EPA-600/2-76-275
October 1976
Environmental Protection Technology Series
STORM WATER MANAGEMENT MODEL:
LEVEL I • PRELIMINARY SCREENING PROCEDURES
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-600/2-76-275
October 1976
STORM WATER MANAGEMENT MODEL
LEVEL I
PRELIMINARY SCREENING PROCEDURES
By
James P. Heaney
Wayne C. Huber
Stephan J. Nix
Department of Environmental Engineering Sciences
University of Florida
Gainesville, Florida 32611
Project No. R-802411
Project Officer
Richard Field
Storm and Combined Sewer Section (Edison, NJ)
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
US ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, US Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views
and policies of the US Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
11
-------
FOREWORD
The US Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution
and it involves defining the problem, measuring its impact, and searching
for solutions. The Municipal Environmental Research Laboratory develops
new and improved technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication is one of the
products of that research; a most vital communications link between the
researcher and the user community.
Combined sewer overflows and urban stormwater discharges are a significant
pollution source. This report describes simplified procedures to enable
decision makers to obtain a preliminary estimate of the magnitude of this
pollution source and the associated costs of control.
Francis T. Mayo, Director
Municipal Environmental Research Laboratory
111
-------
PREFACE
The University of Florida, in conjunction with the American Public Works
Association, developed a methodology to assess the quantity and quality
of urban stormwater runoff and to determine the cost of control for
various levels of control. This methodology provided the basis for a
nationwide assessment of this important problem. These results are des-
cribed in:
Heaney, J. P., W. C. Huber, M. A. Medina, Jr., S. J. Nix and
S. M. Hasan, Nationwide Evaluation of Combined Sewer Overflows
and Urban Stormwater Discharges: Volume II, Cost Assessment,
USEPA, 1976.
This same approach was used in a similar assessment in Ontario. The
results are presented in:
American Public Works Association and University of Florida,
"Evaluation of the Magnitude and Significance of Pollution
Loading from Urban Stormwater Runoff - Ontario," Water
Resources Branch, Ontario Ministry of Environment, Toronto,
1976.
The simplified procedures outlined in this report draw heavily on these
two assessments. The results are intended for policy makers who are
evaluating urban stormwater discharges within a relatively broad frame-
work.
IV
-------
ABSTRACT
The original USEPA Storm Water Management Model (SWMM) provides a detailed
simulation of the quantity and quality of stormwater during a specified
precipitation event lasting a few hours. This model is widely used. How-
ever, it is too detailed for many users. Indeed, there is a need for a
wide range of evaluation techniques ranging from simple to complex pro-
cedures. In particular, the 208 planning effort needs simplified procedures
to permit preliminary screening of alternatives.
In response to this need, four levels of stormwater management models have
been prepared and are being released this year. This initial volume pre-
sents a "desktop" procedure which was developed to do a nationwide assess-
ment of stormwater pollution control costs. The next three models will be
computer based and provide increasing amounts of detail.
The desktop procedure permits the user to estimate the quantity and quality
of urban runoff in the combined, storm, and unsewered portions of each urban
area in his jurisdiction. Using generalized results from the nationwide
assessment, the optimal mix of storage and treatment and its associated costs
may be estimated. Also, comparisons between tertiary treatment and storm-
water management are presented. Lastly, possible savings due to integrated
management of domestic wastewater, stormwater quality, and stormwater quantity
are evaluated.
This report is submitted as part of Grant No. R-802411 by the University of
Florida under sponsorship of the US Environmental Protection Agency. Work
was completed in May 1976.
-------
CONTENTS
Foreword
Preface iv
Abstract v
Figures lx
Tables xi
Abbreviations and Symbols xiii
Acknowledgments xvii
1. Conclusions 1
Demographic Characteristics 1
Quantity and Quality of Urban Runoff .... 2
Cost Assessment Methodology 2
2. Recommendations 4
3. Description of the Planning Area 5
4. Quantity and Quality Analysis 9
Modeling of Urban Runoff 9
Runoff Analysis 9
Stormwater Flow Prediction 9
Dry-Weather Flow Prediction 13
Quality Analysis 16
5. Overall Cost Assessment 23
Methodology 23
Principles 23
Control Technology and Associated Costs 26
Cost of Treatment and Storage 26
Relationship Between Storage/Treatment and
Percent Pollution Control 29
Use of STORM 29
STORM Input Data for Detailed Study of Five Test Cities . 31
VII
-------
CONTENTS (concluded)
Results 34
Storage/Treatment Isoquants 34
Adjustment for Treatment Efficiency 34
Adjustment for First Flush 38
Mathematical Representation of Isoquants 38
Wet-Weather Quality Control Optimization 49
Estimating Number of Overflow Events 55
Overall Cost Estimate. . 55
Overall Results 55
Tertiary Treatment versus Wet-Weather Treatment 64
Potential Savings Due to Multipurpose Planning 67
References 71
Glossary 74
Vlll
-------
FIGURES
Number
1 Hypothetical Planning Area 6
2 Imperviousness as a Function of Developed Population Density. . 11
3 Annual Correction Factor to Account for Depression Storage. . . 14
4 Normalized BOD Loadings vs Developed Population Density .... 18
5 Effect of Street Sweeping Frequency on Annual BOD Concentration
in Urban Stormwater Runoff - Des Moines, LA 19
6 Determination of Least-Cost Combination of Inputs 25
7 Storage/Treatment Configuration Used in STORM Model 27
8 Average Twenty-Five Year Rainfall Frequency for Each Study
Area 35
9 Selected One-Year Rainfall Frequency for Each Study Area. ... 36
10 Monthly Rainfall Distribution for Study Year for Each Study
Area 37
11 Definitional Sketch of Storage/Treatment Isoquants 39
12 Storage/Treatment Isoquants for Percent BOD Removal with First
Flush - Region I - San Francisco 42
13 Storage/Treatment Isoquants for Percent BOD Removal with First
Flush - Region II - Denver 43
14 Storage/Treatment Isoquants for Percent BOD Removal with First
Flush - Region III - Minneapolis 44
15 Storage/Treatment Isoquants for Percent BOD Removal with First
Flush - Region IV - Atlanta 45
16 Storage/Treatment Isoquants for Percent BOD Removal with First
Flush - Region V - Washington, DC 46
ix
-------
FIGURES (concluded)
Number
17 Mean Annual Precipitation in the United States, in Inches,
and Regional Boundaries 47
18 Annual Cost as a Function of Level of Pollutant Control .... 52
19 Effect of Minimum Interevent Time on the Annual Number of
Storm Events 56
20 Relationship Between Percent Runoff Control and Annual Number
of Overflow Events 57
21 Relationship Between Percent Pollutant Control With and Without
First Flush 58
22 Cost Allocation Factor for Five Cities 69
23 Effect of Design Storm and Number of Purposes on Cost Allo-
cation Factor for Various Levels of Control -
Midwestern US 70
-------
TABLES
Number
1 Land Use Analysis 7
2 Land Use by Type of Sewerage System 7
3 Population Served by Type of Sewerage System 8
4 Population Density by Type of Sewerage System 8
5 Effect of Urban Block Size on Curb Length Density and Imper-
viousness Due to Streets 12
6 Wet-Weather Flow 15
7 Dry-Weather Flow 15
8 Pollutant Loading Factors for Desktop Assessment 17
9 Comparison of BOD Loadings 20
10 Dry-Weather BOD Loading 22
11 Wet-Weather BOD Loading 22
12 Installed Costs for Wet-Weather Treatment Devices 28
13 Cost Functions for Wet-Weather Control Devices 30
14 STORM Input Data for Study Areas 32
15 Hydrologic Data for Study Areas 33
16 Values of Parameters for Isoquant Equations for Developed Portion
of the Test Cities 41
17 First Flush Isoquant Coefficients - Unsewered Portion of City 6
in Region III 50
18 Calculation of Optimal Solution - Unsewered Area of City 6
in Region III 51
XI
-------
Number
TABLES (concluded)
19 Approximate Annual Cost per Acre - Combined Areas 53
20 Approximate Annual Cost per Acre - Storm Areas 53
21 Approximate Annual Cost per Acre - Unsewered Areas 55
22 Optimal Total Annual Costs 65
23 Optimal Total Annual Cost per Acre 65
24 Cost Allocation Factors for Multipurpose Systems 68
xn
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
APWA — American Public Works Association
— five day biochemical oxygen demand
— dry-weather flow
— Engineering News Record
— total nitrogen
— total phosphate
— suspended solids
— Storage, Treatment, Overflow and Runoff Model
— Storm Water Management Model
— volatile solids (total)
DWF
ENR
N
P°4
SS
STORM
SWMM
VS
SYMBOLS
A,
tot
AR
Combined sewered area, acres
Storm sewered area, acres
Unsewered area, acres
Total urbanized area, acres
Annual runoff, inches/year
Coefficient
Normalized loading factor for separate sewered areas,
pounds/acre-inch; also cost allocation factor
Coefficient
XI11
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
3 Normalized loading factor for combined sewered areas,
pounds/acre-inch; also parameter in optimal cost
solution
C Annual cost of tertiary treatment, dollars/year
CA Amortized capital cost, dollars/year
CR Runoff coefficient
c Unit cost of storage, dollars/acre-inch
O
c Unit cost of treatment, dollars/inch/hour
c Unit cost of tertiary treatment, dollars/pound
D Sewage flow, mgd
DS Depression storage, inches
d Coefficient
e. Land use distribution as a fraction of the developed area
n Treatment plant efficiency
f Coefficient
f~ Factor for adjustment of pollutant loads, a function of
population density
G Curb length per area, miles/acre
LI
g Coefficient
y Street sweeping effectiveness factor
h Coefficient
I Imperviousness, percent
K Coefficient
k Coefficient
L Pipe length
1 Coefficient
xiv
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
M Pounds of pollutant j generated per acre of land use i per year
M Pollutant loading averaged over different land uses,
pounds/acre-year
Annual dry-weather BOD load, pounds/acre-year
M Pollutant loading on combined sewered areas, pounds/acre-year
M Pollutant loading on separate sewered areas, pounds/acre-year
s
MC Marginal cost, annual dollars/pound
MRS Marginal rate of substitution of storage for treatment
m Coefficient
N Street sweeping interval, days
S
OE Number of overflow events per year
OM Annual operation and maintenance costs, dollars/year
P Annual precipitation, inches/year
PD Population density; persons/acre
PD, Developed population density, persons/acre
p Coefficient
Q Coefficient
q Coefficient
R Percent pollutant control with 100 percent treatment efficiency
R Net percent pollutant removed
R Maximum percent pollutant removed
R* Optimal percent pollutant control prior to using tertiary
treatment
p Proportion of wet-weather load which is controlled
S Storage volume, inches
S* Optimal storage volume, inches
xv
-------
LIST OF ABBREVIATIONS AND SYMBOLS (continued)
s Coefficient
T Treatment rate, inches/hour
T* Optimal treatment rate, inches/hour
TI Treatment rate at which isoquant becomes asymptotic to the
ordinate, inches/hour
T- Treatment rate at which isoquant intersects the abscissa,
inches/hour
TAG Total annual cost, dollars/year
V Volume of storage required for wet-weather quantity control,
inches
v Coefficient
WP Total wet-weather pollutant load, pounds/year
w* Optimal pounds per acre of wet-weather pollutants to control
prior to using tertiary treatment
w. Pollutant removed from i type of sewered area, pounds/
year-acre
y Coefficient
Z Total annual cost, dollars/acre
Z* Optimal total annual cost, dollars/acre
Z Annual cost for primary control unit, dollars/acre
Z Annual cost for secondary control unit, dollars/acre
S
z Coefficient
£ Combined sewer deposition correction factor, pounds/acre-year
xvi
-------
ACKNOWLEDGMENTS
Much of the material for this Level I analysis was developed during the
preparation of stormwater assessments in the US and Ontario. The advice
and guidance of our advisory committees on these assessments was very
useful.
Richard Field of USEPA provided invaluable overall guidance and detailed
critical review of findings throughout the study.
Numerous persons at the University of Florida contributed to this effort.
Gordon Quesenberry developed the storm event definition. Sheikh Hasan
developed most of the cost and performance data and the solution to the
cost allocation problem. Numerous students in undergraduate and graduate
courses evaluated portions of the material. Typing of the numerous drafts
and final report was done by Ms. Mary Polinski.
xvn
-------
SECTION I
CONCLUSIONS
During the past decade, much effort has been expended in identifying
and analyzing the wet-weather pollution control problem. The initial
concern with combined sewer overflows expanded to consideration of
stormwater runoff in general. This study assesses the costs of control-
ling wet-weather pollution to varying degrees. The key question to be
addressed is x^hat is the relative importance of various sources of wet-
weather pollution and how does wet-weather pollution control compare to
dry-weather pollution control? Its impact on receiving water is not
evaluated.
Control of wet-weather pollution is distinctly different than the
traditional dry-weather problem. In wet-weather pollution control, one
would normally use a mix of storage and treatment, not treatment alone.
Thus, new techniques are needed to determine optimal mixes of storage
and treatment. Numerous effectiveness criteria for wet-weather control
have been used, e.g., number of overflows, percent runoff control, per-
cent BOD control. For wet-weather control, the most critical impact on
receiving water does not necessarily occur under low flow conditions.
How should the critical conditions be defined? Basic questions of this
nature arose throughout the study because it is such a relatively nex^
area of concern. Thus, the final estimate could vary widely if some of
these assumptions are changed. Hoxvever, the approach is a fairly general
one and assumptions are stated explicitly. Thus, the interested reader
can refine the estimates as better information becomes available. The
remainder of this section presents conclusions.
DEMOGRAPHIC CHARACTERISTICS
For this level of analysis, each urban area is partitioned into five
categories by type of use (residential, commercial, industrial, other,
and undeveloped) and by type of sewerage system (combined, storm, and
unsewered). The population served by type of sewerage system is also
included.
-------
QUANTITY AND QUALITY OF URBAN RUNOFF
An examination of precipitation patterns led to the division of the country
into five zones for purposes of analysis with the Corps of Engineers' STORM
model: Pacific Coast, Rocky Mountain, Midwest and Texas, South and Southeast,
and Northeast. STORM was run on a representative city for each of these
regions: San Francisco, Denver, Minneapolis, Atlanta, and Washington, DC.
Results from these runs were used in developing the nationwide assessment
methodology and also used to calibrate the elementary technique used for run-
off prediction for the 248 urbanized areas.
Annual wet-weather runoff was generated using a runoff coefficient and
depression storage expressed as a function of imperviousness which in turn
is a function of population density. Dry-weather flow is a function of
population density on the basis of 100 gallons per person-day (379 liters per
person-day).
Analysis of available urban runoff quality data indicates a great number
of disaggregated urban runoff studies from which it is highly difficult
to draw meaningful conclusions as to pollutant loading rates. For
instance, there are no known studies in which both surface and effluent
data have been gathered simultaneously. In addition, there is a wide
variation in the manner in which data are reported (e.g., "average" concen-
trations) and in the amount of related information provided about the
catchment areas (e.g., population density).
On the basis of the available data, pollutant loading estimates were
developed for wet weather for BOD , suspended solids, volatile solids,
total phosphate (PO.) and total nitrogen (N) , and derived as functions
of precipitation, land use and population density; the latter only for
residential land use. Other land uses are commercial, industrial and
open. These estimates indicate that, for the same population density,
loads from combined sewered areas are approximately four times higher
than those from separate sewered areas. Furthermore, higher population
densities in combined sewered areas will increase the ratio even more
because loadings are assumed to be an increasing function of population
density. Annual pollutant loads were calculated for both wet-and dry-
weather conditions, the latter under the assumption of 0.17 pounds per
person-day (0.08 kg per person-day).
COST ASSESSMENT METHODOLOGY
A generalized method for evaluating the optimal mix of storage and treatment
for any desired level of pollutant control was presented. This method can
be used for any city in the United States to obtain a first approximation
of control costs. Five cities (Atlanta, Denver, Minneapolis, San Francisco,
and Washington, DC) were used in the more detailed analysis. The effects of
treatment plant efficiency, first flush, and/or street sweeping are included.
-------
An evaluation was made of the relative desirability of using a mix of
storage with either primary treatment or secondary treatment. The
basic tradeoff to be evaluated is whether primary treatment is sufficiently
less expensive than secondary treatment to offset its lower removal
efficiency which necessitates treating a much larger amount of flow to
effect an equivalent BOD removal.
The annual average percent runoff control and the annual number of over-
flow events were correlated to permit using either criterion as an
effectiveness metric. A simplified procedure for defining a precipitation
event is included in this analysis.
Assessment results (annual costs per acre)are presented by type of
sewerage system. In order to obtain an overall wet-weather pollutant
control of, say, 50 percent in a given urbanized area, the optimal
strategy is to use a blend of control in the combined, storm, and
unsewered portions of the urbanized areas such that the marginal cost of
control in each of these three areas is equal. The results are shown by
type of sewerage system. Knowing this result and the control cost equations
for each type of sewerage system in each urbanized area, the optimal cost
per acre is determined. Lastly, the costs per acre are multiplied by the
acreage in the combined, storm, and unsewered categories to obtain the
final assessment results. The incremental costs for wet-weather control
increase significantly. This is due to the disproportionately larger
control units needed to capture the less frequent, larger storms.
An analysis was made of the possibility of cost allocation among wet-
weather quality control, dry-weather quality control (with excess
capacity and flow equalization), and wet-weather quantity control (with
storage required to reduce runoff rates and volumes). The results sug-
gest that significant savings might be realized, i.e., 70 percent at
low control levels to 30 percent at high control levels.
Lastly, the relationship between tertiary treatment and wet weather
control was examined by finding the percent wet-weather control to
initiate prior to using tertiary treatment. Results indicate that a
portion of the wet weather flow problem should be controlled before
initiating tertiary treatment. BOD removal was used as the effectiveness
metric. Different results would be obtained using nutrient control as the
criterion.
-------
SECTION II
RECOMMENDATIONS
The Storm Water Management Model (SWMM) Level I, desktop analysis is designed
to provide a highly simplified first estimate of urban stormwater pollution
quantities, control alternatives, and associated costs. Accordingly, users
should use these estimates judiciously. A wide margin of error in the
results is to be expected if the default values shown in this document are
used. Local estimates are, of course, the preferable ones to use and can
be obtained at additional costs.
After completing the Level I analysis, one can proceed to the more
sophisticated procedure outlined in Levels II to IV if further refinement
is desired.
-------
SECTION III
DESCRIPTION OF THE PLANNING AREA
For this level of analysis, each urban area is partitioned into five
categories by type of use (residential, commercial, industrial, other
developed, e.g., schools, cemeteries, parks; and undeveloped) and by type
of sewerage system (combined, storm, and unsewered). The population served
by type of sewerage system is also included. Land use information can be
obtained from the local planning agency. Similarly, information regarding
area served by type of sewer system is usually available from the local
department of public works. It is important to carefully delineate the
boundaries of the urban area to be considered. US Census definitions of
"urbanized areas" and Standard Metropolitan Statistical Areas (SMSA's)
include significant portions of land which would not be developed during
the planning horizon. Such lands should not be included in the analysis.
The layout of a hypothetical Section 208, PL 92-500, planning area ±s
presented in Figure 1. The planning area is assumed to be located in the
midwestern US. Information on existing land use by type of use is shown
in Table 1 while land use by type of sewerage system is shown in Table 2.
Population and population density data by type of sewerage system are
shown in Tables 3 and 4, respectively. Mean annual precipitation in the
area is approximately 31 inches (78.7 cm). Each area is served by a
secondary treatment plant. Consideration is being given to installing
tertiary plants in some or all of the cities. The average dry-weather
wastewater flow is 100 gallons per capita per day.
-------
208 PLANNING AREA BOUNDARY
POLITICAL JURISDICTION 3
POLITICAL
JURISDICTION 2
POLITICAL
JURISDICTION I
— CONNECTING SEWER
3mi
— PIPE LINES
(DISTANCE IN MILES)
Figure 1. Hypothetical Planning Area
-------
Table 1. LAND USE ANALYSIS
Hypo- Land Use (acres)
thetical
City Res Comm Indl Oth Undv
1 200 50 50 200 500
2 100 20 10 100 430
3 1,100 200 100 400 2,200
4 400 100 100 400 1,000
5 50 10 5 50 285
6 6,000 1,000 1,000 2,000 5,000
7 400 100 100 400 1,000
Total 8,250 1,480 1,365 3,550 10,415
Total
1,000
660
4,000
2,000
400
15,000
2,000
25,060
Table 2. LAND USE BY TYPE OF SEWERAGE SYSTEM
Area Served by Type of System
Hypo- (acres)
Lhelical
City Undv Comb Storm Unsew
1 500 0 100 400
2 430 0 100 130
3 2,200 0 1,000 800
4 1,000 0 500 500
5 285 0 50 65
6 5,000 6,000 2,000 2,000
7 1,000 0 500 500
Total
1,000
660
4,000
2,000
400
15,000
2,000
Total
10,415
6,000
4,250
4,395
25,060
-------
Table 3. POPULATION SERVED BY TYPE OF SEWERAGE SYSTEM
Hypo-
thetical
City
1
2
3
4
5
6 75
7
Total 75
Population Served by Type
of System
Comb
0
0
0
0
0
,000
0
,000
Storm
1,000
600
10,000
3,000
300
15,000
800
30,700
Unsew
1,000
400
5,000
2,000
200
10,000
200
18,800
Table 4. POPULATION DENSITY BY TYPE OF SEWERAGE
Hypo-
thetical
City
1
2
3
4
5
6
7
Comb
0.0
0.0
0.0
0.0
0.0
12.50
0.0
Population
of System
Storm
10.00
6.00
10.00
6.00
6.00
7.50
1.60
Total
2,000
1,000
15,000
5,000
500
100,000
1,000
124,500
SYSTEM
Density by Type
(persons/acre)
Unsew
2.50
3.08
6.25
4.00
3.08
5.00
0.40
Avg
4.00
4.35
8.33
5.00
4.35
10.00
1.00
Total
12.50
7.22
4.28
8.50
-------
SECTION IV
QUANTITY AND QUALITY ANALYSIS
The purpose of this section is to estimate the quantity and quality of
urban runoff. Precipitation patterns are analyzed to form a basis for
predicting the quantity of urban runoff. A pollutant load predictive
equation is developed which provides the basis for assessing pollutant
loads.
MODELING OF URBAN RUNOFF
The overall goal of urban runoff modeling is to aid in decision making
for the abatement of water quantity and quality problems. Thus, computer
models do not provide "solutions" to problems, in and of themselves.
Rather, they serve as useful tools to those charged with devising such
solutions. Within this context, sub-objectives of the modeling process
may be identified: planning, design, and operation. Models for the
latter category are generally site-specific1'2 and were not considered
during this research study. However, numerous models are available for
planning and design purposes3'1* and two — the Corps of Engineers' STORM5./
and the USEPA Storm Water Management Model (SWMM)7 respectively — are
particularly useful for such tasks. However, they are not unique; several
other urban runoff models are capable of similar tasks.8
RUNOFF ANALYSIS
Stormwater Flow Prediction
Techniques for prediction of runoff quantities vary from very simple methods
of the Rational Method type to sophisticated models of the nature of SWMM.
The Storage, Treatment, Overflow and Runoff Model (STORM) was developed by
Water Resources Engineers, Inc. (WRE), for the Hydrologic Engineering Center
(HEC) of the Corps of Engineers.5'6 The model was designed for planning
purposes, i.e., for long-term simulation of many storm events using an
hourly time step. Techniques used in STORM are relatively simple, relying
on weighted average runoff coefficients and a simple loss function to
predict hourly runoff volumes. Nonetheless, because of the nature of the
continuous simulation involved, it is at a considerably higher level, and
therefore more complex, than earlier, desktop techniques.
-------
Due to the complexities and data requirements of STORM, it is unnecessary
to run the model for all applications. Rather, the methodology is based
upon runs in only five test cities: San Francisco, Denver, Minneapolis,
Atlanta and Washington, DC, as described in Section V. However, these
applications produced useful information regarding the formulation of a
simple runoff prediction method for application to other cities.
STORM computes a runoff coefficient, CR, weighted between pervious and
impervious areas by
CR = 0.15 (1 - 1/100) + 0.90 I/100 (1)
= 0.15 + 0.75 1/100
where I is percent imperviousness and the coefficients 0.15 and 0.90 are the
default values used in STORM for runoff coefficients from pervious and imper-
vious areas, respectively. Note that in equation 1, the effect of demographic
factors (e.g., land use, population density) is incorporated into the
imperviousness, I.
Graham et al. (Washington, DC), the American Public Works Association and
Stankowski (New Jersey) have developed equations to predict imperviousness
as a function of population density.9'10'11 The imperviousness is to be
estimated for the developed portion of the urbanized area only. Also the
weighted average imperviousness and population density were calculated for
nine Ontario cities.12 These results are plotted on Figure 2 along with the
three estimating curves. Also, a tabulation was made of the imperviousness
due to streets alone for various block sizes as shown in Table 5. These
results are also plotted on Figure 2. A comparison of these various plots
and the actual data indicates that the New Jersey11 equation provides a
suitable predictive equation with the population density defined as developed
population density. Thus, the equation used to estimate imperviousness is
(0.573-0.0391 log10PDd)
I = 9.6 PDd u (2)
where I = imperviousness, percent, and
PD = population density in developed portion of the
urbanized area, persons/acre.
The simplified equation for estimating annual runoff (AR) is
AR = (0.15 + 0.75 I/100)P (3)
where AR = annual runoff, inches/year,
I = imperviousness, percent, from equation 2, and
P = annual precipitation, inches/year.
10
-------
persons/hectare
100
50
so
70
80
GRAHAM ET AL.,
WASHINGTON. D.C
NEW JERSEY,
567 MUNICIPALITIES
A WASHINGTON, D.C
D ONTARIO
IMPERVIOUSNESS DUE TO STREETS ONLY
05 10 15 20 25 30 35
DEVELOPED POPULATION DENSITY, PDd , persons/acre
Figure 2. Imperviousness as a Function of Developed Population
Density
11
-------
Table 5. EFFECT OF URBAN BLOCK SIZE ON CURB LENGTH DENSITY AND IMPERVIOUSNESS DUE TO STREETS
Block Size Curb Length Density
ftxft (mxm) Area, ac (ha) ft/ac (m/ha)
660 x 330 5 ( 2.02) 392.0 (298.0)
(201 x 101)
1320 x 660 20 ( 8.09) 198.0 (148.0)
(402 x 201)
2640 x 1320 80 ( 32.40) 99.0 ( 74.6)
(807 x 402)
5280 x 2640 100 (130.00) 49.5 ( 37.3)
(1609 x 807)
Fraction Impervious-
ness Due to Streets
0.150
0.077
0.039
0.019
Assumes 34 ft (10.4 m) wide street.
-------
Equation 3 can be refined by accounting for depression storage as is done
in STORM. For this simplified assessment methodology, the depression
storage is assumed to be as follows:
Land Use Depression Storage, in. (cm)
Impervious 0.0625 (0.159)
Pervious 0.25 (0.635)
For a given land use, the area weighted depression storage, DS, in inches,
is
DS = u.25 - 0.1675 (1/100) 0 £ I <_ 100 (4)
To approximate the effect of depression storage on estimated annual runoff,
one year of Minneapolis, MI-j data were simulated for varying levels of
depression storage. The results, shown in Figure 3, indicate that a factor
for depression storage can be subtracted from the original runoff equation
to yield the final equation for estimating annual runoff, i.e.,
AR = (0.15 + 0.75 I/100)P - 5.234(DS)°'5957 (5)
where AR = annual runoff, in./yr,
I = imperviousness, percent,
P = annual precipitation, in./yr, and
DS = depression storage, in. (0.005 <_ DS <_ 0.30)
The results for the seven hypothetical cities of Section III are shown
in Table 6.
Dry-Weather Flow Prediction
Dry-weather flow is predicted based on an average flow of 100 gallons per
person-day (179 liter per person-day). Upon multiplication by population
density and conversion to appropriate units,
DWF =1.34 PD, (6)
d
where DWF = annual dry-weather flow, in./yr, and
PD, = developed population density, persons/acre.
d
Results of these calculations are shown in Table 7.
13
-------
3.0
o
c
-C
o
c
.2.0-
cc
o
h-
o
o
h-
o
LU
ct
or
o
o
cc
o
h-
co
2:
o
co
CO
UJ
cc
0.
UJ
Q
0.0
0
Figure 3.
O.I
0.2
cm.
0.3 0.4
05
0.6
0.7
i
DEPRESSION STORAGE 0 5957
CORRECTION FACTOR 5.234(DS)'
.005irKDS<:.30iiv
SIMULATED USING ONE YEA'3 OF
HOURLY DATA-MINNEAPOLIS, MINN.
.10 .20
DS, DEPRESSION STORAGE , inches
-7.0
-6.0
•5.0
-4.0 \
o
c
-3.0
E
o
-2.0
-1.0
0.0
.30
Annual Correction Factor to Account for Depression
Storage - Minneapolis, MN
14
-------
Table 6. WET-WEATHER FLOW
Hypo- Annl-
thetical Precp
City in./yr
1
2
3
4
5
6
7
31.0
31.0
31.0
31.0
31.0
31.0
31.0
Wet Weather Flow
(inches/year)
Comb
0.0
0.0
0.0
0.0
0.0
11.4
0.0
Storm
10.3
8.4
10.3
8.4
8.4
9.2
5.2
Unsew
6.1
6.5
8.5
7.2
6.5
7.8
3.7
Wgtd
Avg
6.9
7.3
9.5
7.8
7.3
10.2
4.4
Average
31.0
11.4
8.9
7.2
9.4
Table 7- DRY-WEATHER FLOW
Dry-Weather Flow
Hypo-
Lhetical
City Comb
1
2
3
4
5
6
7
0.0
0.0
0.0
0.0
0.0
16.8
0.0
(inches/year)
Storm
13.4
8.1
13.4
8.1
8.1
10.1
2.2
Unsew
3.4
4.1
8.4
5.4
4.1
6.7
0.5
Wgtd
Avg
5.4
5.8
11.2
6.7
5.8
13.4
1.3
Total
9.7
5.7
11.4
15
-------
QUALITY ANALYSIS
Quality analyses may be performed at several levels of detail, ranging from
an explicit formulation of runoff quality for small subcatchments within a
city to a broad representation of pollutant loads for an entire urbanized
area. It may be necessary to consider the entire spectrum during the course
of a study.
It is unfortunate that perhaps the only consistent remark about urban runoff
quality analysis in general is that data and results of previous studies are
so remarkably inconsistent. Few studies have been made of characteristics
of street litter, and they offer a wide range of values of concentrations
and loads. Effluent data show a similar scatter. However, it is necessary
that a decision be made regarding actual values for use in the analysis.
Table 8 presents a predictive equation developed after a review of available
stormwater pollutant loading and effluent concentration data. ^ The equation
permits one to estimate BOD_, SS, VS , PO, and N loads as a function of land
use, type of sewer system, precipitation, population density, and street
sweeping frequency. Loadings in combined sewer areas are assumed to be 4.12
times as large as loadings in separate areas. They are assumed to vary as a
function of developed population density as shown in Figure 4. The intercept
(0.142) was determined based on data for open space. The exponent (0.54) is
based on the exponent of the imperviousness equation at a population density
of 8 persons per acre (20 persons per ha) such that pollutant concentration
increases as a function of population density. Lastly, the coefficient
(0.218) is based on an average of data points with a PD. ranging from 5 to
15 persons per acre (12 to 37 persons per ha) to yield a value of f_(PD,)
of 0.895 at 10 persons per acre (25 persons per ha). The street sweeping
relationship was derived by making numerous runs of STORM with varying street
sweeping frequencies. The results are shown in Figure 5.
The BOD loadings are compared to dry-weather flow loadings in Table 9 for
residential land use. Storm and combined runoff can be seen to be comparable
to secondary treatment plant effluent, although on a city-wide basis they
would be greater because of higher loadings for commercial and industrial
areas. BOD loads in storm runoff from storm sewered and unsewered areas are
in addition to the dry-weather flow loads. However, BOD loads in combined
sewers contain the portion of the dry-weather load which settles during dry-
weather periods. A detailed study of deposition in combined sewers in
Boston indicated that 10.3 percent of the total daily load is deposited.1"4
Examination of Table 9 indicates that the calculated difference in BOD
loadings between the combined and separate areas is 67 pounds per acre-year,
10.8 percent of the dry-weather load. Thus, the equation appears to provide
a reasonable approximation of combined sewer loads.
The equations indicated in Table 8 may easily be used to calculate loadings
of any of the desired parameters, given the precipitation and population den-
sity of the area of interest. The land use distribution can be used to weight
the pollutant loading factors to give an average over all land uses as follows:
a± ' f (PD ) • Y. (7)
i
16
-------
Table 8. POLLUTANT LOADING FACTORS FOR DESKTOP ASSESSMENT
The following equations may be used to predict annual average
loading rates as a function of land use, precipitation and population
density.
Separate Areas: M - a(i,j) • P • f,(PD.)
s z a
Combined Areas: M • 6(i,j)
P • f2(PDd)
where M » pounds of pollutant j generated per acre of
land use i per year,
P - annual precipitation, Inches per year,
PD, • developed population density, persons per acre,
a,B • factors given in table below,
y - street sweeping effectiveness factor, and
f2(PDd> » population density function.
Land Uses: i « 1 Residential
i = 2 Commercial
1=3 Industrial
i « 4 Other Developed, e.g., parks, cemeteries, schools
(assume PD.
0)
Pollutants: j = 1 BOD , Total
j = 2 Suspended Solids (SS)
j = 3 Volatile Solids, Total (VS)
4 Total PO, (as PO )
5 Total N
Population Function: i = 1 f,(PD
i = 2,3 f,(PDp - 1.0
1=4 f,(PD°) = 0.142
2. a
0.142 + 0.218
PD
0.54
Factors a and g for Equations: Separate factors, u, and combined factors,
g, have units Ib/acre-in. To convert to kg/ha-cm, multiply
by 0.442.
Pollutant, j
Land Use, i 1. BOD5 2. SS 3. VS 4. P04 5. N
1. Residential 0.799
Separate 2. Commercial 3.20
Areas, a 3. Industrial 1.21
4. Other 0.113
1. Residential 3.29
Combined 2. Commercial 13.2
Areas, 6 3. Industrial 5.00
4. Other 0.467
16.3 9.45 0.0336 0.131
22.2 14.0 0.0757 0.296
29.1 14.3 0.0705 0.277
2.70 2.6 0.00994 0.0605
67.2 38.9 0.139 0.540
91.8 57.9 0.312 1.22
120.0 59.2 0.291 1.14
11.1 10.8 0.0411 0.250
Street Sweeping: Factor y is a function of street sweeping interval,
N , (days):
S
(Ng/20 if 0 <_ Ng <
1.0 if N > 20 d
20 days
days
17
-------
persons/hectare
20
2.0-
Q
o
1.0-
40
L_
60
I
e
Of
LAND USE = OPEN
0
10
\
20
80
. l
100
120
140
SEPARATE COMBINED
WASHINGTON, D.C A
DES MOINES O
MILWAUKEE
TULSA V
ROANOKE *
CINCINNATI O
SEATTLE a
WINDSOR <>
CALIBRATION POINT. ®
a
A
O
O
f2(PD)- 0.142 + 0.218
'
30
40
r
50
DEVELOPED POPULATION DENSITY, PD.,persons/acre
60
Figure 4. Normalized BOD Loadings vs Developed Population Density.
^Average loadings for separate and combined areas
on the basis of data from the indicated cities are,
respectively13: 0.069 Ib-BOD/ac-day (0.078 kg-BOD/ha-day)
and 0.27 Ib-BOD/ac-day (0.30 kg-BOD/ha-day).
18
-------
70-|
60-
50-
40-
Q
O
CO
30-
20-
10-
BOD
-©-
.©-
—T
28
—1 ' T~
32 36
,-600
_500
-400
- e
-------
Table 9. COMPARISON OF BOD LOADINGS
Assume residential land use; PD, = 10 persons/acre
(24.7 persons/ha), P = 30 in./yr (76 cm/yr), and
Y = 1.
Separate Areas
Combined Areas
Dry Weather
DWF at 85% Treatment13
Ib/ac-yr
21
88
621
93
kg/ha-yr
24
99
697
105
No correction for deposition in sewers included
in this estimate.
Assuming 0.17 Ib-BOD/person-day (0.08 kg-BOD/person-
day).
where e. = land use distribution as a fraction of the
developed area.
Equation 7 may be applied to separate and unsewered areas and multiplied by
4.12 for use in combined areas. Dry-weather flow BOD loadings are based on
a generation rate of 0.17 pound-BOD per person-day (0.08 kg-BOD per person-
day) for storm and combined systems. To recognize the deposition in
combined sewers as discussed earlier, the difference in pollutant load
between combined and storm sewered drainage is subtracted from the dry-
weather load. When multiplied by population density and converted to an
annual basis, the result is
M =62.1 PD, - e (8)
Dw d
where M^ = dry-weather BOD loading, Ib/acre-yr,
D w
PD, = developed population density, persons/acre,
£ = adjustment factor for combined sewers , Ib/acre-yr,
0 4> storm sewers and unsewered
[3(i,J)-a(i,j)]P-f0(PD,).Y => combined sewers
2 a
20
-------
a(i,j) = combined sewer loading factor (see Table 8),
g(i,j) = separate sewer loading factor (see Table 8),
Y = street sweeping factor (see Table 8),
j = 1 for BOD, and
i is for appropriate land use.
Since equation 8 is used for BOD loadings, the subscript } for parameters a
and 3 will equal 1 (see Table 8). Land use i will refer to the case at hand,
or the equation can be averaged over different land uses.
Results for both dry- and wet-weather conditions for the seven hypothetical
cities are shown in Tables 10 and 11.
21
-------
Table 10. DRY-WEATHER BOD LOADING
Hypo-
thetical
City
1
2
3
4
5
6
7
Average
Table
Dry- Weather BOD
(Ibs/acre-yr)
Comb
0
0
0
0
0
685
0
685
Storm
621
373
621
373
373
466
99
449
11 . WET -WEATHER
Unsew Wgtd
Avg
155 248
191 270
388 517
248 310
191 270
310 566
25 62
266 491
BOD LOADING
Wet-Weather BOD
Hypo-
thetical
City
1
2
3
4
5
6
7
(Ibs/acre-yr)
Comb
0.0
0.0
0.0
0.0
0.0
117.6
0.0
Storm
22.8
18.2
26.8
21.0
18.2
25.5
18.1
Unsew
18.8
16.3
24.2
19.3
16.3
23.6
16.6
Wgtd
Avg
19.6
17.1
25.7
20.4
17.1
80.4
17.3
Average
117.6
24.1
21.7
61.7
22
-------
SECTION V
OVERALL COST ASSESSMENT
This section develops and applies a methodology to estimate the cost of
controlling pollution from urban storm related discharges. Costs of con-
trolling combined sewer overflows, stormwater runoff, and/or providing
tertiary treatment are compared. A general methodology for determining
wet-weather pollution control costs is presented. Then, a procedure is
described for determining the relationship between storage, treatment, and
pollutant control for control of wet-weather flows. Generalized predictive
equations are developed based on relatively intensive studies of five cities:
Atlanta, Denver, Minneapolis, San Francisco, and Washington. •*• ^ Knowing
this "production function" one can determine the optimal combination of
storage and treatment by combining this information with data on the cost
and performance of the available control options. This information is
combined to produce the assessment of costs.
METHODOLOGY
Principles
There are several economic theories which, when applied to environmental
resources management, assist in the decision-making process. One such
theory is production theory, which provides techniques that aid in
evaluating items such as the optimal size of a reservoir for water supply
and flood control, or a wastewater treatment plant for pollution control.15
When the cost of inputs such as the reservoir or treatment plant is known
then the cost of achieving a desired level of output (e.g., water supply
or pollution control) may be determined.
In stormwater management, the inputs may be in the form of a storage
capacity and a treatment rate. Storage is expressed in terms of million
gallons or inches over a certain area, typically the watershed being
analyzed. The unit for treatment is either million gallons per day or
inches per hour, using the same area as storage.
When the degree of wet-weather control is considered as a single output,
it can be expressed either in terms of the percent of the runoff treated
or the number of overflows per year. This is with respect to quantity only
and is therefore dependent upon the input storage capacity and treatment
rate.
23
-------
When dealing with only two inputs it is feasible to use a graphical method
to find the optimal combinations. Isoquants can be constructed which
represent equal levels of output for different combinations of input (see
Figure 6). For example, each isoquant could represent a specific percent
of the runoff treated for different combinations of storage and treatment.
Isoquants have the following properties:15
1. Two isoquants cannot intersect. Intersecting
isoquants would imply two different levels of
output from the same input.
2. Isoquants slope downward and to the right
because as one input increases it takes
less of the other input to achieve the same
level of output.
3. Isoquants are convex to the origin because of
the decreasing ability of one input to be sub-
stituted for another to obtain a given level of
output. This is known as the principle of
diminishing marginal rate of substitution.
Also on Figure 6, a series of parallel lines has been constructed. These
lines represent combinations of input 1 and input 2 which may be achieved
at the same total cost. The lines are known as isocost lines. The slope
of the isocost lines is the relative unit cost between input 1 and input 2.
The most economical combination of input 1 and input 2 to produce a desired
level of output is the point where the isocost lines become tangent to the
isoquant representing the desired level of output.
The line which joins the points of tangency among several isoquants and
the isocost lines is called the expansion path. After the expansion path
has been determined, the optimal combination of inputs can be determined
for any level of output by finding the intersection of the isoquant
representing the desired level of output and the expansion path.
The maximum output for a given cost may be found by constructing the
isocost line for the given total expenditure. The slope of the isocost
line is the relative unit cost of the two inputs. The intercept of the
axis depicting input 1 would be the allowed total cost divided by the
unit cost of input 1. From this information, the isocost line may be
drawn. The point where the isocost line intersects the expansion path
gives the combination of inputs which produces the maximum output at the
given cost.
The next three sub-sections describe
the available storage/treatment options - their
costs and effectiveness;
the production functions for evaluating tradeoffs
between storage and treatment; and
24
-------
ID
CL
C. = UNIT COST OF INPUT I
C = UNIT COST OF INPUT 2
2
EXPANSION PATH
ISOCOST LINE
INPUT 2
Figure 6. Determination of Least-Cost Combination of Inputs
25
-------
0 the solution to the optimization problem yielding
the optimal expansion path for any city.
Given this information, the final assessment methodology is presented.
Control Technology and Associated Costs
A wide variety of control alternatives are available for improving the
quality of wet-weather flows.16'17'18 Rooftops and parking lot storage,
surface and underground tanks, in-line storage, and storage in treatment
units are used to control the flow. Wet-weather quality control alter-
natives can be subdivided into two categories: primary devices and
secondary devices. Primary devices take advantage of physical processes
such as screening, settling and flotation. Secondary devices take advan-
tage of biological processes and physical-chemical processes. These control
devices are suitable for treating stormwater runoff as well as combined
sewer overflows. At the present time, there are several installations
throughout the US designed to evaluate the effectiveness of various primary
and secondary devices. 3 Based on these data, the representative performance
of primary devices is assumed to be 40 percent BOD- removal efficiency and
that of secondary devices to be 85 percent BODj- removal efficiency.
"Storage" devices will typically be used in conjunction with the above
"treatment" devices. The two purposes are interrelated. Wastewater
detained a sufficient time in a storage unit will undergo treatment. On
the other hand, treatment units also function as storage units in that they
equalize fluctuations in influent flow and concentration. DiToro presents
approaches for evaluating the equalization and treatment which occur in both of
these units.1 The STORM model, which was used in this assessment, assumes
the configuration for storage and treatment shown in Figure 7 - No treatment
is assumed to occur in storage and "treatment" is assumed to be complete
removal of all pollutants routed through treatment. Thus, for the purposes
of this assessment, no treatment is assumed to occur in storage and control
costs are assigned accordingly. This assumption tends to underestimate the
costs of storage since all provisions for solids handling are included in
treatment.
COST OF TREATMENT AND STORAGE
Cost data for installed wet-weather treatment devices are listed in Table 12.
Since wet-weather control facilities operate intermittently, annual operation
and maintenance costs are greatly affected by the number of hours the
facility is utilized. As a general rule, a facility will operate a greater
amount of the time if it incorporates storage. An examination of Table 12
reveals that annual operation and maintenance costs are 16.7 percent of the
total annual costs for the contact stabilization unit. In the case of the
swirl concentrator, the percentage is 27.3. Annual operation and maintenance
costs for other units fall in between these two values. Based on this
analysis, it was decided to assume annual operation and maintenance costs as
20 percent of the total annual costs for all treatment devices. Cost
26
-------
RUNOFF
TREATMENT
STORAGE
OVERFLOW
RECEIVING WATER
Figure 7. Storage/Treatment Configuration Used in STORM Model
-------
Table 12. INSTALLED COSTS FOR WET-WEATHER TREATMENT DEVICES
CO
Annual Cost: $/yr
Control Device
Swirl Concentrator
d
Microstrainer
£
Dissolved Air Flotation
Contact Stabilization6
Capacity
mgd (m3/day)
8.9 (34,500)
7.4 (28,700)
25.0 (96,900)
20.0 (77,500)
a b
Amortized Capital '
5,600
14,230
71,706
120,000
Operation and Maintenance
2,100
3,895
16,700f
24,000
Total
7,700
18,125
88,406
144,000
Eased on 8 percent interest for 20 years.
Construction cost. Does not include sludge handling costs.
CField et al^, 1976.20
'Wner, 1974.21
eLager and Smith, 197417 for Racine, Wisconsin adjusted to ENR = 2200.
Operation and maintenance costs based on 480 hours of operation
-------
functions developed for various wet-weather quality control devices are
presented in Table 13. These costs include provisions for sludge handling,
engineering, contingencies and land costs.
All treatment units exhibit economies of scale, i.e., unit costs decrease
as plant size increases. Thus, there is an incentive to build larger units.
The optimal size treatment unit can be found by comparing the savings in
treatment cost of going to a larger unit with the increased piping costs.
For example, if one is comparing building two 10 mgd (37,850 m-Vday) plants
with building one 20 mgd (75,700 nP/day) plant and a pipeline, the breakeven
pipe length, L, is found using
Two plants One plant + pipeline
s(10)Z + s(10)Z = s(20)Z + Q(10)y(L) (9)
where s,z,Q and y = coefficients.
For this level I analysis the number and flow rate of stormwater discharges
in urban areas is assumed to be unknown. Thus, it is not possible to
determine the optimal mix of treatment plants and pipelines. Therefore,
representative treatment costs were used as shown in Table 13.
Review of data on the cost of storage indicated wide variation in the costs
of storage. Thus, the relatively simple relationship shown in Table 13 was
used. Annual storage costs are estimated as a function of gross population
density. The curve was derived using an unamortized capital cost of $0.10
per gallon ($0.02 per liter) for PD = 5 persons per acre (12.4 persons per
ha) and $0.50 per gallon ($0.132 per liter) for PD = 15 persons per acre
(37.1 persons per ha).
RELATIONSHIP BETWEEN STORAGE/TREATMENT AND PERCENT POLLUTION CONTROL
Use of STORM
STORM5'6 was used to evaluate various storage/treatment options for con-
trolling stormwater runoff pollution. This model assumes that the study
area can be characterized as a single catchment from which hourly runoff
is directed to storage and treatment.
STORM uses a simplified rainfall/runoff relationship, neglects the transport
of water through the city and assumes a very simple relationship between
storage and treatment. However, these simplifications are essential if one
hopes to do a continuous simulation. The continuous simulation approach
was used because no general concurrence exists regarding an appropriate
single event that one should analyze. The degree of control can be
expressed in terms of the percent of the runoff treated, the annual number
of overflows, or the amount of pollutants discharged to the receiving waters.
As described in the User's Manual, STORM computes the runoff based on the
composite runoff coefficient and the effective precipitation.5 The
29
-------
Table 13. COST FUNCTIONS FOR WET-WEATHER CONTROL DEVICESa'b'
OJ
o
Device
Primary
Amortized
CA =
or
Control Alternative 1
Swirl Concentrator0'*1'6 1,971.0
Capital
lTm
ism
m
0.70
Annual Cost: $/yr
Operation and Maintenance
OM = pTq
p q
584.0 0.70
Total
TC - sTZ
or sS
s z
2,555.0 0.
70
Secondary
Storage
Microstrainer '
Dissolved Air Flotation
Sedimentation
7,343.8 0.76
8,161.4 0.84
32,634.7 0.70
1,836.0 0.76
2,036.7 0.84
8,157.8 0.70
9,179.8 0.76
10,198.1 0.84
40,792.5 0.70
Representative Primary Device Total Annual Cost = $4,000 per mgd ($1.05/m /day)
Contact Stabilization8
Physical-Chemical
19,585.7 0.85
32,634.7 0.85
4,894.7 0.85
8,157.8 0.85
24,480.4 0.85
40,792.5 0.85
Representative Secondary Device Total Annual Cost = $15,000 per mgd ($3.93/m /day)
High Density (15 per/ac)
Low Density,(5 per/ac)
Parking Lot
Rooftop
51,000.0 1.00
10,200.0 1.00
10,200.0 1.00
5,100.0 1.00
Representative Annual Storage Cost ($ per ac-in) = $122 e
0.16(PD)
k i
T = Wet-Weather Treatment Rate in mgd; S = Storage Volume in mg
ENR = 2200. Includes land costs, chlorination, sludge handling, engineering and contingencies.
Sludge handling costs based on data from Battelle Northwest, 1974.2^
CField et al., 1976.20
Benjes et al. , 1975.2k
®Lager and Smith, 1974.17
Maher, 1974.21
^Agnew et al., 1975.22
VWiswall and Robbins, 1975.25 3
.For T <_ 100 mgd. No economies of scale beyond 100 mgd (378,500 m /day).
:rPD = gross population density, persons/acre.
^One mgd = 3,785 m3/day.
One mg = 3,785 m3.
-------
depression storage must be satisfied before the runoff coefficient is
applied to the precipitation. The amount of depression storage available
in ditches, depressions, and other surfaces is a function of the past
precipitation and the evaporation rates. Each hour that runoff occurs,
the model compares it to the treatment rate. As long as the runoff rate
is less than or equal to the treatment rate all the runoff passes directly
through the treatment plant and storage is not utilized. When the runoff
rate exceeds the treatment rate, the excess runoff is sent to storage. If
excess runoff occurs frequently enough to exceed the storage capacity then
overflow occurs. When runoff falls below the treatment rate then storage is
depleted at the excess treatment rate. The hourly occurrence of treated
runoff, stored runoff, and runoff that has overflowed is tabulated for the
entire record of rainfall. Included in the output is the annual number of
overflow events and the percentage of the runoff that overflowed to the
receiving waters. This type of analysis was carried out for different
storage capacities and treatment rates.
STORM Input Data for Detailed Study of Five Test Cities
STORM requires several input parameters that characterize the urban area
under study. These include hourly precipitation, total area, land use
types and percentages, percent imperviousness and curb length per area for
each land use. Local data used to run STORM on the five study areas were
collected by onsite interviews. The percent imperviousness and length of
street gutters were found by their relationship to population density using
Stankowski's equation for imperviousness11 (see Section IV) and APWA's
equation10 for curb length density, i.e.,
PD
G = 0.0782 - 0.0668(0.839) (10)
LJ
where G = curb length per area, miles/acre, and
J_i
PD, = developed population density, persons/acre.
d
Daily evaporation rates for each month are from a report by Thornthwaite
and Mather.26 The depression storage was assumed to be 0.01 inches
(0.025 cm) for all cities. Street cleaning frequencies are taken from a
1973 survey by APWA of street cleaning practices. A summary of input data
for all of the study areas is given in Table 14. The hydrologic data for
the study areas are shown in Table 15.
Hourly precipitation data were acquired from the US Environmental Data
Service in Asheville, North Carolina. Twenty-five years (January 1948 to
December 1972) of hourly data were obtained for the five test cities. Two
and one-half years (July 1970 to December 1972) of data were obtained for
all stations in the United States.
The frequency distribution of each of the twenty-five years of rainfall was
analyzed for each of the five cities. Little year-to-year variation in the
31
-------
Table 14. STORM INPUT DATA FOR STUDY AREAS
Study Area: Atlanta
Area: 278,400 ac (112,800 ha)
Depression Storage: 0.01 in.(0.025 cm)
Street Sweeping Frequency: 45 days
Daily evaporation rates for each month, Jan-Dec, in in/day (cm/day)
0.01 0.02 0.04 0.07 0.10 0.11 0.10 0.08 0.06 0.04 0.02 0.01
(0.03)(0.05)(0.10)(0.18)(0.25)(0.28)(0.25)(0.20)(0.15)(0.10)(0.05)(0.03)
Study Area: Denver
Area: 187,500 ac (75,900 ha)
Depression Storage: 0.01 in.(0.025 cm)
Street Sweeping Frequency: 46 days
Daily evaporation rates for each month, Jan-Dec, in irVday (cm/day)
0.0 0.0 0.01 0.02 0.04 0.07 0.09 0.08 0.06 0.05 0.03 0.01
(0.0)(0.0)(0.03)(0.05)(0.10)(0.18)(0.23)(0.20)(0.15)(0.13)(0.08)(0.03)
Study Area: Minneapolis
Area: 461,400 ac (186,700 ha)
Depression Storage: 0.01 in»(0.025 cm)
Street Sweeping Frequency: 46 days
Daily evaporation rates for each month, Jan-Dec, in in^day (cm/day)
0.0 0.0 0.0 0.02 0.04 0.06 0.07 0.06 0.05 0.04 0.02 0.0
(0.0)(0.0)(0.0)(0.05)(0.10)(0.15)(0.18)(0.15)(0.13)(0.10)(0.05) (0.0)
Study Area: San Francisco
Area: 435,800 ac (176,400 ha)
Depression Storage: 0.01 in^(0.025 cm)
Street Sweeping Frequency: 46 days
Daily evaporation rates for each month, Jan-Dec, in in«/day (cm/day)
0.01 0.01 0.01 0.02 0.01 0.02 0.02 0.02 0.02 0.02 0.02 0.01
(0.03)(0.03)(0.03)(0.05)(0.03)(0.05)(0.05)(0.05)(0.05)(0.05)(0.05)(0.03)
Study Area: Washington, DC
Area: 316,800 ac (128,200 ha)
Depression Storage: 0.01 in ..(0.025 cm)
Street Sweeping Frequency: 45 days
Daily evaporation rates for each month, Jan-Dec, in in«/day (cm/day)
0.0 0.0 0.01 0.02 0.03 0.05 0.05 0.05 0.03 0.02 0.01 0.0
(0.0)(0.0)(0.03)(0.05)(0.08)(0.13)(0.13) (0.13)(0.08)(0.05)(0.03)(0.0)
32
-------
Table 15. HYDROLOGIC DATA FOR STUDY AREAS
Study Year
Atlanta
Denver
Minneapolis
San Francisco
Washington, DC
Year
1969
1960
1971
1967
1969
Rainfall
in . ( cm)
44.
14.
29.
24.
43.
40
98
29
26
30
(113.
( 38.
( 74.
( 61.
(110.
Imperviousness ,
1/100
0)
0)
4)
6)
0)
0
0
0
0
0
.299
.314
.293
.329
.339
Runoff a
Coefficient, CR
0.374
0.386
0.370
0.397
0.404
Annual Runoff , AR
in. (cm)
16
5
10
9
17
.18
.59
.50
.37
.22
(41.10)
(14.20)
(26.70)
(23.80)
(43.70)
CR = 0.15(100-I)/100 + 0.90 1/100.
DFrom STORM analysis.
-------
distributions was noted, but there was considerable variation among cities.
The annual frequency for twenty-five years of rainfall is shown for the
cities in Figure 8.
In the early stages of the research it became apparent that multiple runs
of STORM would be required on each city to adequately determine the
effectiveness of different storage capacities and treatment rates. It was
also discovered that making STORM runs using the entire twenty-five years
of rainfall for each city was expensive and time consuming.
Since the useful information was in terms of the overall level of control
of the runoff, it appeared adequate to run STORM on a single year if the
results were the same as running STORM for the entire twenty-five year
period. The frequency distributions for the single year chosen for each
city are shown in Figure 9. These may be compared to Figure 8 to see how
the typical year compares to the twenty-five year average. The monthly
distribution for the study year is shown for each study area in Figure 10.
The comparisons indicated that running a single year would be adequate.
RESULTS
Storage/Treatment Isoquants
For each storage/treatment rate combination, there is a value for the
percent of the runoff and pollutants which are "treated." By making
several runs at different combinations of treatment and storage, points
were generated representing different levels of control. Then isoquants
were drawn connecting the points that represent combinations of storage
capacities and treatment rates which give equivalent percent runoff and/or
pollutants "treated." If the concentration of pollutants is constant and
"treatment" efficiency, r\, is 1.0, then percent runoff control is
synonymous with percent pollutant control. Obviously, this is not the case.
Thus, account needs to be taken of
1. treatment efficiency, and
2. variable concentration due to first flush
effects.
Adjustment for Treatment Efficiency
Let R denote the percent runoff control and n equal treatment plant
efficiency. If R denotes the percent pollutant control, then to realize
R-, , one needs to process R, /n of the runoff. Note that R, may be percent
BOD removal, percent SS removal, etc. The following representative treat-
ment efficiencies, in terms of BOD,, removal, were assumed for primary and
secondary devices.
34
-------
cm
200
175
150
>: 125
o
z
UJ
o
Sioo
cc
75
50
25
.3
I
.5
DENVER
SAN FRANCISCO
MINNEAPOLIS
WASHINGTON, D. C.
ATLANTA
0 .02 .04 .06 .08 .10 .12 .14 .16 .18 .20
RAINFALL, in.
Figure 8. Average Twenty-Five Year Rainfall Frequency for
Each Study Area
35
-------
cm
.0
200
175
150
t_
.c
^ 125
o
2
LU
z>
S 100
o:
u.
75
50
.2
1
.3
.4
ATLANTA
MINNEAPOLIS
:NVER
/SAN FRANCISCO
/ WASHINGTON, D. C.
.5
0 .02 .04 .06 .08 .10 .12 .14 .16 .18 .20
RAINFALL, in.
Figure 9. Selected One-Year Rainfall Frequency for
Each Study Area.
36
-------
10
5
J M M J S N
-24
r16
-8
0
E ATLANTA
10
0
._T~L
J M M J S N
r-24
-16
-8
0
E SAN FRANCISCO
<= 10
r24
-16
-8
E DENVER
o
J M M J S N
•0
<
CC
716E WASHINGTON
J M M J S N
I0r
J M M J
MONTH
S N
. I6E MINNEAPOLIS
-8
0
Figure 10. Monthly Rainfall Distribution for Study
Year for Each Study Area
37
-------
Assumed Efficiency,
Treatment Device (BOD,. Removal)
Primary 0.40
Secondary 0.85
Thus, if one desires 25 percent BODr removal with a primary device, then
62.5 percent of the runoff volume must be processed whereas only 29.4
percent of the runoff needs to be processed if a secondary device is
selected. Thus, to convert percent runoff control isoquants to percent
pollutant control isoquants, one uses
Rl
R = A (11)
Adjustment for First Flush
STORM estimates the percent pollutant control as well as percent runoff
control. The STORM runs incorporated the standard first flush assumption
which is used in the model, i.e., the amount of pollutant removal at any
time, t, is proportional to the amount remaining and that a uniform runoff
of one-half inch per hour would wash away 90 percent of the pollutant in
one hour.5 If a first flush is assumed, then storage and treatment can
be operated more effectively because of the greater relative importance
of capturing the initial runoff. The first flush is accounted for by
defining the output in terms of pollutant control directly.
Mathematical Representation of Isoquants
The storage/treatment isoquants are of the form (see Figure 11):
T = TI + (T2 - T1)e~KS (12)
where T = wet-weather treatment rate, inches per hour,
TI = treatment rate at which isoquant becomes
asymptotic to the ordinate, inches per hour,
1 = treatment rate at which isoquant intersects
the abscissa, inches per hour,
S = storage volume, inches, and
K = constant, inch
A relatively large storage reservoir is required to operate the treatment
unit continuously. Thus, first flush effects would be dampened out and the
38
-------
to
LJ
CD
<
01
o
CO
T= T, + (T2-T,)e-K'
T,
TREATMENT, T, in/hr
Figure 11. Definitional Sketch of Storage/Treatment Isoquants
39
-------
effluent concentration from the reservoir should be relatively uniform. Thus,
if stormwater entering the treatment plant has a relatively uniform concen-
tration, then T can be found as follows for 8,760 hours per year:
Ti -
where AR = annual runoff, inches per year,
a = coefficient defined by AR and conversion factors,
and
R = percent runoff control.
By relating the parameters TI , T9~T, and K to the level of control R, one
equation was developed for each of the five cities. The T ~-T, and K terms
versus R were found to be of the following general form:
T2 - Tx = behR (14)
_ f-Q
K = de . (15)
Based on this analysis the following general equation for the isoquants is
obtained:
. (16)
The values of parameters a, b, h, d and f for various cities are presented
in Table 16. The correlation coefficients for each fit were all above 0.99.
The results for the five cities are shown in Figures 12, 13, 14, 15, and 16.
Each figure shows the isoquants calculated by the isoquant equation. Also
shown are some actual data points for a treatment rate of 0.01 inches per
hour and varying amounts of storage. The boundaries of th.e five regions are
shown in Figure 17-
The optimal expansion path can be found using
- MRSST (17)
D
where c = unit cost of storage,
O
c = unit cost of treatment, and
= marginal rate of substitution of storage
for treatment .
40
-------
Table 16. VALUES OF PARAMETERS FOR ISOQUANT EQUATIONS FOR
DEVELOPED PORTION OF THE TEST CITIES
Percent BOD Control with First Flush, n = 1.0.
in
Test City
San Francisco
Denver
Minneapolis
Atlanta
Washington, DC
a
. hr (% R)
(cm hr )
0.0000107
(0.0000271)
0.0000064
(0.0000162)
0.0000120
(0.0000304)
0.0000185
(0.0000469)
0.0000197
(0.0000500)
b
in. hr
(cm hr )
0.002165
(0.005500)
0.001363
(0.003462)
0.001366
(0.003469)
0.002586
(0.006569)
0.001896
(0.004816)
h d
(% R)"1 in."
(cm )
0.03884 211.3
(536.6)
0.04398 185.0
(469.8)
0.04820 241.6
(613.7)
0.04682 190.2
(483.2)
0.04879 228.8
(581.3)
f
(% R)'1
0.03202
0.02792
0.03016
0.03125
0.03393
-------
.00
.000
T, cm/hr
01 02
000
T, cm/hr
004 .008 .012 .016
L_ _ • ' __ 1 - • - 1
.020
.010
TREATMENT ,T, iiVhr
.020
-.40
ANNUAL RUNOFF 9.37 in.
.030
Figure 12. Storage/Treatment Isoquants for Percent BOD Removal
with First Flush - Region I - San Francisco
42
-------
.90-
.00
.80-
.70-
.60-
.50-
.00-
.000
T, cm /hr
.01 .02
T, cm/hr
.000 .094 .008 .0(2 .016
.15-
.10-
.020
-.40
-.30
-.20 (
-.10
.00 |'wpr--|^i =r=-i 1 i 1 .00
.000 .002 .004 .006 .008
T, in/ hr
ANNUAL RUNOFF = 5.59in.
r 1-2
- 1.0
-.60
-.40
-.20
TREATMENT , T, in7hr
Figure 13. Storage/Treatment Isoquants for Percent BOD Removal
with First Flush - Region II - Denver
43
-------
.00
T, cm/hr
.01 .02
T, cm/hr
.000
.90
.80-
.70-
.60-
,50-
.I5H
,004 .008
i . i
.012 .016 .020
,000 .002 .004 .006
T, in/hr
ANNUAL RUNOFF = 10.50 in.
•000
.010
TREATMENT, T, irv/hr
.020
K40
-.30
.008
-.20"
-.10
.00
E
u
to"
ri.2
hi-o
-.80
-.60
E
o
co"
-.40
-.20
..00
.030
Figure 14. Storage/Treatment Isoquants for Percent BOD Removal
with First Flush - Region III - Minneapolis
44
-------
T, cm/hr
.01 .02
T, cm/hr
.000 .004 .008 .012 .016
020
.00
.000
.010 .020
TREATMENT, T, in/hr
-.40
002 .004 ,006
T, in/hr
ANNUAL RUNOFF 16.1810.
.00
.030
Figure 15. Storage/Treatment Isoquants for Percent BOD Removal
with First Flush - Region IV - Atlanta
45
-------
.00
.90
.80-
.70-
.000
T, cm/hr
.01 .02
T, cm/hr
.000 ,004 .008
i 1
.002 .004 .006
T, in/hr
ANNUAL RUNOFF= 17.22 in.
.010
TREATMENT, T, in./hr
.020
Figure 16. Storage/Treatment Isoquants for Percent BOD Removal
with First Flush - Region V - Washington, DC
-------
20 24
Figure 17. Mean Annual Precipitation in the United States, in Inches,
and Regional Boundaries
Weather Bureau Climatic Atlas of the United States, 1968
-------
It is simple to find the optimal expansion path graphically for the five
test cities. Unfortunately, these results need to be extrapolated to
other cities . It appeared that an analytical approach would provide a
more general and consistent procedure. Thus, the isoquant parameters were
adjusted based on the runoff in the city under consideration relative to
the reference city, i.e.,
let AR. = annual runoff in city i, and
AR. = annual runoff in test (reference) city for
J region j (see Figure 17); j = 1,2,3,4,5.
Then, the isoquant coefficients are
AR.
-- ^-3- (18)
J (8.76 x 10 )
AR.
(20)
AR.
dij - AR7V
f . . = f .,
1J J
where a , b.., h. ., d.., and f . . are parameters for city i in region
j and b., h . , d., and f. are the parameters for the test city in
region Jj. The-1 test cities are denoted as follows:
1 San Francisco
2 Denver
3 Minneapolis
4 Atlanta
5 Washington, DC
48
-------
Wet-Weather Quality Control Optimization
The wet-weather optimization problem, assuming linear costs, may be stated
as follows:
minimize
subject to
Z = csS + cTT (23)
T = T± + (T2 - T1)e KS (24)
T,S >_ 0.
Solving this constrained optimization problem yields
S* = max [± In •+ [K(T2 - T^J, 0] (25)
O
where S* = optimal amount of storage, inches,
and
T* = T1 + (T2 - T1)e~KS* (26)
where T* = optimal amount of treatment, inches per hour.
Note that T* is expressed as a function of S*, so it is necessary to find S*
first. Knowing S* and T*, the optimal solution is
Z* = cgS* + cTT* (27)
where Z* = total annual cost for optimal solution, dollars
per acre.
Data needed to estimate TI , T and K have already been presented.
For a primary device, CT = $4,000/mgd = $2,610/acr^"ch ($1.05/m3/day) .
For storage cost,
cc($/acre-inch) = 122 e0-16(PD) (28)
O
where PD = gross population density in persons per acre.
49
-------
The above procedure is applied to the unsewered area of hypothetical city 6
as an example. The isoquant coefficients are determined in Table 17. The
optimal solution is calculated as shown in Table 18.
Table 17. FIRST FLUSH ISOQUANT COEFFICIENTS - UNSEWERED PORTION
OF CITY 6 IN REGION III
Item
Annual
Kunoii , AK
(in./yr) a
Coefficients
b h d
f
Reference City, 10.50 0.0000120 0.001366 0.04820 241.6 0.03016
Region IIIa
Unsewered Portion 7.80 0.0000089 0.001015 0.04820 325.2 0.03016
of City 6
aFrom Table 16.
These results also permit one to decide whether a primary or a secondary
control level is more cost-effective in controlling smaller percentages
of pollution. As seen in Figure 18, a primary control device is less
expensive for low removals (<_ 12 percent), but it loses effectiveness at
higher levels because of the disproportionately large storage requirements.
Costs will be reported for 25, 50, and 75 percent control levels. Thus,
the secondary cost curve can be used in this range. The primary curve will
not be discussed further.
The curves shown in Figure 18 were approximated by functions of the form:
BR-,
Z* = ke (29)
where Z* = total annual cost for optimal solution,
dollars per acre,
k,g = parameters,
R, = percent pollutant removal, 0 <_ RI <_ R , and
R, = maximum percent pollutant removal.
The resulting costs for 25, 50, and 75 percent pollutant control for
combined, storm, and unsewered areas are shown in Tables 19, 20 and 21.
Values of the cost equation parameters are also shown.
50
-------
Table 18. CALCULATION OF OPTIMAL SOLUTION - UNSEWERED AREA OF CITY 6 IN REGION III
Level of
BOD
Type of Control
Control R
Primary 10
25
50
75
Secondary 10
25
50
75
S* - max [-| In (—• [ (
S
c
T
K VT1 CS
240.7 0.00164 7.35
153.1 0.00338 7.35
72.0 0.01129 7.35
33.9 0.03766 7.35
240.7 0.00164 27. L
153.1 0.00338 27.6
72.0 0.01129 27.6
33.9 0.03766 27.6
K)(T -I )], 0)
Unit Costs Optimal
Storage Treatment Storage Treatment Cost - Z*
S** T*h CS CT AnnUal Efficiency
Tl in./ac in./hr-ac $/ac-in. $/ac-in./hr S/ac of Unit, <-.
0.00009 0.00444 0.00065 355 2,610 3.28 0.4
0.00022 0.00875 0.00111 355 2,610 6.00 0.4
0.00044 0.02486 0.00233 355 2,610 14.90 0.4
0.00067 0.06615 0.00467 355 2,610 35.66 0.4
0.00009 0.00993 0.00024 355 9,810 5.87 0.85
0.00022 0.01738 0.00046 155 9,810 10.66 0.85
0.00044 0.04322 0.00095 355 9,810 24.60 0.85
0.00067 0.105J.2 0.00174 355 9,810 54.29 0.85
Net
BOD
Control
Rl
4
10
20
30
8.5
21.2
42.5
63.8
T* - T
-KS*
'"Gross population density of city 6 - 6.67 persons per acre.
dZ* - c(S*) + C(T*)
-------
1000
Zp-Annual cost for primary control unit,
dollars/acre
cost for secondary control
unit, dollars/acre
Zs=4.47e
SECONDARY
0 10 20 30 40 50 GO 70 80 90 100
LEVEL OF POLLUTANT COisJTROL, R,,%
Figure 18. Annual Cost as a Function of Level of Pollutant
Control
52
-------
Table 19. APPROXIMATE ANNUAL COST PER ACRE - COMBINED AREAS
Hypo-
thetical
City
1
2
3
4
5
6
7
Table 20.
Hypo-
thetical
City
1
2
3
4
5
6
7
Cost Equation
k
0.0
0.0
0.0
0.0
0.0
6.52
0.0
Parameters
B
0.0
0.0
0.0
0.0
0.0
0.0392
0.0
APPROXIMATE ANNUAL COST PER
Cost Equation
k
3.96
3.08
4.59
3.34
3.01
5.25
1.78
Parameters
6
0.0377
0.0376
0.0383
0.0379
0.0375
0.0392
0.0372
Annual Control
($/acre)
25%
0
0
0
0
0
17
0
ACRE -
50%
0
0
0
0
0
46
0
Cost
75%
0
0
0
0
0
124
0
STORM AREAS
Annual Control
($/acre)
25%
10
8
12
9
8
14
5
50%
26
20
31
22
20
37
11
Cost
75%
67
52
81
57
50
100
29
53
-------
Table 21. APPROXIMATE ANNUAL COST PER ACRE - UNSEWERED AREAS
Hypo-
thetical
City
1
2
3
4
5
6
7
Cost Equation
k
2.32
2.40
3.78
2.86
2.34
4.47
1.24
Parameters
0
0.0377
0.0376
0.0383
0.0379
0.0375
0.0392
0.0372
Annual Control
25%
6
6
10
7
6
12
3
($/acre)
50%
15
16
26
19
15
32
8
Cost
75%
39
40
67
49
39
85
20
54
-------
Estimating Number of Overflow Events
Some urban areas have used the number of overflow events per year as an
indication of level of control due to different storage/treatment combi-
nations. San Francisco, for example, is concerned with the number of times
the beaches would be closed due to combined sewer overflows. The objective
in this case would be to find the most economical combination of storage
and treatment which would not allow the annual number of overflows to exceed
a predetermined value.
The following procedures were utilized to develop the event definition for
each of the study cities. A one year precipitation record that approximates
its average rainfall distribution was obtained for each city. Precipitation
events were tabulated by varying the number of zero rainfall hours necessary
to divide two separate events.
A "minimum interevent time" is defined as the minimum number of consecutive
zero precipitation hours which must occur between two separate storm events.
By varying the "minimum interevent time" the number of separate storm
events generated is tabulated. The results are shown in Figure 19. Based
on a qualitative analysis, a value of 12 hours is chosen for the national
precipitation event definition. Using this event definition and the results
of the STORM analysis, one can derive the relationship between percent
runoff control and number of overflow events per year. The results, shown
in Figure 20, can be used to transform the final estimates to a base of
events per year. Figure 21 shows the relationship between the percent
pollutant control with and without first flush so that one can convert from
percent runoff control to percent pollutant control.
OVERALL COST ESTIMATE
Overall Results
The only remaining problem is to estimate the area-wide costs for 25, 50,
and 75 percent control. As a first approximation, assume that an overall
25 percent control level is achieved by 25 percent control on the combined
(A,), storm (A~) and unsewered (A,,) areas at annual unit costs (dollars per
acre) of C-, , C~, and Co, respectively. Thus, the approximate total annual
cost, TAG, for city 6 is
(TAC)25 - (C1)25A1 + (C2)25A2 + (C3)25A3. (30)
From Table 2 and Tables 19 to 21, one obtains
(TAC)25 = 17(6000) + 14(2000) + 12(2000)
= $154,000 per year.
55
-------
140
Ui
SAN FRANCISCO
DENVER
MINNEAPOLIS HI
ATLANTA EC
WASHINGTON, D.C
0
0
160
180
Figure 19.
80 100 120 140
MINIMUM INTEREVENT TIME, hours
Effect of Minimum Interevent Time on the Annual Number of Storm Events
200
-------
100
REGION
SAN FRANCISCO I
•— DENVER n
MINNEAPOLIS IE
ATLANTA IZ
WASHINGTON, D,C Y
0
20 40 60 80
% CONTROL OF VOLUME, R
100
Figure 20.
Relationship Between Percent Runoff Control
and Annual Number of Overflow Events
57
-------
100
I.WEST
ROCKY MOUNTAIN
IH MIDWEST
SOUTHEAST
Y NORTHEAST
10 20 30 40 50 60 70 80 90 100
% POLLUTANT CONTROL WITH FIRST FLUSH, %
Figure 21. Relationship Between Percent Pollutant Control With and
Without First Flush
58
-------
Likewise,
(TAC)5Q = $414,000, and (31)
(TAG) = $1,114,000. (32)
The cost of wet-weather control using secondary facilities is
Z* = ke (33)
s
where Z* = total annual cost for optimal solution, dollars
per acre,
k, g = constants, and
R- = percent BOD removal (0 <_ R <_ 85) .
Primary control facilities are not analyzed at this point. If results
indicate an optimal level of control such that primary facilities are
preferrable, then the calculations need to be repeated. The cost of
wet-weather control in terms of pounds of pollutant removed, w, is
Z * = ke (34)
where w = pollutant removal, Ibs/acre-yr, and
M = pollutants available, Ibs/acre-yr.
The marginal cost of BOD removal is
1003 w
dZ
M
dw M
Given these convex cost functions, the optimal mix of control of storm
runoff from combined, storm, and unsewered areas is found by equating
marginal costs. Using equation 35 with the subscript (1) denoting
combined, (2) denoting storm, and (3) denoting unsewered, yields
59
-------
ioo g-V 100
100 B k M^ 100
(36)
100 63w3
100 Bk M
s
If marginal costs for, say, 50% BOD removal in city 6 are compared, data
from Tables 11 and 19 to 21 are used to obtain
M_ 100(0. 0392) (6. 52) .100(0.0392) (0.5)
MC1 ~ 117.6 e
= $1.54/lb-BOD ($3.80/kg-BOD)
MC = 100(0. 0392) (5. 25) &100(0. 0392) (0.5) (3g
/. Z.J • 3
= $5.73/lb-BOD ($14.15/kg-BOD)
100(0. 0392) (4. 47) 100(0.0392) (0.5) (39)
ML3 23.6 e
= $5.27/lb-BOD ($13.02/kg-BOD).
This result indicates that, to achieve 50 percent control, storm and
unsewered areas should be controlled less intensively due to their
relatively high marginal costs and combined sewered areas should be
controlled more intensively because of their relatively low marginal
costs.
The correct solution can be found by solving for w, and w,. as functions
of w2, i.e. ,
Wl = a!2 + b!2W2 ° - wl - Ml (40)
W3 = a32 + b32W2 ° - W3 - M3
60
-------
M 3 k M
Where a
a32
32 M
b!2= <><>
32 M
b32 = 0 (M >
•Jt- P o IJ-o
M , M , M , 3 , 3 , 3 , k , k , k , w , w , w are as
-J- - X- -. j J _ JL ^ , ^i -J J. ^ J _L ii j
defined earlier.
The total wet-weather pollution load, WP, in pounds per year, is
WP = Z M.A. (42)
11
where M. = annual pounds per acre from i area, and
A r -rth
A. = area of i area.
Let p denote the proportion of WP that one wishes to control. Then, the
optimal solution for a given p is found by substituting equations 40 and
41 into 42, or
p (WP) = w^ + w2A2 + w3A3 (43)
p(WP) = (a12 + b12w2)A1 + w2A2 + (a32 + b32w2)A3 (44)
p(WP) - a,?A, - a—A.,
"*2 = b A + A + b A °±W21M2- (45)
2 b!2 1 A2 b32A3
Knowing w*, one can find w* and w* by substituting into equations 40 and
41. l L
The optimal percent control of the i source for control level, p, in
terms of w* is
100 (w*)
(46)
61
-------
If there is no storm sewered area, then
= a.. 0 + b1clwQ 0 < w < M (47)
U -Lj J — J. — 1
M 3 k M
where a = ln [ < ( (> ] • and
3 M
bi3 = (ir> •
Then,
p (WP) = w + wA (48)
or
P(WP) - a A
0±W^M' (49)
An example calculation for city 6 is shown below.
A. Find a, a> a> b, b, and b.
M 3 k M
)(-)] (50
117.6 r/Q. 0392, ,5. 25,
100(0.0392) IV0.0392'V6.52'v 25.
= 39.5.
Likewise a«2 = 0.505 and a,,. = 37.0.
32 Ml
b!2 = <]> >
0.0392117.6>
25. 5;
b12 = 4.61.
Likewise, b32 = 0.925 and b,_ = 4.98.
62
-------
B. Find optimal w as a function of control level, p.
P(WP) -a-a
= P(803,800) - 39.5(6000) - 0.505(2000)
4.61(6000) + 2000 + 0.925(2000)
w* = 25.51p - 7.55 0 <_ w* <_ 25.5.
Then,
Wl = a!2 + b!2w5 <53)
= 39.5 + 4.61(25.51p - 7.55)
= 117.6p + 4.69 0 <_ w* <_ 117.6
and
w* = 23.6p - 6.48 0 <_ w* <_ 23.6. (54)
For 50 percent control,
, -, , Ibs BOD
w* = 63.5 ,
1 acre-yr
«*. 5.20^^^.,
2 acre-yr
._
3 acre-yr
C. Check. Does
w*(A1) + w*(A2) + w*(A3) - p(WP)? (55)
63.5(6000) + 5.20(2000) + 5.32(2000) = 0.50(803,800)?
402,000 = 402,000 OK
D. Find the optimal percent control for areas 1, 2, and 3.
100(w*)
50 = -^ (56)
) - 54-°-
63
-------
Likewise (R2)5Q =20.4 and (R3)50 = 22.5.
E. Find the optimal overall cost per acre.
0.0392(R ),.
(Z1)§Q = 6.52 e (57)
= 6.52 e°-0392(54-0)
(Vso = $54-15 per acre-
Likewise (Z2)|0 = S11-68 Per acre and ^Z3^Q = $10-80 Per acre.
F. Find the total annual cost for 50 percent control.
Total annual cost = I (Z.)* A (58)
1=1
= (54. 15) (6000) + 11.68(2000)
+ 10.80(2000)
= $370,000.
Note that the approximate solution has a total annual cost of $414,000.
Thus, the marginal cost procedure reduced costs by over 10 percent.
Using the marginal cost procedure yields the overall costs per acre
shown in Table 22. The total costs are shown in Table 23. The values of
k and £ are derived using the three data points, i.e., 25, 50 and 75 per-
cent control levels.
Tertiary Treatment versus Wet-Weather Treatment
The optimal mix of tertiary treatment for additional organic pollutant
control and wet-weather control can be found by equating the marginal
cost of tertiary treatment with the marginal cost of wet-weather pollu-
tion control. The estimated total annual incremental cost of a tertiary
treatment plant for additional organic pollutant control is:13
Ctert = 87'000 D' (59)
64
-------
Table 22. OPTIMAL TOTAL ANNUAL COSTS PER ACRE
Hypo-
thetical
City
1
2
3
4
5
6
7
Average
k
0.133
0.063
0.761
0.310
0.320
4.70
0.150
6.13
Table 23.
Hypo-
thetical
City
1
2
3
4
5
6
7
Dollars per Acre
e
0.0375
0.0372
0.0383
0.0379
0.0367
0.0408
0.0372
0.0402
OPTIMAL
25
6.80
6.96
11.00
8.00
6.96
12.95
3.80
11.33
TOTAL ANNUAL
50
17.20
17.39
28.07
20.60
17.39
37.00
9.60
31.85
COSTS
75
44.40
44.78
74.61
53.10
43.48
98.63
24.40
84.37
Total Annual Cost
(thousands)
25%
3.4
1.6
19.8
8.0
0.8
128.5
3.8
50%
8.6
4.0
51.6
20.6
2.0
370.0
9.6
75%
22.2
10.3
134.3
53.1
5.0
986.3
24.4
Total
165.9
466.4
1235.6
65
-------
where C = total annual incremental cost of tertiary
tert
treatment plant, dollars per year, and
D = plant size, mgd.
The tertiary plant increases BOD removal efficiency from n to n fc
so that the additional pollutant removal is (n ^ ~ n )Hr> where
L&2TL S£C JJW
M-. is the annual dry—weather BOD load. Thus, the unit cost of tertiary
,_Dw . . .
treatment is
tert (\ert ~
where c = unit cost of BOD removal, dollars per pound.
tert
Equating the marginal cost of wet-weather control to the unit cost of
tertiary treatment yields
(60)
1006
w
1003 k M
ctert = "If— e
or
M ctert(M)
W* ~ TOOT ln [1006 (k)J (62)
where w* = optimal pounds of wet-weather pollution control
prior to using tertiary treatment.
The optimal percent control in terms of R.. is
R* = max (? In I1QQg (fc)] , 0) . (63)
The overall average BOD loading per acre, M, is
(64)
M =
A
Atot
For city 6, the solution can be found as follows.
A. Find the unit cost of tertiary treatment.
Assume n = 0.95, n = 0.85. For city 6, with a population of
100,000 people and average (Table 7) per capita sewage flow of 100 gallons
per day, the approximate plant size is 10 mgd. From Table 10, >L = 566
pounds BOD per acre-year. Thus,
66
-------
= 87,000 D°-787
Ctert (n - n )M (A) (65)
tert sec Dw
- 87.000 IP0'787
(0.94 - 0.85)566(10,000)
c = $0.86 per pound of BOD.
B. Find the weighted average BOD load per acre.
From Table 11, M, = 80.4 pounds of BOD per acre.
C. Find the optimal level of wet-weather control, R*, to initiate
prior to using tertiary treatment.
Using data from Table 22,
R*=max [±ln IJ , OJ (66)
, 1 r 0.94(80.4) n n1
max L0.0408 ln 1100(0.0408)(4.70)J' J
R* = 33.6 percent.
Thus, for these assumed conditions, approximately 34 percent of the wet-
weather pollution should be controlled prior to initiating tertiary
treatment. While these results are for one specific set of assumptions,
they do suggest that is highly desirable to do this tradeoff analysis
before committing a community to tertiary treatment.
Potential Savings Due to Multipurpose Planning
The cost of wet-weather quality control can be reduced by integrating
this purpose with dry-weather sewage treatment plants and/or storage
facilities for stormwater quantity control. Therefore, it has been
suggested that flow equalization be considered as an alternative to con-
ventional design. The storage volume needed for dry-weather flow
equalization is estimated to be 10 to 20 percent of the average annual
dry-weather flow. Integration of wet-weather quality control with dry-
weather control affords the opportunity for equalization of dry-weather
flow since the wet-weather control, in general, must be accomplished through
a combination of storage and treatment. Therefore, if a sewage treatment
facility is designed on the basis of peak flow, equalization would result
in some excess capacity at this facility. Utilization of this excess
capacity can reduce the treatment capacity needed for wet-weather quality
control.
67
-------
Wet-weather quantity control can be accomplished through storage.
Utilization of this storage for accomplishing wet-weather quality control
can reduce the storage and treatment requirements for wet-weather quality
control. A rough estimate of the potential savings by integrating these
three purposes can be made as follows.
As part of the nationwide assessment13, the proportion of control costs
assignable to wet-weather pollution control were determined assuming that
excess capacity equal to the average daily flow exists in the sewage
treatment plant and that on-site detention of the two-year, twenty-four-
hour storm is required for stormwater quantity control. The cost allocation
factor, a, shown in Figure 22 represents the proportion of single purpose
cost which wet-weather pollution would pay in a multipurpose project for the
five regions. Figure 23 shows how a varies for various assumptions
regarding excess capacity and required storage for quantity control. For
example, if excess capacity exists, and a level of 15 percent BOD control
is desired, then the proportion of the total cost assignable to wet-weather
quality control is about 80 percent (a = 0.80). Based on these results,
the cost allocation proportions shown in Table 24 can be used as a first
approximation.
Table 24. COST ALLOCATION FACTORS FOR MULTIPURPOSE SYSTEMS
Cost Allocation Factor, a, for Various Purposes
/» r uxj-ULctiiu
Control, R
10
25
50
75
2 alone
1.0
1.0
1.0
1.0
I3 and 2
0.70
0.90
0.96
0.98
2 and 3°
0.57
0.60
0.60
0.73
1, 2, and 3
0.30
0.46
0.50
0.70
Sanitary sewage treatment facility.
Stormwater pollution control facility.
"Stormwater quantity control facility.
68
-------
0.0
20
30 40 50 60 70
% POLLUTANT CONTROL, R,
80
90
100
Figure 22. Cost Allocation Factor for Five Cities
-------
2 YEAR (E- V- 0)
1.0
cc
£
o
0.6-
z
o
o
o
CO
o
o
0.4-f
0.2-
0.0-
YEAR (E i V)
100 YEAR (E+V)
E - AVAILABLE EXCESS CAPACITY IN DRY-WEATHER PLANT
V= AVAILABLE STORAGE VOLUME IN STORM WATER QUANTITY CONTROL FACILITY
—i—
20
—i—
40
—i—
60
—i—
90
10
30
50
70
80
% POLLUTANT CONTROL, R,
100
Figure 23. Effect of Design Storm and Number of Purposes on Cost Allocation Factor for
Various Levels of Control - Midwestern US
-------
REFERENCES
1. Bowers, C. E. Harris, G. S. and Pabst, A. F., "The Real-Time Computation
of Runoff and Storm Flow in the Minneapolis-St. Paul Interceptor
Sewers," St. Anthony Falls Hydraulic Laboratory, Memo No. M-118, Uni-
versity of Minnesota, Minneapolis, December 1968.
2. Leiser, C. P., "Computer Management of a Combined Sewer System,"
Office of Research and Development, USEPA Report EPA-670/2-74-022,
NTIS-PB 235 717, July 1974.
3. Brandstetter, A. B., "An Assessment of Mathematical Models for Storm
and Combined Sewer Management," USEPA Report (At Press), 1976.
4. Huber, W. C., "Modeling for Storm Water Strategies," APWA Reporter,
Vol. 42, No. 5, pp. 10-14, May 1975.
5. Hydrologic Engineering Center, Corps of Engineers, "Urban Storm Water
Runoff: STORM," Generalized Computer Program 723-58-L2520, May 1975.
6. Roesner, L. A., et al., "A Model for Evaluating Runoff-Quality in
Metropolitan Master Planning," ASCE Urban Water Resources Research
Program, Technical Memo No. 23, ASCE, 345 E 47 St., NY, NY 10017,
72 pp., April 1974.
7. Huber, W. C., Heaney, J. P., et al., "Storm Water Management Model
User's Manual Version II," Office of Research and Development, USEPA
Report EPA-670/2-75-017, March 1975.
8. DiGiano, F. A. and Mangarella, P. A., eds., "Application of Storm-
water Management Models," USEPA Report EPA-670/2-75-065, June 1975.
9. Graham, P- H., Costello, L. S. and Mallon, H. J., "Estimation of
Imperviousness and Specific Curb Length for Forecasting Stormwater
Quality and Quantity," JWPCF, Vol. 46, No. 4, pp. 717-725, April 1974.
10. American Public Works Association,"Nationwide Evaluation of Combined
Sewer Overflows and Urban Stormwater Discharges, Volume III: Charac-
terization," USEPA Report (At Press), 1976.
11. Stankowski, S. J., "Magnitude and Frequency of Floods in New Jersey
with Effects of Urbanization," Special Report 38, US Geological
Survey, Water Resources Division, Trention, NJ, 1974.
71
-------
12. American Public Works Association and University of Florida,
Evaluation of the Magnitude and Significance of Pollution Loading
from Urban Stormwater Runoff, Ontario, Department of the Environment,
Toronto, 1976.
13. Heaney, J. P., W. C. Huber et al.,"Nationwide Evaluation of Combined
Sewer Overflows and Urban Stormwater Discharges. Volume II: Cost
Assessment and Impacts," USEPA. Report (At Press), 1976.
14. Pisano, W. C.,'tost Effective Approach for Combined and Storm Sewer
Cleanup," in Proc. Urban Stormwater Management Seminar, USEPA Report
WPD 03-76-04, 1976.
15. James, L. D. and Lee, R. R., Economics of Water Resources Planning,
McGraw-Hill, Inc., NY, 1971, 615 pp.
16. Field, R. I. and Struzeski, E. J., Jr., "Management and Control of
Combined Sewer Overflows," JWPCF, Vol. 44, No. 7, 1972, pp. 1393-1415.
17. Lager, J. and Smith, W., "Urban Stormwater Management and Technology:
An Assessment," USEPA Report EPA-670/2-74-040, NTIS-PB 240 697, 1974.
18. Becker, B. C. et al., Approaches to Stormwater Management, Hittman and
Associates, USDI Contract 14-31-001-9025, 1973.
19. DiToro, D. M., "Statistical Design of Equalization Basins," J.E.E.
Div., ASCE, Vol. 101, No. EE6, December 1975, pp. 917-933.
20. Field, R. I. et al., Give Stormwater Pollutants the Spin,
The American City and County, 1976, pp. 77-78.
21. Maher, M. B., "Microstraining and Disinfection of Combined Sewer
Overflows - Phase III," USEPA Report EPA-670/2-74-049,
NTIS-PB 235 771, 1974.
22. Agnew, R. W. et al., "Biological Treatment of Combined Sewer Overflow
at Kenosha, Wisconsin," USEPA Report EPA-670/2-75-019,
NTIS-PB 242 120, 1975.
23. Battelle-Northwest, Evaluation of Municipal Sewage Treatment
Alternatives, Council on Environmental Quality, NTIS-PB 233 489,
1974.
24. Benjes, H. et al., "Estimating Initial Investment Costs and Operation
and Maintenance Requirements of Stormwater Treatment Process," USEPA
Cont. EPA-68-03-2186 (unpublished), 1975.
25. Wiswall, K. C. and Robbins, J. C., "Implications of On-Site Detention
in Urban Watersheds," ASCE Hyd. Div. Conf., Seattle, WA, 1975.
72
-------
26. Thornthwaite, C. W. and Mather, J. R., "Instructions and Tables for
Computing Potential Evapotranspiration and the Water Balance," Drexel
Institute of Technology, Publications in Climatology, Centerton, NJ,
Vol. 3, No. 3, 1957.
73
-------
GLOSSARY
Antecedent conditions: Initial conditions in catchment as determined from
hydrologic events prior to storm.
Combined sewage: Sewage containing both domestic sewage and surface water
or stormwater, with or without industrial wastes. Includes flow in heavily
infiltrated sanitary sewer systems as well as combined sewer systems.
Combined sewer: A sewer receiving both intercepted surface runoff and
municipal sewage.
Combined sewer overflow: Flow from a combined sewer in excess of the
interceptor capacity that is discharged into a receiving water.
Depression storage: Amount of precipitation which can fall on an area
without causing runoff.
Detention: The slowing, dampening, or attenuating of flows either enter-
ing the sewer system or within the sewer system by temporarily holding
the water on a surface area, in a storage basin, or within the sewer
itself.
Domestic sewage: Sewage derived principally from dwellings, business
buildings, institutions, and the like. It may or may not contain ground-
water.
Economies of scale: Unit costs decrease as output increases.
Equalization: The averaging (or method for averaging) of variations in flow
and composition of a liquid.
Expansion path: Locus of points connecting numerous isoquants indicating
the optimal combination of inputs.
First flush: The condition, often occurring in storm sewer discharges and
combined sewer overflows, in which a disproportionately high pollutional
load is carried in the first portion of the discharge or overflow.
Frequency diagram: Curve which relates the number of occurrences of events
to their magnitude.
74
-------
Isocost lines: Lines of equal cost.
Isoquants: Curves representing combinations of the inputs yielding the
same amount of output.
Physical-chemical treatment processes: Means of treatment in which the
removal of pollutants is brought about primarily by chemical clarification
in conjunction with physical processes. The process string generally
includes preliminary treatment, chemical clarification, filtration, carbon
adsorption, and disinfection.
Precipitation event: A precipitation event terminates if zero rainfall
has been recorded for the previous specified time interval.
Primary treatment: Process which removes about 40% of the biochemical
oxygen demand of the waste.
Retention: The prevention of runoff from entering the sewer system by
storing on a surface area or in a storage basin.
Runoff coefficient: Fraction of rainfall that appears as runoff after
subtracting depression storage and interception. Typically accounts for
infiltration into ground and evaporation.
Sanitary sewer: A sewer that carries liquid and water-carried wastes from
residences, commercial buildings, industrial plants, and institutions,
together with relatively low quantities of ground, storm, and surface
waters that are not admitted intentionally.
Secondary treatment: Process which removes about 85% of the biochemical
oxygen demand of the waste.
Storm flow: Overland flow, sewer flow, or receiving stream flow caused
totally or partially by surface runoff or snowmelt.
Storm sewer: A sewer that carries intercepted surface runoff, street wash
and other wash waters, or drainage, but excludes domestic sewage and
industrial wastes.
Storm sewer discharge: Flow from a storm sewer that is discharged into a
receiving water.
Stormwater: Water resulting from precipitation which either percolates
into the soil, runs off freely from the surface, or is captured by storm
sewer, combined sewer, and to a limited degree sanitary sewer facilities.
Surface runoff: Precipitation that falls onto the surfaces of roofs,
streets, ground, etc., and is not absorbed or retained by that surface,
thereby collecting and running off.
Tertiary treatment: Process which removes about 95% of the biochemical
oxygen demand of the waste.
75
-------
Urbanized area: Central city, or cities, and surrounding closely settled
territory. Central city (cities) have population of 50,000 or more.
Peripheral areas with population density of 1,000 persons per acre or
more are included.
Urban runoff: Surface runoff from an urban drainage area that reaches
a stream or other body of water or a sewer.
76
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/2-76-275
3. RECIPIENT'S ACCESSION-NO.
.. TITLE AND SUBTITLE
5TORM WATER MANAGEMENT MODEL:
SCREENING PROCEDURES
LEVEL I - PRELIMINARY
5. REPORT DATE
October 1976 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James P. Heaney
8. PERFORMING ORGANIZATION REPORT NO.
Wayne C. Huber Stephan J. Nix
9. PERFORMING ORGANIZATION NAME AND ADDRESS
)epartment of Environmental Engineering Sciences
University of Florida
Gainesville, FL 32611
10. PROGRAM ELEMENT NO.
1BC611
11. CONTRACT/GRANT NO.
R-802411
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
Project Officer: Richard Field, Phone: 201/548-3347 x503 (8-342-7503)
16. ABSTRACT
The original USEPA Storm Water Management Model (SWMM) provides a detailed simulation
of the quantity and quality of stormwater during a specified precipitation event
lasting a few hours. This model is widely used. However, it is too detailed for
many users. In particular, the 208 planning effort needs simplified procedures to
permit preliminary screening of alternatives. In response to this need, four levels
of stormwater management models have been prepared and are being released this year.
This initial volume presents a "desktop" procedure which was developed to do a nation-
wide assessment of stormwater pollution control costs. The next three models will be
computer based and provide increasing amounts of detail.
The desktop procedure permits the user to estimate the quantity and quality of urban
runoff in the combined, storm, and unsewered portions of each urban area in his
jurisdiction. Using generalized results from the nationwide assessment, the optimal
mix of storage and treatment and its associated costs may be estimated. Also,
comparisons between tertiary treatment and stormwater management are presented.
Lastly, possible savings due to integrated management of domestic wastewater, storm-
vater quality, and stormwater quantity are evaluated.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
*Storm sewers, *Water pollution, Control
simulation, *Cost effectiveness, *Waste
treatment, *Sewage treatment, *Surface
water runoff, *Runoff, ^Combined sewers,
^Mathematical models, Storage tanks,
Methodology, Economics
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Simplified evaluation
13B
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
95
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
. S. GOVERNMENT PRINTING OFFICE: 1976-757-056/5J)02 Region No. 5-11
77
------- |