EPA Municipal
Environmental Research Laboratory, Cincinnati, Ohio, 435 pp., March 1977
(Available from NTIS as PB 265 321). '
- 80 -
-------
169. Young, G. K., "Non-Point Source Impact and Urban Holding Capacity," pp. 1-98
to 1-121 in Proceedings^ Urban Stormwater Management Seminars (Atlanta and
Denver), U.S. EPA, Water Quality Management Guidance report WPD 03-76-04,
Washington, D.C., 482 pp., January, 1976. (Available from NTIS as PB 260 889).
170. Young, G. K., "Decision Perspectives on Urban Storm Water Pollution," Water
Resources Research, Vol. 12, No. 1, pp. 94-100, February, 1976.
171. Espey, W. H., Jr., and D. E. Winslow, "The Effects of Urbanization on Unit
Hydrographs for Small Watersheds, Houston, Texas," A Report to OWRR, TRACOR
Document No. 68-975-U, Austin, Texas, 70 pp., September 25, 1968. (Appendix,
Document No. 68-1006-U, 244 pp.).
172. Everhart, Ralph C., "New Town Planned Around Environmental Aspects," Civil
Engineering, Vol. 43, No. 9, pp. 69-73, September, 1973.
173. Claycomb, Elmer L., "Urban Storm Drainage Criteria Manual from Denver," Civil
Engineering, Vol. 40, No. 7, pp. 39-41, July, 1970.
174. Wright-McLaughlin Engineers, "Urban Storm Drainage Criteria Manual," Denver
Regional Council of Governments, Denver, Colorado, 1969. (In sixteen sections).
175. Hydrocomp International, Inc., "Simulation of Discharge and Stage Frequency
for Flood Plain Mapping in the North Branch of the Chicago River," Palo Alto,
California, 75 pp., February, 1971.
176. Crawford, Norman H., "Studies in the Application of Digital Simulation to
Urban Hydrology," Hydrocomp International, Palo Alto, California, 100 pp.,
September, 1971. (Available from NTIS as PB 206 715).
177. Crawford, Norman H., "Computer Simulation for Design Criteria for Urban Flow
Storage Systems," Hydrocomp International, Palo Alto, California, 109 pp.,
January, 1973. (Available from NTIS as PB 218 827).
178. Linsley, R., and N. Crawford, "Continuous Simulation Models in Urban Hydrology,"
Geophysical Research Letters. AGU, Vol. 1, pp. 59-62, May, 1974.
179. Terstriep, M. L., and J. B. Stall, "Urban Runoff by Road Research Laboratory
Method," J.Hyd.Div.. ASCE Proc., Vol. 95, No. HY6, pp. 1809-1834, November,
1969. Authors' closure to discussion: Vol. 97, No. HY4, pp. 574-579, April,
1971.
180. Stall, John B., and Michael L. Terstriep, Storm Sewer Design An Evaluation
of the RRL Method, Environmental Protection Technology Series EPA-R2-72-068,'
GPO, Washington, D.C., 73 pp., October, 1972. (Available from NTIS as PB 214 314),
181. Terstriep, M. L., and J. B. Stall, The Illinois Urban Drainage Area Simulator.
ILLUDAS. Bulletin 58, Illinois State Water Survey, Urbana, 90 pp., 1974.
182. Lanyon, R. F., and J. P. Jackson, Metropolitan Sanitary District of Greater
Chicago, "A Streamflow Model for Metropolitan Planning and Design," ASCE Urban
Water Resources Research Program Technical Memorandum No. 20, ASCE, New York,
N.Y., 49 pp., January, 1974. (Available from NTIS as PB 232 181). (A truncated
presentation is included in Reference 38).
- 81 -
-------
183. Jennings, Marshall E., and Harold C. Mattraw, "Comparison of the Predictive
Accuracy of Models of Urban Flow and Water-Quality Processes," Proceedings,
National Symposium on Urban Hydrology, Hydraulics and Sediment Control,
Lexington, Kentucky, pp. 1-8, July, 1976.
184. University of Florida, Storm Water Management Model User's Manual. Version II
Environmental Protection Technology Series EPA-670/2-75-017, U.S. EPA,
Cincinnati, Ohio 45268, 350 pp., March, 1975.
185. Bureau of Engineering, City of Chicago, "Computer Simulation Programs,
Development of a Flood and Pollution Control Plan for the Chicagoland Area,"
Technical Report Part 2, 101 pp., December, 1972. (Available from NTIS as
PB 236 645).
L86. Metcalf and Eddy, Inc., Palo Alto, California, "Application of Storm Water
Management Model to Selected Chicago Drainage Areas - Phase 1," February 16,
1971; and "Application of EPA Storm Water Management Model to the Chicago
River System - Phase 2," April, 1972.
187. Roesner, Larry A., David F. Kibler and John R. Monser, "Use of Storm Drainage
Models in Urban Planning," National Symposium on Watersheds in Transition,
June 19-22, 1972, Colorado State University, Proceed ings. American Water
Resources Association, Urbana, Illinois, pp. 400-405, 1973.
188. Water Resources Engineers, "The Use of Stormwater Simulation Models in
Developing a Coordinated Wastewater Management Plan for the City of San
Francisco," Summary Report, prepared for Department of Public Works, City and
County of San Francisco, Walnut Creek, California, 31 pp., September, 1974.
189. Torno, Harry C., "Storm Water Management Models," Urban Runoff, Quantity and
Quality. ASCE, New York, N.Y., pp. 82-89, 1975.
190. Sarma, P.B.S., J. W. Delleur and A. R. Rao, "Comparison of Rainfall-Runoff
Models for Urban Areas," J.Hydrology. Vol. 18, pp. 329-347, 1973.
191. Papadakis, Constantine N., and Herbert C. Preul, "Testing of Methods for
Determination of Urban Runoff," J.Hyd.Djv.. ASCE Proc., Vol. 99, No. HY9,
pp. 1319-1335, September, 1973. Authors' closure to discussion: Vol. 101,
No. HY3, pp. 542-544, March, 1975.
192. Keeps, David P., and Russell G. Mein, "Independent Comparison of Three Urban
Runoff Models," J.Hyd.Div.. ASCE Proc., Vol. 100, No. HY7, pp. 995-1009, July,
1974., Authors' closure to discussion: Vol. 102, No. HY1, pp. 93-94, January, 1975
193. Brown, J, W., M. R. Walsh, R. W. McCarley, A. J. Green and H. W. West, "Models
and Methods Applicable to Corps of Engineers Urban Studies," U.S. Army Engineer
Waterways Experiment Station Misc. Paper H-74-8, Vicksburg,.Mississippi, 420 pp.
August, 1974. "'
194. Marsalek, J., T. M. Dick, P. E. Wisner and W. G. Clarke, "Comparative Evaluation
of Three Urban Runoff Models," Water Resources Bulletin. Vol. 11, No. 2,
pp. 306-328, April, 1975.
195. Huber, Wayne C., "Modeling for Storm Water Strategies," APWA Reporter. Vol. 42,
No. 5, pp. 10-14,.May, 1975.
- 82 -
-------
196. Chow, V. T., and B. C. Yen, Urban Storrawater Runoff; Determination of Volumes
and Flowrates, Environmental Protection Technology Series EPA-600/2-76-116,
GPO, Washington, D.C., 240 pp., May, 1976. (Available from NTIS as PB 253 410).
197. Brandstetter, Albin, Assessment of Mathematical Models for Storm and Combined
Sewer Management, Environmental Protection Technology Series EPA-600/2-76-175a,
GPO, Washington, D.C., 510 pp., August, 1976. (Available from NTIS as
PB 259 597).
198. Hossain, A., J. W. Delleur and A. R. Rao, "Evaporation, Infiltration and
Rainfall-Runoff Processes in Urban Watersheds," Water Resources Center
Technical Report No. 41, Purdue University, West Lafayette, Indiana, 63 pp.»
January, 1974. (Available from NTIS as PB 229 642).
199. Systems Control, Inc., Palo Alto, California, "Use of Mathematical Models for
Water Quality Planning," WRIS Technical Bulletin No. 3, Department of Ecology,
State of Washington, 212 pp., June, 1974.
200. Systems Control, Inc., Evaluation of Water Quality Models; A Management Guide
for Planners, Socioeconomic Environmental Studies Series EPA-600/5-76-004, GPO,
Washington, D.C., 176 pp., July, 1976.
201. Pentland, R. L., P. J. Reynolds and Asit K. Biswas, "Water Quality Modeling:
State-of-the-Art," pp. 481-496, Vol. 3, Proceedings of International Symposium
on Modeling Techniques in Water Resources Systems, May 9-12, 1972, Ottawa,
Canada.
202. Canale, Raymond P., Ed., Modeling Biochemical Processes in Aquatic Ecosystems,
Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan, 500 pp., February, 1976.
203. Gordon, R., and M. Spaulding, "A Bibliography of Numerical Models for Tidal
Rivers, Estuaries and Coastal Waters," Marine Technical Report 32, University of
Rhode Island, Narragansett, R.I. 02882, 55 pages, 1974.
204. Hinwood, Jon B., and Ian G. Wallis, "Classification of Models of Tidal Waters,"
J.Hvd.Div., ASCE Proc., Vol. 101, No. HY10, pp. 1315-1331, October, 1975.
Authors' closure to discussion: Vol. 102, No. HY12, pp. 1776-1777, December, 1976.
205. Hinwood, Jon B., and Ian G. Wallis, "Review of Models of Tidal Waters," J.Hyd.Div..
ASCE Proc., Vol. 101, No. HY11, pp. 1405-1421, November, 1975. Author*1 closure
to discussion: Vol. 103, No. HY4, pp. 453-455, April, 1977.
206. Lombardo, P. S., "Critical Review of Currently Available Water Quality Models,"
Hydrocomp, Inc., 101 pp., July, 1973. (Available from NTIS as PB 222 265).
207. Lassiter, Ray R., Modeling Dynamics of Biological and Chemical Components of
Aquatic Ecosystems. Ecological Research Series EPA 660/3-75-012, GPO,
Washington, D.C., 54 pp., May, 1975. (Available from NTIS as PB 241 987).
208. Parker, R. A., and D. Roop, "Survey of Aquatic Ecosystem Models," The Institute
of Ecology, Washington, D.C., 131 pp., 1974.
209. Kramer, W. P., and R. H. Weisberg, "FORTRAN Graphics Programs for Physical
Oceanographic and Time Series Data,1.1 Marine Technical Report 46, University of
Rhode Island Marine Advisory Service, Narragansett, 92 pp., 1976.
- 83 -
-------
210. Hydrologic Engineering Center, U.S. Array Corps of Engineers, "Water Quality
for River-Reservoir Systems," Computer Programs 401-F2-L2100 and -L2100A
Davis, California, 210 pp., July, 1974.
211. Willey, R. G., "Water Quality Evaluation of Aquatic Systems," ASCE Convention
April 14-18, 1975, New Orleans, Louisiana, ASCE Meeting Preprint 2420, 20 pp.*
212. McBean, Edward A., and Daniel P. Loucks, "Planning and Analyzing of Metropolitan
Water Resource Systems," Water Resources and Marine Sciences Center Technical
Report No. 84, Cornell University, Ithaca, New York, 805 pp., June, 1974.
(Available from NTIS as PB 235 257).
213. Evenson, Donald E., and Donald J. Smith, "Mathematical Modeling of San Francisco
Bay," a paper presented at ASCE Annual and National Environmental Engineering
Convention, October 21-25, 1974, Kansas City, Missouri, ASCE Meeting Preprint
2380, 30 pp.
214. Chen, Carl W., "Concepts and Utilities of Ecologic Model," J.San.EnB,rp;.Div.
ASCE Proc., Vol. 95, No. SA5, pp. 1085-1097, October, 1970."'
215. Orlob, Gerald T., "Mathematical Modeling of Estuarial Systems," pp. 78-128
Vol. 1, Proceedings of International Symposium on Modeling Techniques in Wat-or
Resources Systems, May 9-12, 1972, Ottawa, Ontario, Canada. " "
216. Soice, V. Phillip, "Application of a Water Quality Model to the Denver
Metropolitan Area," pp. 1165-1182, Volume II, Symposium on Modeling
ASCE, New York, N.Y., 1975. "~~
217. Benson, Donald J., "A Coordinated Approach to Water and Waste Management,"
Municipality of Metropolitan Seattle, a paper presented October 12, 1973 at
the 40th Annual Convention, Pacific Northwest Pollution Control Association
Vancouver, B.C.
218. Russell, D. L., J. W. Poirot and H. A. Cornell, "A Study of Water Resources
Under King County's Developing Environmental Management Information System "
a paper presented on June 20, 1974, at the AWWA Annual Conference, Boston *Mas
33 pp.
219. Lombardo, Pio S., and Ronald F. Ott, "Water Quality Simulation and Application "
Water Resources Bulletin. Vol. 10, No. 1, pp. 1-9, February, 1974. *
220. Water Resources Engineers and Kramer, Chin and Mayo, "Urban Runoff and Basin
Drainage Study, Green and Cedar River Basins of Washington," Appendix B of final
report "Program Documentation and User's Guide for Urban Storm Drainage
Simulation Models," for Municipality of Metropolitan Seattle and U.S. Army Cor
of Engineers, Seattle District, approx. 320 pp., July, 1974. Pa
221. Donigian, A. S., Jr., and N. H. Crawford, "Modeling Nonpoint Pollution from th
Land Surface," a report for the Environmental Research Laboratory, Athens
Georgia, Hydrocomp, Inc., Palo Alto, California, 279 pp., February, 1976.
222. Hydrologic Engineering Center, U.S. Army Corps of Engineers, "HEC-2, Water
Surface Profiles," Computer Program 723-X6-L202A, Davis, California, February
1972. ^'
- 84 -
-------
223. Soil Conservation Service, Engineering Division, U.S. Department of 'Agriculture,
"Computer Program for Project Formulation Hydrology," Technical Release No. 20,
May, 1965.
224. Eichert, B. S., "Survey of Programs for Water Surface Profiles," J.Hyd.Div..
ASCE Proc., Vol..96, No. HY2, pp. 547-563, February, 1970. Author's closure to
discussion: Vol. 97, No. HY8, pp. 1237-1238, August, 1971.
225. Walesh, Stuart G., "Simulation in Watershed Planning," J.Hyd.Div.. ASCE Proc.,
Vol. 99, No. HY9, pp. 1383-1399, September, 1973.
226. Montgomery-Water Resources Engineers, "Water, Wastewater and Flood Control ,'.
Facilities Planning Model, Summary Report," LaJolla, California, 38 pp., May, 1974.
.
Section 7
227. Stankowski, Stephen J., Population Density as an Indirect Indicator of Urban and
Suburban Land-Surface Modifications. U.S. Geological Survey Professional f&pet
800-B, GPO, Washington, D.C., pp. 219-224, 1972.
228. Graham, P. H., L. S. Costello and H. J. Mallon, "Estimation of Imperviousness
and Specific Curb Length for Forecasting Stormwater Quality and Quantity,"
J.WPCF. Vol. 46, No. 4, pp. 717-725, April, 1974.
229. Gluck, W. R,, and R. H. McCuen, "Estimating Land Use Characteristics for
Hydrologic Models," Water Resources Research. Vol. 11, No. 1, pp. 177-179,
February, 1975.
230. Jackson, Thomas J., and Robert M. Ragan, "Value of Landsat in Urban Water
Resources Planning," J.Wat.Res.Plan.&Mgt.Div.. ASCE Proc., Vol. ,103, No. WR1,
pp. 33-46, May, 1977.
231. Jackson, Thomas J., Robert M. Ragan and William N. Fitch, "Test of Landsat-
Based Urban Hydrologic Modeling," J.Wat.Res.Plan.&Mgt.Div.. ASC* Proc., Vol. 103,
No. WR1, pp. 141-158, May, 1977.
232. McPherson, M. B., and W. J. Schneider, "Problems in Modeling Urban Watersheds,"
Water Resources Research. Vol. 10, No. 3, pp. 434-440, June, 1974.
233. Meier, Peter M., "Land-Use Model for Service Area Projection," J.Kavit«BnRrg.Div..
ASCE Proc., Vol. 102, No. KB1, pp. 71-85, February, 1976.
234 Seader, David, "Runoff Projections for Fairfax County, Virginia," Urban Runoff.
quantity and Quality. ASCE, New York, H.Y., pp. 139-150, 1975.
235. Seader, David, "Evaluation and Planning Techniques," Environmental Design for
Public Projects. Water Resources Publications, Fort.Coll ins, Colorado, pp. 521-
575, 1975.
236* Walesh, Stuart G., Lois A. Kawatski and Paul J. Clavette, "Land Data
Management System for Resource Planning," J.Wat.Res.Plan.& MRt.Div., ASCE
Proc., Vol. 103, No. WR2, pp. 177-192, November, 1977.
- 85 -
-------
237. William Spangle and Associates, et al., Earth-Science Information in Land-lisa
Planning, Guidelines for Earth Scientists and Planners. U.S. Geological
Circular 721, GPO, Washington, D.C., 28 pp., 1976.
238. James, L. Douglas, "Role of Economics in Planning Flood Plain Land Use,"
J.Hyd.Div., ASCE Proc., Vol. 98, No. HY6, pp. 981-992, June, 1972. Author's
closure to discussion: Vol. 97, No. HY1, pp. 172-173, January, 1974.
239. Grigg, Neil S., and Otto J. Helwig, "State-of-the-Art of Estimating Flood
Damage in Urban Areas," Water Resources Bulletin. Vol. 11, No. 2, pp. 379-39Q
April, 1975.
240. Espey, William H., and David E. Winslow, "Urban Flood Frequency Characteristics '
J.Hyd.Div.. ASCE Proc., Vol. 100, No. HY2, pp. 279-293, February, 1974. '
241. U.S. Water Resources Council, Guidelines for Determining Flood Flow Frequency
Bulletin No. 17 of the Hydrology Committee, GPO Stock No. 052-045-00031-2,
Washington, D.C., 197 pp., March, 1976.
242. Jones, D. Earl, Jr., "Basis for Flood Plain Occupancy Decisions," Urban Runoff
Quantity and Quality,.ASCE. New York, N.Y., pp. 235-246, 1975. ~L
243. McPherson, M. B., "Feasibility of the Metropolitan Water Intelligence System
Concept (Integrated Automatic Operational Control)," ASCE Urban Water Resources
Research Program Technical Memorandum No. 15, ASCE,.New York, N.Y., 110 pp.
December, 1971. (Available from NTIS as PB 207 301). '
244. Gibbs, C. V., S. M. Alexander and C. P. Leiser, "System for Regulation of
Combined Sewage Flows," J.San.Engrg.Div.. ASCE Proc., Vol. 98, No. SA6, pp. 951.
972, December, 1972.
245. Municipality of Metropolitan Seattle, Computer Management of a Combined Sewer
System. Environmental Protection Technology Series EPA-670/2-74-022, GPO,
Washington, D.C., 487 pp., July, 1974. (Available from NTIS as PB 235 717).
246. Giessner, W. R., R. T. Cockburn, F. H. Moss and M. E. Noonan, "Planning and
Control of Combined Sewerage Systems," J.Envir.Engrg.Div.. ASCE Proc., Vol. loo-
No. EE4, pp. 1013-1032, August, 1974.. *
247. Wenzel, H. G., Jr., J. tf. Labadie and N. S. Grigg, "Detention Storage Control
Strategy Development," J.Wat.Res.Plan.&Mgt.Div.. ASCE Proc., Vol. 102, No. WRJ.
pp. 117-135, April, 1976. '
248. Labadie, J. W., N. S. Grigg and B. H. Bradford, "Automatic Control of Large-Scale
Combined Sewer Systems," J.Envjr.Engrg.Div.. ASCE Proc., Vol. 101, No. EE1,
pp. 27-39, February, 1975.
249. Corotis, Ross B., "Stochastic Considerations in Thunderstorm Modeling," J.Hyd
ASCE Proc., Vol. 102, No. HY7, pp. 865-879, July, 1976.
250. Changnon, Stanley A., Jr., and Floyd A. Huff, "Chicago Hydrometeorological Area
Project: A Comprehensive New Study of Urban Hydrome.eorology," First Interim
Report, Program of Advanced Environmental Research and Technology, NSF/RANN,
Atmospheric Sciences Center Illinois State Water Survey, Urbana, Illinois
61801, 69 pp., September, 1976.
- 86 -
-------
251. Tucker, L. S., "Raingage Networks of the Largest Cities," ASCE Urban Water
Resources Research Program Technical Memorandum No. 9, ASCE, New York, N.Y.,
90 pp., March 17, 1969. (Available from NTIS as PB 184 704).
252. Friedland, Alan 0., "San Francisco Precipitation and Flow Measuring Network,"
Urban Runoff. Quantity and Quality. ASCE, New York, N.Y., pp. 205-212, 1975.
253. Brownson, L. D., City of Portland, Oregon, "Development of HYDRA (Hydrologic
Data Retrieval and Alarm System)," paper presented at Engineering Foundation
Conference on Instrumentation and Analysis of Urban Stormwater Data, Quantity
and Quality, November 28-Deceraber 3, 1976, Easton, Maryland, 23 pp.
254. Earl, C. T., et al., Melbourne and Metropolitan Board of Works and Dandenong
Valley Authority, "Urban Flood Warning and Watershed Management Advances in
Metropolitan Melbourne," ASCE Urban Water Resources Research Program Technical
Memorandum No. 30, ASCE, New York, N.Y., 72 pp., June, 1976. (Available from
NTIS as PB 259 549).
255. Envirex, Inc., Methodology for the Study of Urban Storm Generated Pollution
and Control. Environmental Protection Technology Series EPA-600/2-76-145, U.S.
EPA, Cincinnati, Ohio, 326 pp., August, 1976. (Available from NTIS as PB 258 743).
256. Henry, J. G., and P. A. Ahern, The Effect of Storage on Storm and Combined Sewers,
Research Program for the Abatement of Municipal Pollution under the Provisions
of the Canada-Ontario Agreement on Great Lakes Water Quality, Research Report
No. 34, Environment Canada, Ottawa, 106 pp., February, 1976.
257. Lott, David L., "Optimization Model for the Design of Urban Flood-Control
Systems," Center for Research on Water Resources Technical Report CRWR-11, The
University of Texas at Austin, 217 pp., November, 1976. (Available from NTIS
as PB 263 490).
258. Hardt, Richard A., and Stephen J. Burges, "Some Consequences of Areawide Runoff
Control Strategies in Urban Watersheds," Department of Civil Engineering Technical
Report No. 48, University of Washington, Seattle, 81 pp., June, 1976. (Available
from NTIS as PB 261 258).
259. Terstriep, M. L., M. L. Voorhees and G. M. Bender, "Conventional Urbanization
and Its Effect on Storm Runoff," Illinois State Water Survey, Urbana, Illinois
61801, 68 pp., August, 1976.
260. Field, Richard, Russell Bowden and Kathy Rozgonyi, "Literature Review Urban
Runoff and Combined Sewer Overflow," J.WPCF. Vol. 49, No. 6, pp. 1095-1104,
June, 1977.
261. Shubinski, Robert P., and William N. Fitch, "Appraisal of Areawide Wastewater
Planning," J.Wat.Res.Plan.&Mgt.Djv.. ASCE Proc., Vol. 103, No. WR1, pp. 63-72,
May, 1977.
262. Westman, Walter E.3 "Problems in Implementing U.S. Water Quality Goals,"
American Scientist. Vol. 65, No. 2, pp. 197-203, March-April, 1977.
263. Jameson, David L., and Harold V. Kibby, Ecosystem Impacts of Urbanization
Assessment Methodology. Ecological Research Series EPA-600/3-76-072,
Environmental Research Laboratory, U.S. EPA, Corvallis, Oregon 97330, 227 pp.,
July, 1976.
- 87 -
-------
ADDENDUM 1
METROPOLITAN INVENTORIES
by M. B. McPherson
Hydrologic Complexity
Figure 1 is a simplified schematic summary of the interconnections of
water occurrences, uses and processes in the urban hydrologic system. Not shown are
water quality aspects, which would be still more complicated. Considerable detail
is hidden from Figure 1. For example, most of our major cities have dozens and even.
hundreds ptf stormwater drainage catchments with cumulative conduit lengths often
exceeding 1,000 miles, all of which are subsumed under "storm drainage" in Figure 1
for Water Quantities
Figure 2' ^ is a graphical representation of water quantities in a
hypothetical urban area of one-million inhabitants. National averages have been used
to calculate the magnitude of the various components. These magnitudes are thus
individually typical, but their combination in Figure 2 is not. Nonetheless, two
important factp are illustrated: the bulk of the supplied water originates outside
of urban areas; and well over half of all the water handled one way or another is in
private hands. This means that there is typically only a partial overlap in local
government water supply source planning and metropolitan general planning, and that
the usual approach of focusing on public water supply can overlook a much larger user
group. The situation for water pollution control is similar, as over half of the
volumes of discharges to receiving waters are from private corporation lands and the
region impacted by water pollution of tea extends well beyond the metropolis of origin
Cooling water withdrawals of thermoelectric power plants are essentially all
self-supplied. While such withdrawals in urban areas have not been segregated from
national totals,'^' they are quite possibly on the order of three times the non-
thermoelectric industrial withdrawals of Figure 2.
Figure 3' depicts a breakdown of that portion of Figure 2 termed "public
water supply withdrawals". Figure 3 is drawn to scale and is again based upon a
composite of national annual averages that resulted from a reconciliation of estimates
for individual components made by various experts. Thus, the distribution of component
amounts for any given community will vary from those shown. For example, there are
cases where industrial use is minuscule and cases where it predominates. Because of
difficulties in quantity accounting generally, there is considerable uncertainty about
the extent of "unaccounted-for" water and "infiltration and inflow". This is the
result of the necessity to arrive at amounts for these two elements by determining th
residuals or leftovers remaining after the magnitudes of all other elements have been
taken into account. Incomplete system quantity measurements and errors in registrati
of measuring devices also contribute to the accounting uncertainties. Finally, °n
difficulties in calculation arise from the fact that water and wastewater jurisdictio
often differ in size. S
Figure 4 depicts a breakdown of that portion of Figure 2 termed "urban
runoff". Whereas Figure 3 is a composite of water and wastewater national annual
- 88 -
-------
00
lO
ATMOSPHERE
IMPORTED
LOW-FLOW
AUGMENTATION.
(PRECIPITATION)
NOUSTRIAL
COOLING
WATER
(SURFACE WATER
OUTFLOW) ~
URBAN LAND SURFACE
(MANIPULATED)
SURFACE
PULATED)
ZONE OF
AERATION
(CAPLLARY
RISE)
(LAWN 8 CHOP
IRRIGATION)
(SEEPAGE AND
SPRING FLOW)
WATER
SYSTEMS
STORM
DRAINAGE
(DIRECT
EFFLUENT
DISCHARGES)
WAS1EWA1ER
SEWERAGE
(SEWER
LEAKAGE OR
ACCRETION)
(EXPORTED
fc INDUSTRIAL
CONSUMPTIVE
USES)
(SUBSURFACE
GROUNDWATER -
OUTFLOW)
(GROUNDWATER
PUMPAGE)
ZONE OF SATURATION
DEEP INTRUSIONS
(»: Interconnected in Combined Systems)
Notes: Water Quality Aspects Not Shown.
Some Connections/Functions Not Typical.
Salt Water Intrusion Not Shown.
FIGURE 1-URBAN HYOROLOGIC SYSTEM
(Adapted From: "Summary of the Hydrologicol Situation in Long Island, N.Y. as a Guide to Water Management Alternatives",
by O.L. Franke and N.E. McClymonds, U.S. Geological Survey Professional taper 627-F, 1972).
-------
MEAN
ANNUAL
PRECIPITATION
(FROM
ATMOSPHERE)
GROUNDWATER
.SURFACE
sWATER
fiu X(FROM
PUBLIC
WATER
SUPPLY
WITHDRAWALS
(1/3 TO
INDUSTRY)
EVAPOTRANSPI RATION
(TO ATMOSPHERE)
URBAN RUNOFF
(TO RECEIVING
WATER BODIES)
(TO GROUNDWATER
AND EVAPOTRANSPIRATION)
PUBLIC
WASTE WATER
(TO RECEIVING
WATER BODIES)
SCALE;
500,000
f TONS/DAY
SURFACE
WATER
(FROM
OUTSIDE)
SELF-
SUPPLIED,
NON-THERMO-
ELECTRIC,
INDUSTRIAL
WATER
WITHDRAWALS
(ACTUAL
USE
IS
MUCH
GREATER)
INDUSTRIAL
WASTE WATER
(TO RECEIVING
WATER BODIES)
(TO EVAPORATION,
EXPORTS AND
GROUNDWATER)
FIGURE 2 -
ANNUAL AVERAGE WATER QUANTITY BALANCE
FOR AN HYPOTHETICAL URBAN AREA OF ONE-
MILLION INHABITANTS, IN TONS PER DAY
(D
- 90 -
-------
PUBLIC
WATER
SUPPLIES
DELIVERED
TO
DISTRIBUTION
SYSTEMS
SPRINKLING
LAUNDRY a DISHES
DOMESTIC
USE
BATHING AND
PERSONAL USES
TOILET FLUSHING
COMMERCIAL USE
PUBLIC USE
INDUSTRIAL
USE
(ACTUAL USE IS
GREATER BECAUSE
OF IN-PLANT
RECYCLING)
PUBLIC
WASTEWATER
COLLECTION
SYSTEMS
PUBLIC
WASTEWATER
TREATMENT
FACILITIES
'UNACCOUNTED-
FOR WATER
CONSUMPTION
LEAKAGE
AND WASTE
INFILTRATION
AND INFLOW
FIGURE 3- COMPOSITE PUBLIC WATER/WASTE WATER INPUT-OUTPUT BALANCE
(3)
-------
FLOW OVER UNSEWERED LAND
I
»0
MEAN
ANNUAL
PRECIPITATION!
(FROM
ATMOSPHERE)
FLOW
OVER
LAND TO
STREET
INLETS ON
SEWERED
CATCH-
MENTS
UNDER-
GROUND
OOLLECTTON
SYSTEMS
(SEWERS
AND
UNDER-
GROUND
STORAGE)
STORM
SEWERS
MAN-MADE
OPEN
DRAINAGE
CHANNELS
AND
IMPROVED
NATURAL
CHANNELS
AND
DIRECT
DISCHARGE
COMBINED
SEWERS
AND
SURFACE
STORAGE
{PONDS,
LAKES,
DELIBERATE
USE OF
NORMALLY
DRY AREAS
RECEIVING
BODIES
WATER
'(STREAMS,
RIVERS,
LAKES,
OCEAN)
FIGURE 4-COMPOSITE URBAN SURFACE WATER RUNOFF INPUT-OUTPUT BALANCE
-------
averages drawn to scale, and there are reservations about how precise some of these
estimates are, we confront much greater uncertainties with urban runoff. The principal
reason is that there is a substantial body of data on metropolitan water and
wastewater volumes and practically none on storm and combined sewers. The indication
in Figure 4 that national annual average flows from storm sewers are about twice as
large as from combined sewers is based on an indirect indication^' of the population
served by .combined sewers versus wastewater sewers. The indication in Figure 4 that
about one-fourth of urban runoff flows over unsewered land is purely conjectural but
within reason. After leaving sewer systems, runoff may be stored in surface
reservoirs and may pass through artificial and improved channels on its way to
receiving waters, and some of it will enter receiving waters directly. There is no
basis for estimating the division of national annual averages among these three modes.
Further, there are complex mixtures of open drainage channels and surface storage in
many metropolitan areas and it might be next to impossible to assign a typical
simplified chain of occurrence (vis-a-vis Figure 3) to the flow through these elements,
even for a particular metropolitan area. The writer's guess is that nationally less
than one-fourth of urban runoff passes through open artificial or improved channels
and that one-tenth or less is subjected to surface storage on its way to receiving
waters. How much that is subjected to surface storage also passes before or afterwards
through open artificial or improved channels is anyone's guess.
Obviously, an input-output balance of urban surface water runoff pollutant
burdens would be even more complicated, and much less is known about urban runoff
quality than on urban runoff quantity.
No attempt has been made to show the interrelations between groundwater and
surface water in Figure 4. While these are taken into account in Figure 1, exactly
how to depict them quantitatively in Figure 4 presents quite an enigma because of the
complexities involved and the lack of data.
Despite the absence of national indications, as described above, responsible
management would appear to call for definitive breakdowns of urban runoff components,
in terms of quantity and quality, in each metropolis. How can we solve a presumed
problem when we have not assessed its magnitude? Such breakdowns would form a part
of a metropolitan water balance inventory, the subject of the remainder of this
Addendum.
Metropolitan Water Balances
Comprehensive planning faces a number of obstacles, including fractionalized
authority of administrative agencies and balkanized local governments in metropolitan
areas ^' specialist approaches taken individually on each water service component, and
uneven advances in knowledge on various water components. More comprehensive management
approaches would integrate or give due consideration to all aspects of water in a
metropolis, together with related environmental, energy and other relevant considerations,
The need for a comprehensive overview becomes more crucial as the complexity of urban
areas increases, caused by such things as population migration and growth, competition
over available energy supplies, rising expectations of urban dwellers in the face of
inflation of local government costs, and greater public awareness of environmental
issues and enlarging demands for public participation.
Water balance inventories, describing the quantity and quality aspects of
the fate of water from its appearance as precipitation through its departure from a
metropolis as runoff and evapotranspiration, have been advocated as a referencing tool
for planning in metropolitan areas.'°' Such inventories provide a basis for better
- 93 -
-------
recognition of the interrelation, interdependence, and interconnection of the element
of the water resources of. a metropolis. "Attempts to study any inventory system ana t-
from the total system does not enable the investigator to see the total impact of
alternatives."(7'
In Sweden, urban water inventories have been approached from two direction
the "outer system," embodying natural processes governed by seasonal and random
variations; and the "inner system," involved with the conveyance and distribution of
water for uses within urban areas. The collective average annual water budget for
urban areas in Sweden is shown in Figure 5'°^ for the outer system and in Figure
for the inner system. Numerical values indicated are volumes in millions of cubic
meters per year. The total urban water budget can be described as the sum of the
outer and inner systems, Figure 5 plus Figure 6, where the sole connections would be
"via combined sewers" (340 x 10 - m /year) and "receiving waters".
Inventory methods used in San Antonio for the quantification of soil water
surface water and groundwater have been described.'^' Long Island, New York, an '
instance of a nearly "closed" hydrologic system covering about 1,400 square miles
has been the subject of extensive water-budget studies.CIO)
Figure 7^ ' is a water balance, on an annual average basis, for metropolit
Chicago (3,714-square miles). All but the smallest of the seven drainage basins
involved receive flows of water from outside the metropolitan area. Variations withi
individual years were also inventoried. "Water in the metropolitan area is constantl
on the move. The three principal components of the water resource (atmospheric
moisture, surface water and groundwater) are interconnected, and water moves in a
continuous cycle between them. Any comprehensive study of water resources, therefor
must first recognize the existence of this cycle and then define the total water '
system, even while realizing that only a limited part of the water moving through th
system is actually available for use. The situation is further complexed in that th
metropolitan area is part of several larger systems which embrace much of the United
States."v11)
Getting the Act Together
In an effort to facilitate the preparation of water balance inventories a
attempt has been made in Figure 1 to describe, in flow-chart form, the urban hydrolo
system,' ' of which stormwater is a part or a subsystem. Figure 8^^) describes the °
urban stormwater runoff subsystem, interpreted from a drainage standpoint, and deals
with water quantities. In contrast, Figure 9'*^' is a schematic diagram of the
relationship between urban activities and water quality, where urban runoff is a ma fo
element. Not many urban runoff control master plans have been developed, fewer still
have been implemented, and rare are the instances where master plans have integrated
quality control with quantity control. All this despite the increasing importance of
their conjunctive planning.
Strategies for improving water quality have been delineated from the
perspective of the urban planner concerned with stormwater control.^5'
References
1. McPherson, M. B., Regional Earth Science Information in Local Water Management-
ASCE, New York, N.Y., 155 pp., July, 1975. ' "'
(Continued)
- 94 -
-------
PRECIPITATION
2820
IMPERMEABLE
SURFACES
/
270 ,rnr
VIA
COMBINED (340
SEWERS v
1970
EVAPOTRANSPIRATION
<1270
PERMEABLE
SURFACES
GROUNDWATER,
HOUSE DRAINS, ETC.
GROUNDWATER
VIA STORM SEWERS 780
\
FIGURE 5 - GENERAL URBAN AREA WATER BUDGET FOR SWEDEN,
OUTER SYSTEM .(8)
WATER
SUPPLY
SOURCES
USE IN WATERWORKS 38
VIA
340 COMBINED
SEWERS
COMBINED
SEWER
OVERFLOWS
35
RECEIVING WATERS
FIGURE 6- GENERAL URBAN AREA WATER BUDGET FOR SWEDEN,
INNER SYSTEM.(8)
- 95 -
-------
4410 (24.59 INCHES/YR)
.EVAPOTRANSPIRATION
5867 (33.18 INCHES/YR)
PRECIPITATION
SURFACE AND SOIL WATER 4088
:;::x::;PUBLlC WATER SUPPLY 1077 (1961)
DIVERSION 646 (1961 )::-::::::
RUNOFF FROM PRECIPITATION 663
s^-tMttttMtr&ttW*^^
p^^^^^^fW**^^wrw*^^w^r^^^Ki
BASE FLOW 673
GROUND-WATER PUMRAGE 182(1961)
f
ANKAKEi: RIVER 2375
..GROUND-WATER
MNFLOW 21
WITHDRAWAL FROM
SUBSURFACE STORAGE40
WATER IN SUBSURFACE STORAGE
FIGURE 7-AVERAGE ANNUAL WATER BALANCE, METROPOLITAN CHICAGO
(MILLION GALLONS PER DAY, 1961)
(ii)
- 96 -
-------
PRECIPITATION
+W44
STREAMFUDW
VHS)
zz
p
il
INITIAL
DRAINAGE
INTERMEDIATE
STORAGE
OUTFLOW
I*
DEPRESSIONS
STREETS
SUMPS
GULCHES
STORM
SEWERS
MAJOR
DRAINAGE
THUNDER
STORMS
TORRENTIAL
RAINS
HEAVY
RAINFALL
AND
SNOWMELT
GULCHES
CREEKS
STREAMS
FLOOD
PLAINS
LAKES
RESERVOIRS
FLOOD
PLAINS
CREEKS
RIVERS
EVAPORATION
AND LOSSES
4 NATURAL
STORAGE
STREAM FLOW
AQUIFER
RECHARGE
STORAGE|
POTABLE
WATER
TREATMENT
DISTRIBUTION
STORM SEWER
WASTE
WATER
TREATMENT
SUBSURFACE FLOW
OTHER URBAN
SUBSYSTEMS
BUILDING AND
DEVELOPMENT
PI'^LIC HEALTH
AND SAFETY
TRANSPORTATION
SOLID WASTE
RECREATION/
OPEN SPACE
SANITARY WASTE
IRRIGATION
WATER SUPPLY
FIGURE 8-THE URBAN STORM DRAINAGE SUBSYSTEM
(13)
-------
LAND USE DETERMINANTS
Population and
Economic
Growth
Private
Market
Forces
Land Use
Regulation
Public
Services
and Facilities
Natural
Features
ft Constraints
Land Use
Pattern
Form
Density
Use Mix
Open Space
Land Conversion Rate
WATER QUALITY IMPACTS OF LAND USE
Withdrawal
of Surface
and Ground
Water
Loss of
Natural
Ground Cover
And Disturbance
of Environmentally
Sensitive Areas
Net Change
of Stream
Flow
Construction
Related
Sediment
Loads
Silvaculture,
Mining ana
Agriculture
Runoff
Loads
Urban
Runoff
Loads
Stream
Enlargement
and Erosion
Domestic
Waste
Loads
Industrial
Waste
Loads
Total
Non-Point
Load
Water
Quality
Standardsd
Effluent
Limitations
Total
Point
Load
Water Quality
Management
Practices
Quantity
and Quality
of Receiving
Water Prior
To discharge
Quantity
and Quality
of Discharge
Land
Disposal
Wastewoter
Reuse,etc.
Resulting
Quality of
Receiving Waters
FIGURE 9- SCHEMATIC DIAGRAM OF THE LAND-USE/WATER-QUALITY
RELATIONSHIP04*
- 98 -
-------
2. Tucker, L. S., J. Millan and W. W. Burt, "Metropolitan Industrial Water Use,"
ASCE Urban Water Resources Research Program Technical Memorandum No. 16, ASCE,
New York, N.Y., 45 pp., May, 1972. (Available from NTIS as PB 212 578).
3. McPherson, M. B., "Household Water Use," ASCE Urban Water Resources Research
Program Technical Memorandum No. 28, ASCE, New York, N.Y.S 25 pp. + 84 p. append.,
January, 1976, (Available from NTIS as PB 250 879).
4. Sullivan, Richard H., "Inventory of Combined Sewer Facilities," Civil Engineering,
Vol. 38, No. 11, pp. 52-53, November, 1968.
5. McPherson, M. B., "Urban Water Resources," EOS, Transactions of the AGU, Vol. 57,
No. 11, pp. 798-806, November, 1976.
6. McPherson, M. B., "Need for Metropolitan Water Balance Inventories," J.Hyd.Div.,
ASCE Proc., Vol. 99, No. HY10, pp. 1837-1848, October, 1973. Author's closure
to discussion: Vol. 101, No. HY4, p. 409, April, 1975.
7. Koch, C. T., "Systems Description for Urban Water Resources," pp. 19-41,
Proceedings, 16th Annual Conference on Water for Texas, "Urban Water Resource
Planning and Management," Texas A. and M. University, College Station, Texas,
200 pp., September, 1971. (Available from NTIS as PB 210 325).
8. Lindh, Gunnar, "Urban Hydrological Modeling and Catchment Research in Sweden,"
ASCE Urban Water Resources Research Program Technical Memorandum No. IHP-7,
ASCE, New York, N.Y., 33 pp., October, 1976. (Available from NTIS as PB 267 523).
9. Koch, C. T., "Methodology for Managing Resources," J.AWWA, Vol. 64, No. 2, pp. 122-
126, February, 1972.
10. Cohen, Philip, 0. L. Franke and B. L. Foxworthy, An Atlas of Long Island's Water
Resources, New York State Water Resources Commission, Bulletin 62, Albany, N.Y.,
117 pp., 1968.
11. Sheaffer, John R., and Arthur T. Zeizel, "The Water Resource in Northeastern
Illinois: Planning Its Use," Technical Report No. 4, Northeastern Illinois
Metropolitan Area Planning Commission, Chicago, Illinois, 182 pp., June, 1966.
12 Task Committee Report, "Aspects of Hydrological Effects of Urbanization,"
J.Hvd.Div.. ASCE Proc., Vol. 101, No. HY5, pp. 449-468, May, 1975.
13. Rice, Leonard, "Sediment Considerations in Urban Drainage," ASCE National Water
Resources Engineering Meeting, January 24-28, 1972, Atlanta, Georgia, ASCE
Meeting Preprint 1588, 14 pp.
14. Shubinskia Robert P., and Steven N. Nelson, "Effects of Urbanization on Water
Quality," ASCE Urban Water Resources Research Program Technical Memorandum No. 26,
ASCE, New York, N.Y., 34 pp., March, 1975. (Available from NTIS as PB 242 297).
15. Shubinski, Robert P., "Concepts of Urban Runoff Control," Section 5 in "Management
of Urban Storm Runoff," ASCE Urban Water Resources Research Program Technical
Memorandum No. 24, ASCE, New York, N.Y., 92 pp., May, 1974. (Available from
NTIS as PB 234 316).
- 99 -
-------
ADDENDUM 2
THE DESIGN STORM CONCEPT
by M. B. .McPherson
Introduction
Historically, urban areas have been drained by underground systems of
sewers that were intentionally designed to remove stormwater as rapidly as possible
from occupied areas. Discharges from conventional storm drainage sewer facilities
and flood-plain intrusion by structures both tend to aggravate flooding, and
thereby jointly tend to raise the potential for stream flooding damages. The
advantages of local detention storage in lieu of the traditional rapid removal of
storm flows has long been recognized^' and there is evidence that the usage of such
storage is on the rise.' ' However, local deteTition storage has been only
occasionally employed as part of overall flood mitigation, as in Denver^' and
Fairfax County, Virginia.)
Detention storage is recognized as one of the principal means for abatement
of pollution from stormwater discharges and combined sewage overflows.^) There are
opportunities in new land development to incorporate detention storage at the ground
surface.(°' However, for existing drainage systems, there may be few opportunities
to add detention storage except underground, for both combined sewer and separate
storm sewer systems. Because abatement of pollution from the dispersed sources
served by drainage systems began with a focus on combined systems, our knowledge on
storage requirements for them is greater. Recognized very early was the need for
some form of automatic control,''' because of system complexity and the need to
manipulate flows in order to insure their containment as a means for reducing
overflows.
System Performance
When rapid removal of stormwater was the primary objective, hydrological
considerations were mostly restricted to the sizing of conduits. Retardation of
flows and attenuation of their peaks is another matter, as is containment of flows
and their diversion for treatment to abate pollution. Whereas simple concepts
sufficed for conduit sizing, their perpetuation in more complex applications
inevitably leads to lowered reliability.
An attempt is made herein to indicate how precipitation data can best be
employed in planning and design of stormwater facilities under emerging imperatives
Responsibilities of local government operating agencies were formerly restricted1 to*
the management of underground conduit systems that had been deliberately designed to
be overloaded upstream of their outlets rather often, such as once in five years on
the average, with little regard for the impact of system flows on receiving waters.
Thus, the sole criterion of performance was how effectively the land had been
drained. Once containment of flows and their pollutant burdens become additional
objectives, performance criteria enlarge to include the meeting of regulations and
statutes. As a result, public officials will no longer be able to depend only on
the relatively obvious incidence of system overloading evidenced by flooded
basements and streets. They will also have to show concrete evidence of compliance
acceptable to regulatory agencies. This means that they will have to know when
- 100 -
-------
how much and how often allowable limits have been exceeded. Regulatory limits
necessarily have to be founded on some level of frequency, or probability, such as
containment on the average of all but one combined sewer overflow every five years,
or of all but a certain volume of stormwater over a season, or that would permit no
more than a certain maximum pollutant discharge per storm.
The only way an event frequency can be determined is by referring to a
reasonably large series of prior events. A long history of overflow or storm
discharge amounts is almost never available and, even if it were, it would not be
suited for systems revised with new storage and controls. Thus, recourse is made
to simulation of system performance via some sort of calculation, such as the use
of hydrological models. That is, a historical record of rainfall is transformed by
calculation into a simulated historical record of stormwater flow and possibly
quality. Simulation of an existing system thereby yields a synthesized historical
record of pre-modification conditions. System revisions are then targeted to
contain all but some relatively rare events from the synthesized performance record,
as prescribed by regulations. But how will compliance be judged after a system has
been modified to meet a regulation? Of course, monitoring all overflow/discharge
points by means of rainfall-runoff-quality instrumentation would indicate when
violations had occurred, but would not indicate why. Further, reality dictates
that from an economical standpoint, only a sample of overflow/discharge points could
probably be monitored, and simulation of post-modification performance can be employed
to interpret sparsely monitored systems and explain whether or not violations have
occurred and why.
At least a decade ago there was widespread acceptance of the need for
complete storm hydrograph simulation if major system modifications were contemplated,
particularly if new storage was involved. However, as we enter an age of regulation
another need arises, for simulation of the performance of systems modified to meet
regulations. Unfortunately, concepts that were adequate for sizing conduits have
been carried into the sizing of detention storage and now threaten to be perpetuated
in the future testing of compliance performance in flooding relief and pollution
abatement facilities planned for construction. For this reason, it is necessary to
review concepts that have been used for decades, to show why they are not suited for
investigating performance, either expected or experienced. These now outmoded
concepts originated in connection with the sizing of conduits.
Simplistic Traditions
The overwhelmingly used technique for sizing combined and storm sewers is
known as the "rational method".(8) its limitations have been discussed at length
elsewhere,(9^ and present purposes are served by concentrating on its fundamental
premise.
Intuition, theory and laboratory tests indicate that when a rain of a
constant intensity falls on a catchment, the outflow from the catchment will
ultimately rise to a maximum rate, which would be sustained as long as the constant
rainfall intensity continued thereafter. The time required to reach the maximum
outflow rate (to reach a state of equilibrium) is commonly called the "time of
concentration". The equilibrium discharge is merely a fixed fraction of the
constant rainfall intensity in the rational method, with the ratio of equilibrium
discharge to rainfall intensity defined by a constant that is usually interpreted in
terms of type of land use. Because for a given catchment, equilibrium discharge is
a f^ed multiple of input constant rainfall intensity, it follows without reservation
- 101 -
-------
that the frequency* of an equilibrium discharge is identical with the frequency of
its associated constant rainfall intensity.
Storm and combined sewers are usually designed to flow full without
surcharging at some selected frequency, such as on an average of once in five years
or once in ten years. Thus, the only type of rainfall information needed for use
of the rational method is an array of intensities for each of several durations from
which various frequencies can be interpolated. The result is called "intensity-
duration- frequency curves". Such curves have been developed in a number of
municipalities and advantage has been taken of the spatial persistence of curve
values in the preparation of national maps that contain lines of equal rainfall for
given durations and frequencies, such as by Yarnell,'^) Hathaway(Il) and
Hershfield.(12) Certain characteristics of such information must be reviewed next
in order to demonstrate some rather severe limitations.
Intensity-Duration-Frequency Relations
The almost universal data source is the U.S. Weather Service, from its
first-order stations in principal cities.
In principle, one would search the recorded rainfall for each storm at a
given station for the largest catch over a particular duration. Because engineers
have been more interested in the rarer events, and to reduce the data processing and
reporting burden, the USWS processes and reports data for only those storms where at
least some rainfall depths exceed a specified threshold level. The level is
sufficiently low to include 1-year frequency and somewhat lower values. Data from
storms having depths above the threshold are termed "excessive precipitation".
Because for a given duration the approximately largest depth of rainfall is reported
it is therefore the maximum amount for that duration for that storm at that gage. '
Because there is considerable variability in rainfall over time, the selected amount
when expressed as an intensity, is necessarily the average intensity for the given
duration. Hence, we will hereinafter use the term maximum average when describing
excessive precipitation.
In deriving intensity-duration-frequency relations, rainfall values for
each duration are regarded independently from other durations, the first step being a.
separate ranking of values for each duration in descending order of size. A
mathematical fit is made to the array of depths or intensities for each duration and
the line of best fit values are computed. In the last step, fitted values for each
duration for a specified mean recurrence interval (or frequency) are plotted with
intensity as ordinate and duration as abscissa; and smooth fitted lines are drawn
through points of equal frequency.
Starting in 1936 the USWS published maximum average excessive precipitation
for various durations, presumably to facilitate rational method applications, but
which makes it necessary to refer to the National Climatic Center, NOAA Environmental
Data Service, Asheville, N.C., for the true sequence of occurrences. From 1895
through 1935 the USWS published maximum average excessive precipitation as it had
accumulated over time, mostly for a 5-minute interval. The report of a study of
frequency analysis methods includes a listing of such data for 1913-1935 for
Chicago,(13) USed in the succeeding discussion.
*: (The terms "frequency," "return period" and "recurrence interval" are used
interchangeably in drainage practice, as they will be here. Frequency is the
reciprocal of probability).
- 102 -
-------
Listed in Table 1 are the top 14 values of maximum average excessive
precipitation for Chicago, 1913-1935 Casual inspection of Table 1 will reveal the
fact that maximum average depths for a given average return period are not necessarily
from the same storm, e.g. see Rank No. 5. Inspection of actual records of
accumulation would reveal that 5-minute maximum average depths did not necessarily
occur in the first five minutes, and so on for other durations. Also, the record
includes storms that lasted as little as five minutes, with a majority of the 130
storms reported lasting less than 35-minutes, and with very few lasting more than an
hour.
Further, in order to insure a complete data matrix for regression analysis
to develop frequency curves, a dummy value is inserted for all durations beyond the
cessation of excessive precipitation for a given storm. For example, for the storm
of 7/16/14, to be examined later, 1.61-inches had accumulated at the end of the
storm, which lasted 35-minutes, and for frequency analysis purposes a fictitious
value of 1.61-inches would be inserted for durations from 40 to at least 120 minutes
for that storm prior to the ranking of maximum averages. The insertion of dummy
valufes (called "extended duration" values) is consistent with the underlying concept
of the rational method, but further distorts intensity-duration-frequency curves from
the actual history of occurrences. This distortion is illustrated in Figure 1, where
maximum average values of rainfall for a T£ (the average return period) of 4.6-years
and 5.8-years from Table 1 have been converted into intensities. Compared is a
"5-year" curve for Chicago attributed to Eltinge and Towne.^1^ The falloff of
plotted points beyond a duration of 40-minutes, compared with the uninterrupted
smooth curve, reflects the exclusive use of actual data for the former and the
traditional use of extended duration values for the latter. Note that the plotted
points in Figure 1 cease at 80-minutes because only one storm from the record used
had lasted any longer but was a rarer value.
In referring to a frequency curve in the context of the rational method,
some engineers erroneously use the term "storm frequency". For example, values taken
from a 5-year curve or the resulting computed flow rate are often stated to be for
"a five year storm". Considering that a given frequency curve can represent values
from different storms, in a time sequence different than actually occurred, and might
include non-existent dummy values, it is obvious that any reference to the term
"storm" in rational method applications is at least misleading and is certainly
technically imprecise.
Further, it should be noted that the time of concentration varies from
point to point in a catchment area. Thus, using a given frequency curve, various
portions of a drainage area may be designed on the basis of pieces of different
storms that might have occurred several years apart. Hence, the presentation of
water surface profiles or hydraulic gradients from upstream extremities to the
outlet for "design conditions" can be greatly misleading.
In summary, the term "storm" should never be used in connection with the
rational method, whereas the term "rainfall" is adequately vague and innocuous. In
fact, an outstanding limitation of the rational method stems from its complete
independence from storm pattern. To reiterate, the maximums of the several
durations from a given storm record as used in the rational method are not
necessarily in their original sequential order; and the resulting tabulation of
maximuras ordered by size of duration may bear little resemblance to the original
storm pattern.
- 103 -
-------
TABLE 1 - HANKING BY MAXIMUM DEPTH FOR VARIOUS DURATIONS, CHICAGO, 1913-1935
BANK
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AVERAGE
RETURN
PERIOD,
YEARS
23.0
11.5
7.7
5.8
4.6
3.8
3.3
2.9
2.6
2.3
2.1
1.9
1.8
1.6
MAXIMUM DEPTH
5-min.
0.55
(6/26/32)
0.53
(6/29/20)
0.51
(8/11/31)
0.50
(6/20/28)
0.48
(8/8/27)
0.45
(8/11/23)
0.42
(8/27/31)
0.41
(10/5/19)
0.41
(8/13/25)
0.41
(8/2/33)
0.39
(9/5/20)
0.39
(7/7/21)
0.39
(7/20/24)
0.38
(5/18/26)
10-min.
0.96
(8/11/31)
0.94
(6/20/32)
0.88
(6/20/28)
0.80
(8/11/23)
0.80
(7/2/33)
0.76
(7/7/21)
0.74
(10/5/19)
0.74
(8/27/31)
0.71
(8/2/35)
0.68
(8/13/25)
0.68
(5/18/26)
0.68
(6/29/33)
0.67
(6/18/35)
0.66
(7/16/14)
15-min.
1.22
(8/11/31)
1.16
(6/20/28)
1.16
(6/26/32)
1.12
(8/11/23)
1.08
(7/7/21)
1.00
(7/2/33)
0.96
(8/27/31)
0.93
(10/5/19)
0.87
(6/13/26)
0.87
(8/2/35)
0.86
(7/20/24)
0.86
(6/18/35)
0.84
(8/13/25)
0.83
(8/9/14)
20-min.
1.39
(8/11/23)
1.39
(8/11/31)
1.38
(6/20/28)
1.35
(7/7/21)
1.21
(8/27/31)
1.16
(7/2/33)
1.13
(6/13/26)
1.10
(10/5/19)
1.05
(7/20/24)
1.05
(8/13/25)
0.97
(7/16/14)
0.96
(6/29/33)
0.96
(8/2/35)
0.95
(5/18/26)
2 5-min.
1.58
(6/20/28)
1.49
(8/11/23)
1.45
(7/7/21)
1.34
(6/13/26)
1.29
(10/5/19)
1.27
(8/27/31)
1.25
(7/20/24)
1.21
(7/16/14)
1.21
(8/13/25)
1.18
(6/29/33)
1.07
(8/15/34)
1.05
(9/5/20)
1.05
(7/1/27)
1.04
(5/18/26)
IN INCHES
30-min.
1.70
( 6/20/28 )_
1.61
(8/11/23)
1.55
(6/13/26)
1.49
(7/7/21)
1.48
(7/16/14)
1.44
(10/5/19)
1.43
(7/20/24)
1.38
(6/29/33)
1.31
(8/13/25)
1.20
(8/29/28)
1.19
(8/15/34)
1.12
(9/5/20)
1.11
(7/1/27)
1.05
(8/2/35)
FOR STATED
35-min.
1.86
(6/20/28)
1.77
(6/13/26)
1.71
(8/11/23)
1.61
(7/16/14)
1.59
(7/20/24)
1.53
(10/5/19)
1.53
(6/29/33)
1.38
(8/13/25)
1.30
(8/29/28)
1.21
(8/15/34)
1.17
(7/19/31)
1.16
(9/5/20)
1.09
(6/25/26)
1.05
(7/19/16)
DURATIONS;
40-min.
1.92
(6/20/28)
1.85
(6/13/26)
1.81
(6/29/33)
1.77
(8/11/23)
1.75
(7/20/24)
1.24
(7/19/31)
1.23
(8/15/34)
1.22
(9/5/20)
1.13
(6/24/25)
1.10
(9/10/22)
1.08
(7/19/16)
1.08
(8/5/24)
1.06
(8/2/35)
1.03
(6/18/35)
AND DATE OF RAINFALL
45-min.
2.04
(6/20/28)
2.03
(6/29/33)
1.80
(8/11/23)
1.31
(7/19/31)
1.25
(6/24/25)
1.14
(9/10/22)
1.12
(8/5/24)
1.06
(7/11/22)
1.05
50-min.
2.06
(6/20/28)
1.85
(8/11/23)
1.40
(7/19/31)
1.23
(9/10/22)
1.17
(8/10-11/31)
1.15
(8/5/24)
1.12
(7/11/22)
1.02
(6/23/31)
0.87
(8/10-11/31) (6/15/25)
1.00
(6/23/31)
0.98
(6/12/15)
0.82
(6/15/25)
0.74
(9/17/27)
0.73
(4/10/22)
0.80
(9/17/27)
0.79
(3/31/29)
0.78
(4/10/22)
0.78
(6/24/28)
0.74
(8/19/21)
60-min. 80-min. 100-min. 120-min.
2.30 2.30 1.36 1.53
(6/20/28) (8/11/23) (6/23/31) (6/23/31)
2.02 1.55
(8/11/23) (9/10/22)
1.53 1.54
(9/10/22) (8/5/24)
1.27 1.45
(8/10-ll/31)(8/19/21)
1.22 1.23
(8/5/24) (3/31/29)
1.21 1.22
(7/11/22) (6/23/31)
1.13 0.96
(6/23/31) (7/5/30)
1.12
(8/19/21)
0.99
(6/24/28)
0.97
(9/17/27)
0.89
(3/31/29)
0.82
(7/5/30)
0.81
(9/23/26)
-------
. 3
i- CO
g 2
. UJ
<
o:
0
FIGURE 1
TE = 5.8-YEARS (1913-1935), FROM TABLE 1
TE = 4.6-YEARS (1913-1935), FROM TABLE 1
,TC « 5-YEARS (ELTINGE AND TOWNE)
I
I
I
20
40
60 80 100
DURATION, MINUTES
120
140
160
180
- INTENSITY-DURATION-FREQUENCY COMPARISONS, CHICAGO DATA
-------
5
30-35
20-25
10
30-40
20-30
15
20-35
20-35
20
20-40
10-30
25
15-40
5-30
30
10-40
0-30
35
5-40
0-35
~~40 ~~
0-40
Volumetric Considerations
Because the sizing, deployment and operation of detention storage facilities
is determined by the time patterns and amounts of rainfalls selected, the use of
actual precipitation records is greatly preferred, for reasons already given. Listed
in Table 2 are the top 14 storms in terms of total rainfall depth for Chicago, 1913-
1935, for the same record from which Table 1 was derived. Shown are the recorded
depths for successive intervals of time. It may be noted that the larger amounts of
rain occur early, late, and in the middle of storms, and some have double peaks
such as Rank No. 5.
In order to illustrate the mixture of individual frequencies of separate
parts of storms, the actual depths for two storms given in Table 2 have been traced
through Table 1 (and its extension, not shown), and the result is given in Table 3
Note the wide disparity in individual storm component frequencies. Also to be
considered is the distortion in timing of the maximum average intensities:
Occurrence During Storm of Maximum
Aver;
Rank Storm
4 6/13/26
5 7/16/14
which may be checked by inspection of Table 2.
Hopefully, some indication has been presented of the folly of attempting
to define a design storm by using concepts applicable solely to the rational method
Another Example
As a companion to Table 2, listed in Table 4 are the top 14 storms in terms
of total rainfall depth for Philadelphia, 1913-1935. What has been said about the
Chicago data is generally applicable for the Philadelphia data. In addition, because
the writer has been supplied a copy of the complete USWS record by the Philadelphia
Water Department, certain features from that record can be cited
. For the storm of August 7, 1921, 0.65-in. fell between 3:40 p.m. and 5:05 p.m.
0.81-in. between 6:10 p.m. and 7:00 p.m., and the 1.91-in. indicated in Table 4
followed between 10:15 p.m. and 11:40 p.m.
. For the storm of May 27, 1918, 1.28-in. fell between 7:38 p.m. and 8:45 p.m.,
followed by a 0.25-in., and then the 1.62-in. indicated in Table 4 followed betwe
10:18 p.m. and 11:33 p.m. Sa
. For the storm of June 27, 1914, the 1.60-in. indicated in Table 4 fell between
3:21 p.m. and 4:15 p.m. and was followed by 0.56-in. between 5:40 p.m. and 8:10
. Note that in Table 4 the storm of July 10, 1931 with a total depth of 1.57-in. wa
preceded only three days earlier by a storm with a total depth of 1.43-inches
Also, 0.91-in. fell in a little over half an hour on July 14, 1931, followed bv
1.03-in. on July 18, 1931, which fell mostly in a 26-minute period.
- 106 -
-------
TABLE 2 - BANKING BY DEPTH OF STORM, CHICAGO, 1913-1935
RANKING BY MAGNITUDE
OF TOTAL STORM DEPTH
RANK
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AVERAGE
RETURN
PERIOD,
YEARS
23.0
11.5
7.7
5.8
4.6
3.8
3.3
2.9
2.6
2.3
2.1
1.9
1.8
1.6
TOTAL
STORM
DEPTH,
INCHES
2.30
(6/20/28)
2.30
(8/11/23)
2.03
(6/29/33)
1.85
(6/13/26)
1.61
(7/16/14)
1.55
(9/10/22)
1.54
(8/5/24)
1.53
(10/5/19)
1.53
(6/23/31)
1.49
(7/7/21)
1.45
(8/19/21)
1.40
(7/19/31)
1.39
(8/11/31)
1.38
(8/13/25)
-. 1 ! "
0-5
0.22
0.10
0.31
0.08
0.27
0.08
0.12
0.09
0.18
0.32
0.20
0.15
0.26
0.10
^^VH^H
5-10
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.50
.27
.37
.22
.24
.23
.12
.19
.23
.39
.16
.21
.45
.21
10-15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.38
.35
.12
.21
.19
.08
.23
.33
.26
.37
.22
.32
.51
.16
RECORDED
15-20
0.28
0.45
0.05
0.21
0.12
0.04
0.10
0.41
0.09
0.27
0
0.23
0.17
0.27
DEPTH
20-25
0.20
0.32
0.25
0.31
0.33
0.16
0.12
0.17
0.12
0.10
0
0.07
0.41
IN INCHES FOR STATED TIME INTERVAL IN MINUTES
25-30
0.12
0.09
0.28
0.23
0.33
0.15
0.11
0.19
0.08
0.04
0
0.06
0.16
30-35 35-40 40-45 45-50 50-60 60-80 80-100 100-120
0.16 0.06 0.12 0.02 0.24
0.13 0.06 0.03 0.05 0.17 0.28
0.15 0.28 0.22
0.33 0.26
0.13
0.13 0.01 0 0.21 0.44 0.02
0.08 0.20 0.04 0.03 0.07 0.32
0.15
0.03 0.01 0 0.02 0.11 0.09 0.14 0.17
0 0 0.07 0.09 0.38 0.33
0.13 0.07 0.07 0.09
0.07
__
o
H
-------
TABLE 3 - RETURN PERIODS FOR MAXIMUM DEPTHS
OF TWO STORMS, CHICAGO, 1913-1935
RANK
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
AVERAGE FROM
RETURN RANKING
PERIOD, BY DEPTH
YEARS OF STORM
23.0
11.5
7.7
5.8 6/13/26
4.6 7/16/14
3.8
3.3
2.9
2.6
2.3
2.1
1.9
1.8
1.6
1.5
1.4
1.3
1.3
1.2
1.2
1.1
1.0
1.0
1.0
0.9
0.9
0.8
OCCURRENCE OF MAXIMUM DEPTHS FOR GIVEN STORMS
5-min. 10-min. 15-min. 20-roin. 25-rln. 30-min. 35-min. 40-min.
6/13/26 6/13/26
6/13/26
6/13/26 7/16/14
7/16/14
6/13/26
7/16/14
6/13/26
7/16/14
7/16/14
7/16/14
6/13/26
6/13/26
7/16/14
- 108 -
-------
TABLE 4 - BANKING BY DEPTH OF STORM, PHILADELPHIA, 1913-1935
RANKING BY MAGNITUDE
OF TOTAL STORM DEPTH
RANK
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AVERAGE
RETURN
PERIOD,
YEARS
23.0
11.5
7.7
5.8
4.6
3.8
3.3
2.9
2.6
2.3
2.1
1.9
1.8
1.6
TOTAL
STORM
DEPTH,
INCHES
2.70
(6/26/30)
2.09
(8/16/17)
1.91
(8/7/21)
1.82
(7/15/26)
1.82
(4/21/27)
1.76
(7/3-4/26)
1.74
(5/24/33)
1.63
(7/13/19)
1.62
(5/27/18)
1.60
(6/27/14)
1.57
(7/10/31)
1.48
(7/7/31)
1.43
(6/20/13)
1.39
(8/28-29/23)
RECORDED DEPTH IN INCHES FOR STATED TIME INTERVAL IN MINUTES
0-5
0.38
0.41
0.12
0.06
0.07
0.12
0.38
0.06
0.09
0.15
0.25
0.32
0.34
0.08
5-10
0.
0.
0.
Oc
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
32
30
02
04
08
11
49
05
08
22
18
40
19
10
10-15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
--
.32
.37
.08
.05
.13
.18
.29
.14
.03
.21
.20
.46
.20
.11
15-20
0.29
0.23
0.24
0.05
0.09
0.15
0.23
0.28
0.04
0.21
0.16
0.30
0.22
0.13
20-25
0.08
0.49
0.24
0.17
0.08
0.15
0.12
0.04
0.11
0.14
0.16
0.38
0.12
25-30
0.02
0.27
0.17
0.14
0.11
0.09
0.03
0.02
0.23
0.09
0.03
0.10
0.11
30-35 35-40 40-45 45-50 50-60 60-80 80-100 100-120
0.04 0.05 0.10 0.08 0.33 0.69
0.02
0.18 0.04 0.02 0.02 0.06 0.62 0.10
0.13 0.26 0.24 0.13 0.15 0.40
0.07 0.01 0.20 0.24 0.31 0.20 0.23
0.08 0.16 0.15 0.24 0.28 0.05
0.02 0.01 0 0.02 0.05 0.06 0.02 0.02
0.20 0.36 0.23 0.18 0.07
0.20 0.11 0.15 0.08 0.05 0.45
0.01 0.02 0.22 0.20 0.13
0.06 0.02 0.10 0.21 0.20
0.08 0.12 0.15 0.15 0.24
-------
On August 3, 1898, 5.43-inches were recorded for the period between 10:50 a.m.
and 12:35 p.m., only 105 minutes; and on September 14 and 15, 1904, an equal
amount was recorded, 3865-in. in 149 minutes preceded by almost 3-in. which had
ended only about seven hours earlier and most of which occurred in a period of
only one hour.
These examples should give a clear indication of why synthetic storms can
be a liability in detention storage evaluations. Only by referring to actual records
as a basis for simulating performance can multiple events be realistically taken
into account. Additionally, a synthetic storm approach neglects surficial and
subsurface effects of prior storms on the magnitude of runoff and pollutants from a
given storm (commonly called antecedent conditions).
Synthetic Storms
Graphed in Figure 2 is a "5-year" synthetic storm developed in Chicago.(14)
Also graphed thereon are the two storms from Table 2 with average return periods of
4.6-years and 5.8-years on the basis of total storm depth. It is difficult to decide
just how the two actual storms should be placed with respect to time for comparison
and the positions shown are quite arbitrary. The extension of the synthetic storm
to 180-minutes and its overstated total depth (2.28-inches versus 1.61-inches and
1.85-inches for the actual storms) reflect its origin as an intensity-duration-
frequency curve containing extended duration values described earlier.
The method of storm synthesis for Chicago reported in 1957 was criticized
on the grounds that it retained too many of the fallacies and empiricisms inherent
in the rational method to recommend its principle for general use by others. * *'
The authors(14) were asked(l^) why could not actual storms be used instead, in the
later development of general design criteria.(16) «£0 be fair, all this occurred before
high-speed, general-purpose computers were readily available and the processing of a
number of storms rather than only one was simply not practicable at that time.
However, when it came time to prepare a master plan for combined sewer overflow
abatement in metropolitan Chicago that incorporated extensive underground storage (17,18)
hourly precipitation records for 1949 through 1969 for twenty raingages in the
metropolitan area were used in conjunction with over a score of computer models as a
means for determining system component sizes, and necessary waterway improvements and
waterway quality for each plan alternative.(19)
Perhaps because the Chicago synthetic storm was featured in a widely used
handbook ^ that had been first issued in 1960 and last in 1969, there have been a
number of reports and published papers that either adopt the method or allude to its
use. This paper has attempted to indicate why synthetic storms are a poor choice for
simulating system performance, particularly where detention storage of any significance
is involved.
Availability of Chicago data makes possible the illustration of another
synthetic storm procedure. A computer model has been developed for storm sewer design
called ILLUDAS (Illinois Urban Drainage Area Simulator),(20' that had been described
earlier in an ASCE journal.(2i) A design storm procedure for Illinois sites is
recommended in the users' manual,(20) Graphed in Figure 3 are the two storms from
Table 2 with average return periods of 4.6-years and 5.8-years on the basis of total
storm depth. Also graphed are "5-year" synthetic storms for ILLUDAS criteria, setting
the total duration in each case equal to the duration of the actual storm compared.
- 110 -
-------
8
7
^
i 4
O
I
tr
s
t-1 »
Eeo 3
1 1
O
> 2
i-
co
z
LJ
1-
2 I
0
1 T 1 1 1 1 I >
I
ll
M^« j^J
I STORM OF JULY 16
TE =4.6-YEARS(19
ON BASIS OF TOT
DEPTH (1.61-IN.) v
__
-r^'^i
) 20 40
, 1914,
13-1935)
'AL
h
" :
1
/
/
f
1
nrVr
i i
1 \
1
[ »«
^
STORM OF JUNE 13, 1926,
: ': x TE = 5.8- YEARS (1913-1935)
: xv^ *
: ^ ON BASIS OF TOTAL
! i DEPTH (1.85-IN.)
!l"
\
\
\
\
\
\
"~n v
\ M5-YEAR" SYNTHETIC STORM
/ \ y (TOTAL DEPTH = 2.28-INJ
r*/ ^^
^Z ^^*
^'^
^^^
^
^~ 60 80 100 120 140 160 18O
TIME, MINUTES
FIGURE 2-ACTUAL VERSUS SYNTHETIC STORM PATTERNS, CHICAGO
-------
U.D
0.5
[30.4
X
o
Z n -a
( J T
^~ XX* W
t 0.2
LJ
o
0.1
n
1
^"5-YEAR," ILLUDAS CRITERIA
f (TOTAL DEPTH = 1.33- IN.)
1
1
1
T
1
^^
L.
0 10
STORM OF JULY 16, 1914,
/TE-4.6-YEARS
((1913-1935) ON BASIS OF
TUTAL DEPTH
t ^_
1.1.61-IN.)
4 ' , L___
20 30
-
1
40
TIME, MINUTES
U.D
0.5
S04
^
-0.3
X
fc
gjO.2
Q
0.1
0
1 1
"5-YEAR" ILLUDAS
_ f (TOTAL DEPTH = .
i i
CRITERIA
39- IN.)
STORM OF JUNE 13, 1926,
TE = 5.8-YEARS (1913-1935)
1 ON BASIS OF TOTAL DEPTH
1 1
i
I
i
, ,
I
[__--,
, i
4/1 on
-IN.)
\
-
_
0 10 20 30 40
TIME, MINUTES
FIGURE 3-COMPARISON WITH ILLUDAS, CHICAGO
- 112 -
-------
The synthetic storms yield lesser volumes. They have an advanced type of pattern,
with decreasing depth over time, the characteristics of distributions from intensity-
duration-frequency curves such as the one illustrated in Figure 4. It seems evident
that the ILLUDAS synthetic storm criteria also evolved from the rational method
concept.
Because there are inherent non-linearities in most methods for processing
inputs (subtracting assumed abstractions from total rainfall to derive the net
rainfall associated with runoff) for linear models, and dynamic models are non-linear
by definition, the statistics of the input array may differ appreciably from those of
some or all of the arrays for ruroff and quality characteristics. Attempting to
assign a mean frequency of probable occurrence to a "design storm" is meaningless
because of statistical nonhomogeneity of rainfall, runoff and quality.
In an attempt to test the design storm transformation hypothesis, a
graduate student at Purdue University calibrated the ILLUDAS model for a 13-acre
Chicago catchment for which synchronous rainfall-runoff field data had been collected,
and input both a 35-year rainfall record for Chicago("' and ILLUDAS synthetic
storms of various frequencies.(22) Concluded was that the ILLUDAS synthetic storms
of given return periods produced peak runoffs of essentially the same return period
as those generated from the long-term rainfall record, at least for return periods of
25-years or less. Unfortunately, while the student reproduced the table of
accumulated depths for the 35-year record in an appendix of his report, he had
mistakenly used, instead, another table listing maximum values for each duration,
skewing all storm data into advanced patterns. Therefore, the only conclusion that
can be reached is that advanced rainfall patterns give very similar peak discharge
arrays. What happens with actual storm pattern data thus remains to be demonstrated,
because the well-intended Purdue experiment appears to have been the first reported.
However, an inkling of what might be found is suggested in a test of a
synthetic storm used in a preliminary study of stormwater pollution control for a
major metropolitan area.(23) Because a planning model(2^>25) was used for the study
that normally operates on hourly precipitation data, one-hour interval rainfall data
was used for the test. The model was calibrated for a catchment draining into the
Charles River, and then run with a long-term precipitation record for Boston and with
the synthetic storm. From the test it was found that storage sized to contain either
90% or 95% of the total runoff volume resulting from the synthetic storm would be
capable instead of containing only about 607. or 807., respectively, of the total
runoff volume.
So far, we have restricted discussion to the shorter return periods
associated with storm drainage system sizing. A synthetic storm expected to yield
the peak 100-year urban streamflow was developed in a study of the Four Mile Run in
northern Virginia^26' following the scheme developed in Chicago.(' However, the
100-year peak flow obtained from an extrapolation of a series of peak flows in
simulations with a long period of actual storm data was used to check the validity
of the synthetic storm. The 100-year peak flow simulated using the synthetic storm
was only 11% lower, and the synthetic storm was enlarged accordingly for its
application.in subsequent simulations using a much more elaborate model. Demonstrated
in this study was that a simple continuous simulation model can be used to develop
a synthetic storm for later use in a complex single-event simulation model for more
reliable estimation of a rare event such as a 100-year peak flow. Use of this
aoproach for more frequent events would be even more reliable, provided the simple
simulation model was first calibrated with local rainfall and runoff data as in the
case above.
- 113 -
-------
0.5
to
LJ
I
o
Q.
LJ
0
0.4
0.3
0.2
0.1
T-j-R
1
1
- 1
1
1
1
1
1
1
1
1 1 \ \ 1 -j
DR ELTINGE AND TOWNE, "5-YEAR"
STORM OF JULY 16, 1914,
/TE = 4.6-YEARS (1913-1935)
k ON BASIS OF TOTAL DEPTH (1.61-IN.)
/ ^/STORM OF JUNE 13, 1926,
t
1
j
i
j
1
i
£, TE=5.8-YEARS (1913-1935)
ON BASIS OF TOTAL DEPTH
j (1.85-IN.)
i
i ,
"~" i
1
L._, J
1
"""» L ,
(CONTINUES)
i
. i
0
ELAPSED TIME, MINUTES
60
FIGURE 4- STORM PATTERN COMPARISONS, CHICAGO
- 114 -
-------
Some Preliminary Planning Implications
Preferred inputs for modern planning models are reasonable lengths of
actual rainfall records, perhaps spanning at least 20 years. It seems more
reasonable to route rainfall data of local record through a model to arrive at
output parameter frequencies than to synthesize a storm of some assumed probability.
Use of meteorological expedients can conceivably be unnecessarily hazardous and the
results obtained thereby can be extremely misleading.
In the development of the San Francisco Master Plan for abatement of
pollution from combined sewers, 62-years (1907-1968) of hourly rainfall from the
local U.S. Weather Service gage were used(27,28) ^n conjunction with an early
version of an uncomplicated planning model that has since been iraproved'24>25) an(j
extensively used elsewhere.
As mentioned earlier, for the development of the Chicagoland Master Flan
for abatement of overflow pollution and reduction of flooding, 21-years (1949-1969)
of rainfall data from the local network of twenty raingages was used.'^) Some 104
individual drainage basins in the 375-sq. mi. study area were involved, and thirty-
two computer simulation models were developed as part of the planning effort.
Features of both of the above-mentioned master plans are outlined in an
ASCE Program report.(29'
Another example is the use of a 60-year rainfall record with a rainfall-
runoff model. The model was calibrated using 4-year to 10-year periods of concurrent
rainfall and runoff observations for each of 26 partly sewered catchments ranging in
size from 1/2 to 88 square miles, and an imperviousness of 2 to 35 per cent.''^'
Primary interest was on peak discharges.
Some Design Implications
Because there is insufficient evidence to justify reform, it is difficult
to fault the use of synthetic rainfall distributions and simplistic methods in routine
design of storm sewers and detention reservoirs of small size. Major facilities that
cost substantially more are another matter entirely.
Because the cost of running the more elegant design/analysis models per
storm event is high, many defenders of these models champion acceptance of a
svnthetic storm as an expedient to save time and money, but at the expense of
credibility of results. This is not to suggest that all catchments of a jurisdiction
where large capital investments are contemplated should be analyzed using 20 years or
wore of continuous rainfall records on a complex model. Rather, as some of the
leading practitioners do, such a long record should be applied to a calibrated
catchment near the reference weather station to segregate those storms of design
mportance. Because only the unusual occurrences are of design interest, there may
be perhaps only two dozen or so actual storms of concern. To be consistent, any
oroject sufficiently important to call for the use of a complex model should also be
important enough to apply a few storms rather than a single synthetic storm. Thus,
the handful of storms selected on the basis of simulated catchment response become
a family of design storms for use in connection with other catchments in the
Jurisdiction. Officials in charge of urban drainage facilities are hard-put to
explain an artificial synthetic storm's frequency to irate citizens who have been
flooded or to a State official regulating overflows. Defense against storms of
- 115 -
-------
record is rather direct, and in the opinion of the writer the only realistic option
open to a local government official. The temptation to use artificial confections
as input data should be resisted. In addition, the only way to check the expected
performance of a facility is to simulate not only the design loading but greater
possible loadings as well.
Even if very simple procedures are used, e.g. to determine only a peak
flow rate, the collateral, occasional monitoring of computations by means of one of
the more complete models can serve as an auxiliary guide to sharper judgment.
In the United Kingdom, a national study has been undertaken that
demonstrates that economic analysis-planning-design of storm sewer systems is
technically feasible, on a limited scale.^' However, needed in particular would
be development of damage cost-curves and risk cost-curves. Underpinning damage
cost curves would be performance evaluation, which would have as its objective to
simulate conditions of expected system performance "well enough to be able to
measure how well the thing works and thence, if possible, to get an estimate of
value before the design is committed the degree of stress that a system can
cope with is a measure of its performance."(31) This would be a considerable
departure, from conventional evaluation where, for example, pipe sizes are selected
on the basis of negligible hydraulic stress (no surcharge, flowing full: no
flooding). Concerning performance, it is important to realize that the adoption of
a particular design frequency imputes acceptance of a particular risk of system
capacity or water quality level being exceeded. Thus, employment of a "design storm"
freezes the level of risk, and very little information is available to assess the
reality of presumed levels of risk in a cost-benefit context. Unfortunately, little
data is available even in the U.S. for making suitable benefit analyses.(32")
Conclusion
Current storm drainage analysis procedures too often emphasize the use of
a synthetic storm, which is a device for facilitating analysis but at the expense of
reliability. For small projects such an approach is useful and can be appropriate
in helping to define marginal costs among alternatives, but may be acceptable only
when gross differences in levels of protection from flooding or pollution are sought.
However, for more important projects, local officials should make reference to
actual rainfall histories for the planning, design and operation of new facilities.
This recommendation is particularly appropriate where detention storage is involved
and even more so when detention storage is interconnected as in schemes for reduction
of combined sewer overflows. For the latter, and especially where outflow from
storage would be governed by treatment plant capacities, dewatering may take a
number of hours or days.
: When flooding or pollution occur from system overloading, local officials
should have a firm rainfall data base on which to defend themselves against
criticism and liability, and the use of synthetic storms would leave them all too
vulnerable. Similarly, they should be encouraged to make some field rainfall-runoff-
quality measurements to sharpen the validity of whatever analytical methods and
models are used. Lastly, crudities used in preliminary planning should not be
carried into advanced stages of design or into the operation of complex interacting
systems.
- 116 -
-------
Bauer, W. J., "Economics of Urban Drainage Design," J.Hyd.Djv.. ASCE Proc.,
Vol. 88, No. HY6, pp. 93-114, November, 1962.
2. Poertner, Herbert G., Practices in Detention of Stormwater Runoff. APWA
Special Report No. 43, Chicago, Illinois, 231 pp., June, 1974. (Available
from NTIS as PB 234 554).
3. Rice, Leonard, "Reduction of Urban Peak Flows by Ponding," J.Irrig. and Drain
Div., ASCE Proc., Vol. 97, No. IR3, pp. 469-482, September, 1971.
4. Michel, H. L., and W. P. Henry, "Fairfax County Works to Solve Runoff Problems,"
Water & Wastes Engineering, Vol. 13, No. 9, pp. 38-42, September, 1976.
5. Field, Richard and John A. Lager, "Urban Runoff Pollution Control State-of-
the-Art," J.Envir.Engrg.Djv.. ASCE Proc., Vol. 101, No. EE1, pp. 107-125,
February, 1975. Authors' closure to discussion: Vol. 102, No. EE4, p. 863,
August, 1976.
6. Poertner, Herbert G., "Better Storm Drainage Facilities At Lower Cost,"
Civil Engineering, Vol. 43, No. 10, pp. 67-70, October, 1973.
7. Grigg, Neil S., and John W. Labadie, "Computing the Big Picture," Water &
Wastes Engineering, Vol. 12, No. 5, pp. 37-39 & 86, May, 1975.
g^ Design and Construction of Sanitary and Storm Sewers, ASCE Manuals and Reports
of Engineering Practice No. 37 (WPCF Manual of Practice No. 9), ASCE, New York,
N.Y., 332 pp., 1969.
9. McPherson, M. B., "Some Notes on the Rational Method of Storm Drain Design,"
ASCE Urban Water Resources Research Program Technical Memorandum No. 6, ASCE,
New York, N.Y., 74 pp., January 22, 1969. (Available from NTIS as PB 184 701).
10. Yarnell, D. L., Rainfall Frequency-Intensity Data. U.S. Department of Agriculture,
Miscellaneous Publication 204, 1935.
11 Hathaway, G. A., "Design of Drainage Facilities," Transactions. ASCE, pp. 697-
* 730, Vol. 110, 1945.
12. Hershfield, David M., Rainfall Frequency Atlas of the United States, Weather
Bureau, U.S. Department of Commerce, Technical Paper No. 40, May, 1961.
13. Chow, Ven Te, Frequency Analysis of Hydrologic Data With Special Application to
Rainfall Intensities. University of Illinois Engineering Experiment Station
Bulletin No. 414, Urbana, 80 pp., 1953.
14 Keifer, Clint J., and Henry Hsien Chu, "Synthetic Storm Pattern for Drainage
Design," J.Hyd.Div.. ASCE Proc., Vol. 83, No. HY4, pp. 1-25, August, 1957.
15. McPherson, M. B., "Discussion of Synthetic Storm Pattern for Drainage Design "
J.Hvd.Div.. A.SCE Proc., Vol. 84, No. HY1, pp. 49-57, February, 1958. '
16. Tholin, A. L., and C. J. Keifer, "The Hydrology of Urban Runoff," Transactions
ASCE, Vol. 115, pp. 1308-1379, 1960. L-ai>
17. Pikarsky, Milton, and Clint J. Keifer, "Underflow Sewers for Chicago " Civil
Enp-lneering. ASCE, Vol. 37, No. 5, pp. 62-65, May, 1967.
18. Flood Control Coordinating Committee, "The Chicago Underflow Plan - Development
of a Flood and Pollution Control Plan for the Chicagoland Area," State of
Illinois, County of Cook, Metropolitan Sanitary District of Greater Chicago
City of Chicago, 29 pp., December, 1972. *
- 117 -
-------
19. Bureau of Engineering, City of Chicago, "Computer Simulation Programs,
Development of a Flood and Pollution Control Plan for the Chicagoland Area "
Technical Report No. 2, 101 pp. plus 37 exhibits, December, 1972. (Available
from NTIS as PB 236 645).
20, Terstriep, M. L., and J. B. Stall, The Illinois Urban Drainage Area Sjmulator,
ILLUDAS. Bulletin 58, Illinois State Water Survey, Urbana, 90 pp., 1974""*"
21. Terstriep, M. L., and J. B. Stall, "Urban Runoff by Road Research Laboratory
Method," J.Hyd.Div.. ASCE Proc., Vol. 95, No. HY6, pp. 1809-1834, November,
1969. Authors' closure to discussion: Vol. 97, No. HY4, pp. 574-579 April
1971.
22. Veres, James S., "Testing the Design Storm Concept on Oakdale Basin," a paper
presented to Professor J. W. Delleur in partial fulfillment of the course
requirement for CE 697, Department of Civil Engineering, Purdue University
West Lafayette, Indiana, 87 pp., August 8, 1975.
23. New England Division, "Wastewater Engineering and Management Plan for Boston
Harbor Eastern Massachusetts Metropolitan Area," Technical Data Vol. 8,
"Urban Stormwater Management," Department of the Army, Corps of Engineers
Waltham, Mass., 210 pp., October, 1975,
24. Roesner, L. A., et al., Water Resources Engineers, The Hydrologic Engineering
Center/Corps of Engineers, and Department of Public Works/City and County of
San Francisco, "A Model for Evaluating Runoff-Quality in Metropolitan Master
Planning," ASCE Urban Water Resources Research Program Technical Memorandum
No. 23, New York, N.Y.., 73 pp., April, 1974. (Available from NTIS as
PB 234 312 or from the H.E.C., Davis, Cal. 95616).
25. "Storage, Treatment, Overflow, Runoff Model (STORM), Users' Manual," Computer
Program 723-S8-L7520, The Hydrologic Engineering Center, Corps of Engineers
U.S. Army, 609 Second Street, Davis, California 95616, 170 pp., July, 1976.'
26. Shubinski, Robert P., and William N. Fitch, "Urbanization and Flooding, an
Example," pp. 69-73 in Environmental Modeling and Simulation. Office of Research
and Development and Office of Planning and Management, U.S. Environmental
Protection Agency report EPA 600/9-76-016, 847 pp., July, 1976. (Available
from NTIS as PB 257 142).
27. Department of Public Works, "San Francisco Master Plan for Waste Water
Management," City and County of San Francisco, September 15, 1971. (Three
volumes plus an appendix).
28. Department of Public Works, "San Francisco Master Plan for Waste Water Management
Supplement I," City and County of San Francisco, 88 pp., May 15, 1973. '
29. McPherson, M. B., "Innovation: A Case Study," ASCE Urban Water Resources
Research Program, Technical Memorandum No. 21, ASCE, New York, N.Y. 59
February, 1974. (Available from NTIS as PB 232 166). *' P**
30. Johnson, Steven L., and Douglas M. Sayre, "Effects of Urbanization on Floods
in the Houston, Texas, Metropolitan Area," U.S. Geological Survey Water-Res
Investigations 3-73, 50 pp., April, 1973. (Available from NTIS as PB 220 CeS
31. Green, J., A. King and K. Bowden, "Economics of Sewerage Design " Local
Government Operational Research Unit, Royal Institute of Public'Administrati
Report No. C218, 201 Kings Road, Reading, Berkshire, U.K., 77 pp., April,
1975.
32. Grigg, Neil S., and Otto J. Helwig, "State-of-the-Art of Estimating Flood
Damage in Urban Areas," Water Resources Bulletin. Vol. 11, No. 2 DD
April, 1975. ' VVt
- 118 -
-------
ADDENDUM 3
NOMOGRAPHS FOR TEN-MINUTE UNIT HYDRDGRAPHS
FOR SMOL URBAN WATERSHEDS
ASCE Urban Water Resources Research Program
Technical Memorandum No. 32
by
William H. Espey, Jr., and Duke G. Altman
Espey, Huston & Associates, Inc.
and
Charles B. Graves, Jr.
City of Austin
Austin, Texas
December, 1977
American Society of Civil Engineers
345 East 47th Street
New York, N.Y. 10017
- 119 -
-------
PREFACE
by M. B. McPherson
Background
The following technical memorandum is Addendum 3 of a recent ASCE
Program report on "Urban Runoff Control Planning11.'1-) Addendum 1, "Metropolitan
Inventories," and Addendum 2, "The Design Storm Concept," were appended to the
latter report. The present technical memorandum is the first of several that
will contain additional, individual Addenda over the period 1977-1979.
The principal intended audience of the recent ASCE Program report was
the agencies and their agents that are participating in the preparation of
areawide plans for water pollution abatement management pursuant to Section 208
of the Federal Water Pollution Control Act Amendments of 1972 (P.L. 92-500).
While the presentation which follows is also directed to areawide agencies and
their agents, it is expected that it will be of interest and use to many others
particularly local governments. '
ASCE Program
The American Society of Civil Engineers' Urban Water Resources Research
Program was initiated and developed by the ASCE Urban Water Resources Research
Council (formerly the Urban Hydrology Research Council). The basic purpose of the
Program is to help establish coordinated long-range research in urban water
resources on a national scale.
Abstracts of the twenty-eight reports and technical memoranda of
Program for the 1967-1974 period are included in a readily available paper.
The two reports and the six technical memoranda of the regular series completed
since are identified in a recent publication.^) Also included in the latter is
a. listing of all but one of the twelve national reports in the special technical
memorandum series for the International Hydrological Programme; and the last
national report(^) and an international summary^) have been released since.
A Steering Committee designated by the ASCE Council gives general
direction to the Program: S. W. Jens (Chairman); W, C. Ackermann; J. C. Geyer-
C. P. Izzard; D. E. Jones, Jr.; and L. S. Tucker. M. B. McPherson is Program '
Director (23 Watson Street, Marblehead, Mass. 01945). Administrative support is
provided by ASCE Headquarters in New York City.
Unit Hydrographs
Unit hydrographs have been developed from rainfall-runoff field data
for completely sewered drainage catchments in Louisville, Kentucky,(6) and Atla
Georgia.(7) However, there is considerable doubt over the transferability or *
universality of the findings from a single jurisdiction. Moreover, most of the
unit hydrographs developed across the nation have been for partially sewered
catchments where the streamflows measured had included a significant contributio
from non-sewered sectors. An example is the testing of unit hydrographs using **
urban streamflow data from large catchments nearby by the Georgia Institute of
Technology.(8)
- 120 -
-------
Dr. William H. Espey, Jr., the senior author of the following Technical
Memorandum, pioneered in the regionalization of unit hydrographs (the regional
synthesis of unit hydrograph characteristics) for urban streams. His initial work
dealt mainly with streamflow data for urban watersheds located in Texas, Kentucky,
Illinois, Indiana and Ohio. (9) This initial unit hydrograph analysis was
subsequently expanded utilizing the USGS data for Houston, Texas. ^10) His first
syntheses of characteristics were for thirty-minute duration unit hydrographs,
suitable only for direct application to larger-sized urban watersheds. In contrast,
the characteristics reported herein are for ten-minute duration unit hydrographs.
The synthesized equations and associated nomographs are derived from data for
forty-one urban watersheds of which eighteen are in Texas. Flow gaging in nearly
all instances was in streams, a limitation discussed below.
Unit hydrographs have been employed in investigations of the effects of
urbanization or in the improvement of prediction techniques. (H~16) Field data
from urban catchments in Indiana and elsewhere have been used in tests of various
linear process methods, including "instantaneous" unit hydrographs.' I'' A series
of papers has been devoted to the development of unit hydrographs for gaged urban
Local Validation
Users of the equations herein, or of the nomographs representing them,
are very strongly urged to check their local validity by deriving as many unit
hydrographs as possible using rainfall and runoff data from local catchments for
comparison. This is particularly important where applications will be for wholly
sewered catchments, a condition not truly represented by the formulations
presented. One's first reaction might be that if local unit hydrographs can be
derived why do we need the nomographs? What the nomographs provide is a
generality of characteristics so that local unit hydrographs for particular
catchment physical characteristics can more reliably be extrapolated as part of
a family of characteristics for application in catchments where the physical
characteristics differ. Also, the fact that the nomographs are founded on data
from around the country should reinforce the credibility of any indications
obtained using local data.
An excellent manual on unit hydrograph analysis is readily available^ )
together with a computer program user's manual^") for development of unit
hydrographs from field data and for routing flows from one point to another. The
computer program has been used extensively in urban projects of the Corps of
Engineers. Alternative methods for deriving unit hydrographs (linear programming
versus least squares) have been compared. (2*)
As is the case for some of the newer runoff planning tools, access to a
digital computer is not required for the use of the information in this report.
The user can calculate synthetic unit hydrograph parameters directly from the
eauations with a pocket calculator of modest capability or simply use the
nomographs. Mareover, complete program listings have been published for use of
eleven different small programmable calculators for: deriving a unit hydrograph
from an observed hydrograph; converting a unit hydrograph of a given duration to
its equivalent for another duration; and synthesizing a hydrograph using a given
rainfall pattern and a unit hydrograph. W2-* Also included among the 40 complete
program listings are three different methods for routing hydrographs from one point
to another. The cost of the calculators for which programs are listed span
essentially the full price range. A sample of the programs available, but for
water distribution system analysis, is provided in a technical magazine.' "'
- 121 -
-------
We are not advocating ,the use of unit hydrographs as a replacement for
more versatile or comprehensive tools of analysis but as a supplementary element
among the range of tools needed in the development of urban runoff control plans.
For example, use of locally validated synthetic unit hydrographs based on
indicators from the following report would be an improvement over employment of
hypothetical triangular unit hydrographs as descriptors of subcatchment inputs in
some of the more comprehensive runoff models.
Validation Example
The Colorado Urban Hydrograph Procedure (CUHP) has been used extensively
in the design of flood abatement facilities throughout the Denver metropolitan area
When the CUHP was introduced in 1969 it was founded on a very small amount of
field data. On the basis of much more subsequent data, the Procedure was modified
in 1975(2^' and again in 1977.^25' The Executive Director of the Urban Drainage
and Flood Control District of the Denver metropolitan area, Mr. L. Scott Tucker
was asked to check the equations in the following technical memorandum against
local data. (Mr. Tucker is also a member of the ASCE Program's Steering Committee
and Chairman of the ASCE Council). Mr. Ben Urbonas of his staff reported their
findings,' °' summarized below:
Stream Gaging Station
Area, A, mi2
L, mi
S, ft/ft
I, %
4
Hydrographs checked
Hydrograph
CD
(2)
(3)
(4)
(5)
(6)
(7)
S. Platte R.
Tributary
at Englewood
1.03
1.93
0.0076
46
0.8
(1) 8/19/71 (3)
(2) 8/24/72
Peak Discharge
Via This Tech. Memo.
185
204
99
57
100
218
170
S. Platte R.
Tributary
at Denver
0.75 0.51
2.02 1.56
0.007 0.006
42 60
1.0 0.8
8/7/73 (4) 7/21/76
, cfs, Generated and
Via CUHP
164
180
93
53
105
223
176
Sand Creek
Tributary
at Denver
0.270
0.80
0.0069
43
0.6
(5) 7/24/73
(6) 8/7/73
(7) 7/30/74
Actual
Recorded
122
122
59
36
104
252
250
- 122 -
-------
For the South Platte River Tributary at Denver, rather high values of
were selected to account for induced surface storage expected in that particular
case whenever storm sewer capacity is exceeded. (Considerable care should be
exercised in the selection of ^-values because of their profound influence on the
magnitude of unit hydrograph peak discharge and the time to the peak discharge).
No special attempt was made to modify the originally selected values of
J, to force a better agreement between observed and generated peak discharge for any
of the above cases. The tests reported above were for different streams than the
two in the Denver area from which data was used in the development of the
nomographs herein.
For the Sand Creek Tributary at Denver, the "Hydrograph (6)" storm of
8/7/73 had a maximum average 15-minute intensity close to the once in 100 years
on the average historical value. The 100-year design peak flow would be about
335-cfs for this location, using the CUHP, only about one-third higher than the
peak flow recorded for storm "Hydrograph (6)".
Reiterating the point made-earlier, users of this Technical Memorandum
are emphatically urged to check the local validity of the relationships presented
by deriving as many unit hydrographs as possible using rainfall and runoff data
from local catchments for comparison. The exercise of such care will not
necessarily be easy, because there is an inherent difficulty with any runoff model
in the necessarily subjective separation of abstractions (infiltration, depression
storage, etc.) from total rainfall to resolve rainfall excess (amount and pattern),
which is the input from which an equal volume of direct runoff is generated by
models of one kind or another. After analyzing the performance of a variety .of
models, it was concluded that the weakest link is the proper estimation of
rainfall excess/27'
Acknowledgments
The empirical hydrograph equations presented in this Technical Memorandum
were developed as part of the City of Austin Master Drainage Study under the
direction of Charles B. Graves, Jr., Director of Engineering. Pursuant to this
study, a drainage criteria manual was written which contains design discharge
curves to be used in hydrologic studies in and near the Austin, Texas, area.
These curves were developed utilizing synthetic unit hydrographs determined from
the empirical equations along with local rainfall data for application in the
Austin area.
The ASCE Urban Water Resources Research Council is indebted to the
authors for their generous contribution of this report as a public service.
Processing, duplication and distribution of this Technical Memorandum
was supported by grants to ASCE from the Research Applied to National Needs program
of the National Science Foundation. Technical liaison representative for NSF/RANN
is Dr. J. Eleonora Sabadell. Any opinions, findings, and conclusions or
recommendations expressed herein are those of the contributing authors or the
writer and do not necessarily reflect the views of the National Science Foundation.
New References
To the references of Section 4 of the parent report^' should be added
a useful document containing brief summaries of 18 selected publications dealing
with nonstructural measures for reduction of flood losses.^ °'
- 123 -
-------
In the report' ' to which this Technical Memorandum is Addendum 3
reference 55 on page 72 states that the EPA "Areawide Assessment Procedures
Manual" is in two volumes, whereas a third volume on "Best Management Practices"
has since been released. Also, reference 99 on page 75 is available from NTIS
as PB 239 333, and reference 221 on page 84 is available from NTIS as PB 257 089
(July, 1976).
Reference 236 on page 85 of the parent report has been published in
the Journal of the Water Resources Planning and Management Division. ASCE
Proceedings, Vol. 103, No. WR2, pp. 177-192, November, 1977. Other papers in the
same issue that are related to the subject of urban runoff control planning are
on: "Urban Flood Management: Problems and Research Needs," by D. H. Howells
pp. 193-212; "Innovative Management Concept for 208 Planning," by J. w. Bulkley
and T. A. Gross, pp. 227-240; and "Flood Management for Small Urban Streams "
by W. Whipple, Jr., pp. 315-324. Overlooked in a previous issue was a reference
on control of sedimentation via detention basins. (29)
The report of a nationwide evaluation of combined sewer overflows and
urban stormwater discharges (including the cost of control or abatement of
receiving water pollution from such sources) has been released in three volumes
The first volume is an executive summary. C3") Assessments were for all 248 "Urba
Areas" in the US., of which about 46 per cent is undeveloped. Land uses for th **
54 per cent that is developed are approximately as follows: residential, 58%-
industrial, 15%; commercial, 9%; and other, 18%. Also, about 14% of the 'developed
urban land is served by combined sewers and 38% by separate storm sewers, with
the balance containing unsewered storm drainage. "Average annual dry-weather
flow is significantly greater than average wet-weather flow only in the arid are
However, in most parts of the country, dry-weather flows represent 30 to 50 per 8*
cent of the total (wet plus dry) runoff from urban areas. "OO)
Mr. Carl F. Izzard, member of the ASCE Program Steering Committee ha
re-analyzed the original experimental data for full-size street inlets and
developed a graphical solution for the hydraulic design of curb-opening inlets (31}
An overview and assessment of current catchbasin technology has been
reported recently/32'' Included are evaluations of hydraulic and pollutant
removal efficiencies.
To the transport module of the version of the Storm Water Management
Model embodying the fundamental hydrodynamic equations of motion has been added
capability for analyzing alternatives for the abatement of deposition and scour
storm and combined sewers. The project report'33' includes listings of the new *
computer program subroutines that have been added for solids transport
characterization. A related reportv3^' describes procedures for estimating dr
weather pollutant deposition in combined sewers. ^~
Four papers presented at a recent national meeting, under the auspice
single document.' "'
of the ASCE Urban Water Resources Research Council, have been preprinted in SL
A critique of traditional land-use control mechanisms is included tn
report on economic incentives for such control.(3°) An extensive bibliograoh *
occupies almost a third of the content.
National state-of-the-art indications on urban stormwater runoff co
possibilities have been blended with local information in a comprehensive ret>o °^
for the metropolitan Milwaukee region.'3''
- 124 -
-------
References
1. McPherson, M. B., "Urban Runoff Control Planning," ASCE, New York, N.Y.,
118 pp., June, 1977. (Available as PB 271 548, at $5.50 per copy, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161).
2. McPherson, M. B., and G. F. Mangan, Jr., "ASCE Urban Water Resources Research
Program," J.Hyd.Div.. ASCE Proc., Vol. 101, No. HY7, pp. 847-855, July, 1975.
3. McPherson, M, B., and G. F. Mangan, Jr., Closure to Discussion of "ASCE Urban
Water Resources Research Program," J.Hyd.Div.. ASCE Proc., Vol. 103, No. HY6,
pp. 661-663, May, 1977.
4. Ramaseshan, S., and P. B. S. Sarma, "Urban Hydro logical Modeling and Catchment
Research in India," ASCE UWRR Program Technical Memorandum No. IHP-12, ASCE,
New York, N.Y., 21 pp., May, 1977. (NTIS No. PB 271 300).
5. McPherson, M. B., and F. C. Zuidema, "Urban Hydrological Modeling and
Catchment Research: International Summary," ASCE UWRR Program Technical
Memorandum No. IHP-13, ASCE, New York, N.Y., 48 pp., November, 1977.
6 Eagleson, Peter S., "Unit Hydrograph Characteristics for Sewered Areas,"
J.Hvd.Div*. ASCE Proc., Vol. 88, No. HY2, pp. 1-25, March, 1962. Author's
closure to discussion: Vol. 89, No. HY4, pp. 193-203, July, 1963.
7 Black, Crow and Eidness, Inc., Storm andjCpmbined Sewer Pollution Sources and
Abatement. Atlanta, Georgia, Water Pollution Control Research Series 11024,
ELB 01/71, GPO, Washington, D.C., 181 pp., January, 1971.
g. Wallace, James R., "The Effects of Land Use Change on the Hydrology of an
Urban Watershed," Georgia Institute of Technology Report ERC-0871, Atlanta,
Georgia, 66 pp., October, 1971. (NTIS No. PB 206 426).
9 Espey, W. H., Jr., C. W. Morgan and F. D. Masch, "A Study of Some Effects of
Urbanization on Storm Runoff from a Small Watershed," Center for Research in
Water Resources, Technical Report No. HYD07-6501/CRWR-2, Department of Civil
Engineering, The University of Texas, Austin, 109 pp., July, 1965.
10 Espey, W. H., Jr., and D. E. Winslow, "The Effects of Urbanization on Unit
Hydrographs for Small Watersheds, Houston, Texas, 1964-1967," TRACOR Document
No. 68-975-U, Austin, Texas, 70 pp., September 25, 1968. (NTIS No. PB 183 049),
("Appendices Data Compilations," TRACOR Document No. 68-1006-U, 244 pp.,
September 25, 1968. NTIS No. PB 183 050).
11 Dracup, John A., Thomas J. Fogarty and Sharon G. Grant, "Synthesis aud
Evaluation of Urban-Regional Hydrologic Rainfall-Runoff Criteria,"
Environmental Dynamics, Inc., Los Angeles, California, 105 pp., February,
1973. (NTIS No. PB 220 965).
12 Johnson, Steven L., and Douglas M. Sayre, "Effects of Urbanization on Floods
in the Houston, Texas, Metropolitan Area," U.S. Geological Survey Water-
Resources Investigations 3-73, Austin, Texas, 50 pp., April, 1973. (NTIS No.
PB 220 751).
- 125 -
-------
13. Hamra, D. W., C. W. Morgan and H. A. Reeder, "Statistical Analysis of
Hydrograph Characteristics for Small Urban Watersheds," Tracer, Inc., Report
No. T73-AU-9559-U, Austin, Texas, 155 pp., October, 1973.
(NTIS No. PB 228 131).
14. Chien, Jong-Song, "Urban Runoff by Linearized Subhydrograph Method,"
J.Hyd.Div.. ASCE Proc., Vol. 100, No. HY8, pp. 1141-1157, August, 1974.
15. Schulz, E. F., and 0. G. Lopez, "Determination of Urban Watershed Response
Time," Colorado State University, Hydrology Paper 71. Fort Collins, Colorado
41 pp., December, 1974. '
16. Brater, Ernest F., and James D. Sherrill, Rainfall-Runoff Relations on Urban
and Rural Areas. Environmental Protection Technology Series EPA-670/2-75-04"6
GPO, Washington, D.C., 98 pp., May, 1975. '
17. Rao, A. R. , J. W. Delleur and P. B. S. Sarma, "Conceptual Hydrologic Models
for Urbanizing Basins," J.Hyd.Div., ASCE Proc., Vol. 98, No. HY7, pp. 1205-
1220, July, 1972.
18. Stubchar, James M. , "The Santa Barbara Urban Hydrograph Method," pp. 131-141.
Hann, C. T., "Comparison of Methods for Developing Urban Runoff Hydrographs "
pp. 143-148; and Delleur, J. W. , A. R. Rao and H. Hassain, "On Modeling the'
Runoff Process in Urban Areas," pp. 193-208: Proceedings. National Symposium
on Urban Hydrology and Sediment Control, July 28-31, 1975, Office of Research
and Engineering Services report UKY BU 109, University of Kentucky, Lexington
314 pp., November, 1975. '
19. Hydrologic Engineering Center, Corps of Engineers, "Hydrologic Engineering
Methods for Water Resources Development: Volume 4, Hydrograph Analysis "
Publication HEC-IHD-400, Davis, California, 122 pp., October, 1973. (NTIS
No. AD 774 261).
20. Hydrologic Engineering Center, Corps of Engineers, "HEC-1, Flood Hydrograph
Package," Generalized Computer Program 723-010, 609 2nd Street, Davis,
California 95616, 186 pp., January, 1973.
21. Singh, Krishnan P., "Unit Hydrographs A Comparative Study," Water
Bulletin. Vol. 12, No. 2, pp. 381-392, April, 1976.
22. Croley, Thomas E., Hydrologic and Hydraulic Computations on Small ProgrammaM*.
Calculators. Iowa Institute of Hydraulic Research, The University of Iowa -- "
Iowa City, 837 pp., 1977. ($15.95). '
23. Croley, Thomas E., "Hydraulic Computations for Small Programmable Calculators "
Water & Sewage Works. Vol. 124, No. 11, pp. 64-71, November, 1977. S*
24. Schulz, E. F., Ben Urbonas and Bill DeGroot, "improved Version of the Colorad
Urban Hydrograph Procedure," Flood Hazard News (Urban Drainage and Flood °
Control District, Denver), Vol. 5, No. 2, pp. 1-2, August, 1975.
25. Urbonas, Ben, "Colorado Urban Hydrograph Procedure Revisions," Flood Hazard
News (Urban Drainage and Flood Control District, Denver), Vol. 7 ito T -
p. 7, June, 1977. ' * »
- 126 -
-------
26. Urbonas, Ben, Urban Drainage and Flood Control District, Suite 156B, 2480 West
26th Avenue, Denver, Colorado 80211, letter and enclosure dated July 1, 1977,
to M. B. McPherson, 32 pp.
27. Hossain, A., J. W. Delleur and A. R. Rao, "Evaporation, Infiltration and
Rainfall -Runoff Processes in Urban Watersheds," Water Resources Center
Technical Report No. 41, Purdue University, West Lafayette, Indiana, 63 pp.,
January, 1974. (NTIS No. PB 229 642).
28 Hydrologic Engineering Center, UoS. Army Corps of Engineers, "Annotations of
Selected Literature on Nonstructural Flood Plain Management Studies," Davis,
California, GPO No. 1977-789-006/7333, 95 pp., March, 1977.
29 Curtis, David C., and Richard H. McCuen, "Design Efficiency of Stormwater
Detention Basins," J^Wat.Resour.Plan.k Mgt.Piv.. ASCE Proc., Vol. 103,
No. WR1, pp. 125-140, May, 1977.
r»n American Public Works Association and University of Florida, Nationwide
Evaluation of Combined Sewer Overflows and Urban Stormwater Discharges,
Volume I: Executive Summary. Environmental Protection Technology Series
EPA-600/2-77-064a, Municipal Environmental Research Laboratory, U.S. EPA,
Cincinnati, Ohio 45268, 95 pp., September, 1977. (NTIS No. PB 273 133).
31 Izzard, Carl F., "Hydraulic Design of Curb-Opening Inlets," Flood Hazard
News (Urban Drainage and Flood Control District, Denver), Vol. 7, No. 1,
"p~pT~8-ll, June, 1977.
*2 Metcalf and Eddy, Inc., Palo Alto, California, Catchbasin Technology Overview
and Assessment, Environmental Protection Technology Series EPA-t>00/2-77-051,
Municipal Environmental Research Laboratory, U.S. EPA, Cincinnati, Ohio
45268, 128 pp., May, 1977. (NTIS No. PB 270 092).
33. Sonnen, Michael B., "Abatement of Deposition and Scour in Sewers," Final
Report to U.S. EPA Storm and Combined Sewer Section, Water Resources
Engineers, Inc., 710 South Broadway, Walnut Creek, California 94596, 114 pp.,
June 1, 1977.
>4 Energy and Environmental Analysis, Inc., Boston, Massachusetts, Procedures
* for Estimating Dry Weather Pollutant Deposition in Sewerage Systems.
Environmental Protection Technology Series EPA-600/2-77-120, Municipal
Environmental Research Laboratory, U.S. EPA, Cincinnati, Ohio 45268, 92 pp.,
July, 1977. (NTIS No. PB 270 695).
_ Lytiam, B. T., F. C. Neil and F. E. Dalton, "Managing Storm Runoff in the
^ Chicagoland Area," pp. 1-34; W. B. Whipple, Jr., "Urban Runoff Pollution and
Water Quality Planning," pp. 35-59; B. Urbonas and L. S. Tucker, "Stormwater
Quality - What is the Problem?," pp. 60-89; and P. G. Collins and L. A.
Roesner, "Data Analysis for Nonpoint Pollution Control," pp. 90-111: "Urban
Runoff Quality - Measurement and Analysis," ASCE October 17-21, 1977, Fall
Convention and Exhibit, San Francisco, California, ASCE Preprint No. 3091,
New York, N.Y. ($4.00).
f. CONSAD Research Corp., Pittsburgh, Pennsylvania, Economic Incentives for
Land Use Control, U.S. EPA Office of Research and Development, Report
EPA-600/5-77-001, Washington, D.C. 20460, 359 pp., February, 1977. (NTIS
No. PB 265 468).
- 127 -
-------
37. Southeastern Wisconsin Regional Planning Commission, "State of the Art of
Water Pollution Control in Southeastern Wisconsin, Volume 3, Urban Storm
Water Runoff," SEWRPC Technical Report No. 18, P.O. Box 769, Waukesha
Wisconsin 53186, 63 pp., July, 1977. ($5.00). '
- 128 -
-------
NOMOGRAPHS FOR TEN-MINUTE UNIT HYDROGRAPHS
FOR SMALL URBAN WATERSHEDS
(Addendum 3 of the ASCE Program report "Urban
Runoff Control Planning," June, 1977)
by
William H. Espey, Jr., President, and
Duke G. Altraan, Staff Engineer,
Espey, Huston & Associates, Inc.,
3010 S. Lamar
Austin, Texas 78704
and
Charles B. Graves, Jr.,
Director of Engineering,
City of Austin,
P.O. Box 1088,
Austin, Texas 78704
December, 1977
- 129 -
-------
TABLE OF CONTENTS
Pa^e
Introduction ...... ...... ...... ««.. 131
Figure 1 - Definition of Unit Hydrograph Parameters ^33
Watershed Data . 134
Table 1 - Location and Physiographic Description of Watersheds . . 135
Figure 2 - Watershed Conveyance Factor, ^, As a Function of
Per Cent Watershed Impervious Cover, I, and Weighted
Main Channel Manning 'n* Value 138
Table 2 - Classification of Watershed Drainage Systems 140
Development of Empirical Ten-Minute Unit Hydrograph Equations 142
Table 3 - Ten-Minute Unit Hydrograph Equations 144
Nomographs and Examples ........ . 0 143
Figure 3 - Time of Rise (TR) Nomograph ...oo.... 145
Figure 4 - Peak Discharge (Q) Nomograph ......... 146
Figure 5 - Time Base (Tg) Nomograph 147
Figure 6 - Hydrograph Width (W^Q) Nomograph .............. 143
Figure 7 - Hydrograph Width (Wy^) Nomograph . . 149
Figure 8 - Example of Ten-Minute Unit Hydrograph Construction ..... 152
References >.. 151
- 130 -
-------
NOM3GRAPHS FOR TEN-MINUTE UNIT HYDROGRAFttS
FOR SM&.LL URBAN WATERSHEDS
Introduction
Due to rapid runoff response and the resulting flood potential characteristic
of many relatively small urban and surrounding area watersheds, a need is evident
for adequate and efficient description of the dynamic process involved. In some
instances, procedures such as the "rational method" may be sufficient in
describing a hydrologic process in a small watershed but at other times a more
definitive method is required. A short-duration unit hydrograph has considerable
potential for describing the dynamic runoff process of small watersheds.
The basic theory of the unit hydrograph appears to have been suggested first
by Folse (1929). The Boston Society of Civil Engineers (1930) stated that "the
base of the flood hydrograph appears to be approximately constant for different
i
floods, and peak flow tends to vary directly with the total volume of runoff".
Three years later, in 1932, Sherman introduced the basic concept of the unit
hydrograph. The unit hydrograph has found wide acceptance as an outstanding
contribution to engineering hydrology. The general concept of the unit hydrograph
has been summarized as follows (Jforgan and Johnson, 1962):
"for a given drainage area, the time-base of surface-runoff hydrographs
resulting from similar storms of equal duration are the same
regardless of the intensity of rainfall;
"for a given drainage area, the ordinates of the surface-runoff
hydrographs from similar storms of equal duration are proportional
to the volume of surface runoff; and
"for a given drainage area, the time distribution of surface runoff
from a particular storm period is independent of that produced by
any other storm period."
Because the unit hydrograph is such a valuable hydrologic tool, many investigations
have been undertaken to develop synthetic unit hydrograph relationships based on
watershed features.
- 131 -
-------
Because it is generally accepted that the best unit period should be
somewhat smaller than the watershed lag time (time from the center of mass of
rainfall to the peak of the hydrograph), the need for short-duration unit
hydrographs to describe the runoff process for small, quick-responding watersheds
becomes, obvious. Empirical equations given subsequently were developed to aid
users in describing or synthesizing the shape of a ten-minute unit hydrograph for
any small watershed for which adequate rainfall-runoff relationships had not been
previously developed through use of site-specific data.
The following hydrologic parameters were chosen to describe the shape of the
ten-minute unit hydrograph:
T - the time of rise, in minutes;
H
Q - the peak discharge, in cfs;
T_ - the time base, in minutes;
B
W _ - the time, in minutes, between the two points on
the unit hydrograph at which the discharge is
half of the peak discharge; and
W - the time, in minutes, between the two points on
the unit hydrograph at which the discharge is
three-fourths of the peak discharge.
For clarification, these parameters are illustrated in Figure 1. Once the five
parameters have been determined by use of the empirical equations, and-an
auxiliary means for distributing a runoff volume of one inch has been employed,
the ten-minute unit hydrograph can be easily constructed for a watershed.
Dimensionless unit hydrographs, such as those found in the Soil Conservation
Service hydrology manual (USDA-SCS, 1971), can be used as an aid in determining
the basic shape of the ascending and descending portions of the unit hydrograph.
Generally, the lower portions of the ascending and descending unit hydrograph
limbs can then be adjusted to obtain the one inch runoff volume.
From the empirical equations, parallel, scaled nomographs have been developed
to provide a straightforward, simple procedure enabling graphical determination
by the user of the five hydrologic parameters. The nomographs are featured in
this document.
- 132 -
-------
UJ
o
OL
<
O
TIME
FIGURE I - DEFINITION OF UNIT HYOR06RAPH PARAMETERS
- 133 -
-------
Watershed Data
The five hydrologic parameters that describe the fundamental shape of the
ten-minute unit hydrograph were developed from rainfall-runoff data collected
for forty-one small watersheds throughout the United States. Of the forty-one
watersheds, eighteen are located in Texas. Listed in Table 1 are the name,
location, and physiographic characteristics of each of the forty-one watersheds
which have a range in size of drainage area between 0.014 and 15.0 square miles.
As indicated in Table 1, size of drainage area (A), main channel length (L)
main channel slope (S), extent of impervious cover (I), and a dimensionless
watershed conveyancy factor (
-------
TABLE 1
LOCATION AND PHYSIOGRAPHIC DESCRIPTION OF WATERSHEDS
Watershed Name
Bachman Branch
Hunting Bayou at
Cavalcade Street
Stoney Brook Street
Ditch
Sims Bayou at Carlsbad
Street
Cole Creek at Guhn Rd.
Bering Ditch at
Woodway Drive
Berry Bayou at Forest
Oaks Street
Berry Bayou at Gil pin
Street
Berry Bayou at Globe St.
Brickhouse Gully at
Clarblak Street
Halls Bayou at
Deertail Street
Keegans Bayou at
Keegan Road
Willow Waterhole
at Landsdown St.
Watershed Drainage Area, A
Location (Square miles)
Dallas, Texas 10.0
Houston, Texas 1.03
" 0.50
" 4.99
" 7.05
11 2.59
11 11.10
" 3.26
" 1.58
" 2.01
11 5.27
11 5.77
" 1.15
Main Channel
Length, L
(Feet)
28,512
5,800
3,700
12,400
18,500
11,400
18,300
4,500
7,900
12,700
20,600
31,200
1,700
Main Channel
Slope, S
(Ft. per Foot)
0.006
0.002
0.0006
0.0008
0.001
0.0007
0.0016
0.0015
0.0006
0.0011
0.0011
0.0005
0.0006
Impervious
Cover, I
(%)
30.0
27.0
33.0
4.0
4.0
17.0
13.0
9.0
15.0
2.5
2.0
2.0
33.0
Dimensionless
Conveyance
Factor,
0.8
1.25
0.60*
1.30
1.30
0.85
0.65*
1.00
1.00
1.05
1.30
1.30
0.95
*Modified i selection process
(Continued)
-------
TABLE 1 (Continued)
Watershed Name
Waller Creek at
38th Street
Waller Creek at
23rd Street
Helotes Creek at
State Hwy. 16
Brown Creek at
State Fairgrounds
Three Mile Creek
Crane Creek
Walton Run
Little Sugar Creek
at Brookcrest Dr.
Little Sugar Creek
at Burnley Road
Paw Creek at
Allenbrook Drive
Briar Creek at
Sudbury Road
Briar Creek at
Shamrock Drive
Irwin Creek
Tributary
Main Channel
Watershed Drainage Area, A Length, L
Location (Square miles) (Feet)
Austin, Texas 2.31
4.13
San Antonio, 15.00
Texas
Nashville, 11.80
Tenn.
Jackson, Miss. 1.10
0.45
Phila., Penn. 2.17
Charlotte, 0.84
N.C.
11 0.44
" 0.62
0.56
" 0.52
11 0.27
23,080
27,560
35,600
28,512
9,504
4,224
14,784
9,029
5,544
7,022
5,808
5,650
5,914
Main Channel
Slope, S
(Ft. per Foot)
0.009
0.009
0.01
0.0072
0.0079
0.0067
0.0068
0.0175
0.0157
0.0143
0.0112
0.0148
0.0193
Impervious
Cover, I
(7.)
27.0
37.0
6.0
15.7
19.7
27.5
24.7
21.0
14.0
18.0
16.0
20.0
19.0
Dimensionless
Conveyance
Factor,
0.80
0.80
1.00
0.80
0.75
0.80
1.00
0.80
0.80
0.80
0.85
0.85
0.90
(Continued)
-------
TABLE 1 (Continued)
Watershed Name
Silas Creek at Pine
Valley Road
Tar Branch at Walnut
Street
Brushy Creek at U.S.
Hwy. 311
Sanderson Gulch
Tributary
Tuck Drain
Western Outfall
Sewer
17th Street Sewer
Northwest Trunk
Southern Outfall
Southwestern Outfall
Freeman Field A
Freeman Field B(l)
+ Apron
Freeman Field B + Taxi
St. Anne Auxiliary
Field
Godman Field No. 1
Watershed Drainage Area, A
Location (Square Miles)
Winston-Sal em,
N.C.
ii
it
Lakewood,
Colorado
Northglenn,
Colorado
Louisville,
Kentucky
ii
ii
ti
ii
Indiana
ii
ii
Ind iana
Kentucky
0.89
0.59
0.55
0.47
0.067
2.77
0.22
1.90
6.43
7.52
0.015
0.0128
0.014
0.114
0.0205
Main Channel
Length, L
(Feet)
8,554
6,706
5,808
4,752
1,584
22,000
4,900
16,000
34,000
34,200
900
555
1,200
2,600
1,300
Main Channel Impervious
Slope, S Cover, I
(Ft. per Foot) (%)
0.0193
0.0295
0.0271
0.0139
0.0163
0.0009
0.0038
0.0012
0.0014
0.0015
0.009
0.0058
0.004
0.008
0.015
19.0
28.0
37.0
51.0
33.0
70.0
83.0
50.0
48.0
33.0
21.6
100.0
100.0
5.1
22.1
Dimensionless
Conveyance
Factor, $
0.90
0.80
0.90
0.70
0.80
0.60
0.60
0.60
0.60
0.60*
1.00
1.00
1.00
1. 00
1.00
u>
vj
*Modified i selection process
-------
.01 02 03 04 05 .06 .07 .08 .09 .10
.12 .13 .14 .15 .16 .17
WEIGHTED MAIN CHANNEL MANNING n VALUE
FIGURE 2-WATERSHED CONVEYANCE FACTOR, $, AS A FUNCTION OF
PER CENT WATERSHED IMPERVIOUS COVER, I, AND WEIGHTED
MAIN CHANNEL MANNING V VALUE. (Austin/Drainage Manual, 1977)
- 138 -
-------
The effects of urbanization on the unit hydrograph can generally be
accounted for in the percentage of impervious cover (I) and the watershed
conveyance factor (tf). Man-made improvements such as buildings, paved streets,
and parking lots effectively reduce infiltration into the ground and often
increase the volumes and peak rates of runoff. The I and values are related to
these improvements and are indicators of a watershed's runoff response
characteristics.
A brief history concerning the evolution of the 4 factor is appropriate at
this point. Espey e£ &! (1965) found that the introduction of watershed impervious
cover as an index of urbanization was not sufficient for describing the runoff
characteristics of some urban watersheds. It was noted in this early study that
in most cases when the channel had been improved, and/or a storm sewer system
existed, the predicted values of the time of rise (TR) were high compared to
measured values. Therefore, a new urban factor (^) was introduced to account for
the reduction in the time of rise due to channel improvements or storm sewers.
Further studies by Espey e£ al^ (1968) refined the determination of the factor
to include channel vegetation conditions. In the latter study, for several
watersheds, predominately in the Houston area, a significantly shorter response
time was indicated than for watersheds in other areas. The only apparent variable
that could account for the shorter response times was channel vegetation. To
explain the effect of channel vegetation on the time of rise, the watershed
conveyance factor (
-------
*1
CLASSIFICATION
0.6
EXTENSIVE CHANNEL IMPROVEMENT AND
STORM SEWER SYSTEM, CLOSED CONDUIT
CHANNEL SYSTEM.
0.8
SOME CHANNEL IMPROVEMENT AND STORM
SEWERS; MAINLY CLEANING AND ENLARGE-
MENT OF EXISTING CHANNEL.
1.0
NATURAL CHANNEL CONDITIONS.
$2
0.0
0. 1
0.2
0.3
CLASSIFICATION
NO CHANNEL VEGETATION.
LIGHT CHANNEL VEGETATION.
MODERATE CHANNEL VEGETATION,
HEAVY CHANNEL VEGETATION.
= <£,
TABLE 2 - CLASSIFICATION OF WATERSHED DRAINAGE SYSTEMS
(Espey et gl., 1968)
- 140 -
-------
that study the 4> factor was determined for each watershed shown In Table 1 and
used as one of the watershed physiographic characteristics in a statistical
analysis to determine the ten-minute unit hydrograph parameter equations.
Figure 2 was then developed to aid users in the determination of the 4 factor
for urban watersheds.
It must be pointed out that the normal process for selecting a watershed's 4
value (Figure 2) may not be applicable for watersheds with atypical drainage
characteristics. For that reason, judgment must be exercised in the use of
Figure 2 because the data base used in the development of the figure was limited
(Table 2). Future study and collection of additional data may indicate a need
for modification or refinement of the selection process. If flow-modifying
conditions exist in a watershed, such as excessive surface detention storage,
combined sewer systems, or overtaxed storm sewer systems combined with a lack of
surface escape routes, additional hydrologic procedures may be required for the
description of the rainfall-runoff process.
Rainfall-runoff data for numerous watersheds were reviewed for applicability
in describing the composition of the ten-minute unit hydrograph equations. The
major share of this data was obtained from a TRACOR report (Espey, et aJU, 1968)
and a report to the Texas Water Commission (Espey, e£ al., 1965). Reported in
these two documents were reduced runoff data from fifty watersheds providing
average thirty-minute unit hydrographs for the respective watersheds. Data
compilation8 reported by the U.S. Geological Survey were also reviewed, from
hich a few more watersheds were found with adequate hydrologic data that could
be incorporated into the data base. The previously derived thirty-minute unit
hvdrographs were transformed into ten-minute unit hydrographs by the S-curve
procedure, which is described in most comprehensive hydrology texts. The results
£ £he S-curve reductions combined with new unit hydrographs were then studied
and the resulting ten-minute unit hydrographs having the best shape were chosen
- 141 -
-------
for further analysis. From this process the final forty-one watersheds listed
in Table 1 were selected to determine the relationships between physical watershed
characteristics and the shape characteristics of the ten-minute unit hydrograph to
be generalized. Definitions of the five hydrograph parameters discussed earlier
were derived from among the variables affecting the ten-minute unit hydrographs for
the forty-one selected watersheds.
Development of Empirical Ten-Minute Unit Hydrograph Equations
The objective of developing empirical equations to describe the runoff
response characteristics of watersheds (i.e., a unit hydrograph) is to provide a
means for taking known or relatively easily obtained information, such as rainfall
and watershed physical characteristics, and to describe synthetically the runoff
hydrograph for a particular rainfall event. Of course, there are other
considerations that must be accounted for along with the synthetic unit hydrograph
in describing runoff patterns. These considerations include such items as
infiltration rates, rainfall surface-storage, base flow, and other flow-modifying
conditions.
The problem of statistically synthesizing a unit hydrograph is by definition
one of approximating the hydrograph shape. Empirical equations allow the
description of a hydrograph in terms of typical hydrograph dimensions (parameters)
avoiding the difficulties that would have to be overcome in attempting to derive
a mathematical function for the hydrograph. Having described the hydrograph for
each of the forty-one watersheds in terms of five parameters, each parameter was
statistically described as a function of the physiographic features and/or other
unit hydrograph parameters of the watershed.
To describe the five parameters mathematically, a multiple non-linear
regression formula of the following form was used:
A
V -2
Y = KX, Xn . . . AI, ,
- 142 -
-------
where Y is one of the five hydrograph parameters,
X (for i = 1, ..., N) are physiographic characteristics or hydrograph
1 parameters for the watershed, and
K and A (for i = 1, ..., N) are regression coefficients.
The above equation was expressed in the following logarithmic form:
log Y » log K + AI log Xx + . . . A^ log X^
and the method of least squares was then applied to evaluate the regression
coefficients. This method, instead of minimizing the sum of the squares of the
deviations of the function itself, minimizes the sum of the squares of the
deviations of the logarithm of the function.
Table 3 presents the results of the regression analysis in the form of the
ten-minute unit hydrograph equations. For easy reference, brief definitions are
included in Table 3 of the watershed physiographic characteristics and the five
hydrograph parameters. Table 3 also presents the statistical accuracy of the
empirical ten-minute unit hydrograph equations obtained using the data from the
forty-one watersheds, by giving the total explained variation of the logarithmic
oredictions for each of the five equations, where the value 1.000 would be for
a perfect fit.
and Examples
To simplify tne process of synthesizing ten-minute unit hydrographs, a set
Of parallel, scaled nomographs were constructed using the empirical equations, to
facilitate user determination of values for all five of the unit hydrograph
parameters. A ten -minute unit hydrograph can then be constructed resembling the
ofle shown in Figure 1. These nomographs are presented in Figures 3 through 7.
A description of the respective nomographs shown in each of these five figures
will now be given along with an example of their use.
Determination of the unit hydrograph parameters using the developed nomographs
- 143 -
-------
TABLE 3
TEN-MINUTE UNIT HYDROGRAPH EQUATIONS
Total
Equations Explained Variation
T^.IL^V0'25!-0-18*1-57 0.802
Q , 31.62xl03 A0-9V°7 0.936
T_ 125.89xl03 A Q~°'95 0.844
B
W5()- 16.22xl03 A0'93 Q-°-92 0.943
W75» 3.24xl03 A0'79 Q-°'78 0.834
L is the total distance (in feet) along the main channel from the point being
considered to the upstream watershed boundary.
S is the main channel slope (in feet per foot) as defined by H/(0.8L), where L
ia the main channel length as described above and H is the difference in
elevation between two points, A and B. A is a point on the channel bottom
at a distance of 0.2L downstream from the upstream watershed boundary. B is
a point on the channel bottom at the downstream point being considered.
I is the impervious area within the watershed (in per cent).
4 is the dimensionless watershed conveyance factor as described previously in
the text.
A is the watershed drainage area (in square miles).
T_ is the time of rise of the unit hydrograph (in minutes).
j\
Q is the peak flow of the unit hydrograph (in cfs).
T0 is the time base of the unit hydrograph (in minutes).
B
W50 is the width of the hydrograph at 50% of the Q (in minutes).
W?5 is the width of the unit hydrograph at 75% of Q (in minutes).
- 144 -
-------
50,000
O
^ 20,000<
*-
10.OOO
_j
-"
x
o ^o00
z
UJ
_l 2,000
_J
2 1000
Z
X
0 400
Z
2 200
IOO
f-
r c
*- «
^- S
r ^
b <
H_ *"^
"**'
r oc
r uj
r >
= o
r u
r co
r o
^- >
cc.
=- UJ
9- O.
i_ .^
r^ 2
^
h S
= X
r- <"
E_ uj
£T
*
a,
=-
L oc
= o
1-3 <-»
4
r u.
L4 w
= ""Z
= <
=-5 >
=. UJ
1-6 Z
E °
E-8 0 <
= UJ
9 X
S-K) ^
- ui
<
E ^
E-
1.5
>
~ 3
-10 -
^ ' C
~~- ^_^
E~ -^ *~
=-0.9 "" ^H?
~ ""^
=". UJ
- t CO
*
u.
O
=Lo.7 uj
«
r »-
~ /
06 ..^
=-50O
^
^_
|-
=-2OO
^ .
^100
5-
~ ^^ ^
fso "
_ ^~ -^
= / %
5-20 li
r z
I z
1-
i-20 /
=- X
= /
1. /
1-30 /
[/
t~4O
|-50
1- 60
e/
p __
UJ
z
_i
0
z
z
(T
t-
R
c
' - _^_
~~
E-7O
E-80
i-90 FIGURES- TIME OF RISE (TR) NOMOGRAPH
e-100
UJ
Q.
O
_J
CO
U)
o
z
.0005
==.001
|-002
t
I-.005
01
EE_b
|-.02
t
S-.04
-------
Q = 31.62 xl03(A-96)(TR~L07)
i-IO
r'5
r-15
C\J I
0) ~
r -10 £
E -9 "
< L; §
r6 _, -m-" <
LLl ; """ """ Q:
< : 4 x
r o
-3 co
UJ ; _
CO r- Q
z --z ^,
< ~ "*
2 f S
-
-1.0
- 9
=- 30,000
=-20,000
=-
z
^10,000
1-5,000
1- 4,000 __ ^- " "
j~ ^T^^^^
1- 2,000
i: 1,000
h 500
|- 400
1- 300
i-
|- 200
100
- .8
.7
6'
^_ "
£
3
c
E
"^r
K
Ul
CO
o:
u.
o
Ul
3s
(-
E-20
|-
|-25
~
r30
-
t*
|-40
|-50
1-60
5-70
E-80
L-90
^-100
150
E-200
§-250
L300
E-400
-500
FIGURE 4- PEAK DISCHARGE (Q) NOMOGRAPH
-------
PEAK DISCHARGE (Q) cfs
Hinniuinuipiinim|mn i |i p|i| i n 'l""l
o
o
to
o» -J CO
TIME BASE (Tg) minutes
iyni|iiM|i|i|i|ipiyiii|iiiin
JSSM S gSS^SSiS S
ao 5 o p o °^
O
o
DRAINAGE AREA (A) miles
m
0»
bbo b o o o
(0 OD --j »
-------
o
O
UJ
O
IT
X
O
5
ui
a.
-4
E-3
2
30,000
-20,000
-10,000
9
-7
6
._ |E
0>
O
-1,000
-9
-8
-7
-6
-5
-4
=-3
2
! 0
=-2.0
!-30
E-4.0
i- 5.0
60
70
8.0
9.0
10.0
E- 20 0
evj
100.0
=- 200.0
j- 300 0
400.0
500.0
600.0
700.0
8000
9000
1000.0
UJ
a:
ui
c
o
F^ .5
5- .7
^ 8
I 9
^ 1.0
r 1.5
L 2.0
1- 2-5
I 3.0
- 5.0
6.0
- 7.0
-8.0
-9.0
E: 10 o
15.0
100
FIGURE 6- HYDROGRAPH WIDTH (W5O) NOMOGRAPH
- 148 -
-------
c- 30,000
W75=3.24xl03(A79)(Q~78)
- 20,000
*
a
'i
"" -^ 'To
^ ^ --
X
H
9
^g
X
a
a:
o
o
oc
a
X
- 10
^ 20
=3.0
^40
1-5.0
L6.0
-7.0
rs.o
r9.0
710.0
r200
=-30.0
1-40.0
i-so.o
^-60.0
-70.0
rSO.O
: 106.0
-
~
-200.0
- 300.0
-400.0
- 500.0
^- 600.0
L 700.0
L8OO.O
p-.5
- .6
- .7
.6
1.0
20
=- 25
40
50
FIGURE?- HYDROGRAPH WIDTH (W75) NOMOGRAPH
- 149 -
-------
is shown below for a watershed having the following physiographic characteristics:
A =» 4.1 square miles
L - 21,000 feet
S - 0.01 ft/ft
I m 40 per cent and
4 » 0.8
The unit hydrograph time of rise (Tj.) determination is a multiple-step
process that can be followed In Figure 3. Align the scale values for L (21,000
at point a) and S (0.01 at point b) with a straightedge and mark point c on
turning line 1. Align point c and the scale value for I (40 at point d) along a
straight line and mark point e on the intersection with turning line 2. Align
point e and the scale value for $ (0.8 at point f) with a straightedge and read
the time of rise (TR) value at the intersection with its scale (35 at point g).
The peak discharge (Q) of the unit hydrograph can be obtained from Figure 4
by simply aligning the drainage area scale value (4.1) and the time of rise scale
value (35) with a straightedge and reading the Q value from the peak discharge
scale in the middle (2700).
Figure 5 provides the unit hydrograph time base (TR) by aligning the peak
discharge scale value (2700) and the drainage area scale value (4.1) with a
straightedge and obtaining the value from the time base scale in the middle (285)
Hydrograph widths W and W _ can be respectively found by aligning the
drainage area (4.1) and peak discharge scale values (2700) with a straightedge in
Figures 6 and 7 and reading the appropriate values from the hydrograph width
scales in the middle of each nomograph (W - 42 and W = 21).
The above determinations are summarized below for the five ten-minute unit
hydrograph parameters:
TR - 35 minutes
Q - 2700 cfs
T0 » 285 minutes
a
W. » 42 minutes
W m 21 minutes
- 150 -
-------
8 shows a constructed ten-minute unit hydrograph satisfying these five
hydrograph parameter values. The dimensionless unit hydrograph shape presented
in the USDA-SCS (1971) hydrology manual was used to approximate initially the
shape of the ascending and descending portions of the constructed hydrograph in
Figure 8. ^ne un*-t hydrograph's ascending and descending portions below 50
per cent of peak discharge were then adjusted until the represented runoff volume
equalled one inch.
inferences
Boston Society of Civil Engineers, 1930. Report of the Committee on Floods:
journal of the Boston Society of Civil Engineers, V. 17, no. 7 (Sept.)
p. 285-464.
City of Austin Engineering Department, 1977. Drainage Criteria Manual, First
Edition.
Espey, Jr., w- H-» C* W* ^rg*11* and F- D- Masch, 1965. A Study of Some Effects
of Urbanization on Storm Runoff from a Small Watershed: Technical Report
HYD 07-6501, CRWR-2, Center for Research in Water Resources, Department of
Civil Engineering, University of Texas, Austin, Texas.
Espey Jr., W. H., and D. E. Winslow, 1968. The Effects of Urbanization on Unit
Hydrographs for Small Watersheds, Tracor Document No. 68-975-U, including
appendices.
Folse, J. A., 1929. A New Method of Estimating Streamflow: Carnegie Institution
of Washington, Publication 400.
Hanm D. W., C. W. Morgan and H. A. Reeder, 1973. Statistical Analysis of
Hydrograph Characteristics for Small Urban Watersheds, Tracor Document
No. T73-AU-9559-U.
Morgan, P. E«i and s» M* Johnson, 1962. Analysis of Synthetic Unit-Graph Methods:
Journal of the Hydraulics Division, Proceedings, American Society of Civil
Engineers, V. 88, No. HY5, pt. 1 (Sept.), p. 199-220.
U S. Department of Agriculture, 1971. National Engineering Handbook, Section 4,
Hydrology, USDA-SCS.
- 151 -
-------
TR = 35 min
en
ro
2750 f*
40
80
120 160
TIME- minutes
200
240
280
300
FIGURE 8- EXAMPLE OF TEN-MINUTE UNIT HYDR06RAPH CONSTRUCTION
-------
RESEARCH ON THE DESIGN STORM CONCEPT
ASCE Urban Water Resources Research Program
Technical Memorandum No. 33
by
Jiri Marsalek
National Water Research Institute
Burlington, Ontario,
Canada
September, 1978
American Society of Civil Engineers
345 East 47th Street
New York, N.Y. 10017
- 153 -
-------
PREFACE
by M. B. McPherson
Background
The following Technical Memorandum is Addendum 4 of a 1977 ASCE
Program report on "Urban Runoff Control Planning".'*' Addendum 1, "Metropolitan
Inventories," and Addendum 2. "The Design Storm Concept," were appended to the
latter report. Addendum 3^2' was the first of several additional, individual
Addenda to be released over the period 1977-1979.
The principal intended audience of the ASCE Program's June, 1977,
report was the agencies and their agents that are participating in the preparation
of areawide plans for water pollution abatement management pursuant to Section 208
of the Federal Water Pollution Control Act Amendments of 1972 (P.L. 92-500).
While the presentation which follows is also directed to areawide agencies and
their agents, it is expected that it will be of interest and use to many others
particularly local governments.
ASCE Program
The American Society of Civil Engineers' Urban Water Resources Research
Program was initiated and developed by the ASCE Urban Water Resources Research
Council. The basic purpose of the Program is to promote needed research and
facilitate the transfer of findings from research to users on a national scale.
Abstracts of the twenty-eight reports and technical memoranda of the
Program for the 1967-1974 period are included in a readily available paper.(3)
The two reports and the six technical memoranda of the regular series completed
since are identified in a companion publication.'^' Also included in the latter
is a listing of all but one of the twelve national reports in the special technical
memorandum series for the International Hydrological Programme; and the last
national report^5) and an international summary) have been released since.
A Steering Committee designated by the ASCE Council gives general
direction to the Program: S. W. Jens (Chairman); W. C. Ackermann; J. C. Geyer-
C. F. Izzard; D. E. Jones, Jr.; and L. S. Tucker. M. B. McPherson is Program
Director (23 Watson Street, Marblehead, Mass. 01945). Administrative support is
provided by ASCE Headquarters in New York City.
Design Storms
In our Addendum 2, "The Design Storm Concept," in our June, 1977,
report^1' we attempted to indicate the hazards that might be encountered in
interpreting results of analyses based on that concept. However, we were forced
to dwell almost exclusively on the characteristics of rainfall because demonstrate
of the effects of transforming synthetic rainfalls into runoff were lacking. Xhe
report that follows provides the first such demonstration. While the findings ar
site-specific and are certainly not universal, they effectively show the potential
liabilities in the adoption of synthetic storms for the planning and design of
important works.
We have long advocated reference to a long period of historical rainfall
for simulation of important sewered systems. It was not proposed that the entir
- 154 -
-------
record would be used in every simulation exercise. Rather, we suggested that
the about two-dozen actual storms of analysis interest be identified from total
record simulations of token catchments, and that the resultant set of "design
storms" be applied thereafter in analyses for other catchments in the jurisdiction
involved. The report herein describes an alternative method for accomplishing
essentially the same objective. A somewhat similar procedure was reported in
1972(7) for segregating storms of primary interest to facilitate simulation of
urban catchment flows for a long period of record. More recently, a storm
screening procedure applied to the rainfall data for several Canadian cities has
been described. (8)
An alternative to the above approaches has been sought in an attempt to
oreserve the statistically attractive features of continuous hydrological modeling
while reducing the extensive computer costs of such simulations. To this end, a
"fixed recurrence interval transfer technique" has been developed by the
Southeastern Wisconsin Regional Planning Commission.^9) Using this technique,
once a full continuous simulation has been completed for a given catchment
condition, it is possible to explore revised conditions via simulation of a few
selected meteorological intervals of data. In this way, alternative plans and
oro lections can be compared. The technique was developed "for preliminary
screening and assessment of the impact of alternative land use conditions and
structural water control measures". (9) While the technique is illustrated in the
reference with flood flow simulation examples, it is noted that the concept has
been used already in water quality simulations. (10)
Testing of the design storm concept is also under way overseas, for
example in Sweden. (H)
Results have been reported from a preliminary test of design storms
versus actual rainfall data in Denver. (12) The 190-acre catchment is drained by
separate storm sewers and concrete lined channels. A unit hydrograph derived
from field data for the catchment was used in the tests. Concluded was that
design storms can produce results that can vary significantly from the probability
distribution of runoff simulated using actual rainstorm data. Planned are similar
analyses for about fifteen other catchments for which rainfall and runoff data are
available.
Initial indications from an investigation at the University of Illinois
Urbana of the design storm concept will be reported in December 1978. (13)
This writer was advised in private correspondence with the University investigator
that preliminary findings, for a large urban watershed for which the model used
AS calibrated against field data, were analogous to those in Denver.
Both the Denver and University of Illinois correspondents have warned
Hat the magnitudes of assumed rainfall abstractions used in simulations can have
large an effect on the results as the use of 'design storms.
We may state with little fear of contradiction that exploration of the
sign storm question has only barely begun. This is not too surprising when we
nsider that most of the more comprehensive simulation tools emerged in the 1970's,
This report draws on and is an extension of findings from papers
ented at four international conferences. Mr. Marsalek has assembled in this
- 155 -
-------
report the salient results of his research on implications in the use of the desien
storm concept in urban sewered system planning and design. The Hydraulics Research
Division of the National Water Research Institute at the Canada Centre for Inland
Waters has spearheaded a substantial research effort on urban hydrology, resultina
in products of considerable international value, mostly under the Canada-Ontario^
Agreement on Great Lakes Water Quality. In 1976, Mr. Marsalek assembled a
national state-of-the-art report on urban hydrological modeling and catchment
research in Canada(14) for the ASCE Program, which was reprinted in Canada(15)
and included in a Unesco publication.(*&)
The ASCE Urban Water Resources Research Council is indebted to
Mr. Marsalek and the Hydraulics Research Division of the National Water Research
Institute for their generous contribution of this report as a public service.
Processing, duplication and distribution of this Technical Memorandum
was supported by Grant No. ENV77-15668 awarded to ASCE by the U.S. National
Science Foundation. However, any opinions, findings, and conclusions or
recommendations expressed herein are those of Mr. Marsalek or this writer and
do not necessarily reflect the views of the National Science Foundation.
References
1. McPherson, M. B., "Urban Runoff Control Planning," ASCE, New York, N.Y.
118 pp., June, 1977. (Available as PB 271 548, at $5.50 per copy, from
the National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22161).
2. Espey, W. H., Jr., D. G. Altman and C. B. Graves, Jr., "Nomographs for
Ten-Minute Unit Hydrographs for Small Urban Watersheds," ASCE UWRR Program
Technical Memorandum No. 32, ASCE, New York, N.Y., 22 pp., December 1977
(NTIS: PB 282 158). '
3. McPherson, M. B., and G. F. Mangan, Jr., "ASCE Urban Water Resources Reseav
Program." j;Hyd.Div.. ASCE Proc., Vol. 101, No. HY7, pp. 847-855, July, 1975
4. McPherson, M. B., and G. F. Mangan, Jr., Closure to Discussion of "ASCE Urb
Water Resources Research Program," J.Hyd.Djv.. ASCE Proc., Vol. 103 NO avR0
pp. 661-663, May, 1977. ' * *»
5. Ramsseshan, S.t and P. B. S. Sarma, "Urban Hydrological Modeling and Catchmen
Research in India," ASCE UWRR Program Technical Memorandum No. 1HP-12 ASCII *
New York, N.Y., 21 pp., May, 1977. (NTIS: PB 271 300). '
6. McPherson, M. B., and F. C. Zuidema, "Urban Hydrological Modeling and
Catchment Research: International Summary," ASCE UWRR Program Technical
Memorandum No. IHP-13, ASCE, New York, N.Y., 48 pp., November, 197?
(NTIS: PB 280 754).
7. Lee1ere, Guy, and John C. Schaake, Jr., "Derivation of Hydrologic Frequeac
Curves," Ralph M. Parsons Laboratory for Water Resources and Hydrodynamics
Report No. 142, M.I.T., Cambridge, Mass. 02139, 151 pp., January, 1972.
8. Howard, Charles D. D., "Theory of Storage and Treatment-Plant Overflows '
J.Env.Engrg.Div.. ASCE Proc., Vol. 102, No. EE4, pp. 709-722, August
Discussions: Vol. 103, No. EE3, pp. 514-520, June, 1977. Closure: '
Vol. 104, No. EE2, pp. 369-371, April, 1978.
- 156 -
-------
9. Walesh, Stuart G., and Daniel F. Snyder, "Reducing the Cost of Continuous
Hydrologic-Hydraul ic Simulation," Water Resources Bulletin, (in press), 1979.
10* Geiger, W. F., S. A. LaBella and 6. C. McDonald, "Overflow Abatement
Alternatives Selected by Combining Continuous and Single-Event Simulations,"
pp. 71-79 in Proceedings. National Symposium on Urban Hydrology. Hydraulics
and Sediment Control, Report UKY BU111, University of Kentucky, Lexington,
386 pp., December, 1976.
11. Arnell, Viktor, "Analysis of Rainfall Data for Use in Design of Storm Sewer
Systems," pp. 71-86 in Preprints. International Conference on Urban Storm
Drainage. 11-15 April 1978, University of Southampton, England, U.K.
12. Urbonos, Ben, "Some Findings in the Rainfall -Runoff Data Collected in the
Denver Area," Flood Hazard News. Urban Drainage and Flood Control District,
Denver, Vol. 8, No. 1, pp. 10, 11 and 14, July, 1978.
13. Wenzel, Harry 6., "Evaluation of the Design Storm Concept," a paper to be
presented at the session on Urban Hydrologic Systems, Fall Meeting of the
AGU, San Francisco, December 4-8, 1978.
14. Marsalek, J., "Urban Hydrological Modeling and Catchment Research in Canada,"
ASCE UWRR Program Technical Memorandum No. IHF-3, ASCE, New York, N.Y.,
52 pp., June, 1976. (NTIS: PB 262 068).
15. Marsalek, J., Urban Hydrological Modeling and Catchment Research in Canada,
Technical Bulletin No. 98, Canada Centre for Inland Waters, Burlington,
Ontario, 52 pp., 1976.
1$. Unesco, Research on Urban Hydrology. Volume 1, Technical Papers in
Hydrology 15, Imprimerie Beugnet, Paris, 185 pp., 1977.
(Areawide planning agencies should have the following report In their
libraries: Setting the Course for Clean Water, "A Citizen's Guide to the
Section 208 Water Quality Management Program," National Wildlife Association,
Education Division, 1412 16th St. N.W., Washington, D.C. 20036, 64 pp., March,
1978).
- 157 -
-------
RESEARCH ON THE DESIGN STORM CONCEPT
(Addendum 4 of the ASCE Program report "Urban
Runoff Control Planning," June, 1977)
by
Jiri Marsalek
Hydraulics Research Division
National Water Research Institute
867 Lakeshore Road
Burlington, Ontario, Canada L7R 4A6
September, 1V78
- 158 -
-------
TABLE OF CONTENTS
Page
SECTION 1 - SUMMARY AND CONCLUSIONS 160
SECTION 2 - SYNTHETIC DESIGN STORMS 162
Background *..*........ 162
Chicago Design Storms 162
Figure 1 - Synthetic and Actual Storm Hyetographs .... 164
Table 1 - Values of Parameters Used for Chicago Method . 163
Illinois State Water Survey Design Storms 163
Table 2 - Maximum Hourly Rainfalls of Various
Return Periods 165
Table 3 - Median Rainfall Distribution of
Predominant Storms 165
Selection of Actual Events 165
Table 4 - Characteristics of Top-Ranked Actual Storms . . 166
Discussion . . 167
SECTION 3 - SIMULATED PEAK FLOWS 168
Background ..... 168
Actual Catchment Simulations ..... 168
Figure 2 - Recurrence Intervals of Runoff Peaks Simulated,
Malvern Catchment, Actual and Design Storms . . 169
Simulation Time Step 170
Hypothetical Catchment Simulations ...... 17Q
Figure 3 - Comparison of Simulated Peak Flows,
Hypothetical 26-Ha Catchment 171
Figure 4 - Comparison of Simulated Peak Flows,
Hypothetical 130-Ha Catchment 172
Discussion of Results ,. 173
Caveat 175
SECTION 4 - EFFECTS OF DETENTION STORAGE .......... 176
Background 17 g
Simulation of Storage Effects .......... 176
Table 5 - Detention Reservoir Characteristics ...... 177
Figure 5 - Simulated Effects of Storage,
Hypothetical Catchment ............ 178
Discussion .......................... 177
Figure 6 - Runoff Volumes Produced by Selected
Actual and Design Storms ........... 177
Table 6 - Ranks of Selected Actual Storms with
Respect to Runoff Peaks and Volumes
Produced by These Storms ...... 180
Table 7 - Examples of Transformation of
Frequencies by Storage ....... 180
SECTION 5 - WATER QUALITY SIMULATION CONSIDERATIONS 182
Background 182
Antecedent Rainfall Effects 182
Figure 7 - Effects of Dry-Weather Period,
Malvern Catchment 183
Figure 8 - Probability of Dry-Weather Periods,
Burlington, Ontario ......... 185
Discussion 184
SECTION 6 - REFERENCES 186
- 159 -
-------
SECTION 1
SUMMARY AND CONCLUSIONS
This report draws on and extends the findings presented in four recent
papers,(1~^) assembling in one document the salient results of research on
implications in the use of the design storm concept in urban drainage design and
planning. Analysis was restricted to sewered catchments in a southern Ontario
municipality.
Urbanization substantially increases the volumes and peak flows of
surface runoff. The cumulative effect of increased runoff volumes and localized
peaks then contributes to flooding of downstream areas. Such adverse effects of
urbanization are particularly pronounced in the case of catchments in which the
downstream part is developed and the upstream part is undergoing development.
This type of catchment is fairly common in southern Ontario.
To control the increase in stormwater runoff due to urban development
many government regulatory agencies have introduced criteria for urban drainage*
design. Such criteria require various degrees of control of runoff from areas
undergoing development.
In order to meet requirements stipulated by runoff control policies
the frequencies of runoff flows of various magnitudes need to be determined. in
streamflow flood frequency studies, the frequency of occurrence of various
floods is often determined directly from a flow record. Since adequate flow
records are virtually nonexistent in the case of urban catchments, runoff models
are employed to produce surrogate flow records from which the frequencies of
occurrence can be determined. Such an approach was taken here and the frequeueie
were determined in two ways. Firstly, runoff flows were simulated for design
storms of assigned frequencies of occurrence and it was assumed that the
frequencies of the runoff peak flows were identical to those ascribed to the
design storms. Secondly, runoff flows were simulated for a large number of
historical storms and the frequencies of occurrence of various runoff flows were
determined directly from the simulated runoff flow record. Since most sewer
systems are typically designed for return periods of from one to 10 years, such
a range of return periods is of primary interest here.
To establish the frequency of occurrence of runoff peak flows, continuous
runoff simulation was approximated by a series of single-event simulations for 27
selected actual storms. The selection of these storms, which effectively replaced
a 15-year rainfall record, was based on the ranking of all of the storms accordin
to their peak rainfall intensities for several durations. 8
In Section 3, runoff peak flows simulated for synthetic as well as
actual rainfall events are compared for an actual catchment and for several
hypothetical catchments patterned after typical urban developments in southern
Ontario. Although the results obtained are valid only for the conditions
studied, the comparisons give a general indication of the relationship between
synthetic and actual storms and demonstrate some shortcomings of an approach
based on a synthetic rainfall hyetograph. The analysis is restricted to runoff
peak flows on small and intermediate catchments (a maximum of 130-ha).
.160 -
-------
Comparison of runoff peaks simulated for two types of synthetic design
storms and for actual storms of nominally equal return periods produced widely
varying results. One type of synthetic storm produced runoff peak flows much
larger than those simulated for actual storms of corresponding return periods.
The use of another type of synthetic storm resulted in runoff peak flows that were
also larger than those simulated for the actual storms of corresponding return
periods.
The use of design storms simplifies preparation of rainfall data for
runoff calculations and drainage design. It appears plausible that these
subjectively defined storms could be replaced by historical storms which produced
runoff flows of certain frequencies of occurrence on similar catchments in a given
locality. These historical storms would then be used in future drainage design.
To identify such events, one needs to carry out either a continuous simulation of
runoff for a reasonable time period (10 to 20 years), or to carry out single
event simulations for a number of selected events. In the latter case, the
initial catchment conditions may be adjusted as necessary to account for antecedent
precipitation.
Selection of historical storms to be used for design is affected to some
extent by the characteristics of the urban catchment on which such a storm would
be applied. It is therefore recommended that a wide range of catchment parameters
be covered in runoff simulations serving for the selection of historical design
storms. For the design of storage, runoff volumes also have to be considered in
such an analysis. On the basis of simulated performance for a catchment, described
in Section 4, synthetic storms were found to be incapable of representing the true
volume, timing and multiple-peak nature of actual hyetographs. For the analysis
of storage, it is preferable to employ historical storms producing runoff events
of known frequencies of occurrence rather than to use a design storm. Implications
of the concept requiring that urban runoff be held to pre-development levels were
explored through a series of runoff simulations for different degrees of detention
storage.
The last Section of this report is devoted to some water quality
simulation considerations. Under the conditions explored, it is concluded that
the design storm concept and single-event runoff simulation are of limited use
for water-quality oriented design, because of statistical nonhomogeneity of runoff
quantity and quality events.
- 161 -
-------
SECTION 2
SYNTHETIC DESIGN STORMS
Background
Sizing of storm drainage conduits is universally based on the concept
of a design storm. In application, it is assumed that the frequency of
occurrence of a runoff peak flow is identical to that assigned to the associated
rainfall. Design storms have been traditionally represented as a. block rainfall
and more recently by synthetic hyetographs. Such hyetographs are typically
derived by synthesizing generalizations from a large number of actual events.
The concept of a design storm and its use in urban drainage design
and other applications has been a subject of considerable criticism. Particularly
criticized have been attempts to assign mean frequencies of probable occurrence
to storms of various intensities and durations, and the presumption of identical
frequencies of occurrence of coupled rainfall and runoff events has been
questioned because of the inherent statistical nonhomogeneity of rainfall and
runoff processes.(5) Although such criticism seems to be generally justified
the shortcomings of the design storm concept have been only tenuously
demonstrated with actual rainfall data, and only indirectly in conjunction with
runoff calculations. Included herein are findings from a study of the design
storm concent. Two synthetic storm configurations were analyzed, termed the
"Chicago"(6> and "Illinois State Water Survey (ISWS)"(7) design storms. The
Chicago design storm was singled out for attention because several Canadian
municipalities have adopted this type of storm representation as part of their
design criteria for urban drainage. The ISWS design storm was included because
of its association with a well-known computer simulation model developed for the
sizing of storm drains.
Chicago Design Storms
Formulation of the Chicago synthetic hyetograph was presented over
twenty years ago in August, 1957.(°) Adaptation of the method to local rainfall
data elsewhere has been reported, for example, in India,(8) the U.S.(9) and
Canada.(10) jhe Chicago method has been rather wid3ly incorporated in North
American practice because it can be readily derived from available rainfall
intensity-duration-frequency relationships and partly because of limited
alternative approaches. A contributing factor was undoubtedly its inclusion
in a widely used handbook. (H) When the method was presented it was criticized
on the grounds that it retained too many of the fallacies and empiricisms inherent
in the Rational Method to recommend its principle for general use by others.(12)
The fact remains that several Canadian municipalities have included this type of
design storm in their design criteria for urban drainage.
In an attempt to preserve correspondence with actual rainfall events
the Chicago method takes into account the maximum rainfalls of individual '
durations, the average amount of rainfall antecedent to the peak intensity and
the relative timing of the peak intensity. The first step in applying the
method is determination of the time antecedent to the peak intensity, expressed
as a dimensionless ratio. This time, tr, divides the hyetograph into two parts
and is defined as *
tr = tp/T
-162 -
-------
where tp is the elapsed time from the onset of rainfall to the peak intensity and
T is the total storm duration. Values of tr are determined individually for a
number of historical storms and their mean value is used for the design hyetograph.
The hyetograph intensities on either side of the peak are obtained from applicable
local intensity-duration-frequency relationships in. the form
1av
where
is the average maximum rainfall intensity over a duration
and the
constants a, b and c satisfy a fit to the data. Typically, one to six hours is
selected as the total storm duration, I. (However, the choice of T does not
affect the magnitudes of the peak rainfall intensity or the dimensionless time
to peak).
Synthetic hyetographs of the Chicago method type have been derived for
various assigned frequencies for a 15-year rainfall record from the station at
the Royal Botanical Gardens in Hamilton, Ontario. (13) These hyetographs were
adopted for the studies to be reported herein. Figure l(a) is an example, for
a "Two-Year Storm". Figure l(c) illustrates the typical departure of variations
in an actual storm from those represented in a synthetic formulation. (The
computed peak discharges for Storm No. 47 in Figure l(c) were for about a two-year
frequency). Parameter values used for the Chicago method are listed in Table 1.
TABLE 1
Values of Parameters Used
for Chicago Method
Return Period, Years
Parameters
tr:
a:
b:
c:
T, rain.:
1
0.48
22.6
0.78
5
180
2
0.48
29.9
0.80
6
180
5
0.48
43.7
0.84
7
180
10
0.48
55.9
0.86
8
180
jinnots State Water Survey Design Storms
In this procedure, maximum hourly rainfall depths are taken from local
data or from intensity-duration-frequency relationships for various return periods.
Individual hourly depths are then distributed over the selected storm duration in
accordance with normalized relations developed from Illinois data.O*1^,15) yOT
application elsewhere, actual storms are first divided into a number of groups in
accordance with the relative timing of the peak intensity. Distributions over time
are next determined for the predominant group of storms and their median
distribution is used for the design storm.
- 163 -
-------
150--
E
E
100-
ID
K-
50--
z
<
DC
JL
60
120 0 60
TIME .MINUTES
60
(a) CHICAGO "2-YEAR STORM" (b) ISWSM2-YEAR STORM" (c) ACTUAL STORM
No. 47
FIGURE 1-SYNTHETIC ANQ ACTUAL STORM HYETOGRAPHS
- 164 -
-------
As for the Chicago method, the data used to develop ISWS design storms
were from a 15-year rainfall record from the station at the Royal Botanical
Gardens in Hamilton, Ontario. Maximum hourly rainfall depths that had already
been identified in this record(13) were used, from which values for various
return periods were abstracted, Table 2.
TABLE 2
Maximum Hourly Rainfalls of Various Return Periods
(Royal Botanical Gardens, Hamilton)
Maximum Hourly
Rainfall, mm:
Return Period^ Years
_1 _2 _5 10
22 26 33 39
About 30 heavy recorded storms, which are further described in the next
subsection, were used to determine the temporal rainfall distribution. These
storms were divided into three groups in accordance with the distinctive part of
each storm in which the burst of peak rainfall intensity had occurred. The peak
intensity for a majority of the storms occurred within the last third of their
total duration. A median rainfall distribution was determined for this group and
expressed as
Rep - f(TCp)
where Rcp is the cumulative per cent of rainfall, T is the cumulative per cent
of storm time, and f is an empirical function. Numerical values of this
distribution, which was adopted for ISWS design hyetographs, are listed in
Table 3 and an example hyetograph (two-year return period) is shown in Figure l(b).
Again, there are important departures from actual storm variations, such as in the
example of an actual storm in Figure l(c).
TABLE 3
Median Rainfall Distribution of Predominant Storms
(Royal Botanical Gardens, Hamilton)
Cumulative Per Cent of Storm Time, Tcp
Cumulative
Per Cent of
Rainfall, R^t
0 10 JO
0 5 10
3CI 40 M)
15 22 30
60 70
39 56
80
86
90.
96
100
100
Selection of Actual Events
For analysis of certain types of urban runoff problems, particularly
those related to water quality, a very strong case can be made in support of
continuous rainfall-runoff-quality simulation. In some types of analysis, such
-165 -
-------
as for peak flow determination, continuous simulation can be effectively
approximated by using a series of single-event simulations. For the studies
reported here, simulations were made for selected actual storms that would be
likely to cause high runoff peak flows on urban catchments, and the findings are
therefore limited to that realm of analysis.
Selection of actual storms that would be most likely to produce high
runoff peak flows was facilitated by screening the rainfall record to segregate
all storms with either a total rainfall depth larger than 1.25-cm or a ten-minute
intensity larger than 1.5-cni/hr. A total of 34 storms met one or both of these
criteria. Next, the top 20 storm depths were identified for durations of 5-
10-, 15-, 30- and 60-mlnutes. Because a number of the storms contained multiple
martima, the segregation process yielded only 27 storms that net all the selection
criteria. For the purpose of establishing the frequency of occurrence of runoff
peaks on the catchmeats studied, these 27 storms were regarded as a suitable
replacement for the 15-year rainfall record. The basic characteristics of the 27
selected storms are summarized in Tabla 4.
TABLE 4
Characteristics of Top-Banked Actual Storms
(Royal Botanical Gardens, Hamilton)
Storm
Number
44
2
46
10
25
36
47
20
23
6
1
8
39
54
31
29
37
22
35
11
15
53
17
9
32
43
26
Means:
Total
Rainfall, mm
37.8
57.7
31.2
14.2
44.7
20.8
15.3
46.5
22.9
28.7
30.0
30.7
17.0
78.5
27.7
26.4
24.9
32.8
24.4
80.3
26.4
27.2
20.6
25.9
27.9
37.3
23.4
32.6
Duration,
hours
0.5
10.3
1.5
5.4
4.8
1.0
1.3
6.5
0.6
6.3
9.2
0.7
4.5
18.4
2.4
3.4
1.9
5.6
5.6
19.5
3.8
6.6
9.1
2.9
5.2
14.1
6.2
5.8
Antecedent
Dry Weather
Period, Days
8
2
2
6
3
1
1
3
1
6
3
1
3
8
0
10
1
7
2
4
4
5
6
8
18
2
0
4
5-Day Antecedent
Precipitation, mm
0.8
46.2
10.9
15.2
4.8
18.8
8.9
19.1
3.0
8.4
16.3
17.5
19.8
0.5
21.3
0.5
13.7
0.3
13.5
5.3
5.8
3.6
0.3
0
0 i
36.3 *
18.5
11.5
-166 -
-------
In the segregation of storms, the minimum inter-event time was taken as
three hours. That is, a storm event was defined as one where at least three hours
without rainfall occurred before and after the event. On this basis, the average
total rainfall depth was about 33-nm and the average storm duration was about six
hours for the storms selected, Table 4.
Of interest is the relationship between the antecedent dry-weather
period and the antecedent five-day precipitation of these heavy storms. Because
the values of these parameters indicated that catchments in the area studied would
have been fairly dry at the beginning of heavy storms, neglecting the effects of
antecedent precipitation on runoff from the associated storms appeared to be a
safe approximation. This observation contradicts to some extent one of the
objections to the use of design storms but at the same time removes a possible
limitation from the results to be presented.
Diecussion
Explained above were the reasons why it was assumed to be a safe
approximation to neglect the long-term effects of antecedent precipitation on
runoff simulations, for the rainfall record involved. Further justification for
ignoring such antecedent effects in this study is given in the next Section.
However, long-term antecedent conditions indicated by the raingage record from
Hamilton in Table 4 may well be unusual. Also, for small rural or very sparcely
urbanized catchments, not considered here, it would probably be hazardous to
neglect long-term antecedent influences.
The most important objections to the use of design storms of the type
involved here are that they: attempt to summarize variegated storm patterns in
a single hyetograph shape; assemble components of equal individual probability of
occurrence that have a quite different joint probability; and ignore the
possibility that antecedent conditions may vary considerably from storm to storm.
Described in the next Section is the computer model used in the simulations, which
has a single-event capability in the version employed. Therefore, it was
serendipitous to be able to omit the simulation of all antecedent rainfall
required in continuous simulation. That is, because minor long-term antecedent
influences were indicated in the rainfall record used, it was deemed reasonable
to use the single-event model without substantial modification. In what could be
the more usual case where long-term antecedent conditions are more pronounced, it
would be prudent to rely on continuous simulation. In sum, it must be recognized
that the effect of accounting fully for the influence of antecedent precipitation
on runoff for recorded storms, as opposed to its omission in design storms, was
not Investigated in this study. However, antecedent conditions were taken into
account in this study, when required, by adjusting the initial detention storage
and infiltration. That is, because the version of the computer model used does
not maintain water balances between storms, this was done externally to set
initial conditions when required. Little guidance for further refinement is
provided in the literature because the effects of antecedent conditions on urban
runoff are frequently discussed on the basis of conceptual models rather than on
the basis of actual observations.
- 167 -
-------
SECTION 3
SIMULATED PEAK FLOWS
Background
To avoid the shortcomings of the design storm approach, several
researchers have proposed converting the rainfall record into a runoff record
and then determining the frequencies of occurrence of various runoff flows from
the latter record. Such a conversion can be performed by means of urban runoff
models. In the study described here, a version of the Storm Water Management
Model (SWMM) was used for runoff simulation. SWMM is essentially a single-event
urban runoff model, although it has recently been run in a continuous simulation
mode on a test catchment.(16) Its development was first reported in 1971 for
the U.S. Environmental Protection Agency,(17) it is continuously upgraded by the
U.S. EPA,(18) and it has been studied at length in Canada.(19)
The 27 top actual storms and the synthetic storms described in the
preceding Section were used in SWMM simulations for an actual catchment and a
group of hypothetical catchments. Simulations were performed in a single-event
mode.
Actual Catchment Simulations
The test catchment, known as the Malvern catchment,(2°) is a modern
residential area of 23-ha with single-family housing units. The catchment is
fully developed, its imperviousness is 30 per cent, and it is drained by separate
storm sewers. For modeling purposes, the catchment was subdivided into ten
subcatchments varying in size between 0.9-ha and 4.0-ha. The sewer system was
represented by 21 pipes varying in size from 0.3-m to 0.83-ra. The Malvern
catchment was instrumented in 1973 and a large number of rainfall-runoff events
have since been recorded there.
Before proceeding with the simulations to be reported in this
subsection, SWMM was calibrated and verified for the Malvern catchment. About
25 rainfall-runoff events were available for this purpose, for which the return
periods of the two highest observed runoff peak flows were about 1-year. On the
average, the calibrated model underestimated observed runoff volumes by 1% and
observed peak flows by 5%, with standard deviations about the means of 12% and
16% respectively. The raingage serving the Malvern catchment is located about
10-km east of the raingage at the Royal Botanical Gardens, the source of the
15-year record from which the 27 storms cited in Table 4 were obtained and from
which the design storms were derived.
Design storms determined according to the Chicago method and the ISWS
method were then used with the calibrated SWMM, for assigned return periods of
1-year, 2-years, 5-years and 15-years. By definition, the return periods of the
resulting runoff peak flows were assumed to be identical to those assigned the
design storms. For the actual storms, the simulated runoff peak flows were
ranked and their frequency of occurrence determined therefrom. The results of
the Malvern catchment simulations are presented in Figure 2. There are large
discrepancies between the runoff peaks for the historical storms and for the
design storms having the same nominal frequencies. This discrepancy will be
discussed after the results from the simulations for the hypothetical catchments
are presented.
-168 -
-------
0.2
o
o>
I I
2 3 456789 10
AVERAGE RETURN PERIOD, YEARS
15
LEGEND - Chicago Design Storms +
ISWS Design Storms x
Actual Storms
FIGURE 2- RECURRENCE INTERVALS OF RUNOFF
PEAKS SIMULATED, MALVERN CATCHMENT,
ACTUAL AND DESIGN STORMS
- 169 -
-------
Simulation Time Step
The simulation time step was selected on the basis of previous
studies on the Malvern catchment with SWMM.^20^ As a part of those simulation
studies, the effect of the time step on the reproduction of observed hydrographs
was investigated. For high intensity storms, the best reproduction was
obtained for simulations with time steps of one and two minutes. Consequently,
these two time steps were adopted for all the simulations discussed in this
report. Specifically, a one-minute time step was used for storms with a
duration of two hours or less and a two-minute time step was used for storms of
longer duration. Use of such small time steps makes possible the determination
of runoff peak flows from generated hydrographs without any need for interpolation
Hyetographs for the design storms were accordingly discretized into
one or two minute intervals. For the one-hour duration ISWS design storm a
one-minute interval was used, whereas a two-minute interval was used for the
longer Chicago design storm, which was assigned a duration of three hours as
specified by its proponents. For all simulations reported, the simulation time
step was therefore the same as the time interval for the input rainfall data.
Hypothetical Catchment Simulations
Physical catchment parameters strongly influence runoff simulations and
can to some extent influence the selection of rainfall inputs. In order to
explore the effects of such parameters, runoff flows were simulated for a series
of nine hypothetical catchments of widely varying characteristics. These
catchments were patterned after some typical urban catchments located in modern
residential developments in Ontario. Three catchment sizes were used: 26-ha
52-ha and 130-ha. The drainage was maintained at about the same density for all
three cases. Catchment imperviousness was set at three levels: 15%, 30% and
45%. The last two levels are typical for modern residential areas in Ontario.
Because of the more extended sewer system, the lag time for the 130-ha catchment
was longer by about 7 to 14 minutes than that for the 26-ha catchment.
Two types of rainfall inputs were used in runoff simulations: design
storms obtained via the Chicago and ISWS methods described earlier, for four
return periods; and the 27 selected actual storms.
Return periods and the associated peak runoff flows simulated for the
various actual and synthetic storms are plotted in Figure 3 for the smallest
hypothetical catchment (26-ha), separately for each of the three levels of
imperviousness. Figure 4 is a companion illustration, for the largest
hypothetical catchment (130-ha). These and similar plottings for the intermediate
catchment size revealed an attenuation in peak flows per unit area with an increas
in catchment size. This attenuation in peak flow was fairly consistent and
represented about a 13% reduction for the largest (130-ha) catchments over the
smallest (26-ha) for otherwise identical characteristics. It is conceivable that
even larger differences could be encountered in practice, depending on the
relation of the lag times of the catchments involved. The attenuation in peak
flows with increase in catchment size is attributed to the larger concentration
times noted above. The degree of this attenuation increased with the amount of
imperviousness.
Comparison of runoff peaks simulated for actual and synthetic storms
yielded interesting results. For all four return periods, both design storms
.170 -
-------
10 . 1MPERVIOUSNESS, 15%
.00
IMPERVIOUSNESS, 30°>5
.20
.10
IMPERVIOUSNESS, 45%
8 10
2 3 4 5 6
AVERAGE RETURN PERIOD, YEARS
LEGEND - Chicago Storms
ISWS Storms
Actual Storms A
FIGURE 3 -COMPARISON OF SIMULATED PEAK FLOWS,
HYPOTHETICAL 26-HA CATCHMENT
- 171 -
-------
IMPERVIOUSNESS
IMPERVIOUSNESS, 30%
O
IMPERVIOUSNESS, 45%
2 34568
AVERAGE RETURN PERIOD, YEARS
LEGEND Chicago Storms
ISWS Storms
Actual Storms A
FIGURE 4-COMPARISON OF SIMULATED PEAK FLOWS,
HYPOTHETICAL J30-HA CATCHMENT
- 172 -
-------
produced flows larger than those produced by the actual storms for corresponding
return periods. This overestimation was particularly large for the design
storms derived via the Chicago method, which produced peak flows for all three
sizes of catchments that were three-fourths larger, on the average, than those
produced by the corresponding actual storms. The design storms derived via the
ISWS method produced better results, with simulated peak flows only one-fifth
higher, on the average, than those for the corresponding actual storms.
Recall that the ISWS design storms used had a delayed pattern, dictated
by the predominant characteristics of the actual storms from which they were
derived. In an attempt to explore the effect of storm pattern, some simulations
were also performed using the advanced pattern recommended by the ISWS for
drainage design.(^ In contrast to the delayed pattern, starting at the 2-year
level the simulated peak flows for the advanced pattern were lower than for the
actual storms and corresponded, In general, more closely to the actual peak
flows, except at the ten-year level, where they were about one-fifth below the
actual storm rates. The effect of pattern shift was non-linear, as indicated
below, for simulations with 307. tmperviousness:
Ratio of Peak Flow for Advanced Pattern
Return Period, to Peak Flow for Delayed Pattern
Years 26-ha Catchment 130-ha Catchment
1 0.91 0.82
2 0.84 0.80
5 0.69 0.69
10 0.61 0.62
As would be expected, storm pattern evidently has an important influence on the
magnitude of catchment peak flow.
Discussion of Results
Three shortcomings affect simulated hydrographs obtained through use of
Chicago design storms:
All the values represented by a particular intensity-duration-frequency curve
are implied to belong to the same storm, whereas such curves are typically
obtained by a synthesis of data from a large number of storms.
The intensity-duration curves are extrapolated beyond the smallest reported
data duration of 5-minutes, yielding a peak rate exceeding the 5-minute
intensity by about 607.. (However, these high intensities are reduced
somewhat when the hyetograph is divided into constant time intervals).
The description of the time of the peak intensity by a single tr-value,
which is an average of all the tr-values derived from selected storms, is
questionable in view of the probabilistic nature of this parameter. Analysis
of the selected storms reported here yielded a mean tr-value of 0.48 with a
standard deviation about the mean of 0.32, Indicating a large scatter.
While better results were obtained with the ISWS method, there is some
degree of arbitrariness in the definition of these design storms, particularly
in the choice of the storm duration, which affects the magnitude of rainfall
intensities. A one-hour duration had been recommended by the ISWS on the basis
- 173 -
-------
of findings from some simulations of runoff from several urban catchments.
The effect of ISWS design storm duration was investigated in the present study.
For the delayed pattern, when the design storm duration was reduced to one-half
of an hour, the runoff peak flows increased about one-third over those obtained
for a one-hour duration, such as the cases plotted in Figures 3 and 4 for a
one-hour duration. When the design storm duration was raised to five hours, the
runoff peak flows were much smaller than those obtained for a one-hour duration.
The durations of the actual storms used here (e.g., the ones in Table 4) do not
support the hypothesis of a fixed value of one-hour duration. It is evident that
the comparatively better performance of the ISWS design storm for a one-hour
duration reported here may be largely coincidental.
Findings from this study strongly suggest that much more attention
should be given to the rainfall input than has been the normal practice. Results
obtained for the two synthetic design storms differed from each other and from
those for actual storms. The uncertainty in simulated runoff peaks caused by
the choice of rainfall input appeared to be larger than the uncertainty inherent
in the simulation process.
Recall that the actual storms used for runoff simulations were selected
on the basis of the rank of peak intensities for durations of 5-, 10-, 15-, 30-
and 60-minutes. As a means of investigating the efficiency of this process,
correlations were examined between the ranks of peak rainfall intensities for
individual durations and the associated runoff peak flows using the Spearman
rank correlation coefficient. Using the values for all 27 actual storms, the
correlation coefficients were larger than 0.545, which indicates a rank
correlation significant at a 0.01 level of confidence.(21) The segregation of
peak intensities therefore appeared to have been a good selection criterion for
the identification of important historical storms which would be associated with
larger runoff peaks.
Peak flow and peak storm intensity are assumed to be directly related
in the Rational Method. The highest correlation coefficients obtained for
simulated runoff peak flows versus peak rainfall intensities (5-, 10-, 15-, 30-
and 60-minutes duration) of actual associated storms varied between 0.629 and
0.734. This means that, at best, only about half of the linear variation in the
runoff peaks could be explained by the linear variation in the rainfall intensity.
Evidently, other parameters of the rainfall distribution are also important in
the generation of realistic runoff peak flows.
Although the results presented here are valid only for the conditions
described, the methodology used in the selection of actual storms and the use of
frequency graphs of runoff flows may have a general applicability. We plan to
apply these methods to studies of other situations. Graphs of runoff flow
frequencies, analogous to those in Figures 3 and 4, could be used to obtain quick
estimates of runoff peaks from new urban developments or for checking design values,
Effects of storage reservoirs on runoff peaks were not considered in
this Section. As will be shown in the next Section, such storage transposes the
runoff peak flow frequency curve into a series of smaller values.
- 174 -
-------
Caveat
The version of SWMM employed does not accommodate surcharging
realistically in a hydraulic sense. Prior simulations of catchments of the
type involved in this study revealed surcharging of some reaches at the 2-year
to 5-year peak flow level. Surcharging of entire systems would be expected
somewhere between the 5-year and 15-year peak flow in this type of system.
Because incorporation of the WRE transport module, which accommodates surcharging
realistically, would have required considerable reprogramming for the computer
available, and its use would have been much more costly, an expedient was
adopted instead. Conduit diameters, Including those of the Malvern catchment,
were artificially enlarged to sizes that would handle all flows to be simulated
without surcharging. Use of this expedient naturally casts some doubt on the
reliability of the results. However, because all simulations reported here
employed this expedient, it is felt that the comparative results are nevertheless
valid.
- 175 -
-------
SECTION 4
EFFECTS OF DETENTION STORAGE
Background
Various methods of runoff control are employed in modern innovative
design of urban drainage. Use of detention storage is one of the most important
methods and can be used to achieve almost any degree of runoff control, provided
that the costs of such facilities are not prohibitive. In Canadian urban
drainage practice, use of stormwater detention ponds has become popular in
recent years and many storage facilities of this kind have been constructed.
Some of these ponds have a multiple-purpose use, serving not only for runoff
control and reduction of drainage costs but also for recreation. These
detention ponds are frequently constructed as small reservoirs with gently
sloping embankments and drop-inlet spillways.
The hydrological design of detention ponds is not well established.
Regardless, in urban conditions where real damage would occur as a result of
flooding, detention ponds should be designed in a manner analogous to the design
of flood control projects.(22) por SOme urban conditions, the damage would be
little more than a nuisance and often a rather arbitrary decision is made to
limit the probability of this nuisance to some acceptable level. In either case,
however, the hydrological techniques employed should yield information adequate
for the definitions of flood frequencies before and after control.v22) As
discussed later, an approach based on the design storm concept does not meet
this criterion.
One of the most stringent runoff controls is imposed by ordering that
there shall be a zero increase in stormwater runoff due to urban development, in
other words, the runoff peak flows from a developed catchment have to be
restricted to their predevelopment level for rainfalls of certain frequencies of
occurrence. In one case reported in the literature, such a standard was enforced
for rainfalls having a 2-year or shorter return period.(23)
Simulation of Storage Effects
The concept of a zero increase in runoff is not as well defined as it
would appear. Technical implications of this runoff control policy are discussed
in this Section. Using historical storms, the recurrence interval of runoff peak
flows was analyzed for the hypothetical catchment of 26-ha and an imperviousness
of 30%, described earlier. Ten selected runoff hydrographs for this hypothetical
test catchment were then routed through detention storage and, by providing a
sufficient storage capacity, runoff peak flows were reduced to the predevelopment
level for a range of return periods.
Characteristics of the three levels of reservoir storage employed in
the simulations are listed in Table 5. In all three cases, the reservoir was
assumed to be located in the vicinity of the catchment outlet.
- 176 -
-------
TABLE 5
Detention Reservoir Characteristics
Reservoir Designation
R-l
R-2
R-3
Volume,
cu.m.
200
680
1310
500
1650
3020
660
2160
3900
Depth,
m.
0.3
0.9
1.5
0.3
0.9
1.5
0.3
0.9
1.5
Outflow,
cu.m/sec.
0.13
0.74
0.96
0.13
0.74
0.96
0.13
0.74
0.96
Routing of runoff hydrographs through the detention reservoirs was
accomplished by using the modified Puls method.(2*) The outflow hydrograph peaks
were then used to establish new peak flow versus recurrence interval values,
plotted in Figure 5. Included also in Figure 5 are the values from Figure 2,
designated in Figure 5 as "urbanized, no control". For reference, results for
"non-urbanized" conditions are also given in Figure 5.
The peak outflow discharges from storage were affected not only by the
magnitudes of the inflow peaks and the storage volumes, but also by the total
volume of inflow and its time distribution. It can be deduced from Figure 5 that
the criterion of a zero increase in runoff with urbanization is essentially
satisfied for the cases of the two larger reservoirs, where the peak rates are
practically identical with those for the undeveloped catchment, particularly for
recurrence intervals of design interest.
Discussion
Among the various means of runoff control, only detention storage and
its effects on catchment runoff were considered. By providing a sufficient
storage capacity, runoff from an urbanized catchment was reduced to about the
predevelopraent level for a wide range of recurrence intervals (see Figure 5).
The design-storm concept was found to be of questionable use in this
analysis. Multiple-peak storms may produce higher outflow peaks from storage,
if a second consecutive peak arrives when the storage facility is essentially full,
The outflow hydrograph of the Chicago design storm will be affected by the
tr-value. Peak runoff rates from storms with smaller tr-values will be more
attenuated than those with larger tr-values because for smaller tr-values the
inflow peak arrives when the available storage capacity is larger than it would
be for larger tr-values. Special tests were performed using Chicago design
storms with 1-year, 2-year and 5-year average return periods for all three
detention reservoir cases. The resulting runoff peak flows were all higher than
those indicated in Figure 5 for the historical storms ("urbanized, with storage")
with departures of more than 100% for the 1-year design storm and as little as
about 107. for the 5-year design storm. These results were not unexpected, as may
be seen in Figure 6, which Is a plot of frequency of flow volumes without storage
for the related simulations (middle graph of Figure 3). In Figure 6 the
- 177 -
-------
CO
o>
V)
O
0.05-
LU
CL
_
O
z
Z)
cr
LEGEND
Urbanized, No Control A
Urbanized, With Storage
Rl A R2
R3 0
f
Non- Urbanized
1 8
2 3 4 56789 10
AVERAGE RETURN PERIOD,YEARS
15 20
FIGURE 5-SIMULATED EFFECTS OF STORAGE, HYPOTHETICAL CATCHMENT
(26-ha, 30% Imperviousness)
-------
300
fO
~ 200 f
O
>
£ 100
UJ
u.
O
o:
LEGEND
Actual Storms
Chicago Design Storms
ISWS Design Storms
0.5
H 1 I I I I
5 6 7 8 9 10
AVERAGE RETURN PERIOD,YEARS
15 20
FIGURE 6-RUNOFF VOLUMES PRODUCED BY SELECTED ACTUAL AND DESIGN STORMS
(26-ha, 30% Imperviousness)
-------
departures in magnitude of runoff volumes between those for the Chicago design
storms and those for the historical storms also decrease with rarer return period
even though no routing through storage is involved.
Peak flow rates and magnitude of flow volumes are not highly correlated,
as indicated in Table 6, where the top 15 peak flow rates from the 27 actual
storms are associated with a seemingly random ranking of runoff volumes. During
this study it was noted in about half of all cases that storm hydrographs are
transformed differently by storage, with the result that frequencies of occurrence
of Inflows to storage and outflows from storage do not agree. Two of the most
obvious examples of this change in frequencies are given in Table 7.
TABLE 6
Ranks of Selected Actual Storms with Respect to Runoff
Peaks and Volumes Produced by These Storms
Storm
Number
44
2
46
10
25
36
47
20
23
6
1
8
39
54
31
Rank of the Runoff
Peak Produced by
the Storm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Rank of the Runoff
Volume Produced by
the Storm
4
2
6
15
3
12
14
5
8
7
9
10
13
1
11
TABLE 7
Examples of Transformation of
Frequencies by Storage
Storm
Number
(Table 4)
10
54
Average Return Period. Years
Without
Storage
4
1
Outflow From
Reservoir R3
1
8
The above indications substantiate the view that the design storm concept
was found to be of questionable use in this analysis of storage effects.
- 180 -
-------
The requirement of a zero runoff increase due to urban development is
a political solution to flooding problems created by progressive urbanization,
and under some circumstances this requirement may have little technical
justification. Typically, such a requirement is specified for a design storm
of a certain assumed frequency. While some control of urban runoff from
urbanizing catchments is mandatory where increased runoff would contribute to
flooding of downstream areas, this purpose may not be best served by the concept
of zero increase in runoff due to urban development. As interpreted by various
governmental agencies at the present time, this concept appears to be vaguely
defined and impractical. Runoff control measures required under this policy
apply to events of more or less arbitrarily specified frequency of occurrence
typically expressed in terms of design storms.
Urban runoff can be controlled by means of storage to a more Or less
arbitrary degree, including the predevelopment level. For the analysis of
storage, it is preferable to employ historical storms producing runoff events
of reliable frequencies of occurrence rather than to use a design storm. The
Chicago storm, in particular, is not capable of representing the true volume,
timing, and multiple-peak nature of actual hyetographs.
It is recommended that the zero increase in runoff concept be replaced
by a runoff policy in which the degree of runoff control is determined from a
cost-benefit analysis of the control measures. The entire basin and the main
stream have to be considered in such an analysis, since localized runoff control
by detention and uncontrolled, poorly-timed releases of stored runoff may still
lead to flooding in the main stream.
- 181 -
-------
SECTION 5
WATER QUALITY SIMULATION CONSIDERATIONS
Background
Consideration of the water quality aspects of urban drainage discharges
represents a new facet of urban drainage design. As "point" pollution sources
are abated and become less threatening, "non-point" sources of water impairment
become relatively more significant. Among such non-point sources, urban runoff
has been found to be particularly important. For example, a recent study sub-
stantiates that urban runoff can convey high pollution loads, and under certain
conditions can control water quality in receiving waters for extended time
periods.(25)
Some water quality considerations have to be recognized early in the
planning of urban drainage. Such considerations are related to pollution control
at the source and include the planning of land use, type of development and
extent of natural drainage.
Water-quality oriented design of urban drainage is somewhat hindered by
a lack of understanding of runoff-quality processes and by a lack of
authoritative water quality criteria for use in such a design. Ideally, water
quality criteria should be defined as receiving water criteria. In that case,
a given level of water quality (e.g., pollutant concentration) has to be
attained during critical storms and under critical conditions in the receiving
water body. Extensive data needs and complex modeling of the environmental
response retard wider application of receiving water criteria. More often,
effluent control criteria are imposed in Canada. Such criteria can be stated in
terms of annual allowable loads, number of events (e.g., annual average number of
overflows from a combined sewer system), controls for specified events, or maximum
allowable pollutant concentrations, or by prescribing specific control measures
for an area or at an effluent.(2°)
Although the best estimates of runoff quality are obtained from
extensive local field data, such data are virtually nonexistent and quality
simulation models have to be employed. Such simulation models again require
precipitation input data. These data, however, differ from those used in the
hydraulic design of urban drainage. Before discussing the nature of these
differences, mention should be made of the relative accuracy of quantity and
quality models. In each case, the uncertainty in the simulated output depends
upon the uncertainties in the input and in the model description of the processes
which transform inputs into outputs. Uncertainties in precipitation inputs are
comparable for both cases. However, uncertainties in contaminant inputs and in
the descriptions of the actual processes lead to far greater errors in simulated
quality outputs than in simulated quantity outputs.
Antecedent Rainfall Effects
The magnitude of pollution loads conveyed by urban runoff events depends
strongly on the initial conditions of the event, which are thought to be
characterized by the amount of pollutant accumulated on the catchment surface
during the antecedent dry-weather period. This point is illustrated in Figure 7,
which contains runoff hydrographs and pollutographs from the Malvern catchment
for two storms of similar volume and intensity of rainfall. The two hydrographs
- 182 -
-------
2 0
i 10
-§ 20
2 J
J]
2 0
4
OC
4
u.
O
.2
400
200
oc
t-
ui
o
8
U.
O
Z
20 40 60
TIME, MINUTES
.2
800
400
z
UJ
U
Z
o
U
20 40 60
TIME, MINUTES
lo) ONE-DAY DRY-WEATHER PERIOD
Legend
SS
COD
lb) SIXTEEN-DAY DRY-WEATHER PERIOD
FIGURE 7-EFFECTS OF DRY-WEATHER PERIOD,
MALVERN CATCHMENT
- 183 -
-------
agree quite well but the storm with the longer antecedent dry-weather period,
Figure 7(b), produced pollution loads about five times higher than those conveyed
by the other storm, Figure 7(a). Further changes in the magnitude of event
pollution loads might be introduced by street sweeping that effectively reduced
the accumulation of pollutants on the catchment surface between storms.
One possible way to estimate the antecedent dry-weather period is to
analyze the historical precipitation record and to derive therefrom
probabilities of dry-weather durations for individual events. An example of
such a relationship is given in Figure 8. For a catchment in Burlington, Ontario,
a five-year precipitation record was analyzed for the duration of dry weather
period considering only storms with rainfall sufficient to wash off the catchment
surface (in this case, storms with rainfall larger than 2.5-mm). If the
pollutant accumulation rates are known or determined from street sweeping
experiments, the total pollution loads per event and their probability of
occurrence can be determined from Figure 8 by multiplying the accumulation rates
by the number of dry days. Note that this approach assumes that the distribution
of dry periods does not depend on storm characteristics and that any antecedent
storm would have completely washed off the catchment. Such assumptions appear to
be acceptable when considering the large uncertainties involved in runoff quality
computations.
Discussion
Under the circumstances described above, the design storm concept and
single event simulation are of limited use for water-quality oriented design.
Storms producing high runoff flows may produce relatively low pollution loads and
vice versa. Consequently, continuous simulation of runoff quality, quality
control and associated costs should be employed using historical precipitation
data. Such simulations offer a good basis for the selection of the most cost-
effective control alternative meeting water quality criteria. As for design peak
flows, discussed earlier, continuous simulation might be reduced to indicator
catchments, with abbreviated simulations indicated thereby successively applied
to the more numerous other catchments in the jurisdiction involved.
In summary, water-quality oriented drainage design is a new idea which
has not yet gained wide acceptance. Quality considerations related to source
control have to be undertaken in the planning process. Among these considerations
one could name land use, type of development and extent of natural drainage.
Precipitation data requirements for water-quality oriented design are virtually
identical to those for the planning of urban drainage. Subsequently, the quality
design is finalized by means of detailed simulations for selected events of which
the initial conditions have been determined. The design storm concept has little
application in quality design because of statistical nonhomogeneity of runoff
quantity and quality events. Such nonhomogeneity is illustrated by the data
plotted in Figure 7. Typically, historical rainfall events are used in quality
design. Additional data, such as rainfall chemistry and atmospheric fallout
rates may be useful.
- 184 -
-------
EVENT RETURN
5 1 1/12 1/52 PERIOD, (YEARS)
t i t
25i
£20
o
o
DC 15
UJ
Q_
a:
UJ
x _
ui
i
cc 5
I I I I I l I I III ii I I ill
l l l l l l l l l I I l l i l II
0.1 .1 1 10 50 90 99 9999
EXCEEDANCE PROBABILITY (%)
FIGURE 8- PROBABILITY OF DRY-WEATHER PERIODS,
BURLINGTON, ONTARIO
' 185 -
-------
SECTION 6
REFERENCES
1. Marsalek, J., "Runoff Control on Urbanizing Catchments," pp. 153-161 in
Proceedings, Amsterdam Symposium on Effects of Urbanization and
Industrialization on the Uydrological Regime and on Water Quality,
October, 1977, International Association of Hydrological Sciences, IAHS-
AISH Publication No. 123, 1977.
2. Watt, W. Edgar, and Jiri Marsalek, "What the Practising Urban Hydrologist
Needs from the Hydrometeorologist," pp. 15-23 in "Preprint Volume,"
Second Conference on Hydrometeorology, October 25-27, 1977, Toronto,
Ontario, American Meteorological Society, Boston, Mass.
3. Marsalek, J., "Comparison of Runoff Simulations for Actual and Synthetic
Storms," pp. 100-116 in special release of "Storm Water Management
Modeler," the news release bulletin of the SWMM Users Group, U.S. EPA,
Washington, D.C., January, 1978.
4. Marsalek, J., "Synthesized and Historic Design Storms for Urban Drainage
Design," a paper presented at the 11-15 April 1978 International Conference
on Urban Storm Drainage, University of Southampton, England.
5. McPherson, M. B., "Special Characteristics of Urban Hydrology," pp. 239-255
in Prediction in Catchment Hydrology. Australian Academy of Science,
Canberra, ACT, 1975.
6. Keifer, Clint J., and Henry Hsien Chu, "Synthetic Storm Pattern for
Drainage Design," J.Hyd.Div.. ASCE Proc., Vol. 83, No. HY4, pp. 1-25,
August, 1957.
7. Terstriep, M. L., and J. B. Stall, The Illinois Urban Drainage Area
Simulator. ILLUDAS. Bulletin 58, Illinois State Water Survey, Urbana,
90 pp., 1974.
8. Bandyopadhyay, M., "Synthetic Storm Pattern and Run-off for Gauhati,
India," J.Hyd.Div.. ASCE Proc., Vol. 98, No. HY5, pp. 845-857, May, 1972.
9. Preul, H. C., and C. N. Papadakis, "Development of Design Storm Hyetographs
for Cincinnati, Ohio," Water Resources Bulletin. Vol. 9, No. 2, pp. 291-300,
1973.
10. Mitci, Constantin, "Determine Urban Runoff the Simple Way." Water & Wastes
Engineering. Vol. 11, No. 1, pp. 24-26 and 35-36, January, 1974.
11. Design and Construction of Sanitary and Storm Sewers. ASCE Manuals and
Reports of Engineering Practice No. 37 (WPCF Manual of Practice No. 9),
ASCE, New York, N.Y., 332 pp., 1969.
12. McPherson, M. B., "Discussion of Synthetic Storm Pattern for Drainage
Design," J.Hyd.Div.. ASCE Proc., Vol. 84, No. HY1, pp. 49-57, February, 1958.
13. M. M. Dillon Ltd., "Storm Drainage Criteria Manual for the City of Burlington,"
unpublished report, Toronto, Ontario, 1977.
- 186 -
-------
14. Huff, F. A., "Time Distribution of Heavy Rainfall in Storms," Water
Resources Research. Vol. 3, No. 4, pp. 1007-1019, 1967.
15. Vogel, John L., and Floyd A. Huff, "Heavy Rainfall Relations Over Chicago
and Northeastern Illinois," Water Resources Bulletin. Vol. 13, No. 5,,
pp. 959-971, October, 1977.
16. Jewell, T. K., T. M. Nunno and D. D. Adrian, "Methodology for Calibrating
Storrawater Models," J.Env.Engrg.Div.. ASCE Proc., Vol. 104, EE3, pp. 485-
501, June, 1978.
17. Field, Richard, and John A. Lager, "Urban Runoff Pollution Control
State-of-the-Art," J.Env.EngrR.Div.. ASCE Proc., Vol. 101, No. EE1,
pp. 107-125, February, 1975.
18. Torno, Harry C., "Storm Water Management Models," pp. 82-89 in Urban
Runoff. Quantity and Quality. ASCE, New York, N.Y., 1975.
19. Proctor and Redfern, Ltd., and James F. MacLaren, Ltd., Storm Water
Management Model Study. Volume II, Research Program for the Abatement of
Municipal Pollution under Provisions of the Canada-Ontario Agreement on
Great Lakes Water Quality, Research Report No. 48, Environment Canada,
Ottawa, 148 pp., September, 1976.
20. Marsalek, J., Malvern Urban Test Catchment. Volume I, Research Program for
the Abatement of Municipal Pollution under Provisions of the Canada-Ontario
Agreement on Great Lakes Water Quality, Research Report No. 57, Environment
Canada, Ottawa, 55 pp., 1977.
21. Siegel, S., Nonparametrlc Statistics for the Behavioral Sciences. McGraw-
Hill, New York, N.Y., 1957.
22. Linsley, R., and N. Crawford, "Continuous Simulation Models in Urban
Hydrology," Geophysical Research Letters. AGU, Vol. 1, pp. 59-62, May, 1974.
23. McCuen, Richard H., "A Regional Approach to Urban Storm Water Detention,"
Geophysical Research Letters. AGU, Vol. 1, No. 7, pp. 321-322, November,
1974.
24. Chow, V. T., Editor, Handbook of Applied Hydrology. McGraw-Hill, New York,
N.Y., 1964.
25. Colston, Newton V., and Anthony N. Tafori, "Urban Land Runoff Considerations,"
pp. 120-128 in Urbanization and Water Quality Control. American Water
Resources Association, Minneapolis, Minnesota, 1976.
26. Weatherbe, D. G., "Urban Drainage Planning," Chapter 2 in Manual of
Practice on Urban Drainage (Draft), Canada-Ontario Agreement on Great
Lakes Water Quality, Environment Canada, Ottawa, March, 1977.
- 187 -
«U.S. GOVERNMENT PRINTING OFFICE: 1078 620-007/3737 1-3
-------
TECHNICAL REPORT DATA
(Please read Instruction* on the reverse before completing)
1. REPORT NO.
EPA - 600/9 - 78 - 035
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
5, REPORT DATE
October 1978
Urban Runoff Control Planning
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Murray B. McPherson
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
ASCE Urban Water Resources Research Program
23 Watson Street
Marblehead, Mass. 01945
10. PROGRAM ELEMENT NO.
\. £6NtRACT/GRANT NO.
N/ A
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA 600/Z
B. SUPPLEMENTARY NOTES
Section 208 of Public Law 92-500 (Federal Water Pollution Control Act of 1972)
encourages areawide planning for water pollution abatement management, including
urban runoff considerations where applicable. Areawide studies are under way or
planned in just about every metropolitan area. Deadlines for initial areawide
reports are not far off, and it is expected that many of the agencies preparing
reports are presently resolving their projected activities beyond the current
first planning phase. This report has been prepared to assist agencies and their
agents that are participants in the preparation of areawide plans, from the stand-
point of major urban runoff technical issues in long-range planning. Emphasized
Is the importance of conjunctive consideration of urban runoff quantity and quality
and the need to development a factual basis that will support expected reliability
of performance of proposed actions and programs. While not intended as a handbook
for urban runoff control planning, this report delves into some important technical
Issues that are often slighted or poorly handled, such as the utilization of simu-
lation. Recognizing that the ultimate test of any plan lies in its implementation,
topics are viewed from the perspective and experience of the local government level
where Implementation takes place.
Key WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED.TERMS C. COSATI Field/Group
Urban Runoff
Master Planning
Urban Hydrology
Storm Water Runoff
Combined Sewer Overflows
13 B
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. Oi- PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
aaao-t it-73)
------- |