EPA-600/2-75-071
December 1975 Environmental Protection Technology Series
DETENTION TANK FOR
COMBINED SEWER OVERFLOW
Milwaukee, Wisconsin, Demonstration Project
Municipal Environmental Research Laboratory
0fficfi.itf Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/2-75-071
December 1975
DETENTION TANK FOR COMBINED
SEWER OVERFLOW
Milwaukee, Wisconsin, Demonstration Project
by
City of Milwaukee, Wisconsin
Department of Public Works
Milwaukee, Wisconsin 53202
and
Consoer, Townsend and Associates
Consulting Engineers
Chicago, Illinois 60611
Project No. 11020FAU
Project Officer
Clifford Risley, Jr.
U.S. Environmental Protection Agency
Region V
Chicago, Illinois 60606
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
AGENCY
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
ii
-------
FOREWORD
Man and his environment must be protected from the adverse
effects of pesticides, radiation, noise and other forms of pollution,
and the unwise management of solid waste. Efforts to protect the
environment require a focus that recognizes the interplay between
the components of our physical environment--air, water, and land.
The Municipal Environmental Research Laboratory contributes to
this multidisciplinary focus through programs engaged in
• studies on the effects of environmental contaminants
on the biosphere, and
• a search for ways to prevent contamination and to
recycle valuable resources.
This Report studies the environmental effects of combined
sewer overflows and evaluates a potential method for controlling
the combined sewer overflows and reducing environmental contam-
inants.
Louis W. Lefke
Acting Director
Municipal Environmental
Research Laboratory
ill
-------
ABSTRACT
The City of Milwaukee evaluated the merits of detention tanks
as a practical method for abatement of combined sewer overflow
pollutional discharges from urban areas.
A 3. 9 million gallon combined sewer overflow detention tank
was constructed to intercept overflow from a 570 acre segment of the
City's combined sewer area. As part of the evaluation program, an
extensive sewer and river monitoring program was conducted, utilizing
eleven automated monitoring stations. The monitoring program pro-
vided data utilized with a mathematical detention tank model to evaluate
performance of the project detention tank and provides a basis for other
design and planning situations.
Based upon approximately 5 years of data and modeling studies,
detention tanks were shown to be effective in preventing a large portion
of the contaminants found in combined sewer overflow from entering
receiving waters. General information and methods for sizing and
estimating costs of detention tanks for other areas have been developed.
This information was utilized to establish preliminary cost estimates
for providing similar facilities to serve the entire combined sewer area
tributary to the Milwaukee River in the City.
This report was submitted in fulfillment of Project
Number 11020FAU by the City of Milwaukee, Wisconsin under the
partial sponsorship of the Environmental Protection Agency.
Data collection was completed in November, 1972. Work was
completed as of October 30, 1974.
IV
-------
TABLE OF CONTENTS
Page
Abs
Lls
Lis
tract
t of Figures
t of Tables
Acknowledgements
Sec
I
II
III
IV
V
VI
tions
CONCLUSIONS
Project Area
Combined Sewer Overflow Detention Tank
Milwaukee River
RECOMMENDATIONS
INTRODUCTION
PROJECT DESCRIPTION
Project Study Area
Project Area Combined Sewers
Milwaukee River
River Flushing Tunnel
PROJECT MONITORING SYSTEMS
Combined Sewer Monitoring Station Function.
Combined Sewer Monitoring Station Component
Equipment
River Monitoring Station Function
River Monitoring Station Component
Equipment
Rainfall Gauging
Monitoring Station Equipment Suppliers
Station Operation and Maintenance
Monitoring Station Cost Data
DETENTION TANK
General
Detention Tank Component Equipment
Detention Tank Operation
Detention Tank Maintenance
Personnel Requirements
Detention Tank Cost Data
iv
viii
xiii
xvi
1
1
5
9
12
15
19
19
19
23
23
26
30
39
47
47
52
55
55
57
59
59
65
73
74
76
76
-------
TABLE OF CONTENTS
(CONT.)
VII DATA COLLECTION PROGRAM 83
General 83
Rainfall Data 85
Sewer Monitoring g5
Detention Tank Monitoring 87
River Monitoring 87
Laboratory Analyses 91
Analytical Procedures 94
Data Handling 97
Operation and Maintenance Problems and
Recommendations 97
VIII SUMMARY AND ANALYSIS OF DATA 105
Milwaukee River Quality 105
Sewage - Dry Weather Quality 121
Sewage - Dry Weather Flow 133
Rainfall 133
Sewage - Wet Weather Flow (Runoff Coefficient
CR) 143
Sewage - Wet Weather Quality 14g
Sewage - Wet Weather Quality Correlations 155
Detention Tank Performance 156
IX STORM OVERFLOW MODELING 173
Storm Detention Tank Model Description 173
Model Coefficients 175
Model Output - Analysis of Detention Tank
Performance 177
River Water Quality Model 190
River Water Quality Model Description 290
River Water Quality Model Output and Verification
(Dry Weather) 209
River Water Quality Model Output and Verfication
(Wet Weather) 220
VI
-------
TABLE OF CONTENTS
(CONT.)
Pag
X COMBINED SEWER OVERFLOW DETENTION
TANK APPLICATION AND MILWAUKEE
RIVER WATER QUALITY 235
Discussion of Capacilities of Detention Tanks
in Control of Combined Sewer Overflow 235
Factors Influencing Water Quality in the
Milwaukee River 243
Detention Tank Application - City of Milwaukee 257
XI REFERENCES 263
XII APPENDICES 264
VI1
-------
LIST OF FIGURES
Figure No. Page
1 Project Critical Path Flow Chart 17
2 Project Area Location Map 20
3 Location of Combined Sewer Monitoring Stations 27
4 Location of Detention Tank and River Monitoring Stations 28
5 Schematic of Inputs to Combined Sewer Flow Metering 31
bystem
6 Monitoring Component Location Map 32
7 Monitoring Station with Gate Orifice Flow Device-MIS 33
Chamber
8 Location Map - Stations 21, 45 and 49 34
9 Monitoring Station With Flume Flow Device-Station 25 36
10 Monitoring Station With Nozzle Flow Device-Station 29 37
11 Location Map-Stations 33, 37 and 41 38
12 Schematic Flow Diagram for Monitoring Stations 41
13 Sewer Monitoring Housing and Equipment Layout 48
14 Combined Sewer Monitoring Station-Photograph 49
15 Combined Sewer Monitoring Station-Photograph 50
16 River Monitoring Housing and Equipment Layout 51
17 River Monitoring Station-Photograph 53
18 River Monitoring Station-Photograph 54
19 Detention Tank Sectional Plan 61
20 Detention Tank-Section 62
viii
-------
LIST OF FIGURES
(CONT.)
Figure No. Page
21 Detention Tank - Section 63
22 Detention Tank - Photograph 64
23 Log of Accumulated Sediment 70
24 Detention Tank Construction Cost vs. Detention Tank
Capacity 79
25 Control Building and Appurtenances Cost vs.
Engineering News Record Construction Cost Index 81
26 Milwaukee River Sampling Stations 107
27 River Quality At Station 52 vs. Rainfall 115
28 River Quality At Station 52 vs. Rainfall 116
29 Milwaukee River Quality Changes Due to Rainfall 119
30 Milwaukee River Quality Changes Due to Rainfall 120
31 Hourly Variation - Dry Weather
Sewage Quality - Winter Data 129
32 Hourly Variation - Dry Weather
Sewage Quality - Summer Data 130
33 Constituent vs. Time Since Last Storm 131
34 Constituent vs. Time Since Last Storm 132
35 Average Dry Weather Sewer Flows For March-May 1972 134
36 Dry Weather Sewer Flows Following Rainfall for Sept. 135
17-18,1970
37 Schematic of Combined Sewer Flow Metering System 145
38 Effect on Storm Characteristics on Runoff Coefficient 149
39 Effect of Storm Characteristics on Runoff Coefficient 150
IX
-------
LIST OF FIGURES
(CONT.)
Figure No. Page
40 Effect of Pump Out Rate On Detention
Tank Performance 182
41 Effect of Detention Tank Volume on BOD Removal 184
42 Effect of Detention Tank Volume on Suspended Solids
Removal 185
43 Effect of Interceptor Capacity on BOD and Suspended
Solids Loading to Detention Tank 187
44 Temporal D.O. Distribution - Survey I 192
45 Temporal D.O. Distribution - Survey III
(North Ave. Bridge Station) 194
46 Temporal D.O. Distribution - Survey III
(Downstream Stations) 195
47 Temporal D.O. Distribution - Survey II 198
48 Temporal D.O. Distribution - Survey IV 202
49 Notation for Finite Segments 204
53 Study Area 206
51 Model Segmentation 207
52 Observed vs. Calculated Data - Survey I
(Humboldt Ave. ) 210
53 Observed vs. Calculated Data - Survey I
(Cherry Street) 211
54 Observed vs. Calculated Data - Survey I
(St. Paul Avenue) 212
55 Observed vs. Calculated Data - Survey I
(Water Street) 213
-------
LIST OF FIGURES
(CONT.)
Figure No. Page
56 Observed vs. Calculated Data - Survey III
(Humboldt Ave. , Cherry Street) 216
57 Observed vs. Calculated Data - Survey III
(St. Paul Ave. , Water Street) 217
58 Observed vs. Calculated Data - Survey III
(Humboldt Ave. , Cherry Street) 218
59 Observed vs. Calculated Data - Survey III
(St. Paul Ave., Water Street) 219
60 Observed vs. Calculated Data Survey IV
(Humboldt Ave. , Cherry Street) 222
6l Observed vs. Calculated Data Survey IV
(St. Paul Ave. , Water St. ) 223
62 Temporal Dissolved Oxygen Distribution
(Milwaukee River - Humboldt Ave. ) 22%
63 Temporal Dissolved Oxygen Distribution
(Milwaukee River - Cherry Street) 229
64 Temporal Dissolved Oxygen Distribution
(Milwaukee River - St. Paul Ave. ) 230
65 Temporal Dissolved Oxygen Distribution
(Milwaukee River - Water Street)
66 Combined Sewer Overflow BOD and Suspended
Solids Removal As A Function of Tank Size 236
67 Unit Size Removal Efficiencies For Combined
Sewer Overflow Detention Tanks 237
68 Volumetric Efficiency of Combined Sewer
Overflow Detention Tanks 238
XI
-------
LIST OF FIGURES
(CONT.)
Figure No. Page
69 Contaminant Removal vs. Volumetric Efficiency 240
70 Comparison of Contaminant Retention vs.
Volumetric Retention Wet Year 241
71 Comparison of Contaminant Retention vs.
Volumetric Retention Dry Year 242
72 Coliform Count in Detention Tank Overflow
vs. Tank Throughput Rate 244
73 Model Application at Humboldt Ave. - Survey I 247
74 Model Application at Cherry St. - Survey I 248
75 Model Application at St. Paul Ave. - Survey I 249
76 Model Application at Water St. - Survey I 250
77 Model Application at Humboldt Ave. and
Cherry St. - Survey III 252
78 Model Application at St. Paul Ave. and
Water Street - Survey III 253
79 Dissolved Oxygen Deficit Due to Benthal
Oxygen Demand 254
80 Detention Tank Location Map 260
81 Cost Per Unit Area Served For Detention
Tank Construction 262
Xll
-------
LIST OF TABLES
Table No. Page
1 Milwaukee - Combined Sewers 22
2 Combined Sewer Outfalls Into Milwaukee River 24
3 Monitoring Station Identification 29
4 Flume Design Data 40
5 Nozzle Design Data 42
6 Monitoring Station - Flow Instrumentation 43
7 Monitoring Station - Cost Data 58
8 Humboldt Avenue Detention Tank Bid Prices 77
9 ENR Construction Cost Index 78
10 Humboldt Avenue Detention Tank Operating Costs 82
11 General Operating Cost Budget Typical Detention 82
12 Analyses Performed 92
13 Schedule of Analyses 93
14 River Stations 106
15 Seasonal Distribution of Data Collection 109
16 Five Year Average - River Water Quality HO
17 Five Year Average - River Water Quality HI
18 Seasonal River Water Quality 112
19 Seasonal River Water Quality 113
20 Periods When Samples Were Collected At
Frequent Intervals to Study Hourly Variation
in River Quality 117
xiii
-------
LIST OF TABLES
(CONT.)
Table No. Page
21 River Sampling and Storm Event Analysis Dates 118
22 Flow Record - Milwaukee River 121
23 Dates of Dry Weather Sewage Samples 122
24 Dry Weather Sewage Quality 124
25 Dry Weather Conditions - Average Sewage Quality 125
26 Dry Weather Sewage Quality
Dirunal Variation - Winter Samples 126
27 Dry Weather Sewage Quality
Dirunal Variation - Summer Samples 127
28 Precipitation Record - Milwaukee 137
29 Comparison of Rain Gauge Records 138
30 Comparison of Annual Rainfall at U.S. Weather
Bureau Stations in Milwaukee 141
31 Data Summary of Rainfall vs. Runoff 146
32 Average of Wet Weather Sewage Quality
Variations for all Storms Analyzed 152
33 Comparison-Dry and Wet Weather Sewer Quality 154
34 Wet Weather Quality Correlation Coefficients 157
35 Comparison of Data Records For Evaluation of
Detention Tank Performance 164
36 Summary of Raw Data For Tank Performance
Analysis 165
37 Averaged Tank Influent and Observed Tank Overflow 170
38 Detention Tank Performance - Actual vs. Predicted
Discharge 172
xiv
-------
LIST OF TABLES
(CONT.)
Table No. Page
39 Detention Tank Performance Projections 178
40 Significance o£ Source of Rainfall Data 180
41 Effect of Interceptor Capacity on Combined
Sewer Overflow 188
42 Comparison of Bypass Operation of Detention
Tank vs. Plug Flow Operation 189
43 Comparison of River Discharge and Rainfall Accumulation 197
44 Comparison of River Discharge and Rainfall
Accumulation 200
45 Comparison of Dissolved Oxygen Concentrations
During Survey III and IV 224
46 Milwaukee River - Detention Tank Rainfall
Analysis For Project Area (570 Acres) 233
47 Projected Combined Sewer Overflow Loads
For Sept. 15, 1970 Storm 234
48 Calculated Response of Milwaukee River to
Menomonee River 256
49 Milwaukee River Detention Tank Data 261
xv
-------
ACKNOWLEDGEMENTS
The cooperation and assistance received from the following
persons and their organizations during the course of this study
is greatly appreciated:
1. Technical Advisors
Hydroscience Inc., Water Quality Consultants: Interpre-
tation of data; development and analysis of the mathe-
matical river and detention tank models.
Mr. John L. Mancini
Mr. Eugene D. Driscoll
Marquette University Sanitary Engineering Laboratory
under the direction of Dr. Raymond Kipp: Laboratory
analyses; data collection and special studies.
2. U.S. Environmental Protection Agency
Mr. Clifford Risley, Jr., Director
Office of Research and Development
Region V
Chicago, Illinois
Mr. Darwin Wright
Chief of Control and Treatment Integration
Washington, D. C.
Mr. Albert Printz, Jr., Director
Permit Program Division
Office of Water Enforcement
Mr. Ralph Christensen, Chief
Great Lakes Demonstration Program
Mr . Ron Eng
Project Officer's Representative
3. City of Milwaukee, Wisconsin
Henry W. Maier - Mayor
Alderman William R. Drew - President of the Common Council
-Common Council
Herbert A. Goetsch - Commissioner of Public Works
Edwin J. Laszewski - City Engineer
xvi
-------
ACKNOWLEDGMENTS
(CONT.)
Herbert D. McCullough - City Engineer (Retired)
-Department of Public Works Staff -
Mr. Ted Prawdzik, Engineer in Charge
Engineer - Sewers
Mr. Edmund Hirsch, Engineer in Charge
Sewer Engineering Division (Retired)
-Construction and Operations Personnel -
Mr. Robert Burmeister (Retired)
Mr. Franklin Gerschke
Mr. Robert Hirsch
Mr. Marvin Rutkowski
Mr. Jorgen Knudsen
-City of Milwaukee Health Department -
4. Sewerage Commission of the City of Milwaukee
Ray D. Leary - Chief Engineer and General Manager
Lawrence A. Ernest - Laboratory Director
5. United States Weather Service
6. United States Department of the Interior, Geological Survey,
Water Resources Division
7. Consoer, Townsend and Associates
Frederick N. Van Kirk
Gerald I. Brask
Robert P. Biebel
Edwin E. Pick
xv 11
-------
SECTION I
CONCLUSIONS
PROJECT AREA
Project Area Sewer System
1. Observations made during the project study indicate that all
dry weather flow entering the project area's combined sewer system
is normally intercepted by the Milwaukee Sewerage Commission
intercepting sewer for conveyance to the Commission Wastewater
Treatment Plant.
2. Recorded sewer liquid level readings obtained at the project
detention tank equipment building have indicated that the interceptor
sewer has been flowing full or surcharged during storm periods,
indicating that the intercepting devices in the project area and up-
stream are operating at or above the capacity of the sewer.
3. Since maintenance requirements for the type of combined sewer
intercepting device utilized in the project area appear to be minimal
and since these devices suitably control both dry weather and storm
period flows as indicated in items 1 and 2 above, the devices are an
effective and practical method of intercepting combined sewage.
Dry Weather Sewage Quality
1. Average values of wastewater quality parameters presented in
this report, are considered to represent typical dry weather quality
for the project sewer system.
2. Seasonal dry weather sewage quality variations were examined
for spring, summer, and winter (those seasons when large numbers
of samples were collected). Spring seasonal dry weather sewage
quality show the highest contaminant concentrations of all parameters
investigated except coliforms and volatile solids. Dry weather
sewage contaminent concentrations during the winter are higher than
summer values for all parameters except phosphates and total solids.
-------
3. Diurnal quality fluctuations were explored using winter and
summer seasonal data breakdowns. Total phosphate shows a distinct
pattern in both winter and summer data, with early evening values
ipproximately three times early morning concentrations. Total
coliforms show a clear diurnal variation in winter, but not in summer.
This is also the case with nitrogen and suspended solids data.
4. Chloride concentrations during the winter dry weather periods
suggest an increase from approximately 100 mg/1 during the early
morning hours until midday to about twice this concentration during
afternoon and evening hours. This effect is probably due to concen-
trations of chlorides in snow and ice melt entering the sewer system
rather than increases in concentrations in the sanitary discharges to
the sewers. Based on a dry weather flow rate of 2. 0 mgd and a chloride
concentration increase of 100 mg/1 over background concentrations
between 2:00 P.M. and 10:00 P.M. , a daily flushing of approximately
600 pounds of chlorides per square mile is indicated, which is in excess
of the amount of chlorides present in the sewage during non-winter
months.
5. The data indicates no significant effect on dry weather sewage
quality as a result of an atecedent storm. The sewage quality data
indicates that infiltration is not excessive in the project area sewer
system.
Dry Weather Sewage Flow
1. The flow data obtained indicates the normal average dry weather
flow pattern in the test area has a range between approximately 1. 7
and 2. 3 mgd. At extremes, flows have been observed which range
from 0. 9 to 2. 9 mgd. The 570 acre project area is served almost
excl isively by combined sewers. The area, which is residential
and •ommercial in character with an approximate population of 19, 000,
is located on the north side of the City of Milwaukee. Sewage loadings
on a per capita basis are as follows:
Sewage Flow: 105 gallons per day per capita
BOD: 0. 10 Ib. per day per capita
Suspended Solids: 0. 13 Ib, per day per capita
2. Normally, maximum hourly average flows are approximately
50% greater than minimum hourly average flows during a 24 hour period.
Wide variation between maximum and minimum hourly flows, common
to many municipal systems, has not been observed.
-------
3. Variations in dry weather flow, either due to diurnal or seasonal
fluctuations are not considered to be a significant factor in the analysis
and modeling of the storm overflows in the test area.
Wet Weather Sewage Quality
1. During storm conditions the data collected clearly demonstrate
a pattern of quality variation which reflects the existence of a "first
flush" condition. When all available data is considered on an averaged
basis, all of the sewage quality parameters investigated demonstrate
the pattern whereby concentration of the contaminant decreases with time
as the storm event progresses. Certain individual storm events did not
follow this pattern. However, the frequency of occurrence and magnitude
of the deviation from the "first flush" pattern, are not sufficient to
distort the overall picture provided by analysis of the large number
of samples. In many cases when deviations from the pattern occurred
in an individual storm, they could be attributed to a complex storm
pattern, in which a number of radical changes in rainfall intensity
occurred during the duration of what had been considered a single storm
event. The major emphasis in this study -was to identify the broad
aspect of all storm events rather than single individual events and a
clearly defined "typical" pattern has been shown.
2. During a storm event the initial concentrations of many parameters
in the combined sewage are higher than average dry weather concentra-
tions. BOD, COD, total and volatile suspended solids, chlorides,
nitrates and organic nitrogen exhibit higher initial concentrations in
combined sewage than in dry weather sewage flow. Average maximum
concentration observed for these parameters during a storm overflow
event are approximately 1. 5 to 2. 5 times the dry weather average
values. This suggests the conclusion that the primary source of such
contaminants is materials which have settled and been deposited in the
sewer lines, catch basins, gutters etc. prior to the start of the storm.
3. Several combined sewage quality parameters, including ammonia,
total and ortho phosphate, and fecal coliforms, all exhibit maximum
concentrations during storm events which are less than dry weather
averages. This suggests the quite reasonable conclusion that the
primary source of such contaminants is the sewage flowing in the
lines when the storm occurs which is diluted by the storm flow and that
pollutant discharge of these parameters is not storm generated.
4. The data obtained during the project study period indicates that
no significant correlation exists between combined sewage quality
variations and specific storm characteristics, with the exception of
that indicated in Item 1 above.
-------
Wet Weather Sewage Flow (Runoff Coefficient Cr)
1. On the basis of the data analyzed, storm water runoff in the test
area is characterized by a runoff coefficient Cr of 0. 5. Observed
ranges for C were 0. 3 to 0. 8.
2. The effect of duration of the storm event, total volume of rain
per storm, rainfall intensity, and interval since antecedent storm were
investigated. When data includes many storm event? taken as a whole,
the average coefficient, Cr, more closely approximates 0. 5 indicating
that no variation in Cr with any of the above parameters is justified for
purposes of the study.
-------
COMBINED SEWER OVERFLOW DETENTION TANKS
Project Detention Tank Data
1. Based on the cost data obtained during the project, estimates for
various size detention tanks were developed. These construction cost
estimates (based on estimated December 1973 prices) ranged from $1,500,000
for a two million gallon tank to $3,000,000 for ten million gallon tank
assuming comparable design and local conditions to those at the Humboldt
Avenue Facility. These estimates do not include land or right of way costs.
2. An approximate estimate of annual operating cost for a typical
detention tank of the type employed for this project is $30,000. This
operational cost should not be significantly dependent upon detention
tank size but would be directly influenced by the number of storm events
over a given period of peration.
3. The method and equipment employed at the project detention tank
for handling combined sewage solid materials deposited in the tank (this
basically consisted of mixer agitation equipment utilized to resuspend
solids materials during tank dewatering) operated as anticipated. This
method of handling the solids contained in the combined sewer overflows
was established as a reliable and probably suitable alternate to separate
removal of sludge from combined sewer overflow detention tanks.
4. An analysis of the data obtained during the 12 month period between
November 1, 1971 and October 31, 1972 indicated that the project detention
tank prevented approximately 121,000,000 gallons (out of 181,000,000
gallons) of combined sewage, approximately 100,000 pounds (out of 147,000
pounds) of BOD, and 225,000 pounds (out of 321,000 pounds) of suspended
solids from being discharged to the Milwaukee River, from the 0.9 square
mile project area. Removals of the other combined sewage pollutant con-
stituents studied were also significant with the percentage removal of
all parameters being in the same order of magnitude.
COMBINED SEWER OVERFLOW DETENTION TANK MODEL
1. A combined sewer overflow detention tank system model was developed.
This model makes it possible to evaluate both the total quantity of storm
water and pollutants resulting from storm overflows, and the quantities
which can be intercepted by a detention tank. The model will have a
general value in that evaluations may be made for tanks of various sizes,
serving a range of drainage areas, and a variety of rainfall conditions.
Analysis of model predictions versus observed quality variations in over-
flows leaving the detention tank has provided good verification of the
validity of the model including the sedimentation portion of the model
employed in the detention tank program.
-------
2. The removal efficiency (in terms of the amount of pollutant
material entering the River from the tank versus the amount of material
entering the tank) of the project detention tank and detention tanks in
general is directly related to tank size in relation to the size of the
drainage area served. To a lesser extent, the removal efficiency of a
detention tank is also affected by tank pump out rate, interceptor
sewer capacity, drainage area, rainfall and runoff patterns and tank
sedimentation efficiency.
Discussion of Capabilities of Detention Tanks in Control of Combined
Sewer Overflow
1. Detention tanks have been shown by this project to be effective
in preventing a large proportion of the contaminants found in combined
sewer overflow from entering receiving waters.
2. Removal of BOD and suspended solids (in terms of percent pounds
not overflowed versus pounds in tank influent) can range from approximately
30% to in excess of 80% as tank size is increased from one to six million
gallons per square mile of drainage area.
3. The unit removal efficiency in terms of percent removal per
unit volume decreases as tank size increases. Tank unit efficiency
can range from approximately 30 to 15 percent removal per million
gallons per square mile as tank size is increased from one to six
million gallons per square mile of drainage area.
4. Studies evaluating detention tank removal efficiencies of BOD
and suspended solids indicate that the removal due to volumetric retention
is much more significant than the removals due to sedimentation. Removal
due to sedimentation generally increases total removal efficiency by
approximately 5% over removals due to volumetric retention alone.
Thus, although some increase in overall removal effiency can be attained
by designing a combined sewer overflow detention tank which permits
effective sedimentation to occur, the major consideration in designing
the tank to meet a selected annual average removal efficiency, that is
consistent with water quality standards, is volume. Cost effectiveness
considerations should concentrate on maximum volume which can be
achieved at a given cost and be consistent with the needs to meet water
quality standards.
-------
Detention Tank Application
1. Abatement of discharge of combined sewer overflows for the entire
City of Milwaukee is a highly complex problem. Based upon the results ob-
tained from this project, it is evident that detention tanks can be designed
to eliminate pollutional discharges from combined sewer overflows, which
would cause violation of river water quality standards. Thus detention
tank evaluation should be included in any feasibility study for abatement
of combined sewer overflow pollutional discharges.
Although detention tanks have not necessarily been established as the
only method of eliminating pollution from sewer overflows in every case,
the results of this study indicated that they will be a viable and economical
tool.
2. For purposes of demonstrating the cost impact of the problem, and to
facilitate comparison with other methods of abatement, an approximate cost
estimate has been developed for construction of thirteen detention tanks
to receive flows from all combined sewer overflow points on the Milwaukee
River in the City which accounts for 9 of the 27 square mile combined
sewered area of the City. The approximate preliminary cost estimate for
this construction is $29,500,000, based on estimated December, 1973 cost
indices and including the present tank. This estimate does not include
costs for land, right of way, or sewer construction which would add consider-
ably to the cost. The preliminary estimate is not based on a detailed
feasibility study for each location. It is based only on a general visual
survey in the vicinity of each overflow and should be considered only as an
indication of the general magnitude of cost.
It is anticipated that combined sewage pumping stations will be re-
quired at four of the thirteen locations. Based on very preliminary studies,
the costs of these four pumping stations will add approximately $8,500,000
to the above cost for detention tanks.
The use of detention tanks to receive flows from all of the combined
sewer overflow points along the Milwaukee River also requires the con-
struction of sewers for interconnecting the various outfalls tributary to
each of the thirteen detention tanks. Based on preliminary studies, the costs
for these interconnecting sewers is estimated to be approximately $9,000,000.
While the above stated costs are apparently lower than the reported
cost of some alternate plans for pollution abatement, the sewer and land
costs in certain areas of the City would significantly raise the total pro-
ject cost. Further, the estimate includes costs for detention tanks
-------
which were sized based on the same size per unit area as used for the
Humboldt Avenue facility. Such sizing does reduce the discharge of
pollutants approximately 79% on an annual basis, but does not eliminate
the discharge totally. Therefore an indepth feasibility study should
be made for each outlet or group of outlets to determine the best pollution
abatement alternate solution for that segment of the system.
3. General construction cost information relating cost per square
mile to percent removal of combined sewer overflow based on detention
tanks has been developed. Based on one detention tank serving a 0. 5
square mile area the capital cost per square mile varies from
approximately $2, 800, 000 to $3, 200, 000 as percent removal increases
from 50 to 80. Based on one tank serving a 2. 0 square mile area the
capital cost per square mile varies from $1, 000, 000 to $1, 400, 000 as
percent removal increases from 50 to 80. These estimates do not
include costs of sewers, land, right of way, contingencies, or technical
services. As previously pointed out for the total Milwaukee River
combined overflow solution, each particular overflow situation in
Milwaukee and in other municipalities will have to be evaluated in detail
to determine the most desirable pollution abatement alternate for the
particular installation location. The costs developed herein can be
utilized in making such evaluations.
-------
MILWAUKEE RIVER
Milwaukee River Quality
1. In general, river water quality in the section of the Milwaukee
River studied (between North Avenue and Water Street) is relatively
poor and contains relatively high and variable levels of polluting
material. Typically, the Milwaukee River exhibits the following
characteristic concentrations based on the overall analysis of the
average values observed for the periods compiled during the 5 year
study period.
Temperature (°C) =0-22
pH = 7. 5 - 8. 5
COD (mg/1) =35-60
BOD (mg/1) =5-10
Chlorides (mb/1) =20-50
Total Nitrogen (mg/1) =1-2
Total Phosphorus (mg/1) =1-2
Suspended Solids (mg/1) =20-50
Total Coliform (no/ml) = 200 - 40,000
Significant seasonal variation in most parameter concentrations
is exhibited.
2. The Milwaukee River in the area being considered, is subject
to vide variations in flow which can occur over relatively short
inte rvals.
3. "Many factors affect the water quality in Milwaukee
River. Some of these factors have originated geograph-
ically outside the project study area but add to the pol-
lutional loading of river section studied.
Because of these complexities, any effort to evaluate
river quality responses to storm overflows from the combined
sewer system directly using the routine river quality data
obtained during the study would be highly speculative. In
order to obtain a reliable quantitative assessment of the
effect of combined sewer overflows on water quality in the
lower reach of the Milwaukee River, reliance must be placed
on the mathematical model developed during this program to
characterize water quality responses.
-------
Milwaukee River Model
1. A time variable Milwaukee River water quality model has been
developed and verified by observed data. The model effectively
projects quality responses in the relatively complex Milwaukee River
System. The model has demonstrated its value by providing insight
into river water quality influences and quality responses to these
influences.
River Water Model Output and Verification
1. Significant influences on river water quality in the section of the
Milwaukee River studied (North Avenue to Water Street) have been
identified. These influences are briefly summarized below.
(a) Bottom Deposits - Bottom deposits in the Milwaukee River
degrade water quality as measured by dissolved oxygen under all
weather conditions. The degree to which benthal oxygen demand depresses
dissolved oxygen varies with location and -with River flow. The adverse
impact increases as one proceeds downstream with oxygen depletions
attributable to bottom demand at Water Street ranging from 1. 0 to 2. 5
mg/1. The more serious conditions occur at lower river flows.
(b) Menomonee River - The Menomonee River quality has a
variable influence which at times has a substantial effect. Its adverse
influence, which generally occurs during wet weather, can account for
as much as 1. 0 to 2. 0 mg/1 dissolved oxygen reduction in the vicinity
of St. Paul and Water Streets. The Menomonee has a negligible effect
on quality in the stretch of the Milwaukee River upstream of Cherry
Street.
(c) Activity By Algae and/or Macrophytes - At times algae
and/or macrophyte activity is intense and it can significantly affect
dissolved oxygen levels in the River. Daily variations of 3. 0 to 4. 0
mg/1 to as much as 7. 0 mg/1 of dissolved oxygen have been observed,
which are attributable to photosynthetic activity. Variations in this
order result in D. O. levels approaching zero at times when D. O.
would otherwise be in the range of 2. 0 - 4. 0 mg/1
(d) Flushing Tunnel - The flushing tunnel exerts a favorable
influence on river water quality most of the time. A possible
exception to this, tenatively suggested by the data analysis, is that some
adverse effect may be exhibited, if the tunnel is operated during high
river flow periods. A scouring action can then result in release or
disturbance of bottom deposits accompanied by an increase in oxygen
10
-------
demand. Normally however, tunnel operation can increase D. O.
concentrations by approximately 3. 5 mg/1 at Cherry Street, 2. 0 mg/1
at St. Paul Avenue and 1. 3 mg/1 at Water Street, as well as improve
the aesthetic condition of the River.
(e) Combined Sewer Overflows - Under most storm conditions,
combined sewer overflows entering the River between the North Avenue
Dam and the Menomonee River, exhibit a relatively minor direct effect
on quality as measured by dissolved oxygen in that section of the River.
However, the overflows do contribute coliform organisms, BOD and
other pollutants, and these can be expected to have adverse affects on
areas after the river -water leaves the area of study. Solving the water
quality problem in the section of the River investigated in this study
will require a program which considers all of the water quality
influences on the river and not simply the reduction of combined sewer
overflows in the immediate area. However, present regulation and
national goals regarding pollution abatement lend increased importance
to abatement of combined sewer overflows.
(f) Upstream Water Quality Conditions - As previously mentioned
upstream quality conditions account for an estimated 25 - 60 percent
of the oxygen deficits observed in the river segments studied.
2 The studies conducted indicate that elimination of bottom deposits
can have a significant impact on River dissolved oxygen levels. In
terms of required time and probably cost, removal of bottom deposits
is the most economical single measure that would improve the River
water quality as measured by dissolved oxygen. However, due to the
various sources of siltation in the River such as combined sewer
overflows, urban construction and farm erosion, solids will continue
to accumulate. This would make periodic dredging necessary.
Elimination of a portion of these sources of siltation would, of course,
reduce the frequency of solids removal operations. The problem of
dredged sediment will also have to be considered and could be costly.
11
-------
SECTION II
RECOMMENDATIONS
1. The effect of a single detention tank on the Milwaukee River
quality is relatively small due to numerous other influences on the
river. However, the operation of the Humboldt Avenue detention tank
should be continued since, during a year of normal rainfall, it is
estimated that the tank will prevent approximately 93,000,000 gallons
(out of 124,000,000 gallons) of combined sewer overflow, approximately
79,000 pounds (out of 100,000 pounds) of BOD, and 176,000 pounds (out
of 219,000 pounds) of suspended solids from entering the Milwaukee
River. This indicates an operating efficiency of 79% for BOD and
80% for suspended solids.
2. It is recommended that monitoring of the operation of the
detention tank be continued to provide additional data which could
aid the tank operation personnel in improving the tank efficiency in
reducing pollutant discharges to the River as well as being of interest
to others in developing plans for similar facilities. Particular
emphasis should be given to coliform, BOD and suspended solids removal
efficiencies. Other important water quality parameters, such as
ammonia and phosphorus, could be studied.
3. On projects which included a significant amount of automation
and instrumentation systems which are critical to data collection,
special maintenance provisions should be made. The following items
should be given special consideration:
(a) Provisions should be made to include well-trained service
personnel in the project staff, virtually on a full-time basis if data
collection is to be a major function of the project.
(b) Arrangements should be made to insure adequate local
slocking of spare parts.
(c) The use of backup systems for the most critical system
components should be included.
4. Due to the high degree of maintenance required to keep dis-
solved oxygen probes operating on a continuous basis in certain river
water service installations, a manual D.O. collection system should be
initiated unless it is imperative that continuous data is required. As
progress is made in the instrumentation development field, continuously
operated probes may become more practical.
12
-------
5. Consideration should be given to providing a detention tank
influent measuring and sampling device at the Humboldt Avenue tank
location. This will eliminate the dependence on the Humboldt and
Commerce monitoring station for influent measurement and sampling.
6. Planning for improvement of the Milwaukee River quality in the
area studied (between North Avenue and Water Street) should include
the following major factors which influence the river water quality:
(a) Milwaukee River bottom sludge deposits.
(b) Menomonee River water quality.
(c) Algae activity and growth in the river.
(d) River flushing tunnel operation.
(e) Combined sewer overflow discharges to the river.
(f) Milwaukee River water quality upstream of the area
studied.
7. Since the flushing tunnel operation has been demonstrated to be
a favorable influence on river water quality, its operation should be
continued. The present schedule of operation by the Sewerage Commission
appears to be a desirable operational plan. Due to possible detrimental
scouring action, consideration and further study should be given to cutting
off the tunnel operation during high river flow periods even if D. O.
levels are low. Such consideration should be made in conjunction with
decisions regarding sludge deposit dredging.
8. In continuation of comprehensive planning for elimination of the
causes of poor water quality in the Milwaukee River, the river model
developed under this project should be expanded to include water quality
influence outside of the river segment which was studied. The river
model developed demonstrated its value in providing insight into
conditions which influence water quality, some of which were not
readily apparent. An expansion of the model could provide data upon
which construction priorities could be established. That is, those
pollution abatement measures to be taken could be phased in a cost
effective manner such that those with the greatest river water quality
improvement significance would be taken earliest in the pollution
abatement program.
13
-------
9. Planning to evaluate the most desirable method of abating pollution
due to the discharge of combined sewer overflow should include a study
of the construction of detention tanks similar to the Humboldt Avenue
facility at many locations. This report has demonstrated that the cost
effectiveness of detention tanks may be attractive in many instances.
This should be compared to alternate combined sewer overflow pollution
abatement methods. Study into treatment plant and interceptor capacity
should also be included in any comprehensive plan.
10. When evaluating combined sewer overflow pollution abatement
plans for a large area, detailed study involving receiving stream modeling
can often be very valuable in establishing priorities. The plans, as an
early part of any program, should include staged construction or other
remedial measures which would have the most significant influence on
improvement of receiving water quality. The plan could then include
maximum exposure of connected area to combined sewer overflow
storage facilities as early as possible. The area connected to these
facilities on a volume per area basis should be maximized. This will
normally provide the greatest total removal of pollutants.
The planning could be based on higher acreage connection per
unit of volume as part of the initial storage construction program.
This would result in maximizing the effectiveness of storage volumes.
As the construction program proceeds, the acreage connection per unit
volume could be decreased as total storage volume approaches optimum.
11. Since the dredging of bottom deposits was discontinued north of
E. Buffalo Street solids deposits have been accumulating. These
bottom deposits exert an oxygen demand on the River and are one of
the major causes of poor river water quality in the Milwaukee River.
The visual manefestations of the deposits are gasification and floating
sludge solids at times. The depletion of oxygen by these deposits also
creates unfavorable environment for fish and other desirable forms of
aquatic life. Therefore to improve the water quality of the Milwaukee
River, the bottom deposits in the River should be removed because of
the benthal oxygen demand with its resultant reduction in the dissolved
oxygen content in the River water. Investigations into potential dredged
solids disposal problems should be an early step in planning for dredging.
14
-------
SECTION III
INTRODUCTION
Combined sewers, that is, sewers which transport both sanitary
wastes and storm water are prevalent throughout the county, particularly
in the older urban areas. As designed and constructed, these sewers
originally represented an economical solution to the problem of handling
both sanitary and storm flows. All sanitary sewage and storm water
was initially discharged, untreated, directly to nearby receiving waters.
As the population served by these combined sewer systems grew,
the result was a corresponding increase in sanitary sewage discharged
to the receiving waters. As wastewater treatment plants came into
'general use, an attempt was made to intercept all dry weather flows for
conveyance to the plants for treatment. However, during periods of
rainfall, storm water flows may be in excess of the interceptor sewer
capacities. Under these conditions, an adequate outlet must be provided
for the combined sewers to prevent flooding of basements and streets.
To provide the necessary outlet, the combined sewers are permitted to
discharge untreated dilute sewage directly to the receiving waters.
According to a 1964 U. S. Public Health Service Publication,1 more
than 1900 communities in the United States, inhabited by some 59 million
people, are served by combined or by combined and separate sewerage
systems. Studies done in 1967 by the American Public Works Associat-
ion indicated that approximately 29 percent of the nation's total sewered
population is served by combined sewers.
In the 1964 Public Health Service report,! it is estimated that the
annual overflow from these systems contains 3 to 5 percent of the total
raw sewage generated within the areas served, and during storms as
much as 95 percent of the untreated sewage overflows to receiving waters.
The older areas of the City of Milwaukee, like many older urban
areas, are served almost exclusively by combined sewers. In 1966, the
Department of Public Works of the City of Milwaukee at the direction of
the Common Council applied to the Federal Water Pollution Control
Administration for a combined sewer demonstration grant under the
Water Quality Act of 1965. In preparing the application, City Officials
15
-------
envisioned the implementation of a practical, economical approach to the
abatement of pollution caused by combined sewer overflow.
Early in the project planning stage it was determined that this
project would be limited to the study of a detention tank to which
combined sewer overflows could be diverted and detained until low flow
periods when the tank contents could be pumped into intercepting sewers
for conveyance to treatment facilities.
The study of the detention tank concept was undertaken with the
intent of satisfying the following objectives:
1. Characterize stormwater overflows from a
combined sewer system in an urban test area
of Milwaukee, Wisconsin.
2. Characterize the performance of a combined
sewer overflow detention tank in reducing the
pollutional load to the Milwaukee River caused
by rainfall runoff in the test area.
3. Evaluate cost and effectiveness of detention
tanks for abatement of storm overflow
pollutional discharges for urban areas
larger (or smaller) than the test area
studied.
4. Project the impact of combined sewer overflow
detention tanks on the quality of water in the
Milwauke e Rive r.
The various phases of the project and the planned sequence of
these phases is illustrated in Figure 1.
To meet these objectives a study area was defined and a thorough
investigation of the study area combined sewer system and Milwaukee
River, upstream and downstream from the study area, was conducted.
This investigation required the design and construction of eleven sewer
and river monitoring stations. Also designed and constructed was a
3. 9 million gallon capacity detention tank. The effectiveness of the
detention tank in abating storm overflow pollutional discharges was
analyzed in accordance with the project objectives.
16
-------
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVENUE POLLUTION ABATEMENT DEMONSTRATION PROJECT
EXECUTED FACILITIES
DEMONSTRATION GRANT
CONTRACT
FIGURE I
PROJECT CRITJCAL PATH FLOW CHART
-------
The objectives -were met during the course of the study and all
findings are discussed in detail in this Report.
It is planned that the results of this Report can be utilized as part
of an extensive study of the City's entire combined sewer problem.
The results of such a study coupled with the existing well operated and
relatively efficient wastewater treatment facilities will provide a major
step in improvement of water quality in the Milwaukee metropolitan
area.
18
-------
SECTION IV
PROJECT DESCRIPTION
PROJECT STUDY AREA
The Humboldt Avenue Pollution Abatement Demonstration
Project study area, consists of 570 acres containing some 21 miles
of combined service sewers and representing 1/27 of the entire
combined sewered area of the City of Milwaukee. The study area is
an urban section of the City's northeast side bounded generally by
the Milwaukee River on the south and east sides; East Townsend Street
on the north side; and North Third Street on the west. The area is
residential and commercial in character and is served exclusively by
combined sewers with a few separate storm sewers intercepted within
the project area. The location of the project study area is shown in
Figure No. 2.
PROJECT AREA COMBINED SEWERS
The combined sewer system of the project area is relieved,
during periods of precipitation, at six locations by two combined relief
sewers which traverse the area. These sewers also furnish relief
to other local service systems adjacent to the project area. The relief
sewers were constructed at elevations deep enough to permit gravity
discharge from the service sewers. The overflow chambers utilized
do not contain automatic regulating devices or controls. The combined
relief sewers outlet to the Milwaukee River at East Locust Street and
at East Auer Avenue.
Dry weather flows and a portion of the combined flow generated
during periods of precipitation are directed to Milwaukee Sewerage
Commission intercepting sewers for conveyance to the Commission's
Jones Island Wastewater Treatment Plant. These Milwaukee Sewerage
Commission intercepting sewers are referred to as M. I. S. sewers.
The terms intercepting sewer and M. I. S. sewer are used interchange-
ably in this report. Flow to these interceptor sewers (M. I. S. sewers)
is regulated by intercepting chambers (M. I. S. chambers) which divert
all of the dry weather flow and a portion of the combined flow into the
intercepting sewer. Flows in excess of intercepting sewer capacity
enter the combined overflow downstream of the intercepting structure
and discharge into the Milwaukee River.
19
-------
PROJECT AREA,
LOCATION MAP
JONES ISLAND WASTEWATER
TREATMENT PLANT
\
20
-------
The main feature of these M. I. S. chambers is the flow proportion-
ing device which controls the amount of combined flow to the intercepting
sewer. The device consists of an eccentrically pivoted flap gate
mounted in a precast concrete housing set in the dividing wall between
the combined sewer and the interceptor sewer. The gate is designed
to rest at a maximum open position during periods of dry weather flow
to permit passage of all sanitary wastes into the interceptor sewer.
Experience to date is that all dry weather flow has been intercepted.
As the level of flow in the combined sewer increases, due to rainfall
runoff or snow melt, the increased pressure on the larger top portion
of the gate causes the gate to close to a minimum open position. The
gate is so designed that the increased head on this orifice and correspond-
ing higher velocity through it result in the passage of approximately the
same quantity of combined flow as the maximum permissible dry weather
flow. To date recorded flow readings obtained at the project detention
tank equipment building have indicated that the interceptor sewer has
been flowing full or surcharged during storm periods indicating that
the intercepting M.I. S. devices are operating at or above the capacity
of the sewer.
Since the maintenance requirements of these M. I. S. devices
is relatively minor and since they operate effectively as noted above,
they appear to be an effective and economical method of intercepting
flow in combined sewers and directing it to the City's waste-water
treatment facilities.
There are two such intercepting structures (M. I. S. chambers)
within the study area, one located on the 60 inch diameter combined
sewer at North Humboldt Avenue and East Wright Street and the other
on the 60 inch combined sewer at North Humboldt Avenue and North
Commerce Street. The primary outlet for the study area is this 60 inch
combined sewer. The overflow for this combined sewer was a 72 inch
diameter overflow sewer located at North Humboldt Avenue and the
Milwaukee River. It was estimated that an average of approximately
50 to 60 overflow incidents per year occurred at this point.
Further details on the various overflow and intercepting points
in the project area sewer system are given in Section V of this Report.
In Table 1, the project study area combined sewer system can be
compared with Milwaukee's total combined sewer system.
21
-------
Table 1. MILWAUKEE-COMBINED SEWERS
Total Project
Municipal Area
System System
Combined Sewer Length in Miles 550 21
Tributary Area, Square Miles 27 0. 9
Overflow and Intercepting Devices,
Number:
Overflow 109 6
Intercepting 134 2
22
-------
MILWAUKEE RIVER
The Milwaukee River is a shallow, meandering stream flowing
in a generally southerly direction from its source in southeastern
Fond du Lac County to Lake Michigan at Milwaukee. The approximate
690 square mile Milwaukee River watershed primarily consists of
agricultural, industrial, commercial, and residentially developed
areas. For most of its length the river flows through soil areas
carrying along large quantities of silt which results in a brownish color,
characteristic of siltaceous streams.
Within the Milwaukee River -watershed, the City of Milwaukee
and the Village of Shorewood are the only communities served by
combined sewers. The Milwaukee River Technical Study Committee
noted, in their 1968 report to the Mayor of the City of Milwaukee,
that sources of industrial and organic pollution have, with few exceptions,
been controlled north of the Milwaukee County Line. However, between
Keefe Avenue and the mouth of the river there are 62 combined sewer
outfalls, 52 of which discharge to the river between the North Avenue
Dam and Lake Michigan. The locations and sizes of the outfalls are
listed on the following Table in order of their appearance on the River
beginning at Capitol Drive and proceeding south to the mouth of the
River.
To monitor river flows, a discharge gaging station is maintained
by the U. S. Department of Interior, Geological Survey. The station
is located on the river at Estabrook Park, approximately 6 miles
upstream from the river's mouth. Discharge records show that the
river undergoes wide fluctuations in flow rate. The highest flow of
record occurred on March 20, 1918, and August 6, 1924, when flows
of 15, 100 cfs were recorded and the lowest flow of record occurred
on September 8, 1943, when zero flow was recorded. Minimum flows
during any year may range from 0 to 200 cfs and maximum flows from
2, 000 to 15, 100 cfs (theoretical maximum flow has been estimated at
35, 000 cfs).
RIVER FLUSHING TUNNEL
Wastewater treatment facilities were provided in Milwaukee
for the first time in 1925 when the Jones Island plant was placed in
operation. Prior to that time all sanitary wastes were discharged
directly to Lake Michigan and the streams flowing through the City.
By the late 1800's this direct discharge had transformed that section
of the Milwaukee River below the North Avenue Dam into an open sewer.
Unsightly surface pollution was evident throughout the year and during
the summer months -when temperatures are high and river flows are
low the odors emanating from the river became unbearable.
23
-------
Table 2. COMBINED SEWER OUTFALLS INTO
MILWAUKEE RIVER
Location
E. Capitol Drive
E. Keefe Avenue
E. Edgewood Avenue
E. Auer Avenue
E. Auer Avenue
E. Hampshire Street
E. Burleigh Street
E. Chambers Street
E. Locust Street
E. Park Place
E. Bradford Avenue
E. Boylston Street
E. Tunnel Place
N. Pulaski -"Street
N. Humboldt Avenue
N. Marshall Street
N. Holton Street
E. Grady Street
N. of E. Walnut Street
E. Walnut Street
E. Pleasant Street
N. of W. Cherry Street
W. Cherry Street
E. Lyon Street
E. Ogden Avenue
W. McKinley Avenue
W. Juneau Avenue
E. Juneau Avenue
N. of W. Juneau Avenue
W. Highland Avenue
Size
Location
Size
72"
54"
72"
84"
36"
24"
Dbl.
9'-6"x
4'-3"
21"
78"
60"
72"
72"
12"
72"
72"
24"
7'x4'
30"
96"
42"
7'x3'
5'x4'
90"
36' &
18"
6'x3'
60"
42"
42"
84"
9'-3"
x4'-6"
E. Highland Avenue
W. State Street
E. State Street
W. Kilbourn Avenue
W. Kilbourn Avenue
E. Kilbourn Avenue
N. of W. Wells Street
W. Wells Street
E. Wells Street
N. of W. Wisconsin Ave.
W. Wisconsin Avenue
E. Wisconsin Avenue
W. Michigan Avenue
E. Michigan Avenue
N. of W. Clybourn St.
W. Clybourn Street
E. Clybourn Street
N. of W. St. Paul Ave.
W. St. Paul Avenue
E. St. Paul Avenue
E. Buffalo Street
E. Chicago Street
S. First Street
S. Water Street
E. Pittsburgh Avenue
N. Broadway
S. of E. Oregon Street
E. Florida Street
E. Polk Street
E. Harbor Place
E. Bruce Street
36"
46"
60"
36"
60"
54"
30"
45"
48"
18"x30"
24"
30"
54"
42"
24"x26"
30"
Dbl. 48"
30"
6'x3'
8' -6"x4'-0'
42"
6'x4'
24"
24"
24"
30"
30"
60"
54"
30"
24"
24
-------
In 1888 a tunnel was constructed between Lake Michigan and
the Milwaukee River through which large volumes of lake water
could be pumped to the river, discharging just below the dam. The
oxygen rich lake water served to dilute the polluted river water and by
increasing the dissolved oxygen content in the river the obnoxious
odors were greatly reduced. It is believed that the flushing tunnel
operation plus the Corps of Engineers dredging at the upper segment
of the Milwaukee River estuary, which ended in the 1950's main-
tained the river water quality in a relatively acceptable quality from
some aspects.
Until 1964, the flushing tunnel was operated by the City of
Milwaukee with the following minimum schedule:
Hours
Month Per Week
May 20
June 30
July 40
August 56
September 40
October 40
November 20
This schedule was flexible and the hours of operation were readily
increased whenever lower than usual water levels or extended periods
of hot weather were encountered.
In 1964 the flushing tunnel operation was conveyed to the
Sewerage Commission of the City of Milwaukee, which has continued
the practice of river flushing to reduce odors. The Commission
schedules operation of tunnel for 80 hours per week from May through
October, but this schedule is readily altered in response to changing
dissolved oxygen levels in the river. When the oxygen level is above
approximately 5 rng/1 the hours of tunnel operation are reduced but
when the level falls to 3 mg/1 tunnel operation is increased. If the
dissolved oxygen level does not stabilize, or if it continues to drop,
the tunnel is operated on a 24 hour per day basis, and up to seven
days per week.
To determine the actual effect that flushing has on the water
quality of the river below the North Avenue Dam, it was necessary
to determine the rate of discharge of lake water through the tunnel.
The method used for this determination and the results of the study
are discussed in detail in Section VII of this Report.
25
-------
SECTION V
PROJECT MONITORING SYSTEMS
To characterize the stormwater overflow from the project area
combined sewer system, and to evaluate the effectiveness of the
detention tank in reducing pollutional loads to the Milwaukee River,
it was necessary to determine the quantity and quality of the combined
sewage flows generated within the project study area. A full evaluation
of tank effectiveness, on a per acre basis, would have necessitated
the routing of all runoff in excess of intercepting sewer capacity from
the project drainage area to the detention tank. This would have
required the placement of bulkheads at all relief points within the
system. However, the placement of bulkheads would have resulted
in the flooding of basements and possibly some streets.
The alternative to the placement of bulkheads was to determine
the quantity and quality of flow at each overflow point from the project
area sewer system to the relief sewers. To do this, six monitoring
stations were designed for construction within the study area. Two
additional monitoring stations were constructed at the intercepting
sewer structure sites to obtain dry and wet weather flow data. An
additional three monitoring stations were constructed along the
Milwaukee River to develop a profile in the river water quality.
By projecting the data obtained from these monitoring stations
and the detention tank, the impact of the combined sewered area on
the Milwaukee River water quality can be reproduced to provide an
over-all picture or profile of the existing water quality of the river.
Estimates of improvement which might be made due to combined sewer
overflow facilities can also be made. One of the river monitors was
located at the North Avenue Dam, upstream of the project area
combined sewer outfall and detention tank; one was located downstream
at the Cherry Street bridge, after a considerable additional combined
sewer area is encountered; and the third station was located at the
St. Paul Avenue bridge, near the outlet of the river into Lake Michigan,
downstream of most of the area served by combined sewers.
The locations of the sewer and river monitor stations are shown
on Figures 3 and 4. Station identification numbers and locations are
as shown in Table 3.
26
-------
E. TOWNSENO ST.
1__J l_—I U l_l I—J.C.
i ni—i
E. BURLEI6H ST.
u_J u uiu t_i i I
PROJECT
BOUNDARY
N
~ir
COMMERCE AND
HUMBOLDT
FIGURE 3
LOCATION OF COMBINED SEWER MONITORING STATIONS
NOTE: FOR FURTHER DETAIL ON THE SEWER
SYSTEM AT EACH MONITORING STATION
LOCATION SEE FIGURES 5,7,8,9, JO SI I
27
-------
OO
LJLJ
nnn
COMMERCE AND
HUMBOLDT
MONITORING STA
NORTH AVE. 0AM RIVER
MONITORING STATION
[DETENTION TANK SITE
[FLUSHING TUNNEL OUTLET
EAST VINE ST.
EAST BRADY S
FLUSHING
TUNNEL
INTAKE
ST.PAUL AVE. BRIDGE
RIVER MONITORING STA.
NOT SHOWN
CHERRY STREET BRIDGE
RIVER MONITORING STATION
FIGURE 4-LOCATIONS OF DETENTION TANK AND
MONITORING STATIONS
-------
Table 3. MONITORING STATION IDENTIFICATION
Station Number Station Location
Sewer Stations
21 N. Humboldt Avenue and E.
Wright Street
25 W. Ring Street and N. Richards
Street
29 E. Auer Avenue and N. Fratney
Street
33 E. Auer Avenue and N. Humboldt
Avenue
37 N. Humboldt Avenue and E.
Locust Street
41 E. Auer Avenue and N. Booth
Street
45 E. Locust Street and N.
Richards Street
49 N. Humboldt Avenue and N.
Commerce Street
River Stations
52 St. Paul Avenue Bridge
58 Cherry Street Bridge
*64 North Avertue Dam - Downstream
*66 North Avenue Dam - Upstream
The North Avenue Dam monitoring station was designed to allow
monitoring of water quality both upstream and downstream of the
dam.
29
-------
A schematic diagram of the project's combined sewer flow-
metering system is included in Figure 5.
Rainfall gauging stations were also included as part of the project
monitoring systems. A location map indicating the approximate
location of the various pertinent project monitoring components outside
of the project area is illustrated in Figure 6.
COMBINED SEWER MONITORING STATION FUNCTION
The eight combined sewer monitoring stations were designed to
perform a common function; that is, to indicate, record, totalize,
and sample combined sewage flow. The specific components for each
station varied, but in general included the following: pressure bell
type level sensing devices for activating the measuring and sampling
systems; liquid level and flow measuring and recording equipment;
samplers and refrigerated storage compartments; and miscellaneous
piping, valves and accessories required for satisfactory operation.
A description of the function and a list of the equipment for each monitoring
station follows.
Station 21 - Humboldt and Wright
The function of this station was to record liquid levels in an
overflow structure (M.I. S. chamber) by means of a compressed nitrogen
gas bubbler; measure, indicate and record flow through a gate orifice
type intercepting device (M. I. S. device) with indicating, recording and
totalizing equipment; obtain samples of combined sewage in proportion
to flow by an ejector type sampler with frequency control and appurten-
ances; and store samples in refrigerated containers.
Figure 7 shows a typical equipment arrangement used to determine
quantity and quality of combined sewage flow through a gate orifice type
intercepting device. A location map indicating the general location of
this station and the main sewers in the area is indicated in Figure 8.
Station 25 - Ring and Richards
Station 29 - Auer and Fratney
Station 33 - Auer and Humboldt
Station 37 - Locust and Humboldt
The function of each of these four stations was to measure,
indicate and record combined sewage flow through a flume or flow
nozzle; record the liquid level at the crest of the primary measuring
device; record the liquid level at a point upstream from the primary
measuring device; obtain proportional samples of combined sewage
overflow; and store samples in a refrigerated container.
30
-------
FIGURE 5 - SCHEMATIC OF INPUTS TO COMBINED
SEWER FLOW METERING SYSTEM
OT
cr
u
ui
CO
£0
33
cr LU
UJ CC
x <
U. ||0 CC
IU
ct
O
cc
DRAINAGE
AREA—-—
HUMBOLDT & WRIGHT-
STATION 21
FLOW RECORDER
LEVEL RECORDER
HUMBOLDT & COMMERCE-
STATION 49
STORM
DETENTION
TANK—-
UPSTREAM
OVERFLOW
STATIONS
TO M.I.S.
CC
UJ
E
f
o
TO M I S
-OVERFLOWS
TO RIVER
31
-------
FIGURE 6-MONITORING COMPONENT
LOCATION MAP
PIERCE STREET SCHOOL IAIN GAUGE LOCATION
HOLTON STREET RAIN OAUGE LOCATION
PROJECT DETECTION TANI
58 CHERRY STREET MONITOR INC STATION
66 NORTH AVE. DAM
NOHITORIHC STATION
52 ST. PAUL AVE. BRIDGE HONITODING STATION
•59 MILWAUKEE
HARBOR
AREA
JONES ISLAND WASTE-
K»TER TREATMENT PLANT
Z
2
x
o
bf
NUMBERS INDICATE SAMPLING STATION
LOCATIONS-SEE SECTION Htt FOR DETAILS
NOTE- FOR LOCATIONS OF COMBINED SFWER
MONITORING STATIONS IN PRC^CT AREA
ott rIGURE 3
32
-------
M.I. ^(INTERCEPTING)
SEWER
ACCESS PIPI
FLOW DEVICE
60"COMBINED
SEWER
*••"
PRESSURE BELL^
SAM
PIPE
SAMPLER
CAGE
BUBBLER PIPE—'
PLAN
^-PRESSURE BELL
SAMPLER PIPE-CAGE
PIPING
TO STATION
FLOW
DEVICE
SEWER
SECTION B
PRESSURE BELL
60'COMBINED
SEWER
SAMPLER PIPE-CAGE
SECTION A
FIGURE 7
MONITORING STATION WITH GATE ORIFICE FLOW DEVICE-MI. S. CHAMBER
33
-------
60" COMBINED SEWE!
15" COMBINED SEWEF
42" COMBINED SEWER-?
= — = :£
Lo
.p-
24" STORM SEWER
6O" RELIEF SEWER-i
a
w
12" COMBINED SEWER
36 RELIEF SEWER
^•MONITORING STATION
FLUME CHAMBER
SHAFT
STATION 21
is" COMBINED SEWER
24" OVERFLOW SEWER
"7"~
IOZ2LE CHAMBER
18" COMBINED SEWER
to
I
OVERFLOW STRUCTURE
E LOCUST ST.
18 COMBINED
SEWER
STATION 45
<£-24" MIS SEWER
IS^COMBJNED SEWER
^COMBINED SEWER
!MIS CHAMBER
MONITORING STATION
60" COMBINED SEWER
15" x22" SEWER—*-*J| [
11*1
MONITORING STATION
96" RELIEF SEWER-...
TO DETENTION TANK \>
~ ~f-
^COMMERCEJijr
36" MIS
MIS
CHAMBER
12" SEWER
84" RELIEF
SEWER
r=tt
.xkr==*==
>")<£
72" OVERFLOW SEWE
TO RIVER
24 SEWER
STATION 49
FIGURE 8-LOCATION MAR-STATIONS 21,45,849
-------
All equipment in these stations, other than the primary flow
measuring device, was similar to that used in the Wright and Humboldt
Station.
Figures 9 and 10 show a typical equipment arrangement used to
determine quantity and quality of combined sewage flow through a flume
or nozzle. A location map indicating the general location of these
stations and the main sewers in the area are shown in Figures 9, 10,
and 11.
Station 45 - Locust and Richards
The function of this station was to measure, indicate, totalize
and record flows through a 36 inch flume, as well as through a 24 inch
flow nozzle; record the liquid level at a point upstream from each of
the primary measuring devices; obtain combined sewage samples in
proportion to the flow through one of the primary devices; and store
samples in a refrigerated container. Two points of measurement were
required because two combined sewers are relieved at this location.
It was assumed that the overflow characteristics would be similar, thus
sampling facilities were provided for only one of the sewers.
This station included equipment as described for the stations
in preceding paragraphs, except that two sets of flow indicating, record-
ing and totalizing equipment were required as well as four liquid level
recording devices.
Station 41 - Auer and Booth
The function of this station was to measure, indicate, record
and totalize flow through each of two flumes; record the liquid level
at the crest of each of the flumes; record the liquid level at a point
upstream from one of the flumes; obtain combined sewage samples
in proportion to flow through one of the flumes; and store samples in
a refrigerated container.
The station included equipment as described in the paragraph
above for the monitoring station at Locust and Richards, except that
only three liquid level recording devices were required.
A location map indicating the general location of this station
and the main sewers in the area is shown in Figure 11.
35
-------
RELIEF
CHAMBER
u>
OVERFLOW
WALL
UPSTREAM
BUBBLER
EXISTING
RELIEF
SEWER
36"COMBINED
SEWER
SEWER
DROP SHAFT
FLUME
CHAMBER
$&— MONITORING
"•y STATION
^ EXISTING
RELIEF
CHAMBER
FLUME
BUBBLER
LOCATION MAP
PROPOSED ACCESS
PIPING
• / rTO STATION
-SAMPLER
i ACCESS
t
PRESSURE—
BELL
)"RELIEF
SEWER
^SAMPLER ACCESS
PIPE-CAGE
DROP SHAFT
30"MEASURING
FLUME
FLUME CHAMBER
30 RELIEF SEWER
PLAN
30' RELIEF
SEWER
FLUME
CHAMBER
SECTION A-A
SEWER
PLAN
FLUME
BUBBLER'S
^•RELIEF
CHAMBER
•OVERFLOW
WALL
-30 MEASURING FLUME
SECTION B-B
FIGURE 9-MONITORING STATION WITH FLUME FLOW DEVICE
STATION 25
-------
ACCESS
PIPING
PRESSURE
BELL
DROP SHAFT
NOZZLE
CHAMBER
N
-RELIEF
SEWER
C4
(OVERFLOW
-~-rLL7
72"RELIEF
SEWER
12" COMBINED
SEWER
DROP SHAFT
rNOZZLE CHAMBER
/-RELIEF
/ CHAMBER
V30"COMBINED SEWER
PLAN
RELIEF
STRUCTURE
PROPOSED MONITORING
STATION
-I8"COMBINED SEWER
LOCATION PLAN
?~PROPOSED ACCESS
" PIPING
RELIEF CHAMBER
OVERFLOW WALL
NOZZLE CHAMBER
PRESSURE
BELL
-20"MEASURING NOZZLE
8 CONNECTION SEWER
-SAMPLER ACCESS
PIPE-GAGE
SECTION
FIGURE 10-MONITORING STATION WITH NOZZLE FLOW DEVICE
STATION 29
37
-------
94" COMBNED SEWER-
36" RELIEF SEWER-, —
MONITORING STATION-^
--72" RELEF SEWER
V._ NOZZLE CHAMBER-,
1 /r
~^,
\
4.
i
~- —
N
4" RELEF SEWER-p»
00
21 OVERFLOW SEWER
OVERFLOW STRUCTURE
36 COMBNED SEWER -
RAILROAD TRACKS
54 COMBINED SEWER
MONITORING
STATION
FLUME CHAMBER
MONITORING STATION
72" RELIEF
OVERFLOW
SEWER •? \ \
94 OVERFLOW SEWER
-II
18 COMBINED SEWER
DROP SHAFT—
60" RELIEF SEWER-
E. AUER AVE. /\\ f C 30" COMBINED SEWER
^
18" COMBINED SEWER
~i_
/ II
H T^
r ^OVERFLOW STRUCTURE
•-v
J'
-FLUME CHAMBER
42" COMBINED SEWER
" RELEF SEWER-^
E. LOCUST ST.
STA
TION
41
TV7
18" COMBINED^V*
SEWER
1
ui
t-
o
o
z
u
tl
1
1
^
r
r
54" COMBINED
- SEWER
FIGURE II- LOCATION MAP- STATIONS 33,37841
-------
Station 49 - Humboldt and Commerce
The function of this station was to measure, indicate, record
and totalize flow through a flume; record the liquid level at the crest
of the flume; record the liquid level at a point upstream from the
flume; record the liquid level at an existing gate orifice; indicate, record
and totalize the flow through the gate orifice, obtain combined sewage
samples in proportion to flow through the flume; store samples in a
refrigerated container; obtain additional individual samples on a time
interval basis using a separate sampler; and store such samples in
refrigerated bottles.
This station was located immediately upstream from the
detention tank and measured and collected samples of all combined
sewage flows discharging to the tank. To determine the quality of
the combined sewage, at various intervals of storm flows, an additional
sampler was provided at this station which collected individual grab
samples at predetermined time intervals.
All equipment in this station was similar to that used in the
Humboldt and Wright Station, except for the addition of a vacuum type
sampler for obtaining the grab samples and a flume measuring device.
A location map indicating the general location of this station
and the main sewers in the area is shown in Figure 8.
COMBINED SEWER MONITORING STATION COMPONENT EQUIPMENT
Primary Measuring Devices - General
The type of primary measuring device used at each station
depended upon the length and slope of the connecting sewer between
the point of overflow and the relief sewer. Where a sufficient straight
run of sewer existed, a flume was used; where an insufficient straight
run of sewer existed, an open flow nozzle was used. At two monitoring
locations, Humboldt and Wright and Humboldt and Commerce, dry weather
flow and a portion of the combined sewage flow is diverted through orifice
type M. I. S. intercepting devices to the Sewerage Commission interceptor
sewer. The flow of sewage to the interceptor is regulated by these
devices and is maintained at a fairly constant rate by an eccentrically
pivoted orifice gate, the opening of which is controlled by the depth of
flow in the intercepting structure. A head measurement upstream from
the orifice was used to determine and record the approximate flow through
the orifice to the interceptor by means of a cam-type recorder instrument.
39
-------
Under open flow conditions, it was only necessary to measure
the head on either the flume or the nozzle. However, at high flows,
when the connecting sewer was surcharged, it was also necessary to
measure the upstream head. This information, compared with the
downstream reading, provided approximate head loss data for use in
calculating the quantity of flow in the sewer by use of hydraulic formulae.
During these high flow conditions, flows are calculated manually, rather
than automatically provided through the instrumentation system.
Figure 12 is a schematic diagram showing the interconnection of
the various components of a combined sewer monitoring station.
Flumes
The flumes necessary for the required flow measurements were
designed in accordance with the requirements shown in Table 4.
Table 4. FLUME DESIGN DATA
Location Sewer Diameter Capacity
Ring and Richards 30" 13 cfs
Auer and Booth 36 " 21 cfs
Auer and Booth 42 " 31 cfs
Locust and Richards 36 " 21 cfs
Humboldt and Locust 54 " 60 cfs
Humboldt and Commerce 60 " 77 cfs
The capacities shown on the foregoing tabulation were nominal
and were established by the flume manufacturer for heads equal to
that resulting from a depth of flow in the upstream sewer of about 85
percent of sewer diameter. The flumes were further calibrated for
heads resulting from a full depth of flow in the upstream sewer. Flows
exceeded flume capacity on a few occasions.
The flumes were designed for permanent installation in a half
section of sewer, and were installed with appropriate anchorage devices.
The entrance and discharge ends of the flumes have a semi-circular
invert section of the same diameter as the sewer in which each is
installed. The flumes were manufactured of a corrosion-resistant
resin reinforced with fiberglass mat to provide a minimum wall
thickness of 1/8 inch throughout. Each flume was designed to produce
metering heads within 2 percent of the theoretical rating curve for the
physical conditions existing at each location.
40
-------
FLOW RATE
RECORDER-
•FLOW RATE
INDICATOR
FLOW TOTALIZER
NITROGEN
CYLINDERS
LIQUID LEVEL
RECORDER
^OVERFLOW
WALL
SAMPLE
CONTAINER
WITHIN
REFRIGERATOR
LEGEND
AIR LINES
SEWAGE SAMPLE LINE
GATE VALVE
CHECK VALVE
PNEUMATIC RED VALVE
SOLENOID VALVE
•SAMPLER CHAMBER
FILTER REGULATOR
(» WITH DRAIN
^ PRESSURE GAUGE
=0
IB
f
CONSTANT FLOW
PURGE METER
FLOW MEASURING
DEVICE
AIR BUBBLERS
PRESSURE REDUCING
VALVE
FIGURE 12- SCHEMATIC FLOW DIAGRAM FOR MONITORING STATIONS
41
-------
Open Flow Nozzle
The open flow nozzles necessary for required flow measurements
were designed in accordance with the following data:
Table 5. NOZZLE DESIGN DATA
Pipe and Nozzle
Location Diameter Capacity
Auer and Fratney 20" 7. 0 cfs
Auer and Humboldt 36" 31.0 cfs
Locust and Richards 24" 11.6 cfs
The nozzles were calibrated for flows up to the nozzle head
which resulted from full depth of gravity flow in the upstream sewer
section. The contour of the cast iron nozzle causes essentially equal
changes in nozzle head loss for corresponding changes in flow. The
change in head is sensed at a bronze piezometer opening to -which is
attached a gas bubbler pipe.
The manufacturers of the primary measuring device provided
the manufacturer of the liquid level and flow indicating, recording
and totalizing equipment with all the information required for the design
and calibration of the associated instrumentation.
Flow Indicating, Recording and Totalizing Instrumentation
The flow indicating, recording and totalizing equipment for each
primary measuring device (flume, open-flow nozzle, or gate orifice)
•was of the head measuring type. The head -was measured by the use of
a constant flow purge meter assembly using nitrogen gas flowing from
a double pressure reducing station mounted on high pressure cylinders.
The gas cylinders were stored inside the monitoring station housing.
The flow-head signal was transmitted to a direct reading, flow
indicating lenearized in the instrument enclosure and transmitted as
a linear, 4 to 20 milliamp DC electronic signal to the electronic strip
chart recorder and electronic, 8 digit flow totalizer in the enclosure.
Metering range and accessory functions for each device are shown in
the following Table 6.
42
-------
Table 6. MONITORING STATIONS -
FLOW INSTRUMENTATION
Monitoring Station
Humboldt and Wright
Ring and Richards
Auer and Fratney
Auer and Humboldt
Humboldt and Locust
Auer and Booth
Locust and Richards
Humboldt and
Commerce
Type of
Primary
Device
Orifice
One Bubbler
30" Flume
Two Bubblers
20" Nozzle
Two Bubblers
36" Nozzel
Three Bubblers
54" Flume
Two Bubblers
36" Flume
Two Bubblers
42" Flume
Ono "RnlVhl *=» T
V^ilC J_> LIU U J-C -L
36" Flume
Two Bubblers
24" Nozzle
Two Bubblers
60" Flume
Two Bubblers
Orifice
Onf RnViKlor-
Flow Range
0-10. 5 cfs
0-18. 0 cfs
0-7. 0 cfs
0-31.0 cfs
0-78. 0 cfs
0-28. 0 cfs
n 38 0 rfc;
VJ -> O • VJ *_J»O
0-28. 0 cfs
0-11. 6 cfs
0-100 cfs
n_8 ^ r-fs
Accessory
Functions5'
1,3
1,2,3
1,2,3
1,2,3
1,2,3
1,2,3
2
1,2,3
1,2,3,4
Accessory Functions:
1-Control Ejector Sampler
2-Control Upstream Bubbler
Gas Supply
3-Measuring System Shutdown
4-Control Detention Tank
Startup
43
-------
The flow instrumentation for certain primary devices
controlled the operation of ejector type sewage samplers, as indicated
in the preceding Table 6. In such cases, the totalizer would retransmit
pulses representing preselected flow increments, to an electric,
adjustable, predetermining counter. The counter in turn actuated the
proper relays to control valves. The above described system resulted
in a sample being obtained each time a given quantity of flow passed
through the primary measuring device. Control of valves at the sample
container caused the sample to be discharged to it. In addition, the
sample ejection system was activated at the start of a sampling cycle;
that is, when the bubbler valves were opened, and the flow measurement
system was started for a given overflow period.
Control of valves at the sampler caused the sampler to refill
immediately following ejection of each sample. At the end of each
overflow' period, after the measuring and sampling system was shut
down, a final sample remained in the ejector sampler. Controls for
valves at the sample container were therefore provided so that, during
the next overflow period, the initial sample discharged to a drain.
At various stations the instrumentation also controlled the
operation of upstream liquid level recording equipment as indicated
in the preceding Table 6. Where such accessory function was required,
the flow instrumentation included controls for opening a solenoid valve
on the nitrogen gas supply for the upstream liquid level bubblers when
the flow through the primary device reached 95 percent of capacity.
The flow instrumentation also performed the function of shutting
down the gas bubblers and flow measurement systems, by closing the
gas bubbler solenoid valves when the flow dropped to zero. Measuring
systems were started by bell-type pressure switches, as later described.
The zero flow shutoff circuit incorporated an adjustable delay
timer. The timer held the solenoid valves open under zero flow conditions,
for an adjustable period of 0 - 15 minutes. The delay timer •was necessary
to eliminate the surge effects which develop due to varying intensities
of runoff during rainfall events.
Liquid Level Recording Equipment
Liquid levels measured by all flow measurement devices, and
by all upstream liquid level bubblers, were recorded. The flow measure-
ment and upstream liquid level gas bubbler backpressures were piped to
individual, electronic, differential pressure level transmitters. The
4 to 20 milliamp DC liquid level signals were transmitted to electronic,
44
-------
strip chart liquid level recorders. The range on all level records was
0-10 feet, except that the range on the recorder for liquid level in
the 24 inch sewer at Auer and Humboldt was 0-2 feet.
As an auxiliary function, the liquid level instrumentation at the
Humboldt and Wright and at the Humboldt and Commerce monitoring
stations shut down the flow measurement systems when measured
head dropped to a predetermined level. This function was necessary
due to the continuous dry -weather flow at these locations.
Ejector-Type Samplers and Controls
Each ejector-type sampler consisted of a sample chamber
located in the sewage flow channel, pinch-type valves at the chamber
inlet and outlet connections, flexible preassembled gas and sample
tubing connecting the sample chamber to valving and piping within the
monitor station housing, solenoid valves to control sampler operation
in accordance with signals from flow measurement instrumentation,
sample containers and container refrigerator, and all gas and sewage
sample piping and controls.
The sample chambers were sized to obtain a 100 milliliter
sample. The chambers were designed with an inlet check assembly
so that gas under pressure, when applied in accordance with flow
signals, would evacuate the sample to the sample container. Chambers
were fabricated of polyvinylchloride, with a Teflon seat and nylon ball.
Sample entrance openings were 1/8 inch in diameter and the gas inlet
and sample outlet lines were 1/2 inch and 1/4 inch diameter, respectively.
Each sample chamber and valve assembly was equipped with a chain
lifting device to permit its removal from the sewage flow channel
through a 12 inch diameter access pipe.
The sample chamber was connected to the monitoring station
piping by four flexible hose connections. Two of these lines were for
gas to operate the pinch-type valves, the third line conducted high
pressure gas to the chamber for evacuation of the sample. The fourth
line conducted the sample to the sample container.
The sample containers were rigid, self-supporting polyethylene
plastic drums, with a capacity of 10 gallons. The container refrigera-
tors were standard, commercial units.
Vacuum Type Samplers
The station located at Humboldt and Commerce was equipped
with a vacuum type, individual sampler. This vacuum actuated,
45
-------
automatic sampler was capable of collecting 24 individual samples at
predetermined intervals. Each of the 24 individual samples was at
least 250 milliliters in volume when collected at a sampling lift of
15 feet.
The sampler was equipped with a thermostatically controlled
refrigeration unit for cooling and storage of the collected samples
and with a vacuum pump, both of which were enclosed in the monitoring
station housing.
Bell Type Pressure Switch
Each monitoring station included a bell-type pressure switch
for activating the flow measurement system gas bubbler, when overflow
conditions occurred. The switch operated on the principal of compressed
air trapped within a bell by the rising liquid level which mechanically
closed a contact within the monitoring station housing. Contact closure
opened a solenoid valve on the gas supply to the bubbler for the primary
measuring device. The pressure switch would seal in, and would not
close the solenoid valve -when the pressure switch opened. The shutoff
of gas was accomplished by a signal from the flow measurement equip-
ment when the flow rate dropped to zero. To eliminate surge affects,
an adjustable time delay was provided in the pressure switch circuit.
The complete assembly was designed to cause closure at the
remotely located pressure switch when rising water levels at the bell
location resulted in a liquid depth of 1^ inch above the bottom of the bell.
Accessory Equipment - Typical for all Sewer Monitoring Stations
Each sewer monitoring station included the following accessory
equipment:
Ventilation louvers and fan
Thermostatically controlled heating unit
Two 100 watt, incandescent lighting fixtures
Space for two nitrogen cylinders and pressure
reducing valves and piping
Each item of flow and liquid level instrumentation
was provided with an engraved nameplate
describing the function being indicated or
recorded, as well as indicating the primary
device such as "Flume", "Nozzle", "Upstream
Bubbler", etc. All timers and switches were
also identified as to system and function.
46
-------
Combined Sewer Monitoring Station Housing
The monitoring station housings were sized to accommodate
all units and facilities described above. The housings had overall
dimensions approximately 6'-6" high x 7'-0" long x 3'-0" deep.
Figure 13 shows a typical sewer monitoring station housing arrangement.
Photographs of a typical combined sewer monitoring station are shown
in Figures 14 and 15.
The monitoring station housings were generally located on parkways
between sidewalks and curbs. These locations restricted station depth
to approximately 3'-0".
The housings were fabricated of twelve gauge steel, shop primed
with red lead primer and finished with two coats of green enamel. Doors
were provided to permit opening the entire front and rear of each station
for easy access to all equipment, valves and piping. The doors are
equipped with cylinder locks and all locks were keyed alike.
All monitoring station components (flow and liquid level instru-
mentation, valves and piping, samplers, refrigerators, etc. ) were
factory mounted within the housings. All piping was copper tubing,
neatly arranged, and marked or tagged to show line function. Unions
were provided at all valves and at equipment connections to facilitate
removal for maintenance or replacement and all pressure reducing
valves on bubbler supply lines were provided with valved by-passes to
permit manual blowdown. The power supply to each station was single
phase 120/240 volt, 60 cycles. All units were completely factory wired
and connected to a circuit distribution panel.
RIVER MONITORING STATION FUNCTION
The function of the three river monitoring stations was to measure,
indicate and record the dissolved oxygen concentration and temperature
of flow in the Milwaukee River; obtain samples of the river water
automatically; and store samples in a refrigerated container.
RIVER MONITORING STATION COMPONENT EQUIPMENT
Vacuum Type Samplers
Each of the three river monitoring stations was equipped with a
vacuum actuated, automatic, individual sampler. The samplers were
identical to the vacuum sampler previously described for the combined
sewer monitoring station at Humboldt and Commerce.
47
-------
GAS
CYLINDERS
REFRIGERATOR
CD
(APPROX. DIMENSION)
VALVE
ASSEMBLY-
FLOW RECORDING
ASSEMBLY
HEATER
JION
KENT
1 <
to '
z
UJ
S
(APPROX.
[
* 3'-°" ul
APPROX. OlMENSIONlj
h*-A
I ' '!
ill
FRONT VIEW
ELEVATION A
PIPE CONDUITS TO
UNDERGROUND
CHAMBERS
SIDE VIEW
ELEVATION B
FIGURE 13-SEWER MONITOR HOUSING AND EQUIPMENT LAYOUT
-------
FIGURE 14
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVENUE DEMONSTRATION PROJECT
COMBINED SEWER MONITORING STATION
STATION 49
-------
Ln
O
FIGURE 15
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVENUE DEMONSTRATION PROJECT
COMBINED SEWER MONITORING STATION
STATION 41
-------
rVENTILATION ASSEMBLY
-REFRIGERATOR a
SAMPLER
-REFRIGERATOR a
SAMPLER
'"TEMPERATURE
EQUIPMENT
.^SAMPLER
f ACCESS
PIPE
J^-HEATER
SAMPLER
LINES
D.O. a TEMPERATURE
EQUIPMENT
SAMPLER
LINES
FRONT VIEW-ELEVATION A
SAMPLER
ACCESS
PIPES
SAMPLER ACCESS-*
D.O. PROBE
ACCESS PIPE
D.O. PROBE
ACCESS PIPE
SIDE VIEW-ELEVATION B
SAMPLER! LINE
REAR VIEW
FIGURE 16
RIVER MONITOR HOUSING AND EQUIPMENT LAYOUT
51
-------
Dissolved Oxygen Concentration Analyzers
The Cherry Street Bridge and St. Paul Avenue Bridge river
monitoring Stations were each equipped with one temperature com-
pensated, automatic dissolved oxygen concentration analyzer. To
evaluate the effect of the aeration of the river water as it passes over
the North Avenue Dam two such devices were provided at this location;
one located upstream and one downstream of the dam.
The analyzers were installed to provide an accurate, continuous
record of the dissolved oxygen concentrations in the river. However,
due to chronic probe malfunction, very little useful data was realized
from the use of these probes. The problems encountered in the
operation of these analyzers are discussed in Section VII of this Report.
Accessory Equipment - Typical for all River Monitoring Stations
Each river monitoring station also includes the following accessory
equipment:
Ventilation louvers and fans
Thermostatically controlled heating unit
Two 100 -watt, incandescent lighting fixtures
River Monitoring Station Housing
The river monitoring station housings were very similar to that
described for the combined sewer monitoring stations except that they
were slightly smaller due to less space requirements. Figure 16 shows
a typical river monitoring station housing arrangement. Photographs
of a typical river monitoring station are shown in Figures 17 and 18.
RAINFALL GAUGING
In order to ultimately relate U. S. Weather Service records from
outside the project study area to the rainfall pattern within the area, a
rain gauge installation was required.
During the initial stages of the project, the City of Milwaukee
installed a gauge on the roof of the Pierce Elementary School located
near the center of the project area. The gauge was a recording, tipping
bucket type rain gauge. This gauge was later replaced with a recording,
weighing type gauge. Early in the project, vandalism necessitated the
relocation of the gauge to another point centrally located in the project
area.
52
-------
Ln
OJ
FIGURE 17
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVENUE DEMONSTRATION PROJECT
RIVER MONITORING STATION
STATION 65
-------
FIGURE 18
CITY OF MILWAKEE, WISCONSIN
HUMBOLDT AVENUE DEMONSTRATION PROJECT
RIVER MONITORING STATION
STATION 41
-------
In addition to the rainfall data obtained from the project area,
records from the City recording rain gauge located at the Municipal
Building were also collected and studied. This gauge is located directly
south of the project study area.
MONITORING STATION EQUIPMENT SUPPLIERS
The various component equipment was supplied by a general
contractor in accordance with detailed equipment specifications. The
suppliers of the various components are listed below:
Flumes F. B. Leopold Inc.
Open Flow Nozzles BIF
Flow Indicating
Recording Totalizing,
and Ejector Sampling
Equipment Fischer and Porter Co.
Station Fabrication
and Auxiliary
Equipment Fischer and Porter Co.
Vaccuum Samplers Sonford Products
D. O. Concentration
Analyzers Union Carbide Co.
Selection of general contractor and the equipment suppliers was
based upon general municipal bidding procedures and does not constitute
endorsement or recommendation for use by the City of Milwaukee.
STATION OPERATION AND MAINTENANCE
The operation and maintenance of the monitoring stations was a
demanding aspect of this demonstration project. The sampling phase
of the project was frequently hampered by equipment malfunction.
To reduce the amount of monitoring station down-time, the City
assigned two men to assist the Engineer in the operation and maintenance
of the monitoring stations during the final year of data collection. Their
presence substantially improved the results of the sampling program.
These personnel were able to repair, in the field, many pieces of equip-
ment that the manufacturer's service representative had previously
found necessary to remove and return to the factory for repair.
Aside from the equipment operation problems which were encountered
and which required additional maintenance time, the routine operation
and maintenance of the monitoring stations consisted of the following
activities:
55
-------
1. Replacement of empty nitrogen cylinders.
2. Weekly inspection of piping connections to
eliminate nitrogen leaks as they appeared.
3. Monthly change of recorder charts and ink
capsules.
4. Cleaning of composite samplers and vacuum
sampler lines.
5. Weekly check of the calibration of the level and
flow measuring and indicating instrumentation.
To insure that the stations were ready to operate automatically,
the following activities were performed immediately after each rain-
fall:
1. Samples were picked up from the stations
and delivered to the laboratory for analysis.
2. Totalizers were checked and total flows
recorded for use in preparing samples for
laboratory work.
3. Pressure bells were checked and any rags,
debris, etc. caught on them by the rising
sewer level were removed.
4. Nitrogen bubbler lines were blown out to
remove any grease or other substance which
may have collected in the open end of the
lines.
5. The vacuum sampler at the Humboldt and
Commerce monitoring station was
reset.
The routine activities necessary to the operation and maintenance
of the stations, can be accomplished in one or two days per week with
a few hours extra during periods of rainfall necessary to collect samples
and perform the above post-rainfall tasks.
A detailed discussion of problems encountered in operation of the
monitoring stations, together with recommendations for resolving such
difficulties in future studies, is presented in Section VII.
56
-------
MONITORING STATION COST DATA
Detailed plans and specifications for the monitoring stations
were completed early in 1968. Bids were received on March 20,
1968 when the Engineering News-Record Construction Cost Index was
1117. 15. The low bid for the eight sewer and three river monitoring
stations was $278, 818, with changes and alterations to the project
increasing the construction cost by approximately $3, 000. The total
construction cost of $281,818 reflects a per monitoring station cost of
approximately $25,600.
The installed prices for the various major items of equipment
quoted by the low bidder as well as the estimated cost of each monitoring
station is given in Table 7.
57
-------
Table 7. MONITORING STATION COST DATA
Major Equipment Items
Monitoring Stations Including
All Instrumentation $ 121,000.00
Flumes 8,930.00
Open Flow Nozzles 7,633.50
Station Cost Including Major
Equipment Items'1'
Combined Sewer Stations
Station 21 - Humboldt and Wright $ 19,000.00
Station 25 - Ring and Richards 24, 000. 00
Station 29 - Auer and Fratney 24, 000. 00
Station 33 - Auer and Humboldt 25, 000. 00
Station 37 - Humboldt and Locust 31, 000. 00
Station 41 - Auer and Booth 38, 000. 00
Station 45 - Locust and Richards 42,000.00
Station 49 - Humboldt and Commerce 44, 818. 00
River Stations
Station 52 - St. Paul Bridge 10,000.00
Station 58 - Cherry Street Bridge 10, 000. 00
Stations 64 and 66 - North Avenue Dam 11, OOP. 00
TOTAL $ 278,818.00
^Individual Station Costs are based on engineers' estimates - a break-
down of each station's cost was not requested of the Contractor.
58
-------
SECTION VI
DETENTION TANK
GENERAL
The Humboldt Avenue Combined Sewer Overflow Detention Tank
is located at the primary outlet for the demonstration project combined
sewer area. The tank was constructed underground on the north bank
of the Milwaukee River on North Commerce Street just west of North
Humboldt Avenue. Figures 4 and 6 show the detention tank location.
The tank receives combined sewer overflow from a 570 acre study
area shown in Figures 2, 3 and 6. Two relief sewers,which traverse
the area and the Sewerage Commission's intercepting sewer remove
from the system a substantial amount of the total combined sewage
generated within the study area before it reaches the detention tank.
The effective capacity of the tank is approximately 3. 9 million
gallons and the approximate interior dimensions are: 420 feet long, 75
feet wide and 16 feet deep. For design purposes it was estimated that
the tank would receive the following maximum inflow rates:
Approximate 5 year storm flow 270 cfs
Approximate 10 year storm flow 320 cfs
Approximate 20 year storm flow 380 cfs
The above estimated detention tank influent flow rates were utilized
where applicable in selecting and sizing the various equipment components
which were to be incorporated into the project. The maximum values
were utilized in developing the hydraulic elements of the detention tank
and the combined sewerage conduits.
The detention tank volume was selected based on several factors.
The major factors are noted below:
1. Site Capacity - In order to be able to take full advantage
of the available site, several alternate tank layouts were
investigated. Various depths and horizontal shapes were
evaluated. Based upon the site limitations, as well as the
59
-------
anticipated funds available, it was determined that a tank -with
approximately 3. 9 million gallons of capacity would be the maximum
practical size.
2. Volume of Capacity versus Volume of Rainfall Per Storm -
A rainfall histogram based on 16 years of record was developed
for the project. This histogram is included in Appendix VI.
The histogram was utilized to establish the approximate number
of rainfall events for which the tank volume would be exceeded
and thus cause overflow to the River. It was planned that this
number of events, which would exceed the tank capacity, would
not be too large as to prevent effective pollution control but still
be economically reasonable.
3. Economic Considerations - The project was initially
arranged with a specific budget allowance for the detention tank
construction. When the detailed design concept of the tank was
finalized, it was determined that the 3. 9 million gallon capacity
tank might exceed the budget available. For this reason and in
order to provide the maximum practical tank volume for the
budget available, bids were taken on five alternate tank sizes.
Bidders quoted prices on a 3. 9 million gallon capacity and also
on four alternate size tanks of smaller capacity. As can be
noted in the portion of this Section on Detention Tank Cost Data,
the cost per unit of detention tank volume was significantly less
for the largest tank size than for the alternate sizes. This factor
•was weighed heavily in establishing the tank size.
Due to the size of the tank and its depth, considerable uplift
pressure forces were anticipated. During pre-design site investigation,
an extremely hard soil condition was found to exist about 20 feet below
existing grade which made the use of hold-down piles for resisting the
uplift pressures impractical. As a solution to this problem, the final
design called for the placement of approximately eight feet of fill over
the top of the tank to resist the anticipated uplift. The fill area was
landscaped and could be utilized for recreational purposes at some
later date.
The detention tank and its component equipment are indicated in
Figures 19, 20 and 21. A photograph of the tank is included as Figure 22.
60
-------
^OVERFLOW WEIR (DEPTH OF FLOW MEASURED)
V .BAFFLE WAL«*™LORtNATON DIFFUSES
1 ' BAFFLE
73 INLET SEWER
60" SLUICE GATES
COMBINED SEWAGE PUMPS
14" COMBINED SEWAGE
MIXER (TYPICAL)-^
TO M. I S SEWER
(FLOW METERED)
COMBINED SEWAGE
SUMP FOR TANK
DEWATERIN6
LIQUID LEVEL BUBBLER
BAFFLE WALL
BAR SCREEN
PRECHLORINATION OIFFUSER
, \ 96 EFFLUENT SEWER
\ V-SLOPE TO RIVER
EFFLUENT SAMPLER
78" BYPASS SEWER_
TO RIVER
INVERT ELEV 3 3
SECTIONAL PLAN
SEE FIGURES 20 a 21 FOR SECTIONS
DTENTION TANK-SECTIONAL PLAN
-------
NJ
ELEV -IOO
SUPERNATANT SUCTION t ELEV. -5.0-s
PUMP DISCHARGE TO M I. S. SEWER (METEREO) t ELEV -6 5 -^ J
ELEjr._=II.Ql
SUMP SUCTION t ELEV. -13.5 --
ELEV -K
ELEV -16 0
SECTION A-A
FIGURE 20
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVE. DEMONSTRATION PROJECT
DETENTION TANK-SECTION
-------
MIXER HOUSING
54 » 84 OVERFLOW
INVERT ELEV. 10 5
ELEV z.e^'•''''
i CHLORINE DIFFUSER
ELEV -4.0
MIXER IMPELLER
(TYPICAL)
72" INLET SEWER INVERT ELEV 36
SECTION B-B
FIGURE 21
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVE. DEMONSTRATION PROJECT
DETENTION TANK-SECTION
-------
FIGURE 22
CITY OF MILWAUKEE, WISCONSIN
HUMBOLDT AVENUE DEMONSTRATION PROJECT
DETENTION TANK
-------
DETENTION TANK COMPONENT EQUIPMENT
Summary of Equipment Data
Following is a summary of the detention tank component equipment
data:
BAR SCREEN
Manufacture r
Number of Units
Type
Clear Openings, Inches
Capacity, CFS
MIXING EQUIPMENT
Manufacturer
Number of Units
Type
Motor Horsepower, Each
COMBINED SEWAGE PUMPS
Manufacture r
Number of Units
Type
Pump No. 1
Pump No. 2
Pump No. 3
Pump No. I
Capacity, GPM
Total Dynamic Head, Feet
Speed, RPM
Capacity, GPM
Total Dynamic Head, Feet
Speed, RPM
Pump No. 2
Capacity, GPM
Total Dynamic Head, Feet
Speed, RPM
Rex Chainbelt, Inc.
1
Heavy Duty-Mechanically Cleaned
1. 5
380
Mixing Equipment Co. , Inc.
7
Rotary
40
A.G. McKee & Co. (WEMCO)
3
Non-clogging, Vortex Type
Vertical, Pedestal Mounted
Two Speed
Constant Speed
Constant Speed
400
17. 5
700
10
870
400
30
900
17. 5
1, 170
900 1,500
30 10
1, 170
65
-------
Pump No. 3
Capacity, GPM
Total Dynamic Head, Feet
Speed, RPM
SLUICE GATES AND OPERATORS
Manufacturer
Number of Units
Size, Inches
Type
CHLORINATION EQUIPMENT
Manufacture r
Chlorinators:
Number of Units
Capacity (each), Lbs. /day
Chlorine Evaporators:
Number of Units
Capacity (each), Lbs. /day
Chlorine Residual Analyzer:
Number of Units
2, 200
30
3,200
10
870
Rodney Hunt Company
2
84 x 60
Rising Stem
Wallace & Tie man Company
2
8, 000
2
8, 000
Chlorine Leak Detector:
Number of Units
Type
Chlorine Flow Recorder:
Number of Units
Type
Metering Range, Lbs. /day
Indicator Type
Continuous Sampling
Differential Pressure Orifice
0-8000
Direct Read-Out of Remaining
Chlorine
Chlorine Cylinder Hoist:
Manufacturer
Number of Units
Capacity, Tons
Type
Shepard-Niles
1
2
Geared, Chain Driven
66
-------
Chlorine Residual Sample Pump:
Manufacturer Robbins & Meyers, Inc.
Number of Units 1
Type Horizontal, Helical Screw
Positive Displacement
Capacity, GPM 20
Total Dynamic Head, Feet 35
PROCESS MEASUREMENT AND CONTROL EQUIPMENT
Manufacturer Fischer and Porter
Type Bubbler
OVERFLOW SAMPLER
Manufacturer Sonford Products Corporation
Number of Units 1
Type Automatic Vacuum Actuated
Sample Volume 200 ml at 20'
Selection of the various contractors and the equipment suppliers
•was based upon general municipal bidding procedures and does not
constitute endorsement or recommendation for use by the City of
Milwaukee.
Bar Screen
All combined sewage entering the detention tank passes through
a mechanically cleaned bar screen located in the tank influent channel.
The screen is designed to pass a peak flow of 380 cfs. The purpose
of the screen is to remove sewage solids, rags and other debris which
otherwise would settle and pile up on the tank bottom or be drawn into
the tank dewatering pumps. Screenings are discharged to large portable
containers and are then removed to a land disposal site along with all
other refuse collected in the City.
Detention Tank Mixing Equipment
The detention tank serves as a settling basin, therefore the
deposit of solids over the tank bottom must be removed.
One design concept contemplated the use of bottom scrapers for
the removal of settled solids. The final tank layout however made the
use of such scrapers impractical. The final design therefore was not
based on separate removal of sludge but relied on a system of mixers
67
-------
designed to resuspend the settled solids into the liquid within the tank
during the time the tank contents are pumped to the intercepting sewer.
Seven rotary mixers are located within the tank, each driven by
a 40 hp motor. Each stainless steel mixer impeller is 7 feet 6 inches
in diameter and was installed approximately 22 inches above the tank
floor. The impellers were set at thin elevation, which is adjustable
within a 36" range. Studies were not made relating to most efficient
mixing impeller elevation or to operation of only a selected number
of the mixers during certain periods to improve the solids suspension
capability. The mixer drive equipment is located above each mixer
on the tank roof, enclosed in a concrete housing which terminates at
grade. Each chamber is entered through access manholes.
The mixers were designed to resuspend, within one hour after
start-up, all entrapped solids to the following extent:
Solids having settling rate of 1 ft/min or less:
Uniform suspension throughout the basin.
Solids having settling rate of 1 to 4 ft/min:
Off bottom motion.
Solids having settling rate of 4 to 8 ft/min:
Off bottom motion in radius of 30 feet from mixer.
During the first year of tank operation, the mixers satisfactorily
performed their intended function with the result that it has not been
necessary to manually clean the detention tank after receiving approxi-
mately 181 million gallons and detaining approximately 121 million
gallons. Solid deposits, consisting mainly of sand and silt, accumulated
to a maximum depth of approximately 11 inches and -were limited to small
areas, three to six feet in width, along the perimeter of the tank. These
areas of deposit changed location along the tank perimeter during the year,
possibly due to one or more of the following variable factors:
1. Influent flow rate
2. Effluent pumping rate
3. Mixer speed
4. Solids concentration in influent flow
5. Length of detention time
68
-------
A log of the sediment accumulated in the detention tank as of March
20, 1973 is included in Figure 23.
Tests during future operation of the tank could possibly determine
the causes of the deposits and indicate what remedial action is necessary
should Ihe continuing deposit of solids bo deemed undesirable. To
date this is not the case.
One of the seven rotarv mixers is equipped with a two-speed motor
drive. This mixer, located approximately 150 feet from the tank over-
flow weir, is operated at low speed prior to and during periods of tank
overflow to distribute chlorine for disinfection. It is felt that this slow
speed mixer operation may be causing solids loss during overflow, to
increase somewhat.
Each mixer is rated to be suitable for operation as a surface
aerator. The rating indicates that with a tank liquid depth of approximately
48 inches and 20 percent oxygen available in ambient atmosphere, each
mixer is capable of dissolving oxygen at a rate of 10 pounds per hour.
Actual detention tank tests to verify this rating were not performed
since this aeration feature is not the main design function of the mixers.
Studies into the economics and possible treatment benefits of operating
the mixers as aerators may be a subject for future tank operation
evaluation.
Combined Sewage Pumps
Three vertical, dry pit, pedestal mounted combined sewage pumps
are available for dewatering of the tank. Each pump is of the non-
clogging, vortex type with recessed impeller suitable for pumping
combined sewage.
A sump agitation header was installed in the tank sump to prevent
solids build up. Its operation has not proved satisfactory, however it
has not been necessary to replace it.
Measurement and Control Systems
All measurement and control systems included in the operation
of the detention tank, with the exception of the combined sewage pump
discharge measurement system, are of the bubbler tube type utilizing
nitrogen or compressed air. The compressed air system is located
in the control building while the nitrogen system is located in a monitoring
station (Humboldt and Commerce Station) 300 feet from the building.
Back pressure, caused by the level being measured, is piped to electronic
differential pressure level transmitters, which in turn provide 4-20 ma
DC level signals to electronic strip chart recorders.
69
-------
S
jWFFLt V
oooooooo
Q< (X (&
oooooooo 0*6 OQ
-w
•r
®
.*'* X «i
d
T^c'-.. ',
I
J ©
/v,
v.Oi ^
DUFFLE WAU
(£
DATE
A
i
l&-2-72i^
3-20-73J2
?
3
4
1
d
7
?
3
7
3
3
6
4
-------
Influent Flow Measurement System - A flume level signal is
transmitted via a two-wire transmission line from the sewer monitoring
station at Commerce and Humboldt to the detention tank control panel
where it is converted to a flow signal and the flow indicated, recorded
and totalized. The function of this system is to initiate on-off operation
of the bar screen, chlorine evaporators, chlorinators and a protected
water pumps. The measured rate of influent flow also controls the rate
of prechlorination if prechlorination is being utilized.
Tank Liquid Level System- One bubbler tube is located within
the detention tank. The function of this system is to initiate the on-off
function of process equipment. At various tank liquid levels this
system performs the following operations:
1. Turn off mixers and pumps
2. Change speed of two-speed pump
3. Initiate post chlorination system
4. Initiate tank overflow sampling equipment
Throughout the course of the project the influent flow measure-
ment system was subject to continued malfunctioning. The major
problem encountered -was the yet unexplained loss of back pressure
through the bubbler system. This loss of back pressure occurred at
seemingly random moments during periods of sewer overflow and lasted
for as long as 15 minutes per occurance during which time the influent
flow rate could not be recorded and prechlorination of the tank influent
flow on a flow proportional basis was not possible. To provide accurate
influent flow rate data, it was necessary to compute the tank volume
on a per foot basis such that the increase in tank level could then be
converted to rate of flow. This method did provide the required flow
data for all storms except those producing runoff in excess of the 3. 9
million gallon tank capacity. To obtain the influent flow rate during
periods of tank overflow the following measurement system was installed.
Tank Overflow Weir-Head Measurement System - A bubbler tube
was installed at the tank overflow weir to measure, indicate, transmit
and record the liquid level over the crest of the weir. Using hydraulic
formulae the resulting head measurements were converted to rate of
tank overflow. A check between this system and that of the influent
flow measuring system, when functioning, has proven the substitute
system to be as accurate as the system initially intended.
Combined Sewage Pump Discharge System - All combined sewage
pumped from the tank passes through a 12 inch magnetic flow meter.
The meter transmits a flow proportional, 4 to 20 ma DC signal to
indicating, recording and totalizing equipment located at the main
control panel.
71
-------
M. I. S. Sewer Liquid Level System - One bubbler tube -was installed
in a manhole upstream from the point of tank discharge to measure,
indicate, transmit and record the liquid level in the M. I. S. sewer.
The system functions to coordinate the operation of the combined sewage
discharge pumps with the available level in the M. I. S. thus eliminating
the possibility of its surcharge with a resultant undesirable backwater
effect.
Tank Overflow Sampler
A vacuum actuated, automatic, individual sampler is located
in the sampler building at the southwest corner of the detention tank.
The sampler is capable of collecting 24 individual samples at predetermined
intervals and is activated whenever the tank approaches an overflow to
the river. The sampler head is run into the tank through a 12 inch
diameter access pipe to a point approximately 6 feet below the crest
of the tank overflow weir. The sampler collects samples approximately
200 milliliters in volume which are stored in a refrigerated compartment
until transfer to the laboratory.
Chlorination Facilities
It was concluded that liquid chlorine would be utilized for disinfection
of the detention tank overflows and for pre-chlorination to prevent
odors. Liquid chlorine was selected as the chlorine source in lieu of
various chlorine compounds such as chlorinated lime, sodium hypoch-
Jorite or calcium hypochlorite. This selection was based upon the
following factors:
1. Cost of liquid chlorine is less
2. Equipment must be capable of an automatic
operation. The liquid and gas feeding
equipment was best suited for the automated
operation.
3. The system must be capable of feeding large
and variable quantities of chlorine to the
detention tank influent.
The detention tank chlorination facilities include the following
items of equipment for handling and feeding chlorine:
2 8000 pounds per 24 hours evaporators with high and low
temperature alarms.
2 8000 pounds per 24 hours vacuum operated chlorinators.
4 2000 pound chlorine storage tanks with manifold.
1 amperometric-type chlorine residual analyzer and recorder.
72
-------
1 portable amperometric titrator used to calibrate
the residual analyzer and recorder.
1 chlorine leak detector, to continuously sample
the air in the chlorine rooms, and alarm
system.
2 chlorine flow recorders to record use of chlorine
and indicate remaining chlorine available.
DETENTION TANK OPERATION
Combined sewer overflow is directed to the tank by gravity, through
a 78 inch diameter sewer. At the tank inlet the combined sewage enters
either the tank inlet channel or a bypass channel depending upon the
position of two 84 by 60 inch sluice gates. Use of the bypass channel
allows direct discharge of combined sewage to the Milwaukee River,
therefore its use is limited to emergency conditions when the tank must
be kept empty, which up to the time of this report was not necessary.
Upon entering the tank inlet channel, the combined sewage passes through
the mechanically cleaned 1^ inch bar screen. All solid material which
is too large to pass through the screen is removed from the sewage
flow and deposited in 3 cu. yd. portable disposal containers.
During sewer overflow to the tank, the seven rotary mixers are
not operated, except for the low speed operation of the one mixer used
to disperse chlorine, and the tank serves as a settling basin. Should
a storm be of extreme intensity or long duration and generate an influent
flow in excess of tank capacity, this settling results in the removal
of a portion of the setteable solids from the combined flow prior to
overflow to the Milwaukee River.
After each storm the mixers are activated when pumping commences.
The tank contents are thoroughly mixed to resuspend the settled solids
so that the pumps deliver the entire tank contents, including the solids,
into the M. I. S. for conveyance to the Jones Island wastewater treatment
plant.
Chlorination facilities are provided to permit both pre-and post-
chlorination of the tank contents. When pre-chlorinating, chlorine is
administered in the tank inlet channel on a flow proportional basis. Pre-
chlorination is practiced as an aid in odor control. The need for pre-
chlorination is subject to many variables including: 1) time of year
2) type and duration of storm, 3) anticipated time to empty detention
tank.
The post-chlorination system is activated when tank overflow to
the river is anticipated and is administered for the purpose of disinfection.
73
-------
During post-chlorination, the rate at which chlorine is added is controlled
by the chlorine residual as determined by the residual analyzer, as
well as on the flow proportional basis.
The pre- and post-chlorination chlorine solution is distributed
through diffuser headers. The pre-chlorination header is located in
the tank inlet channel. The diffuser runs across the inlet channel just
ahead of the inlet parts. The post-chlorination diffuser is located
approximately 177 feet from the end of the 420 feet long tank. The
diffuser distributes chlorine across the entire 75 feet width of the tank
at the point approximately 12 feet above the tank floor.
Piping is also available to administer chlorine solution at the base
of each mixer for tank content high rate dosage. To date the effects
of utilizing this system have not been studied.
DETENTION TANK MAINTENANCE
During the first year of tank operation, maintenance requirements
were carried out by two employees of the City of Milwaukee. The
experience of these men and manufacturers' recommendations were
utilized in the development of the following list of required maintenance
procedures:
Daily
1. Check chlorine residual analyzer acid supply and add acid
as needed.
2. Check that chlorine evaporator cathodic protection indicator
is in the green range.
3. Replace chlorine leak detector sensitized paper.
4. Check recorder and equipment which is in continuous operation.
Weekly
1. Check operation of sump pump
2. Check air compressor oil level
3. Drain air dryer
4. Visually check ground indicator lights
5. Change chlorine recorder chart and ink pen
6. Renew chlorine residual analyzer buffer solution
7. Check non-continuous operating equipment
Monthly
1. Check seal water lubrication of sewage pumps
2. Inspect insect screen on exterior fresh air intake to
boiler room and clean if necessary.
74
-------
3. Change recorder strip charts and ink capsules on all
recorders at main control panel.
4. Clean chlorine residual analyzer sample cell (more
frequently with extended use).
5. Operate chlorine residual analyzer sampler and
sample pump using manual override (tank liquid
level must be at a minimum of 10 feet to take
samples).
6. Inspect condition of and operate hand-operated chlorine
valves.
7. Operate deluge showers and eyewashers to insure proper
operation.
8. Inspect insect screens on vent piping and clean as
necessary.
9. Inspect unit ventilator filter and clean as necessary.
10. Check oil level in rotary mixer housing.
Quarterly
1. Grease sewage pumps.
2. Oil heat distributor motor on evaporator No. 1.
3. Grease and lubricate bar screen (more frequently
with extended use).
4. Grease rotary mixers.
5. Grease seal water pump.
6. Change oil for air compressor.
7. Purge tank level and M. I. S. level bubbler systems
from blow down block located in metering panel.
8. Lubricate chlorine leak detector blower motor bearings.
Semi-Annually
1. Blow down hot water boiler and heater.
2. Check amperage on all 3-phase equipment.
3. Clean unit heater filter.
4. Flush all floor drains to check for blockage.
5. Clean chlorinator rotometer tubes.
6. Clean air-handling unit filter.
7. Grease chlorine residual sample pump.
8. Clean and lubricate trip motor for vacuum sampler.
9. Backwash vacuum sampler tubing (more frequently
with extended use).
10. Clean chlorine residual sample pump strainer.
11. Test chlorine leak detector vacuum tubes and replace
as necessary.
12. Oil unit ventilator motor.
13. Change air purifier and flow equalizer in chlorine
recorder-controller.
75
-------
Annually
1. Change Oil in mixers.
2. Change Oil in fluid couplings for mixers.
The following maintenance procedures must be performed in
addition to those listed above:
1. Change chart on chlorine residual recorder-
controller after each use.
2. The portable residual titrator cell must be maintained
with distilled water and electrolyte tablets and the
cell must be kept submerged in sample or tap water
to prevent drying of the cell tip.
PERSONNEL REQUIREMENTS
Prior to tank start-up, it was anticipated that two men would be
required to operate and maintain the tank equipment through a three to
six month shake-down period. During this time all equipment was to
have been checked, adjusted and fine tuned to the point that complete
confidence could be placed in the automatic operation of the equipment.
Due to equipment malfunctioning, which extended beyond the first six
months of operation, and to insure that as much reliable data as possible
would be collected, during the one year test period, the City employed
two men on a full-time basis. These men were assigned to operate and
maintain the tank equipment for the entire final year of the data
collection phase of the demonstration project.
It is assumed that normal tank operation, that is considering only
those maintenance procedures previously listed, would require the
attention of one man for approximately 2 hours per day, seven days
per week. During and immediately following a period of rainfall, it
would be advisable to have a man visit the tank site to check that all
automatic operations functioned properly during the influent period.
When it is found necessary to enter the tank or mixer equipment
chambers at least two men should be on duty and the proper safety
equipment should be employed.
DETENTION TANK COST DATA
Construction Costs
Detailed plans and specifications for construction of the detention
tank were completed early in 1969. Bids were received on July 10,
1969 and construction began shortly thereafter.
76
-------
In July, when the bids were received, the Engineering News-
Record Construction Cost Index for the Chicago area registered 1, 367
and the Environmental Protection Agency Index for the Chicago area
registered 135.
Bidders quoted prices on a 3. 9 million gallon capacity tank and
also on four alternate size tanks of smaller capacity.
The prices quoted by the successful bidder, for each of the five
tank sizes, are shown on the following Table.
Table 8. HUMBOLDT AVENUE DETENTION TANK BID PRICES
Approximate Tank Capacity (Million Gallons)
3. 9 3. 3 2. 8 2. 2 1.7
Section I
General
Construction $1,107,860 $1,045,860$ 987,860 $ 930,860 $ 870,860
Section II
Electrical
Section III
Plumbing
Section IV
Heating fe
Ventilation
39,490
38,595
26,940
37, 546
38, 381
25, 789
34, 812
38, 167
24,589
31,480
37,953
23,389
28, 580
37, 736
23, 620
Totals * $1,212,885 $1, 147, 579 $1, 085, 428 $1,023,682 $ 960,796
Cost Per *
Gallon of
Capacity $ 0.31 $ 0.35 $0.39 $ 0.47 $ 0.57
*Costs Include the Cost of the Equipment and
Control Building Associated with the Detention
Tank.
Construction Cost Projections - Other Facilities
The Humboldt Avenue Detention Tank can provide a basis for
estimating costs of similar facilities which may be considered at other
locations. In order to utilize costs at other locations, the Humboldt
Avenue costs can be related to the ENR Construction Cost Index.
77
-------
In order to relate the costs to tanks of other sizes, the Humboldt
Avenue costs can also be related to unit volumes.
To estimate future tank construction costs, it is necessary to
first project the index levels. The following tabulation shows past and
predicted index levels for the Chicago area:
Table 9. ENGINEERING NEWS-RECORD CONSTRUCTION COST INDEX
ENR Base: U. S. Average in 1913 = 100 - Labor and Materials
Increase from
Index Previous Years
December, 1968 1336
December, 1969 1408 5. 3%
December, 1970 1603 13.9%
December, 1971 1837 14.6%
December, 1972, as
predicted by ENR 1970 7. 3%
December, 1972
Actual 1964 6.9%
December, 1973, as
predicted by ENR 2075 5. 6%
In 1969, the 3. 9 million gallon capacity tank cost $1, 212, 885
to construct. Of this amount, approximately $765, 000 was required
for construction and equipping of the Control Building and other basic
facility appurtenances not specifically related to tank volume. Since
a similar building and related equipment would be required regardless
of tank size, that cost can be deducted from the total cost to establish
the basic tank volume cost. Thus the cost of the tankage only for the
Humboldt Avenue Detention Tank was estimated to be approximately
$414, 000. This cost is meant to be the cost of adding the tank volume
to the basic facility of the Control Building, conduits, equipment, etc.
Construction costs, for tanks ranging in capacity from 1. 0 to
10. 0 million gallons, are shown in Figure 24. Included are costs as
existed in 1969 and also costs as predicted for December, 1973
based on the ENR Construction Cost Index. These costs include only
the detention tank cost and do not include the cost of the equipment
building and appurtenances.
When the data presented herein are utilized as the basis to
estimate cost for detention tanks, it may be necessary to incorporate
appropriate factors into the estimates for local conditions which would
influence cost. Consideration should be given to subsurface conditions,
78
-------
TANK CONSTRUCTION COST (MILLION
IO
DOLLARS)
r°
b
1
2
TANK
r OF TANKONL
EQUIPMENT Bl
\ A
. CAPACITY
Y- NOT INCLUDI
JILDING AND A
' €
f (MILLIOf
NG THE COSTO
PPURTENANCES
i i
* GALLONS
F
O"j
\\
Q \
\\
\
P
3
^
a
>
i
mm
O
0
^
\
\
Si
\\
•A
1C. \
»r \
£ \
!» \
i<» \
f* N
»
i
»
4
mot
ZP13C
3BOPI
ma
z£3
oS-
m2ni
K'c
ii —
IO<0
O"1
i«»i
•^|UI
at
v/
V
\
\
\
k
\
\
I
i
i
\
\
\
\
O
m
-\
m
z
H
O
Z
o
m
H
m
z
H
O
_,<^ ~ °
—I <• O C
>W 0 a,
30
C
O
O
-<
ij Q
o
O
-------
available site size, environment protection restriction, etc. The sub-
surface conditions of site of the Humboldt Avenue Detention Tank could
be classified as relatively difficult construction.
The costs presented in Figure 24 do not include the cost for
constructing and equipping the tank Control Building and appurtenances.
Figure 25 shows the estimated construction cost for a detention tank
building and appurtenances, comparable to the Humboldt Avenue facility,
at various ENR index levels.
To estimate total project costs for planning purposes, additional
costs must be added to the construction costs presented in Figures
24 and 25. These additional costs can be estimated on the basis of
percentage of total construction cost, for the purposes of this Report,
as follows:
Engineering (Design, General Inspection
and Detailed Inspection) 12%
Construction Contingencies 5%
Fiscal, Legal and Administrative 8%
Miscellaneous Special Studies (Soil Borings,
Operation and Maintenance Manual,
Engineering Reports in Support of
Financing, Environmental Impact
Study, etc. ) 5%
Total Incidental Cost Factor 30%
'o
The site of the Humboldt Avenue Detention Tank was acquired
by the City prior to the inception of this project and, therefore, an
additional cost for land was not a consideration. Land costs being
so variable from one location to another, this addition to the total
project cost must be separately evaluated for the particular area of
study.
Operating Costs
The following costs of operation are based on a one year period
of operation between November, 1971 through October, 1972,
inclusive. The operating costs presented include costs associated
with the operation and maintenance of the Humboldt and Commerce
monitoring station:
80
-------
co
Z
o
o
FIGURE 25
CONTROL BUILDING a APPURTENANCES COST
VS
ENGINEERING NEWS-RECORD CONSTRUCTION
2200-
x
UJ
o
- 2000
V)
o
o
z
2 1800
H-
o
1600
IX
o
u
ui
QC 1400
oo
ui
Z
- 1200
tC
ui
UI
z
o
z
UI i
COST INDEX
DECEMBER. 1973
( PREDICTED)
"DECEMBER.1972
1.0 1.5
COST OF CONTROL BUILDING a APPURTENANCES
(MILLION DOLLARS)
81
-------
Table 10. HUMBOLDT AVENUE DETENTION TANK
OPERATING COSTS
Replacement Parts and Equipment $ 2, 172. 51
Nitrogen 94. 08
Chlorine 698. 00
Miscellaneous Supplies 1, 154.47
Utilities (Light, Heat, Telephone, Etc.) 10,059.92
Labor 35.717. 35
Total Operating Costs $49, 896. 33
The cost of labor included in the above Table represents
approximately 2.2 man-years. This relatively large amount of
labor time resulted from start-up problems during the first year of
operation, and from the program of having personnel on intensive
duty to insure that as much reliable data as possible would be collected
for incorporation into the demonstration project studies. To perform
the routine operational and maintenance procedures required at the tank,
it is estimated that as little as 0. 50 man-years per twelve-month
period would be required.
Possible incorporation of a connection to the detention tank com-
pressed air system or a separate system at the Commerce and Humboldt
monitoring station would eliminate the cost of nitrogen from the tank
operation. The cost associated with replacement parts and equipment
was, perhaps, 50% higher than v/ould be expected for first-year operation.
A reasonable estimate of operating cost for a typical detention
tank at another location is difficult to develop because costs will vary
with tank size, amount of flow handled per year, policy on require-
ments for 24-hour per day attention at the facility, and other variables.
However, as a guide, the following estimate is presented for general
budget purposes:
Table 11. GENERAL OPERATING COST BUDGET -
TYPICAL DETENTION TANK
Based on 1972 Costs for Labor and Services
Parts and Equipment $ 2, 500. 00
Chlorine and Nitrogen 1, 500. 00
Miscellaneous Supplies 2, 000. 00
Power and Other Utilities 12, 000. 00
Labor 12,000.00
Total Estimated Annual Operating
Cost Budget $30, 000. 00
82
-------
SECTION VII
DATA COLLECTION PROGRAM
GENERAL
Early in the Demonstration Project planning stage an extensive
monitoring program was established. Its purpose was to facilitate
the orderly collection of the data necessary in evaluating the effectiveness
of the detention tank approach in reducing pollutional loadings to the
Milwaukee River caused by overflow from the project study area combined
sewer system. The data to be collected was to be used to characterize
the effect of the resultant reduction of pollutants on the quality of water
in the river and to enable the prediction of the effect of a system of
such tanks on the river water quality. In addition, the data was to be
used to aid in projecting the suitability of this method of handling
combined sewer overflow to areas of varying size and runoff characteristics
as is discussed in Section X of this Report.
To satisfy these objectives, it was deemed necessary to thoroughly
characterize the following conditions which are major influences on
;v and quantity determination:
CHit-TclC 1C I IXjtS tllC HJliUWlllg (_UilUlt J.UJ
quality and quantity determination:
Sewer System
Dry weather flow and quality:
1. Seasonal variation.
2. Average daily variation during each of the four seasons.
3. Four to six hour variation.
4. Influence of time since, and duration of, last rain storm.
Wet weather flow and quality:
1. Average quantity and quality of overflow as influenced by:
a. Time of day storm occurs.
b. Time since ]ast storm.
c. Intensity and duration of the storm.
83
-------
2. Variation in overflow quantity and quality during a specific
storm as influenced by:
a. Time of day storm occurs.
b. Time since last storm.
c. Intensity and duration of the storm.
Milwaukee River
Dry weather quality:
1. Seasonal variation.
2. Variation during the day in each of the four seasons.
3. Influence of the flushing tunnel.
Wet weather quality:
1. Seasonal variation.
2. Variation during the rainfall event.
3. Influence of the flushing tunnel.
The data collection phase of the project with brief interruptions
extended over a five year period, beginning in July, 1967 and continuing
through November of 1972. By the nature of its purpose, it was viewed
as consisting of two stages.
The first, or pre-tank, stage encompassed all data collected
prior to the commencement of detention tank operation. It began with
the collection of baseline river water quality and flow information
from the Milwaukee River and the river flushing tunnel, which served
as background data for the monitoring program. Also included in this
stage was the collection of combined sewer wet and dry weather sewage
quality and quantity data, Lake intake water quality data, and project
area rainfall gauging. Finally in the first stage, other available data
pertinent to this project was compiled for review. This included:
past rainfall records from the National Weather Service recording
station at General Mitchell Field; Milwaukee River discharge data
recorded by the U. S. Department of Interior Geological Survey at its
Estabrook Park gauging station; and river water quality data compiled
by the Sewerage Commission of the City of Milwaukee since 1946.
The second, or post-tank, stage of the monitoring program
consisted of the sampling necessary to evaluate the detention tank per-
formance as a treatment unit and was initiated following tank start-up.
Included in this stage was the collection of river water quality and flow
data; detention tank influent and overflow quality and quantity data; and
84
-------
a continuation of wet and dry weather combined sewage quality and
quantity data and rainfall data collection.
Following is a discussion of the specific monitoring programs
conducted during the five years of data collection. Much of the data
compiled during that time is presented in tabular or graphic form in
Section VIII of this Report and discussions regarding the ultimate use
of the data are presented in Section IX.
RAINFALL DATA
Hourly precipitation data of record, from September of 1948
through November of 1972, for the Milwaukee area was developed.
The recorded data was obtained from the U. S. Weather Service which
maintains a recording gauge at General Mitchell Field. This location
is approximately 8 miles south of the project study area.
Rainfall dataare also recorded at two City maintained rain gauge
installations. One gauge is located within the boundaries of the project
study area, and the other is located atop the Municipal Building, approxi-
mately 2 miles south of the project area.
Rainfalls were classified by intensity, to one-hundredth of an inch,
and by duration for 0 to 10 hours.
Computer processing of the data provided the following information:
1. Printout (-with punch cards optional) indicating each event,
its corresponding average intensity and duration time, and
the time interval to the next event.
2. Each event was classified based on its duration and intensity.
It was then added to a frequency matrix for that month of the
year, for that month of all years, and for all months of all
the years.
3. Printout (with punch cards optional) indicating the frequency
matrix of the individual months over all the years and the
frequency matrix for all the months over all the years.
SEWER MONITORING
Wet Weather Sewer Monitoring
The eight combined sewer monitoring stations were placed in
service in the summer of 1969. However, during the station testing
and trial operation period and continuing through 1971, only a limited
85
-------
number of storms resulting in combined sewer overflow in the project
area were experienced. This small number of high intensity, long
duration storms provided limited opportunities to obtain the data required
to characterize wet weather combined sewage.
Further hindering the study were the numerous equipment malfunctions
encountered throughout the program. Initial installation and operation
of the monitoring equipment appeared to be quite satisfactory, but
frequent malfunctioning began to occur shortly after commencement of
station operation. The problems mainly centered around the flow measure-
ment systems and their electrical and mechanical controls in the monitoring
stations.
The abnormally dry weather conditions and station down-time due
to frequent equipment failures combined to result in a minimal amount
of wet weather sewer flow and quality data during this period.
Fortunately, in the final year of data collection (1971-72) ralnfallm tfee project
study area was in excess of average amount of record. Also during
the final year, an increase in monitoring station operational personnel
decreased the occurances of equipment failure. This, combined with
the above average rainfall, resulted in a much improved sampling
program.
Sufficient wet weather combined sewage data was gathered during
this period of the program to fully characterize the quality and quantity
of the flow generated within the project study area. Composite samples
were collected at the seven upstream stations on a flow proportional
basis. The Commerce and Humboldt monitoring station, located immediately
upstream from the detention tank, measured and recorded the flow rate
and took samples of all sewer overflows discharging to the tank. Flow
proportional composite samples were collected at this station and, in
addition, a sampler was provided which permitted the collection of
individual samples at predetermined time intervals. The timing device
of this individual sampler was normally set to collect samples at 10 to
30 minute intervals.
Dry Weather Sewer Monitoring
Dry weather quality sewer monitoring was restricted to the Humboldt
and Wright and Humboldt and Commerce monitoring stations. It was
observed that the strengths of the dry weather sewage flow incident
to these stations is characteristic of flow generated during dry weather
within the entire project area.
The four seasons of the year were represented in the sampling
program and many of the samples collected were analyzed on a per hour
86
-------
basis in order to render the data as flexible in use as possible. Samples
collected at the Humboldt and Wright station were composite samples
collected over 24-hour periods. At the Humboldt and Commerce station,
however, the sampling equipment available allowed samples to be collected
on either an hourly or composite basis.
DETENTION TANK MONITORING
During periods of tank overflow the detention tank serves as a
treatment unit. To determine the degree of treatment being administered
to the combined sewage, influent flow characteristics were compared
to the characteristics of samples collected at the tank overflow weir.
Overflow samples are obtained using an individual sampler similar to
the one located at the Commerce Street monitor station. Samples
were automatically collected during tank overflows at from 10 to 60
minute intervals.
Flow measurement systems utilized at the detention tank were
described in Section VI.
RIVER MONITORING
During the design, construction and initial operation of the
monitoring stations, base line water quality and flow information was
collected on the Milwaukee River and the river flushing tunnel. This
data was required to provide an overall profile of the river as it encounters
the combined sewer area of Milwaukee and as it existed prior to the
operation of the detention tank.
An important feature of the Milwaukee River, downstream of the
North Avenue Dam, is the flushing tunnel which is used to augment the
river flow. Samples of the lake water at the tunnel intake were collected
on a weekly basis for a total of approximately 8 months during -which
time the tunnel -was in operation. This sampling -was required to
provide the quality data necessary to characterize the effect of flushing
on the river water quality. The full impact of flushing could not be
characterized, however, without first determining the rate of discharge
through the tunnel.
For this determination, Rhodamine B dye was used to measure
the time of travel bet-ween the tunnel intake at Lake Michigan and its
outlet at the river. The Milwaukee Sewerage Commission provided
radio-equipped service vehicles to signal the exact time of dye introduction
into the tunnel intake. Samples were taken from a boat stationed on the
river approximately 100 feet downstream from the tunnel outlet. Two
tests were run for verification using different dye concentrations.
87
-------
During the first test, samples were collected at one minute intervals
and at 15 second intervals during a part of the second tests. The
samples were returned to the laboratory and fluorometrically analyzed.
In addition to the sampling, visual observations were made directly
at the tunnel outlet. In the first test, using 12 gm of Rhodamine B dye
diluted to 1 liter, visual observation of the dye occured 11. 5 minutes
after its introduction at the tunnel intake and sample analysis indicated
a travel time of 12 to 13 minutes from the intake to the point of sampling.
In the second test, 30 gm of Rhodamine B dye diluted to 2 liters
was used. Again, visual observation indicated a travel time of 11. 5
minutes from the tunnel intake to outlet and laboratory analysis of the
samples collected indicated a travel time of 12. 75 minutes to the sampling
point.
Therefore, using a travel time of 11. 5 minutes for discharge
calculations, the tunnel discharge rate was found to be approximately
200, 000 gallons per minute or 422 cubic feet per second. When this
tunnel discharge rate is compared to the normal range of river discharge,
as recorded by the U. S. Geological Survey, the importance of the
tunnel operation to river water quality becomes quite apparent. The
tunnel discharge can account for from 3 to 100 percent of the river flow.
In addition to the flushing tunnel studies, extensive river water
quality background data was collected during a period of approximately
21 months beginning in July of 1967. Samples were taken at various
locations along the river, generally on a five or six day per week basis.
This sampling schedule was interrupted only by severe weather conditions
or thick ice cover on the river which prevented sampling.
During this period of background data collection, one intensive
48 hour sampling survey was also conducted. Samples were taken
manually at two hour intervals at four locations along the river and at
the river flushing tunnel Lake Michigan intake.
To supplement the extensive background sampling, conducted
prior to monitoring station construction, additional wet and dry weather
river samples were taken using the automatic sampling devices available
at the monitoring stations. These samplers are individual, vacuum-
type samplers which permit sampling at predetermined time intervals.
River monitoring station design called for the installation of four
temperature compensated, automatic dissolved oxygen concentration
analyzers. One analyzer was installed at the Cherry Street station, one
at the St. Paul Avenue Station and two at the North Avenue Dam monitoring
station. Installation of two probes at the dam, one upstream and one
downstream,, was deemed to be desirable to evaluate the effect of aeration
above and below the dam.
-------
The four probes originally installed were plagued by numerous
mechanical and electrical problems. The probe manufacturer attempted,
on several occasions, to remedy the problems encountered but due to
continuous malfunctioning the four probes were eventually abandoned -
having provided a limited amount of useful data. Their failure did prove
that the state of the art in the case of the Milwaukee River, is not yet
developed to the degree necessary for automatic operation without
nearly continuous maintenance.
In the spring of 1972, two dissolved oxygen analyzers of different
manufacture, were installed at the Cherry Street Bridge river monitor
station for the purpose of testing and evaluating their performance.
For comparison purposes, river samples were manually collected, were
analyzed for dissolved oxygen content using the Winkler titration method
and the results were compared to the output of each probe. During the
course of their evaluation, both probes exhibited an encouraging degree
of reliability and accuracy when properly calibrated and diligently cleaned.
It was concluded from the evaluation that the advantage of employing
automatic dissolved oxygen analyzers, in lieu of manual collection of
samples, is questionable when considering collection of D. O. data on
the Milwaukee River. The reasons for this is that the Milwaukee River
presents a seemingly unique scum growth problem. During the warm
weather months, when D. O. data is of highest importance, any material
coming in contact with the river water is soon covered with a coating
of an algae-grease substance. As this scum adheres and increases in
thickness over the tip of the D. O. probe the output of the probe is
gradually diminished.
To preserve the accuracy of the probe output, the probe must be
cleaned at least once each day with a dilute solution of hydrochloric
acid. This frequent cleaning results in a more rapid change in the
probe calibration and, therefore the probes must be recalibrated
approximately once each week. To insure that accurate data is being
recorded by the probes, samples must be collected manually on a daily
basis to determine when recalibration is necessary. The time spent
in cleaning and checking the accuracy of the probe could and was better
utilized to collect samples for Winkler analysis.
A continuous record of dissolved oxygen was desirable, initially,
to assist in the evaluation of the effect of combined sewage overflow on
the river dissolved oxygen levels on a random basis. That is, having
a continuous record of the D. O. concentrations would have permitted
a comparison of D. O. concentration changes to combined sewage over-
flow quantities and strengths for any storm of record. In not having a
continuous recording, it was necessary to collect river samples manually
89
-------
during a storm and then only the effect of that particular storm could be
evaluated. A continuous record of D.O. would also be valuable in showing
the effect of the flushing tunnel operation on the D.O. in the river.
To obtain sufficient data to satisfy the requirements of the demon-
stration project, it was necessary to conduct four intensive, dissolved
oxygen sampling surveys. Two of the surveys were conducted during
periods of dry weather, to identify the base line D. O. conditions in the
river, and two surveys were conducted during periods of wet weather to
determine the effect of combined sewage overflow on the D. O. levels.
The four surveys were scheduled such that a wet and a dry weather
survey were conducted prior to initial operation of the detention tank
and the final two surveys were made after the tank had been placed into
service.
The first two surveys were five days in length and the final two
surveys were extended to eight days in length. During the surveys,
samples were collected at various locations along the river at 2 to 4
hour intervals and at least once per day at various spatial locations.
The dissolved oxygen concentration of each sample collected was determined
by the Winkler titration method.
The Milwaukee River locations and the spatial locations at which
samples were collected during these surveys included the following:
Station
Number Identification
Milwaukee 52 St. Paul Avenue
River 58 Cherry Street
59 South Water Street
62 Humboldt Avenue
66 North Avenue Dam -
Upstream
65 North Avenue Bridge
Spatial 40 Flushing Tunnel Intake
Locations 47 Milwaukee River Harbor
81 Menomonee River at
South 2nd Street
82 Kinnickinnic River at
Kinnickinnic Avenue
90
-------
LABORATORY ANALYSES
All samples collected were delivered to the Marquette University
Sanitary Engineering Laboratory for analysis. All chemicals and
laboratory equipment necessary for the intended analyses were readily
available at the lab.
Laboratory analyses of the samples were conducted under the
direction of Dr. Raymond J. Kipp. At all times during the demonstration
project, highly qualified technical personnel were employed to expedite
sample analyses. A qualified technician was employed on a full time
basis for the most of the five years of the project and part-time
assistance was provided by senior students in environmental engineering
or graduate students from the same area.
The parameters chosen for investigation during the project were
classified into groups of descending importance. The degree of importance
of each parameter was based on the relative utility of each in analyzing
and modeling the performance of the storm tank in one case, and the
river impact in the other. Table 12, following, summarizes the analyses
which were performed during the program. Table 13 indicates the order
of importance of each parameter for each system.
The following schedule of analyses, presented in Table 13 was
divided into three parts to indicate the order of data preference and is
interpreted as follows:
(A) To be determined in all cases-data is critical to analysis
of tank performance and river response.
(B) Data is not critical to development or evaluation of system
models. Data is important for developing a comprehensive
picture of system quality, but occasional gaps in data will
not be detrimental.
(C) Data is of limited value to system analysis or shown by prior
data to be of low concentrations and/or small variations in
concentration.
The analytical results from combined sewage and tank overflow
samples were used to characterize performance of the detention tank as
a treatment unit. BOD, suspended solids, and total and fecal coliforms
are pollutants which the tank was expected to remove and the tank model
describes the removal of these parameters. Complete data was desirable
on organic and ammonia nitrogen also. Organic nitrogen was expected
to be affected but could not be modeled due to a lack ot definition between
the suspended and soluble fractions. Ammonia nitrogen was not expected
to be affected by the tank but complete data was desirable because it
represents an oxygen demand in the river.
91
-------
Table 12. ANALYSES PERFORMED
Parameter
Biochemical Oxygen Demand
Chemical Oxygen Demand
pH
Total Coliforms
Fecal Coliforms
Nitrogen Series
Total Organic
Ammonia ( NH^ )
Nitrite (NO2 )
Nitrate (NOj )
Phosphate Series
Total
Ortho
Solids Series
Total
Suspended
Volatile Suspended
Cl2 Demand
Chlorides
Conductivity
Dissolved Oxygen
Temperature
Sewer
Monitor
Stations
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Tank
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Rive r
Monitor
Stations
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Locations
X
X
X
X
X
X
X
X
X
X
X
X
92
-------
Table 13. SCHEDULE OF ANALYSES
Utilized To Establish Priority of Analyses Based
On Project Requirements
Combined Sewer and
Tank Overflow Samples
Biochemical Oxygen Demand
Suspended Solids
Total Coliforms
Fecal Coliforms
Nitrogen Series
Total Organic
Ammonia
(A)
River Samples
Biochemical Oxygen Demand
Chlorides
Total Coliforms
Fecal Coliforms
Nitrogen Series
Ammonia (NH^)
Chlorides
Phosphate Series
Total
Ortho
Volatile Suspended Solids
Cl Demand
(B)
pH
Chemical Oxygen Demand
Total Solids
Nitro gen
Nitrite (NOz)
Nitrate (NO3)
pH
Chemical Oxygen Demand
Nitrogen Series
Total Organic
(C) Nitrite (NO2)=
Nitrate (NOs)
Phosphate Series
Total
Ortho
Solids Series
Total
Total Suspended
Volatile Suspended
The foregoing schedule of analyses was developed after considering
past experience in combined sewage characteristics, tne importance
of each parameter to the tank and river modeling efforts, and the
background river data collected prior to 1969»
93
-------
In developing the river model, parameters utilized included BOD,
ammonia nitrogen, total and fecal coliforms, water temperature and
dissolved oxygen. Analysis for chloride was specified as a means of
quantifying the effect of Lake Michigan sieche activity on the dispersion
characteristics of the lower Milwaukee River.
The laboratory staff attempted at all times to conduct a thorough
analysis of each sample collected. During extended periods of rainfall
or at times of numerous successive storms, an excessive number of
samples were generated and delivered to the laboratory. To ease the
resulting load on the lab staff, without sacrificing data of importance to
the modeling efforts, guidelines were developed for reducing the number
of lab tests required.
The first preference was to reduce the individual number of samples
to be analyzed by compositing individual samples -where this could be
done without compromising data. For example, a single composite
could be made of all short term individual tank influent flow samples or
short term tank overflow samples could be composited to yield one hour
composites.
When further reduction in the analytical load was necessary,
certain parameters were eliminated from the investigation. Parameters
were eliminated in the order of their importance to the study as shown
in the schedules in Table 13.
ANALYTICAL PROCEDURES
Experience gained from the analysis of background Milwaukee
River samples and "grab" samples of the overflow from the project
study area combined sewer system suggested the following analytical
methods be employed in the investigation of the parameters of interest
to the study.
4
Biochemical Oxygen Demand Standard Methods
Dilution technique.
Chemical Oxygen Demand Standard Methods
Dichromate reflux method.
Standard Methods
Glass-electrode electro-
metric method.
Total Coliform Standard Methods
Membrane filter procedure
using M-Endo broth.
94
-------
Fecal Coliform
Total Kjeldahl Nitrogen
Ammonia Nitrogen
Nitrite Nitrogen
Nitrate Nitrogen
Total Phosphate
Ortho Phosphate
Total Solids
Total Suspended Solids
Total Volatile Suspended
Solids
Chlorine Demand
Chlorides
Method as recommended by
Geldreich, Clark, Huff and Best5
MFC medium (Difco) with incuba-
tion at 44.5°C.
Standard Methods
Standard Methods
For river samples and for low con-
centration in combined flow, the
direct nesslerization procedure
was used. A Spectrophotometric
analysis was used with a standard
curve. Samples with higher con-
centrations of ammonia were
analyzed by the distillation and
titration procedure.
Standard Methods
Spectrophotometric procedure.
Standard Methods
Phenoldisulfonic acid method
Standard Methods
Nitric acid - sulfuric acid
Method.
S tandard
Stannous
Methods
Chloride
procedure.
Standard Methods
Standard Methods as revised. In
lieu of the asbestos mat in a
Gooch crucible, a glass fiber
filter was used with a 5% Celite
solution to seed the filter.
Detailed procedure immediately
follows this listing.
Standard Methods
Standard Methods
Laboratory Method.
Standard Methods
Argentometric (silver-
nitrate) Method.
95
-------
Conductivity Specific conductance was measured
with a Hellige, Inc. Conductivity
Meter. Three ranges were avail-
able:
0-50 Micromhos
0-500 Micromhos
0-10,000 Micromhos
Procedure consisted of measuring
the sample temperature, setting
proper temperature on the in-
strument and reading the micromhos
directly with suitable probe
immersed in the sample.
Dissolved Oxygen Standard Methods
Azide modification of the
iodometric method was used for
the laboratory analyses. Samples
collected in the field were fixed
immediately and titrated follow-
ing return to the laboratory.
Water Temperature Field measurement at time of
sample collection.
Suspended Solids Determination Procedure
1. Place a glass filter paper (rough side up) at the bottom of
the gooch crucible.
2. Seat the glass filter paper with about 1 ml of distilled
water.
3. Add 5 ml. of 5% solution of celite (running the solution
down the sides of the cricible).
4. Allow celite solution to settle for 2 minutes and then apply
va cuum.
5. Place porcelain disc on top of filter and celite mat.
6. Wash the filter and celite with 3 separate 10 ml washings
of distilled water.
7. Place in 103°C oven for 24 hours.
8. Place in a 600°C muffle furnace for 20 minutes.
9. Place in a 103°C oven for 24 hours.
10. Place in a desiccator for 1 hour.
11. Weigh the crucible (for tare weight)
12. Filter the sample through the gooch crucible. Quantity of
sample to be determined on the basis of how difficult fil-
tration becomes.
13. Place in a 103°C oven for 24 hours.
14. Place in a dessicator for 1 hour.
15. Weigh crucible and sample to obtain weight of the suspended
solids in the sample.
96
-------
The above described suspended solids determination method
or similar variations have been adopted at many laboratories
throughout the country as an improvement in analysis technique.
DATA HANDLING
Immediately upon analysis, each item of combined sewer
and river quality data was recorded in a laboratory notebook
by the lab technician. The raw quality data was then transfer-
red to IBM 529 forms. These forms were put through an optical
reader which automatically punched the data onto IBM Cards.
Duplicate sets of cards were prepared in this manner for dis-
tribution. This multiple distribution of data cards provided
a guarantee that data would not be lost due to fire, vandalism
or the like. Combined sewer flow rate data, after being taken
from the monitoring station strip chart recorders, was also
punched onto IBM Cards.
The raw data punch cards were then processed to provide
a listing of the data distinguishing between river and combined
sewer, wet and dry weather quality data and combined sewer
flow rate data. The quality data was divided into two parts
separating the parameters into two sections to permit printing
of the data on normal IBM print out sheets.
Listings of all the data accumulated during this project
are referenced in Section XII.
OPERATION AND MAINTENANCE PROBLEMS AND RECOMMENDATIONS
In other sections of this report reference has been made
to operating problems encountered in the monitoring and
sampling systems employed on this project. Details and an
equipment description was given for the monitoring station
in Section V and for the Detention Tank components in Section
VI.
The following is a discussion of the operating problems.
Suggestions and recommendations are presented in order that
such problems may be reduced or eliminated on future,
similar monitoring systems.
97
-------
Instrumentation
Operating problems resulted from relatively frequent failure
of flow indicating, totalizing and recording equipment, and the
related electronic hardware associated with monitoring station control
and sampling systems. The degree of difficulty was probably not much
greater than that involved in a typical water or wastewater treatment
plant incorporating similar equipment. However, such failures in the
stations had a much more severe impact on the demonstration project
than they might have had on the operation of a treatment plant.
Failure frequently meant the loss, or potential loss, of data
which had to be accumulated in a relatively short span of time, and
program timing did not, for the most part, include the luxury of retrieving
lost data at a subsequent time. (An exception to this condition concerned
data lost due to problems with the monitoring stations during the first
year of operation. Delay in the construction of the detention tank moved
back the project timing which provided extra time to obtain supplemental
data from the stations. )
When instrumentation failures occurred, equipment downtime
was excessive for the purposes of this study. Delays were encountered
in the arrival of manufacturers' service personnel, in obtaining repair
parts and in sustaining unit operations even after service visits.
Multiple, simultaneous failures caused complex repair problems, not
easily remedied by station operating personnel.
For future, similar projects, the following measures should be
considered:
1. Provisions should be made to include, well-trained service
personnel in the project staff, virtually on a full-time basis if data
collection is to be a major function of the project. This can be
accomplished by requiring such service as part of the instrumentation
supplier's responsibility, or could be part of the Owner's or Engineer's
staff. A typical service contract, providing for dispatching personnel
from the manufacturer's service organization would not, in most cases,
be an adequate arrangement.
2. Arrangements should be made to insure adequate stocking
of spare parts. Such parts could be purchased and stored by the Owner,
or stockpiled locally by the manufacturer on a guaranteed basis.
3. Consideration should be given to providing backup systems
for the most critical system components.
98
-------
Bell-Type Pressure Switches
Each sewer monitoring station was equipped with a bell-type
pressure switch designed to activate the flow measurement system whenever
overflow occurred. As the sewer liquid level rises over the mouth of
the bell, the column of air trapped in the tubing connecting the bell to
the station is compressed. This causes the tripping of a pressure
sensitive rubber membrane switch at the station, which, in turn,
initiates station operation.
The bells were originally positioned so that the mouth was
approximately 1/2 inch below the level of overflow at each station.
The resulting 1/2 inch compression of the air column was sufficient to
trigger the pressure switch. However, over the four year period of
operation the rubber membrane apparantly hardened, losing its sensitivity.
This problem was not discovered until the last few months of operation
and, therefore, new membranes were not installed as would normally be
advised. For the final few months of station operation the bells were
lowered 1/2 to 1 inch to provide the compression necessary to trigger
the switch.
As the hardening of the rubber membrane could be dependent
upon frequency of use, ambient temperature, gases, etc. , the useful
life of the membrane would vary for different installations. The
sensitivity of each switch should, therefore, be checked frequently to
insure reliable operation.
Because the bells were located directly in the sewage flow path
they were easily clogged by rags, tissue, etc. , and required cleaning
following each overflow event. Also, under certain flow conditions the
bells failed to respond to the rising sewer liquid level. A possible
reason for this is that under high flow rates the flow around the bell
may have caused a pressure differential between the air in the tubing
and the water surging around the bell. With the water pressure below
the bell less than the air column pressure, compression of the air column
would be impossible.
To eliminate this problem, the pressure bells were relocated to
the overflow portion of the chambers. This relocation resolved the
problem. Another solution to the problem could be the construction of
a stilling well adjacent to the sewer. Availability of a stilling well would
also permit the possible use of a float device, or an air bubbler system,
in lieu of the pressure bell.
Nitrogen Supply
Nitrogen gas was utilized in the operation of the monitoring
station flow measurement bubbler systems and composite samplers
99
-------
Nitrogen was chosen for use because it would not induce biological action
in collected samples. It would also not combine with gases present in
sewers to form explosive mixtures.
The use of nitrogen did, however, present some problems in the
operation and maintenance of the stations. Replacement of empty
cylinders was a difficult and time-consuming task which required the
availability of personnel able to lift and carry the heavy cylinders.
A greater problem, directly affecting the data collection effort,
was the loss of nitrogen due to leaks in the system. Leaks small enough
to evade detection were however large enough to empty a nitrogen
cylinder in just a few days. If this was not discovered in time, a small
leak in the system could empty a cylinder and eliminate all sampling and
flow monitoring until a new tank could be installed. This problem was
overcome through a daily visitation schedule established during the final
year of testing.
Sampling Systems
Each of the sewer monitoring stations was equipped with an
ejector type sampler located either directly in the sewer or in the
overflow chamber. It was found that when the sampler rested directly
on the floor of the sewer or chamber, sediment clogged the sampler
inlet ports or the nylon ball valve. When the ball valve did jam no
further sampling could be accomplished. Attachment of 1/2 to 1 inch
thick rubber pads to the bottom of the samplers raised them above the
sediment and partially alleviated the problem. Clogging still occurred
but much less frequently.
The individual vacuum type sampler at the Commerce and Humboldt
monitoring station required occasional backflushing of the tubing to
remove solids drawn up and trapped in the sample lines. Also, due to
the high vacuum necessary to deliver the sample to the monitor station,
rags, tissue, etc. , were occasionally drawn over the face of the sampler.
When this occurred, one or more samples were missed or were drawn
through the obstruction with the solids being filtered out. This problem
could be largely eliminated by installing automatic backflushing systems,
designed to function prior to collection of a sample. However, this
problem occurred infrequently, and the effect on the sampling program
was negligible.
Level Measurement System
The sewer monitoring stations were equipped with removable,
flexible bubbler tube type level sensing devices to monitor sewage flow
100
-------
rates. Plastic bubbler tubes were run through 6 inch diameter access
pipes from the monitor station to the sewer where they were connected
to metal bubbler tubes anchored to the sewer wall.
At three of the eight sewer monitoring stations locations the
plastic tubing was chewed through by rodents and had to be replaced.
To eliminate this problem, the flexible plastic tubing could be encased
with heavy, flexible, metal conduit for its entire length.
In addition to the foregoing the original instrumentation design
included the characterization of the level to flow signal. A motion
characterizer, which was a motion balance pneumatic device, was
connected to the bubbler line and accepted the same pneumatic signal
as the level DP cells. While this characterizer may function well under
continuous service, the intermittent operating conditions plus a some-
time hostile atmosphere gave many problems including the sticking of
moving parts. The use of a function generator operating from the DP
level signal may have eliminated many problems and much of the down
time encountered.
Another problem involved the anchorage of the bubbler piping
system. Metal tubing in sewer was torn from the anchorage during
severe rainstorms and sewer flows. Replacement of the piping utilized
extra heavy anchors.
Detention Tank Influent Flow Monitoring System
The sewer monitoring station at Commerce and Humboldt was
utilized to the greatest extent possible to monitor the tank influent flow
rate. The rate of flow was measured through a 60 inch flume by means
of a level sensing bubbler tube system. A level proportionate electrical
signal was transmitted to the tank where it was converted to flow rate
and recorded.
As previously mentioned in this Section, this bubbler system often
exhibited erratic behavior during the course of the study. Isolation of
various stages of the system for inspection eventually pointed to some
fault at the open end of the bubbler tube as being the most probable source
of trouble.
By the end of the study the conclusion was drawn that the problem
was physical in nature, and related to the position of the bubbler tube
in the sewer. Being positioned directly upstream of the flume the tube
was subject to the full force exerted by the high rates of flow probably
accounting for the resulting pressure drop through the bubbler system.
101
-------
A feasible answer would be to install a stilling well
adjacent to the sewer. The bubbler tube could then be placed
in the well, protected from the velocity of flow in the sewer.
Because of 1) the expected high cost of installing this system
2) the tenure of this project and 3) the existing backup sys-
tems, this theory was not followed to completion.
Because of the importance of maintaining an accurate in-
fluent flow record, it is recommended that at least one backup
measurement system be provided. To provide influent flow rate
data for use in this study, alternate systems including per
foot tank volumes converted to flow rates, and a bubbler system
to measure and record the head over the overflow weir were used.
Corrosive Gases
Sewer gases were found to vent to the monitoring stations
through the cast iron access pipes which carry the sampler,
pressure bell and bubbler lines. These gases proved to be cor-
rosive to the copper tubing, valves and fittings housed in the
monitoring stations.
Over a 4-year period the copper tubing, aside from dis-
coloring, did not appear to be adversely affected by the bases.
However, while the tubing appeared unaffected, gases entering
the monitoring stations from the sewers apparently caused the
brass fittings to crack thus requiring frequent replacement.
Until a cracked fitting could be located and replaced, nitrogen
was lost from the system. If the cracked fitting was on a
sampler line, the loss of pressure in the line could be suf-
ficient to present sampling and, if on a bubbler line, the loss
of pressure could result in the recording of inaccurate sewage
flow rate data.
To eliminate this problem, removable, air-tight covers
could be required over the station end of the access pipes.
The covers must be removable to permit access to the tubing.
Project tenure did not warrant further work in this area.
Dissolved Oxygen Analyzers
As previously discussed in this report, the D.O. analyzers
installed at the river monitoring stations provided little
useful data. The probes originally selected for use proved to
be completely unreliable even with concentrated maintenance ef-
forts afforded them. Use of the four probes was eventually dis-
continued due to the increasing severity of malfunctioning. It
appears that no probe available at the time of contract (1969)
would have been able to fulfill the requirements. Therefore
adjustments were made in data collection to fill the gap in the
desired information.
102
-------
During the project two analyzers of different manufacture were
installed at the Cherry Street river monitoring station and their performance
was evaluated. Both probes operated satisfactorily and provided accurate
data as long as they were well maintained. Unfortunately, maintenance
requirements of these probes included daily cleaning and accuracy
checks against manually collected samples. This high degree of maintenance
was required due to the adverse effect of a scum growth present in the
Milwaukee River during the summer months. It is possible that once
the river water quality is improved available D. O. probes may work
satisfactorily.
For use on this project, the dissolved oxygen analyzers were concluded
to be of minimal value because of the expense that would have been
involved to maintain the probes or obtain continuous D. O. data. The
personnel assigned to maintain the probes could better utilize the time
required for maintenance to collect manual samples for titration.
Manually collected samples, when handled by an individual who has
undergone the proper training, can provide dissolved oxygen data of
consistently reliable quality, however the data would not be of the
continuous variety.
The good results obtained during the evaluation of the two probes
late in the project indicates that in other installations the use of automatic
analyzers should be seriously considered, provided water quality permits.
Continuous D. O. readings make the task of characterizing the river water
quality an easier one but, again, the unique scum growth they were con-
fronted with minimized their worth to this particular study.
Summary
The field of process instrumentation is highly active and progressive.
Developments of new components could make some of the foregoing
suggestions and recommendations irrelevant in a very short time.
Perhaps the most important concept is that of reliability of systems
for monitoring stations such as those involved in these studies as
compared to comparable systems in other public works projects.
Equipment malfunctions at a wastewater treatment plant, a water treatment
plant or some manufacturing facilities may cause inconvenience, or
control problems, but the basic plant function is infrequently lost. This
is not the case with monitoring stations, intended to function as data
collection devices for a short period of time. Instrumentation failures
can mean loss of data which cannot be replaced.
103
-------
In monitoring station programs, suppliers must understand the
somewhat unusual nature of reliability requirements. Appropriate
planning to insure such reliability must be done and stringent guarantees
must be negotiated to assure an absolute minimum of station downtime.
Under natural hydrological situations judgment is used to a great extent;
therefore an unrealistic desire for accuracy can well be sacrificed for
a more dependable flow measurement system. This will enhance judgment
decisions even though some error may still be present. The extreme
variability of natural phenomena make extremely accurate flow measure-
ment unwarranted.
104
-------
SECTION VIII
SUMMARY AND ANALYSIS OF DATA
MILWAUKEE RIVER QUALITY
Data on water quality in the Milwaukee River have been developed
from the rivers sampling program. Sampling locations are identified
in Table 14 and are located schematically in Figure 26.
The following water quality parameters were measured during
the program:
Temperature
pH
Conductivity
Chemical Oxygen Demand
Biochemical Oxygen Demand
Dissolved Oxygen
Chlorides
Total Coliform
Fecal Coliform
Kjeldahl Nitrogen
Ammonia Nitrogen
Nitrite Nitrogen
Nitrate Nitrogen
Ortho Phosphate
Total Phosphate
Total Solids
Total Suspended Solids
Total Volatile Solids
A complete listing of the water quality data from the river survey
is contained in Appendices II and III. These Appendices have been
organized chronologically.
The total inventory of data on river quality developed during this
program, has been subjected to a statistical analysis. This analysis
has identified the following water quality conditions at each station:
(a) 5 year study period averages
(b) Season averages
(c) Hourly variations for pertinent quality parameters
105
-------
Table 14. RIVER STATIONS
Lake Michigan (Tunnel Intake) 40
Harbor (off Jones Island STP) 47
Buffalo 50
St. Paul* 52
State 53
Kilbourn 54
Cherry * 58
South Water 59
Humboldt 62
North Avenue Dam - Upstream * 66
North Avenue Dam - Downsteam* 64
North Avenue Bridge 65
Locust 68
Capitol 70
Esterbrook Railroad Bridge 72
Silver Spring 74
Menomonee River 81
Kinnickinnic River 82
*Fixed Monitor Stations
106
-------
FIGURE 26- MILWAUKEE RIVER
SAMPLING STATIONS
Z
2
o
in
107
-------
In addition, an analysis of the latter diurnal variations in conjunction
with rainfall data was performed to identify the effect of rainfall on BOD
and suspended solids variations in the river. Table 15 presents a matrix
which indicates the locations and periods covered by the river sampling
program from which the statistical analysis was made. The identification
of seasons as used in this analysis is shown on the table. The particular
months assigned to a particular season (e.g. , winter equals December,
January, February), were made on the basis of generally similar climatic
conditions (temperature, rainfall, etc. ) rather than the standard calendar
designation.
The overall raw water quality (i. e. , 5-year average) for each
station is illustrated in Tables 16 and 17. These tables provide a general
indication of concentration levels of the various constituents in the river.
The information should be interpreted cautiously, as the ranges and
variation in values at some stations are due to distortions introduced
by a limited sampling at a specific time of the year. Temperature
variations are a key to such conditions. Coliform data variations are
considered to reflect the effect of sample collection during wet versus
dry weather periods, rather than specific loads at certain places in the
river. The dissolved oxygen variations are related to a number of
factors, saturation changes from wide temperature variations in the
river, effect of upstream waste sources and storm overflows, and
algae production and respiration. From this table and from closer
inspection of available raw data, it is not possible to identify significant
variations in average quality between upstream and downstream stations.
However, this was not the primary purpose of the study, that purpose
being to determine the possible improvement in the water quality of the
River through the withholding of pollutants generated in the combined
sewer area of the City. The effect of the dam impoundment and estuary
can be seen in the dissolved oxygen surveys reviewed in Section IX
since impoundment of algae and pollutants can affect DO levels.
Seasonal raw water quality dataare summarized in Tables 18 and
19 for stations 50, 52, 58, and 62. These stations were selected on
the basis oi seasonal distribution and the data availability and were syn-
thesized therefrom. These tables should be interpreted in conjunction
with Table 15. Some variation is introduced due to the frequency
and distribution of sampling periods. The chloride variation is believed
to reflect the impact of road deicing during the winter. Concentrations
of chloride are generally two to three times greater during the winter
months. Coliform and BOD data are slightly greater during the summer
months. The dissolved oxygen variations are due to the same factors
mentioned above. Ortho phosphorous and ammonia concentrations varied
108
-------
Table 15. SEASONAL DISTRIBUTION OF
DATA COLLECTION
Station
Number
40
47
50
52
53
54
58
59
62
64
65
66
68
70
72
74
81
82
1967
A, W
Su, A, W
Su, A
Su, A
Su,A
Su, A
Su, A
1968 1969 1970
Su, A
S, Su, W
Su,A,W S,A,W S,A
Su,A,W S
Su,A,W S,Su, A S,A
S, Su, A, WS
A A
S
Su, W
S, W
S, W
1971 1972
S,Su
S, Su
S, Su
S, Su
S,Su
S, Su
S, Su
Su
S,Su
S, Su
S = Spring - March, April, May
Su = Summer - June, July, August
A = Autumn - September, October, November
W = Winter - December, January, February
109
-------
TABLE 16
FIVE YEAR AVERAGE - RIVER WATER QUALITY
Parameter
Temperature (°C)
pH
Conductivity (mmhos)
COD (mg/1)
BOD (mg/1)
Dissolved Oxygen (nag/I)9.9
Chlorides (mg/1)
Total Coliform(nu/ml)
Fecal Coliform(nu/ml)
Kjeldahl Nitrogen(nrg/l)
Ammonia Nitrogen(mg/l)
Nitrite Nitrogen(mg/l)
Nitrate Nitrogen(mg/l)
Ortho Phosphates (mg/1)
Total Phosphates (mg/1)1. 04
Total Solids (mg/1)
Suspended Solids (r
Volatile Solids (r
Station Number
40
16
8.2
;) 413
44
4.8
ig/l)9.9
20
il) 250
il) 55
«/l) .97
5/1) .37
5/1) .05
5/1) .38
5/1) -47
5/1)1.04
«/l)
tf/D
47
7.8
262
4.8
6,325
35
1.28
.11
.38
.88
1.22
50
9.1
8.6
52
6.2
7.6
52
3,050
130
1.68
1.12
1.01
52
12.1
8.0
575
42
9.6
6.3
55
7,070
300
2.54
1.29
.10
.63
1.28
2.63
930
45
27
53
14.5
8.1
570
62
6.4
3. 8
2.01
.86
1.03
54
17.3
7.9
37
8.0
6.1
25
33,125
473
1.55
.75
58
14.7
8.2
504
53
8.6
6.9
38
16,000
600
2.14
.84
.09
.65
1.15
2.47
436
38
22
59
7.9
452
6.7
3,120
104
1.41
.08
.37
1.20
1.47
-------
TABLE 17
FIVE YEAR AVERAGE - RIVER WATER QUALITY
Temperature ( C) 12.1
pH 8.4
Conductivity (mmhos) 480
COD (mg/1) 53
BOD (mg/1) 5.8
Dissolved Oxygen(mg/l) 9.4
Chlorides (mg/1) 34
Total Coliform(miAnl) 2,700
Fecal Coliform(nuAnl) 165
Kjeldahl Nitrogen(mg4l.57
Ammonia Nitrogen(mg/l) .71
Nitrite Nitrogen(mg/l) .09
Nitrate Nitrogen(mg/l) . 62
Ortho Phosphates(mg/l)1.01
Total Phosphates(mg/l)l. 47
Total Solids (mg/1) 380
Suspended Solids(mg/l) 23
Volatile Solids (mg/1) 14
64
8.5
985
180
17.6
68
3,000
17
6.25
1.70
5.2
6.
173
34
65
8.5
540
9.0
50
1,050
100
2.11
.10
.95
1.60
2.03
438
21
12
66
8.4
400
9.7
47
3,715
123
2.02
.70
.10
.61
1.34
1.93
450
25
13
68
12.8
8.1
46
6.1
7.7
27
35,800
450
1.77
1.0
1.22
70
10.5
8.2
46
7. 3
10.6
62
1,015
160
1.96
1.08
1.25
72
11.8
8.4
34
5.8
10.7
27.8
475
16
1.24
1.15
74
2.3
8.6
23
3.1
12.9
33.6
6
2
1.36
.90
1.73
81
8.0
412
5.2
742
50
1.24
.09
.51
1.07
1.33
82
7.8
350
5.6
1,700
110
1.14
.60
.07
.33
.83
1.06
-------
Parameter
Temperature (°C)
PH
Conductivity
COD (mg/1)
BOD (mg/1)
Dissolved Oxygen(mg/l)
Chlorides (mg/1)
Total Coliform(nu/ml)
Fecal -Coliform(nu/ml)
Kjehdahl Nitrogen(mg/l) 1.01
Ammonia Nitrogen(mg/l)
Ortho Phosphates
TABLE 18
SEASONAL RIVER WATER QUALITY
Station 5X>
Station 52
Spring
7.4
8.4
43
5.1
1 7.7
29
220
15
1) 1.01
I
.98
Summer
19.8
8.0
38
7.5
4.3
29
7,600
360
2.12
.72
.79
Autumn
7.4
8.4
44
5.1
7.7
29
220
14
1.01
.98
Winter
2.9
8.2
50
5.8
9.3
79
175
50
1.82
1.61
1.23
Spring
9.7
8.1
540
42
6.5
8.7
36.
3,150
160
2.19
1.08
.729
Summer
20.5
7.8
72
7.9
2.7
29
12,050
500
2.0
.68
.87
Autumn
15.4
7.9
28
5.3
2.7
44
7,350
340
2.05
1.09
1. 17
Winter
3.4
8.1
37
7.8
10.2
112
1,539
257
2.65
1.61
1.54
-------
u>
TABLE 19
SEASONAL RIVER WATER QUALITY
Station 58
Station 62
rai ante uej. ^7;
Temperature (°C)
pH
Conductivity
COD (mg/1)
BOD (mg/1)
Dissolved Oxygen(mg/l)
Chlorides (mg/1)
Total Coliform(nu/ml) 5
Fecal Coliform (mi/ml)
Kjeldahl Nitrogen (mg/1)
Ammonia Nitrogen(mg/l)
Nitrite Nitrogen(mg/l)
Nitrate Nitrogen(mg/l)
Ortho Phosphates(mg/l)
Total Phosphates(mg/l)
Total Solids (mg/1)
Suspended Solids (mg/1)
Volatile Solids (mg/1)
ring
9.8
8.3
45
5.1
9.5
32
,650
245
1.94
.91
.66
Summer
22
8.
52
6.6
5.1
29
2.15
.64
.89
Autumn
14.5
8.2
36
5.9
5.4
43
400
23
1.90
.87
1.16
Winter
1.7
8.5
28
3.8
12.1
56
1.58
.83
1.36
Spring
9.8
8.4
560
43
6.0
10.2
31
1,200
130
1.87
.60
.06
.53
.86
2.45
450
22
16
Summer
20.
8.2
410
65
6.6
7.9
27
6,200
190
1.39
.55
.11
.69
1.00
1.40
312
25
11
Autumn
12.4
8.3
59
5.1
8.3
35
2,250
146
1.38
.66
1.02
1.02
Winter
2.4
8.4
48
4.5
12.6
49
1,370
230
1.78
1.30
1.38
1.21
-------
from 0. 5 to 1. 5 mg/1, with the higher levels occuring during the colder
months. This variation could be the result of algal requirements during
the warmer periods of the year, where active growth tends to strip
these nutrients from solution. Concentrations of these nutrients are
quite high, and in a range well able to support significant levels of algal
growth. Both D. O. diurnal variation and visual observations confirm
the presence of significant levels of algae.
At certain times during the program, samples were collected
at frequent intervals to determine hourly variation in water quality.
Table 20 lists the stations, seasons, and years when such information
was collected. Actual average hourly data is listed in Appendix III,
where the raw data has been arranged by season and station. No normal
variation in quality over the course of a day is evident in the data, nor
are any trends indicated. With the exception of coliform concentrations,
which exhibit considerable variation, concentrations of contaminants
remained relatively constant over the course of a day.
The effect of rainfall and urban storm water overflow on general
background river water quality has been explored to a limited degree
by studying data developed from the general river sampling program.
The long term quality data, obtained from daily samples has not yielded
any definitive information. The primary reason appears to be that the
water quality in the Milwaukee River is controlled to a significant
extent by conditions within the upper reaches of its drainage area and
certainly in those areas substantially outside the test area. The occurance
of storm events in the drainage basin either unrelated to or poorly
correlated with similar events in the test area, the time of concentration
for such runoff to reach the test area, and the variability of river flows,
makes any analysis of river quality response quite complex and strongly
influenced by conditions which were beyond the scope of this project.
One analysis was made using data obtained from daily river quality
sampling program, where the sampling period coincided with a storm
event. Table 21 lists those events where such data exists. Figures
27 and 28 plot BOD and suspended solids variations for the period March
3 to 11, 1970. (Data for March 6th and 7th was not available due to
equipment malfunction or sample handling problems. ) This illustrated
graphically the water quality response which resulted from this storm
event. The storm event in question was a significant one, resulting
in a total accumulation of 0. 35 inches over a 10 hour period. Quality
measurements of BOD and suspended solids at Station 52 (St. Paul
Avenue), in the lower reaches of the river show concentrations well
above normally observed levels, and increase further at the time of the
114
-------
Table 20. PERIODS WHEN SAMPLES WERE COLLECTED
AT FREQUENT INTERVALS
TO STUDY HOURLY VARIATION IN RIVER QUALITY
Season Station Number Year
Spring
A 40 1970
B 52 1970,1972
C 58 1970,1972
D 59 1972
E 62 1972
F 65 1972
G 66 1970
Summe r
A 40 1972
B 52 1972
C 58 1969,1972
D 59 1972
E 62 1972
Autumn
A 46 1967
B 50 1967
C 52 1969,1970
D 54 1967
E 58 1967,1969,
1970
F 59 1970
G 62 1967, 1970
H 64 1969
115
-------
FIGURE 27- WVER QUALITY AT 3TA.52 VS RAINFALL
MADCH 1st-lift., l«70
100-
Ist 2nd r 3rd^4th 5th 6th ' 7th 8fh ' 9th 10th ' llth
D^TE IN MARCH-1970
NOTE: DATA FOR MARCH 6ft7 UNAVAILABLE
116
-------
FIGURE 28- RlVtR QUALITY AT STA 5£ VS RAINFALL
0 -r
MARCH Ut.-llth.. It70
IOO
1st 2nd ' 3rd 4th 5th 6th 7th ftth T 9th ' lOth Mlth
DATE IN MARCH-1970
NOTE: DATA FOR MARCH 687 UNAVAILABLE
117
-------
storm event on March 3rd. The precipitation shown on March 1st
is indicated by weather bureau records to be snow (1. 1 inches), and
followed a month where no significant precipitation was recorded. The
sharp increase in river flow between the 2nd and 3rd may in fact represent
either snow melt (average daily temperature between February 21st and
24th was above freezing) or precipitation in the upper drainage basin
of the Milwaukee River.
Table 21. RIVER SAMPLING AND STORM EVENT
ANALYSIS DATED
Station Dates
52 March 3 - 11, 1970
58 May 20 - 27, 1970
52 October 12 - 14, 1969
52 October 27 - 28, 1970
The March 1970 conditions may be compared with several others
for which comparable data is available. Comparable data is presented
in Figures 29 and 30.
From an inspection of such data, it does appear that storm
overflows in the immediate area can have a significant impact on river
quality. However, significant fluctuations also are observed during
periods when no overflows occur, and in some cases (the storm of
Sunday, May 24th, 1970 for example), no apparant effect can be discerned.
This behavior could be affected by the time of day and day of the week.
River quality in the vicinity of the combined sewer overflows in the lower
reaches of the Milwaukee River is influenced as much or more by
events occuring well upstream, of the area of study. Because of all of
these complexities, any effort to evaluate river quality responses to
storm overflows from the combined sewer system, using the river
quality data discussed in this section without the aid of a mathematical
computer model, would be highly speculative. Therefore, in order to
obtain a reliable quantitative assessment of the effect of urban storm
runoff on water quality in the lower reach of the Milwaukee River, the
mathematical model developed during this program must be relied upon
to characterize water quality responses. The results of this modeling
effort are discussed in a later section of the report.
The Milwaukee River, in the area being considered, is subject
to wide variations in flow which can occur over relatively short intervals.
Table 22 presents a condensed summary of flow data for the lower
Milwaukee River. Some 50 years of record indicate a discharge of 45
cfs to be exceeded 95 percent of the time; 172 c fs at 50 percent of the
118
-------
FIGURE 29-
MILWAUKEE RIVER QUALITY CHANGES
DUE TO RAINFALL
I5tr 16th 17th
DATE IN
AT
18th 19th 20th
MARCH-1970
STA-66
20th 21st ?2nd 23rd 24th 25th 2€th 27th
DATE IN MAY - 1970-
AT STA-58
-------
FIGURE 30-
MILWAUKEE RIVER QUALITY CHANGES
DUE TO RAINFALL
12th 13th 14th
-DATE IN OCTOBER-1970
AT STA-52
27th 28th 29th
-DATE IN OCTOBER-1970
AT STA-52
120
-------
time; and 1,400 cfs 5 percent of the time. Bi-hourly flows in excess of
3, 000 cfs are indicated.
In general, river quality data indicates that the Milwaukee River
in the study area contains relatively high levels of polluting material,
which are evidentally attributable to both storm water discharges from
sewer overflows, and upstream runoff which enters the lower Milwaukee
River estuary. Typically, the Milwaukee River exhibits the following
characteristic concentrations:
Temperature
pH
COD
BOD
Chlorides
Total Nitrogen
Total Phosphorus
Suspended Solids
Total Coliform
= 0
- 22°C
= 7. 5 - 8.5
= 35-60 mg/1
= 5-10 mg/1
= 20-50 mg/1
= 1-2 mg/1
= 1-2 mg/1
= 20-50 mg/1
= 200 - 40, 000 nu/ml
Table 22. FLOW RECORD * - MILWAUKEE RIVER
Year
1967
1968
1969
1970
Average
337
255
386
243
Minimum
52
58
73
43
M aximum
2,300
1,620
3, 150
1,900
Average discharge - 56 years, 379 cfs.
*Data reported in cubic feet per second (cfs).
SEWAGE - DRY WEATHER QUALITY
Data taken at intervals during the program have been analyzed
to identify quality characteristics of the sewage in the project area
during periods of dry weather. A tabulation of all data collected is
presented in Appendix IV.
Table 23 identifies those periods during the program in which
suitable dry weather data was obtained. On many occasions during the
program, sampling efforts intended to produce dry weather quality were
not usable for that purpose because of storm events which occurred
during scheduled"dry weather" surveys.
121
-------
Table 23. DATES OF DRY WEATHER
SEWAGE SAMPLES
Sampling Dates of Dry Weather Samples
Stations 1969 1970 1972
49 June 2-13 January 14 to
February 3
July 13-14
July 30 to
August 11
December 17-18
21 February 11-23 March 24 to
August 9*
These data represent daily average from 24-hour composite. All
other data based on 1-to 6-hour composites.
122
-------
The entire body of dry weather quality data was analyzed, and a
summary of average, maximum, and minimum concentrations observed
in the sampling program is presented in Table 24. Average concentrations
appear somewhat low for "typical" domestic sewage, and rather wide
variation between maximum and minimum values are indicated. Some
of the extreme values may be due to the effect of wet weather conditions,
such as residual runoff or effects of limited infiltration. This explanation,
however, is speculative, and no reliable explanation of the range of
concentrations observed is available. Although some of the variation
is also due to anticipated seasonal and daily variation (including weekday
and weekend), these do not account for the observed variations. Some
of the higher values, especially in chlorides, are the result of snow
melt runoff during winter thaws in otherwise dry •weather conditions.
Because of the large number of individual observations for most
parameters, the average values reported are considered to be represent-
ative of "typical" dry weather quality in this system.
Seasonal variations in dry weather quality were examined by
sorting data by season, and analyzing data to determine mean concentrations
and standard deviation for each parameter. This analysis is summarized
in Table 25.
Spring seasonal averages are somewhat limited by the relatively
small number of observations, and must be evaluated with this in mind.
Both winter and summer data have a sufficient number of individual
data observations to provide a reliable comparison. A seasonal average
for the fall months is not presented because of data limitations. Efforts
to supplement the deficiency of data from prior sampling programs,
during 1972 were frustrated by the extreme wet weather conditions
during this period.
Contaminant concentrations during the winter are higher than
summer values for all parameters, except phosphate and total solids.
The magnitude of the standard deviations from mean values attests to
the rather wide variability between individual observations. Spring
seasonal data shows the highest concentration of all parameters, however,
the number of observations represented compared with those for winter
and summer averages suggests caution in drawing firm conclusions.
Diurnal quality fluctuations were explored using winter and summer
seasonal data breakdowns. Data was sorted on the basis of hour of the
day during which the observation was made. Table 26 lists average
values for each quality parameter, each hour of the day for all winter
data. Table 27 does the same for summer observations. Much of the
quality data does not show any clear-cut diurnal pattern, beyond generally
123
-------
Table 24. DRY WEATHER SEWAGE QUALITY
K3
Parameter
pH
COD (mg/1)
BOD 5-day (mg/1)
Chloride (mg/1) Cl
Total Coliform (nu/ml)
Fecal Coliform (nu/ml)
Total Kjeldahl Nitrogen
(mg/l-N)
Ammonia Nitrogen
(mg/l-N)
Nitrate (NO3)(mg/l-N)
Nitrite (NO2)(mg/1-N)
Orthophosphate (mg/l-PO4)
Total Phosphate (mg/l-PO4)
Suspended Solids (mg/1)
Volatile Suspended Solids
(mg/1)
Number of
Observations
Average
Concentration
Maximum
Concentration
Minimum
Concentration
433
382
427
98
338
316
161
160
19
2
159
158
149
7.6
238
112
141
930,000
38,000
20. 7
11.7
0.5
0. 01
10.7
18.4
150
8.7
730
323
826
-
540, 000
47
31. 2
1.6
0. 01
39
60
900
6.7
35
17
3
11,000
400
5.6
1. 0
0. 1
0. 01
1.0
0. 5
5
148
115
700
-------
Ni
Ul
TABLE 25
DRY WEATHER CONDITIONS - AVERAGE SEWAGE QUALITY
Parameter Average
PH
BOD(mgA)
COD (mg/1)
Chlorides (mg/1)
Coliforms(nu/ml)
Total 243
Fecal 21
Nitrogen frng/l-N)
Kjeldahl
Ammonia
Nitrite
Nitrate
T7S
169
309
162
,000
,546
26.7
17.6
-
.48
Spring
Std.DeV. f
.37
58
71
29
87,475
8,248
2.8
2.4
-
.41
Events
12
13
13
13
13
13
13
13
-
13
Summer
Average Std.Dev. # Events
777
94
206
51
261,884
19,887
18.7
9.8
.01
.39
.37
41
104
33.7
285,152*
39,074
6.2
3.0
.000003
.25
2b6
251
252
7
159
137
64
64
2
6
Average
776"
133
301
146
1,629,270 1
54,899
21.4
12.2
-
-
Winter
Std.Dev.
.28
67
154
147*
,628,890*
69,494*
9.2
6.1
-
-
# Events
165
163
117
78
166
166
84
83
-
-
Phoaphate(mg/l-PC>4)
Total
Ortho
Solids (mg/1)
Total
Suspended
Volatile
Suspended
33.4
22.0
845
155
110
10.7
4.6
97.5
55.4
33.6
March, April
13
13
13
13
13
, May
19.4
11.5
471
122*
87
June
12.2
7.2
76.7
135*
103*
, July, August
81
82
6
51
51
15.3
8.4
128
145
123
December,
8.1
4.7
0
91.2
70.8
January,
81
82
1
85
84
February
•Standard Deviation greater than average
-------
TABLE 26
CRY WEATHER SEWAGE QUALITY
K)
DIRUNAL VARIATION
Hour
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
pH
7.4
7.4
7.5
7.6
7.7
7.6
7.7
7.5
7.7
7.6
7.5
7.5
7.7
7.5
7.6
7.6
7.5
7.4
7.6
7.5
7.5
7.5
7.5
7.7
COD
mg/!
193
182
178
167
116
174
57
239
245
304
265
327
322
393
344
376
371
374
323
427
289
320
290
277
BOD
L mg/1
112.2
156.
122.
78.
70.
86.
44.
63.
89.
110.
144,
143.
157.
163.
146.
174.
149.
151.
125.
135.
130.
158.
142.
143.
3
8
2
5
8
9
9
8
0
1
7
1
8
6
3
2
1
1
3
0
7
0
1
Cl
TOOL
Fed
irgTT nu/mlx 1000mg/l
1,005 22
83
90
105
80
147
747
79
95
89
170
132
72
209
259
54
180
221
119
89
92
574
651
870
105
835
50
255
324
1,693
2,041
1,916
1,536
2,004
1,939
2,481
2,192
2,242
1,663
2,204
2,136
3,081
1,914
1,106
45
16
67
3
33
10
3
9
75
52
51
51
97
76
58
53
74
93
75
70
76
50
21
- WINTER SAMPLES
Kjeld.
mg,
16.
14.
17.
10.
12.
15.
28.
24.
29.
23.
21.
30.
18.
20.
33.
21.
19.
26.
19.
25.
11
83
55
21
64
68
68
84
35
79
86
41
24
47
66
46
62
73
84
82
84
NH--N
mg/1
11.00
8.13
9.46
5.50
7.33
11.40
17.75
15.95
17.25
12.78
13.06
13.60
9.61
10.71
17.30
12.74
10.79
15.89
10.89
14.57
Ortho.
mg/1
6.40
3.90
6.34
1.35
3.62
8.00
12.33
7.65
14.45
6.86
8.67
17.80
7.87
6.98
15.95
7.24
9.32
13.50
9.22
9.72
TPO
4-
mg/I
13.08
8.55
12.41
4.00
6.90
17.00
21.32
14.30
25.60
12/50
16.64
35.60
15.90
10.70
31.00
14.51
17.07
24.55
17.92
18.31
TSS
mg/1
91
92
77
92
113
131
153
131
168
113
123
224
211
165
175
140
175
138
159
TVS
mg/1
78
79
67
77
88
116
131
126
147
86
114
181
166
155
145
117
160
121
137
-------
TABLE 27
DRY WEATHER SEWAGE QUALITY
DIRUNAL VARIATION
Hour
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
7.6
7.7
7.6
7.6
7.6
7.8
7.7
7.7
7.7
7.8
7.7
7.7
7.7
7.6
7.6
7.7
7.7
7.5
7.6
7.6
7.6
7.6
7.7
7.6
COD
mg/1
247
158
176
169
182
163
100
85
170
232
288
224
191
217
211
251
253
229
232
270
249
211
233
203
BOD TCOL PeCl
mg/1 nu/mlxiooo mg/T
94.0 108 8
61.0
136.6
84.2
56.9
51.9
48.3
50.2
65.3
104. r
116.0
110.1
95.3
93.1
104.7
105.0
93.7
104.6
133.7
123.0
120.-
99.9
115.2
94.6
105
283
99
149
310
152
190
240
373
378
202
175
334
255
271
236
166
287
384
186
268
403
15
9
22
3
15
7
12
22
21
25
11
17
25
20
20
14
11
19
94
12
19
19
- SUMMER
Kjeld.
mg/l-N
21.94
12.81
14.17
12.32
25.01
21.18
22.62
18.47
18.89
16.89
16.80
15.50
17.56
24.74
17.60
17.70
20.40
SAMPLES
_NH.-N
mg/iTR
9.87
7.07
8.69
7.49
13.44
11.21
12.63
10.93
9.56
8.63
8.60
7.97
9.53
9.85
10.00
9.55
10.20
Ortho.
TPO .
mg/l-PO4 me/f
6.97
3.73
6.34
5.33
14.44
7.62
14.55
12.70
23.20
14.83
7.30
10.67
14.39
16.52
23.00
16.50
5.50
13.70
4.60
9.60
8.60
24.06
13.02
24.27
21.13
33.73
28.53
8.00
19.30
26.86
29.40
40.60
24.00
29.20
TSS
mg/1
8
100
202
55
38
56
176
160
100
92
93
162
39
56
367
164
62
84
125
TVS
mg/
2
80
132
37
24
22
118
133
42
75
65
102
33
40
280
110
34
55
91
-------
lower concentrations during early morning hours. The plots, Figures
31 and 32 show those parameters which show the most evident diurnal
variations. Total phosphate shows a distinct pattern in both winter and
summer data, with early evening values approximately three times early morn-
ing concentrations. Total coliforms show a clear diurnal variation in winter,
but not in summer. This is also the case with nitrogen and suspended
solids data.
Chloride concentrations in winter dry weather flow suggest an
increase from approximately 100 mg/1 during early morning hours
until midday - to about twice this concentration during afternoon and
evening hours. Since the concentrations begin to increase much later
in the day than is observed with the other constitutents, the effect is more
probably due to snow and ice melt than to increased sanitary discharges
to the sewers. Based on a dry weather flow of about 2 mgd and a chloride
increase of 100 mg/1 over background concentrations occurring between
2 PM and 10 PM, a daily flushing of about 600 pounds chlorides per day
per square mile during dry weather is indicated, which is in excess of
the amount of chlorides present in the sewage during non-winter months.
The data shows a value at 10 AM which reaches 747 mg/1 Cl. This
appears to be an abnormal condition occasioned by a measurement during
a heavy thaw, and is considered to be not typical of dry weather conditions.
However, a 1 to 2 hour runoff under these conditions would contribute
approximately 500 pounds Cl per square mile.
A further analysis of dry weather quality data was made to determine
whether variations in concentration were present as a function of time
elapsed since the antecedent storm event. This data is presented in
Figures 33 and 34. If we assume that the low concentrations at time zero
in several cases are due to the presence of storm runoff at the time the
"dry weather" sample was taken, then the data show no significant
effect on dry weather quality as a result of an antecedent storm. This
data also suggests that no significant infiltration occurs in the test area.
Where an appreciable degree of infiltration is present, one would
expect to see a gradual increase in concentrations as the interval to
the antecedent storm increases. Kjeldahl nitrogen concentrations show
such a pattern, but other quality parameters do not support the actual
presence of such an effect. While some of the data imply a decline in
concentration with time, the observed differences are well within the
normal variations observed for dry weather quality.
128
-------
FIGURE 31- HOURLY VARIATION - DRY WEATHER SEWAGE QUALITY
WINTER DATA
40O
I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24
TIME (HOURS) 0= 12:00 MIDNIGHT
129
-------
FIGURE 32 - HOURLY VARIATION-DRY WEATHER SEWAGE QUALITY
SUMMER DATA
400
300
200
IOO
ISO
100
OCHEMICAL OXYGEN
a.
r
30
20
10
600
400
200
\
I ! i I I
CHEMICAL OXYGEN DEMAND
10 12 14 16 18 2O 22 24
TOTAL
PHOSPHATE
ORTHO
, NITROGEN
/1-TOTAL
/ KJ6LDAHL
/•v
SUSPENDED SOLIDS
2 4 68 10 12 14 6 18 20 22 24
TIME (HOURS) 0= 12:00 MIDNIGHT
130
-------
2OO
FIGURE 33- CONSTITUENT VS TIME SINCE LAST STORM
M. C KYd I MN (AND
TIME SINCE LAST STORM
131
-------
FIGURE 34- CONSTITUENT VS TWE SINCE LAST STORM
TIME SINCE LAST STORM
132
-------
SEWAGE - DRY WEATHER FLOW
Flow rates were monitored at sewer station 21 at various times
throughout the program in periods of dry weather, in order to establish
normal dry weather flow conditions.
The flow data developed indicates average dry weather flows
in the test area range between approximately 1. 7 and 2. 3 mgd. Sewage
flows appear to remain relatively uniform over a period extending from
7 or 8 AM until 8 to 10 PM. Flows then decrease to a minimum which
is reached at 1 or 2 AM, and remain relatively uniform until approximately
5 AM. Maximum flows are approximately 100% greater than minimum
flows during a 24 hour period. However, wide variation between maximum
and minimum within hourly flows has not been observed.
Figure 35 plots schematic representations of daily dry weather
flow variations based on data collected during fall months (September,
October, November) and spring months (March, April, May). The fall
average shown is slightly greater than that observed in spring, although
the difference is small; approximately 0. 5 mgd. Winter dry weather
data is not presented here, as it proved to be influenced in an erratic
manner by snow melt under otherwise dry weather conditions. An
analysis of data indicates that snow melt conditions increased "dry
weather" flows by between 0. 3 and a maximum of 1. 5 mgd and is dependent
upon the amount of snow cover, temperature and ice control conditions.
One set of dry weather data, that for the sampling date September
18, 1970, is plotted in Figure 36. This illustrates in detail the data
presented in schematic form in the previous figure. In addition, this
particular set of data was commenced immediately after a significant
storm event, and illustrates a relatively rapid return to essentially
dry weather flow conditions.
Variations in dry weather flow, either due to daily or seasonal
fluctuations are not considered to be a significant factor in the analysis
and modeling of the storm overflows in the test area. Observed variations
are relatively small, particularly so when compared with the combined
flows required to cause an overflow. Although winter flows are somewhat
higher and more variable, the significant storms, i. e. , those generating
significant overflows are relatively rare occurrances in winter.
RAINFALL
The U. S. Weather Bureau's narrative climatological summary
for Milwaukee provides the following comment:
133
-------
FIGURE 35 - AVERAGE DRY WEATHER SEWER FLOWS FOR MARCH-MAY 1972
FLOW RATE
(mgd)
C
^--_
^-^™
"
- .'^">
^
/^
FLOW M.G.O.
— -^
"""*
^^
\
3 6 9 12 15 18 21 24
HOURS nm ... HOURS am.
TIME OF DAY
AVERAGE DRY WEATHER SEWER FLOWS FOR SEPT.-NOV.
FLOW RATE
(mgd) c
^
0 3 i
. -HOUR
r~
f
J
FLOW
U.G.D.
• ii ,
• . -
K--~^^
^
> 9 12 15 18 21 24
S am. .. - . HOURS Dm.
TIME OF DAY
-------
FIGURE 36- DRY WEATHER SEWER FLOWS FOLLOWING RAINFALL FOR SEPT. 17-18 1970
Ul
. I
.2
RAINFALL ,.
(inches)
.3
.4-
.5
FLOW RATE
(mgd)
\
\
12
15
RAIf IFALL
FLOW, MGD
18 21
M
SEPT. 17-
THURSDAY
DATE
12
SEPT 18
FRIDAY
15
18
21
M 3
- -SEPT. 19-
-------
"The average annual precipitation is about 30 inches. About
two-thirds of the annual amount occurs during the growing
season. Since 1841, the wettest year was 1876 with 50. 36
inches, and the driest year was 1901 with 18. 69 inches.
The long-term average annual snowfall is about 46 inches,
but it varies considerably from season to season. During
the winter of 1885-86, 109 inches were measured, while
in 1884-85, the snowfall totaled only 11 inches. The 1967-68
season produced 12 inches.
Thuderstorms occur less frequently and with less severity
in the Milwaukee area than in areas to the south and west.
Hail size is generally 1/2 inch or less, although it has been
noted as large as 2 inches in diameter with unusually
severe storms. The maximum rainfall which has occurred
in a 24-hour period is 5.76 inches in June, 1917. As much
as 0.79 inches has fallen in 5 minutes, 1. 11 inches in
10 minutes, 1. 34 inches in 15 minutes, 1. 86 inches in
30 minutes, and 2. 25 inches in 1 hour. "
Table 28 reproduces U. S. Weather bureau records tabulating
monthly precipitation at the airport (Station 1) since 1932.
The storm detention tank model developed during this project is
designed to operate on an input of hourly rainfall records. Punched
computer cards were available from the U. S. Weather Bureau
providing such data at the airport station located south of the City.
In addition to the above long term record, two rain gauges were
maintained by the City of Milwaukee during the program, one in the
project area, and one approximately two miles south of the area. Data
from these guages were converted to punched cards using the same format
as U. S. Weather Bureau data.
Table 29 compares records of the three sources of rain data,
listing the volume of rain recorded each month. The pattern of variation
from month to month and from year to year are similar, however distinct
differences in individual months and in annual accumulations are evident.
Routine differences between gauges are to be expected - even for gauges
located relatively close to each other. Table 30 lists annual accumulations
at two gauges maintained in the Milwaukee area by the U. S. Weather
Bureau. Annual differences in the order of 10 to 15% are common.
136
-------
TABLE 28
PRECIPITATION RECORD
MILWAUKEE
TOTAL PRECIPITATION
TOTAL SNOWFALL
u>
Year
"1932
1933
1934
1935
1936
1937
1938
1939
1940
M941
1942
1943
1944
1945
1946
1947
1949
1949
• 1950
1951
1952
"1953
1954
"1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1 = 71
RECORD
MEAN
[jan
1.6
1.3
l.or
2.5r
2.5*
3. 12
4. 6f
1.60
1.57
2.50
1.16
2.15
1.40
0.31
1.97
2.26
1 .07
2.59
2.17
2.38
2.08
1.16
0.92
0.6?
0.57
0.88
1.41
2.48
4.04
0.31
2.48
0.66
1.11
3.31
2.06
1.49
0.9ft
1.81
0.41
1.37
1.B3
I Fob
1 0.94
1 1.43
0.55
2.24
2.32
1.7
3.3
2.24
1.33
0, 3
0. 0
0. 6
1. 9
1. 0
0.88
0.29
1.68
1.74
1.39
1.67
0.82
1.62
1.31
1.32
1.43
0.96
0.15
1.98
3.05
1.22
2.04
0.42
0.41
1.04
1.27
1.31
0.56
0.05
0. 3
2.50
1.55
[Mar
2.2C
3.91
1.97
1.98
0.67
1.74
3.29
1 .54
2.07
1.82
1.46
2.48
2.46
1.40
2.88
1.73
3.59
2,57
2.50
3.33
3.67
1'. 18
1.6
1.0
2.3
r.s
0.4
3.03
3.80
3.80
1.69
2.20
3.05
3.61
3.6
1.35
0.31
1 .05
1.62
2.83
2.41
"Apr
0.5<
2 .91
1 .53
3.0"
2.3 =
4.80
0.97
2.81
2.96
1.93
0.11
0.99
3.74
2.89
0,94
3.68
1.91
1.38
3.58
4.91
2.95
2.81
3.27
2.45
4.14
2.70
1.84
3.29
2.92
3.89
1.49
2.54
3.81
3.47
2.67
2.70
2.90
3.42
2.71
1.31
2.71
iMay
1.5
9.5!
2.73
2.i:
2.55
2.70
3.73
1.40
3.80
3.03
4.49
2.88
2.33
5.27
2.14
4.35
4.05
1.72
2.04
3.87
2.86
1.77
1.83
4.29
4.55
3.82
2.07
1.28
4.27
1.25
2.17
1.95
2.57
2.12
2.00
1.80
3.2>
3.05
3.41
0.90
3.19
June
1.6"
2.4'
2.32
4. 34
1.93
2.64
6.93
3.50
7.5
3.4
4.26
2.3
3.42
2.8
2.8
3,98
3.19
3.79
5.1
2.97
4.03
2.65
1.28
4.58
3.87
4.01
1.71
1.67
3.28
1.53
1.33
1.50
1.70
0.85
1.68
7.38
7.79
7.53
3.52
2.67
3.53
July
3.1
4.5
1.10
3.59
0.28
3.06
2.70
0.5
0.9
2.93
3.58
1.54
2.77
2.65
0,95
2.17
2.16
3.46
6.07
3,12
0.69
2.78
5.13
2,10
5.37
1.50
1,02
6,62
3.50
2.91
3.74
2.36
7.66
2.64
3.32
1.35
3.59
6,61
1.93
2.60
2.90
Aug
1.9
1.7
1.43
3.08
5,92
0,80
6,47
5,03
6,68
1.29
4.14
2.31
1.54
4.07
1.63
1.58
0.46
1.06
3.29
2.56
3.59
4.34
3,86
3,62
2,96
2,03
1,71
3.47
7.07
2.35
1.96
2,48
2.62
6.15
3.27
1.23
2.59
0.53
0.64
2,28
2,79
Sept
0.9C
2.51
4.33
1.12
5.59
1.14
6.12
1.53
0,55
9.67
3.43
0.37
3.05
6.27
1.28
6.03
1.24
1.68
1.75
2.75
0.36
1.65
2.78
2.36
0.30
0.88
2.65
2,31
3,25
9,41
1.49
1.78
1.74
6.65
0.48
1.69
3.36
2.18
6.94
1.30
3.12
Oct
4.8
2.8
2.28
1.37
3.77
1.83
0.76
2.43
1.48
2.86
2.44
0,83
0,2'
0,76
1.79
1.85
0.33
1.62
0.55
4, 42
0.17
0.46
3.18
3.57
0.15
1.34
3,24
6.4!
3.06
2.75
2.14
0.34
0.17
2.68
1.76
2.70
0.94
4.48
2.09
1.90
2.28
Nov
0.66
1.03
6.56
3.43
0.34
0.85
1,86
0.33
2.60
0.93
3.27
3.15
1.54
2.34
2.08
2.62
2.44
0.62
1.60
1.99
3.37
0.58
1.06
0.87
1.62
2.88
3.37
2,08
2,12
2.37
0.81
2.17
2.29
2.02
2.70
1.52
2.56
1.14
2.03
2.45
1.99
| Dec
2.1
1.1
1.22
1.42
2.14
1.4J
1.10
0.46
0.95
1.29
2.55
0.99
1.14
1.47
1.54
1.7!
2.50
2.27
2.59
2.26
2.10
1.87
2.64
1.09
1.03
.36
.34
.85
.02
.55
.70
.98
.73
.31
.33
.65
.18
3.02
4.34
1.74
Annual
22.35
35.52
29.02
30.41
30.35
25.82
41.86
23.38
32.64
32.50
32.09
20.76
25.37
31.66
20,89
32.46
24.62
24.72
32.64
36.43
32.69
22.87
35.9-1
27,90
26.35
24.95
20. 17
37.68
40.71
32.81
21.91
19.10
28.18
38.49
27.13
25.85
31.51
33.05
2". 85
26,45
30.04
a lues above (not adjusted fo
nd 1871 for precipitat;
Season
1932-33
1934-35
1935-36
1936-37
1937-38
1939-40
#1940-41
1941-42
1942-43
1943-44
1944-45
1945-46
1946-47
1947-48
19*6-49
*1949r50
1950-51
1951-52
1954-55
1955-56
1956-57
1957-58
1958-59
1959-60
1960-61
1961-62
1962-63
1963-64
1964-63
1965-66
1966-67
1967-68
1966-69
1969-70
1971-72
July|Aug.|SeptJOct |Nov.|Dec.
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
O.Oj 0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0,0
0,0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
(1.0
0,0
0.0
0.0
0,0
0.0
0.0
0,0
0,0
0,0
0.0
0,0
0.0
0.0
0,0
0.0
0,0
0.0
0.0
0.0
0.0
0.0
0.0
0,0
0.0
0.0
0,0
0,0
0,0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
T
0.0
O.Oi
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
T
0.0
0.0
0.0
0.0
0.0
0,0
0.0
0.0
0,0
0.0
T
T
T
T
T
T
0.0
0.0
T
0.0
0.0
0.0
T
0.0
0.0
0.0
0,0
0.0
T
0.0
0.0
T
0.0
T
0.2
T
0.1
0.0
T
T
0.0
0.8
0.0
T
0.0
0.6
T
.0
.9
.0
T
1 .0
.1
.1
.6
. 8
. 5
, 1
.0
7.9
8. 1
9.7
3.5
4.5
0.4
Jan
1.5
10.2
24.8
0.7
7.8
9.5
1.3 7.4
8 .5
6.1
T
12.5
11.9
6.8
5.2
.Oj 2.8
. 1
5.8
12.4
1.3
7.6
2.4
0.3
0.6
2.1
9.3
2.8
2.4
0.9
T
1.4
T
2.0
0.4
0.3
0.7
6.1
fc. I
20. 1
30.7
17.6
5.2
10.6
6.0
6.9
14.3
2.3
7.7
6.5
12.8
8.1
14.5
9.9
1 .2
11.6
14.9
2.7
8.1
26.4
3.0
6.4
3.5
26.3
16.2
13.7
3.8
27.3
15.7
2.8
4.4
1K2
16.7
27.5
19.4
3.9
22.1
8.1
Feb.
13.8
17.0
23.6
1.9
2.9
14.6
2.7
6.4
3.5
9.3
6.7
7.8
5.2
6. 1
Mar| Apr.|May|june|Total
12. « 0.8
9.2 0.4
l.d 13.0
13.2
3.6
15.1
4.3
1.8
9.4
7.1
T
T
0,3
2.5
T
T
T
T
7
5.4 0.0
9.4| 0.9
0.0
3.2
0.0
0.0
0.0
T
0.0
0.0
0.0
0.0
T
0.0
T
12. « T . 0.0
•.5 3.7 0.3' 0.0
8,6
10.3
6.7
9.9
3.0
1 .6
10.8
34.0
2.4
22.2
6.4
3.8 5.7
23.6
24.6
13.1
4.6
11.1
6.0
10.1
7.7
27.1
3.5
0.7
2.0
1 4 . 9i 5 . :
;?•?
4.4
11.7
9.C
4.6
14.3
14,3
O.B
0,0
0.1
0.3
T
1.5
1 .6
14. 3J 7.0
11.2
7.1
19.8
26.7
2. a
7.4
1.2
6.2
10.7
4.0
T
T
4.1
1.3
T
0.4
0.0
5.2
0.0
T
0.0
0.0
0.0
0.0
0.0
0.0
0.4
0.0
0.0
T
0.0
0.0
0.1
T
0.0
T
0.0
0.0
0.0
0.0
0.0
o.o
0.0
0.0
0.0
0.0
0.0
n.o
0.0
0.0
0.0
0.0
o.o
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0,0
0.0
37.3
46.1
74.2
20.2
21.3
42.3
33,7
27.9
54,3
21.0
26.*
30, 1
46,7
49.1
30.0
41,7
79.3
17,6
39.1
33.7
34,4
31.9
63.3
93.3
32.9
69.8
29. 1
42.1
74.0
51.0
59.3
12.1
29.9
39.5
• ted in the Stati
table) are means for the
Location table. Data
r 195J , otherwise Cro
are from (ity Office
Airport !ocations .
19'*1 and for the period July 1950 through
SOURCE-US. WEATHER BUREAU.
-------
Table 29. COMPARISON OF RAIN GAUGE RECORDS
INCHES OF RAIN
Year
1968
(Total
Total
1969
Month
J
F
M
A
M
J
J
A
S
O
N
D
7 Mos.
Year
J
F
M
A
M
J
J
A
S
O
N
D
STA 1
Airport Sta.
U.S. Weather
Bureau
0. 98
0. 56
0. 31
2. 90
3. 28
7.79
3. 59
2. 59
3. 36
0. 94
2. 56
2. 65
) (23.48)
31. 51
1. 83
0. 05
1. 05
3.42
3. 05
7.53
6. 61
0. 53
2. 18
4.48
1. 14
1. 18
STA 4
STA 2
Proj-Area Broad-way
Gauge
6.61
3. 10
2. 29
2.41
1. 03
2. 58
2.68
(20.70)
1.87
0
0. 98
3. 86
3. 70
5. 25
4. 96
0. 26
0. 59
3. 88
0. 97
1. 19
Gauge
2. 80
2. 85
5. 85
3. 26
3. 84
3. 54
1. 22
2. 69
1.92
(22. 32)
0. 90
0
0. 80
4. 10
3. 97
6. 11
3. 81
0.41
0. 24
5. 08
1. 14
0. 08
Inches
Snow at
STA 1
4.6
3. 5
1. 2
0.4
0. 3
11. 6
11. 1
0. 7
6.2
0.7
14. 9
Total
Year
33. 05
27. 51
26. 64
138
-------
Year Month
Table 29. (Continued)
STA 1 STA 4 STA 2
Snow at STA 1
1970
Total
1971
Total
J
F
M
A
M
J
J
A
S
O
N
D
Year
J
F
M
A
M
J
J
A
S
O
N
D
Year
0.41
0. 13
1. 62
2. 71
3.41
3. 92
1. 93
0. 64
6. 94
2. 09
2. 03
3. 02
28. 85
1. 37
2. 50
2. 83
1. 31
0. 90
2. 67
2. 60
2. 28
1. 30
1. 90
2.45
4. 34
26.45
0
0. 06
4.41
2.72
1.72
0. 60
6.44
1.89
1.73
1. 96
21. 53
2.44
7.23
6.48
5.08
1.69
2.71
2. 63
2.42
1. 10
2. 14
3. 17
4. 30
42.39
0
1. 09
0. 21
3. 27
3. 68
1. 97
0. 63
7. 15
2. 18
1. 88
1. 54
23. 60
0. 63
5.72
7.55
0. 83
2. 04
3. 87
2. 25
2. 63
1.46
2. 21
2. 83
4. 76
36.78
6. 0
2. 0
10. 7
5.2
0. 6
19. 6
15. 8
2. 5
18. 1
0.7
6. 1
2.7
139
-------
Year Month
Table 29. (Continued)
STA 1
STA 4
STA 2
Snow at STA 1
1972
Total
J
F
M
A
M
J
J
A
S
O
0. 75
0. 86
2. 57
2.76
2. 33
3. 33
4. 60
4. 82
7. 57
29. 59*
0. 59
0. 26
2. 03
1.46
1. 10
2.79
2. 83
4. 56
4. 33
2.46
19.95**
0.48
0. 15
2. 29
2. 68
1.23
3.47
3. 19
5. 07
6. 07
3. 31
24. 63**
*First 9 months of data included
"^First 10 months of data included
140
-------
Table 30. COMPARISON OF ANNUAL RAINFALL
AT U. S. WEATHER BUREAU STATIONS IN MILWAUKEE
Rainfall (Inches)
Year
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
I960
1961
1962
1963
1964
1965
1966 24.74 27.13
1967 34.02
1968 30.02
1969 33.03
1970 28.85
1971 26.45
1972 29. 59 (Up to Sept.
Mary College
33. 61
33. 61
26.77
26.40
32. 12
36. 34
35. 78
25. 55
40. 10
26.94
25. 35
24. 32
23. 30
32. 50
39.63
24. 89
18. 50
31. 60
Milwaukee i
36.43
32.69
22.87
35.91
27.90
28. 35
24. 95
20. 17
37.68
40.71
21.91
19. 10
28. 18
141
-------
Annual differences between local guages and the weather bureau
record at the airport are in the order of 20%, with the local (project
area) gauges consistently recording less annual rainfall. The year 1971
is a notable exception, where local gauges recorded particularly high
values. In this case the major variations occur early in the year and
may be related to distribution of snowfall.
Hourly rainfall data for the period covered by the program are
recorded in the Appendix. This listing, developed from punched cards,
recording hourly rainfall at all three stations, lists hourly data on those
days where precipitation occurred. Any day on which no precipitation
was recorded is not listed. In addition, a synoptic listing of each
storm event is presented and provides the following data on each storm
event:
Date and hour of day for beginning of event
Duration (Hours)
Average Intensity (Hundredths of an inch/hour)
Total Volume (Hundredths of an inch - 0. 01 inch)
Delta (Interval in hours since end of antecedent storm)
Delta Prime (Interval in hours between centroids of
storm and antecedent storm)
Station records were compared statistically, based on hourly
records to perform a similar comparison with that previously made by
inspection for monthly and annual totals. For the total period of record
each of the local gauge records (STA 2 and STA 4) were compared
individually with the weather bureau record (STA 1) and with each other.
The analysis explored Total Volume, Hours of rain recorded, and
average intensity for maximum hour, maximum 2 hours, and maximum
3 hours.
Based on correlation coefficients determined - the data is obviously
correlated, though, as can be expected, not to very high levels. Best
correlation obtained is for volume, with correlation coefficients of
about 0.8. Hours of rain and intensity correlated to a lesser degree
(coefficient approx. 0. 6). Data from this analysis indicates the local
gauges correlated with each other to a greater degree than either
does with the airport gauge as expected because of the relative gauge
locations. The analysis further confirms the indication that generally
less rain is recorded by local stations than at the airport, based on the
period of study. This may change if longer term averages were used.
The significance of these differences in recorded rainfall on storm
overflow predictions cannot readily be determined from this rainfall
record, since quite opposite effects are observed for individual events
142
-------
and even for monthly totals, than the general average conditions indicate.
However, the use of long term rainfall data in projections is a valid
approach for this study.
The significance of the rain gauge record selected for predicting
storm overflow conditions, will be made by operating the overall model
using each of the rainfall records independently and comparing the
predicted output for significant differences.
SEWAGE - WET WEATHER FLOW (RUNOFF COEFFICIENT CR)
Since the Urban Storm overflow model utilized in this project
operated on hourly rainfall data, an important element of data analysis
was the determination of a runoff ratio or coefficient. For a specific
volume of rainfall which fell on the area during each hour of a storm
event, the percentage which reached and flowed through the combined
sewer system was identified. In addition, significant influences on
this ratio or coefficient were important to identify. Such factors as
rainfall intensity and duration, and the interval since the antecedent
storm would be expected to have some effect on the observed runoff
coefficient. The greater the degree to which such influences can be
identified and quantified, the greater will be the accuracy of the model
using such a factor in the output it predicts.
Data available from the project on rainfall and on combined sewer
flows during storm events were analyzed to develop characteristics for
the runoff coefficient (Cr) for the 570 acre project area. The runoff
coefficient utilized in this program was defined as the ratio of the volume
of storm water reaching the combined sewers, to the volume of rain
which falls on the drainage area. Thus:
~ _ Volume of runoff
r Volume of rain
Analysis of rainfall volume consisted in taking hourly rainfall
intensity data from rain gauging station No. 2 located approximately 2
miles south of the project area and station No. 4 in the project area,
and computing the total volume of rain which fell on the 570 acre drainage
area during each storm event. Each of the rain gauging stations was
treated independently in this analysis. Data from the period November,
1971 through October, 1972 were utilized in this analysis. The analysis
assumes that the rainfall recorded by a gauge represents a uniform
intensity over the area. This would appear to be a reasonable assump-
tion, based on the relatively small size of the area (less than 1 square
mile). Some variation in rainfall patterns was exhibited between the
two rainfall gauging stations, as would be expected. While some
differences can be assumed to be due to the different gauges in use at
the two stations, the gauges rather consistently record somewhat
143
-------
different times, intensities and volumes of rain falling, indicating a
non-uniform distribution of rainfall bet-ween the gauge located in the
project area and the one to the south of the area during most events.
Based on this comparison, it is expected that some lesser degree of
variation also may exist in the various locations throughout the project
area. In the data presented, storm events on March 1, June 12, and
August 14, 1972 were recorded at only one of the stations. This factor
was noted, although it was not particularly significant in the operation
of the predictive model. Such variations were compensated for in the
overall runoff coefficient (Cr) selected for the test area. Table 31
presents a summary of the storm events utilized in the determination of
C , and includes both rainfall recorded at the gauging station and calculated
volume of rainfall.
Runoff in the combined sewers during each of these events is also
presented in Table 31. Runoff volume was calculated from a flow balance
using recorded flows in the combined sewer system during the storm
events analyzed. Figure 37 presents a schematic representation of the
combined sewer system in the test area. There are a series of "over-
flow stations" within the upper reaches of the project area collection
system, as described in Sections IV and V of this report. These were
monitored overflow stations. Experience indicated that they overflow
relatively infrequently, and for quite short periods. They -were treated
as a single element in the analysis.
Normal dry weather flows, and combined flows from the smaller
storms flow into the Metropolitan Interceptor Sewer (M. I. S) at Station
21 and were recorded (Flow meter 21-0) at that point. When flows
exceed the capacity of the M. I. S. intercepting device at that station
(approximately 4. 5mgd), the excess overflows and is diverted to
Station 49. An inline flume in this line (Flow Meter 49-F) was installed
to record flows entering Station 49. As described in Section VII,
mechanical difficulties greatly reduced the amount of information obtained
from this meter. Normal dry weather flows and some combined flow
are diverted into the M. I. S. at Station 49 and are recorded (Flow Meter
49-0) in the same manner as at Station 21. Flows, beyond the
capacity of this outlet at Station 49, overflow and enter the project
detention tank.
Capacity of the M. I. S. intercepting device at Station 49 is approximately
3. 0 to 3. 5 mgd. , resulting in a total interceptor capacity of approximately
8. 0 mgd. Flow enters the M. I. S. sewer lines by passing through a
submerged orifice, the capacity of which increases somewhat with higher
combined sewer flows. Interceptor capacity in the study area is thus
3. 5 to 4. 5 times the dry weather flows.
144
-------
a
UJ
m
FIGURE 37— SCHEMATIC OF COMBINED
FLOW METERING SYSTEM
SEWER
ce
UJ
UJ
V)
u.1
UJ
_i
Ul
a:
o*
Q:UJ
u. w
DRAINAGE
AREA .
STt
(7) FLOW RECORDER
((?) LEVEL RECORDER
UK^IKtAM
570 ACRES OVERFLOW
/^ SJATIONS
1
mON ^\X
21 £
Q
r
/P\ fc TO M.I.S.
^ P
^--FLOW METER 21-0
^^-FLOW METER 49-1
JT r-FLOW METER 49-0
ac.
UJ
e
STATION
49
TO M.I.S.
STATION
99
STORM
DETENTION
TANK-
©
STATION
98
OVERFLOWS
TO RIVER
145
-------
Average
TABLE 31
DATA SUMMARY OF RAINFALL vs RUNOFF
Station 2 - Rainfall
Station 4 - Rainfall
Storm
Date
11-01-71
11-18-71
11-26-71
11-28-71
12-10-71
12-14-71
12-30-71
02-25-72
03-01-72
04-21-72
06-02-72
06-12-72
06-14-72
06-19-72
07-02-72
07-12-72
07-17-72
07-26-72
08-02-72
08-06-72
08-11-72
08-14-72
08-23-72
08-25-72
09-12-72
09-17-72
09-28-72
10-20-72
Duration
(hr)
10
4
11
16
14
10
21
3
-
9
3
-
9
6
5
2
6
12
7
8
6
-
7
6
7
10
5
62
Interval
(days )
2.5
0
2.0
1.5
.8
3.3
2.3
2.0
-
4.2
3.5
-
2.3
4.7
.4
4.5
3.0
3.0
.10
3.7
3.7
-
5.3
1.8
1.3
4.5
3.0
9.6
Volume
(in)
1.46
.04
.41
.61
1.54
1.52
1.26
.07
-
1.07
.41
-
.62
.90
.46
.68
.48
.51
.77
.88
.80
-
. 48
1. 32
1.48
2.41
.78
2.85
Duration
(hr)
9
3
11
20
12
9
17
2
7
13
4
5
9
5
4
2
6
-
7
7
5
3
7
7
9
11
5
62
Interval
{days )
2.5
0
3.5
2.5
.8
4.3
2.3
2.0
5.0
4.2
3.5
9.5
2.3
3.7
.5
5.0
2.7
-
.10
3.3
3.7
3.0
5.3
1.8
1.6
4.4
3.5
9.6
Volume
(in)
1.74
.03
.49
.64
1.36
1.37
1.16
.05
.36
1.06
.24
.86
.73
.66
.52
.43
.46
-
.81
.59
.83
.64
.40
1.13
1.02
1.94
.60
2.25
Runoff
(me!)
_
.27
2. 34
4.93
4.12
9.61
4.98
.50
3.41
6.90
2.64
6.27
9.50
5.54
3.88
3.43
3.04
3.89
7.06
4.62
5.62
5.72
3.48
8.08
8.12
lb.5
4.81
22.46
C2
—
.44
.37
.52
.54
.41
.29
.46
_
.56
.42
.78
.43
.55
.33
.41
.50
.59
.34
.45
.47
.40
.36
.44
.40
.51
C4
_
.58
.31
.48
.58
.45
.33
.60
.61
.50
.71
.47
.66
.64
.48
.52
.43
.56
.51
.44
.58
.56
.54
.52
.55
.52
.68
Storm
Number
M 1
M 2
M 3
M 4
M 5
M 6
M 7
M 8
M 9
M10
Mil
Ml 2
Ml 3
Ml 4
M15
Ml 6
Ml 7
Ml 8
Ml 9
M20
M21
M22
M23
M24
M25
M26
M27
M28
,46
.54
Note: Run off coefficients plotted verses time since
antecedent storm, duration of storm, rainfall volume,
and rainfall intensity are noted in Figures 38 and 39,
Figures 38 and 39 contain data points which were de-
rived by averaging values in appropriate plotting
groups and then utilizing the average values for the
points.
-------
In the original metering concept, the flows of interest in the
program were to be developed as follows:
Total Q in system = C^I.Q + ^49_F + Sum of uPstrearn Overflows
Flow to tank = Q^.p - C^ng
Storm runoff = Total Q - Dry weather Q
However, because of the difficulties with the flow metering
facilities at Station 49-F, a flow at Station 99 was calculated based
on level changes in the storm tank. The flows of interest were then
developed in the following manner:
Total Q in system = C^I.Q + ^49-0 + ^99 + Sum of Upstream Overflows
Flow to tank - 099
Storm runoff = Total Q - Dry weather Q
It is important to note that complete flow data based upon tank
level could be developed in this manner without Station 49-F input,
only when the detention tank was not full, wherein tank level changes
could be measured. When meter 49-F continued to present operational
problems creating difficulties in providing a reliable flow record, provision
was also made to determine flow rates leaving the storm tank by measuring
head on the outlet weir. When meter 49-F was operating comparison
of the two total flows calculated for the system were made. At times
when the tank filled and overflowed, storm tank flow monitoring would
shift from Station 99 to Station 98 at the tank outlet.
For each specific storm event, flows recorded at all stations
were summed up for the storm interval and converted to a total volume
(million gallons). A listing of all recorded flow information is presented
in the appendix. Data in this listing has been arranged chronologically.
In calculating the runoff, where data •was incomplete, some flow assumptions
were made. A frequent assumption was for a flow rate of 3. 5 mgd at Station
49-0 during events where overflow reaching the storm tank occurred.
This assumption is based on observed values during storm events when
such data was available. Dry weather flow values used in the runoff
calculations are based on an average dry weather flow of 2. 0 MGD.
Runoff volumes determined in this analysis are presented in Table
31, together with the rainfall data previously discussed.
147
-------
In the analysis, separate rainfall volumes were computed using
rain gauge stations 2 and 4. A runoff coefficient (C ) was obtained
for each set of rain data, for each storm event investigated, by calculating
the ratio between the total runoff for the storm in question and the rainfall
volume. Calculated values for Cr are presented in Table 31.
Runoff coefficients have been determined using a total of 28 storm
events over one year of operation of the storm retention tank. Coefficient
analysis was limited to those events where sufficiently complete flow
information was available. The flow assumptions made and the justification
for using an assumed value have been discussed above.
Where flow at more than one station was missing, that event was not
utilized in developing a value for Cr. Events which were utilized in as sumptions
are noted in Table 31, and generally yield values for Cr in the same
range as for those events with complete data. In many instances all
monitoring stations did not operate for a particular storm because the
storm did not generate flows in excess of service sewer capacity. The
amount of rainfall necessary to generate an overflow condition varies
somewhat from station to station.
On the basis of the data analyzed, storm water runoff in the test
area is characterized by a runoff coefficient Cr of 0. 5. Observed
ranges for C were 0. 3 to 0. 8.
Rainfall, runoff data, and calculated Cr values were subjected
to an analysis to identify the influence of various storm characteristics
on the runoff coefficient. The effect of duration of the storm event,
total volume of rain per storm, rainfall intensity, and interval since
antecedent storm were investigated. Results of this analysis are plotted
on Figures 38 and 39. When all the storm data was evaluated, no significant
variation in C with any of the above parameters is indicated. The
data obtained was not sensitive enough to determine whether variations
might exist for individual hours within a storm event.
SEWAGE - WET WEATHER QUALITY
A substantial body of basic data was secured by the sampling
program for combined sewage flows during storm events. A
comprehensive program of chemical analyses covering a wide range
of parameters was incorporated in the test program. Sampling was
done at relatively short intervals providing a large number of
individual observations during each storm event, for a large number
of separate events.
148
-------
FIGURE 33- EFFECT OF STORM CHARACTERISTICS
ON RUNOFF COEFFICIENT
Table 31
.*-
o .?
8
u.
8
U
UJ
.5-
©
j
0
o
23456789
TIME SJNCE ANTECEDENT STORM -DAYS
10 II
b i.o-
U)
.8-
Cy
O
-Oj
6 6 10 12 14 16 IB
DURATION OF STORM -HOURS
20
Note: Data regarding individual storm characteristics was
given in Table 31. Data points plotted above are based upon
Table 31 information. However, the points plotted were es-
tablished by plotting the average value of a plotting range.
This range includes one or more individual storm data points
149
-------
FIGURE 39 - EFFECT OF STORM CHARACTERISTICS ON RUNOFF
COEFFICIENT.
I.U
>- 0 .
OEFFICIEN
n e
RUNOFF C
^ C
AVERAGE
0 '4
®
©
9
a\
«
*
.0 .2 4 .6 .8 1.0 12 1.4 Ifi 18 20 22 24 26
-RAINFALL VOLUME, in.
7 _
6_
4_
.
0 •
•
A 1
• •
§
• u
o
o
o
•
'" .00 .0
•
•
r i°
o
8
•
5
5
•
Q
22
o •
°1
•
•
•
fc
i i
• RAIN GAUGE* 2
QRAIN GAUGE #4
•
•
O
o
0
0 .15 .20 .25 .30
RAINFALL INTENSITY in/hf
Note: Data regarding individual storm characteristics was
given in Table 31. Data points plotted above are based upon
Table 31 information. However, the points plotted were es-
tablished by plotting the average value of a plotting range.
This range includes one or more individual storm data points,
150
-------
The basic data is listed in Appendix V, and is organized in
chronological order. Inspection of the raw data and individual plots
of different events has indicated a wide variation in observed concentra-
tions for all parameters. Many of the events, though not all, exhibit
a typical "first flush" pattern. The flushing of accumulated materials
on streets and in sewers and catch basins, etc. , appears to result
in a rapid increase in concentration of some quality parameter, to a
peak value, followed by a decline to some lower value - often lower
than the dry weather concentration - attributable to dilution of sewage
by the relatively less concentrated storm runoff waters.
A computer program was utilized to consolidate and analyze the
mass of available data. In this approach, all quality data from a
storm event was sorted in accordance with elapsed time since the start
of the storm event. Since automatic samplers were actuated by increased
flow at the sewer monitoring station, the time associated with the initial
quality sample was identified as the start of the storm event. Quality
data for each storm event was analyzed individually, and the following
characteristics summarized:
1. initial concentration
2. maximum concentration
3. average concentration - 0 to 30 minutes
4. average concentration - 30 to 60 minutes
5. average concentration - 1 to 2 hours
6. average concentration - all hours in excess of 2 hours
In addition, the average for all storms analyzed was determined.
this data is summarized in Table 32.
The average quality variations summarized in Table 32 were
based on data from 97 individual storm events. The data clearly
demonstrates a pattern of quality variation which reflects the existence
of a "first flush" condition. When all available data is considered
on an averaged basis, every one of the quality parameters investigated
demonstrates this pattern whereby concentration of the contaminant
decreases with time as the storm event progresses.
When storm events are considered individually, not all events
show this pattern, even though the frequency of occurrence and magnitude
of the deviation from the above pattern, are not sufficient to distort the
overall picture provided by analysis of a large enough sample. In many
cases when deviations from the pattern occurred in an individual storm,
they could be attributed to a complex storm pattern, in which a number
of radical changes in rainfall intensity occured during the duration of
what had been considered a single storm event. A more rigorous
definition as to what constitutes a single storm event could address
151
-------
TABLE 32
AVERAGE OF WET WEATHER SEWER QUALITY
VARIATIONS FOR ALL STORMS ANALYZED
Quality
Parameter
BOD (mg/1)
SS (mg/1)
pH -
COD (mg/1)
Chlroides (mg/l-Cl)
Nitrogen
(as N)
Kjeldahl (mg/1)
Ammonia (mg/1)
Nitrite (mg/1)
Nitrate (mg/1)
Phosphorus
as P04
Ortho (mg/1)
Total (mg/1)
Total
Solids (mg/1)
Total Vola-
tile Solids (mg/1)
Coliforms
Total no. /ml (Wide
Fecal no. /ml
Average
Initial
Value
150
397
6.9
439
188
14.6
4.6
.013
.827
8.8
13.4
813
242
range in data,
14,173
Average
Maximum
Value
176
489
7.1
581
210
17.2
5.7
.013
1.005
9.4
15.7
952
289
averages
21,787
Average
0-30 Minute
Value
143
348
6.9
436
185
13.6
4.5
.013
.749
7.6
13.0
775
217
Average
30-60 Minute
Value
128
270
6.9
383
131
11.3
3.9
.011
.565
6.1
10.1
661
149
Average
1-2 Hour
Value
97
192
6.9
280
167
8.4
3.1
.013
.520
5.4
9.1
681
108
Average
Greater than 2
Hour Value
87
193
6.9
256
117
8.6
3.3
.013
.629
4.9
8.4
550
106
arc not representative of data)
13,671
7,946
6,902
6,199
-------
such deviations. To do so, would involve breaking down some
storm events into a series of overlapping events. This ana-
lytical refinement has not been utilized in this project for
several reasons. Doing so would significantly complicate
what is desired to be a simple, direct, and readily useable
analytical tool for identifying the effects of storm runoff
in an urban area and for exploring engineering alternatives
for their control. In addition, the major emphasis in the
program is to identify the broader aspects of "all storm run-
off" rather than one individual event, and a clearly defined
pattern has been shown where all storms are considered.
Table 33 has been prepared to help provide some perspec-
tive both on the quality data for wet weather conditions and
on the relationship of wet weather quality with dry weather
quality. The summary in this table indicates that many para-
meters exhibit somewhat higher initial concentrations than
average dry weather concentrations. BOD, COD, suspended and
volatile suspended solids, chlorides and nitrates exhibit
this effect. While the normal wide variation in dry weather
quality may have some effect on this observation, the varia-
tions in the wet weather concentrations are believed to be
largely due to the fact that the initial wet weather sample was
in fact secured some time after the effects of the storm event
had begun to be felt at the sampling station and was part of
the "first flush" combined sewage concentration. The occur-
rence of a "first flush" phenomenon in the early stages would
result in initial samples having concentrations somewhat greater
than dry weather values.
Table 33 lists a calculated ratio of the maximum concentra-
tion observed during storm runoff compared with the dry weather
average value for the same parameter. For BOD, COD, solids,
chlorides and nitrates and organic nitrogen, maximum storm
values are about 1.5 to 2.5 times the dry weather average.
These constituents may be assumed to be increased by virtue of
materials flushed from streets, catch basins, sewer lines, etc.,
by the storm flow.
Several parameter initial concentrations did not increase
during storm events. Ammonia, total and ortho phosphate, and
fecal coliforms, all exhibited maximum concentrations less than
dry weather averages (total kjeldahl nitrogen has not been con-
sidered in itself, but rather in its individual components
ammonia and organic nitrogen). This suggests the quite reason-
able conclusion that the primary source of such contaminants is
the sewage in the lines when the storm occurs. The observed
ratios of 0.45 for ammonia and 0.58 for fecal coliform (maxi-
mum wet to dry concentration) imply a simple dilution by storm
waters in which they are present in significantly reduced con-
centrations than sewage concentrations. Phosphorus ratios are
0.85 to 0.88 suggesting some contribution by runoff scouring of
sewer lines, but significantly less phosphorus bearing parti-
culate matter accumulates during dry weather than does organic
particulates .
153
-------
01
-p-
TABLE 33
COMPARISON DRY AND WET WEATHER SEWAGE QUALITY
Dry Weather Quality
Average
Parameter Concentration
BOD (mg/1) 112
SS(mg/l) 150
TVS (SS)(mg/l) 115
pH 7.6
COD (mg/1) 238
Cl(mg/l) 141
TKN(mg/l-N) 20.7
Organic (mg/l-N) 9.0
NH3 (mg/l-N) 11.7 1
N02 (mg/l-N) 0.01
N03 (mg/l-N) 0.5
Ortho-P(mg/lPO4)lO . 7
Total-P(mg/lPO4)18. 4
T Col(nu/ml) 930,000
Wet Weather Quality
Initial
Range Value
17-323 *
5-900 *
2-700 *
6.7-8.7
35-730 *
3-826 *
5.6-47 o
4.6-16 *
.0-31.2 o
0.1-1.6 *
1-39 o
0.5-60 o
F Col (nu/ml) 38,000 400-54000 o
150
397
242
6.9
439
188
14.6
10.0
4.6
0.01
0.83
7.8
13.4
14,000
Maximum
Value
176
489
289
7.1
581
210
17.2
11.5
5.7
0.01
1.01
9.4
15.7
22,000
+ 2 Hour
Value
87
193
106
6.9
256
117
8.6
5.3
3.3
0.01
0.63
4.9
8.4
6,200
Ratio
Max
DW Avg
1.57
2.63
2. 51
2.44
1.5
0.83
1.28
0.49
-
2.0
0.88
0.85
0.58
*Initial Wet greater than average dry.
olnitial Wet less than average dry.
-------
WET WEATHER QUALITY CORRELATIONS
Available data on wet weather BOD and suspended solids was sub-
jected to a linear regression analysis to identify the degree to which
characteristic concentrations are dependent on storm conditions. These
parameters were selected because of their significance in the operation of
the storm detention tank, and because their variation during a storm event
would be comparable to variations observed with other parameters.
Individual quality data for the 97 storm events were used in this
analysis. Synoptic rainfall data for storm events associated with quality
records were also incorporated. The objective of this analysis was to
determine the extent of a direct relationship between pertinent quality
parameters and storm characteristics. For example, it is generally
believed that quality of combined sewage will be influenced by the time
elapsed between the current storm event and the antecedent storm. Streets,
catch basins and sedimentation in the sewer lines themselves would be
expected to accumulate greater quantities of contaminants over a long
dry spell, and thus result in higher concentrations in the combined sewer
flow during the next rainfall.
The system simulator model included provision for using varying
values for runoff and quality concentrations, depending on characteristics
of the storm events. The analysis performed was designed to identify
such relationships if such were indicated by the data developed.
The analytical approach employed compared data for two variables and
determined the degree of correlation between them using a linear regression
analysis. A least squares curve fitted to the data was determined, and
variance and correlation coefficients were calculated.
When quality data only is compared, high degrees of correlation were
obtained for both BOD and suspended solids, between the overflow initial,
maximum and zero to 30 minute average values. Correlation coefficients
between 0. 81 and 0. 92 were obtained. High degrees of correlation were also
indicated when BOD and suspended solids concentrations for samples taken
more than 1 hour after the beginning of a storm were compared. Such
results generally confirm the applicability of the quality variation relation-
ships developed in the previous section.
However, attempts to correlate combined sewage quality variations
with storm characteristics were unsuccessful. Comparing maximum observed
BOD with rainfall intensity and duration showed no correlation. Some
correlation with interval since the antecedent storm is indicated, however
the degree of correlation observed is relatively small and can be neglected
•without introducing significant error in calculating BOD variations from
155
-------
storm parameters. For example, the correlation coefficient relating
BOD max to interval between storms is about 0. 33. An expression relating
combined sewage BOD concentrations to interval between storms, would
remove less than 10% of the variance between observed values and the mean
value developed. The effect of such a refinement in calculating input to a
storm tank based on storm data is insignificant.
This is not to say that a clear relationship between storm interval
and wet weather BOD does not exist. Some dependence is in fact indicated.
What the analysis does say is that, under the conditions which prevailed
over the period in which the 97 events took place, whatever relationship
which existed did not prove to be significant.
A comparison of suspended solids with storm parameters gave similar
results.
Table 34 tabulates correlation coefficients developed by this analysis
of the rainfall and quality data. From the results of this analysis, it was
concluded that one is justified in characterizing combined sewer quality
by the average values developed in the previous section, for various time
intervals during a storm event. The data developed provides no basis for
modifying such values based on characteristics of a storm.
DETENTION TANK PERFORMANCE
Where the total volume of combined sewage entering the tank did
not exceed the vailable capacity of the tank, the entire overflow was retained,
and contaminant removal was 100 percent. At times when the storm runoff
volume exceeded the available tank capacity, overflow from the tank occurred.
Removal of contaminants achieved under these conditions was related
to the efficiency of the storm detention tank as a sedimentation device.
Total removal of contaminants during an individual storm event, or over
the course of a year was related to the combined effects of storage and
sedimentation efficiency. In all cases, flows which left the detention tank
to enter the river were chlorinated for destruction of coliform organisms.
Further, any storm waters detained by the tank, were returned to the sewer
system for transport to normal treatment facilities, once wet weather
flows have subsided and sewer capacity was available.
A substantial portion of the overall effectiveness of a detention tank
in reducing pollution reaching a water course, resides in its ability to
retain storm waters and later return them to a treatment system. Detention
effectiveness is determined both by the size of the tank (Storage Volume),
and also by the rate at which the tank can be emptied, following a storm
event. Where the emptying rate is restricted for some reason, the net
effect is one of reducing the effective storage volume of the tank, since a
156
-------
Table 34. WET WEATHER QUALITY
CORRELATION COEFFICIENTS
Sewer BOD Initial
1
2
3
4
5
6
7
Rainfall Volume
mg/1
is Correlated with.
Sewer BOD 0-30 Min Avg (mg/1) 0. 9000
Sewer BOD 30-60 Min Avg (mg/1) 0. 6398
Sewer SS Initial (mg/1) 0. 5821
Sewer SS Maximum (mg/1) 0. 5157
Sewer SS 0. 30 Min Avg (mg/1) 0. 5877
Rainfall Volume1 2 inches -0. 3117
Sewer BOD Maximum (mg/1) 0. 8090
Inches
is Correlated with
Rainfall Volume Sta. 2 inches
0.5366
Sewer BOD 0-30 Min Avg mg/1
is Correlated with.
1
2
3
4
5
6
7
Sewer BOD Initial (mg/1) 0. 9000
Sewer BOD 30-60 Min Avg (mg/1) 0.7021
Sewer SS Initial (mg/1) 0.4162
Sewer SS Maximum (mg/1) 0.4482
Sewer SS 0-30 Min Avg (mg/1) 0. 5429
Sewer SS 30 - 60 Min Avg (mg/1) 0. 3257
Sewer BOD Maximum (mg/1) 0. 9108
Sewer BOD 30-60 Min Avg mg/1
is Correlated with.
1
2
3
4
5
Sewer BOD Initial (mg/1)
Sewer BOD 0-30 Min Avg (mg/1)
Sewer SS 30 - 60 Min Avg (mg/1
Sewer SS Gr 2 hr Avg (mg/1)
Sewer BOD Maximum (mg/1)
0.6398
0.7021
0.4478
0.3121
0.6420
Sewer BOD 1-2 Hr Avg
mg/1
is Correlated with.
1
2
3
4.
Sewer BOD Gr 2 hr Avg (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
Sewer SS 1-2 Hr Avg (mg/1)
Sewer SS Gr 2 Hr Avg (mg/1)
6085
3477
8294
0.5337
157
-------
Table 34. (Continued)
Sewer BOD Gr2 Hr Avg mg/1
is Correlated with.
5
11
12
Sewer SS
1
3
8
9
10
14
Sewer SS
1
3
7
9
10
14
Sewer SS
1
3
7
8
10
14
Sewer SS
3
4
5
7
8
9
11
12
14
Sewer BOD 1-2 Hr Avg (mg/1)
Sewer SS 1-2 Hr Ave (mg/1)
Sewer SS Gr 2 Hr Avg (mg/1)
Initial mg/1 is Correlated
Sewer BOD Initial (mg/1)
Sewer BOD 0-30 Min Avg (mg/1)
Sewer SS Maximum (mg/1)
Sewer SS 0-30 Min Avg (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
Sewer BOD Maximum (mg/1)
Maximum mg/1 is Correlated
Sewer BOD Initial (mg/1)
Sewer BOD 0-30 Min Avg (mg/1)
Sewer SS Initial (mg/1)
Sewer SS 0-30 Min Avg (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
Sewer BOD Maximum (mg/1)
0-30 Min Avg mg/1 is Correlated
Sewer BOD Initial (mg/1)
Sewer BOD 0-30 Min Avg (mg/1)
Sewer SS Initial (mg/1)
Sewer SS Maximum (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
Sewer BOD Maximum (mg/1)
30 - 60 Min Avg mg/1 is Correlated
Sewer BOD 0-30 Min Avg (mg/1)
Sewer BOD 30-60 Min Avg (mg/1)
Sewer BOD 1-2 Hr Avg (mg/1)
Sewer SS Initial (mg/1)
Sewer SS Maximum (mg/1)
Sewer SS 0-30 Min. Avg (mg/1)
Sewer SS 1-2 Hr Avg (mg/1)
Sewer SS Gr 2 Hr Avg (mg/1)
Sewer BOD Maximum (mg/1)
0. 6085
0. 5476
0. 9244
with
0. 5821
0.4162
0. 9227
0. 9054
0.4284
0.4407
with
0. 5157
0.4432
0. 9227
0. 9368
0. 5996
0. 5460
with
0. 5877
0. 5429
0. 9054
0. 9368
0. 6285
0. 5466
0. 3257
0.4478
0. 3477
0.4284
0. 5996
0. 6285
0. 5894
0. 3647
0. 3441
158
-------
Table 34. (Continued)
Sewer SS 1-2 Hr Avg mg/1 is Correlated with. ..
5
6
10
12
Sewer BOD 1-2 Hr Avg (mg/1)
Sewer BOD Gr 2 Hr Avg (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
Sewer SS Gr 2 Hr Avg (mg/1)
Sewer SS Gr 2 Hr Ave me/I is Correlated
4
5
6
10
11
Rainfall Volume
1
2
o o • —
Sewer BOD 30-60 Min Avg (mg/1)
Sewer BOD 1-2 Hr Avg (mg/1)
Sewer BOD Gr 2 Hr Avg (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
Sewer SS 1-2 Hr Avg (mg/1)
2 Inches is Correlated
Sewer BOD Initial (mg/1)
Rainfall Volume St. 4 Inches
Sewer BOD Maximum mg/1 is Correlated
1
3
4
7
8
9
10
Rainfall Duration
4
5
7
8
Rainfall Intensity
y
3
5
8
Sewer BOD Initial (mg/1)
Sewer BOD 0-30 Min Avg (mg/1)
Sewer BOD 30-60 Min Avg (mg/1)
Sewer SS Initial (mg/1)
Sewer SS Maximum (mg/1)
Sewer SS 0-30 Min Avg (mg/1)
Sewer SS 30-60 Min Avg (mg/1)
St. 2 Hrs is Correlated
Rainfall Duration St. 4 (hr)
Rainfall Intensity St. 4 (in/hr)
Rainfall Volume St. 4 (in)
Rainfall Volume St. 2 (in)
St. 2 in/hr is Correlated
Rainfall Delta St. 2(hr)
Rainfall Intensity St. 4 (in/hr)
Rainfall Volume St. 2 (in)
0. 8294
0. 5476
0. 5894
0. 5142
0. 3121
0. 5337
0. 9244
0. 3647
0. 5142
with
-0. 3117
0. 5366
with
0. 8090
0. 9108
0. 6420
0. 4407
0. 5460
0. 5466
0. 3441
with
0. 6006
0.4249
0.4657
0. 6156
with. ....
-0. 3586
0. 3109
0. 8117
159
-------
Table 34. (Continued)
Rainfall Delta St. 2 his is Correlated with.
2
6
9
Rainfall Intensity St. 2 (in/hr)
Rainfall Delta St. 4 (hr)
Sewer BOD Maximum (mg/1)
-0. 3586
0.4293
0. 3240
Rainfall Duration St. 4 hrs is Correlated with
1 Rainfall Duration St. ? (hr) 0. 6006
5 Rainfall Intensity St. 4 (in/hr) 0. 3695
7 Rainfall Volume St. 4 (in) 0.4865
8 Rainfall Volume St. 2 (in) 0.6164
Rainfall Intensity St. 4
in/hr
is Correlated with.
1
2
4
7
8
Rainfall Delta St. 4
3
9
Rainfall Duration St. 2 (hr)
Rainfall Intensity St. 2 (in/hr)
Rainfall Duration St. 4 (hr)
Rainfall Volume St. 4 (in)
Rainfall Volume St. 2 (in)
0.4249
0.3109
0.3695
0.9567
0. 5707
hrs
is Correlated with.
Rainfall Delta St. 2 (hr)
Sewer BOD Maximum (mg/1)
0.4293
0.3323
Rainfall Volume St. 4
in
is Correlated with.
1
4
5
8
Rainfall Duration St. 2 (hr)
Rainfall Duration St. 4 (hr)
Rainfall Intensity St. 4 (in/hr)
Rainfall Volume St. 2 (in)
0.4657
0.4865
0. 9567
0.5366
Rainfall Volume St. 2
in
is Correlated with.
1
2
4
5
Rainfall Duration St. 4 (hr)
Rainfall Intensity St. 2 (in/hr)
Rainfall Duration St. 4 (hr)
Rainfall Intensity St. 4 (in/hr)
Rainfall Volume St. 4 (in)
Sewer BOD Maximum
mg/1
3
6
Rainfall Delta St. 2 (hr)
Rainfall Delta St. 4 (hr)
6156
8117
6164
0.5707
0.5366
0.3240
0.3323
160
-------
Table 34. (Continued)
SS Max. rng/1 is correlated -with.
1 Rainfall Duration St. 2 0. 355
2 Rainfall Intensity St. 2 0. 194
3 Rainfall Delta St. 2 0.004
4 Rainfall Duration St. 4 0. 112
5 Rainfall Intensity St. 4 0.016
6 Rainfall Delta St. 4 -0.003
7 Rainfall Volume St. 4 -0.084
8 Rainfall Volume St. 4 0. 273
BOD Max. mg/1 is correlated with.
1 Rainfall Duration St. 2 -0. 184
2 Rainfall Intensity St. 2 -0. 121
3 Rainfall Delta St. 2 0. 324
4 Rainfall Duration St. 4 -0. 277
5 Rainfall Intensity St. 4 -0. 288
6 Rainfall Delta St. 4 0. 332
7 Rainfall Volume St. 4 -0.297
8 Rainfall Volume St. 2 -0. 227
161
-------
subsequent storm will be more likely to occur with some volume from the
previous one still in the tank.
Evaluation of the overall capabilities of the detention tank must
therefore take into account both capability for retaining overflows, and
capability for removing contaminants by sedimentation in a flow-through
situation. The former element can be handled readily by simple counting
procedures which account for tank levels and overflow volumes at regular
intervals of time.
The latter factor, sedimentation efficiency, under conditions which
exist in a storm detention tank, presents a more difficult situation. The
basic approach to defining removal of settleable contaminants by sedi-
mentation is straightforward, and many examples are available in the
literature. For example, Fair and Geyer° present typical settling curves
for BOD and suspended solids in sewage. Removals due to simple sedi-
mentation can be expressed by an equation of the form:
-kt
Percent removal = a(l - e )
In this expression the constant "a" reflects the fraction of the insoluble
BOD or solids which will be removable by plain settling; the constant k
defines the rate at which the contaminants settle. As the expression
indicates, removals are related directly to sedimentation time (t), or in
the case at hand, the amount of time a particular "batch" of storm runoff
resides in the tank until it is displaced out the overflow end of the tank.
The analysis of the storm detention time as a sedimentation device
presents a quite complex situation. Flows entering the tank vary erratically
in response to fluctions in rainfall from hour to hour during a storm.
Detention time, as a result, is variable. In addition contaminant concentrations
are subject to fluctuations each hour of the storm. One might therefore
anticipate quite different performance results for a storm pattern which
fills the tank slowly and then surges due to high rainfall intensity occurring
just as the tank fills, compared with a storm pattern in which high flow
surges occur when the tank still has reserve capacity.
In order to evaluate performance efficiency of the storm tank during
tank overflow occurrences, both flow and quality data at influent and over-
flow were necessary over the duration of a storm. Table 35 summarizes
the periods when data relating to tank overflow was available. Where
quality data was present, an overflow from the detention tank was indicated.
At such times, i. e. , when tank is full and overflowing, influent flow will be
approximately equal to the overflow rate. The table indicates that complete
data for the detention tank was available during a series of events in
September, 1972. Continued malfunction of the flow meter at Station 49-F,
162
-------
which was to have provided required flow data on the tank prevented
the accumulation of necessary data from this source. However, a backup
source of flow data was available by converting changes in liquid level
in the detention tank reflected by a level recorder to flow rates. Flow data
generated from this information source was designated Station 99.
Much useful data was developed using this alternative; however,
input for evaluating sedimentation efficiency during overflow from the tank
could not be developed since the flow record ended when overflow began.
Flow information during overflow events was recorded when installation
of a supplemental metering device on the tank effluent (Station 98) was
completed. In Table 35, flow data determined from this source are reported
as flows from Station 99 (flow into tank) for simplicity, although the
source of flow data shifted as the overflow began.
There are five listed dates for which both inlet and outlet flow and
quality are available. However, in the data analysis program, the events
of September 20 and 21, 1972 were considered as a single storm event.
Raw data on BOD and suspended solids and on flows from these four events
is tabulated in Table 36. Inspection indicates several gaps in the data
on these four events. Specifically during the event of September 17 - 18,
flow and influent BOD records terminate between 4 and 6 A.M. , due to
equipment malfunction, well before the end of the storm. Influent suspended
solids data is incomplete after 4 A. M. In this case, for analysis of per-
formance, only the first segment of the storm, for which complete data
existed, was considered. During the event of September 20 - 21, 1972,
influent BOD and suspended solids record end prematurely due to equipment
malfunction or sample handling or transport problems. In this case,
concentrations were assumed for the missing values near the end of the
event.
In the overall model describing storm runoff and detention tank
performance, data input is in the form, of hourly rainfall, and contaminant
concentrations based on storm parameters. To evaluate the accuracy
of the component of this model which describes tank performance, the
detention tank model was modified to accept observed flows and contaminant
concentrations at the tank inlet, and to calculate the time variable contaminant
loading leaving the tank.
Input data for this verification analysis was developed from the raw
data listed in Table 36. In order to provide a continuous set of influent
conditions over the period of the storm event, influent flow and concentra-
tions were interpolated and/or averaged to provide hourly input values.
Table 37 summarizes translation of actual influent data to hourly averages
used as input in the storm tank model. In this table, actual data on tank
effluent BOD and suspended solids is also shown. From recorded flow and
163
-------
Table 35. COMPARISON OF DATA RECORDS FOR
EVALUATION OF DETENTION TANK PERFORMANCE
(Hours During Day for Which Indicated Data is Available)
Date
Quality Data at
Station 98
Flow at Station 99 =
11-01-71
11-02-71
11-29-71
12-10-71
12-15-71
03-07-72
03-16-72
03-17-72
03-18-72
03-20-72
03-21-72
04-20-72
04-21-72
06-14-72
07-14-72
07-15-72
08-25-72
08-26-72
09-13-72**(X)
09-18-7Z**(X)
09-19-72**(X)
09-20-72**(X)
09-21-72**(X)
10-22-72**
10-23-72**
1800-2400
0100-0500
0945-1545
0630-1500
0320-1000
1520-2330
1130-2330
0330-1620
1100-2300
1630-1730
1200-1300
2330
0130
1330-2130
2145-2500
0100
2400
0100
0150-0545
0240-2400
1230-1430
1700-2300
0100-0700
1100-2000
1200-1800
1515-1645
None
2200 (11/28 - 0930 (
0200-0700
2330 (12/14) - 0330
0030-1500
0030-1000
None
None
0900-1500
None
None
1200-2330
0630-0730
None
None
2000-2300
None
2330 (09/12) - 0730
1745 (09/17) - 0615
1000-1500
2045-2345
0015-0845
0630-2400
0030-1500
11/29)
(12/15)
(09/13)
(09/18)
*Flow at Station 99 reflects flow into tank (and flow from tank
when it is full and overflowing).
**Dates on which tank overflow data coincide with quality data.
(X) Dates on which influent quality data (Station 49) coincided
with influent flow data.
164
-------
TABLE 36
SUMMARY OF RAW DATA FOR TANK
PERFORMANCE ANALYSIS
Dates Time
9-12-72 2330
2345
9-13-72 0000
0015
0030
0045
0100
0115
0130
0145
0200
0230
0300
0330
0345
0400
0430
0500
0530
0600
0630
0700
0730
9-17-72 1745
1800
1815
1900
1915
1945
2000
2015
2045
2100
2115
2145
Rainfall
(in/hr.)
2 4
.24 .12
.94 .58
.14 .02
.02 .02
.06 .10
.06 .01
.12
.03
.12 .06
.03 .01
.04 .03
.06 .05
BOD (mq/1) SS (mq/1)
Influent Overflow Influent "Overflow
224.0
66.0
110.7
31.3
36.5
40.5
59.5
77.3
31.5 196.6
26.0
40.5
29.0 288.0
114.2
121.0
129.5
409.7
101.5
51.5
370
440
221
168
226
139
222
472
86 480
54
109
126
298
331
265
116
16
Flow
{mgd
at t)
7.1
78.0
65.2
34.4
7.6
1.0
.1
0
7.3
2.1
1.0
.1
13.2
6.5
1.9
1.7
.3
14.2
2.8
0
0
1.4
4.2
2.8
5.7
0
165
-------
TABLE 36
(continued)
SUMMARY OF RAW DATA FOR TANK
PERFORMANCE ANALYSIS
Rainfall
(in/hr.) BOD (mg/1) SS (mg/1)
Dates Time 2 4 Influent Overflow Influent Overflow
9-17-72 2215
2245
2315
2345
2400 .00 .05
9-18-72 0015
0030
0045
0100 .00 .01
0115
0130
0145
0200 1.4 1.04
0215
0230
0245
0300 .68 .56
0315
0330
0345
0400 .02 .10
0415
0430
0445
0500 .00 .01
0515
0530
0545
0600 .02 .03
0615
0630
0645
0700
0715
0730
71.5
107.0
85.2
54.8
35.2
37.8
69.0
40.0 36.1
60.0
74.0
77.0
81.5 94.1
118.6
86.1
120.6
84
187
92
383
268
275
174
134 146
91
136
107
160 320
370
332
306
Flow
(MGD
at t)
0
1.4
0
0
0
0
0
0
96.4
83.8
93.4
47.8
27.3
15.1
8.2
8.2
8.2
8.2
5.3
166
-------
TABLE 36
(continued)
SUMMARY OF RAW DATA FOR TANK
PERFORMANCE ANALYSIS
Rainfall Flow
(in/hr.) BOD (m^/1) SS (mg/1) (MGD
Dates Time 2 4 Influent Overflow Influent Overflow at t)
9-18-72 0745
0800
0815
0830
0845
0900
0915
0930
0945
1000
1015
1030
1045
1100
1115
1130
1145
1200
1240
1340
1440
1540
1700
1755
1850
1945
2040
2135
2325
2400
9-19-72 1000
1030
1045
1100
26.8
74.6
32.6
31.1
53.6
18.8
65.1
100.6
17.6
119.0 27.6 694
200.0 62.6 172
116.5 19.6 182
127.8 23.5 79
99.0 43.1 60
101.5 47.1 73
70.2 20.1 105
37.4
180.3 211
106
99 122
139
119
99
163
194
139
191
177
56
82
276
66
75
158
155
22
104
9,9
5,7
0
167
-------
TABLE 36
(continued)
SUMMARY OF RAW DATA FOR TANK
Dates Time
9-19-72
9-20-72
9-21-72
1130
1145
1200
1230
1245
1300
1330
1400
1430
1500
1700
1800
1900
2QOO
2045
2100
2115
2140
2145
2200
2215
2240
2245
2300
2315
2340
2345
0000
0015
0040
0045
0100
0115
0145
0200
PERFORMANCE ANALYSIS
Rainfall
(in/hr.)
2 4
BOD (mcj/D
Influent Overflow
54.0
64.0
57.0
60.7
69.0
114.5
20.0]
20.0]
20.0]
23.8]
23.8]
23.8]
16.7]
16.7]
16.7]
22.5]
22.5]
22.5]
12.5]
12.5]
12.5]
12.5]
22.8
37.3
23.8
65.0
6.3
7.5
42.5
10.0
11.3
10.0
16.0
7.5
SS (m^/l)
Influent Overflow
127
97
93
80 97
125
118
93
66
89
215 86
166 110
102 106
12.7
115 138
115
Flow
(MGD
at t)
5.7
3.9
2.1
4.3
1.7
.6
.4
.1
26.3
43.1
78.4
34.9
20.0
0
11.5
0
20.1
29.7
24.3
168
-------
TABLE 36
(continued)
SUMMARY OF RAW DATA FOR TANK
PERFORMANCE ANALYSIS
Dates Time
9-21-72 0215
0245
0300
0315
0345
0400
0415
0445
0500
0515
0545
0600
0615
0645
0700
0715
0745
0815
0845
Rainfall
Cin/hr.) BOD (mcf/1) S3 Cmg/L)
2 4 Influent Overflow Influent Overflow
16.5 161
13.0 71
11.2 59
22.5 82
15.0
Flow
(MGD
at t)
20.0
11.5
15.1
20.0
11.5
5.3
5.3
5.3
5.3
5.3
5.3
5.3
5.3
5.3
Composite Sample.
169
-------
TABLE -37
AVERAGED TANK INFLUENT AND OBSERVED TANK OVERFLOW
Date
9/12
9/13
9/17
9/18
9/19
9/20
9/21
Hour
24
1
2
3
4
5
6
7
8
1«
19
20
21
22
23
24
1
2
3
4
5
6
7
10
11
12
13
14
15
22
23
24
1
2
3
4
5
6
7
8
9
Flow
(mg/hr)
1.772
2.075
.179
.002
.195
.022
.410
.075
.006
.295
.058
.029
.145
. 118
.029
.029
.010
3.75
2.937
.883
.341
.341
.110
.206
. 118
. 199
.133
.047
.010
2.531
1. 143
.239
.416
1.125
.656
.731
.350
.220
.220
.220
.220
BOD
in
mg/1 Ibs/hr
140
60
50
10
35
0
60
120
130
300
260
150
80
50
40
40
90
60
55
70
60
50
40
90
140
70
80
10
0
20
23.8
16.7
22.5
25
25
25
25
25
25
25
25
2070
1038
74
0
57
0
205
75
6
740
126
36
97
49
9
9
7
1876
1347
515
170
142
36
154
138
116
88
3
0
422
227
33
79
121
136
152
72
46
46
46
46
BOD
mg/1
150
200
114
50
78
110
86
99
7
11
10
16
7
16
13
11
22
15
15
15
out
]bs/hr
224
325
390
1230
575
312
244
91
141
107
20
55
70
90
79
33
41
27
27
27
SS in
mg/1 Ibs/hr
300
220
225
0
110
140
200
300
300
70
75
80
85
130
230
180
125
120
110
100
180
160
115
105
95
180
120
60
60
100
100
100
100
100
100
100
100
SS
mg/1
478
305
145
320
370
320
300
110
100
140
115
150
70
60
80
out
Ibs/hr
715
1050
3570
2360
1050
910
276
1550
1000
487
1080
825
428
175
146
170
-------
concentration values the total load in pounds per hour leaving the tank was
calculated and is listed.
The detention tank performance with regard to BOD and suspended
solids removal is calculated by the tank model on an hour-by-hour basis.
The program assigns a plug-flow pattern to storm waters passing through
the tank and tracks each hourly input individually. Removals by sedimentation
are assigned to each hourly input on the basis of time of detention in the
tank for that batch of storm water at the point when it is displaced from the
tank. A variety of coefficients for the equation describing sedimentation
efficiency were investigated. All provide generally similar results, but
vary the magnitude of the predicted hourly value to some degree. The
removal equations selected for use in the model on the basis of comparison
of observed versus predicted tank discharge are:
% Removal (BOD) =25(1- e~°' 20t)
% Removal (Suspended Solids) = 40 (1 - e~ 0< 20t)
Table 38 compares actual and predicted discharges of BOD and
suspended solids from the detention tank for the four storm events which
have been analyzed. Additional storm data which could be analyzed would
permit further refinements; however, a comparison of predicted with actual
values indicates that the model developed does effectively account for but
does not duplicate the relatively complex flow and load fluctuations.
171
-------
TABLE 38
DETENTION TANK PERFORMANCE -
ACTUAL VS PREDICTED TANK DISCHARGE (Lbs/Hr)
BOD
Date
9/12
9/13
9/17
9/18
9/19
9/20
9/21
*Using
Hour
24
1
2
3
4
5
6
7
8
18
19
20
21
22
23
24
1
2
3
4
5
6
7
10
11
12
13
14
22
23
24
1
2
3
4
5
6
7
8
9
Flow
(mg/hr)
0
.312
.179
.002
.195
.022
.410
.074
.006
0
0
0
0
0
0
0
0
1.549
2.937
.883
.341
.341
. 110
0
0
0
0
0
1.692
1. 143
.239
.416
1. 125
.656
.731
.349
.220
.220
.220
.220
% Removal (BOD)
Actual
0
224
325
390
1230
575
312
244
91
141
107
20
55
70
90
79
33
41
27
27
27
= 25 (l-e~-
Model*
Prediction
0
403
225
2
197
22
395
71
5
0
0
0
0
0
0
0
0
1158
1384
405
147
139
43
0
0
0
0
0
666
496
111
65
161
92
109
57
35
35
26
31
SS
Actual
0
715
1050
3570
2360
1050
910
276
1550
1000
487
1080
875
425
175
146
20t) % Removal (Solids)
Model
Prediction
0
678
389
4
382
42
739
131
10
0
0
0
0
0
0
0
0
986
4905
1470
493
420
129
0
0
0
0
0
1268
742
132
444
1316
736
C>K6
25^
154
150
88
74
= 40(l-e-°-20t)
172
-------
SECTION IX
STORM OVERFLOW MODELING
STORM DETENTION TANK MODEL DESCRIPTION
The purpose and value of a mathematical model is that it provides
a basis for (1) predicting effects under conditions other than those
directly encountered during the test program; (2) predicting these
effects with a relatively high confidence level, in lieu of "seat-of-the-
pants" judgments, which in situations with complex interactions may
be either highly speculative or impossible to make; and (3) making
such predictions which can be accomplished simply, accurately and
rapidly allowing a wide range of individual alternatives to be explored.
The system model which has been developed utilizing information
during the data collection period of this project makes it possible to
evaluate the quantity of storm water and pollutants resulting from storm
overflows, which can be intercepted by a storm detention tank. The
model will have a general value in that evaluations may be made for
tanks of various size, serving a range of drainage areas, and a variety
of rainfall conditions.
The model which has been developed is a system simulator which
takes an input of hourly rainfall data. It calculates combined sewage
flow and quality on the basis of characteristics of the storm event.
Elements utilized in the calculation include: time of day, dry weather
flow, time since start of current storm event, drainage area, and
interceptor capacity. Provision had originally been made for modifying
both runoff and quality on the basis of storm characteristics (Intensity,
duration, interval since antecedent storm), however analysis of data
obtained in the program indicates that such a refinement is not practical.
Hourly rainfall data is the basic input to the model. Calculations
are made each hour of a storm event, whereby rainfall recorded on
the drainage area is converted to a runoff volume. This storm runoff
is combined with dry weather flow in the combined sewer system and
total sewer flow determined.
173
-------
Based on the interceptor capacity assigned by the program in
relation to total sewer flow, overflow volume either discharged to
the receiving water or entering the detention tank is determined on an
hourly basis. Concentrations of quality parameters of interest are
assigned to the combined sewer flow on the basis of time elapsed since
the start of the storm event. The program calculates for each hour of
the event the pollutant load (pounds) in the storm overflow from flow
and concentration values.
When the detention tank is incorporated, the model program
accounts for the amount of flow and contaminant load which enters the
tank each hour. When available volume of the tank is exceeded by the
overflow volume, the program calculates on an hourly basis the amount
of flow and contaminant leaving the detention tank and entering the
receiving water. The program accounts for both what is retained by
the tank and later returned to the sewer system, and what is removed
through sedimentation during those periods when overflow from the
tank occurs.
Analysis of model predictions verses observed quality variations
in overflows leaving the tank has verified the general validity of
the sedimentation model employed in the program. This model assigns
a plug flow pattern to combined sewage entering the tank and maintains an
inventory of each hour's input. Removals by sedimentation are cal-
culated using the relationships developed, based on the amount of time
each hour of input has taken to pass through the tank to the point when
it overflows and leaves the tank.
The model thus determines both flow and pounds of contaminant
(BOD and suspended solids) at hourly intervals in the overflow from the
combined sewer system, which is discharged to the river. This time
variable loading serves as input for a water quality model of the
receiving water. The effect of a specific storm overflow can then be
projected as an impact on river water quality.
The output format of the model has been arranged so that a
summary of each storm event is listed hour by hour, and totals for the
particular event are listed. This listing includes flow and pounds of
BOD and suspended solids overflowing from the collection system and
entering the detention tank, as well as similar data on whatever leaves
the tank to enter the receiving water.
The input format is arranged such that a variety of engineering
alternatives may be tested readily, in evaluating overall performance
under various conditions subject to either variation from one location
to another, or to control by engineering design. For example, the
174
-------
following parameters defined by the characteristics of the location to
be investigated may be readily adjusted in execution of the model:
(a) drainage area, (b) capacity of interceptor sewers; (c) dry weather
sewer flow, (d) sewage quality, both dry weather and wet weather
conditions, (e) runoff coefficient. Important elements which are subject
to definition or modification by engineering design may also be readily
adjusted to permit measurement of their effect. Examples in this
category are - (a) size of storm detention tank relative to the area to
be served, (b) interceptor capacity, (c) rate at which tank is pumped
out at the end of a storm event.
A fortran description of the model, together with output summaries
of a series of runs is presented in the Appendix.
MODEL COEFFICIENTS
As discussed previously the storm system model utilizes a series
of variable inputs to reflect specific characteristics of the geographical
area being investigated, and of the quality variations which may either
be general in nature or specific to the area in question. These para-
meters, which have been utilized in the analysis of detention tank
effectiveness, are listed below and discussed briefly. The data supporting
their selection has been presented in Section VIII.
1. Drainage Area - The project area in Milwaukee from which
data was obtained, covers approximately 570 acres (0. 9 square mile).
A brief discussion of the project area is included in Section III.
2. Dry Weather Flow - An average daily flow of 2 MGD has been
selected, however model input is based on hourly values over a 24 hour
morning hours, to 2. 3 MGD between 8 A. M. and 10 P.M. These minimum
and maximum flows are the average values established per a typical day,
as indicated on Figure 35. However, on specific days hourly rates of as
low as 0. 9 MGD and as high as 2. 9 MGD have been observed.
3. Rainfall - Hourly rainfall records from a local rain gauge
which accurately reflects precipitation in the area to be investigated, are
used. Records from both local rain gauges in and adjacent to the test area
and from a U.S. Weather Bureau Station located several miles to the
south of the test area have been used in the analysis performed in this study.
4. Runoff Coefficient - A runoff coefficient of 0. 5 has been
utilized in all program runs because of the great number of variables
attendant to natural phenomenon. The value of this coefficient is maintained
at a constant for practical reasons, and is not modified in the program
175
-------
with variations in storm characteristics. This coefficient reflects the
fraction of the total rainfall recorded which reaches and flows through the
sewer system. It may be considered to account for only losses due to in-
filtration or percolation into the ground, ponding and subsequent evaporation,
but also to average out discrepancies, which would be caused by variations
in intensities actually encountered over the test area as a whole compared
th what occurs at the rain gauge.
-. Interceptor Capacity- Interceptor capacity was selected as
8. C MOD for the project area, or approximately 4 times dry weather.
flow. This reflects the amount of flow including runoff from storms,
which can be retained in the sewer system without overflowing at relief
points.
6. Detention Tank Volume - The detention tank installed at the
Hu^boldt Avenue location, has a capacity of approximately 3. 9 million
galione.
7. Tank Emptying Rate - The pump out rate actually experienced
dis— 'ng the test program has proved to be quite variable. To empty
the *ank following a storm event, tank contents are pumped back into
the interceptor and will ultimately reach the wastewater treatment plant.
Th-: sewer line into which the tank contents are pumped frequently
re:- na.ined surcharged for a significant period. The tank discharge pumps
were regulated from a liquid level in this sewer line, so that during the
program the emptying rate was quite variable. Several different emptying
rales were accordingly explored during program execution (24 hr. ,48 hr.
ana ^ 6 hrs. )
8, Wet Weather Sewage Quality- Sewage quality and variations
during combined storm and sanitary sewage flows, have been
characterized by analysis of data from some 97 storm events. Detention
tank effectiveness has been evaluated in terms of BOD and suspended
solids removals. The model utilizes the following values for combined
sewage BOD and suspended solids as a function of time since the start
of the- storm event. These values are based upon data obtained during
the study period.
Time since start BOD Suspended Solids
of Storm _ mg/1 _ mg/1 _
0-30 min. 143 348
30 - 60 min. 128 270
1-2 hours 97 193
over 2 hours 87 192
176
-------
These parameters are not modified other than on the basis of
the above time scale.
9. Removal Efficiency of Detention Tank - Performance
efficiency of the tank is the result of both its ability to retain contamin-
ants by virtue of the storage volume furnished, as well as separation
by sedimentation during periods when overflow is taking place. The
former factor is accounted for by the volumetric inventory maintained
by the model. Removals by sedimentation are based on the following
relationships developed from analysis of observed results.
% BOD Removal = 25 (1-e °* 20t)
% Suspended Solids
Removal = 40 (l-e°-20tj
Where t - time (hours) that the overflow is detained in the tank
prior to overflow.
MODEL OUTPUT - ANALYSIS OF DETENTION TANK PERFORMANCE
A series of sixty (60) runs were made, using the detention tank
model, with a year's hourly rain data as input. The effect of a range
of variables was explored by modifying either the input rainfall data,
or certain aspects of the storm water collection and detention tank
system. The variables investigated by this program included the
rain gauge used as the data source, the amount of precipitation per
year, the emptying rate for the detention tank, tank size, interceptor
size, and the relative effect of bypassing the tank when full. Table 39
summarizes the test conditions and model output for each of thirty
sets of conditions. The model program was executed twice for each
set to provide data on both BOD and suspended solids, including estimated
removals.
Rain Gauge Data
As discussed in Section VIII, rainfall data was obtained from 3
independent stations during the program. Distinct differences were
noted on some storm events and on the total amount of precipitation
recorded. The significance of the difference in recorded rainfall on
the tank performance is shown on Table 40. Station 1 (Airport U. S.
Weather Bureau) compares favorably with station 2 (Holton Street Gauge
in Project Area) in rainfall statistics over the period, and projected
tank performance compares quite closely. Station 4 (Broadway Street
Gauge Adjacent to the Project Area) rainfall data is substantially d:crr it
from the otb"»r two stations and projections of tank performance are
177
-------
TABLE 39
DETENTION TANK PERFORMANCE PROJECTIONS
•vj
00
Run
1
2
3
4
5
6
7
8
9
10 *•*
H **
12 **
13 ••*
14 **
15 **
16 **
17 **
18 **
19 **
20 **
21 **
22*
23*
24*
25
26
27
28
29
30
Rain
Period Sta.
11/1/71
thru
9/30/72
11
II
11/1 m
thru
10/31/72
If
It
II
II
It
'60 (Wet)
•70 (Normal)
•63 (Dry)
•60 (Wet)
M
it
'70 (Normal)
M
ti
'63 (Dry)
I |
M
'60 (Wet)
'70 (Normal)
'63 (Dry)
'60 (Wet)
'70 (Normal)
'63 (Dry)
'60 fWet)
'70 (Normal)
'63 (Dry)
1
It
It
2
ii
H
4
u
M
1
M
M
1
tl
l<
1
II
II
1
"
ti
1
u
It
1
•1
II
1
M
II
Time
Annual To
Rain Empty
(Inches) Tank
36. 38
(11 mo.)
tt
It
35.53
(12 mo. )
u
M
29.88
(12 mo. )
u
u
40. 71
28.85
19. 10
40. 71
I 1
1 t
28.85
u
M
19. 10
M
II
40. 71
28.85
19. 10
40.71
28.85
19. 10
40.71
28.85
19. 10
24 hr.
48 hr.
96 hr.
24 hr.
48 hr.
96 hr.
24 hr.
48 hr.
96 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
48 hr.
No. of Events
No. of Total Which
Events Storm Overflow Hrs. of Hours of Vol. to
To Tank Hours Tank Overflow Rain Tank MG
72
72
72
74
74
74
60
60
60
64
70
52
t>4
64
64
70
70
70
52
52
52
64
70
52
123
123
105
43
5Z
40
244
244
244
261
261
261
224
224
224
186
226
122
186
186
186
226
226
226
122
122
122
186
226
122
450
485
356
121
142
87
18
20
20
18
19
20
13
15
17
11
10
9
33
23
9
33
24
4
25
15
2
11
10
9
16
16
10
10
7
7
56
59
69
58
61
63
48
55
59
47
43
18
109
73
31
117
82
18
57
34
5
47
43
18
84
75
29
38
29
12
594
594
594
488
488
488
464
464
464
547
584
425
547
547
547
584
584
584
425
425
425
547
584
425
547
584
425
547
584
425
179
179
179
181
181
181
144
144
144
128
124
82
128
128
128
124
124
124
82
82
82
128
124
82
173
178
115
102
93
64
= RUNS 22-23 UTILIZE BYPASS OPTION IN TANK PROGRAM
PUMP °UT RATE IS VARIED T0
-------
TABLE 39
fCont. )
DETENTION TANK PERFORMANCE PROJECTIONS
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
1000 Ibs.
BOD to
Tank
145
145
145
147
147
147
118
118
118
105
100
69
105
105
105
100
120
100
69
69
69
105
100
69
142
143
96
84
75
54
Vol.
Leaving
Tank (MG)
57
60
69
57
60
64
40
44
49
46
31
13
93
71
28
85
61
17
53
35
3
46
31
13
62
49
21
34
20
7
1000 Ibs.
BOD
Leaving Tk.
44
46
52
45
47
49
29
32
35
33
21
10
73
53
19
64
4-1
12
43
28
2
35
23
10
44
33
16
24
14
6
Percent
Removal
BOD
69.8
68.2
64. 0
69.4
68. 1
66.2
75. 6
73.2
70. 5
68. 9
78.5
85. 5
30. 7
49.4
81.7
36. 2
5?. 9
88. 0
37.3
59.3
96.4
66. 2
11.2
85. 5
69.2
76.6
83.8
71. 0
80. 9
89. 5
1000 Ibs.
S. S. to
Tank
315
315
315
321
321
321
260
260
260
227
219
150
227
227
227
219
219
219
150
150
150
227
219
150
308
314
210
181
164
118
1000 Ibs.
S. S.
From Tk.
89
93
104
93
96
101
59
64
70
65
43
20
152
109
37
135
91
24
91
53
4
76
50
21
86
67
31
48
29
11
Percent
Removal
S. S.
71.9
70.5
67. 1
71.0
69.9
68. 3
77.3
75.4
73.2
71. 5
80. 3
86. 8
33.0
52. 1
83.6
38.4
58.3
89. 2
39.0
61. 1
97. 0
66.7
77. 1
85. 6
72. 0
78.6
85.2
73.3
82. 3
90. 5
Percent
Volume
Retained
68. 1
66.4
61.4
68.5
66.8
64.6
72.2
69. 4
65. 9
64. 1
75. 0
84.1
27. 3
44. 5
78. 1
31.5
50. 8
86. 3
35.4
57.3
96. 3
64. 1
75.0
84.1
64.2
72.5
81. 7
66.7
78.5
89. 1
Tank
Capacity
(MG)
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4. 0
4.0
4.0
4.0
1. 0
2.0
6.0
1. 0
2.0
6. 0
1. 0
2.0
6.0
4.0
4.0
4.0
4.0
4.0
4.0
4. 0
4.0
4.0
Interceptor
Capacitv
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
4xDWF
2xDWF
2xDWF
2xDWF
6xDWF
6xDWF
6xDWF
-------
Table 40. SIGNIFICANCE OF SOURCE OF
RAINFALL DATA
Airport Gauge Holton Gauge Broadway Gauge
Station 1 Station 2 Station 4
Period
Inches of Rain
Number of Storms
11/01/71-
09/30/72
36. 38
11/01/71-
09/30/72
32. 33
11/01/71-
09/30/72
27.42
To Tank
To Stream
Hours with Flow
To Tank
To Stream
Storm Water (mg)
To Tank
To Stream
Percent Volume Retained
BOD Load (1,000 Ib)
To Tank
To Stream
Percent BOD Removal
SS Load (1, 000 Ib)
To Tank
To Stream
Percent SS Removal
72
20
244
59
179
60
66.5
145
46
68. 2
315
93
70. 5
67
18
231
49
168
53
68.5
137
43
68.9
299
89
70. 3
53
14
201
49
134
40
70. 0
111
30
73. 5
245
60
75.5
NOTE: Using 48 hour pump out rate and tank volume of 4. 0 mg.
Int Cap = 4 x DWF.
180
-------
significantly different. Since Station 2 and 4 are located quite close
together relative to Station 1, the implications are (a) airport rain
data, for which a substantial historical record exists can be used in
evaluation of the performance of a detention tank in the City of Milwaukee
with satisfactory accuracy, and (b) the discrepancy shown by Station 4 rain
data is most likely due to the inherent inaccuracy of the recorded data
and the variations in rainfall which occured between the various station
locations.
During the period November 1971 thru 1972, in which the Humboldt
Avenue detention tank was in service, it is estimated that a total of
approximately 68. 5 percent of the BOD and 70 percent of the suspended
solids in inflows reaching the tank were intercepted and prevented
from reaching the Milwaukee River, as indicated by this analysis.
These estimates are based on an assumed pump out rate such that the
tank would be emptied in 48 hours. In practice however, this rate was
quite variable.
Effect of Pump Out Rate
The significance of pump out rate was examined by comparing
projections of tank performance during 1972, with assumed pump out
rates of 24, 48 or 96 hours. Results of these projections are illustrated
by Figure 40. The comparison shown has been restricted to Station 1
and 2 because of th,e lack of correlation obtained at Station 4.
Pump out rate, expressed as hours of pumping required to empty
the 3. 9 million gallon detention tank is shown to have an effect on overall
tank performance. Suspended solids removal was reduced by about
4 percent (67. 5 percent removal versus 71. 5 percent) as emptying
time increased from 24 to 96 hours. BOD removals were affected by
approximately the same magnitude (65 percent versus 69. 5 percent).
While the pump out rate is an important factor, its effect within
the range shown does not appear to be major when annual performance
is evaluated. The use of a 96 hour period would increase the amount of
BOD discharged to the river by about 5, 000 pounds (45, 000 versus 50, 000
pounds/year) and suspended solids by about 11, 000 pounds (91, 000 versus
102,000 pounds/year). The longer emptying rate would therefore increase
pollutant discharge by approximately 10 percent compared with what could
be achieved with a 24 hour pump out time. On certain individual storms
however the effect would be substantial.
181
-------
FIGURE 40-EFFECT OF PUMP-OUT RATE ON DETENTION TANK
PERFORMANCE.
24 48 72
HOURS TO EMPTY TANK
Rain
Station
Period
Nov. 1-71 to Sept.
30-72
Nov. 1-71 to Oct.
31-72
Pumpout
Time(Hrs)
24
48
96
24
48
96
Percent of
Applied Load
Entering Riv.
BOD
30.2
31.8
36.0
30.6
31.9
33.8
S.S
28.1
29.5
32.9
29.0
30.1
31.7
Percent of
Applied Load
Removed
BOD
69.8
68.2
64.0
69,4
68.1
66.2
S.S.
71.9
70.5
67.1
71,0
69.9
68.8
182
-------
Effect of Tank Volume
The effect of tank size on overall performance in reducing over-
flows on pollutant loads reaching the river has been explored in a series
of runs which project performance for detention tanks ranging in size
from 1 to 6 million gallons. The comparison has been made reducing
tank volume to million gallons/squa re mile of drainage area by correcting
for the test area (570 acres =0.9 square miles). To compare relative
efficiency with various amounts of precipitation, test runs were made
using rain data from a wet, a normal, and a dry year. The years
selected were as follows:
I960 - Wet Year - 40. 7 inches
1970 - Normal Year - 28.9 inches
1963 - Dry Year - 19. 1 inches
The results of this analysis are illustrated on Figures 41 (BOD),
and 42 (suspended solids).
Detention tank size is seen to be of major significance. The
larger the tank, the greater the removal, and over the range of sizes
explored, no sharp break in performance versus size is noted. In
the larger range of tank volumes, significant differences in effect-
iveness are noted between wet and dry years. This is not the case
with smaller volume tanks where the amount of precipitation has a
minor effect on efficiency.
BOD removal efficiencies of 30 to 35 percent are projected for
tank volumes of about 1 mg/sq^mi, which increases to a range of 80
to 95 percent for tank volumes of 6 mg/sq.mi. Comparable effects
are indicated for suspended soJids.
Effect of Interceptor Capacity and Rainfall
Runs 25-30 of Table 39 shows the effect of interceptor capacity
on the combined sewer overflow using a 48 hour tank pump out time
and a 4. 0 mg tank capacity. This effect is demonstrated for a wet,
normal, and dry year of rainfall data. The data used also shows the
effect of rainfall in that the hours of rain and number of storms was
actually larger for the normal year than for the -wet year even though
the total inches is less than the wet year. Analysis indicates that this
is the result of the incidence of higher intensity storms during the wet
year. Under the weather conditions used for this analysis as the inter-
ceptor capacity increases from 2 times the dry weather flow to six times
dry weather flow, the volume of water reaching the tank is decreased
and likewise the BOD and suspended solids ]oad to the tank will be
183
-------
i-
z
UJ
UJ
-I
u_
•z.
— o<
Z U.O
< a: co
q cr<
c»
DD
co
o
ll
uu — —
O CD<
c: S2
UJ OUi
o. oc:
FIGURE 41
EFFECT OF DETENTION
TANK VOLUME ON
B.O.D. REMOVAL
PERCENT APPLIED TO TANK
REMAINING IN DISCHARGE TO RIVER
lUUTo
90%
80%
70%
60%
50%
40%
30%
20%
10%
o
\
N
N
i
\
^
\
?
t
i
V
\
\
^j
x
>s
X
\
\
\
"^*.
-^
WET YEAR (4-0.7")
NORMAL YEAR (28.9")
DRY YE
AR(I9. ")
34567
TANK CAPACITY
(MG/SOMILE)
184
-------
PERCENT OF S.S. IN TANK INFLUENT
COMBINED SEWER OVERFLOW
REMAINING IN TANK DISCHARGE
_)
^^
51
w
0<
so
51
m
2
3
5
6
8
A
— O
z
o
o
I
3J
O
m
o
3)
m
33
m
o
m
z
H
J>
13
TJ
r~
rn
CO
C
(/>
-O
m
z
o
rn
o
0
r~
CO d
C/5
3D
m
S
z
CD
•JQ
m
H
o
^^
^>
r
H
^
z
o
r~
c
3«
m
0
z
m
~n
-n
rn
o
H
O
O
m
m
z
H
O
Z
-n
0
c
:o
m
ro
-------
decreased (See Figure 43). If the load to the tank is lower the load to
(ho stream will also be lower. Percent removals of BOD and suspended
solids slightly increase as the interceptor capacity increases and loading
to the tank decreases for the normal and dry year. During the wet
year, the BOD and suspended solids loadings to the tank are almost the
same as those of the normal year, but the loadings to the stream are
about 10, 000 Ibs to 20, 000 Ibs higher during the wet year. The added load-
ings are, however, the result of less rainfall (total inches) falling over more
hours for the normal year, which reduces the possibility of tank overflow an
increases the percent volume of storm water retained. The percentage
volume being 75 percent for the "normal" year as compared to 64. 1
percent for the wet year as shown in run Nos. 10 and 11 in Table 39.
The difference in percent BOD and suspended solids removal for the
wet year for the different interceptor capacities is not as pronounced
as the normal and dry year (See Table 41), due to the fact that 6xDWF
is a relatively smaller percent increase over ZxDWF when compared
to the total storm water volumes during a wet year than a dry year.
Effect of Bypass Operation of the Detention Tank
Studies were made investigating operation of the detention tank
in a by-pass mode. That is, in a manner by which combined sewage
influent to the tank is by-passed directly to the River, when the tank is
full. This method of operation is compared to the normal operation,
whereby all flow goes through the detention tank and when the tank is
full combined sewage flows through the tank in a plug flow manner and
overflows to the River. In this case the combined sewage is settled and
chlorinated.
Due to the plug flow nature of the detention tank, during certain
storm conditions, the tank effluent can be higher in concentration of
BOD and suspended solids than the tank influent. This results from the
fact that the higher concentration of pollutants in the initial (first flush)
flow to the tank; when reduced by sedimentation, are still higher than
the concentrations of the more diluted, combined sewage which later
enters the tank.
Table 42 compares for a wet, normal and dry year the performance
of the storm detention tank for bypass and plug flow operation. This
shows that on a yearly basis the plug flow operation has a slightly
higher percent BOD and suspended solids removal. The bypass can be
more effective than plug flow for individual storm events but if operated
over the entire year would be slightly less effective because it more
often causes detrimental effects.
186
-------
FIGURE 43
EFFECT OF INTERCEPTOR CAPACITY ON
B.O.D. S3. S. LOADING TO DETENTION TANK
z
H
O
_
< CD
ico
Z 00
< _l
o
o
o
350
300
250
200
150
100
50
WET
NORMAL
DRY
WET
NORMAL
ORV
SUSPENDED
SOLIDS
B.O.D
8
RATIO OF INTERCEPTOR CAPACITY
TO DRY WEATHER FLOW
187
-------
TABLE 41
EFFECT OF INTERCEPTOR CAPACITY ON COMBINED SEWER OVERFLOW
(USING 48 HOUR PUMP OUT TIME AND TANK CAPACITY OF 3. 9 MG)
WET YEAR
PERIOD
RAIN (INCHES)
RAIN (HOURS)
NO. OF STORMS
INTERCEPTOR
No. of
Storms
00
oo Hours with
Flow
Storm.
Water (MG)
BOD (1000*)
Load
% BOD Removal
S.S. (10000)
Load
CAPACITY
To Tank
To Stream
To Tank
To Stream
To Tank
To Stream
To Tank
To Stream
To Tank
To Stream
I960
40.7
547
64
2xDWF* 4xDWF
123
16
450
84
173
62
142
44
69.2
308
86
64
11
186
47
128
46
105
33
68.9
227
65
6xDWF
43
10
121
38
102
34
84
24
71.0
181
48
NORMAL YEAR
2xDWF
123
16
485
75
178
49
143
33
76.6
314
67
1970
28.9
584
70
4xDWF
70
10
226
43
124
31
100
21
78.5
219
43
6xDWF
52
7
142
29
93
20
75
14
80.9
164
29
DRY YEAR
2xDWF
105
10
356
29
115
21
96
16
83.8
210
31
1963
19.1
425
52
4xDWF
52
9
122
18
82
13
69
10
85.5
150
20
6»DWF
40
7
87
12
64
7
54
6
89.5
118
11
% S.S. Removal
72.0
71. 5
73. 3
78.6
80. 3
82.3 85.2
66.8
90.5
* DWF (DRY WEATHER FLOW) = 2 MOD
-------
00
TABLE 42
COMPARISON OF BYPASS OPERATION OF DETENTION TANK VERSES PLUG FLOW OPERATION
( USING
48 HOUR PUMP
OUT TIME AND
WET YEAR
PERIOD
BOD
LOAD
(1000#)
% BOD
s.s.
LOAD
TO
TO
TANK
STREAM
REMOVAL
TO
TO
TANK
STREAM
Plug Flow
105
33
68.9
227
65
1960
Bypass
105
35
66.2
227
67
3. 9 MG TANK
NORMAL
1970
PF
100
21
78.5
219
43
CAPACITY)
YEAR
Bypass
100
23
77.2
219
50
DRY YEAR
1963
PE B
69
10
85.5
150
20
ypass
69
10
85. 5
150
21
S.S. REMOVAL 71.5 66.7 80.3 77.1 86.8 85.6
-------
The bypass method of oper.ition was investigated to evaluate the
effect on removal efficiency of P.OD and suspended solids. In practice
the operation of the tank in this manner, even during certain individual
storms may not be desirable because of the fact that a satisfactory
chlorine contact time would not be available if the detention tank were
bypassed. However, it -would be desirable to study and evaluate the
practicality of including this by-pass mode option in future detention
tanks at other locations. Possibly an alternate chlorination scheme
could be included in these future tanks which would allow utilizing the
by-pass mode during certain selected storms and still allow for dis-
infection of the bypassed combined sewage.
RIVER WATER QUALITY MODEL
Intensive River Water Quality Data
Four intensive surveys of a duration of greater than five con-
secutive days were conducted during the following four periods:
Survey I October 2-7, 1970
Survey II September 11-16, 1970
Survey III May 17-25, 1972
Survey IV August 16-24, 1972
These surveys consisted of the bi-hourly measurement of dis-
solved oxygen and water and air temperature at the Humboldt Avenue
Bridge (62), Cherry Street Bridge (58), St. Paul Avenue Bridge (52),
Water Street Bridge (59), and the Flushing Tunnel intake (40). The
dissolved oxygen content of the samples was determined by the Winkler
titration method. All samples on the river were taken at a depth of
approximately five feet below the water surface and the Flushing Tunnel
intake samples were taken at a depth of approximately two feet below
the water surface. During the latter two surveys, daily measurements
of the water quality parameters previously mentioned in Section VIII
at all water quality stations below the vicinity of North Avenue Dam
were also incorporated into the sampling program.
Discharge data from two U. S. G. S. gaging stations were utilized
to obtain river flow conditions during the above survey periods. Bi-
hourly measurements were available at Station No. 4-0870 which is
located on the Milwaukee River approximately 6.6 miles upstream
from the mouth and gauging approximately 98% of the total drainage
area. Daily discharge measurements were available at Station No.
4-0871. 2 which is located on the Menomonee River approximately
6. 2 miles upstream from the mouth. Flows in the Kinnickinnic River
190
-------
were estimated from a correlation based on limited data taken by
U.S.G.S. during periods of base flow (little or no surface runoff).
MT-. .M
Dry Weather Surveys
Surveys I and III are categorized as "dry" weather surveys.
No rainfall or storm sewer overflows were recorded during these
survey periods in the City of Milwaukee.
Survey I:
During Survey I, bi-monthly flows ranging from 200 cfs to 138
cfs with an average of about 158 cfs were recorded at the Milwaukee
River gaging station. Average Mean daily discharge in the Menomonee
River was about 24 cfs with a range of 32 cfs to 20 cfs. The flushing
tunnel was operated during the following periods:
Date Time of Tunnel Operation
October 2 0800-1530
October 3 No flushing
October 4 No flushing
October 5, 6 0800-1530
The observed dissolved oxygen profiles for the four Milwaukee River station
are presented in Figure 44 for Survey I. At Humboldt Avenue, the effect
of tunnel operation can be readily seen by comparison of DO levels
observed during the weekend with those observed during periods of
tunnel operation. Super-saturated DO conditions observed at this
station generally occured during tunnel operation. The utilization
of DO is observed as one proceeds downstream. A minimum DO
value of 2. 65 mg/1 was observed at Water Street.
Examination of the data shows no distinct pattern of diurnal DO
variations. Percent possible sunshine for each of the survey days in
chronological order were 80, 64, 65, 100, 42, based on climato-
logical data taken at Mitchell Field. Based on these observations
it appears that no significant algal activity occured in the Milwaukee
River during the survey period, although there are times when large
masses of algae cover the river surface.
Survey III:
During Survey III, bi-hourly flows in the Milwaukee River decreased
gradually from a high of about 360 cfs reached during the first survey
day to about 230 cfs at the end of the survey. Average Milwaukee River
191
-------
DRY WEATHER
HUMBOLDT AVE. BRIDGE
CHERRY ST. BRIDGE
ST. PAUL AVE. BRIDGE
WATER ST. BRIDGE
TUNNEL ON
0=0800 10/2/70
OCT. 2-7.I97O
FIGURE 44
TEMPORAL D.O. DISTRIBUTION
SURVEY I
192
-------
flow for the eight-day period was about 270 cfs. In the Menomonee
River, average mean daily discharge was about 36 cfs with a range of
28 cfs, to 45 cfs. The flushing tunnel was not operated at all during
the survey period. Percent possible sunshine for the survey days was
greater than 93% except for May 20, 1972, when it was 75% as recorded
at Mitchell Field.
In order to ascertain conditions at key locations in the study area,
the sampling schedule was expanded over that of the previous dry
weather survey. This expansion included intensive bi-hourly DO sampling
at North Avenue Bridge (65) above the dam. River depth at the station
limited sampling to a depth of approximately three (3) feet below the
water surface. Three new stations in the Menomonee River (81),
Kinnickinnic River (97) and the Milwaukee River Outlet to Harbor (47)
were also incorporated into the daily sampling schedule. Daily meas-
urements of all water quality parameters previously cited at all stations
below North Avenue Bridge were also conducted.
The dissolved oxygen profile observed at the North Avenue station
is presented in Figure 45. Figure 46 presents the DO profiles for
the remaining Milwaukee River stations below the dam. Very wide
variations in dissolved oxygen were observed at North Avenue Bridge
which would usually be characteristic of high photo synthetic activity.
However, the peak DO concentration generally occurred during no
light periods between 8:00 p.m. and 6:00 a.m. At Humboldt Avenue,
below the dam, these variations were attenuated, which can be attribut-
ed to the effect of the dam driving DO levels toward saturation plus DO
added by the Flushing Tunnel. At Cherry Street, wide fluctuations in
DO were again observed.
It can be postulated that the out-of-phase DO variations observed
at North Avenue are a result of the transport of DO variation caused
by photo synthetic activity occurring upstream. Under this hypothesis,
the organisms producing this activity are assumed fixed or retained
upstream (i. e. , rooted aquatic plants). Reaeration over the dam would
attenuate DO fluctuations as observed at Humboldt Avenue. The
reoccurrence of wide DO fluctuations at Cherry Street as specifically
demonstrated late in the day on May 21 and 22 indicate the resurgence
of a viable algal community. Yet during the first three days of the
survey, peak DO levels occurred during night hours. Between
Humboldt Avenue and Cherry Street, there are no recorded waste
discharges. However, during the latter three days of the survey
(Monday to Wednesday), a distinct deterioration of water quality as
shown in Figure 46 occurred at Cherry Street. It is postulated that
the inconsistent DO variations observed at Cherry Street are a result
of a combination of factors, including the transport of DO variation
193
-------
DISSOLVED OXTGEN mg/l
0=9:00 AM
m
CT
H
O
•z.
H
S
m
33
CO
-------
DRY WEATHER
HUMBOLDT AVE. BRIDGE
CHERRY ST. BRIDGE
ST. RftUL AVE. BRIDGE
WATER ST. BRIDGE
NO TUNNEL FLOW
0 = 0900 5/IT/72
MAY 17-25,1972
70 80 90 IOO 110 120 130 140 ISO 160 I
5/25
,THU.
FIGURE 46
TEMPORAL D.O. DISTRIBUTION
SURVEY IE
195
-------
as observed at Humboldt, photo synthetic activity which at times may
have been depressed and some variable unknown source of oxygen
utilization. It should be noted that at times the entire River area
south of the Dam is an estuary of Lake Michigan and that solids can
be trapped in the River area with oxygen required to stabilize the
organic solids. The data available does not allow a differentiation of
these parameters. Comparison of the DO profiles at St. Paul Avenue
and Water Street with that at Humboldt Avenue demonstrate the over-
all degradation in water quality as measured by DO in the Milwaukee
River. A minimum DO level of 0. 0 was measured at Water Street
during Survey III.
"Wet" Weather Surveys
Survey II:
Survey II, performed in September 1970, is categorized as a
"wet" weather survey. A comparison of mean daily discharge in the
Milwaukee and Menomonee Rivers and rainfall data taken at three
key rain gaging stations, one in the study area, one adjacent to the
project area and one at Mitchell Airport are presented in Table 43.
Table 43 shows that the majority of the rainfall occured on
September 15. A maximum flow of 1530 cfs was recorded in the
Milwaukee River on that date. During the survey period, the flushing
tunnel was operated during the following periods:
Date Time of Tunnel Operation
9/11/70 0800 - 1530
9/12/70 No Flushing
9/13/70 No Flushing
9/14/70 0800 - 1530
9/15/70 0800 - 1530
Percent possible sunshine for each of the survey days in
chronological order were 100, 37, 0, 0, 4 and 78%, based on climato-
logical data taken at Mitchell Field.
The observed dissolved oxygen profiles for the four Milwaukee
River stations monitored intensely are presented in Figure 47.
At Humboldt Avenue, a sharp increase in DO with peak values of 10. 15
and 9. 30 mg/1 was observed during tunnel operation on September 11
and 14, respectively. The average DO during the weekend when the
tunnel was not in operation was about 6. 5 mg/1. From the observed
data at Humboldt Avenue, it appears that tunnel operations had a
comparatively diminished effect under the high river flows recorded
196
-------
TABLE 43
COMPARISON OF RIVER DISCHARGE AND RAINFALL ACCUMULATION
Data
9/11/70
9/12/70
9/13/70
9/14/70
9/15/70
9/16/70
Mean Daily
Milwaukee Riv.
130
160
167
160
547
239
(September 1970
Survey II
Discharge (cfs)
Menomonee Riv.
27
22
64
32
628
76
Survey)
Accumulated Rainfall (in. )
Station 2
(Holton Gauge)
-
.24
.16
.24
1.26
—
Station 4
(Broadway Gauge)
-
.15
.16
.21
• 97
»
Station 1
(Airport
-
.16
.18
.22
.86
_
Gauge)
-------
WET WEATHER
HUM30LDT AVE BRIDGE
•••- ••••CHERRY ST BRIDGE
ST PAUL AVE. BRIDGE
•———WATER ST. BRIDGE
EZ3 -TUNNEL ON
0=0 00 9/11/70
SEPT 11-16, 1970
en
UJ
X
O
Q
IU
CO
5
18
16
14
12
10
10
9/11
20
30
9/ia
4
D
50
9/13
60
70
80
9/14
90
IOC 1
9/15
110!
120
C/IG
TIME- HOURS
FIGURE 47
TEMPORAL D.O. DISTRIBUTION
SURVEY H
198
-------
during September 15. The data at the three remaining stations show a
temporal deterioration of DO levels •with minimum values of 1. 8, 0. 2
and 0. 9 at Cherry Street, St. Paul Avenue, and Water Street,
respectively. It is noted that during the low DO periods, higher DO
values were recorded at Water Street than at St. Paul Avenue. This
can be either due to a higher DO level in the Menomonee or the
propagation upstream of higher DO waters from the harbor. However
no dataware available for the Menomonee or the harbor locations
during this survey period. During the initial hours of the large rain-
fall on September 15, DO levels at the last three stations increased.
It therefore appears that the initial effect of this rainfall was
to increase DO levels. This may have been due to increased re-
aeration caused by the sharp increases in flow in combination with
the flushing effect of this intense rainfall. The data also indicates
no significant algal activity occurred during the survey period.
Survey IV:
A second wet weather survey (Survey IV) was performed in
August 1972. River mean daily discharge and rainfall data taken
during the Survey are presented in Table 44.
A maximum flow of 599 cfs was recorded in the Milwaukee
River on August 23. The data demonstrates that flow in the Menomonee
River can vary considerably more than that in the Milwaukee River
under wet weather conditions. This was also observed in the previous
wet weather survey (Refer to Table 43). During Survey IV, the
flushing tunnel was operated during the following periods:
Date Time of Tunnel Operation
August 16-19 0700 to 1500 and 1900 to 0300
August 20 0600 to 1400
August 21-25 0700 to 1500 and 1900 to 0300
With this frequent operation of the tunnel, the average flow in the
Milwaukee River above the confluence with the Menomonee River
during the survey amounted to about 714 cfs, exclusive of any runoff
that occured below the Milwaukee River U. S. G. S. gaging station
located in Esterbrook Park. The flow however may have been no
greater than the amount noted because the sewers may no longer
have been discharging into the River when the peak passed the U. S. G. S.
gauge. This average flow is considerably higher than that for any of
the previous surveys. It is noted that rainfall occured during only
one of the first seven days of the survey. Therefore a major portion
of the survey was conducted under a high flow-no-rain condition.
199
-------
TABLE 44
COMPARISON OF RIVER DISCHARGE AND RAINFALL ACCUMULATION
No
O
o
Date
8/16/72
8 /1 7/72
8/18/72
8/19/72
8/20/72
8/21/72
8/22/72
8/23/72
8/24/72
Mean Daily
Milwaukee Riv.
575
535
432
439
338
351
358
364
491
(August 1972 Survey)
Survey IV
Discharge (cfs) Accumulated Rainfall (in. 1
Menomonee Riv. Station 2 Station 4
(Holton Gauge) (Broadway Gauge)
170
122
98 -
167 0.28 0.34
100
80 -
99 -
64 0. 48 0. 40
442
Station 1
(Airport Gi
-
-
-
0. 35
-
-
-
0.43
-
-------
Percent possible sunshine for each of the survey days in
chronological order were 40, 83, 79, 80, 86, 85, 93, 10, and 76
based on climatological data taken at Mitchell Field.
Intensive DO sampling was conducted at four Milwaukee River
stations and at the tunnel intake. Daily measurements of all water
quality parameters previously cited at all stations below North Avenue
Bridge were conducted. The observed DO profiles for the four Milwaukee
River Stations monitored intensely are presented in Figure 48. Due
to frequent utilization of the tunnel, its effect cannot be distinguished
from the observed data at Humboldt Avenue. Water quality as measured
by DO was generally good at Humboldt Avenue and Cherry Street with
average DO values of 6. 8 and 5. 6, respectively. However,
substantial deteriation of of dissolved oxygen was recorded at the
two downstream stations. Average DO values for the survey period
were 1.4 and 0. 8 at St. Paul Avenue and Water Street respectively
with anerobic conditions occurring at both stations for periods as long
as twelve consecutive hours. These conditions were the worst recorded
at these two stations in the monitoring program. It should be noted
that anerobic conditions were observed before, during and after the
rainfall period on August 18. It therefore appears that the zero DO
conditions recorded are probably not soley due to storm water overflow.
From the observed variations in the DO study it appears that there
was some algal activity during this survey period.
RIVER WATER QUALITY MODEL DESCRIPTION
To provide a basis for the evaluation of the causitive
elements of water quality in the Milwaukee River a two dimensional
(lateral and longitudinal) model of the study area was developed. The
basic principle upon which the model is based is the concept of
continuity of mass; that is, the total mass of each material modeled
must be accounted for, whether dispersed, transported with fresh-
water or net circulation, or reacted away if the material is non-
conservative.
The materials balance can be described in mathematical terms
by a differential equation of the following form:
Ex *2° + Ey 2^£ = Ux^-9 - Uv 3-c + S (1)
where:
201
-------
WET WEATHER
HUMBOUDT AVE. BRIDGE
CHERRY ST. BRIDGE
ST PAUL AVC. BRIDGE.
WATER ST BRIDGE
B?VWI -TUNNEL ON
0- 0900 8/16/72
— 10
g
Q
UJ
8
CO 9
Y^
' !u » '^
V V". ^
• ' \ / \\
\ r. j i vt A i. A \ i
o 10
8/16
WED
20 30 40 50 60, 70 80
8/17
THU
90 100 *IO 120 130
8/18
FRI I
8/19
SAT
e/to
SUN
8/21
MOM
140 150 160170 180 190200 210 220
TIME- HOURS
8/22
TUE
8/23
WED
8/?4
THU
8/25
FRI
FIGURE 48
TEMPORAL D.O. DISTRIBUTION
SURVEY 3Z
202
-------
c = concentration of water quality
t = time
x = distance in longitudinal direction
y = distance in lateral direction
E , E = dispersion coefficients in the longintudinal
and lateral directions, respectively
Ux, U = net advective velocity in longitudinal and
lateral directions, respectively
S = all other sources and sinks of material, C
In large scale systems, it is convenient to apply mass balance,
Equation (1), in a finite difference form. This procedure requires the
division of the water body into a series of finite, interconnecting seg-
ments, and the application of a mass balance equation to each. The
notation for the segments is shown in Figure 49. The mathematical
model which results is a system of simultaneous linear ordinary
differential equations of the form:
c.)
Ek (C - ck)]
VkKkCk + Wk
where:
C^ = concentration of water quality variable
in segment k
V^. = volume of segment k
Q-^ = advective flow from segment k to
segment j
«< kj = finite difference weight
X? kJ = i - kj
E^.- - mixing coefficient between segments
k and j - Ekj Akj
= first-order reaction coefficients in
segment k for water quality variable C
Ek- = dispersion coefficient between segments
k and j
A-^.- = cross-sectional area between segments
k and j
LIT = average of characteristic lengths of
segments k and j
W"k- = source (or sink) of variable C in
segment k
203
-------
K
FIGURE 49
NOTATION FOR FINITE SEGMENTS
204
-------
The finite difference form of Equation (2) is written for each
segment. The equations provide for direct input of material into each
segment as well as the biological utilization or production of material
within a segment. Boundary values are established based on avail-
able data. The system is considered to be time variable, so that the
finite difference equations are solved simultaneously at specified
intervals of time using an Euler numerical integration scheme. The
result of this integration is the concentration of mass calculated for
each of the segments of the model at various times.
The left-hand side of Equation (2) represents the variation with
time of the concentration C in segment k. The right-hand side of
Equation (2) is made up of four parts: (1) the mass entering or leaving
segment k as a result of the advective flow; (2) the mass dispersed into
or out of segment k as a result of turbulent mixing; (3) the first order
decay, if any, of the substance; and (4) the direct sources and sinks
of the substance for segment k.
The area of study is presented in Figure 50. A grid of 111
segments was utilized and covered the Milwaukee River from the
flushing tunnel outlet to the harbor, portions of the Menomonee and
Kinnickinnic Rivers, and Milwaukee Harbor to the existing breakwater
as shown in Figure 51. Primary emphasis was placed on the Milwaukee
River where the four major intensively monitored stations were located.
Segments in the Milwaukee River were generally 300 feet in length.
Those segments of special interest are listed below:
Segment Number Location
1 Humboldt Avenue Bridge
18 Cherry Street Bridge
35 St. Paul Avenue Bridge
40 Water Street Bridge
The variables that were analyzed, utilizing the mathematical model,
were biochemical oxygen demand and dissolved oxygen.
The general parameters used in the model were obtained from
measured data in the area or through verification procedures.
Dispersion is a measure of turbulent mixing which in the area of
concern would be caused by lake seiches, wind effects, the gological
and geometric characteristics of the system and molecular diffusion.
The dispersion phenomenon would have the net effect of driving the
system to a uniform spatial concentration level with the attenuation
of concentration gradients. A dispersion coefficient of 0. 5 ml /day
was utilized for Milwaukee Harbor and the Kinnickinnic, Menomonee
and lower Milwaukee Rivers. Above the confluence of the Menomonee
205
-------
North AVI. Dem
Humboldt Ave. Bridge
Tanral Intake—«•
St. Paul Ave.
Bridge
Water St.
Bridge
MILWAUKEE f
HARBOR /
AREA
I MENOViONrTERiVER
JONES ISLAND
WASTEWATER
TREATMENT. PLANT
Innlckinn
Ave. Bridge
KINNICKINNIC
RIVER
SCALE OF MILES
MODEL BOUNDARIES
BREAK WATER
LAKE MICHIGAN
FIGURE 5O
STUDY AREA
206
-------
FIGURES!
MODEL SEGMENTATION
207
-------
and the Milwaukee Rivers the dispersion coefficients linearly decreased
from 0.5 to 0. 0 mi /day near the vicinity of the North Avenue Dam
at Humboldt Avenue. For BOD, the deoxygenation coefficient (Kj),
utilized in all model simulations was 0. 20 at 20°C.
The reae ration coefficient in the rivers were approximated for
each segment of the model from the following:
K = (DLU) 1/2
H3/2
in which DL is the molecular diffusivity of oxygen in water (0. 81 x
10~4ft2/hour at 20°C), U is the river velocity, and H is the mean
stream depth. In the harbor segments, where velocities could not
be well defined, a more basic formulation for the calculation of K_
ct
was utilized:
Ka = KL/H
in which K-^ is the surface transfer coefficient (feet/day), which was
assumed to be 2 feet/ day in the model simulation. The aeration
coefficient was not allowed to be less than 0. 02/day under any flow
regime. Both the deoxygenation coefficient and reaeration co-
efficients are functions of temperature and may be converted from
their 20°C value in accordance with the following:
KT = K2Q (-!) T - 2°
in which Km is the rate value at any temperature, T, K-20 ^
20°C value and 1 is the empirical base of the relationship. For
KJ and K . r"} was assigned as 1. 04 and 1. 024 respectively.
A benthic oxygen uptake rate of 4 mg/m -day was measured
in the laboratories of Marquette University for samples of Milwaukee
River bottom sludges. This value was used in all river segments of
the model. A benthal uptake rate of 1 gm/m2-day was assigned to
the harbor segments.
When photo synthetic oxygen production was included in the
modeling analysis, this phenomenon was expressed as follows
assuming a sinusoidal variation of oxygen production during the day-
light hours:
P = Pmsin j) forO^t
-------
in which P is the photo synthetic oxygen production rate at time, t,
Pm is the maximum daily photo synthetic oxygen production rate, and
f is the period of sunlight (hours). Respiration of aquatic plants, R,
was incorporated into the model as a constant. Therefore, the net
effect of aquatic plant life of the dissolved oxygen of the system is
expressed as follows:
Pn =P - R
where Pn is the net production of oxygen by the aquatic plant community.
The geometry of the study area was based on U.S. Army Corps of
Engineers Map No. 743, Milwaukee Harbor. Advective flows in the river
was estimated from data taken by the U.S.G.S. at the river stations previously
cited. The flushing tunnel flow rate was estimated to be 422 cfs based on dye
studies.
The only known major waste discharge with the exception of combined
sewer overflows and sewer by-passes in the study area is the Jones Island
wastewater treatment plant. This plant provides secondary treatment and
discharges the treated effluent to Milwaukee Harbor just beyond the mouth of
the Milwaukee River. Average plant flow is approximately 160 MGD with an
average BOD5 load of approximately 20, 000 pounds per day. Other wastewater
treatment plants, bypasses, overflows, and cross connections, both known
and unknown, exist in the Milwaukee River Drainage Basin and effect the study
area water quality.
RIVER - WATER QUALITY MODEL OUTPUT AND VERIFICATION
The verification of Survey I (October 2-7, 1970) is demonstrated
in Figure 52, 53, 54 and 55 where the observed and calculated dissolved
oxygen profiles in the Milwaukee River are presented. The magnitudes
of the model parameters such as dispersion, kinetic reaction rates,
and benthal oxygen demand have already been presented in the previous
section. The flow regime utilized was based on U. S. G. S. data combined
with flushing tunnel operation.
Under dry weather conditions, the major factors contributing
to the utilization of dissolved oxygen in the Milwaukee River have been
estimated to be the benthal oxygen demand and the BODg and DO
deficit loads transported through the boundaries of the model. At
Humboldt Avenue Bridge, the DO deficit boundary conditions were
based on the actual bi-hourly data. Water quality measurements other
than DO and temperature at the four primary Milwaukee River Stations
were not taken during Survey I. The other boundary conditions of
the model were estimated in this case. This was accomplished, in
part, by evaluating all water quality data taken during the entire project.
A BOD5 boundary condition value of 5 mg/1 was assumed at Humboldt
209
-------
TUNNEL ON
OBSERVED DATA
MODEL RESULTS
0 = 0800 10/2/70
TEMPORAL D.O. DISTRIBUTION
OCT. 2-7,1970-NO RAIN
10
10/2/70
FRI.
20 30
10/3/70
SAT
40
50 60 70 80
10/4/70 10/5/70
SUN. WON
TIME-HOURS
90
100
10/6/70
TUE
110
120
10/7/70
WED.
FIGURE 52
OBSERVED V.S. CALCULATED DATA
SURVEYI-HUMBOLDT AVE.
STATION 62
210
-------
1///S/A TUNNEL ON
• OBSERVED DATA
MODEL RESULTS
0 = 0800 10/2/70
TEMPORAL D.O. DISTRIBUTION
OCT. 2-7,1970—NO RAIN
TIME-HOURS
FIGURE 53
OBSERVED VS. CALCULATED DATA
SURVEY I-CHERRY ST.
STATION 58
211
-------
23 TUNNEL ON
• OBSERVED DATA
— MODEL RESULTS
0 = 0800 10/2/70
TEMPORAL D.O. DISTRIBUTION
OCT. 2-7,1970-NO RAIN
10/2/70
FRI.
TIME-HOURS
FIGURE 54
OBSERVED VS. CALCULATED DATA
SURVEY I- ST. FWJL AVE.
STATION 52
212
-------
TUNNEL ON
OBSERVED DATA
MODEL RESULTS
0=0800 10/2/70
20
TEMPORAL D.O. DISTRIBUTION
OCT.2-7.I970-NO RAIN
18
16- -
Ujl
u>
§
8- -
6- -
V//A
10
10/2/70
FRI.
ZO TO
30 40
10/3/70
SAT.
50 60
10/4/70
SUN.
70 80
10/5/70
MON
TIME—HOURS
90 100 110
10/6/70
TUE
(20
10/7/70
WED.
FIGURE 55
OBSERVED VS. CALCULATED DATA
SURVEY I- WATER ST.
STATION 59
213
-------
Avenue during the tunnel off periods. Surveys subsequent to Survey
I indicated that the flushing tunnel water had a BODj- equal to
approximately 80 percent of that measured at Humboldt Avenue and there-
fore, the appropriate dilute BOD^ values were utilized during periods
of tunnel operation. The values used at the other boundary locations
are as follows:
Location BODq DO Deficit
(mg/1)
Menomonee River 2. 5 2. 0
Kinnickinnic River
(Kinnickinnic Avenue
Bridge) 5.0 10.0
Lake Michigan (At
breakwater) 2. 5 0. 0
The higher DO deficit value estimated for the Kinnickinnic
River boundary is based on data taken during later surveys which
showed that oxygen deficit water existed at this location. The
boundary conditions at the Menomonee River and Lake Michigan
are respresentative of background levels.
The model verification for Survey I is considered to be very
good. The boundary conditions at Humboldt Avenue accounted for
approximately 59, 33, and 26 percent of the calculated DO deficits
at Cherry Street, St. Paul Avenue, and Water Street respectively.
It was estimated that approximately 36 percent of the calculated DO
deficits at these three stations were due to the estimated bottom demand
of 4 gm/m /day in the river sections. The model also indicated that
the Jones Island Wastewater Treatment Plant had only a minor influence
on DO levels at the monitoring locations under the conditions of Survey
I. At Water Street the plant load contributed less than 6 percent of the
total calculated DO deficit while having almost no effect at Cherry
Street. The boundary conditions at Lake Michigan, the Menomonee
River, and the Kinnickinnic River had a combined effect of contributing
about 24 and 31 percent to the calculated DO deficit at St. Paul Avenue,
and Water Street, respectively, while having a minimal effect at
Cherry Street.
For Survey III, an expanded sampling schedule was initiated.
As a result, a better definition of boundary conditions based on actual
data taken during the survey was possible. Daily BOD,- values at
Humboldt Avenue varied from 6. 2 to 9. 6 mg/1 with an average value
of 7. 9 mg/1 which was utilized in the model. DO deficit boundary
conditions at this station were based on bi-hourly data. At the
Kinnickinnic River Station, daily measurements showed variations
214
-------
of 4. 0 to 8. 2 rrig/1 (average 5. 3 mg/1) for BOD and 6. 5 to 9. 9 mg/1
(average = 8. 7) for DO deficit. These data -were used to define the
Kinnickinnic River boundary conditions. For the model simulation of
Survey III, the same dispersion field, kinetic reaction rates, benthal
oxygen demands, boundary conditions at the Menomonee River and
Lake Michigan and average waste load from the Jones Island Waste -
water Treatment Plant as that used in the Survey I simulation were
applied. Background flows in the Milwaukee River were approximately
110 cfs greater than the flows recorded during Survey I. The flushing
tunnel was not operated at all during Survey III. A comparison of
observed and calculated data is presented in Figure 56 and 57 for the
Milwaukee River.
Inputs info the model for simulation of Survey III were relatively
constant. Flow in the Milwaukee River gradually increased from a
high of 360 cfs during the first survey day to about 230 cfs at the
end of the survey -with no variation due to tunnel operation. The only
significant variable input into the model was the DO variation at
Humboldt Avenue. Since Survey III was conducted under dry weather
conditions, no direct point waste sources other than the Jones Island
Wastewater Treatment Plant were defined in the study area. As a
result of the relatively constant nature of the inputs into the model,
the DO variation input at Humbolt Avenue is attenuated as one proceeds
downstream due to the effects of dispersion caused by chemical and
physical reactions in the River, and the calculated profile becomes
fairly flat at St. Paul Avenue and Water Street. The Model reproduces
the general downward trend of the data quite well but is unable to
follow the wide short-term fluctuations demonstrated by the data.
The observed data taken during Survey III demonstrated some
peculiar features. Evidence of variable algal activity is indicated.
Unfortunately, due to the original objectives of this project, the
measurement of algae was not included in the sampling program and
therefore, the simulation of the effect of algae can be no more than
speculative. Nevertheless, in order to demonstrate the possible
fluctuations in dissolved oxygen which could be caused by algae, a
model simulation run of Survey III was repeated which included an
expression of photosynthetic effects. A maximum daily photosynthetic
rate of 20 mg/l-day was assumed to vary sinusoidally over a daily
sunlight period of about 14 hours (based on climatological data taken
at Mitchell Field). This model simulation also employed average
respiration rates of 7.4 mg/l-day in all model segments. All other
model inputs were the same as the previous simulation. The resulting
calculated DO profiles are compared to the observed data in Figures
58 and 59. The observed DO variations are roughly approximated at
St. Paul Avenue and Water Street. A similarity in the pattern of
215
-------
•-OBSERVED DATA
--MODEL RESULTS
)s0900 5/17/72
12
10
2
HUM60LDT AVEi
STATION 62
I
O
a
l ' ' ill
L
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170180 190 2CC
E
10
8
6
4
2
0
5/17 5/18 5/19 5/20 | 5/21 5/22 5
WED THU TRI SAT SUN MOW
TIME-HOURS
.
t
-
...
. •• , •
/ . \ -X~\ /\^ **• '
V *;£. \S" " x:-<"x-^
.
* •
III 1 I 1 1 1 1 1 1 1 1
10 2O 30 40 50 60 70 80 90 100 HO 130 150140
5/17 5/18 5/19 5/20 I 5/21 5/22 5
WEt) THU FRI SAT ' SUN MCSI
TIME-HOURS
/23 5/24 I 5/25
ri)E WED THU
CHERRY S?1
STATION 58
~^\_^ « .
Vt.—
* »
. t •
i * ) " "i i i
ISO (60 170180 190 2CC
/23 5/ZA 5/ZS
rot; WED THU
FIGURE 56
OBSERVED VS. CALCULATED DATA
SURVEY HI.-HUMBOLDT AYE., CHERRY ST
216
-------
-OBSERVED DATA
MODEL RESULTS
0=0900 6/17/73
01
LU
i
Q
>.
V)
o
12
10-
8-
ST PAUL AVE,
STATION 52
10 20 30 40 50 60 70 60 90 100 110 120130 140 150 160 170 ISO 190 ZOO
5/17 j 5/18 I 5/19 I 5/20 I 5/21 | 5/22 | 5/23 | 5/24 | 5/25
WED THU FRI SAT SUN WON TUE WED ThU
TIME-HOURS
12
10
WATER ST.
STATION 59
0 10 20 30 40 50 60 70 80 90 100 110 120 130110 IbO 160 170180 190 200
5/17 | 5/18 | 5/19 | 5/20 | 5/21 | 5/22 | 5/23 I 5/24 I 5/25
TUP I wrn ' TMM
WED
THU
FRI
SAT ' SUN ' WON
TIME-HOURS
TUE
WFD
THU
FIGURE 57
OBSERVED VS. CALCULATED DATA
SURVEY HU-ST PAUL AVE., WATER ST
217
-------
OBSERVED DATA
MODEL RESULTS
(WITH PHOTOSYNTHESIS)
0=0900 5/17/72
HUMBOLDT AVE
STATION 62
50 60 70 80 90 100 1(0 120 ISO 140 15
10
5/17
WED
0 40 50 60 70 80 90 100 110 120 130 140 ISO 160 170 l!
iTo ibo ibo
5/18
THU.
5/19
FRI
s/ao
SAT
5/21
SUN
5/22
WON.
5/23
TUE
5/24
WED.
00
5/25
THU.
TIME-HOURS
FIGURE 58
OBSERVED VS. CALCULATED DATA
SURVEYBE-HUMBOLDT AVE., CHERRY ST.
218
-------
•-OBSERVED DATA
- - MODEL RESULTS
(WITH PHOTOSYNTHESIS)
0=0900 5/17/72
Ul
(O
X
o
1
o
w
o
__ . L-
10 20 30". 40 50 60 70 80 90 100 110 120 130 140 ISO 160 170 100 I'jO 200
5/17 I 5/18 I 5/19 I 9/ZO I 5/21 I i)/?? I 5/23 I 5/24 I 5/26
WED THU Ffil SAT SUN MOM TUE WED THO
TIME-HOURS
WATER ST.
STATION 59
0 iol
5/17
WED
. I .^.i i i i ' I I I I ' I I I I I .VI 1 ... . ,
30 10 50 60 70 00 90 100 110 IZO 130140150160 I7O|80 190 200
6/IS I 5/19 I 5/20 I 5/21 I 5/2i I 5/2SI 5/fe4 I S/25
THU ' FRI ' SAT ' SUN ' MON ' TUE ' WED ' TKO
TIME-HOURS
FIGURE 59
OBSERVED VS. CALCULATED DATA
SURVEY TE-ST PAUL AVE,, WATER ST
219
-------
fluctuation was also noted at Cherry Street although here the observed
fluctuations are substantially greater than calculated by the model.
It does appear from this exploratory analysis that short term fluctuations
in DO observed during this survey, are in a significant measure duo
to algal activity in the Milwaukee River.
However, even without the assumed photosynthetic activity, the
model does simulate the average of the observed data. Good agreement
between observed and calculated oxygen levels has been obtained for
two dry weather surveys using a consistent set of dispersion, kinetic,
and benfhal uptake parameters. The model has therefore been
approximately verified under different dry weather flow regimes and
thus provides an analytical framework for the preliminary deterministic
evaluation of water quality in the Milwaukee River.
RIVER WATER QUALITY MODEL OUTPUT AND VERIFICATION
Wet Weather
Survey IV (August, 1972) was conducted under the expanded
sampling schedule and as a result the major model boundaries could
be defined based on actual data taken during the survey. During Survey
IV rainfall was recorded on two of the nine survey days. Approximately
0. 75 inches of rain fell during the survey. Daily BOD values at Humbodt
Avenue varied from 4. 2 to 8. 4 mg/1 for the first 8 days of the survey.
During the last day of the survey a BOD of 13. 8 mg/1 was recorded.
DO deficit boundary conditions at this station were based on bi-hourly
data. BOD boundary conditions at the Kinnickinnic River were based
on daily measurements which showed variations of 4. 0 to 8. 2 mg/1
(average equals 6. 0 mg/1). All DO measurements at the Kinnickinnic
River station were 0. 0 mg/1 with a corresponding average DO deficit
of approximately 8. 6 mg/1.
Survey IV was first simulated -without the inclusion of combined
sewer overflows. The same dispersion field, kinetic reaction rates
and benthal oxygen demands, developed and used in the dry weather
simulation, were applied. Boundary conditions at the Menomonee
River and Lake Michigan as well as the approximate loads from the
Jones Island Wastewater Treatment Plant were also the same as those
previously applied in the dry weather simulations. The flow rate regime
utilized was based on U.S.G.S. data combined with flushing tunnel
operation. No flows due to urban runoff from the immediate drainage
area were included in the river flow and loads. The resulting DO profiles
are compared with observed data in Figures 61 and 62 for the four key
Milwaukee River stations. It can be noted that the observed data is fairly
well simulated, although somewhat out of phase at Cherry Street. However
220
-------
;\ substantial deviation between observed and calculated results is
noted at St. Paul Avenue and Water Street.
A lolal of 62 combined overflown witb an approximate drainage
nrea of r>,K()0 acres drain into the Milwaukee River. Combined sewers
below the North Avenue Dam serve approximately ^» "00 acres of this
total drainage area. During Survey IV, as previously stated, rainfall
was recorded for 2 of the 9 survey days with a maximum sustained
duration of 4 consecutive hours for any one storm. Approximately
0. 75 inches of rain fell during the survey. In order to evaluate the
effect of urban storm runoff on water quality in the Milwaukee River,
it was necessary to determine the overflows and corresponding loads
into the Milwaukee River for the recorded storms. The detention tank
model was used in this regard to interface with the water quality model.
The influent loads and flows to the tank from the 570 acre demonstration
area were calculated for individual storms using the combined sewer
overflow detention tank model. These estimates were then scaled to
each combined sewer based on drainage area and inputed into the
water quality model. For purposes of demonstrating the effect of
storm conditions loads from the Jones Island sewage Treatment Plant
were arbitrarily approximated to increase approximately 10 fold
over average levels during periods of runoff. This loading was not
based upon detailed data from the Treament Plant but was established
arbitrarily.
The effects of combined sewer overflow due to the rainfall
recorded on August 19 on DO levels in the Milwaukee River are shown
by the dashed lines on Figures 60 and 61. As can be seen the effect of
combined sewer overflow on DO is minimal under the conditions simulated.
This may be due to the relatively short duration of the storms occurring
during Survey IV combined with the reduced residence time (approximately
one day) in the test area due to the frequent tunnel operation. Therefore
the discrepancy between observed and calculated DO results at St. Paul
Avenue and Water Street are not accounted for by combined sewer overflow
occurring during the Survey period. The observed data also shows that
anaerobic conditions occurred at these two stations for extended periods
prior to the storm of August 19 as well as after. On the basis of the above
it is concluded that other factors are controlling quality downstream. This
conclusion is further reinforced by a comparison of the averages of data
observed during Survey III (Dry Weather) with that of Survey IV (Wet
Weather) as presented in Table 45.
The averages of BOD measured at Humboldt Avenue during both
surveys were of a comparable order. Due to higher flows occurring
during Survey IV as a result of frequent tunnel operation, reaeration in
the Milwaukee River was calculated to be more than twice that calculated
for Survey III. Under similar conditions (i. e. , boundary conditions at
Humboldt Avenue) it would be expected that lower DO
221
-------
DISSOLVED OXYGEN mg/!
ISJ
K>
ISJ
C/)
c
o
X
£§?§
-Iffl
~
I
II
0>
§c m
c/> r~
ol!*
oo*^
?3- x
^3lX
-4 >C
"==§
r^m
*5
53
?S
H 5
3 -nx
?•"§
-------
. -OBSERVED DATA
— MODEL WITHOUT URBAN RUNOFF
— MODEL WITH RUNOFF FROM
AUGUST 19th RAINFALL
0=0900 8/16/72
en
E
LU
X
o
Q
UJ
ST PAUL AVE
STATION 82
140 150 160 170 180190 200210220
20 30 40 50 60 70 80 90 100 110 120 130
TIME-HOURS
8/E5
FRI
12
O
CO
co 10
WATER ST
STATION 59
0 10 20 30 40'50 60
6/16
WED
...'.. . ' .
J—.k,. i ' i ' *• i t j i-,- i i • i i i •
W 70 80 90 100 110 120 130 140 150 160 170 ISO 190200210220
8/17
THU
8/13
FRI
8/19
SAT
8/20 8/21 8/22
SUN MON TUE
TIME-HOURS
8/23
WED
8/24
THU
8/25
FRI
FIGURE €1
OBSERVED VS. CALCULATED DATA
SURVEY EC-ST PAUL AVE., WATER ST.
223
-------
Table 45. COMPARISON OF DISSOLVED OXYGEN
CONCENTRATIONS DURING SURVEYS III AND IV
Survey III (May 17-25, 1972)
Station DO DO Deficit
Humboldt Avenue 7.22 1.44
Cherry Street 4.35 4.06
St. Paul Avenue 2. 57 6..06
Water Street 2.26 6.39
Average Flow =270 cfs (at Humboldt)
Average K^ = 0.22/day (segments -1-40)
Average Ka = 0. 09/day (segments 1-40)
Survey IV (August 16-24, 1972)
Humboldt Avenue 6.84 1.43
Cherry Street 5.57 2.61
St. Paul Avenue 1. 38 6. 56
Water Street 0. 76 7. 12
Average Flow =714 cfs (at Humboldt)
Average K^ = 0. 25/day (segment 1-40)
Average K = 0.22/day (segment 1-40)
224
-------
deficits would occur during Survey IV. At Cherry Street this is the
case and therefore model simulation of both surveys at this station,
with a consistent set of parameters is achievable. However, DO
deficits observed at St. Paul Avenue and Water Street are lower for
Survey III with anaerobic conditions occuring frequently. This
indicated the presence and influence of an oxygen demanding source
during Survey IV which did not influence the model results during
Survey III. Model simulation without the inclusion of these influences
would result in the discrepancies encountered. However, it has
been demonstrated that the combined sewer overflow occuring during
Survey IV was of too short duration to influence or cause any major
degradation in DO levels.
Alternative possibilities are suggested. The first is the possibility
of bottom scour due to the high velocities induced by the combination
of the increased Milwaukee River flow (average greater than 400 cfs)
combined with tunnel operation (452 cfs) during Survey IV. Such
scouring or stirring of the bottom could result in an increase in BOD
in the overlying water column from the suspension of particles and the
exposure of a fresh surface •which would increase sediment surface
oxygen demand. Such a phenomenon would be quite complex and any
analysis would be very speculative -without further data.
Another alternative possibility would be an increased load due
to the Menomonee River. For the model simulation the boundary
conditions in the Menomonee were assumed to be 2. 5 and 2. 0 mg/1
for BOD and DO deficit, respectively. The Menomonee River station
at the South Second Street Bridge exhibited zero DO levels for 6 of the
9 survey days. This indicates that the Menomonee could contribute
somewhat to the deficits observed at the two downstream stations.
Quantification of the effects of the Menomonee will be presented in the
next section.
The final alternative would be some point source such as a raw
waste discharge in the study area. However, no such discharge has
been identified.
An additional verification analysis was performed on data obtained
during Wet Weather Survey II (September 1970), even though boundary
conditions, were less well defined than they were with Survey IV, which
has just been discussed. The lack of significant impact from the small
storm events in Survey IV, makes it desirable to investigate the
effect of a rather substantial storm event which occurred near the end
of Survey II. Sufficient confidence in the basic model parameters
developed by the verification analysis previously discussed, permitted
application of the river quality model to conditions observed during this
225
-------
survey, wherein some assumed values for boundary conditions were
necessary.
During Survey II, a major rainfall occurred in -which approximately
1" of rain fell, with the bulk of the storm covering a span of about
5 hours, during the morning hours of September 15th. Very brief
rainfalls, of small intensity had occurred earlier in this survey. This
data has been described in a prior section of the report.
For model simulation of this survey, DO deficit boundary conditions
at Humboldt Avenue were based on bi-hourly data. However, water
quality measurements other than DO and temperature at Milwaukee
River stations were not taken during the survey. Therefore, as with
Survey I, other boundary conditions of the model were estimated base,
in part, on all water quality data taken during the entire project. A
BOD boundary value of 6. 5 mg/1 was utilized at Humboldt Avenue which
is representative of the summer average (refer to Table 8. 6, Section
VIII). Appropriate BOD dilutions were applied during periods of tunnel
operation as previously described. At the Menomonee River boundary,
a BOD of 2. 5 mg/1 was applied except for the period of major runoff
on September 15, when a peak value of 6. 5 mg/1 was utilized. A constant
DO deficit of 2. 0 mg/1 was applied at this boundary. At the Kinnickinnic
River boundary, 5.0 mg/1 and 10 mg/1 were utilized for BOD and DO
deficits respectively. DO saturation was assumed at the Lake Michigan
boundaries with a background BOD level of 2. 5 mg/1. The same
dispersion field, kinetic reaction rates and benthal oxygen demands
used in the other survey simulation were applied.
The detention tank model was again utilized to evaluate the influent
loads and flows into the Milwaukee River based on rainfall data secured
from the gauges in the study area. These estimates were sealed to
each combined sewer based on drainage area served, and inputed into
the model outlet below the North Avenue Dam. The storms occurring
on September 13 and 14 were eliminated in the analysis due to their
relatively short duration (one hour) and low intensity.
The rate flow regime utilized was based on U.S.G.S. data (refer to
Table 43) combined with flushing tunnel operation and urban runoff.
During the weekend, when the flushing tunnel was shut down, the average
flow in the Milwaukee River above the confluence with the Menomonee
was about 160 cfs. An abrupt increase in flow was recorded on
September 15 with an estimated peak of 1955 cfs (inclusive of tunnel
flows) occuring at 10:00 A.M. During this date the average daily flow
in the Menomonee was greater than that of the Milwaukee River as
recorded at the U. S. G. S. gaging station. A peak value of 3730 cfs was
estimated in the Milwaukee River below the confluence with the
Menomonee River.
226
-------
A comparison of observed versus calculated delta is presented
in Figures (>?. through <)c>. Overall verification al Cherry Street is good,
although the model overestimated the D. O. somewhat clviring the week-
end when thr flushing tunnel was shut down. The model does accurately
track the general pattern and trend in dissolved oxygen levels under
the influence of the flushing tunnel and the substantial surge in flow which
occurred on September 15th in association with the substantial storm
event. In this verification run, the model projection was extended
beyond the termination of observed data, to project the residual
influence of the storm event before it is displaced from the area of the
river under study.
Verification at St. Paul Avenue and Water Street is good, until
the very end of the test run. Both reflect the abrupt increase in DO
levels on September 15th. This is attributable to increased reaeration
rates associated with the tenfold increase in river flows at that time.
As flows return to lower levels following the storm surge, significant
deviations are once again observed between observed and calculated
data. The discrepancy cannot be explained in terms of BOD concentrations.
The combined sewer overflows, which occurred, do in fact result in a
significant increase in calculated BOD levels based on model projections.
An increase from 5 to 12 mg/1 is calculated for Cherry Street; from 3
to 10 mg/1 at St. Paul; and from 3 to 8 mg/1 at Water Street. However
at the high flow rates which prevail, there is insufficient residence
time in the River for these reactions to occur, before these combined
sewer overflows are flushed from the system. The BOD load from the
combined sewer overflows will exert an oxygen demand, but it: will
occur outside the section of the river being studied.
The close approximation of observed versus projected DO levels,
throughout the survey up to the point where high flows were sustained,
provides an additional basis for attributing the discrepancy to some
phenomenon associated with scour of bottom sediments. At any rate,
the implication is again raised that there are other factors at work
in addition to those utilized in the model. The same substantial deviation
is observed with the higher flow regimes as noted in Survey IV at high
flows. The effect of storm discharges cannot account for this deviation.
During Survey II, the water quality at Water Street was somewhat
better in terms of DO, than occured at St. Paul. This may be due to
the Menomonee River exerting a positive influence at this time,
possibly because of higher' DO levels in that River. The quality
in the Menomonee is more variable and apparently depends to a major
degree on storm events, possibly due to its smaller drainage area.
227
-------
DISSOLVED OXYG_EN-mg/l
01 CD o 53
V)
3
m
Oo
i
i
c
CD
O
m
JJ
o
c
m
ro
oo
2
5!
-------
DISSOLVED OXYGEN mg/l
VO
01
CD
cr
-o
*-C
rn
o
s 2
3D
O
o
HID *
^.O,
"4
8*
9i §
:' ogo w ^
Jo °pSSi
or*
o
•33
-------
TUNNEL ON
OBSERVED DATA
MODEL RESULTS
MODEL WITH STORM TANKS
0=0900 9/11/70
TEMPORAL D.O. DISTRIBUTION
SEPT. 11-17,1970
16
7//A
14--
-v 12--
s io+-
>-
X
o
O 9
Ld
8
CO
o
V/A
9/11
FRI
120 130
9/16
WEO.
140
9/17
THU.
TIME-HOURS
FIGURE 64
SURVEY n-ST. PAUL AVE.
STATION 52
230
-------
-TUNNEL ON
• -OBSERVED DATA
MODEL RESULTS
16
14
10
MODEL WITH STORM TANKS
0=0900 9/11/70
TEMPORAL D.O DISTRIBUTION
SEPTEMBER 11-17 1970
z
UJ
X
o
o
UJ
8
-------
The model projection applied calculated combined sewer overflow
loadings over the section of the River between the Dam and the
Menomonee River, as has been described. An additional model run
was made, assuming that each of the overflow points which contributed
a load was equipped with a detention tank of similar unit size (million
gallons volume /square mile) to the Humboldt Avenue Demonstration
Tank. In this latter case, only the flow and loads which would escape
the storm detention system were applied as loads. All overflows were
retained except for the last hour of the storm on September 15th when
57% of the hourly rainfall was returned. Table 46 indicates projection
storm flow and BOD removed by the project detention tank. Table 46
indicates the scale up of this data to provide loadings to the river
from a series of tanks intercepting discharges to the section of the
river which was modeled. Output projections from this run are shown
on Figures 64 and 65 as dashed lines.
Again, because of the high flows and short residence in the river
the effects of the detention system on this section of the river are
minimal. Slight reductions in river DO are indicated because of the
reduced river flow resulting from detention of combined sewer over-
flows. However, combined sewer overflows do contribute suspended
solids (which over the years adds to the bottom deposit oxygen demand),
coliform organisms, BOD, COD, etc. The deleterious effect of these
will be greater after the water leaves the river.
232
-------
Table 46. MILWAUKEE RIVER - DETENTION TANK
RAINFALL ANALYSIS FOR PROJECT AREA 570 ACRES
Date
9-12-70
9-12-70
9-12-70
9-13-70
9-13-70
Totals
9-14-70
9-14-70
9-15-70
9-15-70
0-15-70
9-15-70
9-15-70
9-15-70
9-15-70
Hour
2100
2200
2300
0100
0200
1700
1900
0100
0300
0400
0500
0700
0800
0900
Lb BOD
To Tank
81
0
49
0
0
130
52
52
39
34
370
1267
379
332
837
Lb BOD
To River
0
0
0
0
0
0
0
0
0
0
0
0
0
0
300
Flow Rate
(cfs) To River
2.7
0
2.5
0
0
6. 2
2.7
2.7
2.0
1.7
18.9
64.9
19.4
17.0
42. 8
Flow Rate
(cfs) To River
0
0
0
0
0
0
0
0
0
0
0
0
0
0
18.4
Totals
3366
300
172. 3
18.4
233
-------
Table 47. PROJECTED COMBINED SEWER OVERFLOW
LOADS FOR SEPTEMBER 15, 1970 STORM
Combined
Model
Segment
Upstream
4
10
16 16
20
23
24
26
28
29
31
34
35
37
38
40
41
Drainage
Area (Acres)
290
719
307
408
407
95
53
96
118
96
134
9
67
93
37
9
35
Drainage Area
Scale Factor
Drainage Area of Tank
5. 09
1. 26
0. 539
0. 716
0. 714
0. 167
0. 093
0. 168
0. 207
0. 168
0. 235
0. 016
0. 118
0. 163
0. 065
0. 016
0. 061
Sewer
Loads
Overflow
(Ib/hr)
Without With
Tanks
1055
451
599
598
140
78
141
173
141
197
13
99
136
54
13
51
Tanks
378
162
215
214
50
28
50
62
50
71
5
35
49
20
5
18
Loads Applied to River at Model Time 25. 01 to 26. 00 Hours
234
-------
SECTION X
COMBINED SEWER OVERFLOW DETENTION TANK APPLICATION
AND MILWAUKEE RIVER QUALITY
DISCUSSION OF DETENTION TANK CAPABILITIES IN CONTROL OF
COMBINED SEWER OVERFLOW
Combined sewer overflow detention tanks have been shown by this
program to be effective in preventing a large proportion of the
contaminants found in combined sewage from entering receiving waters.
Removal achieved is directly related to tank size in relation to the
drainage area served. Removal efficiencies will vary significantly
with individual storm events and will be particularly influenced by
storm size and pattern, and by whether there has been sufficient
time to empty the tank from the previous storm event. However, when
performance over a year's time is compared an evaluation of longer
range effectiveness can be made.
Removals for BOD and Suspended Solids can range from
approximately 30% to in excess of 80% as tank size is increased from
1 to 6 million gallons volume per square mile of drainage area. The
removal efficiency is slightly better in dry years compared with wet
years for all sizes, and the differential between performance in wet
verses dry years increases with tank size. These relationships are
illustrated by Figure 66.
Figure 67 illustrates the decreasing efficiency per unit volume
as tank size increases. Tank sizes of 1 MG/ sq mi show a unit efficiency
(% Removal/MG/Sq Mi) of about 30% for both BOD and COD. This
unit efficiency decreases to approximately 15% removed per unit volume
for a tank providing a volume of 6 MG per square mile of drainage
area.
An analysis of the volumetric efficiency of the combined sewer
overflow detention tank, defined as the quantity of combined sewer
overflow retained by the tank, to the total quantity discharged to it,
has been made. Figure 68 illustrates the effect of tank size on the
235
-------
100
FIGURE 66
STORM WATER BOD AND SUSPENDED
SOLIDS REMOVAL AS A FUNCTION
OF TANK SIZE
TANK VOLUME (MG PER SQ.MI.)
236
-------
FIGURE 67
UNIT SIZE REMOVAL EFFICIENCIES
FOR COMBINED SEWER OVERFLOW
DETENTION TANKS
DRY
YEAR
\
\
Lul
Q.
O
cr
LJ
o.
-I 20
o
:E
LJ
I-
•z.
LJ
O
cr
LJ
CL-
WET
YEAR
-BO.D.
S.S.-
B.O.D.-
8345
TANK SIZE (MG PER SQ.MI.)
237
-------
lOOr
AiNED
UJ
S
O
5
a:
o
U.
O
UJ
o
cc
UJ
OL
60
FIGURE 68
VOLUMETRIC EFFICIENCY
FOR COMBINED SEWER OVERFLOW
DETENTION TANKS
A
'
iv
23456
TANK VOLUME (M6 PER SQ Ml.)
238
-------
total interception and detention of storm water overflows. The removal
of soluable contaminants present in combined sewer overflows will be
directly related to the interception and retention of storm water overflows
which carry them. Thus, the volumetric efficiency of the detention
tank shown by this figure also represents the removal efficiency for all
contaminants not subject to additional removal by sedimentation of
chlorination.
A comparison of volumetric efficiency with total removal
efficiency for BOD and Suspended Solids provides an instructive insight
into the relative significance of sedimentation as a supplement to re-
tention in the overall performance of a combined sewer overflow-
detention tank. Figure 69 plots calculated removal efficiency vs. tank
size based on rainfall data froma.year of normal precipitation at
Milwaukee (Year 1970 - 28.85" rainfall).
The combined sewer overflow detention tank model calculates
only very minor additional removals of BOD and Suspended Solids by
sedimentation, compared with the amounts removed strictly by virtue
of the retention of storm overflows. A similar comparison for a wet
and a dry year is presented by Figures 70 and 71. All comparisons
shown only a relatively small increase in removal of BOD and
Suspended Solids compared with removal due to retention. The greatest
effect of sedimentation occurs during a -wet year (Figure 70) where
removals are increased an additional 5% or so.
It may be concluded that some advantage exists in designing a
detention tank with a layout which permits effective sedimentation to
occur. However, where such design considerations impose a
significant additional cost factor, economic justification may not be
present. Where space or construction constraints dictate, tank shapes
which do not lend themselves to effective sedimentation efficiency can
be validly considered because of such a tank's value in retaining pollutant
flow. The major consideration in designing combined overflow detention
tanks is volume and cost effectiveness considerations should concentrate
on maximum volume which can be achieved at a given cost.
The combined sewer overflow detention tank design provides for
chlorination of all water which overflows the tank and discharges to
the river. Effectiveness of this operation in the destruction of total
and fecal coliform bacteria will be related to chlorine dosage applied
and the chlorine demand of the waste. Two additional factors will be
of major influence on the effectiveness of this operation. The wide
variability in coliform counts in the combined sewerage is important
since the kill effected is in terms of percentage reduction of those
239
-------
PERCENT REMOVAL
o
EW
2
m
o
s*
p
Z
o>
ro
o
3) O
is
2 i
m m
z -<
o 2
-< o
00
o
o
o
DD
O
O
ro
oo
<
O
o
o
§
c
Z
m
m
?o
w
p
PERCENT REMOVAL
ro
o
O)
o
o>
o
o
a
m
o'
> •
o
<
\
^
\
-------
100
FIGURE 70
COMPARISON
CONTAMINANT RETENTION
VS
VOLUMETRIC RETENTION
WET YEAR
80
o
UJ
z
< 60
LJ
DC
\-
LU 40
O
IT
LJ
Q.
BOD
S.S
-VOL
UME
2Q
0
TANK VOLUME (MG PER SQ.MI.)
241
-------
FIGURE 71
COMPARISON
CONTAMINANT RETENTION
VS
VOLUMETRIC RETENTION
DRY YEAR
100
TANK VOLUME (MG PER SQ.ML)
242
-------
organisms initially present.
Flow rates fluctuate very rapidly and over very wide ranges
during some storm events, such that contact times will often be
very short.
From the data available, it is not possible to provide adequate
definition of this aspect of tank performance. Quality data for tank
overflows record total coliform counts ranging from essentially zero,
to several million per 100 ml. Fecal counts behaved similarly, though
with lower total numbers. The primary difficulty preventing a meaning-
ful analysis of coliform data, was the deficiency in flow rate data during
the times when the detention tank was overflowing in the initial stages
of testing resulting from deficiencies in flow measuring equipment.
A preliminary assessment was made utilizing overflow quality
data, and the output of the storm tank model to provide the order of
magnitude of the flow rate at times when the various coliform counts
were observed. The results of this analysis, which must be con-
sidered a relatively low order approximation only, are presented in Figure
72. This plot compares coliform counts in the final discharge from the
tank with the throughput rate expressed in terms of million gallons/
hour.
FACTORS INFLUENCING WATER QUALITY IN THE MILWAUKEE RIVER
Analysis of long term quality data for the Milwaukee River
described in an earlier section of this report, indicated that water
quality in the river section downstream of the North Avenue Dam is
influenced to a substantial degree by conditions in the drainage basin
upstream of the test area. The intensive surveys and the analysis by
mathematical model have made it possible to investigate the impact of
factors which make their presence felt downstream of the dam. In
this analysis, upstream influences in effect isolated by the technique
of establishing the quality of water entering the area of interest, along
with its variations, as a boundary condition for the model. The effects
of the superimposed local conditions can then be identified and quantified.
These above mentioned factors are of course variable with time.
The verification analysis which has been presented in the previous
section has indicated by the match achieved between predicted and
observed data, that the river quality model provides an accurate
quantification of the major influences on dissolved oxygen levels.
A good verification was achieved for dry weather data surveys under
different flow rate regimes and flushing tunnel operation schedules. Wet
Weather survey data verified the model as far downstream as Cherry
Street. The significant divergence between observed and predicted
response at the lower end of the river (St. Paul Avenue and South
243
-------
8
o
u.
UJ
o
UJ
tr
o
o
oc.
UJ
o.
o
u.
o
H '
e
•i
4
3
2
100
I
5
K
-1
4
S
2
10
5
ft
7
6
S
4
3 .
1
t...
-
6
5
4
3
2
OJ-
0.01
® TOTAL COLIFORMS
A FECAL COUFORMS/
f2
i
i
\
A
u
_X"
X
s^
s
®
A A,
a a
A
yt^_
/
4
9 :
"t
7
/
/,
-^
^x
S
****
^ ,
2*
x^
yT
/
RANGE OF TOTA
TOTAL
COLIFORM
(
L
, COLIFORMS IN ^
' CHLOR
\
\
T
?
/
/
/
f
X
s*
,x
s
INATEO
LOW
/
/
jf
/
f
S
/
^
X"
r
jr
\JS
y
^f
^
"*"•"" ^y
^f
^
^x
^r
^
^
^^^
-f*~
' FECAL
COLIFORM
RANGE OF FECAL COLIFORMS
\IN CHLORINATED OVERFLOW
V"
\
\
\
\
X
r
.
X '
v
X,
x^
X
S
y
r
-*""
^
S'
^x1
_^~
^
^
— -^
J^~
*-&
A
^--^
A
^
^^
o
TANK INFLUENT FLOW RATE ( MILLION GALLONS/HR.)
FIGURE 72
COLIFORM COUNT IN DETENTION TANK OVERFLOW
VS.
TANK THROUGHPUT RATE
244
-------
Water Street) is attributed not to inadequacy in the basic quality
model, but rather to external influences which have not been
identified previously, and which the model did not incorporate.
From general observations made during the course of the
project, and particularly from a review of quality data obtained
during the study, the following factors have been identified as
exerting significant influence on water quality in the section of the
river downstream of the dam:
(a) Upstream quality variations in response to loadings,
storm runoff, and flow changes.
(b) Significant algal and macrophyte activity, occuring
at times both above the dam, and in downstream
sections.
(c) Flushing tunnel operation.
(d) Benthal oxygen demand from accumulated sediments.
(e) The Menomonee River.
(fj Urban Storm runoff.
^' Combined sewer overflow
The river quality model has been utilized in performing a series
of application runs to study the significance of these factors. The
results of these application studies are discussed below:
Upstream Conditions
Both observed data from long term river analysis and intensive
surveys, as well as model results indicate that upstream conditions
provide a major load input to the area of the river under investigation.
Relatively high levels on contamination exist in the Milwaukee River
before its waters reach the study area. At times when oxygen
resources are high and BOD levels are low in this incoming water,
better quality conditions will prevail in the test area. Total river
flow has a major influence on oxygen resources both because of increased
reaeration rates at high velocity, and because of reduced residence
time in the test area, for oxygen demanding substances to exert their
full effect. As an example of the magnitude of this impact, it has
been calculated that water quality conditions in the water reaching
Humboldt Avenue accounted for approximately 59 percent, 33 percent,
and 26 percent of the oxygen deficits at Cherry Street, St. Paul Avenue
and Water Street, respectively, during dry weather Survey I.
Photo synthetic Activity
Activity by algae and macrophytes is at times intense. Diurnal
fluctuation in dissolved oxygen in the order of 7 mg/1 were observed
245
-------
in May 1972 at the North Avenue Bridge, upstream, of the dam. These
variations, as well as the algal cells are transported into the test
section of the river and at such times significantly influence oxygen
levels. Daily variations in dissolved oxygen of about 3 mg/1 occur
in the lower reaches of the river, such that oxygen levels are depres-
sed to or close to zero at times when concentrations would otherwise
be several mg/1 higher.
Flushing Tunnel Operation
The flushing tunnel is a controllable element, and one which
exerts a significant influence on quality.
The model verification for Survey I was used as a basis of com-
parison to identify the effect of flushing tunnel operation. During
Survey I the tunnel was operated three out of the five survey days.
Two alternate tunnel operation schemes were studied using the water
quality model. This necessitated a reevaluation of boundary conditions
at Humboldt Avenue and appropriate changes in the flow regime.
For the first scheme, it was assumed that the tunnel was not
operated at all during the survey period. DO deficit at Humboldt
Avenue was assumed constant and equal to the average of the data
observed at Humboldt Avenue during the weekend for Survey I when
the tunnel was not in operation. A BOD boundary condition of 5 mg/1
•was assumed at Humboldt Avenue. The second scheme approximated
the effect of daily operation (7 day operation) of the tunnel for 7. 5 hour
(0800 - 1550). Boundary conditions at Humboldt Avenue during the
weekend were modified to reflect this daily tunnel operation based on
data observed during those days -when the tunnel was operated. All
other input parameters to the model were the same as those applied
for the original verification run. The effectiveness of flushing tunnel
operation can be readily seen by comparison of the calculated profiles
at the four key Milwaukee River stations as presented in Figures 73
through 76. It is estimated that tunnel operation for 7. 5 hours per day
every day improves DO levels by about 3. 5, 2. 0, and 1. 3 mg/1 at
Cherry Street, St. Paul Avenue, and Water Street, respectively.
Also note the calculated improvement in DO levels as projected under
daily tunnel operation as opposed to weekend shut-down as was
predicted during Survey I.
The favorable influence of the flushing tunnel is the result of both
the additional oxygen resources introduced by virtue of the high DO
content of the water delivered by the tunnel, and also because of the
higher atmospheric reaeration rate in the river, which the tunnel
operation induces because of increased velocities. However, our
246
-------
MODEL VERIFICATION
NO TUNNEL OPERATION
—-DAILY TUNNEL
TUNNEL ON
0= 0800 10/2/70
20
18!
16
en
Ld
C3
X
o
o
LU
O
co
co
o
14
10
10
10/2
FRI
20
30
10/3
SAT
40 50 60 I ro 80
I 10/4 10/5
I SUN I MON
TIME-HOURS
90
100
10/6
TUE
110
120
10/.'
Note: The significant Deviation between the
Daily Tunnel Curve and the Model Verification
Curve are due to the fact that the tunnel was
only actually operated 3 of the 5 days of Survey
FIGURE 73
MODEL APPLICATION AT
HUMBOLDT AVE-SURVEY I
STATION 62
247
-------
MODEL VERIFICATION
NO TUNNEL OPERATION
DAILY TUNNEL OPERATION
TUNNEL ON
0 = 0800 10/2/70
20
18
16
o. 14
X
o
a
LU
O
V)
Q
10
4
2h
—L_
10
yj~*'v
&<£rf*.'^
10/2
FRI
_L
20 30 40 50 60 70 80
10/3 io/4 io/5
SAT SUN MON
TIME-HOURS
90
iOO
10/6
TUE
no
120
IQ/7
weo
Note: The significant Deviation between the
Daily Tunnel Curve and the Model Verification
Curve are due to the fact that the tunnel was
only actually operated 3 of the 5 days of Survey I.
FIGURE 74
MODEL APPLICATION AT
CHERRY ST.-SURVEY I
STATION 58
248
-------
MODEL VERIFICATION
---- NO TUNNEL OPERATION
-- DAILY TUNNEL
TUNNEL ON
0=0800 10/2/70
20
18-
16 -
10
o>
£
l
LU
>
X
O
Q
UJ
O
-------
•MODEL VERIFICATION
NO TUNNEL OPERATION
DAILY TUNNEL OPERATION
TUNNEL ON
0 = 0800 IO/2/70
20
18
16
^ 14
o>
£
Z l2
U)
g 10
b
in
<2 6
Q
^" ^
10
10/2
FRl
20
3O
10/3
SAT
40
50 60
10/4
SUN
70 80
10/5
MON
90
100
10/6
TUE
120
10/7
WED
TIME-HOURS
Note: The significant Deviation between the
Daily Tunnel Curve and the Model Verification
Curve are due to the fact that the tunnel was
only actually operated 3 of the 5 days of Survey
FIGURE 76
MODEL APPLICATION AT
WATER ST. —
STATION
SURVEY I
59
250
-------
analysis suggests that indiscriminate tunnel operation may have an
adverse effect at times. During Survey IV for example the flushing
tunnel was operated 16 hours/day on all days at a rate of 425 cfs.
River flows, as recorded above the North Avenue Dam, averaged
greater than 400 cfs. Total flows in the river below the dam therefore
exceeded 825 cfs with sustained peaks of more than 1, 000 cfs. Com-
parison of the data of Survey III (Average flow equals 270 cfs) with
that of Survey IV (refer to Table 45) show that higher DO levels existed
for Survey III at St. Paul Avenue and Water Street. As described in
the wet weather verification section this indicated introduction of
additional sources of DO deficit during Survey IV. A distinct possibility
is that sufficiently high velocities, either by scouring or stirring up
bottom sediments, increase oxygen requirements associated with
bottom sediments. Therefore, the beneficial effect of tunnel operation
under high flow conditions is somewhat questionable under present conditions.
Additional investigation would be required before a definite conclusion can be
made.
Sludge Deposits
Visual observations of active gasification, as well as model
verification using the relatively high benthal oxygen demand of 4 mg/
m /day, confirms the significance of accummulated sediments on water
quality. An analysis was performed to define the effect of the elimination
of bottom sludges. The verification of Survey III was used as a basis
of comparison. The benthal oxygen demand of 4 gm/m2/day was eliminated
from all river segments in the model. All other model parameters
were unchanged. The projected improvement in DO levels under the
conditions of Survey III can be seen in Figures 77 and 78. The
elimination of bottom deposits has increased impact as one proceeds
downstream with an improvement of approximately 1. 8 mg/1 calculated
at Water street. Under lower flow conditions the elimination of the
estimated benthal demand would result in an even greater improvement
in DO quality. Figure 80 presents a plot of the estimated DO deficit
at Water Street due to a bottom demand in all river segments of
4 gm/m2/day for a range of the average flows observed in the study.
This bottom demand may increase if scouring action or other
disturbance occurs. These results are based on a steady state analysis.
The Menomonee River
The Menomonee River has a variable influence on the Milwaukee
River. Its impact may be relatively slight during dry weather periods
with low flows and reasonably good quality. Under at least some wet
weather conditions, its influence is very significant.
251
-------
MODEL VERIFICATION
MODEL WITHOUT BENTHAL
OXYGEN DEMAND
0=0900 5/17/72
12
10
UJ
w
x
o
Q
UJ
O
to
HUMBOLDT AVE
STATION 62
0 10 20 30 40 50 6O 70 60 9O 100 NO 120 130 140 150 160 170 ISO 190 200
5/17 I 5/18 | 5/19 | 5/20 | 5/21 | 5/22 | 5/23 | 5/24 | 5/25
WED THU FRI SAT SUN MON TUE WED THU
TIME-HOURS
il i I I l I I 1 I 1 I 1 J 1 1 1 1 1 __,
0 10 20 30 40 50 60 70 80 90 100 110 120 130(40150160 170 180 190 200
5/17 5/18 5/19 5/20 5/21 5/22 5/23 5/24 5/25
WED THU FRI SAT SUN MON TUE WED 1 H'J
TIME-HOURS
FIGURE 77
MODEL APPLICATION AT
HUMBOLDT AVE. a CHERRY ST.- SURVEY
HE
252
-------
MODEL VERIFICATION
MODEL WITHOUT BENTHAL
OXYGE-N 'DEMAND
0=0960 5/17/72
LJ
13
X
o
D
Ul
O
co
CO
Q
ST PAUL AVE
STATION 52
1 - I
JL.
JL
JL.
JL.
'. i
J_
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 EOO
5/17 | 5/IB | 5/19 | 5/20 | 5/21 I 5/2Z I 5/Z3 I 5/34 I 5/25
WED 1HU FRI SAT SUN WON TUE WED THU
TIME-HOURS
12
10
WATER S7
STATION 59
0 10 20 30 40 50 60 70 80 90 100 110 (20 130140 150160 170 18019O 20
-------
WATER STREET
D.O. DEFICIT (mg/l) 3
Ln
WATER STREET
BOTTOM DEMAND3 4gm/mZ-Day
Survey I
Survey HI
Survey IST
100 300 300 400 500
MILWAUKEE RIVER FLOW (CFS)
HUMBOLDT AVENUE
600
700
FIGURE 79
DISSOLVED OXYGEN DEFICIT DUE TO BENTHAL OXYGEN DEMAND
-------
For example, during each of the wet weather intensive surveys,
flows in the Menomonee River reached very high levels on one or more
days. Mean daily flows as high as 400 and 600 cfs were recorded on
individual days, providing a total flow of the same magnitude as Milwaukee
River flows at those times. Further, extremely poor quality was
observed, with zero dissolved oxygen levels. Under such conditions,
it is obvious that poor quality in the Menomonee River will result in a
degradation of quality in the Milwaukee River in the vicinity of St. Paul
Avenue and Water Street.
A steady-state model analysis of the Milwaukee River was made
using data developed during the various intensive surveys. Such an
analysis permits investigation of unit responses to various loading
factors. In this particular analysis the effect of the Menomonee River
on oxygen deficit at the test stations on the Milwaukee River was
calculated for each of the flow regimes which prevailed during the
particular surveys. In Table 48, calculated oxygen deficits in the
Milwaukee River due solely to the loading imposed by the Menomonee
are listed. The deficit response shown is based on an assumed BOD
concentration of 10 mg/1 and oxygen deficit of 10 mg/1 in the
Menomonee. This comparison is intended to illustrate the magnitude
of the influence under the assumed quality conditions in the Menomonee.
At times of higher quality, in the Menomonee, the resulting deficits
in the Milwaukee would be proportionately less.
The analysis illustrates both the complex nature of the quality
response and the magnitude of the impact which could occur. The
Menomonee has little or no effect on oxygen deficiency at Humboldt
Avenue or Cherry Street, particularly during higher river flows.
However, it can influence oxygen deficits in the vicinity of St. Paul
Avenue and Water Street by as much as 1 to 2 mg/1 even under the lower
flow conditions.
Combined Sewer Overflow
The effect of combined sewer water overflows in the section of
the Milwaukee River which was studied, is relatively complex. Although
BOD loads discharged to the river in this manner will tend to reduce
the dissolved oxygen level, there are so many other significant
influences on quality in the river section between the North Avenue Dam
and the outlet at Lake Michigan that overflows which directly discharge
into this section of the river do not have a major effecton dissolved oxveen
levels in the River. >&
Analysis of wet weather survey data and model projections indicate
that small overflows have a barely detectable effect. When model
255
-------
Table 48. CALCULATED RESPONSE OF
MILWAUKEE RIVER TO MENOMONEE RIVER
Survey - October, 1970 (Dry Weather)
Milwaukee River Flow - 160 cfs
Menomonee River Flow - 24 cfs (15 percent of Milwaukee River)
Calculated Dissolved Oxygen Deficit in Milwaukee River from
Menomonee River DO Deficits in mg/1
At
Humboldt
Cherry
St. Paul
Water
Menomonee
River
BOD=10 mg/1
0. 0
0.06
0. 66
0. 73
Menomonee
River
Deficit=10 mg/1
0. 00
0. 07
0. 89
0.99
Total DO Deficit for
Milwaukee River Due
To Menomonee River
Influence
0. 0
0. 13
1. 55
1. 72
Survey - May, 1972 (Dry Weather)
Milwaukee River Flow - 290 cfs
Menomonee River Flow - 35 cfs (12 percent of Milwaukee River)
Calculated Dissolved oxygen deficit in the Milwaukee River from
Menomonee River BOD and Deficit:
DO Deficits in mg/1
Humboldt
Cherry
St. Paul
Water
0. 0
0. 01
0. 38
0. 51
0. 0
0. 01
0.71
0. 95
0. 0
0. 02
1. 09
1.46
Survey - August, 1972 (Wet Weather)
Milwaukee River Flow - 714 cfs
Monomonee River Flow - 107 cfs (15 percent Milwaukee River)
Calculated Dissolved Oxygen Deficit in the Milwaukee River from
Menomonee River BOD and Deficit:
DO Deficits in mg/1
Humboldt
Cherry
St. Paul
Water
0.0
0. 0
0. 34
0.74
0. 0
0. 0
0. 54
1. 19
0. 0
0. 0
0. 88
1.93
256
-------
projections are made with and without the combined sewer overflow
loads entering the river. During the major storm event which occurred
during Survey II, the potential effect on dissolved oxygen concentrations
from the large increases in BOD, was compensated for by the large
flow increases which accompanied the storm. The higher flows exert
a dual effect of increased reaeration rates and higher flush out rates.
In fact model projections indicate that for this particular set of
events, detention of combined sewer overflows by a system of tanks
will actually result in a similar DO value in the river.
This effect is due to the reduction in river flow rate due to the volume
of storm water detained.
Under prevailing conditions, combined sewer overflow into the
river between the Dam and the Menomonee River does not for the most
part significiantly degrade dissolved oxygen in the immediate area.
There may be some particular combination of events, storm size, and
pattern, river flow, etc. , where a substantial effect would be observed,
but this will not be common. Combined sewer overflows will contribute
coliform organisms, BOD, COD, etc. , however, their deleterious
effect will be felt after the water leaves the River and enters Lake
Michigan. Combined sewer overflows will additionally contribute a
^considerable amount of suspended solids which will contribute to the
increase of bottom sediments, which create benthal demand.
There seems to be little question that combined sewer overflows
are largely responsible for the degraded conditions observed in this
section of the Milwaukee River. Overflows have contributed over the
years to the high benthal demand observed, and certainly contribute
prominently to the water quality in the Menomonee River and in the
Milwaukee River above the study area. All of these factors have been
shown to have substantial effects on quality in the section of the river
under study.
DETENTION TANK APPLICATION - CITY OF MILWAUKEE
Abatement of discharge of combined sewer overflows for the
entire City of Milwaukee is a highly complex problem. The solution
could involve the construction of detention tanks, similar to the Humboldt
Avenue facility, at many locations. However, a comprehensive plan
for elimination of such overflows will require the evaluation of other
methods, and the ultimate plan may incorporate a variety of measures,
selected for suitability according to the magnitude and nature of the
problem in various parts of the City.
Although detention tanks have not necessarily been established
as the best method of eliminating combined sewer overflows in every
case, the results of this study indicate that they will be a viable and
257
-------
economical tool. For purposes of demonstrating the cost impact of the
problem, and to facilitate comparison with other methods of abatement,
a cost estimate has been developed for construction of thirteen deten-
tion tanks to receive flows from all combined sewer overflow points
on the Milwaukee River in the City. These detention tanks will receive
the combined sewer overflow from the approximately 5800 acres
tributary to the Milwaukee River in the City or about 9 of the 27 square
miles of the combined sewer area in the City.
Based on estimated December, 1973 price indexes, the approxi-
mate preliminary cost estimate for this construction is $29, 5000, 000.
This cost includes the construction of 13 detention tanks, with their
associated control buildings and equipment as indicated on Table 49.
The estimate does not include costs for land, right-of-way, sewer
construction, contingencies or additional treatment facilities which
would also add a considerable amount to the cost.
It is anticipated that combined sewage pumping stations will be
required at four of the thirteen locations. Based on very preliminary
studies, the costs of these four pumping stations will add approxi-
mately $8, 500, 000 to the above cost for detention tanks.
The use of detention tanks to receive flows from all of the
combined sewer overflow points along the Milwaukee River also requires
the construction of sewers for interconnecting the various outfalls
tributary to each of the thirteen detention tanks. Based on preliminary
studies, the costs for these interconnecting sewers are estimated to
be approximately $ 9, 000, 000.
The preliminary costs as discussed herein for a City-wide system
of detention of combined sewer overflows represent only a part of the
total economic impact of such a system. Treatment facilities would
be strongly affected by detention and pump-out arrangements. The
degree of effect would be a function of total storage volumes and pump-
out rate.
Evaluation of the costs of treatment facilities as a consequence
of a comprehensive detention system is beyond the scope of this Report.
However, it can be concluded that the implementation of a combined
sewer overflow storage plan for the City of Milwaukee will require a
concurrent evaluation of treatment plant capacity.
This preliminary estimate is not based on a detailed feasibility
study for each location. It is based on a general visual survey in the
vicinity of each overflow. In that survey, general locations for detention
tanks were determined based on available open space, degree of property
improvements, and access to the location from the sewer system.
258
-------
Several assumptions were made for estimating purposes for each
location:
1. Connecting sewer sizes were assumed to be the same as
the diameter of the overflow, or overflows, to be connected to each
tank.
2. Average depth of connecting sewers was assumed to be
about 10 feet in all cases.
3. Tank volume for each location was based on the sizing
used for the Humboldt Avenue tank under the assumption that water
quality criteria will be met with the removals possible with that sizing.
That is, tank volume was adjusted upwards or downwards in proportion
to the relationship between the area tributary to each site, and the
effective area tributary to the Humboldt Avenue site. This approach
also implies constant rainfall and runoff characteristics for all
locations.
4. Tank construction costs for all locations were based on the
unit costs associated with the Humboldt Avenue tank as presented in
Section VI, Variation in foundation problems, and consequential effects
on construction costs could not be considered.
A preliminary location of 13 detention tanks is indicated in
Figure 80. Each tank is sized to receive the overflow from several
of the 62 combined sewer outlets to the River. These outlets have an
approximate total combined sewer overflow drainage area of approximately
5800 acres.
A listing of the thirteen detention tanks, the approximate combined
sewer acreage which would contribute to each tank, tank size, and the
estimated costs of the tanks are included in Table 49.
General cost information relating cost per square mile to percent
removal of combined sewer overflow based on detention tanks has been
developed. Based on one detention tank serving a 0. 5 square mile
area the capital cost per square mile varies from approximately
$2, 800, 000 to $3, 200, 000 as percent removal increases from 50 to 80.
Based on one tank serving a 2. 0 square mile area the capital cost varies
from $1, 000, 000 to $1, 400, 000 as percent removal increases from 50
to 80. This data is illustrated graphically in Figure 81.
259
-------
FIGURE 80
PROJECT AREA
LOCATION MAPS
O-SYMBOLS INDICATE
APPROXIMATE
DETENTION TANK
LOCATIONS (13)
JONES ISLAltb WASTEWATER
THEATMEW PLANT
260
-------
Table 49. MILWAUKEE RIVER DETENTION TANK DATA
Tank
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
Size of
Contributing
Area (Acres)
690
610
320
570
800
420
95
755
855
200
210
205
95
Tank Capacity
(Million Gallons)
4.7
4.2
2.2
3.9
5.8
2.9
0.7
5.2
5.9
1.4
1.4
1.4
0.7
Estimated
Cost*
$ 2,600,000
2, 500,000
2, 100,000
2,400,000
2,400,000
2, 200,000
1, 800,000
2, 600,000
3, 000,000
2, 000,000
2, 000,000
2, 000,000
1, 800,000
TOTAL
Use
$29, 400,000
$29, 500,000
Based on December, 1973 prices.
261
-------
100
^l
o
O">
o
CO
co
cO
ci
Q
00
I
LU
cc
I- 50
LU
g 40
LU
Q-
Id 3°
I-
<
S 20
X
o
cc
Q.
Q.
FIGURE 81
COST PER UNIT AREA SERVED
FOR
DETENTION TANK CONSTRUCTION
o
0.5
1.0
1.5
2.0
2.5
3.0
NOTE: COSTS INCLUDE ONLY
DETENTION TANK COST
NOTE SEWER MODIFICATION
Etc. NOT INCLUDED
COST IN MILLION DOLLARS PERSQ.MILE
262
-------
SECTION XI
REFERENCES
1. Pollutional Effects of Storm Water and Overflows from
Combined Sewer Systems, Public Health Service
Publication No. 1246. Washington, D.C., 1964.
2. Sullivan, R.H. Assessment of Combined Sewer Problems.
Combined Sewer Overflow Abatement Technology, U.S.
Federal Water Quality Administration. Chicago, Illinois,
U.S. Government Printing Office, June 1970. Pages 107-
121.
3. The Milwaukee River - An Inventory of Its Problems, An
Appraisal of Its Potential. Milwaukee River Technical
Study Committee. Milwaukee, Wisconsin. 1968.
4. Standard Methods For the Examination of Water and Waste-
water. 13th Edition. American Public Health Association,
Inc. 1971.
5. Geldreich, E.E., Clark, H. F. , Huff C. B. and Best, L. C.
Fecal-Coliform Organism Medium for the Membrane Filter
Technique. Journal of AWWA. Vol. 57: Pages 208-214, 1965.
263
-------
SECTION XII
APPENDICES*
I. Description of Detention Tank Model
II. General River Water Quality Data
III. Seasonal, Hourly, Diurnal Variation In River
Quality Data
IV. Dry Weather Sewage Quality Data
V. Wet Weather Sewage Quality Data
VI. Rainfall Frequency Histogram - Station
1 Data 1949 - 1964
VII. Sewage Flow Data
VIII. Rainfall Data
*Note: Appendix I is included as part of this Report. Other
Appendices are not included but are on file with the City
of Milwaukee, Department of Public Works.
264
-------
APPENDIX I
DESCRIPTION OF DETENTION STORM TANK MODEL
General Comments
A mathematical computer model was developed to
describe the performance characteristics of a storm detention
tank. The computer model consists of a main program and five
sub-programs. A listing of these programs is given later in
the Appendix. A description of all the programs is given
below. A computer printout of the program is included following the
program descriptions.
Main Program
The main program is responsible for coordinating the
activities of all sub-programs and reading in and printing out
all the necessary data to assess the performance of the storm
detention tank. The program operates on hourly rainfall data
and calculates the quality and quantity of runoff generated.
A comparison of the quantity of runoff plus the normal dry
weather flow is made with the interceptor capacity. If this
capacity has not been exceeded, all the combined sewage is re-
tained in the sewer system and directed to the treatment
plant. When interceptor capacity is exceeded, the excess
265
-------
portion is directed to the storm detention tank. During the
intervals when it is not raining a certain amount of the tank
contents, depending on the pump out rate, is returned to the
treatment plant. If the runoff is greater than the interceptor
capacity and storm detention tank capacity, the excess is
discharged to the stream. The user of the model has the choice
of determining two types of tank performance, bypass or plug
flow. In plug flow each hourly increment retains its iden-
tity and the first element into the tank is the first one out
of the tank. In a bypass assumption , the tank is circumvented
when full. The model keeps track of the following items which
are necessary to assess performance: hourly rainfall, quantity
and quality both to the tank and to the stream, number of
storms with flow to tank and to the stream, number of hours
of flow to tank and to the stream, and the total number of
hours of rainfall.
Following a description of the various subprograms
an illustration of the input in appropriate format is
presented.
266
-------
Subroutine Bread
This program is used when there is both a volume
of storm water in the detention tank at the onset of the
event under investigation and when the tank performance
characteristics are assumed to be plug flow. The purpose
of this subroutine is to read into the computer the initial
contents of the detention tank. The initial contents are
read in the following manner. The first card gives the
number of "layers" present in the detention tank at the
outset. A layer is defined as an hourly increment of in-
flowing storm water. Since the characteristics are plug
flow no mixing occurs and each increment of wastewater must
maintain its own identity. The succeeding cards describe
the characteristics of each layer. One card is used for
each layer. The following information is contained on each
card: quantity of layer in million gallons, strength of con-
taminant in pounds, and time duration of layer in hours.
When the contents are completely read and checked, control is
transferred to the main program.
267
-------
Subroutine Enter
This program is used when tank performance is assumed
to be plug flow and either one of the following conditions ex-
ist: it has rained during the past hour and the interceptor ca-
pacity has been exceeded, or it has rained during the past hour
and the interceptor capacity has not been exceeded but the tank
was not empty prior to the past hour. The purpose of this
subroutine is to enter into the computer memory the new "layer"
if it has rained during the last hour and update information
relative to old layers still held in the detention tank. This
new layer data is calculated in the main program and a read
statement is not necessary. The quantity of the new layer is
obtained by subtracting the interceptor capacity from the sum
of the wet weather and dry weather sewer flow. The dry weather
flow is an inputted variable and wet weather flow is determined
by calculation from the rainfall drainage area, and runoff co-
efficient. In this subprogram, the tank will accept the total
volume of new input, even in cases where it may be full. The
amount (pounds) of contaminant in the new layer is determined
by calculation from the volume and concentration which is an
inputted variable. The third identifying characteristic (time
268
-------
duration) of this layer is set to zero hours. In addition the
subroutine increases by 1 hour the time duration of all old
layers withheld in the detention tank. When these two functions
are complete control is transferred to the main program.
Subroutine Outof
This program is used when tank performance is assumed
to be plug flow and the volume entered into the detention tank
has exceeded the tank capacity. It should be recalled that
a capacity check was not made in subroutine enter but rather
this subroutine accepted any volume which was entered. The
purpose of this program is therefore, to calculate the strength
and quantity of wastewater that cannot be retained in the tank.
An impliction of this plug flow is that the first layer to
enter the tank is the first layer to leave when capacity has
been exceeded. Since averages are not computed, the calculation of
the total flow being discharged without regard to treatment is straight-
forward. It is merely the sum of the quantities of each layer minus the
capacity of tne tank. The total flow leaving the tank must-
be compared with a s-ummation of flow from each layer added
one at a time to determine how many layers and/or a portion of
269
-------
a layer are/is being forced out of the tank. The determina-
tion of the total pounds discharged to the river is more in-
volved however. Associated with each layer portion thereof
is an identifiable strength in pounds. Contaminant removal
in the storm tank is due to two factors, settling and deten-
tion. Retention can be defined as volume which is prevented
from overflowing to the river due to excess tank availability.
In addition, each layer has associated with it a certain re-
moval due to settling, the amount dependent on the time in-
terval in the tank. The fraction of each layer removed is
—kt
calculated from the equation R«A(l-e) where A and k are
inputted variables for each specific contaminant, and t is the
time which is continually updated for each layer. Accordingly,
this equation is applied to every layer or portion thereof when
it is discharged. The fraction removed by settling is assumed
to be removed continuously from the tank. The fraction not
removed by sedimentation is the strength discharged to the
river when the tank capacity is exceeded. The total strength
therefore is the summation of the individual strengths asso-
ciated with the discharged quantities.
This program in addition keeps track of the number
of layers at any specific time and also the characteristics
270
-------
of each layer which are flow, quality, and residence time in
the tank. Upon completion of these tasks control is trans-
ferred to the main program.
Subroutine Withd
This program is used when tank performance is as-
sumed to be plug flow and the following condition exists: it
has not rained during the past hour and the tank is not empty.
The purpose of this program is to withdraw an amount of waste-
water during dry periods and route it to the treatment plant.
The amount withdrawn hourly is held constant (even though in practice
it may be varied) and is calculated from an inputted variable describing
the number of hours to pump out the tank contents. The flow leaving the tank is
compared with a summation of flows from each layer added one
at a time to determine how many layers and/or a portion of a
layer is/are being returned to the treatment plant. Associa-
ted with each layer/or portion thereof is an identifiable
strength in pounds. The total strength withdrawn is a sum-
mation of the individual strengths associated with the dis-
charged quantities. This program in addition keeps track of
the number of layers remaining at any specific time, and also
271
-------
the characteristics of each layer which are flow, quality,
and residence time. When the remaining layers are updated
for time duration, control is transferred to the main pro-
gram.
Subroutine Head
This program is used to print up column headings
and keep track of margins and lines remaining on a page. It
also points up the date and counts the pages of output.
Illustration of Fortran Card Input to Computer Model
Card
Number Column Variable
1-5
NSET
6-10
TYPE
1-10 ILAST
Description and Comment
Lateger variable right justi-
fied. Number of sets of data.
A set can be no greater than
366 days but may be less. The
sets must be consecutive.
Integr variable right justified.
It is a description of tank
performance. If equal to 1
plug flow is assumed. If equal
to 2 bypass is assumed.
Integr variable right justified.
Time interval in hours from the
last storm to the beginning of
the period under consideration.
272
-------
Card
Number Column Variable
11-20
21-30
51-60
EMPT
AREA
31-40 CAPIN
41-50 TANKC
TANK
1-10 DWF(l)
11-20 DWF(2)
21-30 DWF(3)
31-40 DWF(4)
41-50 DWF(5)
Description and Comment
Integr Variable right justified.
Number of hours it takes to em-
pty the retention tank.
Floating Point Variable. Drain-
age area in acres.
Floating point variable. The
capacity of the interceptor sys-
tem in million gallons per hour.
Floating point variable. The
tank capacity in million gallons.
Floating point variable. The
quantity of combined sewage at
the beginning of the period un-
der consideration in million
gallons.
Floating point variable. Aver-
age dry weather sewer flow at
1:OOAM.
Floating point variable. Aver-
age dry weather sewer flow at
2:OOAM.
Floating point variable. Aver-
age dry weather sewer flow at
3:OOAM.
Floating point variable. Aver-
age dry weather sewer flow at
4:00AM.
Floating point variable. Aver-
age dry weather sewer flow at
5:00AM.
273
-------
Card
Number Column Variable Description and Comments
3 51-60 DWF(6) Floating point variable. Aver-
age dry weather sewer flow at
6:OOAM.
3 61-70 DWF(7) Floating point variable. Aver-
age dry weather sewer flow at
7:OOAM.
3 71-80 DWF(8) Floating point variable. Aver-
age dry weather sewer flow at
8:OOAM.
4 1-10 DWF(9) Floating point variable. Aver-
age dry weather sewer flow at
9:00AM.
4 11-20 DWF(IO) Floating point variable. Aver-
age dry weather sewer flow at
10:OOAM.
4 21-30 DWF(ll) Floating point variable. Aver-
age dry weather sewer flow at
11:OOAM.
4 31-40 DWF(12) Floating point variable. Aver-
age dry weather sewer flow at
noon.
4 41-50 DWF(13) Floating point variable. Aver-
age dry weather sewer flow at
1:OOPM.
4 51-60 DWF(14) Floating point variable. Aver-
age dry weather sev/er flow at
2:00PM.
4 61-70 DWF(15) Floating point variable. Aver-
age dry weather sewer flow at
3:00PM.
274
-------
Card
Number Column Variable Description and Comments
4 71-80 DWF(16) Floating point variable. Aver-
age dry weather sewer flow at
4:00PM.
5 1-10 DWF(17) Floating point variable. Aver-
age dry weather sewer flow at
5:00PM.
5 11-21 DWF(18) Floating point variable. Aver-
age dry weather sewer flow at
6:00PM.
5 21-30 DWF(19) Floating point variable. Aver-
age dry weather sewer flow at
7:00PM.
5 31-40 DWF(20) Floating point variable. Aver-
age dry weather sewer flow at
8:00PM.
5 41-50 DWF(21) Floating point variable. Aver-
age dry weather sewer flow at
9:00PM.
5 51-60 DWF(22) Floating point variable. Aver-
age dry weather sewer flow at
10:00PM.
5 61-70 DWF(23) Floating point variable. Aver-
age dry weather sewer flow at
11:00PM
5 71-80 DWF(24) Floating point variable. Aver-
age dry weather sewer flow at
Midnight.
6 1-10 C Floating point variable. Run-
off coefficient.
275
-------
Card
Number Column Variable
11-20
21-30
31-40
1-2
24-25
WC(1)
WC(2)
WC (3)
1-80 HED(l)-
HED(80)
MONT 2
4-5 DAY2
7-10 YEAR2
21-22 MONTI
DAY1
27-30 YEAR1
Description and Comments
Floating point variable. Aver-
age concentration of contami-
nant during first hour of storm.
Floating point variable. Aver-
age concentration of contami-
nant during second hour of storm.
Floating point variable. Aver-
age concentration of contami-
nant during third hour of storm.
Aphanumeric format. Any suita-
ble title for the first set of
data.
Integer variable right justified.
Ending month for 1st set of data.
Integer variable right justified.
Ending day for 1st set of data.
Integer variable right justified.
Ending year for 1st set of data.
Integer variable right justified.
Beginning month for 1st set of
data.
Integer variable right justified.
Beginning day for 1st set of
data.
Integer variable right justified.
Beginning year for 1st set of
data.
276
-------
Card
Number Column Variable
1-10
PCENT
11-20
RATE
10
1-5
LAYER
11
11
11
1-10
11-20
21-30
FLOW(l)
POWL(l)
ITIME(l)
12 and following:
Description and Comments
Floating point variable. It is
the asymptote in the settling
equation R=PCENT (1 - e )
which defines tank performance
in 1st. It is expressed as a
percent.
Floating point variable. It is
the "k" term in the same sett-
ling equation. The units are
(I/hour).
Integer variable. When the per-
formance of tank is assumed to
be plug flow and the tank is
not empty at the beginning of
the set of data the following
cards are required. Layer to
the number of plug flow elements
initially in the tank.
Floating point variable. Quan-
tity of flow in million gallons
of first elemental layer.
Floating point variable. Pounds
of contaminant in first elemen-
tal layer.
Integer variable right justified.
Time duration in hours of first
elemental layer.
One card is repeated for each
additional elemental layer. The
format is the same as Card 11.
277
-------
Card
Number Column Variable
13
13
13
13
13
13
13
13
13
13
13
7-8
9-10
11-12
13
14-16
17-19
20-22
23-25
26-28
28-31
32-34
YEAR
MONTH
DAY
LSKQ
IHR(l)
IHR(2)
IHRC3)
IHR(4)
IHR(5)
IHR(6)
IHR(7)
Description and Comments
Integer variable right justified.
Year rainfall occurred during
1st set of data
Integer variable right justified.
Month rainfall occurred during
2nd set of data
Integer variable right justified.
Day rainfall occurred during 3rd
set of data.
Describes what part of day rain-
fall occurred LSEQ=1, it is Ml
LSEQ=2, it is PM.
Rainfall during first hour ex-
pressed in .01 inches.
Rainfall during second hour ex-
pressed in .01 inches.
Rainfall during third hour ex-
pressed in .01 inches.
Rainfall during forth hour ex-
pressed in .01 inches.
Rainfall during fifth hour ex-
pressed in .01 inches.
Rainfall during sixth hour ex-
pressed in .01 inches.
Rainfall during seventh hour ex-
pressed in .01 inches.
278
-------
Card
Number Column Variable
13
13
13
14
13
35-37 IHR(8)
38-40 IHR(9)
41-43 IIIR(IO)
44-46 IHR(ll)
47-49 IHRC12)
14 and following:
Description and Comments
Rainfall during eighth hour ex-
pressed in .01 inches.
Rainfall during ninth hour ex-
pressed in .01 inches.
Rainfall during tenth hour ex-
pressed in .01 inches.
Rainfall during eleventh hour ex-
pressed in .01 inches.
Rainfall during twelth hour ex-
pressed in .01 inches.
As many cards as necessary are
used to read into the computer
all the rainfall during the first
set of data. Same format as Card
13.
It is important that a card be
used only when there is rainfall
for at least one or more hours
in the twelve hour period. A
blank in columns 7 and 8 signifies
the end of rainfall record in this
set. Therefore a blank card must
follow the last rainfall card.
If there is more than one set of data repeat cards
seven and following for each set.
279
-------
K)
00
o
DEFINE FILE 1 <366, 27, U, IXV>
INTEGER OUT, PRGE, VR, MO, DflV, EMPT
INTEGER TVPE, V'EflRl, VERR2, MONTI, MONT2, DflVl, DflV2
DI MENSION WCC 5 >, INPHRC 27), HED<20), IHR <12),DWF< 24 >
COMMON FLOUC100),POUNDC100), ITIME<188)
i? FORMfiT<:2c:i2, IX, 12, IX, 14, 16X))
18 FORMflT<:8F19. 2>
1080 FORMRTC2I5>
2000 FORMRTC2I18, 6F18. 3>
2001 FORMflTXIHl, //, 10X,''THE TIME INTERVflL SINCE THE LflST STORM IS', 18, 3
IX, ' HOURS' , /, 18X, ' IT TRKES' , 18, 3X, ' HOURS TO EMPTV THE TflNK' . /. 18X. '
2THE DRRINflGE RRER IS", F18. 8, 3X, ' RCRES', /, 10X, ' THE CRPflCITV '"'F THE
3INTERCEPTOR IS' , F10. 1, 3X, ' MIL. GflL. /HOUR ' , /, 18X, ' THE CRPRCITV OF
4THE STORM RETENTION TflNK IS' , F18. 1, 3X, ' MILLION GflLLON4!' )
2003 FORMRTC18X, "THE ' ,
2''VOLUME OF STORM WRTER IN THE TflNK flT TIME ZERO IS', F18. 1, 3X. ' MILL
31ON GRLLON'>
2002 FORMRTC 18X, 'THE RUNOFF COEFFICIENT IS',F10
12, A10X,'THE BOD CONCENTRRTION DURING THE FIRST HOUR IS', F18. 0, 3X.
2'MG/L',/,18X,'THE BOD CONCENTRATION DURING THE SECOND HOUR IS'.F18
18X, 'THE BOD CONCENTRflTION flFTER TWO HOURS IS',F18.
2006
2007
3000
4000
4001
4002
4003
4004
5000
5001
5002
6000
IS flSSUMED TO
IS flSSUMED TO
3. 8, 3X, 'MG/L'
40, 3X, ' MG/L' >
FORMRTC10X,'TflNK PERFORMflNCE
FORMRTC10X, 'TflNK PERFORMflNCE
FORMRTC6X, 312, II, 1213)
FORMATC28X'NUMBER OF STORMS WITH FLOW
FORMRTC' +' 47X' TO' 7X, I4>
FORMRTC21X'NUMBER OF HOURS WITH FLOW
FORMRTC'+'46X'FROM' 6X, 14)
FORMRTC28X'NUMBER OF HOURS WITH
BE PLUG FLOW-' ')
BE BVPRSS')
TRNK=')
TRNK='
FORMRTC STORM
FORMRTC" SUB
FORMRTC' GRflND
FORMRTC 14, 13"-'
7000 FORMRTC1H+, 14X,
8000 FORMRTC20R4)
1NP = 2
OUT = 3
PRGE=8
LINE=60
RfllNFflLL ='15)
2CF8. 3, Fll. 0, 1X)
TOTflLS '45C'. •' ).
TOTflLS '45C'-'X
TOTflLS '45 (•'*•").
12' -'12)
F4. 2, 3X, F8. 3, 2X, F9. 0, IX, FS. 3, 2X, F8. 9)
2
-------
C
C
C
C
C
C
BLS
C
C
C
C
C
C
C
C
C
C
C
C
C
C
INITIflLIZE GRflND TOTflLS
TQI=0. 0
TQE = 0. 0
TWLT=0. 0
TWLS = 0. 0
INITIflLIZE STORM TOTflLS
01=0. 0
QE = 0. 0
WLT = 0. 0
ML 5=0. 0
QI = FLOW INTO STORM RETENTION
OE = FLOW FROM RETENTION TflNK
WLT = BOD INTO STORM RETENTION
WLS = BOD FROM STORM RETENTION
TflNK
TO RIVER
TflNK TO RIVER
IF THE flBOVE FOUR VRRIflBLES flRE PREFIXED BY fl T THEY REFER TO
C IF PREFIXED BV flN S THEV
INITIflLIZE COUNTERS
NOM=0
IST=i
REDER TO SUBTOTRLS
IST=CUMULflTOR FOR THE NUMBER OF STORMS
NOM=CUMULflTOR FOR HOURS IT HflS
REflD < I NF, 1000 ;•• NSET , TVPE
NSET = NUMBER OF SETS OF DflTfl.
DflVS BUT MflV BE LESS. THE SETS
RfllNED DURING fl PRRTICULRR STORM
fl SET CflN BE NO GREflTER THflN 366
MUST BE CONSECUTIVE
TVPE=DESCRIPTION OF TflNK CHflRflCTERISTICS
IF 1 IT IS PLUG FLOW
IF 2 IT IS BVPflSS
IF 2 IT IS COMPLETELY MIXED
RERDCINP. 2000>ILflST, EMPT, flREfl,
REflD CI NP, ISXDWFCn, 1=1, 24 >
REflD CI NP, 18>C, CWCa:-, 1=1- 3)
ILflST = TIME INTERVAL SINCE LflST
EMPT = NUMBER OF HOURS IT TflKES
flREfl = DRRINflGE flREfl (flCRES)
CRPIN = INTERCEPTOR SEWER CflPflCI
TRNKC = CflPflCI TV OF THE TflNK <
OflPIN, TflNKC, TflNK
STORM < HOURS)
TO EMPTY THE TflNK
TV
4 MG)
C TflNK IS QUflNTITV OF WflTER IN TflNK RT THE BEGINNING OF RUN
C
DWF = RVERRGE DRV WEflTHER SEWER
FLOW
-------
C C = RUNOFF COEFFICIENT
ouRsucci} RRE RVERRGE BOD VRLUES OF RLL THE STORMS FOR THE FIRST THREE H
WRI TEC OUT, 2601) ILRST, EMPT,RRER,CRPIN, TflNKC
WRITECOUT, 2002) C, CWCCI), 1=1, IO
WRITECOUT,2003) TRNK
GO TO C2004,2005), TVPE
2004 WRITE COUT,2896)
GO TO 2068
2005 WRITE COUT,2607)
2098 CONTINUE
C XXX = RMOUNT OF FLQWO1G> PUMPED OUT OF TRNK PER HOUR
XXX=TRNKC/EMPT
DO 999 NSETN=1, NSET
RERDCINP,8000) CHEDCI), 1=1,28)
C RERD ENDIMG DRTE IN COLUMNS U-19 RND BEGINNING DRTES IN 21-30
RERD CINP, 17)MONT2, DRV2, VERR2, MONTI, DRV1, VEflRl
RERD CINP,18) PCENT,RRTE
C RRTE= K RRTE COEFFICIENT IN THE SETTLING EQUflTION
C PCENT=EFFICIENCV OF STORM TRNK SETTLING PERFORMRNCE
C INITIRLIZE SLIBTOTRLS
SQI=0. 0
SQE = 8. 0
SWLT = 0. 0
SWLS = 0. 8
NSTQI=0
NHRQI=8
NSTQE=0
NHRQE=S
NRRIN-0
C NSTQI=NUMBER OF STORM EVENTS WHEN RUNOFF EXCEEDED SEWER CRPRCITV
C NHRQI= NUMBER OF HOURS WHEN RUNOFF EXCEEDED SEWER CRPRCITV
CEEDEDNST£3E= NUMBER OF STORM EVENTS WHEN THE STORM RETENTION TflNK WflS EX
C NUMBER OF HOURS WHEN THE STORM RETENTION TRNK WfiS EXCEEDED
C NRRIN=NUMBER OF HOURS IT RRINED
LRVER=0
TPOUN=D.
C LRVER=NUMBER OF PLUG FLOW ELEMENTS WITHIN THE TflNK
C TPOUN= TOTRL POUNDS OF MRTERIRL IN TflNK RT STflRT
GO TO C31, I<2), TVPE
-------
31 IF CTRNK . GT. 0. > CfiLL SREflD < TRNK, ILflST, TPOUN, LflVER.i
32 CONTINUE
INUM=IDRTE=0
DO 11 1=1, 366
11 MR I TE C 1' I ) < I NPHR C J > , J = l, 27 )
20 RERDC INP, 3800;- VR, MO, DRV, ISEQ, < IHR< J).- J=±, 12)
IFCVR. LE. 0> GO TO 25
VR=1900+VR
I = I DRTE < DRV, MO, VR > - 1 NUM
RERDCl' I>dNPHRc:J>, J=l, 27 >
INPHRCl>=DflV
INPHR<2>=MO
INPHRC3>=VR
J = 0
IFCI5EQ. EQ. 2>J=12
DO 16 K=l, 12
16
WR I TE C 1' I ) ( I NPHR ( J > , J = l, 27 >
GO TO 26
25 NUM = IDRTECDflV2, MONT2, VEflR2> -- 1 NUM
DO 200 N=l, NUM
RERDCl'N>DflV, MO, VR, aNPHRCI), 1=1, 24)
DO 200 1=1,24
IFCINPHRCI). GT. 0> GO TO 60
IF CDRV . NE. 0> ILflST=ILflST+l
IFCDRV. EQ. e> ILRST = lLRST+24
XX=XXX
IF CDRV. EQ. 0> XX=24. *XXX
IF CDRV. EQ. 0> 1=24
INTVL=1
IF CDRV . EQ. 0 > 1NTVL = 24
GO TO C43, 44 >, TVPE
44 TRNK=TRNK-XX
IFCTRNK. LT. 0. > TRNK=0.
GO TO 45
43 IF CTRNK . GT. 0. > CRLL WITHD •; TflNK, ILflST, TPOUN, LRVER, XX, INTVL)
-------
0ST 01 00 < '0 '31 ' 011X3) dl
'0=dlIX3 < 8 '31 ' 01IX3> dl
•0=011X3 < '0 -11 ' 01IX3> d.I
09T
a3iN3 nyo
09T 01 00 < '9 31 ' dMaiO 'IN«1) dl £9
09T 01 00
00
^>I '3* dl
£=III <£ "30 'f
MON=III
T+MON=MON 09
003 01 00 3
0=MON
T+1SI=1SI
0 '0=
0 •0=
0 '0 = 30
0 '0=10
waois 3ziiy.uiNi
30+305=305
10+105=105 8S
siyioians 3indwoo
T+301SN=301SN dl
T+Ii31SN=I01SN dl
W3a 'SIM 'Bin -iiM 'H3 <000s 'ino>3iian
01 00 dI St>
-------
THNK=TRNKC
GO TO (73, 74), TVPE
74 EXITP=EXITQ*WCCI11)*8. 34
GO TO 150
73 CflLL OUTOF (EXITQ, LfiVER, EX UP, TPOUN, PCENT, RfiTE"
150 CONTINUE
190 RflIN = INPHRCI)/100. +0. 005
CflLL HEflDCPflGE,LINE,HED)
VR=VR-1900
WRITECOUT, 6000) 1ST, MO, DHV, VR
VR=VR+1900
WRITE (OUT,7000) RflIN
N>
00
Ln
0.
001)
001)
Q T' E M P, F- T E M P, E XI T Q, E X IT P
NHRQI=NHRQI+1
NHRQE=NHRQE+1
IF CQTEMP . GT
IF (EXITQ . GT
QI=QI+-QTEMP
WLT=WLT+PTEMP
QE=QE+-EXITQ
WLS=EXITP*WLS
ILRST=0
CONTINUE
REM=(SWLT-SHLS>/SWLT
CflLL HEflDCPflGE, LINE,HED>
WRITECOUT, 5001>SQI, SNLT, SQE, SWLS, REM
CflLL HERD(PflGE,LINE,HED>
WRITECOUT,4000>
WRITECOUT,4001>NSTQI
CflLL HEflDCPflGE,LINE,HED>
WRITECOUT,4002)
WRITECOUT,4001>NHRQI
CflLL HEflDCPflGE,LINE,HEDJ
WRITECOUT,4000)
WRITECOUT,4003>NSTQE
CflLL HEflDCPflGE,LINE,HED)
WRITECOUT,4002)
WRITECOUT,4003)NHRQE
CflLL HEflDCPflGE,LINE,HED)
WRITECOUT, 4004)NRflIN
LINE=60
-------
= TQEH-SGE
TWLT=TWLT+SMLT
TWLS = TMLS+-SWLS
999 CONTINUE
R E M= C TIJL T-T WLS > /TWL T
CRLL HERCXPRGE, LINE, HECO
WRITE<:OUT, 5002>TQI, TWLT, TOE, TWLS, REM
CRLL EXIT
END
SUBROUTINE SRERDCTRNK, ILRST, TPOUN, LRVER)
INTEGER OUT-
COMMON FLOwc:i00), POUNDC100), mriE
INP=2
OUT=3
10 FORMRTt:i5>
11 FORMRTc:2F10. 0, I10>
12 FORMRT C'THIS IS RN ERROR MESSRGE'' )
C RERD INITIRL CONTENTS OF TRNK
M RERDCINP, 10 > LRVER
™ QTEMP=0.
TPOUN = 0.
DO S 1=1, LRVER
RERD C I NP, 11 > FLOW ( I > , POUND < I ) , I T I ME < I >
TPOUN=TPOUN-HPOUNDa >
3 QTEMP=QTEMP+FLOWU'>
IFCCGTEMP-TRNK). GT. C. 01*TRNIO > WRITE, POUNDC100), ITIME<100)
INP = 2
6TEMP=0.
DO 5 1=1, LRVER
QTEHP=QTEMP+FLOWC I )
IF CQTEMP . GT. XXX) GO TO 6
-------
5 CONTINUE
TRNK>0.
DO 8 1=1, LflVER
FLOW CI>=0.
POUND CI>=0.
8 i TIME <:n=0
LRVER=0
TPOUN = B.
RETURN
6 J=I-1
IF <:,T. EQ. 0>QJ=0
IF CJ . GT. 0> QJ=QTEMP-FLOW
RRT 1 0= C XXX-Q J ) /FLOW < I >
FLO W < I > =FLOW < I > * ( 1. -RRT 1 0 ')
POUND C I >=POUND< !>* GO TO 9
DO 7 K=I, LRVER
L = K-J
a FLOW=FLOW
^ POUNDCL>=POUND
7 CONTINUE
LRVER=LRVER-J
9 TPOLIN = 0.
DO 14 1=1, LflVER
ITIME=ITIME
RETURN
END
SUBROUTINE HERDciPRGE, LINE, HED)
INTEGER PRGE
DIMENSION HEDC1>
LINE=LINE+-1
IF<:LINE-50;-20, 10, 10
18 LINE=0
PRGE=PRGE+-1
WRITEC3, iee>CHED, 1=1, 20 >, PRGE
WRITEC3, 200 >
-------
I3V)
FLOW TO
BOD TO
TflNK
TflNK '>
LBS/HR
00
00
LflVER, QTEMP, PTEMP)
ITIME<10@)
WRITEC3, 300>
WRITEC3, 400 >
100 FORMRTC'1'19X, 20fl4, 14X' PflGE'
200 FORMRTC' STORM RRIN
2 ' BOD TO ', 'WflT',
3'ER TO BOD TO WflTER FROM
300 FORMRTC' NUMBER" 15X'TflNK
2,' STRERM TflNK
1' TRNK STRERM')
400 FORMRTC12X'CIN/HR) CMG/HR)
2' CLBS) CMG)
20 RETURN
END
SUBROUTINE ENTERCTflNK, TPGUN,
COMMON FLOWC106>, POUND(100),
IFCQTEMP . LE. 0. ) J=LRVER
IFCQTEMP . LE. 0. > GO TO 14
LRVER=LRVER4-1
FLOWCLRVER>=QTEMP
TflNK=TRNK+OTEMP
POUNDCLRVER>=PTEMP
TPOUN=TPOUN+PTEMP
ITIME=0
J=LRVER-1
IF CJ . EQ. 0) RETURN
14 DO 13 1=1, J
RETURN
END
SUBROUTINE OUTOF CEXITG,LflVER,EXITP,TPOUN,PCENT,RflTE)
INTEGER OUT
COMMON FLOWC100>, POUNDC100), ITIMEC108)
OUT=I<
INP=2
QTEMP=0.
DO 5 1=1, LflVER
QTEMP=QTEMP+FLOWCI)
IF CQTEMP . GT. EXITQ) GO TO 6
FLOW FROM'
-------
5 CONTINUE
WRITE COUT, 12;-
12 FORMflT (.' THIS IflN ERROR MESSRGE'>
RETURN
6 J=I-1
IF CJ. EQ. 8)QJ=8
IF CJ . GT. @> QJ=DTEMP-FLOW(I>
RRTIO=C:EXITQ-QJ>/FLOW< n
FLowa>=FLOM*«:i. -RRTIOJ
EXITP=RRTIO*POUND =POUND C I > * < 1. -RflT 10 )
IFCI. EQ. i> GO TO 9
DO 8 K=l, J
RMRIN-1. -CPCENT/10e. )*
K, DO 7 K = I, LflVER
g L=K-J
FLOWCL>=FLOW=POUND
ITIME<:L>=ITIME<:K>
7 CONTINUE
LRV'ER=LflVER-J
9 TPOUN=TPOUN-EXITP
EXITP=DUMMV
RETURN
END
PIP>
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-75-071
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
DETENTION TANK FOR COMBINED SEWER OVERFLOW
Milwaukee, Wisconsin, Demonstration Project
5. REPORT DATE
December 1975 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
City of Milwaukee, Wisconsin, Department of Public Works
and Consoer, Townsend & Associates*
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Public Works
Bureau of Engineers
Municipal Bldg. Room 612
841 North Broadway
Milwaukee, Wisconsin 53202
10. PROGRAM ELEMENT NO.
1BB034 ROAP/Task 21-ASY-077
11. CONTRACT/GRANT NO.
11020 FAU
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
Final
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
*(360 E. Grand Ave, Chicago, Illinois 60611)
16. ABSTRACT
The City of Milwaukee evaluated the merits of detention tanks as a practical method
for abatement of combined sewer overflow pollutional discharges from urban areas. A
3.9 million gallon combined sewer overflow detention tank was constructed to inter-
cept overflow from a 570 acre segment of the City's combined sewer area. As part of
the evaluation program, an extensive sewer and river monitoring program was conducted,
utilizing eleven automated monitoring stations. The monitoring program provided data
utilized with a mathematical detention tank model to evaluate performance of the
project detention tank and provides a basis for other design and planning situations.
Based upon approximately 5 years of data and modeling studies, detention tanks were
shown to be effective in preventing a large portion of the contaminants found in com-
bined sewer overflow from entering receiving waters. General information and methods
for sizing and estimating costs of detention tanks for other areas have been developed,
This information was utilized to establish preliminary cost estimates for providing
similar facilities to serve the entire combined sewer area tributary to the Milwaukee
River in the City.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
^Combined sewers
^Overflows
^Sedimentation
Data aquisition
Cost comparison
Abatement—pollution
Evaluation
Detention tank
River model
Modeling studies
Influences
Storm runoff
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
308
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
290
6USGPO: 1976 — 657-695/5359 Region 5-1
------- |