WATER POLLUTION CONTROL RESEARCH SERIES • 11023FDB 09/70
Chemical Treatment
of
Combined Sewer Overflows
ENVIRONMENTAL PROTECTION AGENCY • WATER QUALITY OFFICE
-------
MATER POLLUTIOil CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe tlie results and progress
in the control and abatement of pollution of our ?!ation's waters. They provide
a central source of information on the research, development and demonstration
activities of the Hater Quality Office of the Environmental Protection Agency,
through in-house research and grants and contracts with the Federal, State,
and local agencies, research institutions, and industrial organizations.
Triplicate tear-out abstract cards arc placed inside the back cover to facili-
tate information retrieval. Space is provided on the card for the user's
accession number and for additional key words. The abstracts utilize the
HRSIC system.
Inquiries pertaining to Hater Pollution Control Research Reports should be
directed to the Mead, Project Reports System, Planning and Resources Office,
Research and Development, Water. Quality Office, Environmental Protection
Agency, Washington, D.C. 20242.
Previously issued reports on the Storm and Combined Sewer Pollution Control
Program:
11034 FKL 07/70 Storm Hater Pollution from Urban Land Activity
11022 DflU 07/70 Combined Sewer Regulator Overflow Facilities
11024 EJC 07/70 Selected Urban Storm Water Abstracts, July 1QG8 -
June 1970
11020 --- 08/70 Combined Sewer Overflow Seminar Papers
11022 DI1U 08/70 Combined Sewer Regulation and Management - A Manual
of Practice
11023 --- 08/70 Retention Basin Control of Combined Sewer Overflows
11023 FIX 08/70 Conceptual Engineering Report - Klngman Lake Project
11024 EXF 08/70 Combined Sewer Overflow Abatement Alternatives -
Washington, D.C.
11024 FKJ 10/70 In-Sewer Fixed Screening of Combined S"wer Overflov/s
11024 EJC 10/70 Selected Urban Storm Hater Abstracts, first Quarterly
Issue
11023 --- 12/70 Urban Storm Runoff and Combined Sewer Overflow Pollution
11023 DZF 06/70 Ultrasonic Filtration of Combined Sewer Overflows
11024 EJC 01/71 Selected Urban Runoff Abstracts, Second Quarterly Issue
11020 FAQ 03/71 Dispatching System for Control of Combined Sewer
Losses
11022 EFF 12/70 Prevention and Correction of Excessive Infiltration
and Inflow into Sewer Systenrs - A f-'anual of Practice
11022 EFF 01/71 Control of Infiltration and Inflow into Sewer Systems
To be continued on inside back cover
-------
CHEMICAL TREATMENT OF COMBINED SEWER OVERFLOWS
Study of Flocculant Treatment and Disinfection of Milk River
Pumping Station Combined Sewer Overflows at
Grosse Pointe Woods, Michigan
Environmental Protection Agency
Water Quality Office
The Dow Chemical Company
Midland, Michigan
Contract No. 14-12-9
Program No. 11023 FOB
September 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C., 20402 - Price $1.50
-------
This report has been reviewed by the
Environmental Protection Agency, Water
Quality Office, and approved for
publication. Approval does not signify
that the contents necessarily reflect
the view and policies of the Environmental
Protection Agency, nor does mention of
trade names or commercial products
constitute endorsement or recommendation
for use.
PROTECTION AGENCY
-------
ABSTRACT
Overflows of .sanitary sewage with stormwater contribute
to pollution and deterioration of water resources. The
description of a typical pumping station, the characteristics
of combined sewage overflows, and chemical treatment of
overflow with chemical flocculants and disinfectants are
included herein.
Combined sewage from a residential community of 3990 acres
in northeast Detroit, Michigan, is collected at the Milk
River Pumping Station. During eight years of record, overflow
rates did not exceed 410 cfs (184,000 gpm) for 90 percent
of the pumping time. Overflow volumes did not exceed 5.5
million cubic feet (41.1 million gallons) for 90 percent
of the occurrences. The average number of days of pumping
per year (41) is about equal to the average number of days
per year having precipitation >0.2 inches (45).
Twenty-two analyses were performed on consecutive time-
weighted samples of influent and effluent during thirty-
three storms over a two-year period. Biochemical oxygen
demand and suspended solids decreased after initial flushing
of the sewers. Chlorine demand values remained relatively
constant throughout storms. The Milk River channel and
the immediate receiving bay were severely polluted.
Cationic polymeric flocculants and flocculant aids signi-
ficantly improved removal of suspended solids from combined
sewage as demonstrated by tests in beakers and a long-tube
sedimentation column. Adequate disinfection of the combined
sewage before discharge to the channel is possible. The
performance of the existing basin can be improved by the
use of staged continuous pumping at lower rates and the
addition of baffles for improved flow distribution.
This report was submitted in fulfillment of Contract 14-12-9
between the Environmental Protection Agency, Water Quality
Office, and The Dow Chemical Company.
-------
TABLE OF CONTENTS
Pai
ABSTRACT
TABLE OF FIGURES
LIST OF TABLES
1. CONCLUSIONS 1
2. RECOMMENDATIONS 7
3. INTRODUCTION 9
4. DESCRIPTION, HISTORY AND DEVELOPMENT OF THE MILK
RIVER DRAINAGE BASIN 13
A. DESCRIPTION OF DRAINAGE AREA 13
B. POPULATION 15
C. DEVELOPMENT OF WASTEWATER HANDLING FACILITIES... 15
D. BASIS OF DESIGN OF TRUNK SEMERS 20
E. ADMINISTRATION OF WASTEWATER HANDLING
FACILITIES 26
5. BACKGROUND DATA AND ANALYSIS 27
A. LONG-TERM RAINFALL-RUNOFF PATTERNS 27
B. MEASUREMENT OF RAINFALL AND RUNOFF 30
C. ANALYSIS OF COMBINED SEWER OVERFLOW RATES 39
D. ANALYSIS OF COMBINED SEWER OVERFLOW VOLUMES 42
E. FACTORS AFFECTING COMBINED SEWER OVERFLOWS 44
F. INFLUENT AND EFFLUENT QUALITY 46
G. BIOLOGICAL QUALITY OF MILK RIVER AND LAKE
SAINT CLAIR 60
6. TREATMENT OF COMBINED SEWAGE WITH POLYMERIC
FLOCCULANTS 75
A. FLOCCULATION STUDIES 75
B. JAR TEST SCREENING 76
C. FULL-SCALE APPLICATION 81
D. STREAMING CURRENT DOSAGE CONTROL 86
E. STABILITY OF PURIFLOC FLOCCULANTS 89
F. TOXICITIES OF PURIFLOC FLOCCULANTS 93
-------
7. LONG-TUBE SEDIMENTATION STUDIES .................... 99
A. SEDIMENTATION .................................. 99
B. DEVELOPMENT OF A LONG-TUBE SEDIMENTATION
DEVICE ......................................... 103
C. PROCESSING OF SOLIDS CONCENTRATION DATA ........ 103
D. VARIABLES ...................................... 106
E. FLOCCULANT SYSTEMS ............................. 109
F. SUMMARY ........................................ 118
8. HYDRAULIC MODEL .................................... 121
A. INTRODUCTION ................................... 121
B. ANALYSIS OF EXISTING BASIN ..................... 124
C. IMPROVEMENTS TO THE MILK RIVER PUMPING STATION.. 125
D. QUALITATIVE DYE STUDIES ........................ 126
E. QUANTITATIVE DYE STUDIES ....................... 131
F. EFFECT OF BAFFLES AT REDUCED FLOW RATES ........ 133
G. OPTIMUM CORRELATION OF FLOCCULAT ION , LONG-TUBE
SEDIMENTATION AND MODEL TEST RESULTS ........... 137
H. PREDICTED PERFORMANCE AT THE MRPS .............. 139
9. DISINFECTION ....................................... 145
A. OBJECTIVES AND METHODS ......................... 145
B. DISINFECTION PILOT PLANT ....................... 147
C. DISINFECTANTS AND THEIR APPLICATION ............ 149
D. ANALYTICAL PROCEDURES .......................... 151
E. EXPERIMENTAL PROGRAM ........................... 152
F. RESIDUAL AND DEMAND vs. DISINFECTANT DOSE ...... 153
G. BACTERIAL POPULATION vs. DISINFECTANT DOSE ..... 155
H. RESIDUAL vs. CONTACT TIME ...................... 155
I. BACTERIAL SURVIVAL vs. CONTACT TIME ............ 155
J. EFFECT OF DISINFECTANTS ON CHEMICAL
FLOCCULATION ................................... 156
K. LONG-TERM RESIDUAL OF BrClg IN COMBINED SEWAGE.. 157
L. CHLORINE DEMAND OF THE MILK RIVER .............. 159
M. DISINFECTION OF MILK RIVER COMBINED SEWER
EFFLUENT ....................................... 159
N. CORRELATION DATA ............................... 163
VI
-------
10. ECONOMIC ANALYSES 165
A. GENERAL 165
B. FLOCCULATION AND SEDIMENTATION 166
C. DISINFECTION 168
D. SUMMARY 172
11. ACKNOWLEDGMENTS 175
12. REFERENCES 177
13. GLOSSARY OF TERMS AMD ABBREVIATIONS 183
ABSTRACT CARD
WATER RESOURCES SCIENTIFIC INFORMATION CENTER FORM
VI 1
-------
TABLE OF FIGURES
Page
1 Topographical Map of Milk River Drainage Basin 14
2 Sanitary Waste Disposal for Areas Northeast of
Detroit 19
3 Flow Diagram - Milk River Overflow System 21
4 Overflow System Basin at Milk River Pumping Station.. 22
5 Harper Woods Separate Sanitary & Grosse Pointe Woods
Combined Sewerage System 23
6 Stormwater Trunk Sewers of Harper Woods 24
7 Yearly Water Inventories for Milk River Basin 29
8 Outline of the Milk River Drainage Area 31
9 Precipitation Data for the Milk River Project -
Event Number 29 33
10 Pumping Data for the Milk River Project -
Event Number 29 34
11 Mass Diagram for the Milk River Project -
Event Number 29 35
12 Distribution of Storm Overflow Rates (1960-1968) .... 41
13 Distribution of Storm Overflow Volumes (1960-1968) .. 43
14 Laboratory Trailer on Site at Milk River Pumping
Station 47
15 Interior of Laboratory Trailer 48
16 Floor Plan & General Arrangement of Mobile
Laboratory Trailer 49
17 Effluent Sampler 51
18 Automatic Analytical Equipment 53
19 Analysis of Combined Sewer Overflow - Suspended
Solids and Biochemical Oxygen Demand 61
20 Analysis of Combined Sewer Overflow - Biochemical
Oxygen Demand and Chlorine Demand 62
VI 1 1
-------
21 Milk River Channel and Estuary with Zones of
Pollution 65
22 Bottom of Milk River Channel Partially Exposed
During Dewatering 69
23 Outline Plan of Milk River Drain Showing Cross
Section Profiles at Selected Stations 72
24 Flocculation of Stormwater During Event Number 7 .... 80
25 Schematic - Polymer Feed System 83
26 Streaming Current Measurements - Event Number 27 .... 88
27 Percent Viscosity Retained (1% Solution) 92
28 Settling Paths of Discrete and Indiscrete Particles.. 99
29 Sedimentation Contour Map: Depth from Bottom vs.
Time 100
30 Sedimentation Contour Map: Log Depth from Surface
vs. Log Time 101
31 Reduced Concentration Profile: Fraction Initial
Depth-Average Concentration vs. Overflow Rate 101
32 Fractional Loss Profile: Fraction of Solids Not
Captured vs. Overflow Rate 102
33 Schematic of Long-Tube Sedimentation Device 104
34 Correlation of Optical and Gravimetric Solids for
System 5A 105
35 Sedimentation Contour Map of Optical Solids as
Functions of Depth from Bottom and Time for
System 5A 112
36 Sedimentation Contour Map of Optical Solids as
Logarithmic Functions of Depth from Surface and
Time for System 5A 113
37 Reduced Concentration Profile of Optical Solids for
System 5A 114
38 Fractional Loss Curve of Optical Solids for
System 5A 116
39 Milk River Pumping Station - Basin Model Inlet 123
40 Milk River Pumping Station - Basin Model Discharge... 123
IX
-------
41 Schematic of Milk River Pumping Station Model 128
42 Model Dye Test Using Pump Number 7 at 75 gpm 132
43 Model Dye Test Using Pump Number 7 at 28 gpm 134
44 Model Dye Test Using Pump Number 7 at 40.8 gpm 135
45 Model Dye Test Using Pump Number 7 at 20.5 gpm 136
46 Effect of Baffles at Reduced Flow Rates 138
47 Distribution of Loss Along Weir 140
48 Average Loss Along Prototype Effluent Weir 141
49 Predicted Loss from the MRPS Prototype with
Five Weirs 142
50 Predicted Area Required for Removal of Class I
Solids 143
51 Disinfectant Reaction Chamber and Detention Coil .... 147
52 Schematic of Parallel Operation of Disinfectant
Pilot Plant 148
53 Effect of Sodium Hypochlorite on Flocculation by
PURIFLOC C31 158
54 Stations in the Milk River Channel Sampled for
Chlorine Demand and Bacterial Populations 160
55 Profile of Chlorine Demand in the Milk River
Channel 161
56 Profile of Bacterial Population in the Milk River
Channel 162
-------
LIST OF TABLES
Page
I Population of Harper Woods and Grosse Pointe Woods
(1920-1960) 15
II Population of Harper Woods and Grosse Pointe Woods
(1961-1968) 16
III Milk River Combined Sewer Time of Concentration .... 25
IV Summary of Total Yearly Flows in Milk River Basin
(1961-1968) 28
V Calculated Station Constants 30
VI Measurement of Storm Volumes 32
VII Precipitation Data for the Milk River Project -
Event Number 29 36
VIII Pumping Data for the Milk River Project -
Event Number 29 37
IX Distribution of Storm Overflow Rates (1960-1968) ... 40
X Distribution of Storm Overflow Volumes (1960-1968).. 42
XI Theoretical Rainfall Intensity and Volume to Cause
Overf 1 ow 45
XII Days of Pumping and Various Excess Precipitation
at Station M-3 46
XIII Summary of Influent Water Quality 55
XIV Summary of Effluent Water Quality 57
XV Suspended Solids, Biochemical Oxygen Demand, and
Chlorine Demand for Selected Influent Samples 63
XVI Composite Summary of Ecological Surveys 67
XVII Quality and Character of Biological Sampling
Locations 68
XVIII Physical Characteristics of the Milk River
Channel 71
XIX Rate of Pumping 73
XI
-------
XX
XXI
XXII
XXIII
XXIV
XXV
XXVI
XXVII
XXVIII
XXIX
XXX
XXXI
XXXII
Volume of Pumping
Conditions for Dispersion and Flocculation
Classification of PURIFLOC Flocculants
Relative Flocculant Activities as Measured by
Turbidity
Settleable Solids for Selected Stormwater
Samples of Event Number 23
Comparison of Streaming Current and Other
Qualitative Criteria of Flocculation
Recommended Maximum Storage Times of PURIFLOC
Flocculants
PURIFLOC Flocculants Evaluated for Toxicity
and Storage
Fish Toxicities of PURIFLOC Flocculants
LDrn Values for Laboratory-Confined Rats
Optical and Gravimetric Solids as Functions of
Uncorrected Depth and Time for System 5A
Calculated Sedimentation Parameters for System 5A. . .
Comparison of Removal Efficiencies of Five
Flocculant Systems Based upon Initial and Final
Time Averages Over All Depths for Optical and
Gravimetric Solids
73
76
77
78
85
87
89
94
95
96
107
108
110
XXXIII Values of Final/Initial Optical Solids Extrapolated
from Reduced Concentration Profiles at Selected
Overflow Rates for Five Flocculant Systems 115
XXXIV Fractional Losses of Optical Solids at Selected
Overflow Rates for Five Flocculant Systems 117
XXXV Similitude of Prototype and Model 122
XXXVI Nominal Detention and Overflow Rates in Milk River
Retention Basin 124
XXXVII Visual and Formal Dye Tests at Various Flows,
Baffle and Weir Placement 129
XXXVIII Effect of Baffles Upon Detention Time at Various
Rates of Pumping 137
XT 1
-------
XXXIX Advantages and Disadvantages of Various
Disinfection Systems 150
XL Summary of Disinfection Experimental Program 152
XLI Demand, Residual, and Bacterial Kill for Combined
Sewage 154
XLII Time Required to Kill 99.99% of Total Coliforms in
Combined Sewage 156
XLIII Long-Term Residuals of Disinfectants in Combined
Sewage 1 57
XLIV Proportional Costs and Benefits at Various Flow
Rates 165
XLV Relative Solids Removal as a Function of Overflow
Rate and Volume 167
XLVI Predicted Removals from Modified Existing Basin .... 169
XLVII Predicted Removals from Optimized Basin 169
XLVIII 7.5% NaOCl Required to Treat Various Volumes at
10 mg/1 171
IL Fixed Costs of Disinfection 172
L Summary of Cost Analysis of Various Treatment
Systems 173
XT
-------
SECTION 1
CONCLUSIONS
BACKGROUND DATA AND ANALYSIS (5)
1. During eight years of record, approximately 39 percent
of the total rainfall falling on the Milk River basin
was collected in the Milk River Sewer System. Approxi-
mately 53 percent of this runoff was pumped to the Milk
River Drain as combined sewage overflow.
2. Less than 8 percent of the total sanitary flow collected
in the Milk River Combined Sewer was discharged along
with storm runoff to the Milk River Drain during storm
overflow pumping events.
3. During eight years of record, approximately 74 percent
of the volume discharged to Lake Saint Clair at Milk
River was pumped at a rate less than or equal to 410
cfs (184,000 gpm) (17 percent of 10 year design flow).
Overflow rates were equal to or less than 410 cfs for
over 89 percent of the pumping time.
4. During eight years of record, overflow volumes at the
Milk River Pumping Station did not exceed 5.5 million
cubic feet (41.1 million gallons) for 90 percent of
the occurrences.
5. The average number of days of storm pumping per year (41)
is approximately equal to the average number of days
per year having precipitation in excess of 0.2 inches (45).
6. The initial biochemical oxygen demand of 18 combined
sewage influent samples at Milk River had an average
value of 110 mg/1 and varied from 23 mg/1 to 376 mg/1.
Final samples averaged 42 mg/1 and ranged from 17 mg/1
to 120 mg/1.
7. Concentrations of suspended solids diminished significantly
during the course of combined sewage overflows at Milk
River. Initial influent samples for 18 storms averaged
249 mg/1 while final samples for the same storms averaged
67 mg/1. Initial samples ranged from 766 mg/1 to 32
mg/1 and final samples ranged from 169 mg/1 to 24 mg/1.
8. The chlorine demand of 56 combined sewage influent samples
at Milk River had an average value of 7.7 mg/1 and varied
from a minimum of 3.5 mg/1 to a maximum of 12.2 mg/1.
-------
9. Total phosphorus concentrations in 64 Milk River influent
samples varied from 0.3 mg/1 to 3.7 mg/1 with an average
of 1.42 mg/1.
10. The Milk River channel including some of the immediate
receiving bay areas of Lake Saint Clair are severely
polluted as characterized by heavy sludge deposits,
the existence of only highly pollution-tolerant life,
high coliform counts, reduced dissolved oxygen, gas
production in the channel, and other qualitative observa-
tions.
11. A transitory zone exists out into Lake Saint Clair between
the severely polluted Milk River channel and the clean
water areas of the Lake.
12. No measurable changes in the bottom-dwelling biota or
the extent of sludge deposition could be observed in
the Milk River channel or Lake Saint Clair as a result
of treatment by polymeric flocculants at the Milk River
Pumping Station.
TREATMENT OF COMBINED SEUAGE WITH POLYMERIC FLOCCULANTS (6)
1. Coagulant aids specifically 15 mg/1 of Baroid Hectorite®
clay can be utilized with the cat ionic flocculant systems
to produce significantly improved turbidity removal
from Milk River Pumping Station combined sewage.
2. Two of the cationic f.l occul ants, PURIFLOC® C31 and
PURIFLOC C32, consistently exhibited excellent flocculation
activity and effective turbidity removals at concentrations
of approximately 10 mg/1 during the early to mid-storm
periods when suspended solids loadings were relatively
high.
3. During the mid- to late-storm periods, when suspended
solids loadings were relatively low, PURIFLOC C31 and
PURIFLOC C32 exhibited relatively poor flocculation
acti vi ty.
4. Final overhead turbidity was relatively constant after
treatment by PURIFLOC C31 and PURIFLOC C32 during all
storm periods; final overhead quality is apparently
independent of initial sample turbidity.
5. The rate of floe formation can be increased significantly
by the addition of anionic polymer flocculants to the
cationic flocculant systems, but may result in a corre-
sponding decrease in solids capture efficiency.
-------
6. The anionic and nonionic polymeric flocculants exhibited
no significant flocculation activity and were not effective
in reducing influent turbidity.
7. Because of the inherent settling inefficiency of the
retention basin, the difficulty in controlling flocculant
concentrations during short periods of pumping, and
lack of a control, quantitative evaluation of the full-
scale treatment studies was not possible.
8. Fair to good flocculation of combined sewage was observed
in the Milk River Retention Basin during the full-scale
treatment studies but sedimentation of these flows was
not poss ible.
LONG TUBE SEDIMENTATION STUDIES (7)
1. Accurate rate data can be collected from long-tube sedi-
mentation devices if flocculation is external to the
column and sedimentation is conducted under dynamic
conditions to prevent deposition.
2. Optical solids are more rapidly and more precisely deter-
mined than are gravimetric solids as indicators of
sedimentation processes in small vessels.
3. Sedimentation of flocculated solids present in combined
sewer overflow is Class I (discrete particles).
4. Concentration requirements of chemical flocculants can
be estimated adequately on the basis of "jar" tests.
5. Efficiencies of solids removals can be estimated from
initial and final depth-average solids concentrations
or, preferably, extrapolated directly from reduced con-
centration profiles at selected overflow rates.
6. Optimized systems of flocculants were capable of removing
80 to 95 percent of optical and gravimetric solids under
the idealized conditions of this study.
HYDRAULIC ANALYSIS AND MODEL STUDIES (8)
1. Effective removal of appreciable amounts of lightweight
suspended solids cannot be accomplished in the existing
Milk River storm overflow system because of short-circuiting,
non-uniform transverse flow distribution, and inadequate
retention time for plain settling.
-------
2. Installation of variable rate pumping to provide continuous
or staged operation from 0-410 cfs (0-184,000 gpm) would
considerably reduce the average pumping rate for over
89 percent of the pumping time. The mean flow rate
for eight years of record was estimated to be 187 cfs
(84,000 gpm). The minimum sized pump at Milk River is
305 cfs (137,000 gpm).
3. Retention efficiency in the existing Milk River Basin
is approximately 42 percent at a flow of 410 cfs. This
corresponds to an effective retention of 8.9 minutes
with a theoretical detention of 21 minutes. Optimum
baffling of the basin resulted in an increase in the
effective retention to approximately 20 minutes for
about 96 percent efficiency.
4. Placement of some transverse weirs in the Milk River
Retention Basin would significantly improve flow distri-
bution and improve retention efficiency and solids removal,
particularly at the most frequently occurring reduced
f1ow rates.
DISINFECTION (9)
1. Differences in the killing quality of the several halogen
disinfectants tested were not sufficiently pronounced
in the available observations to permit conclusions
on their relative merits.
2. Flocculation of combined sewage with PURIFLOC C31 was
not adversely affected by disinfection with sodium hypo-
chlorite in concentrations of up to approximately 40 mg/1.
3. The high chlorine demand of bottom sludges in the Milk
River channel indicates that residual halogen would
not persist in the channel under existing conditions.
4. BrClij mixtures maintain a residual similar to C12 in
combined sewage samples maintained under quiescent conditions
for up to 24 hours.
5. An average concentration of approximately 10 mg/1 of
halogen as Cl^ appears sufficient to remove 99.99% of
the total coliform in the Milk River combined sewage
influent in an effective time of about 15 minutes.
-------
ECONOMIC ANALYSES (10)
1. Where the capital cost of combined sewage treatment
facilities varies approximately in proportion to plant
capacity and where the occurrence of flow rates follows
a log normal distribution similar to that at the Milk
River Pumping Station, the cost of polymeric treatment
starts to increase very rapidly at about 20 percent of
the 10-year design flow of 2450 cfs (1,100,000 gpm).
2. Seventy-five percent of the flows at Milk River can be
treated with NaOCl to provide an average total coliform
concentration of about 1000/100 ml at a cost of approxi-
mately $0.012/1000 gallons. Disinfection equipment
installation cost would be approximately $20,625 and
chemical costs, less operation and maintenance, would
be approximately $7,700 annually.
3. The existing Milk River Pumping Station and retention
basin can be modified by the addition of weirs and variable
rate pumping to achieve an average suspended solids
removal of 63 percent.
4. An idealized basin designed for a rate equal to 20 percent
of the 10-year design storm can be designed to achieve
an average suspended solids removal of 81 percent.
-------
SECTION 2
RECOMMENDATIONS
1. The analyses of combined sewer overflow systems should
take into account the frequency of occurrence of overflow
rate, so that proper weight can be assigned to flows
in the analysis of hydraulic loading of treatment systems
or conveyance systems. Systems must be designed to
operate most efficiently under prevailing conditions
and still have the capacity to handle peak flows.
2. Considering the extremely high coliform counts in the
discharge area of the Milk River drain into Lake Saint
Clair and considering the use of the water for limited,
body contact recreational purposes, it is recommended
that the combined sewer overflow from the Milk River
Pumping Station be treated with sodium hypochlorite
for significant reduction of bacterial levels.
3. One of the 305 cfs (137,000 gpm) storm pumps at Milk
River should be replaced with a flexible pumping system
having a variable capacity of from approximately 100
cfs through 305 cfs, to decrease the average rate of
flow through the retention basin during most storm events.
4. At least two weirs should be placed across the width
of the existing Milk River retention basin to increase
the effect of retention in the basin by reducing the
effect of horizontal short-circuiting and recirculation .
5. Major components of chemical feed equipment utilized
on this Contract should be incorporated, along with
a new flocculator, the recommended weirs and variable
speed pumping to increase the suspended solids removal
to approximately 39 percent at the Milk River Pumping
Station.
6. Analytical data should be developed on the separated
storm sewage from Harper Woods and compared with the
combined sewage of Grosse Pointe Woods. Both areas
have similar population, topography, area, and residential
character, thereby providing the means by which the
effect of sanitary sewage on storm runoff could be measured
-------
SECTION 3
INTRODUCTION
Most sanitary sewerage systems in the United States collect
not only domestic sewage and industrial wastes, but also
receive large quantities of surface runoff during periods
of significant rainfall. In 1967, the Federal Water Quality
Administration of the U. S. Department of Interior estimated
that of the approximately 200,000,000 persons in this country,
about 125,000,000 (62.5%) were served by combined or separate
sanitary sewers. They also estimated that approximately
36,000,000 of these persons were connected to sewerage systems
which collect not only sanitary wastes but also all-,or part
of the storm runoff from the areas which they serve .
There are two main reasons for the widespread existence
of these combined sewers. Firstly, enclosed storm sewers
preceded public water supplies and when central plumbing
came into general use, the logical points of discharge for
household wastes were the already existing storm sewers.
Secondly, even after the water carriage principle of sewage
disposal was adopted, the short-term cost of constructing
one sewer, usually just slightly larger than that required
for storm flow, was considerably less than the cost of constructing
separate sewers for sanitary and storm sewage.
Since it is not economically feasible to construct adequate
treatment facilities to care for the entire flow of combined
systems for all rainfall conditions, it has been common
practice to restrict the flow in interceptor and other main
sewers to some predetermined value. During peak storms,
the rate of runoff can approach 100 times the dry weather
sanitary flow but most sewerage systems are designed to
contain a maximum of about eight times, or more frequently
about four times the dry weather flow^. Relief structures
are provided at suitable points along these sewers to permit
the excess combined sewage to over flow to available water
courses.
These overflows of sanitary sewage mixed with stormwater
contribute significantly to the pollution and deterioration
of our water resources. Because of the extreme variability
of rainfall and runoff patterns, overflow structure designs,
maintenance practices, sewer construction standards, the
existence of other major sources of pollution, and numerous
other important factors, it is extremely difficult to quantitate
the overall extent and effect of this source of water pollution.
To further complicate matters, the compositions of most
combined sewer system discharges vary considerably with
time during a single storm and from storm to storm.
-------
The first flush of a significant storm carries large quantities
of loose suspended and soluble material from pavements and
landscaped areas3>4,5 Street deicing compounds are washed
from pavements during winter thaws. These materials are
added to the sanitary flow along with the large volumes
of stormwater which may resuspend additional accumulations
of grit, sludge and other material previously settled in
catch basins and sewers^. Large volumes of this mixture
then overflow into convenient water courses causing serious
bacterial pollution, aesthetic deterioration, and often
undesirable biological changes in the receiving waters^.
At least three general solutions to this pollution problem
are avai Table.
1. The sanitary and storm sewers can be separated.
2. The combined sewage overflow can be collected in
some type of reservoir and drained back to the
sewage plant for treatment during dry weather.
3. The combined sewage overflow can receive some
appropriate type of treatment allowing water of
acceptable quality to be discharged to the
receiving waters.
In 1967, the Federal Water Quality Administration estimated
that the total cost to separate the combined sewers in this
country on both public and private property would be
approximately $48 billion^. These estimates make it quite
clear that separation of sanitary and storm sewers, while
still the most desirable solution where feasible, is not
a general solution to the combined sewer overflow problem.
Numerous public agencies and private enterprises are currently
studying various methods and processes for solution of the
problem, utilizing the principles of storage with subsequent
treatment before discharge. Because of the variability
in volume, pollution load, and relatively infrequent occurrence
of combined sewer overflows, it appears reasonable to expect
the most economical systems to have wide flexibility, low
capital cost, and an operating cost that can vary with the
load. Because of the cost and limited availability of land
in the areas of greatest need, economics would appear to
favor smaller plants. Some of these advantages are available
in systems utilizing chemical treatment.
10
-------
During 1968 and 1969, The Dow Chemical Company, through
Project No. 11030 FOB with the Federal Water Quality
Administration of the Department of Interior, was involved
in a comprehensive research and development program aimed
at evaluating the applicability of various available chemicals
for treatment of a major combined sewer overflow. The work
was carried out at the Milk River Pumping Station (MRPS)
in Grosse Pointe Woods, Michigan, in cooperation with the
intercounty Milk River Drainage Board.
11
-------
SECTION 4
DESCRIPTION. HISTORY AND DEVELOPMENT OF
THE MILK RIVER DRAINAGE BASIN
Before reviewing the results of the experimental work con-
ducted for the Milk River Project, it is extremely important
to gain an understanding of the Milk River combined sewer
overflow problem. This understanding provides the framework
for a rational analysis of the problem and outlines some
of the major considerations which must be taken into account
in the ultimate design of combined sewer overflow treatment
facilities. Following is a description of the character
of the area within the limits of the drainage basin and
of the system by which the storm and sanitary sewage from
the area is conveyed and handled.
DESCRIPTION OF THE MILK RIVER DRAINAGE AREA
The Milk River Drainage Basin, comprised almost entirely
of the cities of Harper Woods and Grosse Pointe Woods, is
an essentially completely developed residential area of
approximately 3990 acres located in suburban northeast Detroit,
Michigan. Both cities are zoned for limited local retail
business which is concentrated along four main arterial
highways, Vernier Road (M-29), Mack Avenue, Kelly Road and
the Northerly Extension of the Edsel Ford Expressway (1-94).
The only other notable physical features of the basin are
a private golf course (Lochmoor Country Club), a large retail
business area (Eastland Shopping Center) and several rela-
tively smaller school sites and municipal parks.
As can be seen from Figure 1, the topography of the basin
is quite flat and generally slopes toward the Milk River
channel which runs northeasterly between Mack Avenue and
Lake Shore Drive into the community of Saint Clair Shores.
The topography of the basin generally falls easterly toward
the Milk River channel from a maximum elevation of approximately
600 feet above sea level (USGS datum) along Kelly Road in
Harper Woods and falls northwesterly from an elevation of
585 to 590 feet above sea level toward the Milk River channel
from a ridge line running parallel to and west of Lake Shore
Drive in Grosse Pointe Shores. The Milk River channel is
the lowest point in the drainage area with an elevation
of approximately 575 along the banks of the channel at the
point where the Milk River crosses into Saint Clair Shores
between Mack Avenue and Lake Shore Drive along the northerly
limit of the drainage district.
13
-------
Figure 1
Topographical Map of Milk River Drainage Basin
LAKE ST. CLAIR
E. EOSEL FORD FREEWAY^
HARPER ^
-------
POPULATION OF THE MILK RIVER DRAINAGE BASIN
Following is a summary of data on population growth for
the cities of Grosse Pointe Woods and Harper Woods.
TABLE I
POPULATION OF HARPER WOODS AND
GROSSE POINTE WOODS (1920-1926)
(U. S. Bureau of Census)
Harper Woods
73 (Gratiot)
858
9,148
19,995
Grosse Pointe Woods
961 (Lochmoor)
2,805
10,381
18,580
As can be seen from the data, population growth in both
communities was very rapid between the years 1940-1960.
Development in both communities for the years for which
the Milk River Pumping Station has been in operation can
be seen in Table II. These data were taken from the
publication, Population and Occupied Dwelling Units in
The Detroit Region, compiled by the Southeast Michigan
Council of Governments.
These data,
i nformati on
occurred between
Pointe Woods and
based on present
coupled with the U. S. Bureau of the Census
show that maximum development of the communities
1940-1960. As noted previously, both Grosse
Harper Woods are almost completely developed
zoning standards. This is reflected in
the decreasing number of new dwelling units being occupied
each year. The present population density of the area is
approximately 11.2 persons per acre.
HISTORY AND DEVELOPMENT OF WASTEWATER HANDLING FACILITIES
IN THE MILK RIVER DRAINAGE BASIN
The original topography of the basin had several natural
watercourses which served as the drainage system for surface
water runoff (Figure 1). Drainage for the area that is
now Harper Woods followed a small shallow ditch known as
the Girard Drain. This stream originated in the southwest
area of Harper Woods and flowed northeasterly across the
city into the northern section of the present city of Grosse
Pointe Woods. A second watercourse, the Black Marsh Drain,
flowed northerly from the southerly part of what is now
Grosse Pointe Woods to a point south of Lochmoor Road. The
watercourse at this point became known as the Milk River Drain
15
-------
TABLE II
POPULATION OF HARPER WOODS AND GROSSE POINTE WOODS (1961-1968)
HARPER WOODS GROSSE POINTE WOODS TOTAL
Year - as
of July 1
1961
1962
1963
1964
1965
1966
1967
1968
Popul
20,
20,
20,
20,
22,
22,
22,
22,
ati on
100
600
500
800
000
200
600
650
Dwel 1 i ng
Units
5,410
5,550
5,750
5,950
6,230
6,200
6,300
6,320
Popu
18
19
19
21
21
21
22
22
lati on
,800
,300
,900
,100
,200
,600
,000
,200
Dwel
Un
5,
5,
5,
5,
6,
6,
6,
6,
1 i ng
its
300
450
600
880
080
120
240
290
Popul
38,
39,
40,
41 ,
43,
43,
44,
44,
a t i o n
900
900
400
900
200
800
600
850
Dwel 1 ing
Units
10,710
11 ,000
11 ,350
11 ,830
12,310
12,320
12,540
12,610
-------
The Milk River continues to flow northward and joins with
the Girard Drain just east of Marter Road at the north edge
of Grosse Pointe Woods. From this point, it meanders in
a generally northeasterly direction for approximately 6000
feet through Macomb County in what is now the city of Saint
Clair Shores and finally discharges into Lake Saint Clair
just south of the extension of Nine Mile Road. None of
these streams carried any appreciable quantity of dry weather
flow but were filled to some extent by water which backed
up from Lake Saint Clair.
Prior to construction of the Milk River Pumping Station,
the Black Marsh and Milk River drains served as the point
of discharge for the combined sewage from the community
of Grosse Pointe Woods. This sewage was collected in a
combined sewer system, part of which was designed and constructed
prior to 1929. These early improvements committed the city
of Grosse Pointe Woods to a course of using combined sewerage
facilities. About 98 percent of the sewers are presently
combined in the community with almost saturated development
by present zoning standards.
The city of Harper Woods, in comparison, was initially developed
in an area adjacent to the city of Detroit. Sanitary sewers
were constructed to discharge into the Detroit collection
system. As the area in the Milk River drainage district
developed, Harper Woods remained committed to separate sewers.
At the present time, with almost saturated development,
approximately 90 percent of the sanitary wastes from the
community is collected in separate sanitary sewers.
In 1942, a 72-inch sanitary interceptor was constructed
through the central part of Grosse Pointe Woods (see Figure 2).
This interceptor, paralleling the basic courses of the Black
Marsh-Milk River Drains, extended north to Eight Mile Road
(Wayne-Macomb County line) from what is known as the Fox
Creek Enclosure in Grosse Pointe Farms. The Kerby Road
pumping station, located at the Fox Creek enclosure and
Grosse Pointe Woods Interceptor Junction was modified to
lift the incoming flow from the Grosse Pointe Woods interceptor
to the Fox Creek enclosure for eventual discharge into the
Detroit sewerage system. Construction of this interceptor
permitted most of the combined sewers in Grosse Pointe Woods
to be connected to the Detroit sewerage system, thus eliminating
most of the discharge of dry weather sanitary flow into
the open drains.
17
-------
Figure 2
SANITARY WASTE DISPOSAL FOR AREAS NORTHEAST OF DETROIT
City of Detroit
oo
GROSSE POINTE
ARMS
MACOM8 CO.
INTERCEPTOR
WOODS
GROSSED
POINTE SHORES
TORREY ROAD
STATION
MILK RIVER CHANNEL 8
PUMPING STATION
APPROX. SCALE:
0'
10,000'
-------
During the three years following construction of the 72-inch
interceptor, both the cities of Grosse Pointe Woods and
Harper Woods initiated considerable construction of new
sewers to improve their systems. In Grosse Pointe Woods,
a new pumping station was constructed at Torrey Road, and
improvements were made to an existing station at Marter
Road and Hollywood Avenue. Both stations were designed
to discharge dry weather sanitary flow by gravity into the
72-inch interceptor.
The Torrey Road Pumping Station collected almost all of
the sanitary sewage from Harper Woods and some of the combined
sewage from Grosse Pointe Woods. During periods of heavy
rainfall when the interceptor was filled to capacity, the
sewage would back up into the Station until automatic controls
started storm pumps which discharged the combined sewage
into the open Milk River Drain. The Torrey Road Station
had three storm pumps with a combined capacity of 275 cfs
(124,000 gpm) and the Marter Road Station had a pumping
capacity of approximately 200 cfs (89,800 gpm).
The rapid increase in population of both cities required
major improvements to be made to their trunk sewers beginning
in 1954. It was estimated at this time that about three-
fourths of the ultimate population of these communities
had been reached. The Milk River and Black Marsh Drains
were becoming increasingly septic from combined wastewater
overflow discharges. In 1956, two additional pumps and a
weir gate were installed in the Torrey Road Pumping Station
by order of the Michigan State Board of Health. The pumps
were installed to permit discharge at the first flush of
combined sanitary sewage and storm runoff into the interceptor
with the weir gate closed.
It had become apparent by this time that major construction
was required to alleviate wastewater handling problems in
the area. The pollutional problems associated with the
combined wastewater overflowing into the open drains leading
to Lake Saint Clair were of particular concerns. In 1958,
construction was started on collector sewers which generally
followed the course of the Black Marsh, Milk River and Girard
Drains. These sewers were designed to collect almost all of
the sanitary, storm and combined flows from the communities
of Harper Woods and Grosse Pointe Woods. This flow was
then discharged to a new pumping station built at the inter-
section of the Milk River and Girard Drains. The facility,
called the Milk River Pumping Station (MRPS), was designed
to pump the dry weather flow from the two communities into
the 72-inch interceptor by means of three sanitary sewage
pumps each having a capacity of about 11.1 cfs (5,000 gpm).
19
-------
During storm periods, the combined sanitary and storm sewage
from both Grosse Pointe Woods and Harper Woods is lifted
by means of one or a combination of from two to seven storm
pumps into an open concrete settling and skimming basin
adjoining the pumping station, the effluent from which
overflows into the Milk River Drain and eventually into
Lake Saint Clair. A flow diagram of the system is shown
in Figure 3 and a view of the settling basin is shown in
Figure 4.
The Torrey Road Pumping Station remained as part of the
new system. Until 1968, it continued to discharge the dry
weather from from Harper Woods and part of Grosse Pointe
Woods directly into the 72-inch interceptor. The storm
overflows, which originally discharged into the open drain,
were now transported directly to the MRPS. In 1968, this
operation was revised and the Torrey Road Pumping Station
now discharges both dry weather and storm flows directly
to the MRPS. The present major storm, sanitary, and combined
sewers of Grosse Pointe Woods and Harper Woods are shown
in Figures 5 and 6.
The operation of the system can be summarized. Prior to
1968, the MRPS handled principally dry weather flow from
the combined system of Grosse Pointe Woods and lifted it
into the 72-inch interceptor. After 1968 the Harper Woods
dry weather sanitary flow was pumped from the Torrey Road
Pumping Station to the MRPS. The MRPS serves as a sanitary
lift station to the interceptor for all dry weather flow
from the communities of Grosse Pointe Woods and Harper Woods.
During storm periods, the Station serves as a combined sewer
overflow pumping station for all stormwater from Harper
Woods, combined wastewater from Grosse Pointe Woods and
sanitary wastewater from Harper Woods which cannot be pumped
into the interceptor during storm periods.
BASIS OF DESIGN OF THE MILK RIVER TRUNK SEWERS
In general, combined and storm sewers in the Milk River
Drainage Basin area have been sized for runoff calculated
according to the "rational" method.
This method is usually expressed as:
Q = CIA
Q = Rate of runoff, cubic feet per second*
C = Runoff coefficient, % imperviousness
I = Intensity of the rainfall, inches per hour
A = Drainage area, acres
*1 cfs = 448.8 gpm
20
-------
Figure 3
FLOW DIAGRAM - P/IILK RIVER OVERFLOW SYSTEM
(Not To Scale)
Storm Flow Overflow
Bar Screen
Dry Weather
Bar Screen
City Water For Flushing Basin
(175,000 Gal ±/Cleaning)-
3 Storm Pumps at
184,000 GPM
(410 CFS)
STORM
WATER
WET
WELL
16 - 0" Combined
Sewer
SANITARY
WET WELL
4 Storm Pumps at
137,000 GPM
(305 CFS)
3 Sanitary Pumps
at 5000 GPM
(11.1 CFS}
-Parshall Fluma W/Rec. Meter
To Detroit
Interceptor
\
Taintor Gates
-Outlet Scum Baffle
Outlet Weir
SKIMMING &
SETTLING BASIN
3,800,000 Gal.
508,000 Cu. Ft.
x x xx x x x
2 Dewatering fr
Pumps at /1
25,000 GPM^
(55.7 CFS)
800'-0" of 6 ft. Conduit
-8 Hand-operated Mud Valves
Gravity Sewer For Draining Basin
-------
Figure 4
Overflow System Basin at Milk River Pumping Station
-------
Figure 5
HARPER WOODS SEPARATE SANITARY 8
GROSSE POINTE WOODS COMBINED SEWERAGE SYSTEM
APPROX. SCALE: l"= 800'
i
ro
OJ
MILK RIVER
CHANNEL
PUMPING
STATION
GROSSE POINTE
WOODS
TORREY ROAD
'STATION
KERBY ROAD
STATION
SIZE OF SEWER IN INCHES
GRADIENT IN PERCENT
J
TO DETROIT'S SYSTEM VIA FOX CREEK ENCLOSURE
-------
Figure 6
STORMWATER TRUNK SEWERS OF HARPER WOODS
N
APPROX. SCALE: i"=800'
TO MILK RIVER
PUMPING
STATION
Legend:
SIZE OF SEWER IN INCHES
GRADIENT IN PERCENT
24
-------
Harper Woods storm sewers are designed to carry the runoff
from a storm of an intensity which would most probably occur
once in five years. The Grosse Pointe Woods combined sewer
system is designed to carry the runoff from a storm of an
intensity which most probably would occur once in ten years.
In the design of the MRPS and trunk sewers, the calculated
runoff from Harper Woods was adjusted for the ten-year frequency
rainfall at the point of entry into the Grosse Pointe Woods
system. In estimating the runoff, intensity curves similar
to those developed by the city of Detroit for ten-year frequency
storms were used. Specifically, rainfall intensity was
calculated to be:
I = 164
t + 26.5
where I = Intensity, inches per hour
t = Duration of the storm, minutes
A runoff coefficient of 0.20 was used for all park, school
and church areas. All other areas were considered to have
a runoff coefficient of 0.30.
The design engineers calculated the time of concentration
at various points in the sewer system by estimating the
average velocities in existing sewers to be approximately
3.0 fps and by computing the velocities in the trunk sewers
using the calculated runoff and gradients for the proposed
trunk sewers. An inlet time of 20 minutes was assumed.
Based on these assumptions, the time of concentration, including
inlet time, was calculated for the following locations:
TABLE III
MILK RIVER COMBINED SEWER TIME OF CONCENTRATION1
Time of Concentration
Location (Minutes )
Torrey Road Pumping Station 61.0
Vernier and Mack 84.0
Vernier Road and Milk River Drain 89.0
Milk River Pumping Station 92.0
The calculated rate of total runoff at the MRPS based on
a ten-year frequency storm, was calculated by the design
engineers to be 2450 cfs. From these calculations, the
influent trunk sewer diameter required to handle the design
flow of 2450 cfs (1,100,000 gpm) was calculated to be 16
feet at an average gradient of 0.15 percent.
25
-------
ADMINISTRATION OF WASTEWATER HANDLING FACILITIES IN THE
MILK RIVER DRAINAGE AREA
In June of 1955, under the provisions of the State of Michigan
Drain Acts, the cities of Harper Woods and Grosse Pointe
Woods petitioned the Wayne County Drain Commission to finance,
construct and operate major sewerage facilities connected
wi th:
(a) The Girard Drain
(b) The Black Marsh Drain
(c) The Milk River Drain
As seen from Figure 1, the Girard and Black Marsh Drains
are located entirely within Wayne County. The Milk River
Drain, on the other hand, flows through the city of Saint
Clair Shores in Macomb County. Under the provisions of
the law, intra-county drainage boards were established for
the Girard and Black Marsh Drains and an inter-county drainage
board established for the Milk River Drain.
The Inter-County Drainage Board is made up of the Drain
Commissioner of Wayne County, the Drain Commissioner of
Macomb County and one representative from the State of Michigan
This group, called the Milk River Drainage Board, admin-
isters the Milk River Drain including operation of the MRPS
and other facilities, while the Wayne County Drain Commission
actually operates the MRPS. The Torrey Road Pumping Station
is still operated by the city of Grosse Pointe Woods. Most
other facilities, except for the trunk sewers constructed
under petition to the Wayne County Drain Commission are
operated by the cities of Grosse Pointe Woods and Harper
Woods .
26
-------
SECTION 5
BACKGROUND DATA AND ANALYSIS
Understanding of the problem and identification of signi-
ficant variables at the Milk River Pumping Station (MRPS)
are necessary before rationalizing the potential of treatment
by any chemical systems. The significant variables include,
but are not restricted to, rainfal1-runoff patterns, runoff
rates and volumes, and runoff quality.
LONG-TERM RAINFALL-RUNOFF PATTERNS
The long-term relationships between rainfall, runoff, and
operation of the pumping station must be understood before
studying specific rainfall, runoff and quality data. These
relationships can best be seen in Table IV and in Figure 7. The
relationships between rainfall, runoff, stormwater pumping, and
sanitary sewage pumping at the MRPS is tabulated for eight
years.
Several comments are in order:
1. Only calendar years of records are shown.
2. Total rainfall for the basin was calculated using data
from the Detroit City Airport rain gauge station.
3. Dry weather flow was calculated using the lowest seven-
day average for each year.
The total yearly volume of stormwater discharged to Lake
Saint Clair significantly is almost equal to that pumped
from the MRPS through the Detroit Interceptor sewer. The
total annual rainfall varied by almost 50 percent within
the eight years shown; the average yearly runoff varied
from 34 percent to 51 percent. These differences reflect
the extreme variability in storm patterns and indicate the
necessity for evaluating and analyzing runoff data from
selected storms rather than from averaged hydrological data.
Runoff coefficients for project storms of record varied
from 15 percent to 60 percent. Rainfall and pumping hydrographs
have been constructed for each project record storm to establish
individual storm runoff hydrograph patterns.
27
-------
TABLE IV
SUMMARY OF TOTAL YEARLY FLOWS IN MILK RIVER BASIN (1961-1968)
Cumulative Cumulative
Rainfall Rainfall
Year (Inches) (10° Gal)
Total Total
Storm Pump Sanitary
Discharge Flow Pumped
(1Q6 Gal) (IP6 Gal)
Estimated
Total Dry
Weather Flow
(1QP Gal)
Stormwater Total
Pumped Thru Storm
San. System Runoff *
(IP6 Gal) (IP6 Gal) Runoff
1
1
1
1
1
IV)
CO 1
1
1
961
962
963
i 54
965
966
967
968
33
26
19
26
32
29
33
36
.80
.51
.15
.39
.47
.31
.83
.63
3
2
1
2
3
3
3
3
,490
,740
,980
,720
,350
,060
,479
,780
813
575
331
575
1,047
586
738
1 ,471
711
671
739
735
1 ,072
1,041
1 ,287
1 ,598
325
327
341
389
533
478
630
1 ,160
386
344
398
346
539
563
657
438
1 ,199
919
729
921
1 ,586
1 ,149
1 ,395
1 ,909
0.34
0.34
0.37
0.34
0.47
0.37
0.40
0.51
Average 29.76
3,075
767
982
523
459
1 ,226
0.39
-------
Figure 7
YEARLY WATER INVENTORIES FOR MILK RIVER BASIN
J
•
TOTAL VOLUME, BILLION GALS | (*~*7
4
3
2
1
n
—
— -
':•':'. •
^
%,
•::::.
%
E
7Z2
TO
ST
ST
OR
TAL RAINFALL
ORM WATER PUMPED THROUGH
ORM PUMP DISCHARGE
Y WEATHER SANITARY FLOW
^
§
///
///
n
^
///
/ / /
SANITARY
^
SYSTEM
•'.•':':'.•
- —
\\
1961 1962 1963 1964 1965 1966 1967 1968
YEAR
-------
MEASUREMENT OF RAINFALL AND RUNOFF
The rainfall hydrographs were constructed from data obtained
from three weight-type rainfall gauges located within the
Milk River Drainage area. Each of these three rainfall
stations is assumed to have a weighted effect upon the average
rainfall for the entire drainage area according to a method
suggested by Thiessen1.
The location of each of the three gauging stations in the
drainage area is shown in Figure 8. They are designated
W-31 (Torrey Road Pumping Station), W-32 (Detroit Edison
Eastland Substation), and W-33 (Milk River Pumping Station).
These stations were selected for several reasons: good areal
coverage, optimum site conditions, ready accessibility for
maintenance and checking, and location within fenced areas
under the direct supervision of a responsible agency. The
locations were inspected by the ESSA, found acceptable as
gauging sites, and incorporated as part of the Detroit Regional
Rain Gauge Network. The station charts are processed and
evaluated by the Office of the State Climatologist in Lansing
as part of the Regional Network Program and the data published
in the monthly rainfall reports of the SEMCOG Network.
The calculated station constants (Thiessen method) for each
of the three program gauging locations are presented in
Table V. These constants are applied to the point rainfall
data in the calculation of the rainfall hydrographs.
TABLE V
CALCULATED STATION CONSTANTS
SIa t i o n Area (Thiessens Method) Station Constant
W-31 1572 acres 0.39
W-32 1616 0.41
W-33 802 0.20
3990 1.00
Storm runoff hydrographs were constructed using a method
that provided an acceptable degree of accuracy using existing
Station metering equipment. The method selected permitted
us to easily analyze the pumping data, including all storm
overflows from the time that the Station was placed in
operati on.
30
-------
Figure 8
OUTLINE OF THE MILK RIVER DRAINAGE AREA
31
-------
The seven axial flow storm pumps at Milk River have propeller
type impellers. Three have a nominal capacity of 410 cfs
(184,000 gpm) and four have a nominal capacity of 305 cfs
(137,000 gpm). The exact capacity is a function of the
total head against which the impeller is pumping at the
time. At Milk River, the change in total head is determined
by the change in suction head, because the discharge elevation
is fixed. Since the elevation of the water in the storm
well is continuously recorded, the change in total head
on the pump can be determined. The elapsed time of operation
for each pump is also recorded and so a method for calculating
pump flow is available. By determining the pump capacity
from the manufacturer's characteristic curve, based on the
elevation in the wet well, and multiplying this capacity
by the incremental time of operation, the total flow can
be calculated for any time period.
A convenient means of checking flow rates was available
since on some occasions, the storm pumps operate but the
settling basin does not overflow. The volume of water in
the basin was measured and checked against the pumped volume
as calculated below. Three separate storms in which the
basin did not overflow were used to check the accuracy of
the method. The results are listed in Table VI.
TABLE VI
MEASUREMENT OF STORM VOLUMES
Date of Volume (cu ft) from Volume (cu ft) %
Storm Pumping Calculations From Basin Volume Difference
2-15-67 297,000 286,000 +3.7
3-22-67 515,000 495,000 +3.9
4-18-67 260,000 262,000 -0.8
Using the methods described above, rainfall and pumping
hydrographs could then be constructed for any storm occurring
in the Milk River Drainage basin. Data from all record
storms for which quality data has been obtained have been
processed in this manner.
Typical rainfall and runoff hydrographs are shown in Figures
9 and 10. Note the intermittent pumping sequence in Figure 10.
The rainfall and pumping mass diagrams are shown in Figure 11;
the accumulative rainfall and pumping data are shown in
Tables VII and VIII. This is characteristic of most storm
pump operations at the station and indicates that the minimum
32
-------
Figure 9
CO
CO
°'rfT PRECIPITATION DATA FDR THE MILK RIVER PROJECT
°"Jj EVENT NU.V3ER 29
"jt-
0-60.
o-sa
o-^a
0.54
o-sa
0-50.
K o-^a
^ 0.44
S 0-4E
^ 0-40.
Lfi Q.'io
y 0.^
^J
DATE EV£NT BEGAN 6X20/69
TIW.E EVENT BEGAN IS 3AV1
TOTAL PRECIPITATION = 0-44 INCHES
TOTAL VOLUME = S454B17- CU FT
i
^ 0-341
P °!^
< C-24
CL C'2E
^ 0-2Q
^.! o-ia
0-14
0-1Q
o.oa
o-oa
0-04
c-oa
O'On
'
S:
, 1 1
1 ' 1
1
11 1 I 1 1 t ' 1 1 f 1 1 t l I I 1 1 1 1 1 i t 1 t 1 ' 1 1 1 1 : ! t I 1 1 I 1 1 1 I 1 1 1 1 1 ( ( 1 1 1 1 ' I I 1 I 1 ( 1 1 1 I I I I ! 1 i 1
1—4 T— *
TI;/,E
-------
CO
950 -i
900
870
840
810
780
750
720
690
GGO
E30
g GOO
8570
\J, 54G
K 5i°
S^ 4BO
i_ 450
u- 390
M 3£0
~j 330
'-' 300
270
240
210
ISO
120
SO
GO
30
0
Figure 10
PUMPING DATA FOR THE MILK RIVER PROJECT
D/ENT KIM3ER 29
DATE EVENT BEGAN 6/20/69
TIME EVENiT BEGAN IS 3 = 20AM
TOTAL STGRM VOLUME PUMPED = 160O370- CU FT
RUNOFF COEFFICIENT = 0-247
A
TII-C
-------
CO
en
'
42500 .
40000 .
37500 .
35000 .
32500 .
. 30000 .
t —
L.
• 27500 .
U
0 25000 .
H 22500 .
51 20000 .
P 175G3 .
^J 15000 .
< 12500 .
3"" iOOOO .
75CO.
5000 .
2500.
00}
C
Figure 11
MILK RIVER PROJECT
MASS DIAGRAM OF EVENT NUMBER 39
DATE EVENT BEGAN S/20/63
TIME EVENT BEGAN IS 3: 20AM
n M i ATivr RATN,TAI i
n Ml 11 ATTVF H MPAPT x x
^r^~~
/ \j* ™ *£ 'f * *V* ** '^
^PrftfffPifJfZV^^il' 'l II i t t i i [ i i i i : i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i t i i i i i i i i i i i i i i i i
rt£G6Mivwuw*r^ rn t t t f 1 i t 1 1 i 1 — t t t t 1 ! 1 ~t •! T 1 — t t T- T \ • t"1 1 i 1 t i 't t t 1 t t r i t 1 — t 1 i ( t T-T T1 T t r T r T I * t t
) 1 2 3 4 5 & 7 E 9 10 11 2.2 13 14 15 IS 17 IB 19 20 21 22 23 24
DURATION CF STORM (HOURS)
-------
TABLE VII
PRECIPITATION DATA FOR THE MILK RIV6R PROJECT
EVENT NUMBER 29
DATE EVENT BEGAN 6/20/69
TIME PRECIPITATION
OVER AREA FOR
HOUR ENDING
INCHES/HOUR
3AM 0.247
4AM 0.049
5AM 0.057
6AM O.C69
7AM 0.019
SAM O.COO
9AM O.COO
10AM 0.000
11AM 0.000
CUMULATIVE
PRECIPITATION
OVER AREA
INCHES
0.247
0.297
0.355
0.425
0.445
0.445
0.445
0.445
0.445
VOLUME OF
WATER FOR
THE STORM
CUMULATIVE
VOLUME
FOR STORM
1000 CU FT 1000 CU FT
3594.
724.
839.
1012.
282.
0.
0.
0.
0.
3594.
4319.
5159.
6172.
6454.
6454.
6454.
6454.
6454.
EVENT TOTALS
TUTAL PRECIPITATION ^ G.44 INCHES
TOTAL VOLUME = 6454817. CU FT
36
-------
TABLE VIII
PUhPlh'G DATA FOR THE I-'.ILK RIVER PROJECT
EVENT NUMBER 29
DATE EVENT BEGAN 6/?0/6°
TIME EVENT BFGAM IS 3 20AM AND is PRINTED OUT IN TEN MINUTE INTERVALS
AVERAGE DRY FLOW FOR 10 MINUTE PERIOD IS 1650 f.U FT
TIME TOTAL VOLUME
FOR TEN
MINUTE PERIOD
1000 CU FT
CUMULATIVE
STORM
VULUI-iE
1000 CU FT
FLOW
CFS
3. 10AM
3.20AM
3.30AM
3.40 AM
3.50AM
4. 0AM
4.10AM
4.20AM
4.30AM
4.40AM
4.50AM
5. 0AM
5.10AM
5.20AM
5.30AM
5.40AM
5.50AM
6. Ofl'.-i
6.10AM
6. 20 Ah
6.30AM
6.40AM
6.50AM
7. 0AM
7.10AM
7.20AM
7.30AM
7.40AM
7.50AM
8. 0AM
8.10AM
(3.20AM
8.30AM
8.40AM
8.50AM
9. 0AM
9.10AM
9.20AM,
9.30AM
9.4^ AH
9.50A..X
0 . 00
0.00
0.00
0.00
0.00
0.00
o.co
0.00
0.00
0.00
0.00
0.00
0.00
0.00
158.70
168.58
161.47
14V. 6 2
42. 11
0.00
0.00
0 .00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
133. IB
164.84
156.5?
134.8S
0.00
0.00
o.oo
0.00
0.00
0 . 00
0.00
c.oo
-0.83
-0.83
-0.83
-0.63
-0.83
-0.83
-0.83
-0.83
-0.83
-0.83
-0.83
-0.83
-0.83
-0.83
157.87
326.46
487.93
637.56
679.67
679. b'7
679.67
679.67
679.67
679.67
679.67
679.67
670.67
679.67
679.67
812.86
977.70
1134.24
1269.10
1269.10
1269.10
1269.10
1269.10
1269.10
1269.10
1269.10
1269.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
264.51
280.97
269.12
249.37
70.19
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
221.97
274.74
260.39
224.76
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
37
-------
TABLE VIII (Cent.)
T I (••!"- TOTAL vrUVl-
FOR TEN
ni;"UTE PR i-'T nn
].0no CU FT
CU!"-'IH.ATJ VF
STOP!"
VOLUME
lOOn CU FT
FLO'-J
CFS
If). Of-1-'
lO.lOA'"'
10.2"/-!
!0.30/'>i>
l(i.4P-?li
!0.50'/-l;
11. 0 /'l--
ll. 10, 'vi
11.20/X-i
ll.30.t-f'
11.40.AI.
.11 .to «•• '•-;
12. (>.'•••••
12.10P>->
12.20F;;
] 2.:-iOPi'i
12.4np:.'
12.bOPi"
1. f'Pr-'
1 . inpi-i
1. 2f)Pf
1.30P-;
i .^op,-;
1. .50P'-i
2. OPf;
2.10Pi-,
2.20p;.
2.30P;M
2.4-OPi-
2.50P.-
3. OP'-
3.1 OP;-;
3. 20 Pi-;
3.30PM
3.40PI--.
3.50PI'i
0.00
0.00
o.oo
o.i-.o
0.00
0.00
0.00
0.00
0.00
0 .00
0.00
0.00
0.00
0 . C 0
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
o.oo
20.94
Iri2.?9
104.70
o.oo
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1269.10
1?6°.10
12 6°. 10
12A9.10
1269.10
126Q.1Q
126Q.10
1269.10
1269.10
1269.10
1?69. 10
1269.10
1269.10
1269.1 0
1269.10
126°. 10
1269.10
1269.10
12 6". 10
1269.10
1269.10
1.269.10
126°. 10
1290.04
14^2.34
1547.05
1547.05
1547.05
1547.05
1547.05
1547.05
1547.05
1547.05
1547.05
1547.05
1547.05
0.00
0.00
0.00
0.00
0 . 00
0.00
0.00
0.00
0 . 00
0.00
o.oo
0.00
0.00
o.co
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0 . 00
34.90
253.82
174.51
o.oo
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
T TOTALS
TOTAL STi.'Hi-i v/riLlir-'E =
1600370.
CU T;T
38
-------
sized pump at Milk River has a rate which draws down the
sewer storage at a rate greater than the sewer intrusion
rate for most storms. This characteristic results in a
pulsed or intermittent flow into the settling basin.
By averaging the area under each pulse of the curve in
Figure 10 and extending the time-rate function until a continuous
curve is developed, an estimate of the average influent flow
rate can be made. The long-term average for Storm No. 29,
for example, is 51 cfs (22,900 gpm), significantly less
than the existing minimum capacity of 315 cfs (141,000 gpm).
A considerable improvement in pumping efficiency could be
achieved by containing short-term, high-rate peak flows
in upstream storage while maintaining low average pumping
rates at the Station.
ANALYSIS OF COMBINED SEWER OVERFLOW RATES
Because of the nature of combined sewer overflows, it is
necessary to measure the magnitudes and frequencies of occurrence
of both overflow rates and volumes. The problems associated
with hydraulic loading, sludge production, and chemical feeding
can then be assessed and an efficient, economical and workable
design effected. The influent pollutional load and its
final impact on the receiving water can then be estimated.
An analysis of records at the MRPS for the period 1960 through
1968 is contained in Table IX. The percent of time a selected
flow rate was equal to or less than a given value was calculated.
The overflow rate did not exceed the capacity of one large
pump (410 cfs; 184,000 gpm) for almost 90 percent of the
occurrences. A rate of 3 x 410 cfs only occurred an additional
8 percent of the time. The capital cost would be almost
tripled over that to handle 410 cfs if treatment capacity
were provided for 1230 cfs (551,000 gpm). The corresponding
increase in operating time would be approximately 8 percent.
More significantly, the total volume treated would be increased
by about 15 percent.
The logarithmic distribution of flow rates is shown in Figure 12.
The percent of time that the flow rate equaled or exceeded any
given value can be extrapolated from this plot. The geometric
mean overflow rate is extrapolated as that rate of pumping
which would not be exceeded for 50 percent of all occurrences.
The mean overflow rate of 187 cfs (83,900 gpm) is less than
one-half the capacity of one large pump (410 cfs) in the
existing facility. Even one small pump (305 cfs) is too large
to handle the majority of storm overflow rates without periodic
storage in the wet well. This fact was also apparent from the
39
-------
TABLE IX
DISTRIBUTION OF STORM OVERFLOW RATES (1960-1968)
Accumulative Total Accumulative
Pump Rate Duration Duration Volume Volume % Volume
Combinati on (cf s) (m i n ) (mi n) % Time (cu ft) (cu ft) at Rate Shown
1/0 305 20,955 20,955 60.09 383,476,500 383,476,500 44.76
0/1 410 10,237 31,192 89.45 251,830,200 635,306,700 74.16
2/0 610 826 32,018 91.82 30,231,600 665,538,300 77.68
1/1 715 889 32,907 94.37 38,138,100 703,676,400 82.14
0/2 820 405 33,312 95.53 19,926,000 723,602,400 84.46
3/0 915 133 33,445 95.91 7,301,700 730,904,100 85.31
2/1 1,020 275 33,720 96.70 16,830,000 747,734,100 87.28
1/2 1,125 126 33,846 97.06 8,505,000 756,239,100 88.27
4/0 1,220 29 33,875 97.14 2,122,800 758,361,900 88.52
0/3 1,230 101 33,976 97.43 7,453,800 765,815,700 89.39
3/1 1,325 136 34,112 97.82 10,812,000 776,627,700 90.65
2/2 1,430 125 34,237 98.18 10,725,000 787,352,700 91.90
1/3 1,535 225 34,462 98.82 20,722,500 808,075,200 94.32
4/1 1,630 16 34,478 98.87 1,564,800 809,640,000 94.51
3/2 1,735 127 34,605 99.23 13,220,700 822,860,700 96.05
2/3 1,840 85 34,690 99.48 9,384,000 832,244,700 97.14
4/2 2,040 53 34,743 99.36 6,487,200 838,731,900 97.90
3/3 2,155 58 34,801 99.80 7,499,400 846,231,300 98.78
4/3 2,450 71 34,872 100.00 10,437,000 856,668,300 100.00
-------
5000
3000
2000
Figure 12
DISTRIBUTION OF STORM OVERFLOW RATES (1960-1968)
RATE.CFS
1000
800
600
400
200
100
O.I
10 30 50 70 90 98
PERCENT OF RATES EQUAL TO OR LESS THAN
99.9 99.99
-------
pulsed operation. A large portion of the influent flow to
Milk River could be treated under relatively steady-state
conditions if pumping could be varied between 100 and 305 cfs.
The total capacity of the sanitary pumping systems is only
11 cfs and is generally not available during times of storm.
This analysis is limited to the existing Milk River facility
but several useful possibilities are suggested.
1. A similar correlation can be developed from rainfall
distribution values corrected for runoff character-
istics for design of new combined sewage treatment
installations.
2. Optimum or most economical treatment or pumping
analyses can be made on the basis of the most
probable rainfal1-runoff curves developed in 1.
ANALYSIS OF COMBINED SEWER OVERFLOW VOLUMES
It is also necessary for several reasons to evaluate variations
in flow volumes in addition to flow rates. Firstly, the cost
of retaining a given percentage of storms for discharge into
the Detroit interceptor system during dry weather may be more
economical than installing and operating treatment facilities,
chemical inventory requirements, and to some extent, feed rate
capacities which are a function of total storm volumes.
The distribution of storm overflow volumes for the years 1960
through 1968 is shown in Table X and plotted in Figure 13.
DI
STRIBUTION
Vol ume
106 Cu Ft
0-
.06-
.12-
.25-
.50-
1.00-
2.00-
4.00-
8.00-
16.00-
1
2
4
8
16
32
.06
.12
.25
.50
.00
.00
.00
.00
.00
.00
1
3
7
15
31
63
127
OF
TABLE X
STORM OVERFLOW VOLUMES (1960-19
Vol ume
106 Ga
0-
.45-
.90-
.87-
.74-
.48-
.96-
.92-
.84-
.68-
1
3
7
15
31
63
127
255
Total
1 Occurrences
.45
.90
.87
.74
.48
.96
.92
.84
.68
.36
1
9
23
37
64
83
71
33
15
2
Cumul ati ve
Total
1
10
33
70
134
217
288
321
336
338
68)
01
Total/100
0.0030
.0296
.0976
.2071
.3964
.6420
.8521
.9497
.9941
1 .0000
42
-------
GO
32 i-
16
8
Figure 13
DISTRIBUTION OF STORM OVERFLOW VOLUMES (1960-1968)
VOLUME, IO$CU. FT.
.5
.25
.125
.0625
1
I
1
I
O.I
10 30 50 70 90 98
PERCENT OF VOLUMES EQUAL TO OR LESS THAN
99.9
-------
The geometric mean pumped volume (50 percent of total occur-
rences) is approximately 1.2 million cu ft (8.98 million gal);
90 percent of the pumped volumes do not exceed 5.5 million
cu ft (41 .1 mill ion gal ) .
FACTORS AFFECTING COMBINED SEHER OVERFLOWS
Another important aspect of the Milk River combined sewer
overflow problem is that of understanding the factors which
cause the overflow structure to operate.
From studies of the Pumping Station records, it has been
estimated that the Milk River sewerage system has a storage
capacity of approximately 600,000 cu ft (4,490,000 gal).
dry weather flow for 1968 was estimated to be
(2,240 gpm). Under present operating practice,
flow pumped to the sanitary interceptor before
storm operation is about 22 cfs (9,870 gpm).
The average
about 5 cfs
the maximum
switching to
By inserting these values in the conventional storage formula
(Rate of Inflow - Rate of Outflow = Rate of Storage), the
following relationship is derived.
Dry Weather Sanitary Flow and Storm Runoff - Sanitary Pumping
Rate of Storage
, f . rrn 09 -p 600.000 cu ft
5 cfs + CIA - 22 cfs = - 1 (sec) -
CIA - 10».000 + I?
CIA - t (min) 17
Assuming an average runoff coefficient of 0.35 and a drainage
basin area of 3990 acres, the intensity of rainfall required
to cause an overflow becomes:
I =
1
t (min)
Between 1960 and 1969, the median duration for all storms
at Milk River was between three and four hours and 86 percent
of the storms did not exceed eight hours. The minimum duration
of importance would be the Time of Concentration or approximately
1.5 hours. Following is a tabulation of rainfall intensities
required to cause overflow for a reasonable range of storm
durations .
44
-------
TABLE XI
THEORETICAL RAINFALL INTENSITY AND VOLUME TO CAUSE OVERFLOW
Storm Duration Runoff (CIA) Storm Intensity Total Rainfall
(Minutes) (cfs) (Inches/Hr) (Inches)
90 128 0.092 0.138
120 100 .072 .144
180 73 .052 .156
240 59 .042 .168
300 50 .036 .180
360 45 .032 .192
420 41 .029 .203
480 38 .027 .216
These calculations show the effect of storm duration both
on the storm intensity required to cause overflow and on
the accumulated rainfall required to cause overflow. From
this analysis, it appears that the overflow at Milk River
is more a function of total rainfall than of rainfall intensity
Surface storage and wetness factors would undoubtedly increase
the total rainfall requirements shown in Table XI.
The number of days of pumping at Milk River compared with
the number of days of various excess precipitations at a
nearby gauging station are shown in Table XII. The data
were not separated as to individual storms. Some error
was introduced when storms or pumping runs extended into
more than one day. The average number of days of pumping
per year for the eight year period is about the same as
the average number of days per year of precipitation slightly
in excess of 0.2 inches. Considering surface storage and
ground wetness, the data supports the rainfal1-runoff analysis
shown in Table XI.
Coupled with the rainfall-runoff distributions suggested
earlier, the storage-runoff analysis can provide one rational
approach to the analysis of combined sewer overflow systems.
45
-------
TABLE XII
Year
1961
1962
1963
1964
1965
1966
1967
1968
DAYS OF PUMPING AND VARIOUS EXCESS
PRECIPITATION AT STATION M-3
Number of Days of Precipitation
Measurable<. 1"<. 2"<.3"
116
103
87
102
127
95
102
93
77
70
54
61
74
64
72
67
52
40
30
40
53
45
49
47
40
30
23
27
42
32
34
32
No. Days of
Storm Pumping
61
44
23
36
45
26
41
53
Average
78
67
45
32
41
INFLUENT-EFFLUENT QUALITY
The quality of the combined sewage should be evaluated before
a complete statement of the Milk River problem can be outlined.
This was necessary both for characterization of the influent
and effluent sewage and for evaluation of the various chemical
treatment systems. Under the terms of the Contract, character-
ization required 22 analyses on consecutive time-weighted
samples from both the influent and effluent.
Laboratory Trailer
Laboratory
Milk Ri ver
facilities were not available at or near the
site and had to be built. A mobile laboratory
was designed and constructed for immediate use during the
Contract and future use by the Federal Water Quality Administration
at the conclusion of the Contract.
The space requirements were estimated on the basis of the
expected analytical load during storm events, and consultation
with FWQA personnel experienced in mobile laboratory design
and operation. It was decided that the most practicable
unit was a commercial semi -trai1er. The solid construction
of such units provides a stable platform without restricting
mobility or load capacity.
The semi-trailer selected was 40-foot long, 8 feet wide,
and 13 feet from ground to roof. It is the largest trailer
presently allowed on the roads of most states without special
permits. The interior capacity is approximately 2500 cu ft;
it has almost 300 sq ft of floor space. The sides and roof
are an aluminum skin over a steel frame. There are two
46
-------
doors on one side of the trailer and four windows (16" wide
x 24" high) located at eye level. In addition to the normal
"landing gear" on the front of the unit, a set of leveling
jacks is provided behind the rear wheels of the trailer
to give additional support and stability while the trailer
is on site. Two sets of access steps and platforms with
handrails are designed to be removed from the trailer and
placed in storage boxes beneath the trailer.
An outside view of the completed trailer in place at the
Milk River Pumping Station is shown in Figure 14.
Figure 14
Laboratory Trailer on Site at Milk River
The interior units of the laboratory were designed compactly
because of the limited floor area available. The equipment
required for similar analytical tests were grouped together
in specific areas to reduce traffic problems. All of the
equipment required for the bacteriological testing was grouped
together at the front of the trailer; ovens and other equipment
needed for solids determination were placed at the rear.
Each location had a conveniently located sink. A centrally
located air compressor and vacuum pump were connected to
headers along three sides of the trailer.
The design of the fume hood was a special problem. A negative
pressure would develop in the trailer if ventilating air
came from within, so a unit was obtained that allows a portion
of the vent air to come from the outside. Special equipment
was designed to heat this outside air in cold weather.
47
-------
Conventional heating and air conditioning units of the type
that are used with most trailers were not suitable for the
laboratory. Heat from furnaces, hot plates and other equipment
required a unit with a large, flexible cooling capacity.
A custom-built heating-venti1ating and air-conditioning
(HVAC) unit designed to fit on the nose of the trailer was
installed. A hood with ventilating fan was installed over
the muffle furnace and drying ovens to reduce the load on
the HVAC equipment as well as to carry off fumes and odors.
The conditioned air is conveyed through ceiling ducts for
the full length of the trailer and is distributed through
continuous
or cooli ng
large area, low rate
can be controlled by
diffusers
adjustment
Local
of the
heating
diffusers
A view of the interior
is shown in Figure 15;
is shown in Figure 16.
looking toward the
a plan view of the
front of the trailer
trailer interior
Figure 15
Interior of Laboratory Trailer
Sampling
One of the most difficult problems at Milk River was that
of sampling. Flow rates varied from 305 to 2450 cfs
(137,000 to 1,100,000 gpm). Influent sewage depths varied
from 2 to 17 feet with no dry well available for positive
head devices, and a representative effluent sample had to
be obtained from an inaccessible weir approximately 210
feet in length.
48
-------
MOBILE
Figure 16
LABORATORY TRAILER
2. REFRIGERATOR
3. B.O.D. INCUBATOR
4. MUFFLE FURNACE
5. FUME HOOD
6. WATER HEATER (Under counter)
7. LABORATORY GLASSWARE WASHER
6. AIR COMPRESSOR (Under counter)
9. VACUUM PUMP (Under counter)
10. WATER STILL (On wall)
II. DRYING OVENS (2)
12. SMALL WATER BATH
13. LARGE WATER BATH
vo
STEPS AND PLATFORM
FLOOR PLAN & GENERAL ARRANGEMENT
-------
The basin effluent was sampled by four 1-inch vertical suction
lines spaced evenly along the effluent weir. The suction
lines drew a sample from points between the skimming baffle
and effluent weir at a depth above the bottom of the skimming
baffle and just below the outlet weir. The four sampling
pipes fed into a common header through equal lengths of
pipes using the same fittings to equalize the head loss
in each line and permit samples which make up the composite
to be of equal volume. Another line was run to the Milk
River to permit sampling and calibration of automatic analytical
equipment when the basin was empty.
Influent samples were obtained by means of a submersible
pump suspended in the wet well beyond the bar screens within
the transition structure between sewer and wet well. All
main sampling lines in both the influent and effluent systems
were 2-inch in diameter and flowed constantly during the
sampling period.
Because of the importance of sampling, automatic samplers
were designed and constructed specifically for the work
at Milk River. The samplers were designed to collect adjustable
grab samples from the continuously flowing 2-inch pipe stream,
composite them for variable periods and hold them in a refri-
gerated compartment for periods up to about three hours.
The sampling program was controlled by a continuous punched
tape program which varied the collection time of each composite,
the number of grab samples in each composite, and each of
the variables from one sampling time to another. The size
of each grab sample was controlled externally. Construction
of the effluent sampler can be seen in Figure 17. The influent
sampler was identical except for the tape control unit and
reader which was fabricated from basic components in contrast
to the use of commercially available units as seen on the
effluent sampler.
Analytical Methods
The following analyses were performed on the Milk River
influent and effluent samples. The procedures as outlined 2
in Standard Methods for the Examination of Water and Wastewater
werefollowed except for the modificationsnoted below.The
order and abbreviations are identical to those used in Tables
XIII and XIV.
50
-------
Figure 17
Effluent Sampler
51
-------
1. Date - calendar day of record.
2. Storm (STRM)
3. Sample (SMPL)
4. Time
5. pH - Continuously monitored and recorded.
6. Turbidity (TUR) - Continuously monitored and recorded
using a surface scatter turbidimeter.
7. Temperature (TEM) - Influent temperature only was
continuously monitored and recorded.
8. Dissolved oxygen (DO) - Continuously monitored and
recorded using a dissolved oxygen analyzer.
9. Suspended solids (SUS SOL) - Filtered through a glass
mat and dried at 103°C.
10. Volatile suspended solids (VOL SOL) - Filtered through a
glass mat and fired at 600°C.
11. Total solids (TOT SOL) - Evaporated at 103°C.
12. Total volatiles solids (TOT VOL SOL) - Fired at 600°C.
13. Settleable solids (SET SOL) - By volume using an Imhoff cone
14. Chlorides (TOT CL) - Determined by the Volhard method.
The Volhard Method B as given in the llth Edition of
Standard Methods was used because of the expected high
concentration of chloride ions.
15. Total hardness (TOT HARD) - EDTA titrametric method with
methylthymol blue as the indicating agent
16. Calcium hardness (CA HARD) - EDTA titrametic method with
methylthymol blue as the indicating agent.
17. Chlorine demand (CL DEM) - Excess chlorine was
determined by amperometric titration.
18. Biochemical oxygen demand (BOD) - Winkler method with
azide modification to remove interference from nitrites.
19. Nitrogen (TOT N) - Samples were returned to Midland for
analysis. They were preserved by adding 40 mg/1 Hg++
and frozen as recommended in a Report to U. S. Public
Health Service^. The method of analysis used was the
total Kjeldahl method for ammonia and organic nitrogen.
20. Total phosphates (TOT P) - Samples were returned to
Midland for analysis by the stannous chloride method.
21. Oil and grease (OIL) - Samples were quick-frozen and
returned to Midland for analysis by the Soxhlet
Extraction Method
52
-------
22. Inorganic carbon (INR C) - Infrared carbonaceous analyzer.
23. Organic carbon (ORG C) - Infrared carbonaceous analyzer.
24. Total coliforms (TOT COLI) - Membrane filtration and
incubation at 35°C for 18-24 hours on Bacto-m-Endo
broth MF.
25. Fecal coliforms (FEC COLI) - Membrane filtration and
incubation at 44.5°C for 24 hours on Bacto-m-FC broth.
26. Fecal Streptococci (FEC STRP) - Membrane filtration and
incubation at 35°C for 48 hours on Bacto-m-Enterococcus
agar.
27. Storm volume (STRM VOL) - Calculated from station records
(corrected for changes in elevation in wet well).*
28. Conductivity (not tabulated) - Influent stream only was
continuously monitored and recorded.
The analyses listed above were performed within three hours
of the time of sample collection except for those which
were performed in Midland on preserved samples.
Calibration of the automatic pH and dissolved oxygen recorders
was extremely difficult to maintain under the conditions
of intermittent operation and variation in solids content
of the Milk River combined sewage. The surface scatter
turbidimeter performed reliably but was subject to plugging
by large solids. The automatic analytical equipment in
the effluent sampler building is shown in Figure 18.
Figure 18
Automatic Analytical Equipment
53
-------
Analytical Results
Results of the analyses of influent and effluent samples
from the Milk River retention basin are summarized in
Tables XIII and XIV. Each sample is a composite of one
minute grab samples collected continuously during a pumping
sequence. Several factors affect the nature of these data
and should be considered in quantitative evaluation. Some
of the more important factors are:
1. Intermittent operation of the pumping station causes
widely varying sewer velocities, unstable vertical
concentration gradients, and poor mixing in the
influent combined sewage. Since the flow depth
in the sewer varied from 2 to 17 feet, the optimum
influent sample point was necessarily selected on
the basis of empirical testing and weighted judgment.
2. The effluent sample points were more ideally located
but other factors led to difficulties in interpreting
the effluent quality data. Specifically these factors
were :
a. Non-uniform distribution of flow across the
effluent weir. Each pump or pump combination
had a unique flow distribution pattern across
the weir.
b. Circulating flows were created within
the basin during many pumping sequences as a
result of flow diversion by the effluent weir.
c. Effective detention times in the basin varied
considerably depending on the pumping configura-
ti on.
d. In normal operating practice, the Milk River
retention basin never reaches a hydraulic steady-
state condition.
e. Settled solids flushed from the Milk River retention
basin are not removed during the winter months. This
condition also existed whenever weather or other
circumstances prevented flushing of the basin between
events. These accumulated solids were scoured
during many operations causing effluent samples
to contain material deposited during prior storms.
For the reasons noted, evaluation of the quality data is
limited to a study of certain characteristics of the influent
sewage.
54
-------
TAB1.K XIII
MILK RIVER PROJECT - SUMMAP.Y OF INFLUENT WATER QUALITY
DATE
10 1
10 18
10 14
10 18
10 18
1C 18
il 13
11 15
11 15
11 16
11 16
11 16
11 16
12 13
17 '.3
12 13
12 1-i
12 27
01 17
', 1 17
01 17
Cl 17
01 18
Cl IB
01 IS
Cl 18
01 13
•J3 2*
r,» 24
';3 2*
")•» 74
i3 24
03 25
cn
en
STRM
68
63
08
08
68
60
68
68
63
68
68
66
68
68
64
68
68
68
69
6V
69
69
69
(,9
69
69
69
69
t9
69
69
t.9
69
13
14
14
14
14
14
15
15
15
16
16
16
16
17
17
17
18
19
20
20
20
20
20
20
20
20
20
21
21
21
21
21
21
SKPL
01
01
02
03
0*
05
01
02
03
01
02
03
04
01
O?
01
01
01
01
02
03
04
05
06
07
08
09
01
02
03
O4
05
06
T!ME CH
2300
1612 6.7
1712 7.2
1812 7.2
1912 7.2
2012 7.2
0530
0550
1 350
1935
2020
20»S
2113
0305 7.5
03<.5
0445 6.2
1130 8.0
1830
1910
2025
2125
2310
0/30
GcOO
1 135
1440
1800
1125 7.1
1200 7.4
122? 7.2
1J?2 7.2
1422 7.4
1000 7.6
TUR TEM
1 T II of
J 1 U " r
54
54 66
19 66
19 66
19 66
19 66
45
45
45
48
48
48
21 48
47 47
<,! 46
54 46
63 47
47 46
27 50
oo sus
SOL
1
1
247
3.2 72
5.5 74
5.0 141
4.3 6
4.3 26
444
206
57
32
43
85
50
7.8 766
7.6 116
7.0 60
3.8 95
107
466
137
261
96
102
5
24
87
46
3.2 138
5.1 139
5.6 216
5. ft 107
5.6 119
4.7 24
VOL
SUS
SUL
153
27
38
81
2
12
314
160
32
19
22
36
28
564
SB
24
56
43
284
58
94
84
59
5
22
S7
46
77
86
125
42
42
21
TOT
SOL
497
487
604
281
277
721
400
285
261
243
230
240
1450
365
361
1401
36B5
3056
1956
919
770
558
528
522
585
570
1296
1272
604
503
504
598
TOT
VOL
SOL
171
173
264
ioa
94
428
230
112
85
89
94
100
611
139
85
209
136
403
192
161
163
127
115
142
165
164
231
212
157
100
118
176
SET
SOL
5
1
2
8
0
0
15
10
2
0
1
2
2
50
6
I
3
0
21
2
2
0
2
0
0
4
0
1
1
3
1
3
1
IOT
CL
70
55
68
62
43
44
36
25
39
32
32
21
21
263
73
103
707
335
1372
S23
348
305
216
2C6
174
163
160
323
303
163
154
138
TOT
HARD
MG '
72
139
94
106
107
1E2
56
75
94
89
ft 6
92
179
148
224
169
232
136
142
189
148
165
IB?
208
233
240
260
144
112
112
CA
HARD
L^^___
^^^^
72
142
90
74
78
106
54
69
87
74
58
71
117
I 14
109
109
134
186
132
104
93
11 1
114
99
136
160
232
212
112
312
116
192
CL
OEM
10. 1
6.8
6.8
8.8
7.0
6.6
10.9
11. 1
9.3
8.9
10.8
11.8
12.2
10.2
6.2
5.3
R.5
6.9
10.0
11. 1
9.9
11.5
8.9
10.3
10. I
7.7
7.5
5.0
5. 1
4.3
4.5
BOO
72
95
89
121
34
31
216
99
120
23
23
32
33
360
105
52
93
60
376
93
60
47
44
49
50
99
96
101
100
97
42
48
35
TOT
N
4.8
6.6
3.4
4.9
4.3
2.8
3. 1
1.5
2.7
2.8
3.0
2.8
2.8
5.0
5.0
5.0
5.7
5.3
7.6
4.2
2.0
10.2
2.3
2.0
2.3
1.5
2.6
7.5
9.2
4.1
3.6
5.1
6.3
TOT
P
2.3
1.8
1.3
1.5
1.6
1.5
1.1
1. I
1.6
1.8
1.4
1.3
1.3
1.0
1.4
1.2
1.4
1.3
1.0
0.9
0.8
0.3
0.8
0.6
1.6
2.7
2.9
1.4
3.1
2.3
2.5
2.4
2.0
OIL
21 .0
21.5
19.5
38. I
7.3
7.0
76.0
25.0
2.2
2.9
7.2
6.0
11.7
28.0
25.0
4.4
27.5
14.0
99.9
22.5
10.0
11.1
1 1.0
6.7
10.0
10.6
1 1.5
27.9
4.4
33.6
13.5
10.1
11.4
INR
C
11
7
12
11
16
in
10
8
5
6
7
11
11
15
29
29
13
14
13
27
CRG
C
I
1
t)9
256
284
67
27
24
32
27
262
62
33
49
31
20
26
49
36
84
78
75
44
39
31
TOT
COL I
NO. / » V w ••.
FEC
COLI
i nnui
4* 24. 0
19
9
30
2
3
14
t.
I
18
28
24
17
9
2
1
7
3
10
1
1
0
0
0
1
1
0
6
3
1
1
1
3
7.9
0. I
0.8
0.2
0.5
1.0
0.6
0.6
3.0
4.9
4.0
1.2
0.7
0.4
0.4
0.8
0.8
1.6
0.5
0.3
C.O
0.0
0.1
0.8
.3
.3
1.5
.7
.4
.1
.1
.4
FEC
STRP
X 1 06
. T90
.650
.060
.570
.0?6
.029
.640
. 190
.079
.200
.058
.290
.102
.760
.490
.140
.560
.270
.640
.370
. i I 3
.003
.co
15
15
16
lo
13
10
17
17
17
18
19
70
20
?0
20
?0
20
20
20
2O
21
21
21
21
21
21
01
Cl
02
C3
04
05
Cl
02
03
01
02
03
04
01
02
03
01
Cl
01
02
03
04
05
06
07
08
O9
01
0!
03
04
05
06
-------
TABLE XIII
MILK RIVER PROJECT - SUMMARY OF INFLUENT WATER QUALITY (CONTINUEU1
04 re
34 2
0* 2
04 7.
34 2
D4 2
0* 17
0* 17
04 17
0* 17
04 21
C4 21
04 21
OS 18
OS 18
05 16
CS 18
06 1
06 5
06 5
06 A
Oft 3
06 8
C6 20
C6 20
06 20
36 23
06 30
07 11
07 17
07 17
STRM
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
6«
69
69
69
69
22
22
22
22
22
23
23
23
23
2*
24
24
25
25
25
25
26
27
27
28
28
28
29
29
29
30
31
32
33
33
SMPL
01
02
03
04
05
01
02
03
04
01
' 7
_3
01
02
03
04
01
01
02
01
02
03
01
02
03
01
01
01
01
02
TIME
0300
0400
0500
0600
OBOO
1348
2010
2103
2156
1330
1630
1830
0714
0916
1053
1212
0630
0340
Ofi20
1018
1448
1730
0521
0752
1349
1207
1810
1210
1515
PH
u.l
7.8
7.1
7.0
7.1
7.5
7.5
7.5
7.4
7.5
7.3
7.3
7.4
7.4
7.4
7.6
7.7
7.4
7.3
7.4
7.5
7.1
TUR
i tit
•1 1 U
46
37
34
55
32
34
36
31
29
78
40
38
TEM
_ c
" r
40
40
40
61
61
59
59
38
38
47
47
64
48
00
I
1
4.7
5.3
5.B
5.7
5.4
5.5
5.0
2.5
4.6
4.3
4.9
4.8
5.6
3.6
4.0
4.2
6.3
8.4
6.9
5.7
4. IS
5.1
5.0
SUS
SOL
179
267
1B8
145
69
258
176
216
72
57
76
52
190
24
20
40
114
44
119
70
78
367
98
53
183
121
364
233
169
VOL
sus
SOL
162
98
104
66
57
76
35
176
20
16
30
412
87
37
89
37
36
128
43
37
85
64
166
97
65
TOT
SGL
1422
720
S60
420
36B
674
769
556
432
802
589
536
541
216
202
204
149
528
342
318
258
237
551
218
275
332
347
525
502
'.10
TOT SET
V(U SOL
SOL
361
364
132
230
168
204
174
159
156
250
173
160
211
85
67
79
532
364
2CO
186
199
157
193
86
96
113
180
205
169
144
4
5
2
2
1
3
2
2
1
0
0
12
3
1
2
16
4
1
6
3
2
4
1
2
4
1
6
5
2
TOT TOT CA
CL HARD HARD
284
220
170
96
82
131
124
89
57
145
124
113
110
39
36
32
8 4
85
67
56
53
46
57
28
44
39
20
57
28
20
280 199
130 121
111 98
IC5 95
140 133
192 137
179 141
169 118
216
3B6
24'.
223
137
83
70
79
64
16B
134
135
117
136
117
82
109
117
13;,
141
92
131
164
254
161
181
100
69
45
57
165
123
97
103
103
124
82
67
107
110
105
111
85
117
CL
OEM
9.0
8.4
8.6
8.9
7.3
6.9
6.4
4.8
7.3
10.1
9.2
122
7.0
6.5
4.7
5.2
8.3
5.4
9.7
3.5
3.9
5.9
6.3
7.2
6.7
BOO
93
92
50
26
21
128
83
34
17
53
36
38
56
27
24
43
62
44
48
31
18
71
29
48
54
100
45
40
TOT
N
5.8
2.3
1.6
2. 1
3.2
7.3
8.5
5.6
3.8
13.0
9.9
9.0
4.5
2.9
2.7
3.4
9.1
4.8
6.5
6.3
3.9
4.6
2.6
5.4
5.0
4.1
5.1
TOT
p
1.0
l.l
0. 7
0.5
0.5
2.1
2.2
1.3
1.2
3.7
3.0
2.6
1.2
0.8
0.8
1.0
1.1
1.1
0.9
1.3
1.8
1.3
0.6
0.7
0.4
2.7
0.9
1.2
.7
1.4
OIL INR
C
23.0
26.0
12.3
2.2
2.8
33.4
25.2
9.3
0.4
12.9
6.8
5.4
12.1
3.4
2.8
5.6
21.0
18.6
7.0
17.4
12.6
11.6
17.1
fl.2
13.8
13.5
4.0
19.5
5.6
6.7
45
22
16
16
14
21
19
27
26
4
11
13
17
2
2
1
20
14
6
12
12
17
6
2
7
17
12
8
0
8
ORG TOT FEC FfC
C COLI COLl STRP
51
69
44
27
24
54 3 1.5 .990
83 4 1.1 .870
45 2 1.0 .160
42
32
36
26
86
23
17
21
80
76
32
67
42
33
62
29
44
47
33
94
46
32
I
8
2
2
5
1
1
1
3
2
4
2
3
2
2
2
4
6
8
4
.5
5.0
1.4
1.0
1.0
.1
.1
. 1
1.5
1.4
2.4
l.l
1.6
0.6
0.6
1.9
1.8
4.2
5.4
3.6
. 140
.360
.220
.069
.270
.072
.059
.071
.460
.250
.067
.160
.180
.410
.003
.170
.490
.770
.280
.210
CUM
STfiRM
VOLUME
1 1)-* F T
4M2.59
7U3.81
1 ?•;••.. 35
12'i4. jS
15S1.72
35. S3
579.95
B'vO.46
12034.89
0.00
5i8.21
935. GO
103.66
659.30
1122.30
6157.47
4003.61
4t,2.37
408.72
?8.49
1 134.55
2177.62
0.00
679.67
1600.37
436.95
1300.12
940.31
1423.24
4ori.oi
STORK
SNPL
22
22
22
22
2*
23
23
23
23
24
24
24
25
25
2S
25
26
27
27
28
23
28
29
?9
29
30
31
32
33
33
01
02
01
04
05
01
02
03
04
01
02
03
01
02
03
04
01
01
02
31
U2
03
01
02
03
01
01
01
Cl
02
cn
-------
TABLE XIV
MILK RIVER PROJECT - SUMMARY OF EFFLUENT WATER QUALITY
DATE STRH jyPL TIME PH TUR TEN
JTU
22 68
6h
68
(7
( 7
( 7
'7 77 68
' 1 12 68
68
73 68
( 7
C7
f-7 71 68
C 7 23 68
C8
f. 8
CM
C8
08
C8
Cd
(.".
C3
16 68
16 60
16 68
16 68
I 7 68
17 63
19 SB
lr< tb
1? 68
10 0» »,»
13 03 68
13 04 68
1 J 18 68
15 68
10 68
15 68
15 t9
15 63
09
09
09
09
09
09
O9
OT
09
11
I I
11
1 1
II
11
12
12
12
13
13
15
15
15
II 16 63 16
16 68 16
11 16 68 16
01
03
04
05
06
07
08
09
10
01
07
01
04
05
06
1 :.' 1 1 68
17
01
01
02
03
04
05
01
02
03
01
2228
2247
2302
2117
2347
0017
004 /
OOSO
1500
2126
2141
2226
2326
0035
0400
01 1445
07 liOO
03 1515
01 7115 10.0 31
02 2330 10.8 30
03 0010 10.8 28
1619 7.4 63
0545
0600
0615
1345
1400
2020
2035
2050
0320
7.5
7.5
7.5
7.5
7.5
48
47
44
37
25
33
33
3*
1)
03
I
1
2.3
2.2
2.1
1.5
6.1
6.1
6.0
6.5
6.4
6.2
6.0
6.2
sus
SOL
248
250
188
94
108
62
102
1 >3
110
404
207
58
38
81
46
Ifll
19
149
148
132
1 38
116
170
290
220
104
136
108
99
138
VOL
sus
SOL
155
162
144
95
111
72
07
90
74
75
33
16
3
15
14
73
12
49
64
48
90
50
128
132
102
72
90
58
49
78
TOT
SOL
492
474
367
322
370
281
244
2'j2
285
835
'•Tj
331
305
305
384
519
386
403
421
406
309
507
360
393
373
298
346
288
270
288
TOT
VOL
SOL
160
182
126
114
110
51
n
102
186
80
79
80
83
136
IBB
143
146
168
174
108
208
184
195
174
135
173
123
102
117
SET
SOL
I
5
2
1
I
0
•)
I
5
1
2
1
1
1
3
1
1
0
0
0
4
4
4
4
2
1
2
2
3
TOT
CL
50
40
41
40
37
40
25
18
11
279
27.2
270
287
279
270
57
39
39
43
28
28
44
32
25
32
18
25
32
28
28
rnr
HA^D
Mf* /I
Wl> / L
206
125
104
82
94
90
68
54
66
77
156
190
159
112
112
139
110
82
112
83
98
76
91
96
83
79
77
CA
HARD
84
90
9ti
91
91
67
76
69
65
4H
50
55
89
132
95
99
68
84
93
71
69
190
75
60
54
62
67
70
71
80
CL
DEM
7.6
6.2
5.9
4.8
7.5
8.6
9.3
9.5
9.6
8.2
6.9
7.7
5.3
5.4
6.1
5.1
7.5
7.8
7.6
9.8
13.5
13.9
6.9
9.6
9.8
9.3
6.7
8.1
8.4
8.3
8.6
BOO
126
123
52
55
38
30
25
26
24
62
04
55
34
43
32
58
50
30
3 1
62
60
43
93
103
105
101
71
76
45
48
54
TOT
N
6.3
4.8
5.5
3.6
3.4
3.4
2.8
2.3
3.1
4.4
1.7
2.8
2.1
2.6
2.0
5.4
2.6
3.2
3.4
7.1
6.3
3.0
4.6
2.3
2.5
2.7
3.6
3.7
2.8
2.8
3.3
TOT
P
4.8
3.9
3.6
4.2
3.2
3.3
3.0
2.6
2.9
2.2
2.8
2.7
2.1
2.2
2.7
2.0
2.0
2.0
1.0
.0
.0
.1
.6
.5
.5
.4
.6
OIL I MR
C
31.0
21.0
18. n
13.0
18.0
22.0
28.0
25.0
27.0
21.0
19.0
9.4
11.2
5.1
hj
• r.
15.0 13
13.9 10
11.7 10
30.6
56.8
33.3
25.5
R.9
13.0
12.3 8
17.2 8
19.8 7
ORC TOT
C COL I
1 NO / 1
FEC
COL I
nnHi
2 0.6
1
9
3
1
O
0
1
7
1
2
3
4
0
23
9
7
57 2
56 0
55 I
8
147 4
114 6
101 5
70 4
92 4
43 10
43 12
53 19
1.4
1.6
0.5
2.1
0.5
1.4
0.2
0.2
3.8
0.2
0.4
0.5
0.8
O.I
9.2
6.8
2.0
Oa
. 0
0.6
0.3
0.3
1.7
0.8
0.8
0.6
0.8
8.5
1.2
1.7
1.4
FEC
STRP
X 10^
.288
.313
.216
.099
.134
.073
.006
.003
.094
.001
.170
.115
.041
.052
.006
.150
.067.
.042
.070
.330
.080
.010
.390
.210
.270
.130
.134
.081
.180
.210
.160
CUM
STORM
VOI.U"E .
IOJ Ft3
500.79
081.07
1031.75
1031 . 75
1031.75
1031.75
10V. .75
2761.18
3731.68
1591 .94
3*90.40
7399.11
122BH.64
1578-.. 81
19POS.67
17;. 4ft
550.57
1268.23
2 3 1 O . 8 1
910.46
910.46
1473.77
65/.93
699.05
903.65
1001.50
163 f .67
1690.37
501.65
673.61
1160.16
STRM
SMPL
09
C9
09
09
09
09
09
09
09
11
11
11
1 I
11
11
12
12
12
1 2
13
13
13
14
15
15
15
15
15
16
16
lo
01
03
04
Ci
C6
07
OS
09
10
01
C2
C3
C4
C5
C6
Ol
02
03
C4
01
02
03
Cl
01
02
03
04
05
01
C2
C3
5.6 246 198
723 182
203 164
on
—I
-------
TABLE XIV
MILK RIVER PROJECT - SUMH'.RY OF EFFLUENT WATER QUALITY
(CONTINUED)
DATE
•M n
01 17
•VI 17
01 18
01 18
01 18
.! IB
)1 19
03 24
J3 24
03 24
03 24
03 24
03 74
•04 02
34 0?
•>'. 02
04 C2
04 02
04 C2
04 n
04 1 7
04 17
04 21
05 IB
05 19
05 i. 8
05 18
)5 18
06 05
06 05
06 OR
06 08
:6 OB
cn
CO
STRM V(-L
69
69
69
69
69
69
69
69
fc-J
69
69
69
69
69
69
t9
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
69
20
20
20
20
20
20
20
20
21
21
21
21
21
21
22
22
22
22
22
22
2',
23
23
24
25
25
25
75
25
27
27
28
28
2b
01
02
03
04
05
06
07
03
Cl
02
03
04
05
06
01
02
03
04
05
06
01
02
03
01
01
02
03
04
05
01
02
01
02
03
TIKE
2010
2145
2230
0130
0230
1135
1440
1800
1215
1230
1245
1445
1500
1515
0310
0325
0417
0432
0447
0800
1348
2010
2103
1455
0747
0916
1043
1212
1636
0340
0820
1037
1448
173C
PH TUR
TEH 00
J I U ~r |
8.4
7.5
7.5
7.4
7.4
7.4
7.0
7.1
7.1
7.2
7.0
7.1
7.5
7.5
7.5
7.3
7.2
6.4
7.4
7.4
7.5
7.5
7.8
50
50
46
65
65
t>3
7.1
35
62
63
15
45
30
26
28
28
12
42
30
32
10.4
8.3
8.4
5.3
4.6
4.5
1.5
7.0
5.7
5.7
5.3
1.5
4.0
3.5
4.2
4.7
4.7
7.7
7.0
6.4
36 3.5
36 4.3
37 3.7
SUS
SOL
432
260
167
76
75
18
18
418
354
254
151
225
194
314
153
177
191
98
133
594
360
44
75
60
34
10
30
45
167
36
175
100
G8
VOL
SUS
SOL
265
133
ioo
57
51
11
•7
8
196
197
136
57
131
100
252
120
37
66
38
26
8
22
28
95
23
76
47
37
TOT
SOL
2S75
1667
126'J
005
595
542
534
512
1043
904
938
743
646
555
900
898
754
627
564
488
1023
680
423
711
427
271
224
217
570
506
318
385
287
259
TOT SET
VOL SOL
SOL
325
187
184
132
120
131
1O8
134
287
238
249
134
162
134
256
362
186
114
161
290
278
165
221
228
144
05
64
67
117
328
195
271
184
184
4
2
3
3
I
0
0
8
9
6
6
2
4
6
4
4
2
1
2
9
2
0
4
5
2
0
2
4
2
4
2
2
TOT
CL
119.-;
823
557
334
241
185
192
174
262
220
237
220
174
156
316
333
248
195
202
142
124
50
57
135
71
50
39
36
64
82
71
63
49
46
TOT
HARD
ur / 1
HL» / L
222
487
166
164
142
174
170
177
148
140
132
112
116
175
183
144
122
122
130
158
114
218
319
123
86
79
76
310
152
127
139
109
123
CA
HARD
167
140
124
102
101
114
113
127
160
133
130
140
128
264
140
170
135
109
110
121
111
101
156
230
93
73
71
63
224
103
91
117
95
112
CL
OEM
9.1
7.6
7.8
8.5
9.3
9.4
8.5
11.4
7.2
7.4
6.7
4.R
4.3
4.7
8.9
8.6
9.7
8.7
4.4
6.4
8.4
3.4
5.3
6.1
6.0
8.2
5.0
5.7
BOO
171
116
97
101
91
07
79
75
159
140
105
72
69
65
129
68
75
84
45
38
103
42
14
39
47
28
34
39
56
79
37
70
32
31
TOT
N
8.6
6.0
4.3
1.2
2.3
1.7
0.8
2.5
9.4
6.5
11.7
6.2
5.6
7.3
3.4
4.2
4.6
4.8
3.2
5.8
a.4
7.7
3.3
5.8
3.7
3.4
2.5
2.6
7.7
7.6
5.8
7.3
3.0
3.9
TOT
P
1. 1
1.0
0.9
1.2
1.0
1.0
1. 1
1.9
l.l
1.8
2.2
2.6
2.7
2.4
0.7
0.8
0.6
0.7
0.5
0.4
1.2
0.8
1.0
2.4
1.0
0.8
0.8
1.0
0.8
1.4
1.4
0.9
0.8
0.8
OIL
83.6
37.5
35.3
18.7
13.5
10.2
7.3
5.7
56.3
42.7
30.2
20.7
18. 1
20.9
14.5
1 ?.9
17.1
12.0
5. I
4.5
30.2
14.1
0.0
17.6
12.5
5.5
6.8
4.6
5.7
12.0
9.3
15.5
17.0
11.7
INR
C
19
10
10
a
7
9
9
12
21
20
16
16
15
14
25
36
15
26
19
28
15
19
16
32
8
2
2
j
6
18
12
14
2
4
ORC TOT FEC
C COLI COLI
197 2 1.3
96 2 0.7
72 3 0.4
49 1 0.2
49 2 0.3
36 0 0.?
24 1 0.2
25
97
101
91
57
58
50
83
29
47
44
22
20
117
71
24
45
44
31
12
25
33
56
24
65
37
35
1
2
2
2
2
2
2
1
2
1
7
4
3
1
0
1
6
1
5
2
4
0.4
0.7
0.7
0.7
0.6
0.4
0.4
1.0
0.6
0.3
3.0
2.3
1.0
0.6
0.2
0.9
2.9
0.7
2.9
8.3
2.5
FEC
STRP
V 1 0^
A i y
.059
.034
.018
.014
.076
.087
.059
.063
.420
.300
.310
.260
.096
.190
.420
.230
.095
.260
.220
.091
.060
.049
.056
.600
.160
.094
.200
.100
CUH
STORM
VOLUME .
1 rt-> c r
Id r \
618.77
1086.09
173J.28
3176.24
3239.95
6272.79
685 1.59
7265.50
0.00
20i.40
583.26
639. 15
790.81
671V. 95
645.45
70). SI
842.81
1015.93
1254.35
1551.72
3S.93
579.95
12134.89
935.80
56'>.06
6SU. 30
1122.30
195li.40
6152.47
462.37
8/1.09
361.28
1134.55
2149.13
STRM
SKPL
20
20
20
20
20
20
20
20
21
21
21
21
21
21
22
22
22
22
22
22
23
23
23
24
25
25
25
25
25
27
27
28
23
28
01
C2
03
04
C5
06
07
C3
Cl
C2
03
.04
05
06
01
C2
C3
04
05
CS
Cl
C2
03
01
01
02
03
C4
05
01
02
01
02
C3
-------
TABLE XIV
MILK RIVER PROJECT - SUMMARY OF EFFLUENT HATER QUALITY (CONTINUED)
DATE STflM :>OL Tl«fc fH TUK TEH
JIU
06 10 69
06 20 6
-------
One of the most significant characteristics of the overflow
at the MRPS was the generally poor quality of initial storm
samples. This is shown in terms of BOD and suspended solids
during Event 20 in Figures 19 and 20. The plot which is
fairly typical portrays the change of sewage quality with
time during a relatively continuous pumping event. Note
also, that chlorine demand remained relatively constant
during the entire storm period, even when BOD values were
extremely high.
Summaries of values of suspended solids, BOD, and chlorine
demand for initial and final influent samples for all events
in which two or more samples were available are summarized
in Table XV. The averages of these initial and final values
show the same characteristics as the continuous samples shown
above. BOD and suspended solids values diminish significantly
while chlorine demand remains relatively constant.
BIOLOGICAL QUALITY OF THE MILK RIVER AND LAKE SAINT CLAIR
The main objective of this work was to establish the biological
character of the Milk River channel and the immediate receiving
area of Lake Saint Clair. This was accomplished through
ecological surveys of bottom dwelling (benthic) organisms
and evaluation of the overall water quality and physical
features of the area. These studies were conducted primarily
to determine the condition of the Milk River Channel and
Lake Saint Clair before and after full-scale treatment of
combined sewer overflows with polymeric flocculants. Results
of the work are presented at this time because they are
helpful in gaining an understanding of the total Milk River
combined sewer overflow system.
Benthic Organisms as Pollution Indicators
Aquatic organisms differ in their response to changes in
water quality. Some forms survive and flourish in heavily
polluted waters while other forms are unable to withstand
even the slightest amount of pollution. This is true for
plankton and bottom organisms as well as for fish. Benthic
organisms are relatively stationary in the aquatic ecosystem
and have been used extensively as indicators of water quality.
Unlike fish and planktonic organisms which can selectively
move throughout the aquatic environment, bottom-dwelling
organisms have limited mobility and thus are valuable
indicators of past and present water conditions.
A general guideline in pollution biology is that a clean
water environment will support many different kinds of organisms
but, the numbers of individuals representing each kind are
low because of predation and competition for food and space.
60
-------
400". .
350...
300-..
250...
500-..
ISO-..
Figure 19
ANALYSIS OF COMBINED SEWER OVERFLOW
600.
550...
500...
450-..
INFLUENT
MIL* RIVER PROJECT
EVENT NUMBER EO
DATE EVENT BEGAN
TIME EVENT BEGAN 5 AM
SOLIDSr
BOO*
100-.
50...
H f
H 1 1 h
OJMUUSTIVe STORM VOLUME — 1OOO CU- FT-
-------
Figure 20
cr>
360..
330-.
300.,
270..
240..
210-.
1BO-.
2 150-.
120-.
90-.
60-.
30..
o..
8
ANALYSIS OF" CQVailsED SEWER OVERFLOW
INFLUENT MILIS RIVER PROJECT
EVENT NUMBER EO
DATE EVENT BEGAN Ix-X^BS
TIME EVENT BEGAN 5 AM
BOD. MGx1_
OCORINE DEMAND* MG^L
r • • r
• ••••*••••«•••••••
o o o o o o o o o o o o o o o o o o
.H
CUMULATIVE STORM VOLUME -- 1OOO CU> FT.
-------
TABLE XV
SUSPENDED SOLIDS, BIOCHEMICAL OXYGEN DEMAND, AND
CHLORINE DEMAND FOR SELECTED INFLUENT SAMPLES
Suspended Solids
mg/I
Biochemical
Oxygen Demand
mg/1
Chlorine Demand
mg/1
Event
8
9
11
12
14
. 15
16
17
20
21
23
24
25
27
28
29
33
Date
7-
7-
8-
8-
10
11
11
12
i -
o
o ~
4-
4-
5-
6-
6-
16-
22-
16-
19-
-18
-15
-16
-13
17-
24-
17-
21-
18-
68
68
68
68
-68
-68
-68
-68
69
69
69
69
69
5-69
8-69
6-20-
7-
17-
69
69
Initial
351
248
404
181
72
444
32
766
466
138
258
57
190
114
119
357
233
Final
156
110
46
110
26
57
50
116
46
24
72
52
40
44
78
53
169
Initial
53
126
62
58
95
216
23
360
376
101
128
53
56
62
48
71
45
Final
40
24
32
31
31
120
33
52
96
35
17
38
43
44
18
48
40
Ini
9
7
8
15
6
10
8
10
3
7
9
7
4
7
4
5
7
tial
.2
.6
.2
.1
.8
.9
.9
.2
. 5
.7
.0
.3
.5
.0
.7
.4
.2
Fi
8
9
6
17
6
9
12
5
10
4
8
6
9
6
8
3
7
nal
.0
.6
.1
.6
.6
.3
.2
.3
.1
.5
.9
.4
.2
.5
.3
.5
.6
Average
249
67
110
42
8.3
8.0
-------
The diversity of forms is substantially reduced in environ-
ments polluted by organic wastes due to unfavorable habitat;
absolute numbers of pollution tolerant organisms increase.
The life cycles of bottom-dwelling organisms are highly
variable from periods of a few days to a year or longer.
A qualitative examination over a long term of the different
kinds of organisms present ideally would include sampling
at least once during each of the annual seasons. Three
ecological surveys of bottom fauna were made during the
course of the project. Two surveys were made in 1968 prior
to the application of flocculants for treatment. The third
survey was made on August 12, 1969, after the flocculant
treatment program had begun.
Lake and Channel Sampling Program
The biologicalevaluation of water quality involves comparison
of data from clean water areas with those of suspected polluted
areas. Sampling points were selected at significant points
throughout the channel and lake area receiving the discharge
of the channel (Figure 21). Every attempt was made to obtain
representative samples from each area which would indicate
the extremes as well as the overall condition.
Off-shore sampling points in the lake were located by two
intersecting transit readings from reference positions on
shore. The lake sampling points close to shore and the
points in the channel were located by visual notation of
physical reference points. The sampling sites therefore
are relatively the same in the different surveys. Three
additional samples were added to the second survey on the
basis of the first results; a fourth new station was included
in the last survey.
Three different samplers (Petersen, Ponar, and Ekman) were
used in the surveys depending on the bottom conditions.
The Ekman dredge, which is only applicable in soft silty
material, was used for sampling in the channel areas. All
the samples in the lake were taken using either the Petersen
or Ponar dredge. Compositing was usually done to obtain
a representative sampling of the biological life present.
In many cases a single complete sample could not be obtained
because of the unfavorable bottom conditions such as cobble
and rock. All samples were carefully washed through a standard
30 mesh screen and preserved with formalin in labeled jars
for later laboratory examination. Observations of the nature
of the bottom condition at each sampling point were recorded
and water samples were collected. Dissolved oxygen and
bacteriological determinations were made at each sampling
point.
64
-------
FIGURE 21
fllLK RIVER CHANNEL AND ESTUARY WITH ZONES OF POLLUTION
14
14
©
14
13
Note:
13
GROSSE PTE
inYACHT CLUB
12 /
r~~l Severely Polluted
xxx Sludge Deposits
14
Transition Zone
xxx Silt Deposits
Cleorwoter
1500' 3000'
I INCH -1500'
65
-------
Additional qualitative sampling of the bottom was done to
determine the degree and extent of bottom deposition of
organic material. Qualitative sampling in the lake consisted
of visual observation of numerous dredge samples in selected
areas. The qualitative sampling in the channel bottom was
more extensive and included the development of channel profiles
at various cross sections. These profiles were developed
from data obtained with a long indexed pole at spaced intervals
on selected cross sections of the channel.
Benthic Character of Milk River and Lake Saint Clair
A composite summary of the ecologicalsurveys is presented
in Table XVI. A summary of the dissolved oxygen, coliform,
and bottom material descriptive data are contained in Table XVII
Based on these findings, each sampling station was subjectively
rated to be in a clean water condition, a polluted condition,
or in an intermediary transition condition.
There were diverse populations, including a variety of clean
water organisms, at Stations 1, 2, 3, 5, 6, 7, 16, 17 and 18.
These locations were designated as clean water stations.
The life-forms at Stations 1 and 7 seem to be depressed
slightly. This is likely a result of the nature of the
substrate rather than of water quality. The bottom at these
stations was gravel and rock. This type of bottom supports
a specific type of life and collection of representative
samples is difficult. These off-shore samples represent
the natural benthic populations in Lake Saint Clair which
are relatively influenced by pollution.
Stations 4 and 15 fall into the zone of transition. Although
significant numbers of life-forms are present, they are
being suppressed by some form of pollution. The number
of clean water organisms present at these stations is reduced
to only a few; pol1ution-tolerant forms have increased greatly
in number. The bottom material at Stations 4 and 15 consisted
of a deposition of fine organic ooze in contrast to the
sandy clay of the clean water points. The pollution in this
area is more probably in the form of substrate alteration
rather than deterioration of water quality.
The remaining sampling points (Stations 8 through 14) show
definite signs of gross organic pollution. The diversity
of species is reduced to only a few tolerant organisms which
are evident in great numbers. These stations located in
the Milk River channel and the immediate areas of the lake
are definitely greatly influenced by overflow from the MRPS.
The dissolved oxygen is low and very high bacteria counts
66
-------
TABLE XVI
COMPOSITE SUMMARY OF ECOLOGICAL SURVEYS
Station
1
2
3
5
6
7
16
17
18
4
15
8
9
10
11
12
13
14
Ave . No .
Organi sms
30
91
92
64
109
34
128
124
-
199
124
117
75
59
227
116
61
64
Ave. Organism
Diver si ty
8
12
14
9
15
8
12
10
12
8
8
5
3
2
1
1
1
1
Subjective
Classification
C
C
C
C
C
C
C
C
C
T
T
Po
Po
Po
Po
Po
Po
Po
C = Clean water
T = Transition
Po = Polluted
67
-------
TABLE XVII
QUALITY AND CHARACTER OF BIOLOGICAL SAMPLING LOCATIONS
CD
00
Sample
Station
1
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
IS
SPRING SURVEY
Dissolved Total
Oxygen Col iform
cnq/1 1/100 ml
13.1
12.8
12.0
12.6
12.6
13.6
13.6
10.2
10.0
8.6
4.8
4.2
-
-
-
-
-
<100
<100
<100
400
<100
<100
<100
870.000
76.000
57,000
13.000
11 .000
-
-
-
-
-
- 1968
Observation of
Bottom Material
Clean gravel and rock
Clay and sand
Clay and sand
Gravel , sand,
slight silt
Gravel , sand and clay
Clay and sand
Gravel and sand
Debris, some sewage
odor
Cl ean sand
Gravel and sand
Sludge ooze, sewage
odor
Sludge ooze, sewage
odor
Sludge ooze, sewage
odor
Sludge ooze, sewage
odor
-
-
.
Dissolved
Oxygen
mg/1
8.9
9.2
9.0
9.3
9.4
9.1
9.3
8.7
9.1
9.3
8.2
6.8
5.7
5.8
9.2
9.4
9.7
.
FALL SURVEY
Total
Col iform
1/100 ml
125
102
112
78
65
30
134
100
34
94
150
10.400
1 .800
2,300
TN
128
N.C.
.
- 1968
Observation of
Bottom Material
Gravel and rock
Clay and sand
Clay and sand
Heavy organic silt
ooze
Clay and sand
Clay and sand
Clean sand
Organic ooze, sewage
odor
Sand, some silt
Sand, some silt
Sludge ooze, sewage
odor
Sludge ooze, sewage
odor
Sludge ooze, sewage
odor
Sludge ooze, sewage
odor
Organic silt, heavy
ooze
Gravel and sand
Rock and sand
.
01 ssolved
Oxygen
nig/I
8.2
8.2
7.6
6.2
7.2
QUALITATIVE SURVEY - 1969
Total
Coli form
0/100 ml
Observation of
Bottom Material
170 Clay and sand
Heavy organlc silt
ooze
60 Clay and sand
1 70 Organ 1c ooze .
sewage odor
,on Dark sand
380 slight odor
3700 Sludge ooze,
clay, debris
Heavy organic s1lt
ooze
54 Gravel and sand
110 Gravel and sand
270 Debris and sand
-------
are present at these stations. The partially exposed bottom
of the channel approximately midway between the Pumping
Station and Lake Saint Clair is shown in Figure 22. This
condition existed during a partial dewatering of the channel
in the summer of 1968.
Figure 22
Bottom of Milk River Channel Partially Exposed During
Dewatering
Slight improvements in the conditions of the Milk River
channel were evident in results from the fall surveys. A
few tolerant midge larvae in addition to the dominant population
of sludge worms were evident. This is probably a direct
result of physical recruitment from channel flushing during
the summer months. Life in the channel is periodically
replenished by new organisms from the cleaner lake waters
during the flushing process.
69
-------
The quality of the lake water is undoubtedly influenced
by other effluents including effluent from the Chapaton
Combined Sewer overflow which enters the lake just north
of the Milk River channel. Sample Station 8 is located
directly off the mouth of this drain. Results for samples
from this station vary considerably between surveys. This
station was initially characterized by a few cleanwater
organisms and moderate diversity in the population placing
it in no worse than a transition condition. Greatly reduced
diversity and definite indications of severe pollution were
evident in the fall surveys. The lake area near the Chapaton
Drain prior to the spring survey was disrupted due to the
construction of a new pumping station at that time.
The influence of the poor quality water from Milk River
channel on the biota of the lake is shown by looking at
Stations 11, 10, 4, 5, and 6. Station 11 which is located
directly in the mouth of the channel is grossly polluted
with heavy organic sludge deposits and is characterized
by great numbers of tolerant sludge worms. Moving out into
the bay to Station 10 beyond the outfall pipe, the sludge
deposits are gone, undoubtedly due to the wind and wave
action, and yet the water quality is such that poor biological
populations still exist. Further out into the lake, at
Station 4, a transition to better conditions is definitely
evident. A few cleanwater organisms are present and the
overall population has become more diverse.
At Stations 5 and 6 cleanwater organisms are abundant and
the population has become more diverse. Station 6 is even
better than 5 but these differences could likely be in sampling
variation.
The conditions found at each station and the overall inter-
pretations are shown in Figure 21. The limits of these
areas, particularly in the transition zone, cannot be sharply
defined without further work. Further discussion of the
bottom depositions are made in the next section.
Qualitative Observations
A qualitative survey of the nature of the bottom material
in the channel and lake was made by observing conditions
at the sampling stations and at other strategic points.
As already noted, the bottom of the Milk River channel is
extensively covered with sludge deposits. The same conditions
were observed in qualitative samples taken from other boat
canals which enter directly into the bay area near the discharge
of the Milk River. Heavy sludge deposits were also found
in the bay throughout the protected area created by the
pier shown in Figure 21. No sludge deposits were observed
70
-------
south of a line extending from the end of the pier to a
point south of sampling point 11 at the south side of the
channel mouth. The wind and wave action apparent washes
clean the area south of this line. Deposition is pushed
back into the channels or out into the lake depending upon
wind direction.
Channel profiles developed from cross-sectional depth sampling
are shown in Figure 23. Sludge deposits exist at any location
where the flow conditions in the channel are favorable for
deposition of material, e.g. the inside bank of curves,
back eddies, etc. The extent of sludge deposition was so
significant that it was impossible to assess changes brought
about by chemical treatment of the storm overflow in the
Mi 1k River.
Effect of Milk River Channel on Storm Overflows
The Milk River Channel has a unique role in the overall
effect on the combined sewer overflow from the MRPS. As
shown, significant sludge deposits are in evidence in the
channel. These deposits undergo rapid decomposition during
the summer months causing serious odor problems and depletion
of dissolved oxygen in the entire length of the channel.
In an effort to evaluate the cause of these deposits in
some quantitative way, the flow conditions in the channel
were analyzed considering the conditions of pumping which
prevailed from 1960 through 1969. Before summarizing the
results of the analyses, however, it is necessary to point
out certain physical characteristics of the channel.
TABLE XVIII
PHYSICAL CHARACTERISTICS OF THE MILK RIVER CHANNEL
Length (to Lake Saint Clair) 6,000 ft
Area (cross-sectional) 625 sq ft
Estimated Volume (to Lake Saint Clair) 3,750,000 cu ft
28,100,000 gal
Mean Depth 7.5 ft
Volume of Milk River Retention Basin 512,000 cu ft
3,830,000 gal
71
-------
Figure 23
OUTLINE PLAN OF MILK RIVER DRAIN
SHOWING
CROSS SECTION PROflLES AT SELECTED STATIONS
INO 0>TC
JTBUCIUBt
0
I
180'
i
360'
i
I INCH = 180'
72
DATE: OCTOBER, 1968
-------
Based on these approximations and an analysis of the fre-
quency distribution of rate of pumping and pumping volumes,
the following data were calculated.
TABLE XIX
RATE OF PUMPING
Fl ow
(cfs)
187*
305
410
610
715
2450
Rate
Occurrence
U)
50
61
91
93
95
100
*Geometric Mean
Channel Velocity
(fps)
0.30
0.49
0.66
0.98
1.14
3.92
TABLE XX
VOLUME OF PUMPING
Detention (min)
Channel
335
205
152
102
88
25
Basin
46
28
21
14
12
4
Total Volume Pumped
Cu Ft
1,000,000
1 ,200,000*
2,000,000
3,000,000
4,000,000
Gal
7,480,000
8,970,000
14,960,000
22,440,000
29,920,000
Volume Occurrence
(%)
40.3
50.0
61 .3
75.2
83.4
*Geometric Mean
The Milk River Channel is serving as an effective treatment
system in its own right as clearly shown by the above data
and supported by the results of the channel sampling program.
It is probably significantly more effective in preventing
fresh solids from getting into Lake Saint Clair than the
existing retention basin. Of the combined sewer overflows
which occurred from 1960 through 1969, 95 percent produced
channel velocities equal to or less than 1.14 fps while
91 percent of the overflows produced velocities equal to
or less than .66 fps. These velocities alone would permit
deposition of most sewage type solids, but another important
73
-------
factor is noted in Table XX. In over 83.4 percent of the
storms, the volume of the storm overflow was not sufficient
both to fill the retention basin and to displace the volume
in the Milk River Channel up to Lake Saint Clair. Since
there is no dry weather flow in the Milk River, most suspended
solids pumped into the channel are allowed to settle under
quiescent conditions for periods of from several hours to
several weeks.
74
-------
SECTION 6
TREATMENT OF COMBINED SEWAGE WITH POLYMERIC FLOCCULANTS
One of the primary objectives of the Milk River Contract
was to study the problem of chemically treating large volumes
of continuously flowing combined sewage at the Milk River
Pumping Station (MRPS) with relatively low concentrations
of organic water-soluble polymeric flocculants and measure
the clarification potential of the process. The work involved
a program of laboratory, pilot, and full-scale testing designed
to demonstrate the practicability of the process and to
outline the problems associated with the evaluation and
operation of such a system.
In general, this type of chemical treatment can be divided
into two operations, flocculation.and sedimentation. Floccula-
tion involves mixing the combined sewage with a polymeric
flocculant under proper conditions to generate a floe with
desirable settling properties^. Sedimentation involves
the removal of the flocculated material along with other
heavy solids under quiescent conditions. The clarified overhead
is discharged to the receiving water. Flocculation studies
at the MRPS included preliminary screening of flocculants
in jar tests, assessment of optimum flocculant systems in a
long-tube sedimentation column, and full-scale application
of flocculants. The sedimentation studies involved measure-
ment of settling rates of samples of storm flow flocculated
using a long-tube settling column and evaluation of retention
efficiencies of the basin using a hydraulic model of the
Milk River Retention Basin.
FLOCCULATION STUDIES
A study of flocculation of any waste with chemical flocculants
requires evaluation of at least three factors:
a. The capability of the material(s) selected to
react with the wastewater and form a settleable
floe.
b. The reaction times required between chemical addition,
dispersion, and floe formation.
c. The conditions of mixing required to develop
satisfactory or optimum floe.
Experience in the treatment of sanitary sewage, rivers, and
industrial wastes has demonstrated the importance of providing
adequate time for dispersion of flocculants. The time required
75
-------
for dilute but viscous polymeric flocculant solutions to
disperse and react in most wastewaters under the condition
of flash mix (high shear gradient) is approximately 0.5 to
1.0 minutes. Flocculation of most wastewaters can be accom-
plished under conditions of gentle mixing in an effective
time of from 1 to 3 minutes. It has also been found that
a reasonable estimate of the potential clarification efficiency
of a polymeric flocculant treatment system can be obtained
by measuring overhead clarity in a sample flocculated and
settled in a 1500 ml beaker under standardized conditions
of mixing.
The flash mix required to accomplish polymeric flocculant
dispersion corresponds to an approximate shear gradient greater
than 50 sec-1 Or a Phipps and Bird lab stirrer speed of approxi-
mately 90 rpm in a 1500 ml beaker. A shear gradient of
approximately 10 sec~l or a Phipps and Bird mixing stirrer
speed of about 40 rpm in a 1500 ml beaker is required for
f1occulation. The effective retention times and mixing require-
ments for most polymeric flocculants in wastewater systems
are summarized in Table XXI.
TABLE XXI
CONDITIONS FOR DISPERSION AND FLOCCULATION
C r j t e r i o n Pi spersion F1occulati on
Mi nimum time 0.5 1
Optimumtime 1 3
Mixingconditions Flash Gentle
Approximate shear gradient
(sec'1) 50-80 >10
Gang stirrer setting (rpm) 90 40
The initial work at Milk River was devoted to screening various
types of polymeric flocculants for their ability to react
with combined sewage to form a settleable floe using these
guidelines for mixing conditions and reaction times.
JAR TEST SCREENING
The standard conditions for jar testing adopted for the screening
of flocculant efficiencies were: one minute of flash mix
at 100 rpm, three minutes of flocculation mix at 40 rpm,
and five minutes of settling at 10 rpm using a Phipps and
Bird stirrer. These values are approximately the same as
the effective values shown for optimum conditions in
Table XXI. A 1000 ml volume of the material to be flocculated
76
-------
was contained in each of six 1500 ml beakers. Qualitative
measurements of flocculation rate, floe size, settling rate,
and overhead clarity were routinely recorded. Initial sample
turbidity and final overhead turbidity were measured instru-
mental ly.
The screening studies were designed to evaluate the potential
effectiveness of polymeric flocculants having a broad range
of chemical characteristics. The PURIFLOC products investigated
during this study are classified according to ionic properties
and basic formulation in Table XXII.
TABLE XXII
CLASSIFICATION OF PURIFLOC FLOCCULANTS
An ionic Cat ionic Nonionic
A21 Sodium Polystyrene C31 Polyamine Nil Polyacrylamide
Sulfonate
A22 Polyacrylamide C32 Polyethylen- N12 Polyacrylamide
i m i n e
A23 Polyacrylamide ET 721 Polyacrylamide*
*Experimental flocculant, no PURIFLOC designation
The eight flocculants were evaluated at various concentrations
for potential clarification efficiency utilizing the jar
testing procedures outlined above. The tests were conducted
on samples obtained at various periods during Events Nos. 1-8.
The results of the laboratory jar tests in terms of flocculation
efficiency as measured by overhead turbidity in treated
and untreated samples are summarized in Table XXIII. The
effect of storm period on relative influent turbidity and
flocculation efficiency are noted. The following conclusions
were made based on these initial screening tests:
1. The anionic and nonionic polymeric flocculants
exhibited no significant flocculation activity and
were not effective in reducing influent turbidity.
2. Two of the cationic flocculants, PURIFLOC C31 and
PURIFLOC C32, consistently exhibited excellent
flocculation activity and effective turbidity
removals at concentrations of approximately 10 mg/1
during the early to mid-storm periods when suspended
solids loadings were relatively high.
77
-------
—I
CD
TABLE XXIII
RELATIVE FLOCCULANT ACTIVITIES AS MEASURED BY TURBIDITY
BAROID
Event
No.
1
1
1
2
2
3
3
3
4
4
5
6
7
7
7
7
7
8
23
25
Raw
375
450
160
235
160
260
280
215
170
98
235
178
258
220
290
200
175
370
100
180
Settled
Control
350
425
110
200
100
245
203
160
160
95
220
125
220
190
265
195
125
180
76
155
A21*
(2
350
425
110
-
100
220
-
-
160
95
-
120
-
190
260
190
-
155
76
155
A22* A23*
!.0 mg/ll
-
-
100
220
-
160
-
95
-
-
-
-
260
-
-
155
-
155
280
425
no
-
100
230
-
160
145
95
210
-
-
-
260
190
-
-
76
.
C31* C32*
~7T5 moTT)
62
95
110
75
85
30
65
90
60
80
68
72
50
90
70
98
70
62
75
85
67
100
110
90
76
62
48
60
55
80
55
68
40
52
50
70
60
90
75
80
Nil*
TZTO
350
-
100
-
-
160
-
95
-
125
-
-
180
170
-
130
-
140
N12*
350
-
100
245
203
-
-
95
215
-
-
170
240
190
-
130
76
155
ET-721
(10 mq/1)
375
175
-
-
-
100
-
90
-
-
-
135
140
120
-
60
-
-
HECTORITE ,+ C32*
15+10 mg/1 res .
;
-
-
-
40
40
22
35
45
21
-
-
10
17
35
47
65
50
*PURIFLOC products
-------
3. During the mid- to late-storm periods, when suspended
solids loadings were relatively low, PURIFLOC C31
and PURIFLOC C32 exhibited relatively poor flocculation
activi ty.
4. Final overhead turbidity was relatively constant
after treatment by PURIFLOC C31 and PURIFLOC C32
during all storm periods; final overhead quality
is apparently independent of initial sample turbidity.
The characteristic decrease in suspended solids and turbidity
of combined sewage during the course of a storm overflow
limited the potential effectiveness of a single flocculant
treatment system. Previous experience had shown that in
systems of low solids, floe building nuclei could be added
to supplement the natural solids and achieve high efficiency
clarification. During Events Nos. 3 and 4, two mont-
morillionite clays, BAROID CERCLAY® and BAROID HECTORITE®,
were tested as coagulant aids. The BAROID HECTORITE clay,
a calcium-magnesium silicate, was found to be far superior
in effectiveness to the BAROID CERCLAY, an aluminum silicate.
The significant improvement in removal of turbidity obtained
by addition of BAROID HECTORITE clay to PURIFLOC C32 over
the PURIFLOC C32 system alone is shown in Figure 24. During
screening studies of the clay-cationic flocculant system
a minimum dosage of approximately 15 mg/1 of clay was necessary
to achieve a significant increase in effectiveness over
the cationic flocculant alone. Clay concentration was critical
especially when compared with effective polymeric flocculant
concentrations. The optimum concentration of polymer flocculant
dosage is quite broad, even in the dual systems. It was also
observed that clay addition was more effective when it preceded
flocculant addition by approximately 30 seconds of intervening
mixing.
In addition to the clay-cationic polymer flocculant systems,
several other dual and tertiary systems were also evaluated
as well as several cationic blends which exhibit the properties
of a dual cationic-anionic system. Those systems which
provided flocculation activity and turbidity removal efficiencies
greater than that of the cationic polymer flocculants alone
were:
a. Ferric chloride + PURIFLOC A23
b. PURIFLOC C31 + PURIFLOC A23
c. BAROID HECTORITE clay + PURIFLOC C31 + PURIFLOC A23
d. BAROID HECTORITE clay + PURIFLOC Nil + PURIFLOC C31
79
-------
00
o
300
Figure 24
MILK RIVER STORM EVENT
6/25 TO 6-26-68
SETTLED CONTROL^
ISmg/lBAROID
HECTORITE + IOmg/l C32
4:40
5:40
12:40
-------
It was also noted during the course of these system screening
studies that very rapid formation of floe occurred when
anionic and cationic polymeric flocculants were used in
a dual system. Rapid floe formation also occurred when
PURIFLOC Nil was used between the addition of clay and the
addition of the cationic flocculant. This capability to
significantly increase the rate of floe formation can be
utilized to great advantage in systems where flocculation
time is short. Increased flocculation rates appear to be
accompanied by decreases in solids capture efficiency.
The following conclusions were made from the dual and tertiary
system polymer screening studies:
1. Coagulant aids, specifically 15 mg/1 of BAROID
HECTORITE clay, can be utilized with the cationic
flocculant systems to produce significantly improved
turbidity removal from the MRPS combined sewage.
2. The rate of floe formation can be increased signifi-
cantly by the addition of anionic polymer flocculants
to the cationic flocculant systems, but may result
in a corresponding decrease in solids capture efficiency
FULL-SCALE APPLICATION
Treating a periodic waste flow ot up to a million gallons
per minute with polyelectrolyte flocculants is a problem
of monstrous proportions. It was therefore necessary to
design a flocculant feed system on a much larger scale than
had been previously applied in the treatment of continually
flowing municipal wastewaters. A frequency distribution
was developed for the rates of pumping for eight previous
years of record at the MRPS. The combined sewage flow was
less than or equal to 680 cfs (305,000 gpm) for 90 percent
of the time; the flow was ^.1100 cfs (493,000 gpm) for 95
percent of the time. The maximum flow possible is 2450 cfs
(1,100,000 gpm). A cost-benefit analysis was made to
determine the amount of capital required to treat 90, 95,
and 100 percent of anticipated volume. It was decided that
the upper limit of flow to be treated with flocculants would
be 1100 cfs (the 95 percent occurrence). The capital outlay
to treat the final 5 percent of the flow situations is double
that for the first 95 percent of flow.
Two different chemical dispersal systems were required:
one for dispersion of dry anionic and nonionic polymers,
and the other to dilute and disperse liquid cationic polymer.
A sub-contract was let to the Hague Equipment Company of
Evansville, Indiana, to design and construct an automatic
dispersing unit for dry chemicals capable of a continuous
81
-------
flow of TOO gpm. Previous experience by The Dow Chemical
Company with this supplier permitted direct expansion of
the basic components of this automatic disperser from the
usual 25 gpm capacity up to a 100 gpm capacity. Operation
of this unit for dispersing dry anionic flocculants was
trouble-free.
Dilution of the relatively large amounts of cationic flocculants
was best accomplished using an in-line blender. The cationic
flocculant in liquid form was first pumped through a positive
displacement pump into the in-line blender. Water was intro-
duced and the two phases instantly mixed in a high shear
enclosure. The polymer solution was stored after dilution
in a holding tank. Depending on the polymer concentration
required, flow rates up to 100 gpm were possible using the
in-line blender. This system proved to be very effective
for dispering cationic flocculants during several events.
Seven variable speed chemical metering pumps were required
to pump the dispersed polymer into any combination of the
seven storm lift pumps. Gear pumps with a capacity of 50
gpm fitted with variable speed drives were reliable and
effective. A schematic of the complete feed system is presented
in Figure 25.
Based on the initial jar test evaluations, PURIFLOC C31
and PURIFLOC C32 were selected for the first full-scale
feedings of flocculant. There were four events (Nos. 16,
23, 27, and 30) in which flocculants have been applied through-
out. There were six other events in which continuous feeding
was not achieved because of various technical problems or
sampling difficulties.
Selection of the proper point for flocculant addition was
critical. The difficulty of selection is compounded by
the relative high flow rates and minimal detention times
available for flocculation at the MRPS. Flocculant was
first fed full-scale during Event 15. A dosage of 18.8
mg/1 PURIFLOC C32 was fed for the entire 21 minutes of the
third pumping. The point of flocculant addition was located
about half way up the casing on the discharge side of storm
pump #3. Samples from various basin points during this
feed contained small-medium sized floes. Two problems were
immediately apparent: (1) a significant portion of floating
floe was noted, and (2) much floe was "popping" over the
effluent weirs.
82
-------
Fiqure 25
SCHEMATIC - POLYMER FEED SYSTEM
DRY
POLYMER
H,0
AUTO POLYMER
DISPERSER
7-50 GPM
METERING
PUMPS
oo
GO
110 GAL.
LIQUID
CAT. ST.
-------
A condition of floating floe is usually indicative of excessive
concentrations of flocculant or improper mixing conditions
with possible air entrainment. Results from jar tests conducted
during this event demonstrated that 20 mg/1 PURIFLOC C32
was not an overdosed condition. Improper mixing was con-
cluded. Careful examination of the floating floes showed
small bubbles of air attached to each floe particle. This
air was probably trapped on the floes during the violent
aeration immediately after discharge of the pumps onto the
distribution plane of the basin (first concrete apron).
The addition point was changed to a position near the second
apron downstream of the first apron. Visual observation
of the basin operation showed that the aeration effect was
reduced at this new addition point. This second addition
point was used when PURIFLOC C32 was fed at a dosage of
15 mg/1 during Event 17. Very small floes were observed
in the basin; good flocculation was apparent in the jar
tests. The shear gradient was apparently too low at this
addition point to effect proper flocculation.
The third and last addition point tested was located at
the first apron in the vicinity of the violent agitation
created by the pump discharge. The shear gradient was relatively
high at this point. Very good rapid mix conditions were
available for proper dispersal of flocculant. The area
of the second apron provided a relatively lower shear gradient
in which floe could grow. Excellent flocculation was noted
using this addition point. The previously observed flotation
problems were minimal. This addition point was utilized
for all of the remaining full-scale additions.
Samples collected in the basin during all subsequent appli-
cations of flocculant contained medium to large floes. Capture
of fine solids into the larger floes was fair. The single
flocculant systems of PURIFLOC C31 or PURIFLOC C32 were
used during the early periods. Visual observation of samples
collected in the vicinity of the overflow weir indicated
extensive carry-over of floe.
It was known that acceptable flocculation was occurring
in the basin, but proper sedimentation of these floes was
not occurring. How could flocculation be quantified? A
good measure of the extent of flocculation of a particular
system was afforded by the measurement of the volumetric
settleable solids; i.e, the volume of settleable matter
contained in an Imhoff cone recorded after one hour of sedimen-
tation expressed as volume of sludge/volume of sewage (ml/1).
Untreated influent samples were initially assumed to contain
84
-------
less settleable solids than a treated sample from the basin
or the effluent sample. Little sedimentation was thought
to be occurring in the basin. This was not always the case,
however, even when several samples were averaged to minimize
the inherent time lag between influent and effluent samples.
Selected data from samples collected by the automatic samplers
during Event 23 are shown in Table XXIV.
TABLE XXIV
SETTLEABLE SOLIDS FOR SELECTED
STORMWATER SAMPLES OF EVENT 23
Settleable Solids, ml/1
Influent Effluent
First pumpage 3.2 8.7
Second pumpage 1.7 2.5
Thi rd pumpage 1.5 1.5
The average values of five samples collected in intervals
at both the influent and effluent stations for the same
Event showed an influent value of 9.7 ml/1 and an effluent
value of 7.9 ml/1. This positive degree of removal is not
in agreement with the above data.
Another possible method to quantify the flocculation process
was to measure sludge depths and develop profiles of the
basin bottom. Because of dramatic quality changes both
within single events and between two or more events, no true
"control" situation existed, and this method was impractical.
Because of the inherent settling inefficiency of the Milk
River basin and the practical difficulty of controlling
flocculant concentrations during short periods of pumping,
optimum full-scale flocculation was difficult to achieve.
The extreme variations of sewage quality and the inherent
inefficiency of the basin made performance evaluation without
a control impossible. In addition, the model and the long-
tube sedimentation portions of the study had shown that
the hydraulic inefficiencies of the system will not permit
satisfactory settling to occur.
For these reasons, considerable effort was devoted to the
long-tube settling tests in order to establish those parameters
necessary to engineer and design an efficient basin for
chemical treatment of combined sewer overflow.
85
-------
STREAMING CURRENT DOSAGE CONTROL
An automatic control device for chemical addition would
greatly aid application of flocculants for two reasons:
(1) a shortage of qualified operators in the waste treatment
field, and (2) the extreme variation in waste quality. Previous
experience with the streaming current detector within The
Dow Chemical Company^*? nas shown that it provides a means
of controlling the concentration of cat ionic flocculants
added to domestic sewage. This device develops a streaming
flow of fluid confined in an annular space which shears
the electrical double layer. The movement of this diffused
double layer creates a measurable streaming current indicative
of the character of the electrical charges of colloidal
suspens ions.
Streaming current measurements were made during laboratory
f1occulations of a sample collected during Event 27. The
results are presented in Table XXV. For Influent No. 1,
when the solids concentration was high, it was apparent
that the streaming current closely followed the traditional
charge approach to flocculation; that is, optimum flocculation
was seen to occur at or near (slightly on the positive side)
the point of zero streaming current. It was interesting
to note the non-linearity of the streaming current with
increasing dosages of PURIFLOC C31. This response is shown
in Figure 26. The streaming current of raw sewage is usually
in the range of -20 to -25 microamps.
For Influent No. 2, with a much lower solids loading, the
condition of near zero streaming current did not relate
to acceptable flocculation. It was seen that 15 mg/1
PURIFLOC C31 reversed the charge of the system, but that
clay was necessary to provide bridging sites in this low
available surface-site system.
Optimum flocculant concentration apparently could be related
to streaming current, but optimum flocculation during low
solids conditions required a given quantity of flocculant
aid in addition to the polyelectrolyte. Since the dosage
of flocculant aid from previous jar testing appeared to
be constant, an automatic system to predict polymer demand
and feed polymer seems practicable.
86
-------
TABLE XXV
COMPARISON OF STREAMING CURRENT AND OTHER
QUALITATIVE CRITERIA OF FLOCCULATION
Influent No. 1 (High solids, turbidity = 70 JTU)
Streaming Floe
Sample Current (u amps) Size Clarity
Blank
5 mg/1 C31
10 mg/1 C31
15 mg/1 C31
20 mg/1 C31
50 mg/1 C31
Influent No.
Blank
10 mg/1 C31
15 mg/1 C31
20 mg/1 C31
30 mg/1 C31
-12.2
-12.0
-10.8
- 0.6
+ 0.8
+ 8.4
2 (Low sol
-14.0
- 8.4
+ 0.5
+ 2.5
+ 7.0
15 mg/1 Baroid Hectorite
+ 15 mg/1 C31 + 0.4
Key: Floe Size
PP = Pinpoi
S = Small
nt
None
PP
S
M
M-L
S-M
ids, turbidity =
None
S
S
S-M
S
M
Clarity
P = Poor
F = Fair
P
P
P
F-P
F
F-P
35 JTU)
P
P
P
P
P
F
Rating
5
4
4
2
1
3
4
3
3
2
3
1
M = Medium
L = Large
87
-------
-20
Figure 26
STREAMING CURRENT MEASUREMENTS
EVENT NO. 27
0
10 15 20 25
DOSAGE, PURIFLOC C3I , mg/l
30
88
-------
STABILITY OF PURIFLOC FLOCCULANTS
Over normal periods of storage, the PURIFLOC flocculants
are quite stable. Periods of storage should not exceed
loss of activity due to slight natural
This limitation is more important
are high or after a package or
PURIFLOC flocculants can be
adversely affected by temperatures over 48.8°C (120°F) either
in bulk as dry powders or liquids, or in dilute solutions.
Bulk powder or liquid products generally will not approach
this temperature unless confined in unventilated
or stored or utilized near steam lines or
one year to avoid
degradation of flocculant
when storage temperatures
container has been opened.
spaces,
heating equipment.
The dry PURIFLOC flocculants are hygroscopic and will absorb
moisture from the atmosphere. Opened bags should be stored
only in low humidity environments and for short periods
of time to prevent caking. The liquid PURIFLOC flocculants
should be kept from freezing and preferably stored inside
to prevent weathering of the containers and to reduce increases
in viscosity due to lowering temperatures.
The recommended times of maximum storage for bulk PURIFLOC
products and selected solutions are summarized in Table XXVI.
TABLE XXVI
RECOMMENDED MAXIMUM STORAGE TIMES
PURIFLOC F.LOCCULANTS
OF
Product
PURIFLOC Nil
PURIFLOC N12
PURIFLOC A21
PURIFLOC A22
PURIFLOC A23
PURIFLOC C31
PURIFLOC C32
ET-721
Bulk Product
One year
One year
One year
One year
One year
One year
One year
One year
Stock Solution
Cone. % Storage
3.0
3.0
3.0
2.0
0.5
All
All
5.0
1
2
1
1
1
2-3
2-3
1
month
weeks
month
month
month
months
months
month
Laboratory
Working Soln
0.1
0.1
0.1
0.1
0.05-0.1
0.1 -10.0
0.1 -10.0
0.1
The dry powders of PURIFLOC Nil, N12, A21, A22, A23, and
ET-721 flocculants should not be stored for longer than
12 months. Stock solutions are stable for a period of at
least one month, with the exception of PURIFLOC N12. More
89
-------
dilute (0.05%) solutions should be prepared daily by dilution
of the concentrated stock solutions. Liquid cationic flocculants
are considered stable for 12 months or longer from time
of manufacture. Dilute solutions of the cationic flocculants
are stable for at least one month.
Some method of determining polymer degradation is necessary
to express "storage stability" or "shelf life." The most
easily measured parameter of degradation is the change in
viscosity of a flocculant solution. Other parameters such
as molecular weight (derived from viscosity measurements)
and flocculant activity on selected substrates are also
used. Unfortunately, viscosity does not completely char-
acterize a flocculant without other parallel analyses to
determine flocculant activity. A laboratory evaluation
of flocculation activity is best for evaluating the storage
stability of such materials, but there is no universal substrate
that has adequate chemical and physical characteristics that
can be used to satisfactorily screen all types of flocculants.
Caution should be exercised when relating viscosity to
effectiveness as a flocculant. A material classed as a
flocculant may have a specified level of viscosity and still
have no activity as a flocculant. This can be due to contam-
ination of the prime flocculant with materials which may
cause gelation, precipitation, or other reactions which
may or may not alter viscosity, but cause considerable loss
of activity.
Under controlled laboratory conditions, viscosity as a para-
meter for studying stability or degradation of a flocculant
material is accepted as an indication of what will occur
in actual practice. Loss of viscosity under controlled
conditions of storage can be attributed to flocculant
instabilities.
The polyacrylami de-based materials are the most susceptible
to degradation of the several PURIFLOC flocculants. PURIFLOC
N12 is adversely affected when stored as a solution. It
degrades to 50 percent of its original viscosity in two
weeks. The dry product is the least stable of the products
considered, retaining about 75 percent of its original activity
after one year.
PURIFLOC flocculants Nil, A22, and A23 also degrade under
the same conditions as N12, but to a lesser extent. Solutions
of PURIFLOC A22 degrade to about 83 percent of their original
viscosity in two weeks. The dry powder will retain essentially
all of its original activity for one year. PURIFLOC A23
90
-------
has the same stability characteristics as PURIFLOC A22.
The stability of solutions of PURIFLOC Nil is about the
same as PURIFLOC N12. It is manfactured as a non-hydrolyzed
polyacrylamide. In solution it will slowly hydrolyze into
a form chemically similar to PURIFLOC N12. An activity
of 93 percent of the original is retained after one year.
The viscosity of PURIFLOC A21 is not clearly related to
its activity as a flocculant. A test to measure the efficiency
of this flocculant consists of adding the flocculant to
a concentrated aqueous slurry of MINCO BOND® clay, allowing
floe to form, and timing the fall of the solids-liquid interface
over a predetermined distance.
Dry PURIFLOC A21 degrades to about 65 percent of its original
flocculant activity after one year of storage. A one percent
solution degrades to less than 50 percent of its original
activity in two weeks. Stability improves considerably
as the concentration of the solution is increased. The
maximum recommended concentration (3.0 percent) will remain
nearly 100 percent active for at least one month.
At 30°C (86°F), PURIFLOC C32 retains about 80 percent of its
original viscosity after three months storage. For the same
storage time at 10°C, 90 percent viscosity is retained.
About 65 percent of the original viscosity of diluted solutions
of PURIFLOC C32 remains after three months storage at 30°C.
The temperature stability of PURIFLOC C31 is similar to
that of PURIFLOC C32. After three months bulk storage at
room temperature, PURIFLOC C31 retains about 90 percent
of its original viscosity. Dilute solutions of PURIFLOC C31
retain about 65 percent of their original viscosity after
three months storage at 25°C. Reductions in viscosity of
1 percent solutions of PURIFLOC C31 and PURIFLOC C32 upon
storage are shown in Figure 27.
Storage stability is necessary for design of flocculant
dispersion and storage facilities. Liquid organic polymeric
flocculants are similar to other common inorganic chemicals
in respect to the types of storage tanks, piping and pump
equipment used. Common materials of construction with the
exceptions of zinc or aluminum and their alloys can be used
for handling the organic flocculants. Solutions of organic
flocculants can tolerate metal ions such as calcium, magnesium,
and iron at concentrations normally found in water systems
used in preparing and diluting stock and working solutions
of flocculants. An excess of certain metal ions developed
as corrosion products in the flocculant system can reduce
flocculant activity. Trivalent cations, certain divalent
cations, and some high molecular weight organic cations
react with anionic flocculants.
91
-------
ro
Figure 27
PERCENT VISCOSITY RETAINED (1% SOLUTION)
1001
90
80
70
60
50
PERCENT
VISCOSITY RETAINED
30
20
10
0
0 10
20
C3I AT 25°C
30
40
I
50
DAYS
60
70
80
90
100
-------
Bulk handling and dissolution of the flocculants require
specialized equipment. Excessive shear rates, extreme tempera-
tures, and make-up water of poor quality should be avoided.
The minimum dissolution time ordinarily required for maximum
concentrations are on the order of 10 minutes. Mixing and
detention time can become critical in sizing tanks for large
scale operations. Efforts to reduce the time of dissolution
by using high-shear-rate mixing equipment usually leads
to degradation of flocculant by chopping the long polymer
molecules into shorter, less effective chains. High-shear
equipment includes propeller type mixers exceeding 350 rpm
in tanks of moderate volume (300 gallons). High shear rates
may also occur in piping designs with many 90° bends and
other obstructions. Long sweeping bends or hoses are desirable
to minimize localized high flow velocity points. Air spargers
can be installed to provide agitation in lieu of propeller
mi xers.
TOXICITIES OF PURIFLOC FLOCCULANTS
In addition to obtaining effective f1occulation , there are
two other important criteria for a combined sewage treatment
utilizing polyelectrolyte flocculants.
1. The flocculants used must be relatively non-toxic
to humans, animals and aquatic organisms.
2. The effective shelf life of the dry flocculant and
its solution must be sufficient to prevent a
significant loss of treatment activity between events.
Following are the results of studies conducted by The Dow
Chemical Company relating to the toxicity and storage of
the polymeric flocculants used in the Milk River Combined
Sewer Overflow Project. These data are specific to the
flocculants listed in Table XXVII.
Fish Toxicity
Flocculants are generally classified as to their ionic nature,
functional group or chemical structure. The PURIFLOC series
can be divided, from the standpoint of toxicity, into polyacryl
amides, SPSS, and polyamines, with the functional groups -NH?
and -COO", -S03~, and -NR3+ respectively.
Nitrogenous cationic flocculants apparently can limit the
ability of fish to obtain oxygen from the water. These
flocculants apparently flocculate the secretion from the
gills of the fish, thereby coating the oxygen-absorbing
membranes and reducing the transfer of oxygen. The starting
monomer for PURIFLOC C31 has a fish toxicity maximum safe
93
-------
TABLE XXVII
PURIFLOC FLOCCULANTS EVALUATED FOR
TOXICITY AND STORAGE
Flo ecu!ant Descri ption
PURIFLOC N17 polyacrylamide - potable water grade
PURIFLOC Nil polyacrylamide - minimum hydrolysis
PURIFLOC N12 polyacrylamide - low hydrolysis
PURIFLOC A21 sodium polystyrene sulfonate
PURIFLOC A22 polyacrylamide - 30% hydrolysis
PURIFLOC A23 high molecular weight polyacrylamide -
25% hydrolysis
PURIFLOC C31 polyalkylene polyamine
PURIFLOC C32 polyethylenimine
ET-721 low hydrolysis polyacrylamide made partially
(developmental) cationic with dimethylamine via the Mannich
reacti on
limit in excess of 100 mg/1 . After polymerization the same
chemical, due to its resulting flocculant characteristics,
has a fish toxicity maximum safe limit of 2 mg/1 of polymer
in solution.
The polyacrylamide flocculants dissolved in water in gross
amounts are toxic to fish but in a different manner than
the polyamines. The polymer is not chemically toxic, but
the increased viscosity of the water caused by the polymer
decreases the mobility of the fish thereby causing suffocation
The maximum safe limit of acrylamide monomer is greater
than 100 mg/1 which is essentially the same as the polyamines.
Fish toxicities were compiled by the Midland Division Waste
Control of The Dow Chemical Company and listed in Table XXVIII
This table is a guideline for situations in which there may
be residual polymer in flocculant-treated waters.
As can be seen from Table XXVIII, the maximum safe limits
of polyacrylamide-based anionic flocculants are from 500
to 1000 mg/1. These high residuals seldom if ever occur
in normal flocculation practice. With the exception of
PURIFLOC A23, 1000 mg/1 of polyacrylami de-based flocculants
can be tolerated by fathead minnows without adverse effect..
A concentration of 2500 mg/1 of PURIFLOC A22 was 100% fataV
Sodium polystyrene sulfonate (PURIFLOC A21) also is quite
viscous in solution and apparently has the same effect on
minnows as polyacrylamide.
94
-------
TABLE XXVIII
FISH TOXICITIES OF PURIFLOC FLOCCULANTS*
mg/1 in Solution
Polymer Maximum Safe Partial Kill 100% Fatal
PURIFLOC N17 1000
PURIFLOC Nil 1000
PURIFLOC N12 1000
PURIFLOC A21 1000
PURIFLOC A22 1000 -- 2500
PURIFLOC A23 500 -- 1000
PURIFLOC C31 2.0 3.0 4.0
PURIFLOC C32 2.0 3.0 5.0
ET-721 2.0 3.0 10.0
*Acute toxicity to fathead minnows, Pimephales promelas,
in Lake Huron water at 50°F after 72 hours.
The amounts of flocculant that can be tolerated in a receiving
stream can be increased if there is sufficient turbidity
in the form of suspended solids. Excess flocculant can
be adsorbed onto the solids and removed from the water.
Cationic flocculants of the type discussed in this report
are readily adsorbed on substrates such as various types
of clay, river silt, charcoal, CaCOo, and other materials
which possess anionic surface charges. The degree of adsorption
is dependent on the number of sites available on the substrate,
the molecular structure of the polymer, the pH and alkalinity
of the system, and other parameters.
The maximum safe limit of PURIFLOC C31 could be increased
from 2 mg/1 to 15 and 25 mg/1 by addition of 50 and 100
mg/1 of bentonite clay, respectively. The maximum safe
limit of 2 mg/1 in the case of PURIFLOC C32 was increased
to 5-10 mg/1 and 10 mg/1 by addition of 50 and 100 mg/1
of bentonite clay, respectively. The acrylamide-based
flocculants are also readily adsorbed on suspended solids
providing conditions conducive to flocculation exist^.
A recent study by The Dow Chemical Company has shown that
cationic polymers are effectively removed from river water
and raw sewage by adsorption on suspended solids. It was
found that maximum adsorption of flocculants on suspended
solids occurred during flocculation. Fish t o x i cities of
the cationic polymers must be based on the amount of polymer
remaining in solution after flocculation and not on the
total concentration initially applied.
95
-------
Human and Animal Toxicity
All of the PURIFLOC flocculants listed above are low in
oral toxicity and low in degree of health hazard from handling.
None of the polymeric flocculants noted are absorbed through
the skin in toxic amounts. Prolonged skin contact results
in no significant irritation although repeated or prolonged
gross contact may result in mild contact dermatitis. The
effects of eye contact may range from essentially no effect
to mild transitory irritation, depending on the flocculant.
In general, human experience has shown that these products
are safe to handle but reasonable care and personal cleanliness
should be practiced to avoid possible skin and eye contact.
Table XXIX contains LDf-Q values for laboratory rats exposed
to the PURIFLOC series. These data were obtained from reports
issued by the Biochemical Research Laboratory of The Dow
Chemical Company. Human LD50 values can be estimated by
multiplying the LD5Q value for rats by the human body weight
in kilograms.
TABLE XXIX
LD5Q VALUES FOR LABORATORY-CONFINED RATS
F1occulant LD5Q (grams/kilogram)
PURIFLOC N17 4.0 +
PURIFLOC Nil 4.0 +
PURIFLOC N12 4.0 +
PURIFLOC A21 2.0
PURIFLOC A22 4.0 +
PURIFLOC A23 4.0 +
PURIFLOC C31 1.0 - 2.0
PURIFLOC C32 1.1
ET-721 4.0
Specifications for the manufacture of polyacrylamide include
a maximum acrylamide monomer content of <0.05 percent on
a product basis. This is a safe limit for potable water
uses and is within the Public Health Service recommendations.
Sodium styrene sulfonate monomer is not toxic, but forms
aldehydes after prolonged storage which are slightly irritating
to skin and nose membranes. The presence of monomer in
polyamine polymers is insignificant due to the reactivity
of these compounds in polymerization. Essentially no detectable
monomer concentration remains.
96
-------
In the preparation of ET-721, a nonionic acrylamide is made
cationic via aminomethylation (Mannich reaction). The amino-
methylating process depends upon the reaction of dimethylamine
(DMA) with acrylamide via an intermediate, which gives the
polymer its cationic character. DMA as a gas, or dissolved
in water, is toxic. In the preparation of ET-721, the DMA
reacts completely to form part of the polymer molecule thereby
eliminating its availability as the parent compound. The
finished polymers are as non-toxic as the polyamines PURIFLOC
C31 and C32.
97
-------
SECTION 7
LONG-TUBE SEDIMENTATION STUDIES
SEDIMENTATION
"Sedimentation is a waste treatment process whereby suspended
and coagulated particles of a density greater than that of
the liquid medium are removed."' Under quiescent conditions,
the factors influencing sedimentation of discrete suspended
particles are their size, shape, and density. The viscosity
and density of the suspending medium may also require consider-
ation. The settling velocity of a discrete particle can
be defined on the bases of these characteristics and is
assumed constant during the sedimentation process.
Complications may arise in the case of sedimentation of
suspended particles from combined sewer overflows. The
particles are essentially non-spherical, polydisperse, and
possess variable degrees of surface charge. They also may
be of a flocculent nature, i.e., increasing in size and
density, and changing in shape during sedimentation. The
settling velocity of such indiscrete* particles is not constant.
The settling path of a discrete particle is linear in an
ideal tank. The corresponding path of an indiscrete*(f1occulent)
particle is curvilinear. Discrete solids are denoted as
Class I solids; indiscrete solids are denoted as Class II
solids. The settling paths of these two classes of particles
are portrayed in Figure 28.
Figure 28
Settling Paths of Discrete and Indiscrete Particles
a)
a
Discrete
Particle
Indi screte
(Flocculent)
Particle
Di stance (Time)
*Not to be confused with indiscreet, i.e., lacking prudence,
which could also apply to certain exasperating experimental
situations.
99
-------
The extent of discreteness must first be determined before
the efficiency of removal can be predicted for a given system.
Removal of discrete particles is dependent only on the settling
velocity (or overflow rate). Removal of indiscrete particles
is a function of both settling velocity and detention time.
Laboratory experiments are generally advisable to evaluate
sedimentation for an unknown suspension. These usually
consist of measuring the concentration of suspended solids
in a sedimenting fluid at various depth increments and time
intervals in a vertical cylinder. The initial profile
of solids concentration must be uniform. The concentration
of solids in the upper regions of the suspension thereafter
decreases as sedimentation proceeds. Limiting (or maximum)
sedimentation contours of concentration or of removal as
functions of depth and time are represented in Figure 29.
Figure 29
Sedimentation Contour Map: Depth from Bottom vs. Time
o
CO
a
a;
a
Concentration
Removal
Time
Such contour plots are useful in determining the extent
of removal of solids from an ideal settling tank at a given
depth-time coordinate.
A measure of the discreteness of a given suspension can be
determined graphically by assuming an exponential variation
of settling velocity and detention time^. The previously
described contours form a family of parallel lines in a
logarithmic plot of depth vs. time as shown in Figure 30.
100
-------
Figure 30
Sedimentation Contour Map:
Log Depth from Surface vs. Log Time
-------
The depth/time quotient corresponds dimensionally to a
settling or overflow rate (L/T)*. If indeed the system
under consideration is Class I (discrete) then all data
should plot along a single profile in such a graph. The
spread of the data
from discreteness.
is, therefore, a measure of the deviation
The final step in
is to compute the
the analysis of the sedimentation process
area under the reduced concentration profile
at selected overflow rates. The fraction of solids not
captured at a given overflow is equal to the corresponding
cumulative area. The fractional loss profile is illustrated
in Figure 32.
Figure 32
Fractional Loss Profile:
Fraction of Solids not Captured vs. Overflow Rate
•— XJ
O QJ
O Q.
to
c o
o
•I- +J
4-> O
Settling (Overflow) Rate
The average detention times at selected stations across
an effluent weir of a settling basin can be determined from
dye dispersion studies3'?. The local overflow rate at each
station (basin depth-average detention time) is then used
to determine the predicted local value of the fraction of
solids not captured and ultimately an average fraction for
the entire basin.
Five systems of flocculants as established by the flocculant
studies were evaluated using a long-tube sedimentation device
A description of this device, its operation, and performance
is contained within this chapter. Removals of optical solids
This quantity can be reported as a velocity, L/T, or surface
loading (L3/TL^). It is not a weir overflow rate.
*Thi
102
-------
were predicted for each system for selected overflow rates
and combined with results from the hydraulic model studies
for design purposes.
DEVELOPMENT OF A LONG-TUBE SEDIMENTATION DEVICE
A usable long-tube sedimentation device (LTSD) is a requisite
for studying sedimentation processes4'5. The dimensions
of such a device should be sufficient to approximate full-
scale conditions and to minimize wall effects.
The long-tube sedimentation device used in these studies
was made of a clear plastic cylinder having a wall thickness
of 1/4 in., an inside diameter of 5 in., and a length of
12 ft. A schematic of the LTSD is presented in Figure 33.
The tube contained 50 liters of liquid when completely filled.
It was fitted with stainless steel paddles with one-inch
square blades placed vertically at 4 in. intervals along
a central drive shaft. Sampling ports were located at one-
foot intervals and were covered with large syringe caps.
The holding tank has a capacity of 35 gallons, sufficient
for two 50 liter charges to the LTSD plus associated jar
tests. Sample charges were transferred to the column by
a centrifugal pump.
Samples of 100 ml volume were collected by hypodermic syringe
at preselected depths and times. A sample grid of six depths
(1, 2, 4, 6, 8, and 10 feet from bottom) and eight times
(0, 4, 9, 16, 25, 36, 49, and 64 minutes) was established
for the majority of the systems studied. A sludge sample
was collected at the end of 64 minutes at a depth of 0 feet.
, p c
Optical solids were rapidly determined1' ' using a photometric
probe that could be immersed directly into each sample. Optical
solids were defined as absorbance x 1000. Gravimetric solids
were determined after filtration through Gooch crucibles and
subsequent drying to constant weight. A typical correlation
is shown in Figure 34.
PROCESSING OF SOLIDS CONCENTRATION DATA
The raw data of optical and gravimetric solids concentrations
as functions of depth and time must be processed prior to
graphical analysis. An IBM 1130 computer was used for both
processing and plotting**.
Withdrawing of each sample reduces the total volume of the
system. The depth coordinate is corrected for this depletion
A finite time is required to withdraw each sample. The
time coordinate is corrected accordingly.
103
-------
Figure 33
LONG-TUBE SEDIMENTATION DEVICE
V
\
!\
i \\
i
1
i i
^
s~
( (
( v.
}
HOLDING
TANK
4
£
r
^
\
)
.x
N
\
j
(
TRANSFER
PUMP
^
^
^
* i
1 1 —
10-
~ 9 —
^ o
t 8-
0
CO
2 7 —
o
Li.
i- 6-
ii
LU
LU
~ 5 —
CO
cr. .
* 04
Q_
LU
1 3~~
co 2_
1 —
0-
^ MOTOR DRIVE
' (0-IOOrpm)
1
1
I
|
•
i
i
i
i
J/
^
n
D+O
a-f-a/
n 1 n
LrfU
n
0+0
D-4-D „
IT1/
tir-y
rrr-r^li
' ! 1
i
i
i
i
i
i
i
I
/'
'
^•AGITATOR
(4" SPACING OF
BLADES)
/-SEDIMENTATION
COLUMN
(5"l.D.x 12')
SAMPLING
DRAIN
DEPTHS: (0), 1,2,4,6,8, IOFT.
TIMES: 0,4,9,16,25,36,49,64,MIN.
104
-------
Figure 34
MILK RIVER PROJECT - LONG TUBE SEDIMENTATION STUDY
CORRELATION OF OPTICAL AND GRAVIMETRIC SOLIDS
SYSTEM 5A
STORM FLOW 11-18-63
AVG INIT OPTICAL SOLIDS - 120 CA X 10003
AVG INIT GRAVIMETRIC SOLIDS = 206 MG PER L
20 MG PER L PURIFLOC C31 * 20 MG PER L BAROID HECTORITE CLAY
250,.
o
in
350
400
GRAVIMETRIC SOLIDS, MG PER L
-------
Removals are calculated from the concentration data and
reported as fractions. The coordinates of the reduced profile
are also machine calculated. The resulting corrected data
and computed parameters are then punched on cards for subsequent
graphical analysis.
A typical grid of values of optical solids as functions
of uncorrected depth from bottom and uncorrected time of
sampling is presented in Table XXX.* The initial (time
zero) depth-averaged concentrations of optical and gravimetric
solids, respectively, were 119.5 and 206.2. The initial
profile at time zero is uniform. Concentrations decrease
with time as the solids settle to the bottom of the column.
The data corrected for volume depletion and sampling delay
are presented in Table XXXI. Other quantities have been
calculated for use in the graphical presentations. Removals
of optical and of gravimetric solids are calculated as fractions
of the initial depth-averaged concentrations. The settling
rate is calculated by dividing the depth from the surface
by the time of sampling.
VARIABLES
Dry weather flow was the flocculate used during the "shake-
down" of the LTSD and associated sampling procedures. Samples
of combined sewer overflow taken during the first pumping
were used to evaluate all subsequent flocculant systems.
The flocculants used were: PURIFLOC C31 , and/or PURIFLOC
A23. The coagulant aids used were: ferric chloride and/or
BAROID HECTORITE clay. Optimum concentrations of all chemicals
to be used in subsequent LTSD studies were determined on
the basis of jar tests.
Flocculation was initiated in the early studies by injecting
incremental portions of solutions of the selected flocculants
by syringe through the sampling ports of the LTSD. This
procedure proved unsatisfactory due to poor dispersion.
All subsequent f1occulations were conducted by flash mixing
a single total dose of the flocculant in a full-tube charge
of stormwater (50 liters) and then pumping the mixture to
the LTSD.
Sedimentation under quiescent conditions resulted in deposition
of flocculent solids on the agitator blades. This deposition
produced skewed profiles of solids concentrations. A condition
of dynamic sedimentation was established in all subsequent
*Nomenclature provided at end of report.
106
-------
TABLE XXX
11-18-69 20 MG PER L PURIFLOC C31 •*• 20 MG PER L BAROID HECTORITE CLAY
RAW DATA - SAMPLING TIMES AND DEPTHS ARE UNCORRECTED
DEPTH* FT
TIMEt MIN
1.
2«
4.
6.
8.
10.
OPTICAL SOLIDS (ABSORBANCE X 1000)
0.
123.
124.
122.
126.
118.
104.
1
1
4.
24»
23.
122.
1
1
18.
11.
94.
9.
127.
128.
116.
107.
72.
38.
16.
126.
114.
96.
63.
28.
24.
25.
90.
72.
43.
32.
18.
26.
36.
18.
20.
15.
13.
12.
8.
49.
11.
8.
7.
6.
10.
9.
64.
8.
5.
4.
2.
2.
3.
DEPTH* FT
TIME* MIN
1.
2.
4.
6.
8.
10.
INITIAL AND CALCULATED CONSTANTS
GRAVIMETRIC SOLIDS (MG/L)
0.
229.
201.
200.
219.
202.
186.
4.
250.
236.
242.
236.
216.
186.
9.
243.
238.
232.
209,
146.
89.
16.
251.
217,
161.
137.
60.
58.
25.
163.
135.
90.
63.
46.
51.
36.
45.
40.
47.
34.
37.
34.
49.
24.
23.
16.
18.
13.
15.
64.
24,
12.
9.
8.
7.
10.
P = 6
0 = 8
DO = 142.75 (IN)
DF = 127.00 (IN)
TSP = 3.0 IMIN)
OSO = 119.500
GSO = 206.166
(A X 1000)
(MG/L)
TDEPL = 0.02440 (FT/SPL)
ADEPl. = 0.02734 (FT/SPL)
107
-------
TABLE XXXI
11-18-69 20 MG PER L PUR1FLOC C31 + 20 MG PER L BAROID HECTORITE CLAY
CORRECTED DATA AND CALCULATED PARAMETERS
N M
1 1
1 2
1 3
1 4
1 5
1 6
2 1
2 2
2 3
2 4
2 5
2 6
3 1
3 2
3 3
3 4
3 5
3 6
4 1
4 2
4 3
4 4
4 ?
4 6
5 1
5 2
5 3
5 4
5 5
5 6
6 1
6 2
6 3
6 4
6 5
6 6
7 1
7 2
7 3
7 4
7 5
7 6
8 1
8 2
8 3
8 4
8 5
8 6
DIM)
1.
2.
4.
6.
8.
10.
1.
2.
4.
6.
8.
10.
1.
2.
4.
6.
3.
10.
1.
2*
4.
6.
8.
12.
1.
2.
4.
6.
8.
10.
1.
2.
4.
6.
8.
10.
1.
2.
4.
6.
8.
10.
1.
2.
4.
6.
8.
10.
DBOT
1.000
2.000
4.000
6.0CO
8.000
10.000
1.000
1.972
3.945
5.917
7.890
9.863
1.000
1.945
3.890
5.835
7.781
9.726
1.000
1.917
3.835
5.753
7.671
S.S89
1.000
1*890
3.781
5.671
7.562
9.453
1.000
1.863
3.726
5.589
7.453
9.316
1.000
1.835
3.671
5.507
7.343
9.179
1.000
1.808
3.617
5.425
7.234
9.042
DSUR
10.895
9.895
7.895
5.895
3.895
1.895
10.731
9.759
7.786
5.813
3.841
1.868
10.567
9.622
7.677
5.731
3.786
1.841
10.403
9.485
7.567
5.649
3.731
1.S13
10.239
9.346
7.458
5.567
3.677
1.786
10.075
9.212
7.348
5.435
3.622
1.759
9.911
9.075
7.239
5.403
3.567
1.731
9.747
8.938
7.130
5.321
3.513
1.704
LDSUR
1.037
0.995
0.897
0.770
0.590
0.277
1.030
0.989
0.891
0.764
0.584
0.271
1.023
0.983
0.885
0.753
0.578
0.265
1.017
0.976
0.878
0.751
0.571
0.258
1.010
0.970
0.872
0.745
0.565
0.251
1.003
0.964
0.866
0.739
0.558
0.245
0.995
0.957
0.859
0.732
0.552
0.238
0.988
0.951
0.852
0.725
0.545
0.231
T(N)
0.
0.
0.
0.
0.
0.
4.
4.
4.
4.
4.
4.
9.
9.
9.
9.
9.
9.
16.
16.
16.
16.
16.
16.
25.
25.
25.
25.
25.
25.
36.
36.
36.
36.
36.
36.
49.
49.
49.
49.
49.
49.
64.
64.
64.
64.
64.
64.
TSPL
0.000
O.OOC
0.000
0.000
0.000
0.000
0.875
0.791
0.708
0.625
0.541
0.458
1.708
1.625
1.541
1.458
1.375
1.291
2.875
2.791
2.706
2.625
2.541
2.458
4.375
4.291
4.206
4il25
4.041
3.958
6.206
6.125
6.041
5.958
5.875
5.791
8.375
8.291
8.208
8.125
8.041
7.956
10.375
10.791
10.708
10.625
1C. 541
10.458
LTSPL
-0.057
-0.101
-0.149
-0.204
-0.266
-0.238
0.232
0.210
0.187
0.163
0.138
0.111
0.458
0.445
0.432
0.419
0.405
0.390
0.640
0.632
0.623
0.615
0.606
0.597
0.792
0.786
0.781
0.774
0.768
C.762
0.922
0.916
0.914
0.909
0.905
0.900
1.036
1.032
1.029
1.026
1.022
1.019
OS
123.
124.
122.
126.
118.
104.
124.
123.
122.
118.
111.
94.
127.
128.
116.
107.
72.
38.
126.
114.
96.
63.
28.
24.
90.
72.
43.
32.
18.
26.
18.
20.
15.
13.
12.
8.
11.
8.
7.
6.
10.
9.
8.
5.
4.
2.
2.
3.
GS
229.
201.
200.
219.
202.
186.
250.
236.
242.
236.
216.
186.
243.
238.
232.
209.
146.
89.
251.
217.
161.
137.
60.
58.
163.
135.
90.
63.
46.
51.
45.
40.
47.
34.
37.
34.
24.
23.
16.
18.
13.
15.
24.
12.
9.
8.
7.
10.
ROS
-0.029
-0.037
-0.020
-0.054
0.012
0.129
-0.037
-0.029
-0.020
0.012
0.071
0.213
-0.062
-0.071
0.029
0.104
0.397
0.682
-0.054
0.046
0.196
0.472
0.765
0.799
0.246
0.397
0.640
0.732
0.849
0.782
0.849
0.832
0.874
0.891
0.899
0.933
0.907
0.933
0.941
0.949
0.916
0.924
0.933
0.958
0.966
0.983
0.983
0.974
RGS
-0.110
0.025
0.029
-0.062
0.020
0.097
-0.212
-0.144
-0.173
-0.144
-0.047
0.097
-0.178
-0.154
-0.125
-0.013
0.291
0.568
-0.217
-0.052
0.219
0.335
0.70P
0.718
0.209
0.345
0.563
0.694
0.776
0.752
0.781
0.805
0.772
0.335
0.820
O.S35
0.883
0.88B
0.922
0.912
0.936
0.927
0.883
0.941
0.956
0.961
0.966
0.951
SETR
oo
OD
00
CD
09
CO
2.044
2.054
1.832
1.550
1.181
0.679
1.03C
0.936
0.829
0.655
C.458
0.237
0.603
0.5S6
0.465
0.358
C.2^4
0.122
0.390
0.363
0.295
0.224
0.151
0.075
0.270
0.250
0.202
0.153
0.102
0.050
0.197
0.182
0.146
0.11C
0.073
0.036
0.149
0.138
0.110
0.083
0.055
0.027
108
-------
studies. This condition consisted of operating the agitator
at 5 rpm during sedimentation to prevent deposition of solids.
This dynamic condition is also an approximation to that
existing for gravity sedimentation during conventional sewage
treatment.
FLOCCULANT SYSTEMS
Five flocculant systems were evaluated using samples of
combined sewer overflow. These systems are summarized in
Table XXXII. Sedimentations of controls not treated with
flocculants are also reported for comparison. Initial and
final values of optical and gravimetric solids are averaged
for samples at all depths except those containing the sludge
layer.
System 1
The initial "shake-down" of the LTSD and associated procedures
was conducted on the dry-weather flow of 6-5-69. The ternary
system of BAROID HECTORITE clay + PURIFLOC C31 + PURIFLOC A23
in the ratios 20:20:0.75 mg/1 was very effective. Dispersion
of the chemicals, however, was not uniform when applied by
series injection to the long-tube. All mechanical systems
and physical analyses were satisfactory. The same ternary
system described above was evaluated on combined sewer overflow
from Event 27 on 6-5-69. The same concentrations of chemicals
were used. The flocculation and subsequent sedimentation
resulted in very acceptable solids removal and an exceptionally
clear supernate.
System 2
A system of 20 mg/1 PURIFLOC C31 was evaluated on stormwater
for the second pumping of Event 29. Flocculation in the
initial holding tank and subsequent transfer to the long-
tube sedimentation device was much more effective than direct
introduction of the flocculant to the tube. Tank flocculation
was used in all subsequent studies.
System 3
A dual system of PURIFLOC C31 and PURIFLOC A23 was evaluated on
the storm flow of 8-1-69. Optimum doses of the two chemicals
as determined by jar tests were: 30 mg/1 PURIFLOC C31 and
1 mg/1 PURIFLOC A23. In addition to the control, sedimentation
was evaluated under dynamic conditions of 5 rpm agitation
and quiescent sedimentation at 0 rpm. Solids deposition
on the paddles at 0 rpm resulted in local anomalies in the
solids concentration contours. Dynamic sedimentation at
the 5 rpm condition was established in all subsequent evaluations
109
-------
TABLE XXXII
COMPARISON OF REMOVAL EFFICIENCIES OF FIVE FLOCCULANT SYSTEMS BASED UPON INITIAL
AND FINAL TIME AVERAGES OVER ALL DEPTHS FOR OPTICAL AND GRAVIMETRIC SOLIDS
Optical Solids^
Gravimetric Solidsv
stem
1
f7[\
1 A^
IB®
2
2A
2B
3
3A
3B
4
4A
5
5A®
Flocculants
Date (mg/1)
6-5-69 None
20 Clay + 20 C31 +
.75 A23
II
6-20-69 None
20 C31
ii ii
8-1-69 None
30 C31 + 1 A23
II II
8-16-69 None
30 Fe + .5 A23
11-18-69 Control
20 C31 + 20 Clay
Ini ti al
Ao
81.4"
88.8
69.6"
96.6"
84.8
55.7
110. F
75.3"
120.5"
123.8
126.1
149.6"
119.5"
Final Removal
A /A — A \
64 ( o 64)
62.0"
_£5\
5. 8^
4.4
59. f
48. F
15. 1
77.6"
2.7
17. F
48.6"
16.0
87.3
4.0"
^ Ao /
.238
.93F
.937
.388
.428
.724
.30T
.964
.85F
.612"
.873
.414"
.967
Initial
Co
79.4
90. F
67.4
80.7
87.2
84.3"
121 .2
187.7
192. F
212.8
299.0
182.7
206.2"
Final Removal
64 ( o 64^ Comments
\
56.8
-^
5.8^
5.3
59.8
23.0"
5.1
37.6"
8.1
18.0
43.8
17.0"
83. F
11.7
> Co /
.28F
.93F
.92T Dry weather flow
.299" Flocculation 1n
.73F Column
.937 Tank
.69F Sedimentation at
.95F 0 rpm
.906 5 rpm
.794
.943
.543"
.943
^Absorbance x 10° ^mg/1 ^Final values at 25 minutes
initial solids include clay flocculant
-------
System 4
An inorganic metal cation and an anionic polyelectrolyte
were evaluated on the storm flow of 8-16-69. As determined
by jar tests, optimum concentrations were: 30 mg/1 ferric
iron (as Fed.,) and 0.5 mg/1 PURIFLOC A23. This system
did not produce the most rapidly settling floe but was signif-
icantly better than the control. The concentration profiles
were well spread over the time intervals of sampling. This
was contrasted to the rapid sedimentation experienced in
the ternary system described previously in which the profiles
were closely spaced at the early time intervals.
System 5
A dualsystem of a clay and a cationic polyelectrolyte was
evaluated on the storm flow of 11-18-69. Optimum concen-
trations as determined by jar tests were: 20 mg/1 PURIFLOC
C31 + 20 mg/1 BAROID HECTORITE clay. The order of addition
used in the long-tube test was cationic polyelectrolyte
followed by clay. The reverse order was also effective.
The linear correlation of optical and gravimetric solids
for the sample used in the evaluation of System 5A is presented
in Figure 34. Similar correlations were found for the other
samples. The contour maps and reduced concentration profiles
described below were developed using optical solids. Gravi-
metric solids can also be used as a criterion but are more
subject to experimental variation.
The contours of optical solids concentration as functions
of depth from the bottom and sampling times for System 5A
are presented in Figure 35. Concentration decreases with
increased settling time and increased depth from the bottom.
The contours are generally linear except for those of low
concentration where slight variations are magnified.
The optical solids concentrations for the same System 5A
are plotted as functions of the logarithms of depth from
the surface and sampling time in Figure 36. This system
follows Class I sedimentation since the majority of the
contours have unit slopes.
The reduced concentration profile of System 5A is presented
in Figure 37. The overflow rate is calculated by dividing
the depth from the surface by the time of sampling. The
establishment of a single curve is another technique to
verify Class I sedimentation.
Ill
-------
F i gure 35
13..
11..
10..
LL
ft
d
h
H
O
CD
o
C£
IL
I
H
CL
UJ
Q
a..
7..
G..
8..
3 . „
H..
1. .
MILK RIVER PROJECT - LONG TUBE SEDIMENTATION STUDY
SEDIMENTATION CONTOUR MAP
+ SYSTEM SA
STORM FLO'/N 11-13-69
AVG INIT OPTICAL SOLIDS = 120 CA X 10003
AVG INIT GRAVIMETRIC SOLIDS = 206 MO PER L
u 20 MG PER L PURIFLOC C31 + 20 MG PER L BAROID HECTORITE CLAY
134
133 124
a
+
OPTICAL SOLIDS. ABSORBANCE X 1000
i t i i i. i i i i i
+
i i i i i
& S 7 B O V)
SETTLING TIME. 0-1 HR
u ia
14 IB
-------
1-25..
Figure 36
MILK RIVER PROJECT - LONG TUBE SEDIMENTATION STUDY
SEDIMENTATION CONTOUR MAP
SYSTEM SA
STORM FLOW 11-18-69
AVG INIT GRAVIMETRIC SOLIDS =C206 MG°PER L
20 MG PER L PURIFLOC C31 + 20 MG PER L BAROID HECTORITE CLAY
U.
!JJ
O
<
LL
O
QL
IL
I
Q.
LU
D
O
O
.J
i-oo..
•75..
•SO..
•as..
1" s ,
' '
OPTICAL SOLIDS. AQSORBANCE X 1000
0-00..
—I—
1.00
-•50
-•3S
0-CO -E3 'SO -75
LOG SETTLING TIME, 0-1 HR
i-25
-------
Figure 37
MILK RIVER PROJECT - LONG TUBE SEDIMENTATION STUDY
REDUCED CONCENTRATION PROFILE
SYSTEM 5A
STORM FLOW 11-18-63
AVG INIT OPTICAL SOLIDS = 120CA X 1000D
AVC INIT GRAVIMETRIC SOLIDS = 206 MG PER L
20 MG PER L PURIFLOC C31 + 20 MG PER L BAROID HECTORITE CLAY
LL
0
.75
1-25
1-5
1.75
OVERFLOW RATE. FT PER MIN
-------
The fractional loss of optical solids as a function of overflow
rate for System 5 is shown in Figure 38. Fractional loss
values correspond to the cumulative area under the reduced
concentration profile of Figure 37. At an overflow rate
of 0.67 fpm one-half of the initial solids are retained and
one-half are lost.
All Systems
Values of final/initial optical solids for all five flocculant
systems are given in Table XXXIII for selected overflow rates
as estimated from the reduced concentration profiles. The
clay flocculant amounted to 15-25 percent of the initial
solids present for the three systems marked (*). In the
strictest sense removal should be based only on the initial
sewage solids and be independent of any additional solid
flocculants. It was not feasible in the current study, however,
to attempt any fractionation of sewage and flocculant solids.
The tabulated values for these systems therefore also include
removal of the solid clay flocculant.
TABLE XXXIII
VALUES OF FINAL/INITIAL OPTICAL SOLIDS
EXTRAPOLATED FROM REDUCED CONCENTRATION
PROFILES AT SELECTED OVERFLOW RATES FOR
FIVE FLOCCULANT SYSTEMS
Final/Initial Optical Solids (%)
at Indicated Overflow Rates (fpm)
System
1A*
IB*
2A
2B
3A
3B
4A
5A*
.25
5
9
62
27
5
20
47
33
.5
16
35
68
33
13
32
86
80
.75
30
75
73
41
27
47
95
95
1 .0
49
91
80
54
52
67
98
96
*Initial solids include clay flocculant.
115
-------
Figure 38
MILK RIVER PROJECT - LONG TUBE SEDIMENTATION STUDY
FRACTIONAL LOSS CURVE
SYSTEM 5A
STORM FLOH 11-18-69
AVG INIT OPTICAL SOLIDS = 120 CA X 1000}
AVG INIT GRAVIMETRIC SOLIDS = 206 MG PER L
20 MG PER L PURIFLOC C31 + 20 MG PER L BAROID HECTORITE CLAY
I I 1
OVERFLOW RATE, FT PER MIN
-------
Fractional losses of optical solids (that is, not captured
by sedimentation) at selected overflow rates for all five
flocculant systems are given in Table XXXIV. The fraction
of solids lost increases as expected as the overflow rate
increases for all systems. Although the experimental results
are limited, certain observations are possible.
TABLE XXXIV
FRACTIONAL LOSSES OF OPTICAL SOLIDS
AT SELECTED OVERFLOW RATES FOR FIVE
FLOCCULANT SYSTEMS
Fractional Losses of Optical Solids (%)
at lr\d i c ated Overflow Rates (fpm)
System
1A
IB
2A
2B
3A
3B
4A
5A
The flocculant systems giving the best overall performance
based on the fractional loss criterion are systems 1A and
3A. Fractional losses for these systems were the lowest
of the five systems studied. The flocculated materials
formed by Systems 3B, 4A, and 5A were more easily lost at
the higher overflow rates. The same flocculants were used
in Systems 1A and IB. The medium flocculated in System
IB, however, was dry-weather flow and may reflect fundamental
differences in solids composition. The results of Systems
2A and 2B are questionable because of non-zero intercepts
at zero overflow rate for their reduced concentration profiles.
In the region of low overflow rates (0-.25 fpm) small variations
in measurements of optical and particularly gravimetric
solids are magnified. The corresponding samples were collected
at extended times when the majority of the solids had already
settled.
.25
3
4
59
26
4
17
23
11
.5
7
12
62
28
6
21
47
36
.75
12
27
65
31
11
27
62
54
1 .0
19
41
68
35
18
34
71
65
117
-------
SUMMARY
The objective of this phase of the Milk River Contract was
to obtain sedimentation rate data for combined sewer overflows
treated with specified chemical systems. The sedimentation
rate was then to be correlated with detention time data
derived from the hydraulic model studies. Removal of suspended
solids was then to be predicted for specified chemical and
hydraulic conditions.
The initial requirement for these studies was the development
of a workable Long-Tube Sedimentation Device (LTSD) for
evaluation of the sedimentation process. Such a device
of sufficient dimensions to approximate full-scale conditions
and minimize wall effects and sample volume was constructed.
The device consisted of a long tube of 1/4 in. clear plastic,
5" ID x 12', having a capacity of 50 liters. Paddles were
spaced 4 in. apart along a central shaft.
Flocculation-sedimentation studies were conducted on samples
of combined sewer overflow to evaluate the engineering aspects
of flocculant dispersion, mixing, and sample collection
during sedimentation. Flocculation was accomplished most
uniformly in a holding tank before transfer to the long
tube. Dynamic sedimentation at 5 rpm minimized deposition
of sedimenting solids on the paddles. Such deposition occurred
at 0 rpm and results were skewed.
The feasibility of measuring concentrations of solids by
an optical technique was evaluated as a supplement to the
conventional gravimetric technique. The correlation between
optical and gravimetric solids was linear in the ranges
studied. The optical technique was found to be much faster
and more reliable than the gravimetric technique.
A program of specified sampling times, depths, and volumes,
and procedures to implement such a program were developed.
A grid of six sampling depths (1, 2, 4, 6, 8, and 10 ft.)
and eight sampling times (0, 4, 9, 16, 25, 36, 49, and 64
min.) was adequate for definition of a given system. The
duration of manual sampling by syringe was the limiting
factor.
Long-tube sedimentation studies were conducted on samples
of both flocculated and unf1occulated combined sewer overflows.
Sufficient volumes of overflows were collected during repre-
sentative storm periods. Preliminary jar tests were conducted
to determine optimum chemical levels for each selected system.
118
-------
The sedimentation behaviors of unf1occulated combined sewer
overflows were evaluated as controls for comparison with
chemically-treated overflows. The sedimentation behaviors
of combined sewer overflows were evaluated using the most
effective flocculant systems as determined from the Flocculant
Studies.
The sedimentation data were then correlated with hydraulic
data. This involved a data processing program and a graphical
presentation program. The regime of sedimentation (Class
I or II) was determined by developing contour maps of suspended
solids removal as a function of sedimentation time and depth.
Discrete sedimentation (Class I) of fully-flocculated solids
predominated. Since Class I sedimentation predominated,
reduced concentration profiles were developed. Fractional
loss profiles were then developed by determining the cumulative
areas under the reduced concentration profiles.
119
-------
SECTION 8
HYDRAULIC MODEL
INTRODUCTION
It was essential to consider some of the full-scale problems
associated with hydraulic deficiencies and flow distributions
before results of the flocculant jar tests and the long-
tube sedimentation tests could be extrapolated to estimate
full-scale removal efficiencies. It was not possible to
study these variables in the prototype basin within the
constraints of the project because of the infrequency, unpre-
dictability, and variability of storm events. A scale hydraulic
model of the Milk River Pumping Station (MRPS) was constructed,
therefore, to duplicate in model scale any flow condition or
pumping configuration possible in the prototype installation.
The model was designed with geometric similitude to a scale
of 20:1, fabricated in sections, and assembled in the flocculant
feed room at the MRPS. The model hydraulically simulated^
the operation of the prototype at equal Froude numbers (V /gL).
The similitude factors used in the model design are given
in Table XXXV. Several views of the model are shown in
Figures 39 and 40. Pumping and loading characteristics
of the storm pumps were simulated by varying the condition
of influent flow. Fresh water was introduced to the model
from a constant head tank through a header and any combination
of up to seven pump columns. These columns were arranged
in the horseshoe pattern of the storm pumps of the prototype.
Experiments were performed to improve the settling characteristics
by changing detention time, inlet and outlet conditions,
energy dissipation in the approach channel, and elimination
of short-circuiting.
Sedimentation or flocculating tests in the model were impractical
because of anticipated problems in pumping and size control
for scale-up. Modifications of the model were directed
towards establishing uniform flow across the width of the
basin, elimination of short-circuiting and back flow, and
dissipation of potential energy by selected placement of
weirs and baffles.
The prototype was originally designed to capture floating
material and heavy grit. Suspended solids of low density
are not effectively captured under normal operating conditions.
The present configuration of the MRPS does not provide a
regime favorable to sedimentation of lightweight or even
flocculated solids.
121
-------
TABLE XXXV
SIMILITUDE OF PROTOTYPE AND MODEL
Scale = 1:20
Model
Item
Reynolds No.
Velocity (fps)
Length (ft)
Width (ft)
Depth (ft)
Volume (gal )
Time
2
Froude No.
Flow (gpm)
( one smal 1 pump)
(one large pump)
(al 1 seven pumps )
Overflow Rate
Prototype
668,000
0.202
580
150
10
3.8 x 106
1
.266 x 10"5
137,000
(305 cfs)
184,000
(410 cfs)
1,095,000
(2450 cfs)
2,700
At Equal
Froude No.
7,500
0.0456
29.0
7.5
0.5
475
.224
.266 x 10"5
75.4
103
611
604
At Equal
Velocities
33,600
0.202
29.0
7.5
0.5
475
.0500
5.32 x 10
336
459
2,720
2,700
(gpd/sq ft)
1 N - LVp
N
Fr " gL
122
-------
Figure 39
M1lk River Pumping Station - Basin Model Inlet
The constant-head tank, the discharge pipe from the tank and a portion
of the basin including the influent pump column ports and bays, the
distribution plane, terminating at the submerged weir - the white sloping
transverse band on the model, and baffles protecting the discharge ports
in the basin bottom are shown.
Figure 40
Milk River Pumping Station - Basin Model Discharge
The discharge end of the model is shown with the skimming baffle and
effluent weir. There are ten sampling slots, uniformly spaced along
the length of the effluent weir numbered A-0 to J-9, starting at the
left (lower center).
123
-------
ANALYSIS OF EXISTING BASIN
Effective removal of appreciable amounts of lightweight
suspended solids cannot be expected in operating the existing
MRPS as noted above. Sedimentation of lightweight suspended
solids requires relatively gentle, well distributed, steady-
state input of flow. The arrangement of the pumps coupled
with intermittent pumping at high rates combine to prevent
transverse distribution of the influent flow. Short-circuiting
occurs almost immediately and persists throughout the length
of the basin, resulting in fast flowing streams overflowing
the effluent weir at various locations along its length.
Dead spots and areas of backflow have the effect of reducing
the active sedimentation area of the basin. The dead spots
are caused by diagonally-impinging short-circuiting streams
and reversed flows which are caused by the angular horizontal
configuration of the effluent weir and skimming baffle.
The nominal detention time and overflow rates during operation
of one small, one large, or all four small and three large
pumps at the MRPS are shown in Table XXXVI.
TABLE XXXVI
NOMINAL DETENTIONS AND OVERFLOW RATES
IN MILK RIVER RETENTION BASIN
Flow Nominal Detention Overflow Rate
(cfs ) (gpm) (min ) (f ph)
305 137,000 27.8 18.5
410 184,000 20.7 24.8
2450 1,100,000 3.5 148.0
Nominal detention is found by dividing the capacity of the
basin in cubic feet by the input in cubic feet per minute.
The nominal detention time available for sedimentation by
plain settling (no f1occulation) or for flocculated solids
is limited, even if more favorable conditions of flow existed.
Overflow rates (volume input in cubic feet per hour divided
by basin surface area in square feet) of from 18.5 to 148
fph are also significantly high for plain sedimentation
or Class I sedimentation (fully flocculated solids) to occur.
As seen earlier, simultaneous operation of all seven pumps
occurs only very infrequently. Rectangular sedimentation
124
-------
basins are generally designed to a length to width ratio
(L/W) of not less than about 5:1 to reduce the possibility
of short-circuiting. The L/W ratio of the basin at the
MRPS is about 2.5:1.
The active depth of the basin is about nine feet and is
satisfactory for sedimentation. Depth is reduced, however,
to about 3.2 feet for a distance of 75 feet as the influent
flows over the distribution plane prior to entering the
basin proper. The rate of progression down the distribution
plane is 38 fpm (one small pump operating) and is sufficiently
low to permit sedimentation of some grit and possibly some
lightweight solids. Since the rate of progression in transverse
distribution is not uniform the effectiveness of the area
over the distribution plane is unpredictable.
IMPROVEMENTS TO THE MILK RIVER PUMPING STATION
A number of modifications to improve the capability of the
MPRS were suggested based on studies of plant records, current
operating conditions, and dye tests in the model. Continuous
(staged) pumping would considerably reduce average pumping
rates. Placement of transverse weirs would essentially
prevent short-circuiting and reduce the possible adverse
effects of density currents. Both modifications would increase
the efficiency of the basin to handle solids removal for
specified flows and suspended solids loadings either in
plain settling or after chemical treatment.
Combined sewer overflows amount to about 5.5 million cubic
feet (41.1 million gallons), 90 percent of the time. The
storm sewers can contain five million gallons and the wet
well about 0.3 million gallons. There are four small and
three large stormwater pumps at the MRPS rated at 305 and
410 cfs (137,000 and 184,000 gpm), respectively, at pumping
heads of 33 feet. Continuous pumping of combined sewer
overflow at approximately one-third the capacity of one
small pump is feasible and desirable when possible, especially
in sedimentation of chemically treated suspended solids.
Reduced pumping rates (102 cfs, 45,800 gpm) can be accomplished
by replacement of one of the small pumps with one allowing
a variable discharge rate, or by adding a drain for recycling
about two-thirds (203 cfs, 91,100) of the input from one
small pump to the wet well without impairing the ability
to deliver combined sewage at the rated capacity of all
seven pumps.
Satisfactory transverse flow distribution of influent cannot
be accomplished at the MRPS by reducing pumping rates or
by tandem pumping within the existing design. The addition
of baffles, weirs, or other flow distributing devices is essential
to provide rapid transverse distribution of basin influe.nt.
125
-------
Gross short-circuiting and reverse flowing streams could
be visually observed in the prototype and in the unbaffled
model. The short-circuiting, dead spots, and reverse flows
in the unbaffled model when operating one small corner pump
at a reduced flow, equivalent to about 82 cfs (36,800 gpm)
input to the prototype is shown in Figure 45. Much the
same pattern of uncontrolled distribution occurs in the
model (or prototype) at higher rates of flow.
Adverse effects are lessened when operating a more centrally-
located pump. No pump is located on the long axis of the
prototype basin, however, and storm flow is skewed under
any condition of pumping. Operation of two opposing pumps
provides input excessive for good sedimentation under existing
conditions. Uncontrolled flow restricts prediction of solids
removal even though the detention times may compare favorably
with those found when weirs are used.
QUALITATIVE DYE STUDIES
Qualitative tests of the dispersion of Bismark brown dye
were conducted for a range of influent flows on the open
model prior to testing the placement of weirs and baffles.
The need for modifications for obtaining improved detention
efficiencies in the model were emphasized by the results
of these tests. Detention efficiencies could be improved
by reducing short-circuiting and hydraulic heads, and by
preventing back flow caused especially by the configuration
of the skimming baffles.
These objectives were approached by optimizing arrangements
of baffles and/or weirs in the model. Two weirs, each with
15 equally-spaced 90° notches, were fabricated and tested
in the model. The need for more and varied weirs and baffles
was stressed by the results of these preliminary tests.
Several additional weirs and baffles were fabricated from
aluminum stock.
The 90° notches in the original weirs were also satisfactory
in the design of new weirs to be placed on the distribution
plane and in the vicinity of the upstream end of the basin
proper (weirs A, B, E). Notches in A and B weirs were made
somewhat deeper than the original weirs. This reduced the
water level at inactive pump columns when these weirs were
used on the distribution plane. A and B weirs were similar
and made in pairs with staggered notches. The overall heights
of the B weirs were somewhat less than those of the A weirs.
Weir E was similar to A and B but was higher than an A weir
126
-------
for effect in basin placement. Weir G was designed with
fifteen 55° notches for use in the general vicinity of the
middle of the basin. Weir H was designed with twenty-two
40° notches. The notches in G and H were cut deep to avoid
converting kinetic energy into potential energy.
Qualitative and quantitative dye tests were performed using
the improved weirs and baffles in various combinations and
locations. The identity and location of the weirs and baffles
in model tests are shown in a plan outline of the model
basin in Figure 41. Notations of operating conditions and
comments are recorded. Selected data from the sketches
are tabulated in Table XXXVII. Most tests were performed
with only one small corner pump (#7) in operation. Flow
from this location was considered most difficult to distri-
bute transversely. Conversion factors between prototype
and model are tabulated in Table XXXV.
Tests performed with baffles placed on the distribution
plane or in the basin were unsuccessful in preventing short-
circuiting or in substantially consuming the hydraulic head.
The inclusion of a curved weir circumscribing the pump discharge
bay provided some limited flow distribution and consumption
of hydraulic head. The weir used in these tests was evidently
too low to be effective and its use was discontinued in
favor of pairs of weirs placed across the distribution plane.
In most subsequent tests combinations of A and/or B weirs
were used.
In some runs a chain curtain was also located on the distribution
plane or in the basin. It did not appear to contribute to
transverse distribution of the dye front nor to appreciable
consumption of hydraulic head. Water overflowed the inactive
pump ports with either one small pump or one large pump
in operation due to height of the A weir in runs where two
A weirs or one B and one A weir were used. Therefore attention
was directed to evaluation of performance when B weirs
were tested.
Placement of weirs in the basin was important in reducing
velocity head and in developing transverse distribution
of dye. Conditions improved as the number of weirs was
increased. Three weirs in the basin seemed necessary when
operating one small pump to effectively complete transverse
distribution of the dye before it reached the effluent weir.
The weir farthest downstream was most effective in providing
uniform flow to the overflow weir when it had many (22)
40° V-notches. Such a weir (H) was also very effective
in preventing backflow of water caused by the V-shaped con-
figuration of the effluent weir. It was also found desirable
127
-------
Figure 41
SCHEMATIC OF MILK RIVER PUMPING STATION MODEL
These letters
refer to weirs
used and locate
their exact posi-
tion In the model.
The flow retea for the
two pumps used were:
r\imp 4: 103 gpin
Pump 7: 75."I t'-;™.
If two V-notch weirs (W) were
used between sections 1 ap.d 2
then the V-notcli''S were
staggered.
These nunbera refer to the
depth of t.-.e water In Inches
et that location.
This loca:es the bottom
'of the submerged wclr.
These arbitrary numbers
locate the Joints on the
model.
These nu/r.bers refer to
the tltr.e in seconds when
/the dye arrived at that
location.
128
-------
TADLE XXXVII
VISUAL AND FO:WAL DYE TKS73 ' "'. VARIOUS FLOWS,
BAF?Lt: AND WEIR PLACEMENTS. JUNE 18-20, 1969.
FORMAL
DYE TKST
HUN
!
1
2
3
14
5
6
7
1
2
3
1
5
6
7
8
9
10
11
12
13
in
15
16
17
18
19
20
21
22
23
21
25
26
27
28
29
30
31
32
33
PUMP
7
7
7
7
7
7
7
7
7
7
t
1,7
7
7
7
4
7
7
7
7
7
7
7
7
7
t
1
4
1
7
7
7
1
GPM
75
75
75
75
75
75
75
75
75
75
103
178
75
75
75
103
75
75
75
75
75
75
75
75
75
103
103
103
103
75
75
75
103
WEIRS A-ll BAFFLES K-N !HEAD DETENTION
LOCATION
C
C
C
C
A, 22
B,10
K M N
M N
A, 7
A, 7
A, 7
A 11 3/8
B 11 1/8
B 1.1 1/8
B 11 1/8
B 11 1/8
B 11 1/8
B 11 1/8
B 11 1/8
B 11 1/8
B 11 1/8
B 11 1/8
B 10 3/1
B 10 3/1
B 10 3/1
B 10 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
B 11 3/1
1
E 22
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
A 21 1/8
e. 21 1/8
B 20 3/1
E 20 3/1
B 21 3/1
B 21 3/1
E 21 3/1
B 21 3/1
B 21 3/1
B 21 3/1
B 21 3/1
B 21 3/1
E 21 3/1
E 21 3/1
L>
L
F
F
F
F
F
F
F
A
A
A
E
E
E
3
G
0
C
0
G
G
G
G
G
G
G
G
G
G
G
0
H
H
E
E
E
E
E
E
E
E
L .
E
E
E
1
K,J
G
G
G
G
G
G
G
G
G
G
0
0
0
G
b
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
G
H
H
H
H
H
H
H
H
H
H
H
H
H
)!
H
TIMK. SKC
L
160
180
180
150
105
180
170
160
155
230
125
90
205
120
150
120
210
11 0
130
135
165
138
170
165
86
100
120
89
170
ill
170
116
C
125
130
205
195
150
120
180
195
175
200
210
R
115
180
120
150
150
180
160
190
185
210
125
-
205
iro
150
120
190
110
110
135
165
138
150
165
86
100
120
89
200
]11
170
118
129
-------
TADl.C XXXVII (Coiit.)
VISUAL AND PO!'LO'.«,
BAFFLE AND WEIR PLACEMENTS NOV. 11-13, 1969, DEC. 12, 1^6'J.
KOP.V.AL
DYE TEST
11
12
13
14
15
16
17
RUN
34
35
36
36A
37
38
39
40
PUMP
1
1
1
1
7
7
7
7
7
QPM
28.6
20.6
23-5
23.5
1C3
4o.8
20.5
40.0
75.0
WEIHS A-H
LOCATION
1
B i: 3 '4
B 11 3/4
B 11 3/4
B 11 3/4
B 11 3/4
B 11 3/4
-
-
3 21 3/4
3 21 3/4
3 21 3/4
3 21 3/4
3 21 3/4
3 21 3/4
-
_
2
E
E
G
E
E
-
-
3
18 3/4
-
_
4 | ,
G
G
G
G
-
.
H
H
H
K
H
H
-
-
i
1
HEAD DKT:-::TION
1 L
39,-
415
46;
465
13:
255
I8c
_
L 5 J *
'I2C
69:
-
39C
41?
420
57C
135
255
33:
.
Run 36 G 18 3/4 in. downstream from 2.
Footnotes:
L - left C = center R » right
Refer to Figure ill for location of C-welr
Numbers In Location 1 refer to the distance of the weir downstream from i
Refer to Figure 11 for typical location of baffles X-N
F Is located at 2 - the Joint between sections 1 & 2 which Is the bottom
of the submerged weir.
Not all baffles used In comblnatl.cn with all weirs
Location 1 Joint between model sections 1 and 2
2 Joint between model sections 2 and 3
3 Joint between model ne:tlons 3 and 4
4 Joint between mode], sections 4 and 5
5 Joint between model sections 5 and (6-7).
130
-------
to place an E-weir close to the end of the submerged weir
(upstream end of basin) and another weir (G) with 55° V-notches
4 inches from the division between model sections four and
five. This arrangement of weirs in the lower reaches of
the basin plus two B weirs on the distribution plane was
an effective combination.
QUANTITATIVE DYE STUDIES
The degrees of effectiveness of the individual weirs and
their placement were evaluated in more detail in quantitative
tests of the distribution of dye in the model. Contour
maps of dye absorbances (proportional to concentration)
were prepared for several quantitative studies. Dye concen-
trations were measured with a probe or fixed unit colorimeter.
The abscissae of these plots are the cumulative times (min)
after the dye fronts are first observed to overflow the
effluent weir. The detention times are the times the dye
required to reach the effluent weir. The ordinates are the
ten positions across the effluent weir (0 = A, 1 = B, ...,
9 = J). The contoured absorbances (x 10^) represent successive
contours of dye concentration, "frozen" as they crossed the
effluent weir and continue undisturbed through an imaginary
projection of the model.
The effectiveness of the weir design and placement in the
model when one small corner pump (7) was in operation is
illustrated in the contour map (Figure 42) of dye concen-
tration from the highly significant Run 31. Run 33 was
performed to provide data for operation of one large off-
center pump (4). Results obtained for pump 4 were satis-
factory even though optimization of flow control for a corner
pump was expected to be the most difficult. The lack of
good flow distribution was probably due to the location
of the pump. Run 37 was therefore made with 103 gpm flowing
from pump 7.
Comparison of Run 37 with Run 33 shows some improvement
in detention time, 135 seconds versus 118 seconds, when
using the corner pump. Both contour patterns were similar
and showed irregularities for periods between approximately
one and three minutes (Run 37) and four minutes (Run 33).
The 103 gpm input to the model would therefore be excessive.
Support for this contention is found in other contour maps.
Higher weirs on the distribution plane cannot be used to
improve this situation, without increasing the height of the
(inactive) pump columns to avoid their flooding. Increasing
the height of the pump columns is impractical.
131
-------
T r-
Figure 42
MODEL DYE TEST
H i i i h
CO
ro
.2 0 1 •; in 1-1 19 33 35 37 37 1Q 33 33 3S >¥•. ^3 17 IS ,.--. 9 i:T in 10 "i & G 3
£,
B ..
u n"
7 5 5 3 a -'.
la
0 ^a fi, j/J vlGlv30 ^3-: ^37 ^37 J33 xa3 ^lEtogn ^33 .x!9 ^16 ^1-1 ^1-1 ^in i]l; x-n/-I X9 „'•= ,- „'•=
o.
s:
i
I 0? 'J I U i.4 IB 33 117 37 37 2-1 aipffl
X * X /A X IX H VkX XV ^V
1u \ \
... .v 2-1 aipffi ai is 17 is is y iwi 4 4 u G a a /; /. •
k x x v ^~if xx'xx^ixxrjxxxhxxxvx
s \ \
^ fc I C 3 !j 13 j.a .-^3 30 iG 3B P.5 33 33 3-'.29a 1!
" — " * * ' * ICf * «--.•» x x x x x ^x
19 i7 IS 10 iinin H 1 7 f, 5 a 5 -'.
\
in
: 1 C a G il 181 3.3 1G\ 13 OG 33 33 33 ,1-V 33 Cfc. 1.9 17 i.3 13 il 4 S 7 G . !i 'a ri 4
3..
3..
in
in
I- JC .1 _1 ,3 _B ^g,'_y .IS/19.3-: .37 a6.t1-:..V. 33/9_17 >7 17 13 1? _il ^1.10 _ 7 .7 .6 ..G ..'i
'ih'
i i. s '. / f ia ^5 ,ia .TO pa -:s ;n ns SK\ is 1.7 IB 1-1 n • ,.3 *.n in in 7 7 & a 3 '.
SK\
'v
; I
p< i i a s/.in i
/
c..
ilG .?7 33 3J3 ."prja 30 17 IS 1? in iJJ 10 9 S 7 7 5 53 .1
" ' ' " » =5 « x , x -x jtj x » » x » x » < .
CUMULATIVE TIME, MINUTES
—i 1 1 i i 1 i i i 1 r-H
0 1 S 3 •! E G 7 9 3 10
TEST S» RUN 311 PUMP 7, ^LCW 7S GPM. CF.TFNTIDN 1^1 SE'CI. 6XHO/'B3
-------
There may be good reason, however, to postulate that the
increased detention time experienced in Run 37 was due to
using a corner pump for this amount of flow, rather than
pump 4 (which had been used for all other similar inputs).
In which case it may be assumed that the location of the
B weirs further downstream on the distribution plane would
have been more effective when using pump 4 and 103 gpm input.
The contour map of dye concentrations from Run 35 (Figure 43)
is also highly significant in that it represents model performance
for reduced flow of 28 gpm (equivalent to 116 cfs [52,000 gpm]
in the prototype) or approximating one-third the capacity
of one small pump. Good flow control and a total detention
time (elapsed time after adding dye to the center of the
pump port until the dye disappears over the effluent weir)
of about 28 minutes (equivalent to about 125 minutes in
the prototype) is indicated by the contour pattern. Per-
formance in the model for a flow of 40.8 gpm (equivalent
to 166 cfs [74,500 gpm] in the prototype) or approximately
one-half the capacity of one small pump is shown in the
contour map of Run 38 (Figure 44). Good flow control again
is indicated for a total detention of about 16 minutes (equivalent
to 71 minutes in the prototype). The contour map of Run 39
(Figure 45) represents model performance for a flow of 20.5
gpm (equivalent to 83 cfs [37,200 gpm] in the prototype)
with no weirs used in the test. The total detention time,
however, is the same in Run 38 where weirs and double the
input were tested. Runs 40 to 42 were performed at other
flows without weirs in the basin. The total detentions
are only indications of the improvements that can be expected
in the standard and average detentions determined by computation.
EFFECT OF BAFFLES AT REDUCED FLOW RATES
It has already been noted that intermittent operation of the
MRPS leads to inefficient, non-steady state settling conditions
in the retention basin. Variable speed pumping at rates
below 305 cfs would permit relatively steady-state pumping
and a considerable increase in basin detention for up to
about 45 percent of the total flow at the MRPS.
Dye studies in the model at the reduced flow rates have
shown that baffling can improve retention efficiency in
the basin markedly. A summary of the model dye studies
showing the increase in basin detention time achieved for
plug flow under unbaffled and optimally baffled conditions
is contained in Table XXXVIII.
133
-------
Figure 43
MODH. DVE TEST
iiatu)i«
* 'tmiimnu/
iasa>isis)UBi39iz'Ba
' '
/ar \ r \'t
'j30-1 ;i?°/ tf3 4s0 1° •"" J
' L///
1 = " *" *" *™ *
i
U l\o\\
* *" *
7a7DB>
4i95
'
iBaB
'
f°f .*».* ;v ." .* ." 9ia .«• .u> ."•
^ ' •- x
CO
/'/./'
7 f rltj v
y////// ./ /
T9I ^J /SO 40 a89/ %° f* a08 _^?° »W a10 »S J° »F »Q
/ BOni
an• ,
}WO
CUMULATIVE TIME. MINUTES
*" /"J ••/ ^ f "f -la -17 •" --ai '
// 1 ' / 1 I \ /
r r -r &* /jos-'!s "V* •"/" j
III if//
JSB I," jTB I »0 ,86 I m»| JB/ » JQ^li J« J
I I
H 1 I 1 t
I I t
U U 19
ic t7 ta ta
TEST A. RLM 35• PlAP 71 FU> 36 GPU. OETENTITM 415 SET. li-li-S2
-------
Figure 44
MODEL DYE TEST
H H
H - 1
1
H »-
B. .
a..
7. .
5..
•4. .
a..
a..
i..
o..
HO_S>.
\\\
a3i33\oi±riait
• K.'-k • A • • •
A x
faT l\ k \" V • • ifij •
iiJD;.oo)^ai^tf2i££t40i<3i«;ixufliaGii?i03/cc
CUMULATIVE TIME. MINUTES
H H
H - 1
1
1
1 - h
H
O 1B34SG79O1O
TEST Ci RLN 3Bi RM3 7, FLOW 40.8GPK
-------
Figure «5
MOOU OYE UST USING PUHP NUMBER 7 at ?0.5 GPM
Cftte: ll/l? :••>
Teat No. : i :.U11 ;;0_ : ly
Pump Nc . : 7
20.5 KP^i
Weirs or Paf:'lcs:
Nuno
2il: BJnmnrk Oro.r.
Cone. , JC: 10
Vol. . ce: KD
Water Tempers'jre . *P:
53
Con'Jnent.a:
Vatcr level it Inactive
pump discharge poria;
«'ntr*r rtlcl nc^ ovci flo1-'
cffhirr.t, •-•f.-ir :r
sampling points Q,
9, 10
At 21U 3ec ?2~:jU.-.^
atarled
At :)'!C aec I.^tvy coriccn-
trotlon t'/orflc*'lng
At fWO CL-C color r-.cdiu.Ti
to llr.hi-, I;a-c^y
At 1200 ace 30:npl!n5
stopped
At 12GO rec fairly even
distribution of ll^nt
to p.oJlUT. dye frrm
Lochtlcn i to cfflu-jnt
weir
136
-------
TABLE XXXVIII
EFFECT OF BAFFLES UPON DETENTION TIME AT VARIOUS
RATES OF PUMPING
Tes
No.
All
A
C
B
No
D
F
E
G
t Run
No.
Baffles
35
38
37
Baffles
39
40
41
42
Date
(1969)
11/11
11/13
11/13
(1969)
11/13
12/12
11/13
12/12
Model
Rate
(gpm)
28
41
103
21
40
75
103
83
163
305
410
37,200
73,100
137,000
184,000
Detention (min)
E q u i v .
Gross
28
41
103
116
166
410
52,000
74,500
184,000
66.1
42.4
20.1
30.3
13.4
8.9
Theoreti cal
76
52
21
101
53
21
The retention times obtained under the un
baffled conditions plotted against the th
in the basin are shown in Figure 46. The
efficiency from baffling is clear. From
the optimum baffling configuration would
compromise between cost, increased effici
limitations is necessary. The hydraulic
25 to 100 percent of the overflow volume
system under increased rates of flow must
baffled and optimally
eoretical retention
increase in retention
a practical standpoint,
be costly so a
ency and hydraulic
problem of passing
through a baffled
also be cons idered.
OPTIMUM CORRELATION OF FLOCCULATION, LONG-TUBE SEDIMENTATION
AND MODEL TEST RESULTS
Prediction of removals of discrete fully-f1occculated (Class I)
solids from settling basins can be based on batch settling
tests and on model dye tests. The analysis of sedimentation
conducted in long vertical tubes is discussed in a previous
section of this report. Dye tests in the hydraulic model
were analyzed as described below.
A typical dye test was conducted on a basin model by injecting
a slug of dye or other marking substance into the basin
influent such that it was mixed uniformly into all elements
of fluid entering the basin during a very short time interval
following time zero. Effluent was sampled periodically
at a number of stations preferably uniformly spaced along
137
-------
Figure 46
EFFECT OF BAFFLES AT REDUCED FLOW RATES
1,000
800
600
400
— O
200
100
80
60
40
NO BAFFLES
THEORETICAL
PUMPING RATES, CFS
(PROTOTYPE)
20
10
I
1_J I
I
1
.7
2 4 6 8 10 20 30
DETENTION TIME, MINUTES ( PROTOTYPE) x 10
50
138
-------
the discharge weir. All effluent samples were analyzed
for dye concentrations. The average detention time of dye
was determined for each station2. The local overflow rate
(basin depth divided by average detention time was deter-
mined for each station. The predicted local value of the
fractional solids not captured, (F ), was determined. A typical
plot of the local values of F vs.^distance (1) along the
weir, is shown in Figure 47. ^The average value of local
F is defined as:
F - y [^
b " w
This quantity represents the fraction of the initial solids
which is predicted to overflow the basin under the conditions
of the dye test.
A basin operating at an overflow rate different than that
used during the dye test should have approximately similar
flow patterns in a kinematic sense. In such a case, all
local average detentions will change by a constant factor,
which is equal to (test overflow rate)/(new overflow rate).
Under the assumption that this is true, predictions of solids
loss at overflow rates can be made by multiplying the observed
values of detention times by this overflow factor, and then
carrying the calculations through as before.
Correlation of results from flocculation and sedimentation
in long-tube tests and the dye tests are developed sequentially
using the procedure described earlier in this report.
Figure 48 is plotted from a model dye test (Run 31) with
all weirs in the basin. Calculations are made from data
following the procedure described earlier2. The average
losses are also given in Figure 48. The predicted loss of
suspended solids from the MRPS model and the loss relative
to MRPS basin flow are presented in summary, in Figure 49.
PREDICTED PERFORMANCE AT THE MRPS
The area required for removal of fully flocculated Class I
solids as a function of overflow rate are shown in Figure 50.
Intercepts with the 60,000 sq ft ordinate are the predicted
removals for various flows to the existing MRPS basin. An
area of 80,000 sq ft would be required for 85 percent removal
for a normal event handling 305 cfs (137,000 gpm) combined
sewage. Approximately double the area is required for a
10 percent increase in removal from 85 to 95 percent. It
can readily be seen that for influents of 102 cfs containing
75 to 250 mg/1 suspended solids, the predicted effluent
quality would be 15 to 50 mg/1 from the existing MRPS. The
139
-------
Figure 47
DISTRIBUTION OF LOSS ALONG WEIR
4567
WEIR STATIONS
L
Fb AV LOSS
END OF WEIR
WEIR STATIONS ASSUMED EQUALLY SPACED
AVERAGE ALONG WEIR = Fb =0.12
-------
Figure 48
AVERAGE LOSS ALONG PROTOTYPE EFFLUENT WEIR
FRACTION OF SOLIDS
NOT CAPTURED, Fp
.4
.3
.2
0
AVG.Fb
ALONG WEIR
A
0
B
I
C
2
D
3
E
4
F
5
G
6
H
7
8
J
9
0.343
0.181
0.103
0.031
END OF
WEIR
RUN MODEL (GPM)
a. * 25
b. 38 40.8
c. 31 75
d. ** 102
PROTOTYPE(CFS) Fb
102 .031
166 .103
305 .181
410 .343
* RESULTS AVERAGED FROM RUNS 31 a 35
** RESULTS COMPUTED FROM RUN 35
-------
Figure 49
PREDICTED LOSS FROM THE MRPS PROTOTYPE WITH FIVE WEIRS
SYSTEM: 20mg/l BAROID HECTORITE CLAY
20mg/l PURIFLOC C3I
COMBINED SEWAGE 166 mg/1 GRAVIMETRIC SOLIDS
CLASS I SEDIMENTATION OF FULLY FLOCCULATED SOLIDS
MODEL FLOW, gpm
50 75 100
0
50
40
30 —
25
125
150.
LOSS, PERCENT |
t i
20
10
8
7
6
5
4
3
50
60
70
80
REMOVAL ('RETENTION),%J
1
90
92
93
94
95
96
97
98
100 200 300 400
PROTOTYPE FLOW.cfs
500
99
600
142
-------
OJ
Figure 50
PREDICTED AREA REQUIRED FOR REMOVAL OF CLASS I SOLIDS
FLOCCULATED WITH B.H.CLAY 20mg/l 8 PURIFLOC C3I 20mg/l
50.0
OVERFLOW RATE, fph
30.0
20.0
10.0
8.0
6.0
4.0
3.0
2.0
1.0
50,000 100,000 200,000
102 cfs
64.4
79.2
90.0
96.5
SOLIDS REMOVAL,
PERCENT
500,000
AREA, SQ.FT.
1,000,000
5,000,000
-------
MRPS could be modified using flocculants to provide a
fully-flocculated influent at 187 cfs and achieve approxi-
mately 91 percent removal of suspended solids.
The predicted removal efficiencies are based on realistic
basin hydraulic conditions simulated in the model and also more
idealized settling conditions in the long-tube sedimentation
studies. Additional investigation of sedimentation rates
developed under various conditions of long-tube mixing would
be required before finalization of basin design.
144
-------
SECTION 9
DISINFECTION
The evaluation of the effectiveness of chlorine and other
disinfection systems in the treatment of combined stormwater
is of considerable interest at this time. Current information
on the disinfection of combined sewer overflows is sparsel»2,
and the effect of untreated combined sewer overflow on public
health is of serious concern. The problem has been recognized
only in recent years and now methods must be found to treat
this flow effectively and economically. Coupled with the
interest in disinfecting combined sewer overflows, there is
also considerable interest in improving the safety practices
associated with disinfection. The need for safer alternatives
for chlorine gas is becoming more apparent as population
densities in urban areas around water and wastewater treatment
increase. The press of urban and suburban development has
made once undesirable land near sewage treatment plants
or pumping stations now more acceptable. The situation
at Milk River combined sewer overflow pumping station is
a case in point. There are potential dangers of maintaining
large inventories of chlorine required to adequately treat
resultant waste flows. Some larger metropolitan centers
such as New York, Chicago, and New Orleans, have already
converted to relatively safe, though significantly more
expensive sodium hypochlorite^»4.
Four major problems must be dealt with in the chemical treat-
ment of combined sewer overflows: (1) the tremendous volumes,
(2) the relatively short time available for treatment,
(3) control of large quantities of potentially dangerous
chemicals commonly used for disinfection, and (4) extremely
variable rates of flow. The ultimate disinfectant should
be safe to handle, relatively efficient, and reasonably
economi cal.
OBJECTIVES AND METHODS
The primary objective of the disinfection work at the Milk
River Pumping Station (MRPS) was to evaluate chlorine and
other disinfectants for the treatment of combined sewer
overflow. In the course of fulfilling this objective, data
were also gathered on dry-weather sewage to provide a basis
for evaluating treatment efficiencies. A number of parameters
considered were: disinfectant dispersion, disinfectant concentra-
tion, and contact time.
145
-------
Dispersion actually encompasses three items: the form of
introduction (gas vs. liquid, minor stream vs. total stream),
type of diffusion, and degree of mixing. It is also important
to consider efficiency as a function of concentration. It
is desirable to have a chlorine residual sufficient to continue
killing pathogens after treatment has been completed, but
not strong enough to endanger desirable aquatic organisms
such as fish. Another critical parameter is contact time.
Concentration (C) and contact time (t) are related to kill
efficiency (E) in the following manner5'6:
E = K t Cn
where K incorporates certain demand-exerting characteristics
of the sewage, such as HLS or other reducing compounds,
and n is a rate constant of the reaction.
All of the disinfectants used in this study, regardless of
their source (BrClr, Cl?, NaOCl , etc.), immediately hydrolyze
to form hypohalous acids in solution. The undissociated
hypohalous acid species is the most effective form of halo-
genated disinfectants according to current theory. In the
case of chlorinated systems, it is generally accepted that
the neutral HOC1 molecule is able to penetrate the cell
much more readily than its charged conjugate, reacting with
vital enzymes and thereby destroying the cell.
Secondly, the amount of hypochlorous acid formed is dependent
on the source and the degree of dissociation, which is sensitive
to pH. Hypochlorous acid dissociates in the following manner:
HOC1 v -^ H+ + OC1"
7 -8
where K = 3.2 x 10 . The concentration of undissociated
hypochlorous acid decreases rapidly with increased pH°.
The pH of the sanitary sewage flow at the MRPS is typically
around 7.4. Combined sewage overflow pH varied from 7.0-7.5.
Fielding et al.9 have pointed out that many investigators
today are ignoring the effect of pH. At pH 6.0, 97 percent
of the HOC1 is present in an undissociated form; at pH 7.4
only 50 percent undissociated HOC1 is available.
A third factor is the effect of reducing materials such
as nitrogenous compounds in the sewage. These convert HOC1
to chloramines which are thought to be about 20 percent
as effective as HOC1. Conversion is reported to be a maximum
at pH 8.3!°. This phenomenon does not present a problem
with bromine compounds; bromamines do form but they seem
to be about as effective as HOBrll.
146
-------
DISINFECTION PILOT PLANT
Combined sewage overflows are heterogeneous systems which
can vary abruptly in chemical and physical character. Control
systems and experimental systems measured on different days
make comparisons difficult. A parallel pilot plant, using
one side as a control and one for experimentation, was designed
to avoid this problem. This system also permitted direct
comparison of any two disinfectants. A schematic view of
this arrangement is shown in Figure 51. The pilot plant
flow was pumped from the wet well of the pumping station
to a constant head tank which had a detention time of less
than two minutes.
Figure 51
Disinfectant Reaction Chamber and Detention Coil
A variable speed pump serving each plant conducted the flow
into a reactor chamber (10" dia., 5' high) fitted with a
variable speed stirrer and baffled to increase mixing efficiency
(Figure 51). Five disinfectant addition points were provided
along the side of the reactor to allow variance of reaction
time. The disinfected flow then proceeded to a coil of
polyethylene tubing (1.5" ID, 1500' long) which was equipped
with four sample points along its length (Figure 52). Detention
time in the coil was about 23 minutes as determined by timing
the passage of dye fronts between the disinfection injection
points and the sampling ports.
147
-------
Figure 52
SCHEMATIC OF PARALLEL OPERATION
OF DISINFECTANT PILOT PLANT
FLOWMETERS
SAMPLE 4
.47 GAL/MIN
SAMPLE 4
.47 GAL/MIN
SAMPLE 3
.69 GAL/MIN
SAMPLE 2
POLYETHYLENE
DETENTION
COILS
00
to
CV1
'.93 GAL/MIN
SAMPLE I
GAL/MINI
REACTORS
.PUMP
DISINFECTANT
CONSTANT
HEAD TANK-
SAMPLE 3
.69 GAL/MIN *
SAMPLE 2
.93 GAL/MIN'T
SAMPLE I
,.99 GAL/MIN'
to
OJ
DISINFECTANT
RAW SAMPLE
INFLUENT
148
-------
The effluents of both plants were equipped with a flow meter
which provided accurate monitoring of the flows through
both plants. Effluent flow was maintained at 5 gpm for
each plant. Knowing the effluent flow and the flow out
each sample point, the total flow through the reactor could
be determined and chemical feed rate could be adjusted to
maintain the desired feed concentration. The total flow
through each reactor was approximately 8 gpm.
These flows and the pilot plant design represented an attempt
to achieve 5-25 minutes contact time which was felt to be
necessary for adequate disinfection. The flows also had
to be of sufficient velocity to prevent undue settling of
solids in the contact coil. These considerations then had
to be incorporated in the construction of a plant of reason-
able size.
Attempts to automate sampling proved unsuccessful; manual
sampling was accomplished without difficulty. Two 100 ml
samples were collected simultaneously from each plant at
the appropriate time and sample point. In those experiments
involving variable contact time, the object was to sample
the same slug of flow as it proceeded through the plant.
The detention times determined from the dye studies allowed
this to be done fairly accurately.
DISINFECTANTS AND THEIR APPLICATION
Three disinfection systems were of prime interest:
(1) chlorine, (2) sodium hypochlorite, and (3) a chlorine-
bromine mixture (BrCl5). A system in which 10 mg/1 Br~
was injected ahead of Clp application was also examined.
The various advantages and disadvantages of these systems
are summarized in Table XXXIX. Chlorine has a significant
economic advantage although it is probably the most hazardous
of the systems considered. Recent accidents involving large
chlorine gas release in populated areas have caused much
concern by public health officials. Chlorine and the other
gaseous halogen systems can release harmful gases into the
atmosphere during turbulent dispersion. If the concentration
is high and applied at a fast rate with accompanying dispersing
turbulence, this atmospheric accumulation can be considerable.
All of the gaseous systems are highly corrosive.
Because of its common usage, chlorine is useful as a per-
formance standard against which other disinfectants were
measured. Sodium hypochlorite has received some attention
recently as a replacement for chlorine and warranted examin-
ation in this study3»4. The NaOCl system is advantageous
149
-------
TABLE XXXIX
ADVANTAGES AND DISADVANTAGES OF
VARIOUS DISINFECTION SYSTEMS
Chlorine
Advantages
Low cost
Stable
Disadvantages
Hazardous
Corrosive
Gas evolution
NaOCl
Safe
Non-corrosive
(except to Al)
No gas evolution
Large quantities
easily dispersed
Expensive
Higher capital investment
Poor storage stability
BrCl
Fast
Stable
Hazardous
Corrosive
Gas evolution
NaBr + Cl
Fast
Hazardous
Dual system
Corrosive
Gas evolution
from a safety standpoint, and can be rapidly dispersed in
large quantities. However, it is unstable, expensive
(relative to chlorine), and usually requires substantial
investment. The NaOCl system is the preferred disinfectant
at the Milk River Pumping Station principally because of
the high population density in that area. The use of bromine-
chlorine mixtures (specifically BrCl5) as disinfectants
has not been reported extensively in the literature. Consider-
able work with these mixtures has been done by The Dow Chemical
Company. These systems may be potentially superior to chlorine
in efficiency and therefore require reduced volumes to achieve
comparable results. While BrClc shares many physical properties
with chlorine which make it hazardous, a safety advantage
might be realized in the reduced volumes required'^.
150
-------
Chlorine is generally fed as a gas and may be applied either
to a minor stream or to the total stream. The pilot plant
was set up to feed chlorine from a 15 Ib. tank through a
minor stream arrangement. The BrClr was fed in a similar
manner except that it was drawn from its storage cylinder
as a liquid. The liquid quickly evaporated in the feed
line with the aid of a heat tape and was fed through the
minor stream aspirator in the same manner as the chlorine.
Liquid BrClr was withdrawn to prevent changes in the ratio
of bromine to chlorine which would occur if the chemical
were withdrawn in the vapor state. Both the chlorine and
the BrClr systems were controlled by a cylinder mounted
chlorinator which was capable of delivering up to 1.5 Ib.
Clp/24 hours.
Sodium hypochlorite was fed via a microfeed pump as a 5.25
percent solution through a small diffuser inserted in the
reactor in place of the aspirator. Sodium bromide was fed
as a 0.1 N solution in a manner similar to that of NaOCl .
ANALYTICAL PROCEDURES
A number of analyses were performed on samples collected
from the pilot plant. These included: chlorine demand^,
chlorine residual, total coliform, fecal coliform, fecal
strep, nitrogen'^, total carbon^, suspended solids^.
The nature of this work was such that a spectrophotometric
technique for total chlorine residual was considerably more
convenient than the conventional amperometric technique.
To verify the validity of this test and its correlation
to the amperometric^, page 440s a standard curve was run
with NaOCl in distilled water. The two chemical delivery
systems (NaOCl microfeed pump and the cylinder mounted chlorinator)
were then calibrated by both methods by running tap water through
the plant. There was good agreement between the methods.
The spectrophotometric technique consisted simply of collecting
the sample (about 100 ml) and immediately dosing with about
1-2 grams reagent grade potassium iodide. The intensity
of the developed color is proportional to chlorine residual
concentration. Measurements were made in less than one
minute at 420 nm using a spectrophotometer. Acidification
of the sample was not considered necessary prior to analysis
for residual chlorine because: (1) pH was not exceptionally
high (usually about 7.0) and was not significantly affected
by the potassium iodide additions due to the buffer strength
of the sewage, (2) the amount of potassium iodide added was
far in excess of stoichiometry to insure complete reaction,
and (3) suspended solids were not exceptionally high (about
151
-------
100 mg/1 in sanitary flow; 200 mg/1 in storm flow) so that
significant adsorption of color bodies onto the solids was
not a problem.
A second 100 ml sample was also collected at the same time
as the residual sample in a bottle containing about 0.5
ml sodium thiosulfate which arrested further disinfection.
The samples were then analyzed for total coliform, fecal ,r
coliform, and fecal streptococci via the MILLIPORE® technique
EXPERIMENTAL PROGRAM
Initial work was performed using sanitary sewage. Runs
were made using chlorine, sodium hypochlorite, and BrClr,
at conditions of 10 mg/1 as Cip, 100 percent mixing speed,
and fed at the lowest disinfection addition point on the
reactor. The initial variable was contact time. From this
point the work proceeded into two areas: varying the concen-
tration of disinfectant and studying the effect of disinfection
on f1occulation. This experimental program was essentially
duplicated on storm sewage.
A summary of the experimental program for storm and dry-
weather flow is shown in Table XL. For the purposes of
discussing the data, reference is made again to the schematic
layout of the plant as presented in Figure 52.
TABLE XL
SUMMARY OF DISINFECTION EXPERIMENTAL PROGRAM
Variable Time Variable Dose
System Constant Dose Constant Time
C12 vs C12 S S
C12 vs NaOCl S*, C* S
C12 vs BrCl5 C C
NaOCl vs BrCl5 S, C S
NaOCl alone S, C C
NaBr + C12 vs NaBr + C12 S
NaBr/NaOCl vs NaOCl S
NaBr/NaOCl vs C12 S S
BrCl,- vs C12 + NaBr C
*S = Sanitary sewage
*C = Combined sewage
152
-------
RESIDUAL AND DEMAND vs. DISINFECTANT DOSE
The use of chlorine residual as a feedback device to control
disinfectant dose is a practice carried over from water
treatment. Unfortunately, wastewater systems are much more
complex and maintenance of a given residual level is difficult
at best. This is due primarily to continuous and erratic
fluctuations in the concentrations of those substances which
directly exert a demand on dissolved chlorine (e.g. H^S).
Frequently, if a residual is measured after 10-30 minutes
contact, and it is fluctuating noticeably in that period,
elaborate dampening devices must be employed to prevent
increasing harmonic oscillation in dose correction. Not-
withstanding these considerations, residual is still the
best measure of dose efficiency available. From an economic
as well as a public health standpoint, a goal in disinfection
should be to maintain the lowest residual to give the desired
bacteri al kill.
A number of experiments were performed in this work in which
residual chlorine was monitored at a given contact time
(1.3 and 6.4 minutes) while dosage was varied from 2.5 to
15 mg/1 Cl?. The data gathered on combined sewer overflows
parallel tne sanitary data at the lower doses. Chlorine
and BrClr are almost equal at high doses (10-15 mg/1 Cl?)
in terms of residual while limited NaOCl data indicated
possibly slightly lower residuals than that found in sanitary
f 1 ow.
One of the factors which influences reaction rates and the
extent of reaction is reactant concentration. Demand may
be influenced therefore by disinfectant dose. The data
in Table XLI support this supposition. This is an alternative
way of comparing the results shown in Table XL since demand
is merely the difference between residual and dose at any
given time. This type of consideration is necessary from a
design engineering viewpoint to determine inventory capacities.
The important point to note in Table XLI is that demand
usually does increase with dose.
Another variable which is important to disinfection control
is the range of variability of chlorine demand. Analysis
of 56 influent samples has shown that the chlorine demand
had an average value of 7.7 mg/1 and varied from a minimum
of 3.5 mg/1 to a maximum of 12.2 mg/1 for a maximum fluctuation
of 8.7 mg/1 from storm to storm. For any selected storm,
the maximum fluctuation in influent chlorine demand was
6.2 mg/1 .
153
-------
en
TABLE XLI
DEMAND, RESIDUAL, AND BACTERIAL KILL
FOR COMBINED SEWAGE*
Contact Time
(Minutes )
ci2
NaOCl
BrCl5
0
1.3
6.4
14.4
23.0
0
11 .3
6.4
14.4
23.0
0
1.3
6.4
14.4
23.0
Demand at Dose of Residual at Dose of
2.5 5 10 15 2.5 5 10 15
2.0 2.7 4.6 7.0 0.5 2.3 5.
2.3 5.0 6.3 6.1 0.2 0.0 5.
10
3.
1.5 2.0 7.2 9.1 1.0 3.0 3.
3.
4.
10
2.5 3.4 5.0 7.5 0.0 1.6 5.
2.5 5.0 4.8 8.0 0.0 0.0 4.
4.
4.
6 8.0
8 8.9
9
7 5.3
1
5
6 7.5
9 7.0
7
4
Dose to Kill 99.99%
TC FC FS
8.1 8.7 15.0
6.1 7.5 13.1
8.0 15.0 10.8
8.8 8.4 6.5
9.3 14.0 12.0
*A11 demands, residuals, and doses to kill 99.99% of bacteria
indicated are in units of mg as C1-/1.
TC = Total Coliforms
FC = Fecal Coliforms
FS = Fecal Streptococci
-------
BACTERIAL POPULATION vs. DISINFECTANT DOSE
The principal objective of a disinfecting system is to kill
pathogenic organisms. As pointed out previously in this
report, the kill efficiency of a given disinfectant system
is influenced by disinfectant concentration and contact
time. Research by The Dow Chemical Company has shown that
brominated systems may be very effective at reasonable doses
in comparatively short contact time.
The experiments of the previous section, where contact time
was held constant at 6.4 or 1.3 minutes, and dose varied
from 2.5 to 15 mg/1 Clo. were monitored for bacterial concen-
tration. The disinfectant dose required to give 99.99%
kill at these contact times was measured. It was found
that when short contact time is available (<10 minutes),
dosing in the range of 8-15 mg/1 as chlorine was required.
The more dilute stormwater generally required less chemical
to achieve a given percent kill than sanitary flow.
RESIDUAL vs. CONTACT TIME
If it is desirable to keep chemical dosing to low levels for
protection of receiving waters or because of high chlorine
demands of the receiving water, contact time becomes a very
important parameter. This is especially important at Milk
River where sludge deposits exert a significant chlorine
demand. A number of experiments were conducted in which
disinfectant dose was held at a constant level (arbitrarily
10 mg/1 as Clp) and a slug of flow was sampled as it progressed
through the plant. Samples were collected and treated in
the manner previously described. All systems, with the
exception of NaOCl and BrClr in sanitary flow, left residuals
in the range of 5-6 mg/1 as C12- There was no appreciable
decrease in residual over the times monitored (Table XLI)
indicating that most of the chlorine demand is satisfied
within 1.5 minutes. Similar results were obtained for sanitary
f 1 ow.
BACTERIAL SURVIVAL vs. CONTACT TIME
The experiments described in the previous section were also
monitored for changes in bacterial numbers. It is important
to know the amount of contact time required for various
systems to achieve a certain kill at a given dose. Data
of this type are summarized from experiments on storm flow
in Table XLII. The average times necessary for various
systems to kill 99.99% of total coliform are shown.
155
-------
TABLE XLII
TIME REQUIRED TO KILL 99.99% OF TOTAL
COLIFORMS IN COMBINED SENAGE*
Initial
Initial Demand
Run # BrC15 Cl_2 NaOCI Total Coliform mg C12/1
C-12 4.4 - 4.5 19 x 106 11.3
C-ll - 4.8 5.0 10 x 106 10.2
C-9 1.3 7.4 - 7.6 x 106 10.2
C-7 8.8 8.1 - 8.2 x 106 9.5
C-5 - - 8.5 4.5 x 106 10.7
C-3 - - 17.7 13.8 x 106 7.7
C-2 - - 21.0 6.8 x 106 6.2
C-l - - 14.7 8.5 x 106 5.7
*A11 values in units of minutes using concentrations
equivalent to 10 mg as Cl^/l-
No significant differences in the time required to achieve
a 99.99% kill were observed. The data obtained during run
C-9 showed a significant decrease in time of kill for BrCl5
over Cl 2 but this was not supported by the other data. Con-
siderably more work would be needed to quantitate any real
differences.
EFFECT OF DISINFECTANTS ON CHEMICAL FLOCCULATION
Because of the short time available to treat stormwater,
it may be necessary to disinfect simultaneously with other
chemical treatment such as f1occulation. Jar test experiments
were conducted on sanitary flow to determine the effect
of disinfection on flocculation by polyelectrolytes.
Both a cationic (PURIFLOC C31) and an anionic (PURIFLOC A23)
flocculant system were tested. Using the optimum flocculant
dose determined from jar testing, NaOCI was added just prior
to polymer addition at variable concentrations or intervening
mixing times. After flocculation and settling were completed,
the percent optical solids removal from the overhead was
determined from turbidimetric measurements. This was taken
as an indication of flocculation activity. The removal of
optical solids (percent) was calculated using the following
formula:
(1 - A/AQ) x 100 = % Removal of Optical Solids
156
-------
where A is the absorbance of the experimental sample and
A the absorbance of the raw sample. The experiments were
carried out in series of liter quantities using a gang stirrer
In the cationic system the optimum dose of PURIFLOC C31
was 40 mg/1 and NaOCl was first added in concentrations
from 0 to 80 mg/1 as C12 with one minute intervening rapid
mix (100 rpm). No impairment to flocculation was observed
until 40 mg/1 as Clp was exceeded as shown in Figure 53.
The NaOCl dose was then held constant at 10 mg/1 as C12
and the intervening mixing time between halogen and poiymer
addition varied between 0 and 5 minutes. No effect on
flocculation was observed.
The anionic system usually requires a multivalent cation,
in this instance 25 mg/1 Fe^+, to promote flocculation.
The iron was added, stirred 5 minutes at 80 rpm, followed
by NaOCl in concentrations of 0-80 mg/1 as Cl?. After one
minute rapid mix, the polymer was added. Flocculation was
not affected even up to 80 mg/1 as Clp. In the next experi-
ment, NaOCl was held at 15 mg/1 as C12 and the intervening
mix time varied. Again, flocculation was unaffected.
It was concluded that at doses of disinfectant (<40 mg/1
Clp) chemical flocculation can be carried out simultaneously
without impaired performance.
LONG-TERM RESIDUAL OF BrC15 IN COMBINED SEHAGE
A study was made to compare the long-term effect of Clp
and BrClr residuals in combined sewage. Samples of combined
sewage were taken from each of the two pilot plants after
treatment with 10 mg/1 BrCU and CK, respectively. The
samples were held for a period of 24 hours and the residuals
checked. The results of this study are shown in Table XLIII.
TABLE XLIII
LONG-TERM RESIDUALS OF DISINFECTANTS IN COMBINED SEWAGE
Residuals in mg/1
Time (Hours) C1? BrCIc
1 5.7 5.1
4 4.6 3.8
20 2.3 2.3
157
-------
Figure 53
EFFECT OF SODIUM HYPOCHLORITE ON
FLOCCULATION BY PURIFLOC C31
20 40 60 80
NaOCI CONCENTRATION (mg/l CI2)
100
158
-------
Sunlight effects were minimal. Temperature of the samples
increased from the temperature of the flowing sewage to
approximately room temperature. The samples were not mixed
so the effect of turbulence was minimal. As expected, BrCl,-
maintained a residual similar to that of the C12 during
the 24 hour period.
CHLORINE DEMAND OF THE MILK RIVER
The Milk River is a channel about one mile in length, connecting
the MRPS to Lake Saint Clair. It receives almost no dry-
weather flow. An ideal condition would be to maintain a
slight (0.5-1.0 mg/1 C12) residual in the channel but the
highly reducing environment precludes this. Anaerobic decom-
position of the bottom sludges causes considerable generation
of methane, H^S and NH...
In order to assess the conditions in the channel with respect
to chlorine demand and bacterial population, two samples
were collected at each of the five stations shown in Figure 54.
The first sample was taken about one foot below the surface
(in the approximate middle of the channel); the second sample
was bottom sludge. Five and thirty minute demands were
determined for both samples; bacterial population was deter-
mined for the overhead sample.
The bottom sludges had extremely high demands. The profile
in chlorine demand through the channel of the overhead sample
is shown in Figure 55. The demand declined as the channel
outlet was approached. The magnitude of the demands by
the overheads and bottom sludges demonstrates the impractic-
ability of maintaining a residual in the channel. Disinfection
of stormwater will have to occur in the settling basin since
the extreme chlorine demand will consume any residual as
soon as the flow enters the channel. The populations of
various bacterial forms at each station are shown in the
histogram of Figure 56. Populations of all three classes
were relatively constant except in the clean water of the
lake.
DISINFECTION OF MILK RIVER COMBINED SEHER EFFLUENT
Based on theoretical detention times in the existing settling
basin, 99.99% of the initial bacterial population could
be killed at the MRPS for 95 percent of the overflow periods.
An adequate supply of stable disinfectant would be assumed
available. The pumping rate was also assumed constant for
a given storm. In reality, intense storms of short duration
would require increased pumping rates and result in shorter
159
-------
Figure 54
Stations in the Milk River Channel
Sampled for Chlorine Demand and Bacterial Population
The Milk River
\
160
-------
Figure 55
PROFILE OF CHLORINE DEMAND IN THE
MILK RIVER CHANNEL
30
o
E
i
Q
UJ
Q
20
15
10
I I 5 MINUTE
ill 30 MINUTE
234
STATION NUMBER
161
-------
15
fO
O
Figure 56
PROFILE OF BACTERIAL POPULATION IN THE
MILK RIVER CHANNEL
A TOTAL COLIFORM
B FECAL COLIFORM
C FECAL STREPTOCOCCI
10
o
o
CO
LU
GO
0
ABC ABC ABC ABC ABC
12345
STATION NUMBER
162
-------
contact periods in the channel and higher disinfectant con-
centrations would be required. The frequency of pumping
is less and the channel is full toward the end of a given
storm. A new demand value must be considered in such cases.
Accurate dosing during a storm event is a statistical problem
of huge proportions.
A statistical analysis of the frequency of volumes pumped
for all storm events occurring at the MRPS over an eight
year period reveals that 95 percent of pumping time involves
volumes less than 8 million cubic feet (59.8 x 106 gal).
Further, 95 percent of the time the pumping rate is less
than 715 cfs (321,000 gpm). Since the basin volume is .508
million cubic feet (3.8 x 10" gal), the theoretical contact
time will be at least 11.8 minutes. Even with detention
inefficiencies, it appears that this is sufficient time
to achieve 99.99% kills at a dose of 10 mg/1 as C12 with
any of the systems.
Because of extensive short-circuiting in the basin, modi-
fications of the hydraulic design would be required to approach
the theoretical detention times. The disinfectant then
could be added at the pump discharge. Adequate dispersion
would be available, contact time would be sufficient, and
simultaneous flocculation would not be adversely affected.
CORRELATION DATA
During the course of the work at the MRPS, considerable
data were collected on influent storm flow, e.g. various
forms of suspended solids, bacterial counts, chlorine demand,
etc. The correlations between several of these variables
were determined via computer. The main objective was to
discover a statistically significant (^.0.70) correlation
between chlorine demand and some easily or at least immediately
measured parameter. The disadvantages of the present system
of gauging chlorine demand by delayed residual information
has been previously discussed.
Correlations between chlorine demand and other parameters
were extremely poor. Total coliform did not correlate to
the solids data. Fecal coliform displayed the highest corre-
lations with the solids and nitrogen data, but no relation
to time of storm duration.
It was concluded that correlation of disinfection variables
with easily measured parameters at the MRPS was not possible.
163
-------
SECTION 10
ECONOMIC ANALYSES
GENERAL
Numerous factors affect the ultimate cost of combined sewer
overflow treatment systems. Analysis of the frequency dis-
tribution of flow rates for the Milk River Pumping Station
(MRPS) suggests a rapidly decreasing economic benefit if
overflow treatment capacity is increased. Some measure
of this relationship is established in the following analysis
of the MRPS data. The analysis is valid for many combined
sewer treatment systems where the following characteristics
prevai 1 .
1. Capital costs vary approximately in proportion
to plant capacity.
2. The occurrence of flow rates follows a logrithmic
normal distribution similar to that at the MRPS.
Data presented in Table XLIV are derived from the frequency
distribution analysis shown in Table IX. A flow of 2450
cfs was assumed to represent the peak plant capacity for
a ten-year storm (100%). The percentage of peak flow for
each of the rates shown was then calculated. Benefits were
assumed to be the percent of total volume treated; treatment
of all volumes equaled 100 percent.
PROPORTIONAL
Rate*
(cfs)
305
410
610
915
1220
1535
1840
2155
2450
TABLE XLIV
COSTS AND BENEFITS AT
Cost
% Peak
Rate
12
17
25
37
50
63
75
88
100
VARIOUS FLOW RATES
Benefit
% Volume
Treated
45
74
78
85
89
94
97
97
100
*1 cfs = 448.8 gpm
165
-------
It can be seen from Table XLIV that the rate of benefit
decreases significantly when plant size reaches approxi-
mately 20 percent of peak flow, suggesting that an economic
breakpoint in combined sewage treatment facilities may occur
at about that point. A final breakpoint would have to be
developed from cost analyses of specific alternate plans.
The foregoing relationship, however, provides a rational
beginning for economic evaluation of the Milk River combined
sewer overflow treatment problem.
FLOCCULATION AND SEDIMENTATION
Unlike the design of most conventional sewage treatment
plants, designs of combined sewer overflow treatment systems
must consider the logarithmic normal distribution of flow
rates. Accurate estimates of treatment efficiency and cost-
benefit relationships cannot be derived without this consider-
ation. A plant capacity once established must be examined
for efficiency over the expected range of flows. These
flows must be weighted according to their probable frequency
of occurrence.
A condensation of the contributing effects of variable rates
and volumes of flow at the MRPS to the degree of solids
removed by the PURIFLOC C31-BAROID HECTORITE clay system
is contained in Table XLV. The lowest flow rate is that
for which the existing retention basin could operate with
about 96 percent solids removal efficiency; the highest
flow rate is that which is near the economic breakpoint
for capital investment. Note that the 50 percent rate occurs
near the lower limit.
The flow volumes were selected to cover a range overlapping
the 50 percent volume occurrence. It should be noted that
the relative magnitudes of storm volumes can influence the
predicted removals. If the volume retained by the existing
basin is large relative to the volume overflowed into the
Milk River, then the actual removals may be greater due
to a longer effective detention time. As the volume retained
in the basin becomes a smaller fraction of the total volume
treated, then the removals should approach the predicted
removals .
The selected rates of overflow, including the statistical
mean (50 percent occurrence) tabulated in the first column
and the corresponding percentages of occurrence included
in the second column were taken from Figure 12. The predicted
percentages of solids removal in the third column were taken
from Figure 50. If an equal solids removal of 85 percent
is desired, then the required net areas for the selected
overflow rates as taken from Figure 13 are included in the
fourth column.
166
-------
TABLE XLV
RELATIVE SOLIDS REMOVAL AS A FUNCTION OF OVERFLOW RATE AND VOLUME
, Predicted Net Area to Predicted Solids Removed (fraction
Overflow Occurrence Solids Removed Achieve 85% in. Existing Basin at Treated Volumes
Rate iOverflow in Existing Solids Removal Million cu fj .512 1.2 2
Jcfs) Rate (%) Basin (%) (1000 sq
1
2
3
4
5
6
7
8
102
187
305
410
490
Taken from
Taken from
Taken from
25 96 605
50 91 60
70 80 80
80 65 110
90 52 130
Figure 12.
Figure 50.
Figure 13.
Percentages assuming 50% rate occurrence - 50%
on average
solids concentration of 166 mg/1.
ft) % Occurrence-1 > 35
526
2.57
43
2.4
33
2.1
24
1.7
19
1 .4
t
Figures this
column apply
to equivalent
volume of
bas in
c
volume occurrence as 100%;
50
121
6.0
1008
5.7
77
5.0
56
4.0
45
3.2
t
Figures
col umn
to mean
80
202
9.9
167
9.4
128
8.3
93
6.7
76
5.4
this
apply
volume
, tons)
of
4
90
404
19.9
333
18.9
256
16.6
187
13.5
151
10.8
of basin
actual solids removed (T)
Net area of existing basin
Percentage
of 50% rate occurrence - 50% volume
occurrence
Tons of solids removed.
50% rate occurrence - 50% volume occurrence.
Notes
One-third small pump
rate.
Mean rate.
One small pump rate.
One large pump rate.
20% of 2450 cfs.
i.e., 10-year design flow.
based
-------
Several selected volumes of overflow and the corresponding
percentages of occurrence as taken from Figure 13 head the
remaining tabulations. Two tabulated predictions of solids
removal are contained in Table XLV for each selected rate
and volume of overflow. The upper value is the relative
solids removed on a weight basis as a percentage based on
a 100 percent value for the 50 percent rate - 50 percent
volume occurrence. The lower value is the actual weight
(tons) of solids predicted to be removed. This analysis
is based upon actual detention times determined from the
hydraulic studies.
The effects of variable design flow rates, variable storm
volumes, and settling basin areas on suspended solids removal
are included in the foregoing analysis.
Several systems appear worthy of further investigation. They
are:
1. Modified MRPS System - This system involves replacement
of one 305 cfs pump with a variable system, installa-
tion of two transverse baffles, construction of a
flocculation system designed for 85 percent removal
of suspended solids at a rate of 102 cfs, and
disinfection.
2. New Plant - This system includes pumping, construction
of a flocculation system with a settling basin designed
for 85 percent removal of suspended solids at a rate
of 490 cfs, and disinfection.
The data used in calculating the net removal efficiency for
each of the two systems described above are given in Tables XLVI
and XLVII. Steady-state conditions are assumed for each
generalized class interval. This factor must be taken into
account in the final design.
DISINFECTION
Selection of a disinfection system for installations such as
the MRPS requires careful consideration of several important
factors.
Firstly, a decision must be made whether to use gas phase
systems, e.g. Cl~ or BrClr, or a liquid phase system involving
NaOCl. Considerable discussion in the literature has con-
cerned the economic merits of Clo vs. NaOCl.
The conversion from the use of Cl? gas to solid NaOCl by ,
the city of New York was discussed by Steffensen and Nash
who concluded: "the total annual costs using liquid chlorine
or sodium hypochlorite are substantially equal. The use
of sodium hypochlorite appears desirable as the difficulties
168
-------
TABLE XLVI
PREDICTED REMOVAL FROM MODIFIED EXISTING BASIN
% Volume
Predi cted
Solids
Overflow
Rate
(cfs)
102
187
305
410
490
>490
Overf 1 ow
Rate
(cfs)
102
187
305
410
490
>490
% Volume x Mean
% at Rate Storm Volume
Removal Shown (10° cu ft)
96
91
80
65
52
30
PREDICTED
0.08
.14
.20
.18
.15
.25
1 .00
TABLE
REMOVALS
0.096
.168
.240
.216
.180
.300
1.200
XLVII
FROM OPTIMIZED
% Volume
% Volume x Mean
% at Rate Storm Volume
Removal Shown (10° cu ft)
96
95
93
91
85
50
0.08
.14
.20
.18
.15
.25
1.00
0.096
.168
.240
.216
.180
.300
1.20
Total Solids
at Rate Shown
(Ibs)
994
1740
2486
2237
1864
3107
12430
BASIN
Total Solids
at Rate Shown
(Ibs)
994
1740
2486
2237
1864
3107
12430
Removed at
Rate Shown
(Ibs)
954
1583
1989
1454
969
932
7881
Predicted
Solids
Removed at
Rate Shown
(Ibs)
954
1653
2312
2036
1584
1553
10092
169
-------
and hazards involved in the delivery and^handling of •
materials are considerably less." Baker stated that
the
in
such comparisons only the cost of a chlorinator vs. a
diaphragm pump is considered and the cost of the entire
feed system (provision for storage, etc.) which for the
NaOCl system can be considerable. He concludes that, "Differ-
ences in chemical costs favor the selection of liquid chlorine
in calculating operational costs, and more than offset capital
expenditure differences in nearly every case when capital
costs are calculated for the entire chemical feed system."
The proximity of inhabited structures to most combined sewage
overflow installations, such as the MRPS, safety dictates
use of the safer NaOCl system for disinfection.
The extent of disinfection to be provided must also be selected.
A kill of 99.99 percent of total coliforms can be accomplished
successfully within the residence time required for flocculation
of the combined sewage at the MRP-S without measurable detriment
to flocculation or clarification of the overflow. A kill
of 99.99 percent was arbitrarily chosen to provide an effluent
concentration averaging about 1000 total coliforms per 100
ml. This level is comparable to standard levels established
in many recreational areas.
The chlorine demand in the Milk River channel has previously
been shown to be extremely high. This condition prevents
utilization of the channel for disinfection by residual
chlorine in the plant effluent even though sufficient contact
time is available before discharge into Lake Saint Clair.
Dilution of plant effluents in large streams or lakes has
a similar effect in reducing the effectiveness of chlorine
residuals in sewage plant effluents.
From these considerations and the results of the disinfection
studies, satisfaction of the average chlorine demand of
the influent combined sewage at the MRPS might be the best
level to attempt disinfection. The chlorine demand did not
vary greatly within a single storm or between distinctly
separate storms. Disinfection with NaOCl at a level of
approximately 10 mg as C12/1 might effect satisfactory total
coliform levels without attendant dangers of residual chlorine
occurring during periods of low chlorine demand.
The final decision to be made is the selection of flows
for optimum treatment. The distribution of flow rates at
the MRPS produces a rapidly diminishing degree of treatment
per dollar of invested capital beyond flow rates approximately
equal to 20 percent of the ten-year design flow (490 cfs).
170
-------
If chlorine feed facilities were provided for an overflow
rate of 490 cfs, approximately 75 percent of the total volume
pumped from the MRPS would contain total coliform levels of
about 1000 per 100 ml or less. The additional 25 percent
of the volume would also be treated with NaOCl up to the
capacity of the 490 cfs system. Effluent coliform counts
in this fraction, however, would vary upward from 1000 per
100 ml, depending on the rate of flow and the resulting
concentration of disinfectant.
Disinfection with NaOCl involves five major cost elements:
land, storage facilities, chemical feed systems including
pumps, chemical costs, and operation and maintenance costs.
Fixed costs are due to land, storage facilities, and the
chemical feed system; costs of chemicals, and operation
and maintenance vary with the degree of treatment. Land
costs are not considered in this report since they vary
considerably with plant location.
The volume of 7.5 percent NaOCl solution required for treatment
of volumes of overflow shown, at a concentration equivalent
to 10 mg/1 as Clp is shown in Table XLVIII. Commercial
bleach (trade %) is diluted 1:1 to provide 7.5 trade percent
available chlorine. This is necessary to maintain maximum
solution effectiveness over long periods. Dilution can
be made with potable water in the tank while providing air
sparging or other mixing techniques.
TABLE XLVIII
7.5% NaOCl REQUIRED TO TREAT VARIOUS VOLUMES AT 10 mg/1
Volume Treated 7.5 Trade % Available
(IP6 cu ft) do Required (gal) % Occurrence
3 1572 75
6 2360 14
9 3146 6
12 3932 3
15 7864 1
It may be most economical to provide underground storage for
approximately 10,000 gallons of 7.5 percent NaOCl solution.
Sufficient disinfectant to treat several large storms and
the capability to treat five to six storm volumes of more
probable occurrence would be provided. Replenishment of
storage would be required about every one to two months
based on average yearly storm frequencies.
171
-------
The disinfection system should be designed to treat storm
overflow rates of from 100 to 490 cfs with NaOCl at a level
of 10 mg as C12/1. This would require a chemical feed system
capable of pumping and dispersing a 7.5 percent NaOCl solution
at rates from 6 to 30 gpm. Two all-titanium or all-Teflon
centrifugal feed pumps capable of delivering from 6 to 20 gpm
appear reasonable when considering flexibility and standby
protection.
An estimate of the installed costs to provide disinfection
for the Milk River Combined Sewer Overflow according to
the system described above is included in Table IL.
TABLE IL
FIXED COSTS OF DISINFECTION
10,000 gal lined storage tank $ 7,500
Two chemical feed pumps 6,000
Controls, meters, piping, and diffusers 3,000
Total Installation Cost $16,500
Overhead, profit, and contingencies (25%) 4,125
Total Cost $20,625
Assuming a life of 20 years and an interest rate of 8%,
the annual cost for disinfection equipment would be $1,796.
Average yearly storm pumping is approximately 767 million
gallons making the fixed costs for disinfection approxi-
mately $0.002/1000 gallons. Assuming an average cost of
NaOCl of $0.012/pound of available Cl?, the total yearly
cost for NaOCl would be about $7,700 Or $0.010/1000 gallons.
Treatment of 75 percent of the flows at the MRPS with NaOCl
to provide an average total coliform concentration of about
1000/1 would cost approximately $0.012/1000 gallons. Installa-
tion cost of disinfection equipment would be approximately
$20,625; chemical costs, less operation and maintenance,
would be approximately $7,700 annually.
SUMMARY
A cost summary showing the fixed and operating costs associated
with the existing station, the proposed modified station
and a new plant, is contained in Table L.
172
-------
TABLE L
SUMMARY OF COST ANALYSIS OF MILK RIVER TREATMENT SYSTEMS
Present
Net SS Installation , .
Design Rate Removed Costs ^_ Flocculatlon Disinfection Total Costs '
(cf s) (%) Fixed Operati rig Fi xed Operating Fi xed Operati ng Fi xed Operati n"g"
Existing - - 4.190.0003 185,000 - - 20,625 7,700 4,210,625 192,700
Plant
Modified „ 102 63 - - 256,320 190,000 20,625 7,700 4,466,945 382,700
Plant1'2
New 490 81 - - 3,775,000 341,000 20,625 7,700 7,327,625 533,700
Plant
Includes present cost of Milk River Pumping Station less cost of the retention basin.
o
Assumed that sludge can be discharged by gravity to the Detroit interceptor.
3Corrected to 1969 dollars by ENR Index Factor (1.73)
Pumping Station 3,532,000
Retention Basin 658,000
6
Does not include cost of additional sludge flushing water.
-------
SECTION 11
ACKNOWLEDGMENTS
While it is impossible to acknowledge all agencies and
organizations who have contributed to the success of
the Milk River Project, special mention must be made of
the following:
PARTICIPATING AGENCIES
Milk River Drainage Board
Wayne County Drainage Commission
SUBCONTRACTORS
Dorr-Oliver, Inc. - Stamford, Connecticut
Pate, Him and Bogue, Inc. - Detroit, Michigan
ORGANIZATIONS PROVIDING
TECHNICAL ASSISTANCE
City of Grosse Pointe Woods, Michigan
Detroit Edison Company
ESSA, Office of the State Climatoligist - Lansing, Michigan
Department of Interior - FWQA, Detroit Program Office
Grosse He, Michigan
Southeast Michigan Council of Governments
tl. S. Coast Guard - Detroit, Michigan
Grateful acknowledgment must also be extended to numerous
individuals, too many to list here, but without whose help
this work could never have been completed.
175
-------
SECTION 12
REFERENCES
INTRODUCTION (3)
1. Problems of Combined Sewer Facilities and Overf1ows-1967,
Report No. WP-20-11, U. S. Department of the Interior,
Federal Water Pollution Control Administration,
Washington, D.C. 1967.
2. A Preliminary Appraisal of the Pollution Effects of
Stormwater and Overflows from Combined Sewer Systems,
U. S. Department of Health, Education and Welfare, Public
Health Service, Washington, D. C., 1965.
3. Pol 1utional Effects of Stormwater and Overflows from
Combined Sewer Systems: A Preliminary Appraisal"^U~! S.
Department of Health, Education and Welfare, Public
Health Service, Washington, D.C., November 1964.
4. Institution of Civil Engineers, Symposium on Storm Sewage
Overf1ows, William Clowes and Sons, Limited, London 1967.
5. Weibel, S. R., R. J. Anderson, and R. L. Woodward, "Urban
Land Runoff as a factor in Storm Pollution," J. Water
Pollution Control Federation, 36_, July 1964, pp. 914-924.
6. Water Pollution Aspects of Urban Runoff, Report No.
WP-20-15, U. S. Department of the Interior, Federal
Water Pollution Control Administration, Washington, D. C.
January 1969.
7. Weibel, S. R., Weidner, R. B., Christiansen, A. G., and
Anderson, R. J., "Characterization, Treatment and
Disposal of Urban Stormwater," paper presented at Third
International Conference on Water Pollution Research,
Munich, Germany, Section 1, Paper No. 15, 1966, p. 15.
DESCRIPTION. HISTORY AND DEVELOPMENT OF THE MILK RIVER DRAINAGE
BASIN (41
1. Pate, Him and Bogue, Inc., "Report on Milk River Drainage
Basin, Pumping Station and Collecting Sewers," for The Dow
Chemical Company, November 1967, FWPCA Contract No. 14-12-9
2. Southeast Michigan Council of Governments, "Population and
Occupied Dwelling Units in the Detroit Region," 1969.
177
-------
BACKGROUND DATA AND ANALYSIS (5)
1. "Hydrology," Part 1, Section 4. Soil Conservation
Service National Engineering Handbook, U. S. Department
of Agriculture (1964).
2. Standard Methods for the Examination of Water and
Hastewater, 12th Ed., American Public Health Assn.,
Inc., 1740 Broadway, New York, N.Y. (1965).
3. Jenkins, David, "A Study of Methods Suitable for the
Analysis and Preservation of Nitrogen Form in an
Estaurine Environment," Report to U. S. Public Health
Service, Region IX, Water Supply and Pollution Control
Division, SEAL Report No. 65-13.
TREATMENT OF COMBINED SEWAGE WITH POLYMERIC FLOCCULANTS (6)
1. "Flocculation of Suspensions of Solids with Organic
Polymers - A Literature Survey," Mineral Processing
Information Note No. J, Warren Spring Laboratory,
Ministry of Technology, Stevenage, Herts, June 1965.
2. Cardwell, P. H., "Adsorption Studies Using a Streaming
Current Detector," J. Col loid Interface Sci . . 22_,
430-7(1966).
3. Priesing, C. P., Wolfe, R. V., Sack, W. A., and Kelman, S.,
"Plant-Scale Polyelectrolyte Treatment of Wastewater Using
Streaming Current Control," J. Water Pollution Control
Fed., 4_1, 1524-32(1969).
4. Black, A. P., Birkner, F. B., and Morgan, J. J.,
"Destabilization of Dilute Clay Suspensions with
Labeled Polymers," J. of American Hater Works Assn.,
ET7, 1547-60 (1965).
5. McCollister, D. D., Oyen, F., and Rowe, V. K.,
"Toxicologic Investigations of Polyacrylamide,"
Toxicology and Applied Pharmacology, 7_, No. 5, 649(1965).
6. Rebhun, M., Narkis, N., and Wachs, A. M., "Effect of Poly-
electrolytes in Conjunction with Bentonic Clay on Contaminants
Removal from Secondary Effluent," Wat. Res. 3. 345-55(1969).
178
-------
LONG-TUBE SEDIMENTATION STUDIES (7)
1. Daniels, S. L., "The Utility of Optical Parameters in
Evaluation of Processes of Flocculation and Sedimentation,"
Chem. Eng. Prog. Sym. Ser. 65 (97), 171-6 (1969).
2. Daniels, S. L., "Differential and Integral Flocculation,"
presented at the Symposium on Chemistry and Applications
of Polyelectrolytes, 158th National Meeting, ACS,
New York, September 10, 1969.
3. Fiedler, R. H., and Fitch, E. B., "Appraising Basin
Performance from Dye Test Results," Sewage and Industrial
Hastes, 31,, 1016-21(1959).
4. O'Connor, D. J., and Eckenfelder, W. W., Jr., "Evaluation
of Laboratory Settling Data for Process Design," Chap. 2-2,
in Biological Treatment of Sewage and Industrial Hastes,
Vol. II. J. McCabe and H. H. Eckenfelder, Jr., Editors,
Reinhold, New York, 1958.
5. Teot, A. S., and Daniels, S. L., "Flocculation of
Negatively Charged Colloids by Inorganic Cations and
Anionic Polyelectrolytes," Env. Sci . Tech. 3, 825-9
(1969).
6. Thirumurthi, D., "A Break-through in the Tracer Studies
of Sedimentation Tanks," J. Hater Pol. Con. Fed. 41,
R405-18 (1969).
7. "1130 Numerical Surface Techniques and Contour Map
Plotting (1130-CX-11X) Programmer's Manual", IBM
Application Program. H20-0357-0, IBM, Technical
Publications Dept., 112 E. Post Road, Hhite Plains,
N.Y., 10601, 1967.
HYDRAULIC MODEL (8)
1. Anderson, N. E., "Design of Final Settling Tanks for
Activated Sludge," Sewage Horks Journal. Vol. XVII.
No. 1, (Jan. 1945).
2. Fitch, E. B., Lutz, H. A., "Feedwells for Density
Stabilization," J. Hater Pollution Control Federation
32^, 147(1960).
3. Fitch, E. B., "Sedimentation Process Fundamentals,"
Trans. Am. Inst. Mining Engrs.. 223. 129(1962).
4. Fitch, E. B., "Flow Path Effect on Sedimentation,"
Sewage ajid Industrial Hastes, 28, 1(1956).
179
-------
5. Fitch, E. B., "The Significance of Detention in
Sedimentation," Sewage and Industrial Hastes, 29,
1123(1957).
6. Hazen, A., "On Sedimentation," Trans. Amer. Soc.
Civil Engr.. ^3, 45(1904).
7. Kincaid, R. G., "Special Design Features of Water Works
Facilities Handling Highly Turbid Waters," Proceedings,
American Society of Civil Engineers 79. Separate No. 309
(Oct. 1953).
8. Oliver, R. H., "Specifying Clarifier Size Based on
Batch Laboratory Tests," Reprint 5204, Dorr-Oliver, Inc.
9. Rankin, R. S., "Increasing the Capacity of Existing
Treatment Plant Facilities," courtesy J. American
Water Works Assn., 47, No. 4(April 1955T
10. Reynolds, E. C., Jr., "Sludge Washing and Thickening
Eliminate Need for Grit Removal," Wastes Engineering
(Dec. 1956).
11. Schlichting, H., Boundary Layer Theory, 1st Ed., McGraw-
Hill, New York (1955), Chapters XIX and XX.
DISINFECTION (9)
1. Baker, R. J., "Characteristics of Chlorine Compounds,"
J. of the Water Pollution Control Federation. 41(3)
482(1969).
2. Dunbar, D. D., and Henry, J.G.F., "Pollution Control
Measures for Stormwater and Combined Sewer Overflows,"
J. of the Water Pollution Control Federation. 38(1) ,
9(1966).
3. Ellis, J. G., and Dvorkovitz, V., "Stable Solid
Disinfectant Compositions," U.S. Patent 2,815,311,
Dec. 3, 1957.
4. Evans, F. L. , et al., "Treatment of Urban Stormwater
Runoff," J. of the Water Pollution Control Federation, 40(5)
R162(1968T
5. Fielding, G. H., et al., "Disinfection of Resistant Spores
with Hypochlorous Acid," paper presented at the ACS
Conference, New York, September 9, 1969,
6. Johannesson, J. K., "Studies of the Action of Monobromamine
on Escherichia Col i," New Zealand J. of Science, 2_,
pp. 499-505(1 959~T7~
180
-------
7. McKee, J. E., et al., "Chemical and Colicidal Effects
of Halogens in Sewage," J. of the Water Pollution Control
Federation. 3_2, (8) 795 (1960).
8. Morris, J. C., "The Chemistry of the pH Factor in Pools
and Its Relation to Reactions with Nitrogenous Substances,"
paper presented at the Water Chemistry Seminar of the
National Swimming Pool Institute, Chicago, 111,
January 11-14, 1964.
9. Pavia, E. H., and Powell, C. J., "Chi orination and
Hypochlorination of Stormwater at New Orleans," paper
presented at 41st Annual Conference of the Water
Pollution Control Federation, Chicago, 111.,
Sept. 23, 1968.
10. Phelps, E. B., Stream Sanitation, John Wiley & Sons, Inc.,
New York, 1944, p. 203.
11. Sawyer, C. N., Chemistry for Sanitary Engineers, McGraw-
Hill Book Co., New York p. 251 (I960).
12. Steffensen, S. W., and Nash, N., "Hypochlorination of
Wastewater Effluents in New York City," J. of the Water
Pollution Control Federation, 39^ (8) 1381 (1967).
13. Wilson, W. S., and Miles, A. A., Principles of Bacteriology
and Immunity, The Williams and Wilkins Co., Baltimore, 1964,
pp. 152-67.
14. Van Hall, C. E., and Stenger, V. A., "An Instrumental
Method for Rapid Determination of Carbonate and Total
Carbon in Solutions," Analytical Chemistry. 39^ 503 (1967).
15. Standard Methods for the Examination of Water and
Wastewater, 12th Ed., American Public Health Assn., Inc.,
1740 Broadway, New York, N.Y. 10019, (1965).
16. Techniques for Microbiological Analysis," Technical
Brochure of the Millipore Filter Corporation, ADM-40(1965).
ECONOMIC ANALYSES (10)
1. Baker, R. J., "Characteristics of Chlorine Compounds,"
J. of the Water Pollution Control Federation, 41(3)
482(1969).
2. Steffensen, S. W., and Nash, N., "Hypochlorination of
Wastewater Effluents in New York City," J. of the Water
Pollution Control Federation, 39 (8) 1381 (1967).
181
-------
SECTION 13
GLOSSARY OF TERMS AND ABBREVIATIONS
GENERAL
°C
cfs
cu ft
ENR
ESSA
°F
fph
fps
ft
FWQA
gal
gpd
gpm
i n
LD50
mg/T
ml
ml/1
mi n
MRPS
nm
rpm
sec
SEMCOG
sq ft
USGS
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
BACKGROUN
BOD
CA
CL
COLI
DEM
FEC
HARD
INR
N
ORG
P
SET
SMPL
SOL
STRM
STRP
SUS
TEM
TOT
TUR
VOL
_
-
-
-
-
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
degrees Centigrade
cubic feet per second
cubic feet
Engineering News Record
Environmental Sciences Service Administration
degrees Fahrenheit
feet per hour
feet per second
feet
Federal Water Quality Administration
gal 1 ons
gal 1 ons per day
gal 1 ons per mi nute
inches
lethal dose, 50% mortality
mi 1 1 ig rams per 1 i ter
mi 1 1 i 1 i ters
m i 1 1 i 1 i t e r s per liter
mi nutes
Milk River Pumping Station
nanometer
revolutions per minute
seconds
Southeastern Michigan Council of Governments
square feet
United States Geographical Survey
D DATA AND ANALYSIS
Biochemical Oxygen Demand
cal cium
chloride or chlorine
col i f o r m
demand
fecal
hardness
inorganic
nitrogen
organi c
phosphorus
settled
sampl e
sol ids
storm
Streptococci
suspended
temperature
total
turbidi ty
volatile or volume
183
-------
LONG-TUBE SEDIMENTATION STUDIES
ADEPL - actual depletion (ft/spl)
DBOT - corrected depth from bottom (ft)
DF - final depth of suspension in LTSD (in)
D(M) - uncorrected depth from bottom (ft)
DO - initial depth of suspension in LTSD (in)
DSUR - corrected depth from surface (ft)
GS - gravimetric solids
GSO - initial average gravimetric solids (mg/1)
L - length dimension
LDSUR - log]0 DSUR
LTSD - long-tube sedimentation device
LTSPL - log,n TSPL
M - depin sampled (1 ,...,P)
N - time sampled (1,...,Q)
OS - optical solids (absorbance x 1000)
OSO - initial average optical solids (absorbance x 1000)
P - total depths sampled
Q - total times sampled
RGS - removal of gravimetric solids
ROS - removal of optical solids
SETR - settling rate (DSUR/TSPL) (ft/min)
T - time dimension
TDEPL - theoretical depletion (ft/spl)
T(N) - uncorrected time (min)
TSP - sampling period at each depth (min)
TSPL - corrected time (1/10 min)
HYDRAULIC ANALYSES AND MODEL STUDIES
c
c.
h
H
L
N
N
t
W
V
Re
Fr
concentration of suspended solids
concentration of suspended solids in feed
fraction of suspended solids not captured at
value of h/t
of suspended solids not captured in
of suspended solids not captured in
selected
fraction
fracti on
basi n
distance
depth of
bas in
1 oca!
below surface in batch test
continuous settling basin
(Use volume/surface area for basin for varying depth)
length along basin discharge weir
Reynolds number
Froude number
settling time in batch test
Width of weir
characteristic velocity of basin (arbitrarily chosen
to be the overflow rate)
density of fluid
viscosity of fluid
184
-------
DISINFECTION
A - absorbance (white light)
A - initial absorbance (white light)
c - concentration of disinfectant
E - kill efficiency
K - empirical constant
K - dissociation constant
n - rate constant
t - contact time
185
-------
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
l_
1
1
1
t
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
BIBLIOGRAPHIC: The DON Chemical Company. Chemical
Treatment of Combined Sewer Overflows.
EPA Publication No. 11023FDB09/70
ABSTRACT: A typical pumping station and settling basin.
chemical treatment with chemical flocculants and
disinfectants are described. Average number of
days of pumping per year (41) is about equal to
average number of days with precipitation *0.2
inches (45).
Twenty-two analyses of consecutive time-weighted
33 storms over a two-year period. Biochemical
oxygen demand and suspended solids decreased after
was relatively constant. The discharge channel
and immediate receiving bay were severely polluted.
Cationic polymeric flocculants and flocculant aids
from combined sewage in the laboratory. Adequate
disinfection of the combined sewage before dis-
charge is possible. Performance of the existing
basin can be improved by the use of staged con-
tinuous pumping at lower rates and the addition of
baffles for improved flow distribution.
14-12-9 between the Environmental Protection Agency.
Water Quality Office, and The Dow Chemical Company.
BIBLIOGRAPHIC: The Dow Chemical Company. Chemical
Treatment of Combined Sewer Overflows.
EPA Publication No. 1 1023FDB09770
ABSTRACT: A typical pumping station and settling basin.
Chemical treatment with chemical flocculants and
days of pumping per year (41) is about equal to
average number of days with precipitation ±.0.2
inches (45).
Twenty-two analyses of consecutive lime-weighted
significantly improved removal of suspended solids
from combined sewage in the laboratory. Adequate
disinfection of the combined sewage before dis-
charge is possible. Performance of the existing
tinuous pumping at lower rates and the addition of
baffles for improved flow distribution.
Water Quality Office, and The Dow Chemical Company.
BIBLIOGRAPHIC: The Dow Chemical Company. Chemical
Treatment of Combined Sewer Overflows.
EPA Publication No. 1 1023FDB09/70
ABSTRACT: A typical pumping station and settling basin.
chemical treatment with chemical flocculants and
days of pumping per year (41) is about equal to
average number of days with precipitation >0.2
inches (45).
Twenty-two analyses of consecutive time-weighted
samples of influent and effluent arc reported for
initial flushing of the sewers; chlorine demand
was relatively constant. The discharge channel
and immediate receiving bay were severely polluted.
significantly Improved removal of suspended solids
from combined sewage in the laboratory. Adequate
disinfection of the combined sewage before dis-
charge is possible. Performance of the existing
basin can be improved by the use of staged con-
tinuous pumping at lower rates and the addition of
baffles for improved flow distribution.
This report was submitted in fulfillment of Contract
14-12-9 between the Environmental Protection Agency.
Water Quality Office, and The Dow Chemical Company.
KEY WORDS
Combined sewage
Cost analysis
Disinfection
Flocculation
Hydraul ic design
Hydrologic data
Overflow
Sedimentation
Settling basins
Sewage
Storm runoff
Water analysis
KEY WORDS
Combined sewage
Cost analysis
Disinfection
Flocculation
Hydraulic design
Hydrologic data
Overflow
Sedimentation
Settling basins
Sewage
Storm runoff
Water analysis
KEY WORDS
Combined sewage
Cost analysis
Disinfection
Flocculation
Hydraul ic design
Hydrologic data
Overflow
Sedimentation
Settling basins
Sewage
Storm runoff
Water analysis
-------
1
5
4cri-M>ian Number
2
Subjci-1 h'ifid S: Group
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
THE DOW CHEMICAL COMPANY, 2020 Dow Center, Midland, Michigan 48640
Title
CHEMICAL TREATMENT OF COMBINED 'SEWER OVERFLOWS
]Q Authors)
The Dow Chemical
Company
16 P*°>ect Designation
Contract No. 14-12-9
11023 FOB
2] Note
22
Citation
23 Descriptors (Starred First)
*Cost analysis, *Disinfection, *F1occulation, *Hydraulic design,
*Hydrologic data, *0verflow, *Sedimentation, *Settling basins, *Sewage,
*Storm runoff, *Water analysis - Benthic fauna, Bottom sediments,
Colloids, Coliforms, Design storm, Great Lakes, Halogens, Precipitation
intensity, Rainfal1-runoff relationships
25
Identitiers (Starred First)
*Combined sewage, *Milk River, *Detroit, *Michigan
27
Abstract
A typical pumping station and settling basin, the characteristics
of combined sewage overflows, and chemical treatment of overflow with
chemical flocculants and disinfectants are described. The average
number of days of pumping per year (41) is about equal to the average
number of days per year having precipitation >.0.2 inches (45).
Twenty-two analyses of consecutive time-weighted samples of influent and
effluent are reported for thirty-three storms over a two-year period.
Biochemical oxygen demand and suspended solids decreased after initial
flushing of the sewers; chlorine demand values were relatively constant.
The discharge channel and the immediate receiving bay were severely
polluted.
Cationic polymeric flocculants and flocculant aids significantly improved
removal of suspended solids from combined sewage in the laboratory.
Adequate disinfection of the combined sewage before discharge
The performance of the existing basin can be improved by
staged continuous pumping at lower rates and the
improved flow distribution.
the
addition of
is possible.
use of
baffles for
This report was submitted in fulfillment of Contract 14-12-9
between the Environmental Protection Agency, Water Quality
Office, and The Dow Chemical Company.
Ab*
tractor
s.
L.
Daniel
s
Institution
The
Dow
Chemical
Compa
ny
WR, 10? (REV JULY 1969)
WRSIC
AATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.5 DEPARTMENT OF THE INTERIOR
WASHINGTON. O. C. 20240
• 6»0: 1Q6Q-359-330
-------
Continued from inside front cover....
11022 — 08/67
11023 — 09/67
11020 --- 12/67
11023 --- 05/68
11031
11030
11020
11020
11020
11020
11020
11023
11020
11020
— 08/68
DNS 01/69
DIM 06/69
DES 06/69
--- 06/69
EXV 07/69
DIG
DPI
DGZ
EKO
11020 —
11024 FKN
08/69
08/69
10/69
10/69
10/69
11/69
11020 DUF 12/69
11000 — 01/70
11020 FKI 01/70
11024 DDK 02/70
11023 FDD 03/70
11024 DMS 05/70
11023 EVO 06/70
11024 — 06/70
Phase I - Feasibility of a Periodic Flushing Systen
for Combined Sewer Cleaning
Demonstrate Feasibility of the Use of Ultrasonic
Filtration in Treating the Overflows from Combined
and/or Storm Sewers
Problems of Combined Sewer Facilities and Overflows,
1967, (WP-2C-11)
Feasibility of a Stabilization-Retention Basin in Lake
Erie at Cleveland, Ohio
The Beneficial Use of Stonn Hater
Mater Pollution Aspects of Urban Runoff, (l.'P-HO-lD)
Improved Sealants for Infiltration Control, (WP-20-1S)
Selected Urban Storm Water Runoff Abstracts, (!!P-20-21)
Sewer Infiltration Reduction by Zone Pumping, (DAST-9)
Strainer/Filter Treatment of Combined Sewer Overflows,
(UP-20-16)
Polymers for Sewer Flow Control, (l'P-20-22)
Rapid-Flow Filter for Sewer Overflows
Design of a Combined Sewer Fluidic Regulator, (DAST-13)
Combined Sewer Separation Using Pressure Sewers, (CRD-4)
Crazed Resin Filtration of Combined Sewer Overflows, (BAST-4)
Storm Pollution and Abatement from Combined Sewer Overflows-
Bucyrus, Ohio, (DAST-32)
Control of Pollution by Underwater Storage
Storn and Combined Sewer Demonstration Projects -
January 1970
Dissolved Air Flotation Treatment of Combined Sewer
Overflows, (V:P-20-17)
Proposed Combined Sewer Control by Electrode Potential
Rotary Vibratory Fine Screening of Combined Sewer
Overflows, (PAST-5)
Engineering Investigation of Sewer Overflow Problem -
Roanoke, Virginia
Microstraining and Disinfection of Combiner' Sewer
Overflows
Combined Sewer Overflow Abatement Technology
------- |