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

-------