v
    WATER POLLUTION CONTROL RESEARCH SERIES 11020 FAQ 03/71
   Dispatching System for  Control
                 of
      Combined Sewer Losses
ENVIRONMENTAL PROTECTION AGENCY • WATER QUALITY OFFICE

-------
                  WATER POLLUTION CONTROL RESEARCH  SERIES

The Water Pollution Control  Research  Reports  describe  the  results  and  progress
in the control  and abatement of pollution of  our Nation's  waters.   They  provide
a central source of information on the research, development  and demonstration
activities of the Water 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 are 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
WRSIC system.

Inquiries pertaining to Uater Pollution Control  Research Reports should  be
directed to the Head, 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 Water Pollution  from Urban  Land  Activity
11022 DMU 07/70        Combined Sewer Regulator Overflow Facilities
11024 EJC 07/70        Selected Urban Storm Water Abstracts,  July  1968 -
                       June  1S70
11020 — 08/70        Combined Sewer Overflow Seminar Papers
11022 DMU 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 - Kingman Lake Project
11024 EXF 08/70        Combined Sewer Overflow Abatement Alternatives  -
                       Washington, D.C.
11023 FD3 09/70        Chemical Treatment of  Combined  Sewer Overflows
11024 FKJ 10/70        In-Sewer Fixed Screening  of  Combined Sewer  Overflows
11024 EJC 10/70        Selected Urban Storm Water 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
                                  To be continued on  inside  back  cover

-------
 DISPATCHING SYSTEM FOR CONTROL OF
           COMBINED SEWER LOSSES
                          by
        Minneapolis-Saint Paul Sanitary District
                        now the
               Metropolitan Sewer Board
                  St. Paul, Minnesota
                        For the
           Environmental Protection Agency
                  Water Quality Office
                Demonstration Grant 11020 FAQ
                       March, 1971
NOTE: On August 1, 1970, the Minneapolis-Saint Paul Sanitary District
      was absorbed by the Metropolitan Sewer Board. The Metropolitan
      Sewer Board carries on all functions of the former Sanitary District.
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price (1.75
                    Stock Number 5601-0103

-------
                EPA Review Notice
This report has been reviewed by the Water
Quality Office, EPA, and approved for publication,
Approval does not signify that the contents
necessarily reflect the views and policies of the
Environmental Protection Agency, nor does mention
of trade names or commercial products constitute
endorsement or recommendation for use.
 mv-"	~
-------
                   ABSTRACT
   Impressive reductions in combined  sewer overflow
pollution of the Mississippi River in Minneapolis and
Saint  Paul  have  been  effected  by a regulator  control
system.  Working entirely within  the limits  of the
existing interceptor sewer  system, and with relatively
minor modifications to selected major combined sewer
regulators,  incidence of overflow was reduced by 66%
and duration of overflow by  88% during most of a
rainfall season. Computer  simulation techniques using
actual  rainfall  data  indicate  that the  amount  of
overflow volume reduction achieved is the equivalent
of a $200 million separation  project. The efficiency of
collection was improved by  about 20% at controlled
regulators.  The  reduction  in  volume of combined
overflow to the  river  is estimated to be between 35%
and 70% during the runoff  season. The unmodified
combined sewer system captured about 65% of the
urban runoff.  Where  modified, the system captured
about 77% of the urban runoff.
   A mathematical model has been prepared that will,
with rain gage data as input, perform rainfall runoff
analysis,  diversion  of  combined  sewer runoff hydro-
graphs at regulators, and  routing of diverted  hydro-
graphs through the interceptor system. This model will
assist in  operation of the  system to retain combined
sewer flows and utilize the maximum flow capacity of
the existing interceptor sewer system. Part II  of this
report describes the model.

   The 1.75 million dollar project includes a computer-
based data acquisition and control system that permits
remote control of  modified combined  sewage regu-
lators. Data from  rain gages,  regulator control devices,
trunk sewers and  interceptors, and river quality moni-
tors  provide  real-time  operating information. Time
varient quality  data  from  key locations in the sewer
system were obtained  by automated analysis  of nu-
merous hourly samples.

   Work  on the  "Dispatching System for Control of
Combined  Sewer  Losses" began  in  1966, and  the
system has been operated since April, 1969.

   This report was submitted in fulfillment of Demon-
stration Grant 11020 FAQ,  supported  in  part by the
Environmental  Protection  Agency,  Water  Quality
Office.
                                                   in

-------
                               TABLE OF CONTENTS-PART I
                                                                                          Page
     Abstract	   iii
     List of Figures	   vii
     List of Tables	    x
  I.   Summary, Findings, and Conclusions	     1
 II.   Introduction  	     9
          Minneapolis-St. Paul Climatology, River System, and Sewer System	     9
          Minneapolis-St. Paul Sewage Works Studies, 1956-1961	   11
          Proposed Regulator Control System  	   12
          MSSD Personnel   	   13
III.   Studies of Sewage Strength During Dry and Wet Weather   	   15
          Sampling Program   	   15
          Automated Analysis  	   16
          Data Handling	   16
IV.   Installation of Gaging System  	   27
          In-Sewer Work by Regulator Crew  	   27
          Contract 664A   	   27
          Rain Gages	   27
 V.   Regulator Modifications   	   31
          Contract 677  	   31
          Post-Construction Evaluation	   34
VI.   Process Control and Communications Equipment	   39
          Processing System   	   39
          Leased Line System   	   41
          Communications System	   43
          Calibration Procedures	   44
          Inplant Data Acquisition System	   44
VII.  River Monitoring	   47
          Site Selection	   47
          Contract 666  	   47
          Installation, Operation, and Maintenance  	   48
          Findings and Conclusions  	   52
VIII. Computer Analysis of Storm Runoff and Mathematical Model of the
          Regulator-Interceptor System	   53
IX.   Operation of the System   	   55
          Data Acquisition Programs  	   55
          Modes of Operation   	   59
X.   System Effectiveness   	   69
          Rainfall Events Analyzed   	   69
          Regulator Performance	   72
          Projected Seasonal Effectiveness  	   75
          1970 Spring Thaw   	   78
          Computer Analysis of Simulated  Rainfall  	   78
          Cost-Effectiveness of the Control System  	   80
XI.   Recommendations	   81
XII.  References 	   85
XIII. Acknowledgements  	   87
     Appendix A	  89
     Appendix B	111

-------
                             TABLE OF CONTENTS - PART II

                                                                                       Page


  I.   Introduction	 123

 II.   General Concept of Model  	 125

III.   Specific Modeling Techniques  	127
     A.   Hydrology Phase  	 127
          1)    Impervious Analysis	 127
          2)    Pervious Analysis  	 128
     B.   Diversion Phase	 129
     C.   Routing Phase	 132
          1)    General Discussion	 132
          2)    Method of Characteristics	132
          3)    Progressive Average Lag Method	 134

IV.   Discussion of Modeling Techniques	 135

V.   Model Performance for Natural Events	 137

VI.   Critique of Model  	   143

References	 145
Appendix A—Operator's Manual  	147
Appendix B-Definition of Model Parameters	 167
Appendix C-Listing of Model (UROM-9)	173
                                            VI

-------
                               LIST OF FIGURES-PART I

Figure                                                                                    Page
 1. Minneapolis St. Paul Control and Monitoring System	   3
 2. Drawing of Typical Control and Monitoring Regulator	   4
 3. Central Station Process Control Equipment  	   5
 4. Performance of Modified Regulators for Simulated One Hour Rains	   6
 5. Utilization of Minneapolis Northwest lnterceptor-1960 and 1969	   7
 6. Routes of the Principal Intercepting Sewers	   10
 7. Utilizatiorrof Joint Interceptor  	   11
 8. Regulator Crew and Vehicle	   12
 9. Regulator Crew Servicing Automatic Sampler	   15
10. Technicon Autoanalyzer used for Analysis of Sewage Samples  	   16
11. Autoanalyzer Data Key Sheet	   18
12. Technicon Autoanalyzer Report of Scaled Values	   19
13. STORET Retrieval of Hourly Sewage Analyses  	   19
14. Machine Plot of Hourly COD Data  	   20
15. Frequency Distributions—COD at Minneapolis Southwest Interceptor	   21
16. Minimum, Maximum, and Mean-24 Hour COD  at Minneapolis Southwest Interceptor	   21
17. Three-Dimensional Plot of 24-Hour COD Data for Treatment Plant Raw Wastewater  	   22
18. Sewage Flow Meter Reading Key Sheet	   22
19. Comparison of Treatment Plant COD Loadings by Month	   23
20. Flow Meter  Recorder Chart Showing Effect of Rainfall on the Minneapolis Northwest Interceptor   23
21. Plant Raw Wastewater COD Loadings-All Hours and Stormf low Hours	   24
22. Sewage Flow and COD Concentration Compared for a Dry Day and a Wet Day	   25
23. Pole-mounted  Cabinet  	   27
24. Transmitting Weighing-type Rain Gage	   28
25. Locations of the System's Nine Transmitting Rain Gages  	   29
26. Location of Control and Monitoring Regulators	   32
27. New Venturi Meter Control Gates at Minneapolis Northwest Interceptor  	   31
28. Hydraulic Power Operated Gate at the Trout Brook Regulator  	   33
29. Inflated Fabridam Diversion Weir at the Trout Brook Regulator   	   33
30. Inflated Fabridam and Dry Weather Outlet at Otis and Marshall Regulator   	   33
31. Inflated Fabridam at St. Peter and Kellogg Regulator   	   34
32. Hydraulic Power Units  	   34
33. Control Valves, Piping, and Blower for Inflatable Dam	   37
34. Digital Equipment Corporation PDP-9 Computer	   39
35. Control Data 853 Disk Drive	   40
36. Potter Line Printer  	   40
37. Interface Cabinet  	   40
38. Schematic Diagram of Outplant Stations and Leased Line System  	   42
39. Calibration of  an Outplant Address using the CHKOUT Program	   44
40. Plant Operator's Logging Equipment	   45
41. River Quality Monitor Analyzer Unit	   48
42. River Quality Monitor Locations  	   49
43. River Quality Monitor Pump Screen  	   48
44. Industrial Molasses River Monitor	   48
45. Chloride Data  from River Monitor on a Thawing Day   	   51
                                             VII

-------
46. Rain Report Scale Factor List	   55
47. Data File as Stored on the Disk  	   56
48. River Quality Monitor Report	   56
49. Regulator Control Report  	   57
50. Interceptor Report  	   57
51. Interceptor Monitoring Stations	   58
52. Rain Report  	   59
53. Use of SCNFRQ Routine to Control Scanning Frequency	   60
54. Fabridam and Gate Control List	   60
55. Gate Control Report  	   61
56. Data Analysis Report	   62
57. Machine Plotted Rain Gage Data for July 1-2, 1970  	   63
58. Machine Plotted Interceptor Sewer Flow Depth Data for July 1-2, 1970	   64
59. Machine Plotted Trunk Sewer Flow Depth for July 1-2, 1970	   64
60. Ogive-1960 Data for Minneapolis East Interceptor	   65
61. Ogive-1970 Data for Minneapolis East Interceptor	   65
62. Assigning Data Addresses to Recorder Pens Using SELREC  	   66
63  Estimated Disposition of Seasonal Runoff on Combined Sewered Areas in
    Minneapolis and St. Paul.                        	   77
64.  Flow Meter Chart-Minneapolis East Interceptor  	   78
65.  Model Predicted Overflow Reduction	   79
66.  Model Data from Simulated  Rainfall   	   80
                                              VIM

-------
                                 LIST OF FIGURES-PART II
Figure                                                                                    Page
 1.   Map of Twin City Area Showing Location  of  Interceptor Sewers, Major Inlets,  Diversion
     Structures, and Rain Gages	123
 2.   Map of Twin City Area Showing Major Watersheds and Thiessen Polygons	127
 3a.  Impervious Loss Function   	128
 3b.  Pervious Loss Function	128
 4a.  Impervious Loss Rate for 1 In./Hr. Rainfall Rate  	128
 4b.  Pervious Loss Rate for 1  In./Hr. Rainfall Rate	 128
 4c.  Combined Low Rate for 1 In./Hr. Rainfall Rate on a Watershed  1/3 Impervious   	129
 5.   Typical Type 1 Diversion Structure	130
 6a.  Typical Type 1 Diversion Rating Curve with Fabridam Up	131
 6b.  Typical Type 1 Diversion Rating Curve with Fabridam Down	131
 7.   Illustration of Flow Clipping Action of Diversion Structure	131
 8.   Forward and Backward Characteristics and Definition of Grid Points	133
 9.   Characteristic at Downstream Boundary  	133
10.   Results  of  Routing Irregular  Hydrograph by the Method of Characteristics and  by the
     Progressive Average Lag Method	133
11.   Progressive Average Lag Routing	134
12a.  Temporal and Areal Distribution of Storm on Day 207 (July 26, 1969)  	137
12b.  Temporal and Areal Distribution of Storm on Day 211 (July 30, 1969)  	137
12c.  Temporal and Areal Distribution of Storm on Day 218 (Aug. 6,  1969)	137
12d.  Temporal and Areal Distribution of Storm on Day 267 (Sept. 24, 1969)	137
13a.  Inlet Hydrograph and Hydrograph Diverted to the Interceptor Sewer at Minnehaha Diversion
     for Event of Day 211  	 138
13b.  Inlet Hydrograph and Hydrograph Diverted to the Interceptor Sewer at Minnehaha Diversion
     for Event of Day 218  	 138
13c.  Inlet Hydrograph and Hydrograph Diverted to the Interceptor Sewer at Minnehaha Diversion
     for Event of Day 267  	138
14a.  Metered and  Model Predicted Flows for Event of Day 207	 139
14b.  Metered and  Model Predicted Flows for Event of Day 211	139
14c.  Metered and  Model Predicted Flows for Event of Day 218	140
14d.  Metered and  Model Predicted Flows for Event of Day 267	140
15a.  Effect of Fabridam on  Model  Predicted Flows Reaching Waste Treatment Plant for Event of
     Day 207	141
15b.  Effect of Fabridam on  Model  Predicted Flows Reaching Waste Treatment Plant for Event of
     Day 211	141
15c.  Effect of Fabridam on  Model  Predicted Flows Reaching Waste Treatment Plant for Event of
     Day 218	141
15d.  Effect of Fabridam on  Model  Predicted Flows Reaching Waste Treatment Plant for Event of
     Day 267	141
                                              IX

-------
                               LIST OF TABLES-PART I

Table                                                                                   Page
 1. Reduction in Incidence  and  Duration  of Regulator  Overflows Accomplished  by  Regulator
    Demonstration Program   	    2
 2. Minneapolis-St. Paul Monthly Precipitation  	    9
 3. USGS Data for Mississippi and Minnesota Rivers through Water Year 1966	    9
 4. Minimum, Mean, and Maximum River Flows, cfs	   10
 5. Interceptors in Minneapolis and St. Paul  	   10
 6. Summary of Sewer Sampling Program  	: .  . .  .   15
 7. Tabulation of Additions and Modifications—Contract 677	   35
 8. Trunk Sewer Diversion Weir Data	   36
 9. Number of Measurement and Control Functions  	   43
10. River Quality Monitor Parameters	   51
11. Mississippi River Discharge	   52
12. Interceptor Level Measurements, Locations, and Sizes	   59
13. Scanning Time Requirements—Minutes	   66
14. Historical Precipitation Data	   70
15. 1969 Monthly Precipitation Data  	   70
16. Total Rainfall at Each Gage-16 1969 Rains	   70
17. Maximum 60 min. Intensity at Each Gage-16 1969 Rains	   71
18. Maximum 30 min. Intensity at Each Gage-16 1969 Rains	   71
19. Total Rainfall at EachGage-11 1970 Rains	   72
20. Maximum 60 minute Rainfalls at Each Gage-11 1970 Rains   	   72
21. Summary of  1969 Overflow Incidence Reduction	   73
22. Summary of  1969 Overflow Duration Reduction	   73
23. Summary of  1970 Overflow Incidence Reduction	   74
24. Summary of  1970 Overflow Duration Reduction	   74
25. Overflow Volumes at Eleven Regulators From Mathematical Model with Actual Rainfall Data. .  .   75
26. Overflow Control—Mathematical Model Prediction versus Real Data Indication	   75
27. Project Effectiveness in Reducing Frequency and Duration of Overflow	   76
28. Precipitation as Snowfall for Winter of 1969-1970  	   78
29. Reduction of Combined Sewer Overflow During 1970 Spring Thaw  	   78
30. Simulated Rainfall Data used for Mathematical Model Analysis	   79

-------
I.  Summary, Findings and Conclusions
Summary
   This project was conceived during the period  be-
tween  1956 and 1960 when the sewer and interceptor
systems of Minneapolis and Saint Paul were intensively
investigated as  part of a study  for the Minneapolis-
Saint Paul Sanitary District on "Expansion of Sewage
Works  in  the  Minneapolis-Saint  Paul Metropolitan
Area."  These  investigations  revealed that the  inter-
ceptor system was not being fully  utilized  for con-
veyance of combined sewage  during periods of rainfall
and  runoff. The  net effect  of the multiple regulator
settings all over the Twin Cities  was that combined
sewage was often needlessly diverted to the Mississippi
River when interceptor capacity was available. More-
over, the  studies revealed that rainfall seldom, if ever,
occurs uniformly over the Twin Cities and that this
rainfall  diversity  could be used to advantage in  re-
ducing or eliminating combined sewage overflows if a
means of controlling a variable diversion regulator were
available.

 ,  In 1965 the Minneapolis-Saint Paul Sanitary District
took  over the  responsibility  for maintenance  and
operation of combined sewage  regulators in Minnea-
polis  and  Saint  Paul from  their  respective  Sewer
Departments in an effort to improve the performance
of the  regulators and reduce combined sewer overflow
pollution  by more  intensive monitoring. A special
Regulator Crew  was organized and equipped for this
purpose.
   After legislation in the Water Quality  Act of 1965
provided  the possibility of obtaining Federal funding
assistance, the Minneapolis-Saint Paul'Sanitary District
was prepared and made an application to the Federal
Water  Pollution  Administration  for  a "Facilities
Demonstration Grant" to demonstrate  an improved
method  for  controlling  the  discharge from  a  sewer
system carrying storm water and sewage. This  appli-
cation was accepted in April, 1966,  and  the Minnea-
polis-Saint Paul  Sanitary District obtained the first of
the FWPCA Demonstration Grants, a sum of $870,750,
to constitute 50% funding for a "Dispatching System
for Control of Combined Sewer Losses." The project
was described in the application as follows:
      "The Cities of Minneapolis and Saint Paul are
   generally served by combined sewer systems.  Over-
   flows  from the  sewer  system  to the Mississippi
   River,  caused by rainfall or thaw, occur at more
   than  ISO points  in  the  system.  The  losses  of
   domestic and industrial wastewater are  estimated to
   be up to six per cent of the annual wastewater flow
   and possibly from  six to  fifteen per  cent of the
   annual average solids and BOD.  The  losses  occur
   mainly  in the  spring, summer and fall. No assess-
   ment of the effect of overflows on  the bacterial
   condition of the river has been made.
i
    The  proposed system envisions extensive  modi-
 fications to diversion structures at approximately 15
 points along the  interceptor system, minor modifi-
 cations at most  of the  remaining diversions,  and
 construction of five new 'Master Diversions' at key
 interceptors, along with  telemetry and  supervisory
 control systems  to  allow monitoring of overflow
 interceptor  utilization. Maximum utilization of the
 interceptor system and treatment facilities will then
 be  possible, considering  rainfall distribution  and
 other factors. The routing or dispatching of mixed
 storm and wastewater flows can be controlled based
 on  locations of storms in the system, time of day
 and other factors.  At diversions with  high  pollu-
 tional load,  priority cah be given  to avoiding losses
 during  run-off.  The fifteen diversions  carried ap-
 proximately 80 per cent of the Sanitary District's
 load in 1960.
    Past  estimates of the  cost of separation of all
 sewer  systems in  both   cities have  ranged from
 $30,000,000 for  a low  degree  of separation to
 $100,000,000* for fairly complete separation which
 precludes such a program in the near future.  The
 proposed system  will considerably reduce losses at a
 nominal cost.
    The   pre-construction   and   post-construction
 evaluation will take full  advantage of modern data
 handling and analysis techniques with development
 of  correlations,  time  series analysis and  mathe-
 matical models. All data  will be in accordance with
 the USPHS  "STORET" system to facilitate use of
 data by other agencies. Real  time data processing
 techniques will be used to  direct the operation of
 the system,  demonstrating  its use in the field of
 pollution abatement, and providing useful data for
 future planning  and operations  of the proposed
 system  or those which might be designed for other
 areas for reduction of  combined  sewer  system
 losses.
    The  proposed project will demonstrate  a new
 technique of instantaneous observation and control
 of  interceptor  system  performance based on ade-
 quate information  to  drastically reduce losses of
 combined wastes. Information gathered in operation
 of  the  system  will provide  a  basis for further
 reduction of losses by using trunk sewers for storage
 and the  facilities constructed will allow such  a
 measure to be attempted. Post-construction  evalua-
 tion will provide information  which will allow the
 method  to  be adapted  to  other large combined
 sewer systems of differing configuration and clima-
 tology.
    Since the majority of losses  of wastes  occurs
 during the recreational season, considerable benefit
 to the Mississippi River, where it passes through the
 populated area, will accrue. "
More recent estimates place the cost for complete separation
at $600,000.000.

-------
     The first year of  the  project was spent in  pre-
  paration  of plans and  specifications for  the major
  contracts involved in the project. Preliminary in-sewer
  work for the gaging  system was also done by Sanitary
  District  personnel. Construction phases of the project
  began  in   1967 and  Regulator Modifications  were
  completed in 1968.
     Figure 1 shows, on a  map of the metropolitan  area
  all the elements of  the system, including the existing
  interceptors. There  are 18  control  and monitoring
  stations  on the regulator-interceptor  system,  nine
  monitoring rain gages and  five river quality monitors.
     Most  components  of the  regulator control  and
  monitoring system are shown  in Figure 2, an artist's
  rendition  of a regulator typical of  those that were
  modified.  The  underground  vault provides a suitable
  location for all  the electronic and motorized equip-
  ment  which   is connected  to  electric  power  and
  telephone circuits. Depth of flow is measured upstream
  from  the  diversion  dam and  downstream  from the
  regulator  gate  in  the  outlet pipe   leading to  the
  interceptor. Ths gate is  operated  by  an  operating
  cylinder and an oil hydraulic  power system in the vault
  and can be remotely controlled. The  diversion dam is
  inflated with air and can also be remotely controlled.
  Water can flow over the inflated dam, which  does not
  completely close off the trunk  sewer. The dam can be
  deflated to afford almost full hydraulic capacity of the
  pipe.
     The central  control station for the  system is located
  in the Administration Building at the Minneapolis-Saint
  Paul Sanitary District Wastewater Treatment Plant. The
  central  part  of the  system  is  the  process control
  equipment, the functional parts of which are shown on
  Figure 3. The same type of equipment has been used to
  control  electric power  systems, refineries,  chemical
  plants,  steel   mills  and many  other processes. In
  addition to accomplishing control and monitoring of
  the outplant system, the system is capable of logging
  treatment plant operating data.


  Findings and Conclusions
     Real operation of the system began in April, 1969.
t During  the 1969 runoff season operation was some-
  what  impeded  by  a  major  flood on the Mississippi
  River and  later by "hardware bugs" that  prevented
  complete acquisition of data. Additionally, the rainfall
  occurring  during 1969 was  considerably  less  than
  normal. Nevertheless,  enough data that could be  used
  for purposes  of evaluation  was  accumulated.  More
  recent data from April and May, 1970, supplements
  that from the initial  operating season.
     The data collected from each regulator allows one
  to  determine  when the depth of flow exceeds the
  height of the Fabridam or any other height, such as the
  height of  the  fixed  weir that was  replaced by the
 Fabridam. Based upon the actual data, tables showing
 performance as measured by reduction in incidence of
 overflow  and by reduction in duration of overflow
 were prepared.  Incidence  and duration of overflow
 were determined for pre-project conditions with fixed
 height diversion weirs and post-project conditions with
 fully inflated Fabridams as the diversion weirs.

   The  results  of these  analyses are summarized  in
Table 1.

                      Table 1
 REDUCTION IN INCIDENCE AND DURATION OF
 REGULATOR OVERFLOWS ACCOMPLISHED BY
    REGULATOR DEMONSTRATION PROGRAM

                        % Reduction   % Reduction
                          in Incidence  in Duration
16 1969 Rains	  65
11 1970 Rains (April and May)  49
Total of 27 Rains	  58
88
87
88
   A reduction of incidence, obviously, could only be
accomplished when the modifications at the regulator,
primarily the Fabridam, allowed complete retention or
diversion of combined sewer runoff that would not
have been possible with the old fixed weirs. Reductions
in duration were calculated  by comparing total regula-
tor hours of duration for the two cases. The duration
figures are high primarily  because  many of the old
weirs were set only  slightly higher  than dry weather
flow, and consequently the trailing  or  falling limb of
the storm hydrograph would cause overflow for several
hours after the bulk  of the storm water had passed.
Interpretation  was used to  prevent distortion of dura-
tion data that could  be caused, and usually a depth
flow within a few inches of  the weir height was used as
the point of overflow cessation.
   For purposes of evaluating performance, 16 rainfall
events  were selected  from  the  1969  rainfall season.
Although more rains occurred during the 1969 rainfall
season, the monitoring system did not acquire data on
all of them, and some data  that  was gathered was lost.
Particularly troublesome were hardware problems in
the disk  handler which were completely corrected on
July 10, 1969.
   The 16  selected  rains do  provide a representative
sampling, since they include 56 per cent of the rainfall
which  occurred during the period between  April 26
and November 16.
   This 1969  data  was supplemented  by data from
eleven rains occurring in April and May of 1970. When
the 27 means of  total rainfall at each  of the system
gages are added, the sum is 15.59 inches, so the data
sample analyzed  is  almost equivalent to  a normal
rainfall season, at least as measured by precipitation.

-------
             INTERCEPTOR  ION I TOR I KG  STATIONS
             MSSD  RIVER QUALITY MONITORS
             MSSO  REMOTE  TRANSMITTING  RAIN GAUGES
             CONTROL  AND  MONITORING  REGULATORS
	l
CO

-------
          UNDERGROUND EQUIPMENT VAUL
POWER OPERATED GATf

-------
 DISC FILE
                                      PDP-9 COMPUTER
                                        I-O TELETYPE
  INTERFACE EQUIP.
Mill   Ml
  7 LEASED LINES
       FROM
OUTPLANT STATIONS
                                                                                        LINE PRINTER
  OPERATORS AND LOGGING
         TELETYPES
                                         TREND RECORDERS
                                                AND
                                         OPERATORS PANEL
 The central part of the system is the process control equipment.  The same type of equip-
 ment has been  used to control electric power systems, refineries, chemical  plants, steel
 mills and many other processes.  This system can control the sewer system  and has also
 been designed to take readings of treatment plant performance.
PLANT CONTROL ROOM
                            Figure 3.  Central Station Process Control Equipment

-------
   Although these analyses certainly indicate that the
control  system has great  potential and has improved
the capture efficiency of the combined sewer-regula-
tor-interceptor system,  it  is desirable to be  able  to
speak in terms of overflow volume kept out of the river
by the system and still entering the river during  large
storms in spite of the system. The mathematical model
of the sewer-regulator-interceptor  system provided  a
convenient method of simulating pre- and post-project
conditions  and was used with data from certain rain
days  to  calculate reduction in volume  of overflow
occuring from the  operation  of  the system.  Total
overflow  volumes  from  11  major  regulators were
reduced  by from 98 to 30  per  cent  during  seven
individual rains by operation of the system according
to this analysis.
   The mathematical  model was also used with simu-
lated  rainfall  data  in  an attempt to  evaluate  the
effectiveness of the control system in terms of rainfall
intensity and  duration. Figure 4  shows how rainfall
intensity effects regulator performance.
           20
   Analysis  of interceptor monitoring data  indicates
that full capacity of the Minneapolis Southwest and
East Interceptors has been utilized on several occasions
during rainstorms. The capacities of the Minneapolis
Northwest  and  Joint  Interceptors  are being utilized
much more fully than in 1960 when previous gaging
data was 'collected.  Evidence suggests that the Joint
Interceptor at the treatment plant has been operated at
near  full   capacity  on  several  occasions.  Figure 5
compares the utilizations of the Minneapolis Northwest
Interceptor during 1969 and 1960.
   A system  tor controlling overflows from combined
sewer regulators has  been installed and operated. Very
significant  reductions  in combined  sewer overflows
have been effected and  documented through use of this
control  and data  acquisition system since the  spring of
1969.  Overflow  control benefits  will  continue  to
accrue  so  long  as  the  system  is  maintained  and
operated.  Moreover, the effectiveness of the system
will increase  as operating experience accumulates and
problems  encountered  are  solved.  The  use of  the
                            .25
                                                                             1.00
                                             .50              .75
                                   ONE HOUR RAINFALL INTENSITY, IN./HR.
                  Figure 4.  Performance of Modified Regulators for Simulated One Hour Rains
                                                                                              1.25

-------
10
                                             MAX.STORM FLOW
                                                 7-8-69 8.25'
                                        MAX. STORM FLOW 5-20-60 6.31
                                      MAX. DRY WEATHER FLOW 1969 4.2'
                                MAX. DRY WEATHER FLOW 4-22-60. 5-13-60 3.3
                                             9'-6" x 9'-6" SECTION
                                                NORTHWEST INTERCEPTOR
                                             AT V/ENTURI METER
MINNEAPOLIS
                                           40                   60
                               % OF TIME DEPTH OR FLOW EQUALS OR EXCEEDS

                     Figure 5.  Utilization of Minneapolis Northwest lnterceptor-1960 and 1969
                                                                                                        100
 operating   rainfall-runoff-diversion-routing   mathe-
 matical model developed for the system when available
 as a  real-time  operating aid, will  also  increase  the
 effectiveness of the overall system.
    Data acquisition and operating experience will make
 it possible  to up-date,  improve, and even expand the
 existing mathematical model. A logical  extension of
 the model,  incorporation of flow  data  feedback,  is
 discussed in Part II of this report.
    The overall flexibility of the computer based system
 will permit operation to meet changing conditions in
 the tributary urban  area.  The data acquired  by  the
 system  will  be of  great value  in planning future
 modifications  of the  existing combined sewer  and
 interceptor system.  The mathematical model  can be
 modified and used to test and evaluate new ideas  and
 alternative solutions to various problems related to the
 combined sewer system and the interceptor system.
    Considering the present concern about the polluting
 effect  of separate  storm sewer  discharges and  the
 increasing body of  evidence  supporting this concern,
 this system has had an effect on  the magnitude of
                     storm related pollution greater than that which would
                     occur from  a  200  million  dollar sewer separation
                     program which would require years to complete.
                       Among the accomplishments and findings of this
                     project can be listed the following:

                       1. The  number of regulator overflow occurrences at
                     controlled  regulators was reduced by 58 per cent. In 27
                     rainfall events 164 overflow discharges were prevented.
                       2. The  total combined sewer overflow hours  at
                     controlled  regulators were reduced  by 88 per cent.  In
                     27 rainfall  events 1036 overflow hours were prevented.
                       3. In  7  rainfall events analyzed for overflow volume
                     at  11 controlled   regulators,  overflow  volumes  for
                     individual  rains were reduced by from 98 to 30  per
                     cent, with a net reduction of 51 per cent. Except for
                     intense storms occurring only two or three times per
                     year,  overflow  volumes  at  these  regulators  can  be
                     reduced  by about 75 per cent or more.
                       4. Estimates place the total annual overflow volume
                     that  can be  saved by the  system at 230  MG.  To
                     duplicate  this  reduction  by  a separation program.

-------
 10,000 combined  sewered  acres  would  have to be
 separated at a probable cost of at least 200 million
 dollars.
   5. Data acquisition and computer analysis indicate
 that  the  runoff  capture efficiency  of  the  combined
 sewer system was improved  by  about 20 per cent at
 controlled regulators. The  reduction  in  volume of
 combined  overflow to the  river  is estimated to be
 between 35 per cent and 70 per cent during the runoff
 season. The  unmodified combined sewer system  cap-
 tured  about 65  per cent of the urban  runoff.  The
 modified  system  captured 77 per cent of urban runoff
 at the modified locations.

   6.  A network  of nine telemetering rain gages have
and  are providing  urban  rainfall  information of a
quantity and quality never before available. Rainfall
diversity over the metropolitan area has been shown to
be very common.
   7.  Monitoring  of the existing interceptor system has
shown that it is now being more effectively utilized for
conveying storm water runoff.
   8.  River quality  monitors along the  urban reaches
of the Mississippi River have shown that the physical
quality of  the  water is  not  appreciably affected  by
storm and combined sewer discharges.
   9.  Analysis of sewage at many points in the sewer
system has  permitted  the  establishment of diversion
priorities  to help reduce  combined  sewer overflow
pollution.
   10. The  control system was shown to have practi-
cally  eliminated overflow from spring thaw runoff at
the controlled regulators.
   11. A  working,  verified,  and calibrated  mathe-
matical   model  of  the  urban  hydrology  and  the
combined sewer regulator-interceptor system has been
prepared  and is available for studies and as a sewer
system operating  aid.  This model  is thoroughly de-
scribed in Part 11 of this report.
   12. The regulator monitoring system  has permitted
continuous surveillance of  the major combined sewer
regulators.


   13. The  program has demonstrated the efficacy and
value of a  computer  basec  control  and monitoring
system for  overall efficiency of data acquisition and
control and  monitoring functions.

-------
 II.  INTRODUCTION

Minneapolis-St. Paul  Climatology, River System,  and
Sewer System.

   Minneapolis and St. Paul, commonly known as the
Twin  Cities,   are  located  side  by  side  along  the
Mississippi River in Eastern Minnesota. The climate is
temperate, with temperature ranges between extremes
of 99°F and  -32°F. Climatological standard normal
(1931-60) precipitation  is 24.78 inches annually. Of
this, 16.07 occurs during the months May  through
September. Table 2 shows the maximum, normal,  and
minimum precipitation in Minneapolis-St. Paul for each
month.
  A recent publication by the University of Minnesota
(1) indicates that series of annual precipitation amounts
are  approximately normally distributed. In  this report
a somewhat longer period of record  was studied for
Minneapolis-St. Paul  than the climatological standard
period, 1931-60. St. Paul data from 1837 to 1933  and
Minneapolis data from   1934 to  1965  provided  a
combined 129 year record. The range in annual total
precipitation  is great, 39.48 inches, between a  maxi-
mum of 49.69 inches in 1849 and a minimum of 10.21
inches in 1910. The  129 year mean annual precipita-
tion is 26.85 inches.
                             The  Mississippi River enters the Twin Cities from
                           the  North  1812 miles above  its mouth.  The river
                           approximately  bisects  the Metropolitan  area, flowing
                           within,  between, or along Minneapolis or St. Paul  for
                           25 miles. The Minnesota River flows into the Missis-
                           sippi River near the middle of its urban reach at River
                           Mile  1799.  Lock and  Dam No. 1 on the Mississippi
                           River is about two miles upstream from the Minnesota
                           River confluence. The Mississippi  River watershed is
                           about 16,000  square  miles.  The corps  of Engineers
                           maintains a 9 ft. navigation channel in the river, with
                           the  head  of navigation being located in  the upper St.
                           Anthony Pool at River Mile 1811.
                             Upstream from St.  Anthony Falls, the Mississippi
                           River  flows through  primarily glacial  drift.  Down-
                           stream  from  St. Anthony  Falls, the  River  flows
                           through a narrow gorge for about 10 miles. Below  the
                           Minnesota River confluence, the Mississippi occupies a
                           portion of a wider valley formed by a large pre-glacial
                           river.
                             Table 3 and Table 4 summarize USGS river data for
                           the  Mississippi  River and Minnesota River. Minnesota
                           River  near  Carver is  36  miles  upstream  from  the
                           Mississippi River. Mississippi River near Anoka is about
                           6 miles upstream from Minneapolis. Mississippi River at
                           St.  Paul  is about  6   miles  downstream  from  the
                           Minnesota River.
                    Jan.
                    Table 2

  MINNEAPOLIS - SAINT PAUL PRECIPITATION
                  (1931 -1960)
Feb.   Mar.   Apr.   May   June   July    Aug.  Sept.   Oct.
                               Nov.   Dec.
MINNEAPOLIS
Maximum
Normal
Minimum

1.65
0.70
0.11

2.66
0.78
0.14

3.37
1.53
0.48

3.53
1.85
0.62

7.87
3.19
0.74

7.80
4.00
1.26

7.10
3.27
0.11

6.60
3.18
0.43

7.53
2.43
0.41

5.64
1.59
0.26

5.15 1.99
1.40 0.86
0.27 0.06
                                               Table 3
                      USGS  DATA FOR MISSISSIPPI AND MINNESOTA RIVERS
                                    THROUGH WATER YEAR 1966
  Drainage area, sq. mi.
  Period of record, years
  Period of record mean
     day, cfs
  Period of record maximum
     day, cfs
  Period of record minimum
     day, cfs
 Mississippi River
    Near Anoka
     19,100
         35

      7,053

     91,000

        586
Minnesota River
  Near Carver
    16,200
        32

     3,174

   117,000

        79
Mississippi River
   at St. Paul
    36,800
        68

    10,080

   171,000

       632

-------
                                                Table 4
                         MINIMUM, MEAN AND MAXIMUM RIVER FLOWS, CFS
                                                          Minnesota River
                                                            Near Carver
Mississippi River
Near Anoka
Calendar
Year
1962 	
1963 	
1964 	
1965 	
1966 	 	
1967 	
1968 	

Mln.
	 1700
	 1750
	 1820
	 2900
	 2200
	 1500
	 1500

Mean
8232
5854
5856
13730
11800
7500
7700

Max.
39400
22100
23700
90300
42200
40600
20600
                                                        Min.
                                                         350
                                                         390
                                                         385
                                                         372
                                                         482
                                                         245
                                                         185
        Mean.
         6028
         2760
         2080
         7782
         3336
         3381
         4217
  Max.
 39400
 14400
 12900
112000
 16000
 19300
 37200
                               Mississippi River
                                 at St. Paul
Min.
2020
2370
2170
3410
2260
1950
1750
Mean
14590
8546
7820
22380
15730
11100
12400
Max.
56200
31300
32300
171000
52200
52100
25500
   Minneapolis, St. Paul, and most of their suburbs are
served  by a single large waste water treatment plant
operated by the Minneapolis-Saint Paul  Sanitary Dis-
trict (MSSD). The extensive interceptor system leading
to a single large treatment plant was feasible because of
favorable  topographic  and  geological  conditions.
Ground elevations in  the entire  region served  permit
the  use  of  gravity  flow.  A  sandstone formation
underlying much of the area  permitted  the construc-
tion of large interceptor sewers by economical tunnel-
ing methods. With the favorable ground elevations and
the deep interceptors, very little pumping is required.
The total area served is 144,000 acres, of which 54,000
acres are in the core cities of Minneapolis and St. Paul.
Total estimated tributary  population is  1,400,000,
with a population equivalent based on BOD of approxi-
mately 2,400,000. Interceptor sewers were constructed
during  1934-1938  by Minneapolis,  St. Paul, and  the
MSSD.  Figure  6 identifies the  major existing inter-
ceptor sewers.  Flow from the three Minneapolis Inter-
ceptors  is  metered  primarily for  purposes of cost
apportionment.
Figure 6.  Routes of the Principal Intercepting Sewers

   Table 5 indicates the section and capacity of the
interceptors at their downstream ends.  Twenty-eight
miles  of branch  interceptors were constructed  by the
two Cities, and a nine mile jointly used interceptor was
constructed by the MSSD. The replacement cost of this
interceptor system is in excess of $100 million, with
the replacement cost of the Joint Interceptor alone in
excess of $25 million.

                     Table 5
         INTERCEPTORS IN MINNEAPOLIS
                AND SAINT PAUL
   Interceptor         Section       Capacity, MGD
Minneapolis NW     9'6" x 9'6"           230
Minneapolis SW     6'0" x 6'0"             75
Minneapolis E      6'0" x 6'0"             85
Miss. River Blvd.    3'9" x 5'10"            40
W. 7th St.          5'0" x 6'0"             56
West Side          3'6" x 5'6"             34
Joint              9'6" x 10'0"          620
                   Double Barrel

   The system was designed to provide for the sewage
flow  for  the year   1970   from   a   population  of
1,400,000. Design capacity  for the  interceptors was
calculated at 55  percent  for sanitary sewage and 45
percent for storm flow.
   Sewage regulators accomplish the diversion of flow
from trunk  sewers into the  interceptor system. They
also control or limit the quantity of combined sewage
which  enters the interceptor during periods of runoff.
About 150 separate regulators and diversion structures
are used  in the two cities. The regulators are of various
types,  float  operated  gates, tipping  gates,  orifices,
leaping weirs, and simple diversions. The original design
of the regulators was  intended to allow some  storm
water  to enter  the  interceptor  system. The 1928
Report of  the  Metropolitan Drainage  Commission
describes  this "first-flush"  of  storm  water  as that
quantity  assumed to be caused by a rainfall intensity of
0.04 inches per  hour. The rate of  runoff was  deter-
mined  by the Rational Method with assumed coeffici-
ents ranging from 0.2 to 0.8.
                                                   10

-------
   Most suburbs are served by separate sewer systems
but  discharge into either combined sewers or inter-
ceptor sewers which carry combined flow.
   Minneapolis  has carried on a  program of  sewer
separation since  1933. Separation  is  limited to  street
catch basins and is not, however,  100% effective. The
separation work  has been accelerated during the past
10 years and  is being continued. As of 1969 the city is "
nominally 75% separated, that is served  to some extent
by separate storm drainage. Saint Paul is still primarily
served by combined sewers, with only about 20% of its
sewered area served by a separate sewer system.

Minneapolis-Saint  Paul  Sewage  Works  Studies,
1956-1961
   In  May   1956, the  Board  of  Trustees  of  the
Minneapolis-Saint  Paul  Sanitary District  authorized  a
five  year program  of research  and development  on
future metropolitan area sewage  works needs. Because
the  existing  interceptor  system  would  certainly  be
utilized  as the backbone of a future expanded system,
engineering investigation was undertaken  to determine
its current utilization. A gaging  program accumulated
flow depth data at  12 locations  during parts of 1958,
1959,  and  1960  to  supplement  the  flow records
available  from  meter stations  on  the Minneapolis
Interceptors and at  the Treatment Plant. Data was also
obtained from about  20 non-recording rain gages read
daily by volunteer observers. Frequency distributions
of recorded flow depths were made and maximum flow
depths  during dry  weather flow periods and during
storm flow periods were noted.
   Figure 7 shows a cumulative frequency distribution
or ogive for  the Joint Interceptor  at Third St. and
      14.
      10
    a.
    Ul
    o
                                        MAXIMUM STORM FLOW 10-23-59   9.0*
                                     MAXIMUM DRY WEATHER FLOW 8-19-59 6.8*
                                             13-X12' BOX SECTION
                          MINNEAPOLIS - SAINT PAUL SANITARY DISTRICT JOINT INTERCEPTOR
                                         AT THIRD ST. t COMMERCIAL ST.
                                % OF TIME DEPTH OP PLOW EQUALS OR EXCEEDS
                                 Figure 7.  Utilization of Joint Interceptor
                                                   11

-------
Commercial St. in St. Paul. This location is about 2.4
miles upstream from the Treatment Plant, and only a
few  inlets to the Joint Interceptor are  downstream
from it. That storm  flow conditions occur only a small
percentage of  time is evident  from the figure. The
volume of  storm water flowing  in the  interceptor
during this small  percentage of time is dependent upon
the  amount  of  rain and  the functioning  of  sewage
regulators. At  this  location  and  at others  in  the
interceptor system it was found that the capacity of
the pipe was far  from being fully utilized even during
heavy rainstorms.
   As part of the studies an  inspection  was made of
each  regulator in the system. About 120  regulators
consisting of float operated gates, tipping  gates,  ori-
fices, leaping weirs, and simple diversions were known
to be in use at that time. Many were of relatively small
size, on 10 inch  and smaller sewers. In general it was
found that the  regulators were not functioning  as
originally  planned. The original design was usually not
conducive to regular maintenance so non-functioning
mechanical  devices  and obstructed  openings were
found. Changing flow conditions and operational prob-
lems  had  also  necessitated  modifications  at  some
regulators. In both cities some regulators were found to
be  permitting  overflow to  the  River  during dry
weather. Poor hydraulic flow  conditions, where velo-
cities and  turbulence were great, caused some of these
instances.  Other regulators were found to be operating
under flow conditions where only a very small amount
of storm flow would initiate overflow. The net effect
of all the  regulator operating conditions was such that
the  interceptors  seldom or  never operated at full
capacity. Conveying  capacity  available for storm flow
was  unused at the same time that overflows would  be
occurring at many regulators.
   It was determined that 15 major regulators diverted
an estimated 80 percent of the total dry  weather flow
arriving at the Treatment Plant. These large regulators
were also  the ones causing most problems and at which
most combined sewer losses occurred. The structures at
the metering stations at the ends of the three Minne-
apolis Interceptors  also act somewhat as regulators,
since the  interceptor can be  relieved to the river by
overflow weirs  and bypass gates. These  15  regulators
and 3 interceptor metering stations were  recognized as
18  key locations for  effecting  control  of  combined
sewer losses.
   Another  recommendation  of  the  report was  that
"All  regulators which  are adjacent to  interceptors
should  be  placed   under   the  control  of a single
authority  since they play the most important  role in
controlling interceptor use and overflows affecting the
river. In  addition to  permitting readjustment of  all
regulators to optimum  settings, central control would
also assist in preventing unnecessary by-passing of raw
sewage, thereby improving river conditions within  the
Cities."
   In 1965 Minneapolis  and St. Paul enacted ordi-
nances which  transferred  to  the  Sanitary District the
control  of and responsibility for the  operation  and
maintenance  of combined storm and  sanitary  sewer
diversions and  regulators  from which overflow can be
conveyed  directly into the Mississippi  River. A regu-
lator crew was established by the Sanitary District to
carry out this newly assigned responsibility. The crew
was equipped with a service truck, necessary tools and
equipment,  and  extensive  safety  equipment.  Their
initial task was to catalog and describe all  regulators
and  make physical  improvements necessary to permit
them to  be  safely  inspected  and maintained  on  a
regular basis.  Figure 8 shows some  of the men with
their truck.  The  ordinances also provided  for  the
establishment,  construction,  maintenance,  and  the
operation of a regulator control system by the Sanitary
District and the payment of costs for the system.
         Figure 8. Regulator Crew and Vehicle


Proposed Regulator Control System
   The analyses of the interceptors and regulators and
rainfall  data  resulted in the proposal of a  regulator
control  system contained in Volume III of the "Report
on the Expansion of Sewage Works in the Minneapolis-
Saint Paul  Metropolitan  Area".  (2) The   proposed
system was described thusly:
      "It is proposed that a comprehensive control
   system be initiated to provide an integrated system
   of regulator operations which will take advantage of
   the uncontrollable factors affecting the system ...
     Any part of the collection system which can  be
   made variable will provide for more efficient use of
  facilities in the future...
     It is  proposed that a system of supervisory
   control  be installed at key  regulators and  other
  points in the system to provide regulator operation
   based on interceptor utilization.
                                                    12

-------
     The operation of regulators at present cannot be
   based  on interceptor level and cannot be varied to
   meet actual conditions...
     The system of present  regulators  has a fixed
   hydraulic capacity with the variations of quantity
   admitted to  the interceptor sewer varying with the
   head at each particular  regulator and sometimes
   with the functioning of a gate con tolled by level in
   the trunk sewer.
     The float regulator shown in old literature can
   now be replaced with modem power operated gates
   controlled by a supervisory system from one control
   point.  Gate  positions,  flows, and levels in  major
   sewers can be telemetered simply and accurately to
   rapidly  provide  a  dispatching  operator with  a
   picture of instantaneous functioning of the inter-
   ceptor system.
     Jt is suggested that such a system would prove
   immediately  useful in  reducing  loss of sanitary
   sewage at regulators and provide for day by day
   adjustments to meet changing conditions.
     Along with the supervisory control system, it  is
   proposed  that  modifications   be  made  to  non-
   automated regulators  to provide better operation
   and simplify maintenance. It  is proposed that a
   system of telemetering be installed to provide data
   on  interceptor  sewer and  trunk sewer levels for
   operation of the supervisory control, determination
   of frequency and quantity  of overflow of sanitary
   sewage in actual operation  and to provide data for
   design of future modifications. . .
     The proposed system will allow full utilization of
   existing  interceptor sewers. It  will also  insure  a
   reduction of the quantity and frequency of sanitary
   sewage overflow to the river. ..
     The central control  of regulators will allow the
   choice by an operator  of points of bypassing raw
   sewage in the system based on  river conditions at
   the time . . .
     The proposed system  could  also provide auto-
   matic  readout of data for analysis, either manually
   or with  computer techniques, to further  improve
   operation of the system ... In addition, should the
   separation programs set forth by the central cities
   not  fully materialize,  the  control system would
   provide  the   best  means of minimizing  loss of
   sanitary  sewage  without  major expenditures for
   further modification of regulators".

   The Water Quality Act  of 1965  included legislation
which  provided  Federal Water Pollution  Control Ad-
ministration matching  grants for demonstration pro-
jects using new or improved methods for controlling
the  discharge  from combined  sewer systems.  The
Sanitary  District, being prepared  as  a  result of the
earlier  studies,  made an  application  for  a Facilities
Demonstration  Grant in April of 1966. The  Sanitary
District obtained the first such grant for its proposed
project  for  installation  and evaluation of "A Dis-
patching System for  Control  of  Combined  Sewer
Losses." The grant amount of $870,750 is 50 percent
of the $1,741,500 total project cost.

MSSD Personnel
   Because of the size and nature of the project, which
involved some new approaches, a separate department
or area  was essentially created by the Sanitary District.
The Project Engineer for the Regulator Demonstration
Program  had  been responsible  for  the  previously
initiated Regulator Crew and their duties. This respon-
sibility  was retained, and, for the  most part, all  other
time was  devoted to  the  Demonstration Program.
During  the period from July, 1967 to June, 1969, the
Project  Engineer had the assistance of another full time
Assistant  Engineer. Since June,  1969,  only one Engi-
neer has  worked  on  the  Project. Since September,
1967, the Project  group has included a Data Analyst-
Logger  Operator who  was  familiar with  the  sewer
system,  having  worked for three months with the
Regulator Crew, and who had data processing training.
A secretary was assigned to the  Demonstration Pro-
gram during most of the Period.
   The  Regulator Crew was utilized for certain  work
and did all of the sewer sampling.  The  crew contained
four men until  the summer of  1968 when week-end
sampling was carried on. Its size was increased to five
at that  time and has remained so because of the added
work load.
   Two chemists from the  MSSD  Laboratory  were
utilized extensively  for operation of  the  Technicon
Autoanalyzer.  They received training  on  the  equip-
ment, did all of the testing preceding the full scale use
of  the  equipment and operated it during the  initial
sampling  periods.  They later trained   part time and
temporary  student  employees  who were used  for
operating the equipment seven days per week during
1968.
    Five temporary student employees were utilized for
Autoanalyzer operations  and other purposes during
1968.  During  the  summer  of  1966  one  student
employee was used; during the summer of  1969 three
student employees were hired. All temporary help were
students at the University of Minnesota, most engineer-
ing, chemistry,  and mathematics majors. Student em-
ployees have performed  most  of the routine  River
Monitor maintenance since the monitors were installed.

   One  of  the  electricians from  the Maintenance
Department of the Sanitary District was assigned  to the
Project in May, 1968, to  be used as needed. He  has
been used between one-half and three-quarter time for
telemetry  maintenance, river  monitor maintenance,
calibration  and other  electrical  maintenance work. A
 pipefitter was employed on the Project full time for
 about three months.
                                                   13

-------
 III.  STUDIES  OF  SEWAGE STRENGTH  DURING
 DRY AND WET WEATHER
   In the sewer systems of Minneapolis and St. Paul
 variations in composition and  quantity of flow are
 greater  than  the  more well  defined variations which
 occur at the Treatment Plant. Because of this continu-
 ous  variability, single  grab  samples  or  composite
 samples  can not describe  the conditions existing
 adequately enough  to assist in real-time  operation.
 Therefore a combined sewer sampling  and  analytical
 program was initiated.
   The  intent of  the sewer sampling program was to
 accumulate a large enough volume of  data  to enable
 statistical methods  to be used  in  describing diurnal
 variations in dry weather sewage composition and the
 effects of short term  transient events, namely rainfall
 and  runoff events.  Knowledge of relative pollutional
 character of waste waters by location and time of day
 is useful in  determining operating procedures from a
 pollution control  standpoint. When rainfall  conditions
 will   not  permit  operation  for  maximum  diversion
 because  of limited  interceptor capacity, local differ-
 ences  in pollutional  strength  of  waste water can
 determine priorities for diversion and bypassing. Data
 from the runoff events  provide  information useful in
 determining the character of combined sewage over-
 flow  and evaluation  of the  magnitude of pollution
 from this source.
6  summarizes the sampling  program,  showing  the
number of days of sampling at each location.
                   Photo By Kent Koberstein
                   Courtesy Minneapolis Sunday Trioune
Figure 9.  Regulator Crew Servicing Automatic Sampler
                                                     Table 6-SUMMARY OF SEWER SAMPLING PROGRAM
 Sampling Program                                         Regulator Demonstration Program Sampling Data
   _„ "                                                       November 1966 to September 1968
   SERCO automatic samplers were used to obtain                                            .
 hourly grab samples. The samplers hold 24 bottles in a                                                 ys
 rack; each bottle is connected to an individual sample
 tube running to the weighted sample head. A portable     £ ant Secondary Effluent	  12
 vacuum pump is used to evacuate the bottles. A clock     £|a^ ^nmary E
 releases a  clamp on one rubber sample tube each hour             '	
 to collect  a  grab  sample. Usually five samplers were          '    ' '	
 used and five different locations sampled at one time.     _
 Each morning, the MSSD's Regulator Crew would visit     _  _    g v  '	  co
                               ,                      ot. "eter « NeiioQQ	  b^
 each sampler, remove the twenty-four marked sample     We$t Seventh |n^       	  43
 bottles, drain the  sample tubes, install a new set of     D.   0 D                                    __
                                                      Rice & Rondo	  55
 bottles, recharge the vacuum and reinstall the sampler     Mjssjssippi River B|vd. lnterceptor	  35
 in thesewer.                                           Otjs & Marsha||  	  33
   Sampling was initiated in  November,  1966, and     Wabash & Cromwell                           11
some sampling was done through October,  1967, with     Wabash & Eustis                              65
a total  of  633 station days sampled at 14  different     Minneapolis East Meter	  77
stations in the sewer and interceptor system. Sampling     Qak Street                                   73
began again in March, 1968, and lasted until September     Second & Main                               67
 15,  1968, with a total of 832 station days  sampled.     31st & Randolph                              61
 Figure 9  shows  members of  the  Regulator Crew     Minneapolis N.W. Meter	  80
withdrawing one of the samplers from a manhole  to     26th & Seabury                               42
remove  the samples and recharge the sampler.             Portland & Washington	  18
   A total of 24 stations were sampled—16 combined      Camden	  61
sewer regulators, the 3 Minneapolis Interceptors where      Minneapolis S.W. Meter	  80
storm flow relief can occur, 2  other St.  Paul Inter-      38th & Edmund 	  47
ceptors, and 3 Treatment Plant locations, the raw flow,     39th & Minnehaha	  41
the primary effluent, and the secondary effluent. Table          TOTAL	         .... 1465
                                                  15

-------
   The Regulator Crew usually devoted the first two
hours of each day to servicing samplers. They were met
at  the  last station serviced  and the samples were
transported  to the MSSD Laboratory. Sample days
usually ran from 8 A.M. to 8 A.M. Somewhat fewer
than  the  maximum  possible 35,160 samples were
collected,  because of occasional equipment problems
leading  to loss  of a  sample.  Sampling was resumed
during August and September, 1969, when the gaging
and monitoring system was operating, but little rainfall
occurred during this period.
   The samplers have  been used frequently within the
plant and by the University of Minnesota on pilot plant
sampling also.


Automated Analysis
   To run analyses on an additional 120 samples daily
was beyond the capabilities of the Sanitary District
Laboratory,  so  automatic  analysis equipment was
acquired.  The  equipment  used  was  the  Technicon
Autoanalyzer, a portion of which  is pictured in Figure
10. Two MSSD  chemists attended  a  special training
course in its use and were able to work out difficulties
encountered in  adapting  the  equipment  for  use  in
analysis of sewage. The operation of the equipment
and its testing prior to sampling have been reported.(3)
   Five  or six analyses  were made on each  sample
collected, at  a  rate  of  30  samples per hour. The
analyses primarily used were chemical oxygen demand,
chloride,  phosphate,  ammonia  nitrogen,  and  total
Kjeldahl  nitrogen, although  the equipment has also
been  used to measure  iron, urea,  nitrite  nitrogen,
nitrate nitrogen,  and ABS. The equipment has also
been used for analysis of treatment plant samples and,
after modification, for analysis of river water samples.
   A  "common standard"  containing all of the com-
ponents  measured  was  developed.  Standards at  5
concentration levels  were  inserted  after  every 24
samples. Daily plant composites for raw  flow, primary
effluent,  and secondary  effluent  were also  run  in
triplicate. The number of discrete samples analyzed ran
up to  159 each day.

Data Handling
   At an average  of 5 analyses per sample,  up  to 795
discrete analyses were made daily. The analysis results
were recorded on continuous strip chart recorders, with
peaks on  a  continuously traced line representing an
optical  density for each sample and standard run.
Calibration curves prepared from the optical densities
observed for  various standard solution concentrations
could have  been used  to manually  convert  sample
optical density to  mg/l. However,  this would have
                  Figure 10. Technicon Autoanalyzer used for Analysis of Sewage Samples
                                                   16

-------
been  time  consuming  because  of  the  number  of
analyses and because standards were repeated between
sets of 24 samples to account for base line variations.
  As  an  alternative to the manual method a machine
processing method  was developed. A  key sheet was
designed  to be used for keypunching  of raw data onto
cards, Figure 11.  Sample  identification and standard
concentrations and optical density values for standards
and samples were read from  the  recorder graphs and
entered onto the key sheet. A program was written to
fit a curve to the standard concentrations and  optical
densities  and to update  the  curve as standards were
repeated  to  account for base line drift. The program
calculates the concentration  values  and  generates a
report of scaled values,Figure 12. A program was also
written to punch  the data  onto cards  in  a  format
acceptable to the STORE! System. The programming
was done for  MSSD  by a programming and systems
consultant  and data  processing computer time was
rented from a data processing center.
   Now after almost two years of sampling there have
been  approximately 200,000 discrete analyses, with
about  150,000 on sewage samples. Over 6,000 samples
and 30,000 analyses were for the  plant raw waste. On
the average  about  1,200 samples and  6,000 analyses
were obtained from each of 21 other locations in the
sewer system.
   Storage  within   the  STORET System was easily
accomplished.  An example of a retrieval report, gen-
erated in Washington by the STORET System is shown
in Figure 13. Hourly sewage strength data for a few days
at one of the control and  monitoring regulators in  St.
Paul  are listed. The  STORET  analysis program was
used to obtain the number of observations, maximum,
minimum,  mean,  and standard  deviation  for  each
analysis at each station by month, by quarter, by year,
and by period of record.
   Variations in concentration  occurring on any par-
ticular day are often large as Figure 14,a machine plot
from  the STORET system, shows. COD concentration
in mg/l is plotted for a regulator  in St. Paul over a 70
day period beginning on March  18, 1968. Wet and dry
weather analysis results are included in the data. There
are up to 24 data points plotted vertically for each day,
and COD concentration  on any  day varies consider-
ably.   Recurring periods  of  higher  plotting  density
indicate the  lower variability of COD on Saturday and
Sunday at this regulator.
   Using  data center facilities and the services of our
programming consultant  the data were  further pro-
cessed. Distributions by hour were made for all  data at
each station. Frequency distributions have been plot-
ted for some hours for each analysis at each station.
The  diurnal variations in  COD shown on Figure  14
were  not found to  be regular, for most hours of the
day  were distributed quite widely.  Some frequency
distributions skewed on rectangular  coordinate paper
plotted  normally on semi-log paper.  Figure 15 is an
example  of one  of these  plots, showing  frequency
distributions of COD data from the Minneapolis S.W.
Interceptor for four hours.
   Plots of maximum, minimum, and mean  have been
prepared on 24-hour graphs to illustrate  diurnal vari-
ations.  These  functions  for COD  data  from  the
Minneapolis S.W. Interceptor are shown in  Figure 16.
The distribution of values within the extremes could be
quite varied, so it is informative to be able to illustrate
both at once. The 24-hour and frequency distribution
plots were combined  in  a  few  instances  to create
three-dimensional plots such as Figure 17, which shows
the range and frequency distribution of hourly treat-
ment plant raw wastewater COD data. Included in the
plot are both  wet and dry weather values.
   Continuously recorded flow information was  also
available so loadings could be calculated. The plant  raw
waste flow and sewage flow from each of the three
Minneapolis Interceptors is metered  and recorded at
the treatment plant on  24-hour circular charts. Hourly
readings from these charts were transferred to specially
prepared  key  sheets.  Figure 18.  Flow data from
November, 1966, through June, 1969, has been punch-
ed onto cards.  Over 180,000 hourly flow values have
been transferred and recorded  in this manner.

   Distributions of plant and interceptor flows by hour
have been made and plotted. Statistical analysis runs
producing  number  of  observations, maximum, mini-
mum, and mean have also been made and 24-hour plots
have been drawn, and  some  three-dimensional plots
have been  prepared. Diurnal flow variations at four
places  in  the system have thus been well  described.
This will also assist in  formulating operational strate-
gies. Flow variations with season have been shown by
distributions by month.
   Loadings in  pounds per hour were  calculated from
the hourly Technicon analysis results and the hourly
flow values and distributed by hour of day, by day of
week, and by  month  for the plant raw  waste. Plant
hourly flow  values used in  calculating  loadings  are
those taken  through the treatment plant. Primarily
because of a deficiency  in grit channels, portions of
high  flows caused by storm  runoff which exceed  safe
limits are bypassed. Bypass flow volumes are  estimated,
but are not included in the loading calculations.
   Comparison  of the  by month frequency distribu-
tions of hourly plant loadings indicates that  there is no
predictable effect of storm flow on plant loading. In
1967 the months of May and August had 0.61 in.  and
2.79 in. of rainfall respectively. In 1968 the same  two
months  had  3.74  in.  and 0.75  in. of rainfall.  The
distribution of  hourly  COD  loadings  for May, 1968,
with over six times the rainfall of May, 1967,  plotted
to the left of May, 1967, indicating lower loadings.  The
distribution for August, 1967, with almost  four times
                                                   17

-------
                                                                      MINNEAPOLIS-SAINT PAUL SANITARY DISTRICT
00
DATE C
DATE A



Tl
CD
'
C
g
3
D)
Kl
CD
a
Bl
CD
co
I
CARD CODE
OH
0,2
0,3
0|1
0|2
Oi3
OLL
NAL
1 RUN NUMBER
0)1
0|1
on
0,2
0|2
01?

0,1
0|2
0 1 3

0 1 1
0|2


0)1
0|2
0|3


OH
0|2
0|3

0


4

0(3
Oi3
ni3

0,4
0|4
0|4


0|5
0|5
0|5

ECT
YZEI
ED
1



STATION
CODE


N


\





/




\







\

'






0)6
0|6
0|6

n




\

i




\






ANALYSI
UJ
o
8



Q
4
U
>
1






































\























,

\

| MONTH
1






















' \

















o





























\


















M
J
^ < 11
0 *" >
•<
C
! 1






















i





























t \

! 1
t
. V
< C
I !
AUTOMATIC ANALYZER LOADING AND RESULT
ANALYST

e
4
«t
3
3
C


>

>


>

/




V

1

v

,



\

t



\

f

/








COMMON STANDARD
OPTICAL DENSITY










COMMON STANDARD
CONCENTRATION
ANALYZER READINGS BY HOUR
1 1 1
1 ,1, 1
i i
- 61
! ,! 1
i
i
i l 1

!

i
1 ' 1
, , ,

1
,

i
,
I !

1

i
1 ll '


!
L,l
i

-H


i
-i_
i

I

1

1

|
01 	

1 1



















I |










1
! 1
1 1









1
1 1 l
l 1 1
	 EU 	
1 1
I l


i i i

i i
i i
i i

l 1 I
! 1B
t < i

i u
-M-1
i i






i i
1 i.
1 1

H-H














1 	







l



























U





L.










1 1
I
i 1
!
1
i 1


1 1
1

1
1
1 1 1

i i
i
i

i

i






-L-L-
-L-L-
! 1
I) J 1


-L-L.
1 1

1 1 1


1 1
I 1

1 1 1






L_1_L_
I 1

1 1
LJ_1_
1 1

1 !

l 1
















C2




1

1

1





J















1














,


1 j 1
-4




i



i j i



u

L-,

M





-.-J-L-














1 1 1
i i l
i i
1 l l
•'.V-1-
1 1 1

1 J '

1 1 1

1 J 1
1 1 1
1 1 1

1 P-
1,1, 1
1 1 1

1 ,1 1
1,1, 1
1 1 1

H-H-

PACE




OF


COMPOSITE
RAW






<;









,







U




























PRI









1
,
1


1
SEC

III !
ill 1
III l
l 1
1 1 1
1 1 1




i
,
1
,
Ml 1
1 1 1
i i l

l J. 1 1


,
1
1
1 1
1 1 1
l l






L-



,
|

1 1
III 1 l
1 1 l l










1
1 1
1 1 1
l l







H-H





u,.u,u.u..


-------
                                                        lUfiOAt    UT        Ofc*

                                                     PPfi	"<*e
                                                    Tl'jPiiMt        CHICHI 01
                                                     P(K             CL
                                                     -C/l         '  "C/l
  — TTTT 	 r


?|-CO -^ '* '-I 1.0
21-00 ^/rt.. _£.?N.C



•
tl




J

ii


_L

* . i i r if
u . *. » i ^ •.

•1 	 r^— #

t .1'*.

I . !M
i .-:;? j






.•'0
"of





X'.l
1 141.0
1 1R.Q
              Figure 12. Technicon Autoanalyzer Report of Scaled Values
     StnREI RETRIEV1L  OtTE  68/12/20
     1214C
                                          26<>cia
                                         IBRUT 60 DDK REGUKTOR
                                         26 MINNESOTA
                          0033^
       0«IE   TIDE DEPTH  COD
       FROM    OF        LOMIEVEL
        TO    D»T  FEET    "»C/l
     6T/06/0<) 01  00
              02  OH
              0)  00
              0*  00
              Of  00
              06  30
              07  00
     67/06/12 09  00
              10  00
              11  00
              12  OC
              13  00
              15  00
              16  00
              17  00
              18  00
              19  00
              20  00
              21  00
              22  00
              23  00
              ?«  00
     67/06/13 01  00
              02  00
              03  00
              0*  00
              OS  00
              06  00
              07  00
     67/06/1* 06  00
              09  00
              10  00
              U  90
              12  00
              13  00
              It  00
              15  OC
              16  00
              17  00
277.1
222.3
2*6.0
229.3
327.0
186.D
330.0
392.3
»26.0
Ml.O
525.0
298.0
458.0
397.0
»95.0
426.0
41*.0
391.0
440.0
293.0
197.0
216.3
351.0
J57.0
317.0
487.0
244.0
251.0
295.0
273.3
362.0
37S.O
416.0
355.0
324.0
         03610
       AfMQNK
        NH3-N
13.710
12.300
10.600
 9.730
 B.400
 7.50C
 8.300
 J.600
 5.50C
 a. 200
11.500
11. COO
 9.ZOC
 8.800
 5. 300
 7.600
 9.100
 9.400
 5.300
 9.0CO
10.00C
 9.400
13.100
 a. ceo
 5.70C
 5 . eoc
 4.70"
 5.400
 3.100
 9.500
11.200
 9. SCO
11.500
12.30C
15.100
12.30C
11.100
13.200
11.20P
  "0653
SOLP04-T
  P04
  MC/L

   15.00
   15.30
   10.80
   10.00
    7.90
    7.00
    7.no
    4.6C
    7.10
   12.00
   13.20
   21.50
   26.80
   21.40
   12.70
   19.70
   22.50
   21.90
    8. 31
   20.50
   19.On
   17.50
   17.60
   12.60
    B.30
    8.50
    7.70
    8.20
    4.-SO
   14.00
   11.10
   11.20
   15.5C
   11.7r
   21.80
   16.50
   n.or
   21.BO
   14.3C
  00940
CHLORIDE
   CL
  "0/1

      65
      61
      65
      77
      70
      54
      85

      47
      85
     102
      96
      77
      89
      55
      70
      66
      62
      51
      71
      64
      58
      65
      61
      58
      62
      64
      78

      73
      98
      67
      95
      «2
     170
      84
      95
      68
      92
                                                                     SAINT  PAUL
                                                                     I11CR050
                  Figure 13. STORE! Retrieval of Hourly Sewage Analyses

                                                19

-------
           2G0010

          PHOLEAI CFCEK fiECULOTOn

          ZS  MrNNELSdTfl


          UPPER  MISSISSIPPI

          5flIIMT  FflUL
o
ED
o
in
on
on
O
a
       o
       o
       r>
       o
       a
       o
       o
       o
       o
       o
       CM
       O
       o

       d
       a
       LO
o
a
o
       o
       o
       o
       o
       o  (j
            LO
30
                                                              SO
                                                                 60
70
                                                                                             80
           STRRTING DRIF; rs

           TIME  DflYS
                        G8  3    18
                             Figure 14. Machine Plot of Hourly COD Data
                                               20

-------
                       COD ANALYSIS S.W. INTERCEPTOR-August - September '68
 40
     600
Parts Per Million
                                                                                              1000
                 Figure 15. Frequency Oistributions-COD at Minneapolis Southwest Interceptor
                    MPLS. S.W. METER-August - September '68-25-30 Observations/Hr.
1000
 900
   12  1   2   3456   7   8   9  10  11  12   1   23   45    6    1   8   9   10   11  12
           Figure 16. Minimum, Maximum, and Mean-24 Hour COD at Minneapolis Southwest Interceptor
                                               21

-------
                                                         NOTE:  Data  includes wet and dry weather
                                                                flows. Flow data available to calculate
                                                                pounds per hour.
                                                                                           Number of
                                                                                           occurrences
Figure 17. Three-Dimensional Plot of 24-Hour COD Data for Treatment Plant Raw Wastewater
                    MINNEAPOLIS - SAINT PAUL SANITARY DISTRICT
D»TEOF MADIIM:
LOCATION
Hi
§
0
5
1
2
1
2
1
2
I
2




1





1
2
1
1
1
2
1
?
' '








' *
STATION NO
,















44-


'

ANALYUS
~
:































1

o
Z Ml
< <
u O
•
O 0


















0»Y OF TEAR
	 1... At' 	 	
















44.

£
1
L
3
3


















SEWAGE FLOW METER

1 2








1



















2





|
1





345








3



1 1
| i
, j













4



1

j ,

, ,









S


: 1






READING
READINGS BY HOUR
678








6



















7




,


I i











i











9 10 11








9



















10


1


f
,




, .







11


1 1
, .







rtst
12
[
. ,

.




12

.
















































                     Figure 18.  Sewage Flow Meter Reading Key Sheet
                                        22

-------
the rainfall  of  August,  1968, plotted  to the right,
indicating heavier  loadings. The months of  June in
1967 and 1968 had 7.53 in. and 6.78 in. respectively
and the frequency distributions nearly coincide. Figure
19, shows the comparison of COD loadings  for May
and August of 1967 with 1968.
          COD-PLANT RAW WASTEWATER,
 180
 160
           20
                                        100
                 40     60      80
                   1000's LBS./HR.
   MAY 1967 ••••     AUGUST 1967 •
   MAY 1968	     AUGUST 1968— _ .
Figure 19. Comparison  of  Treatment  Plant
          Loadings by Month.
                                                120
                                              COD
   The effect of rainfall and runoff on the combined
sewer system is readily observable on  the flow meter
charts for Minneapolis Meters.  Figure  20, shows the
effect on the Minneapolis N.W. Interceptor of 0.26 in.
of rain which occurred on the evening of September
24,  1969. The time during which  combined runoff
water and sewage  is flowing through the meters can
easily be identified and punched onto cards. The card
format used contains a card code number, the. station
coding, the year, month, day, and the hour combined
sewage flow  starts and the day and  hour it stops. With
this  information, a separation of wet weather Techni-
con loading data at the treatment plant has been made
and distributions made on the wet weather data.  The
same has been  done at the Minneapolis  Meters and
could be done for concentration data at all  the other
regulators where sampling was carried on.
   The frequency  distribution of  274 hourly  COD
loading values representing storm flow  periods at the
Treatment Plant is  compared in Figure 21, with the
frequency distribution  of 4138 hourly COD  loading
values representing both dry weather and storm flow
periods. In general, most of the COD plant loadings fall
within the same usual range during  both dry weather
and  storms.  Similar dry weather-storm comparisons
were  made  for  calculated  hourly  plant  loadings of
Total Kjeldahl Nitrogen, Total Soluble  Phosphate, and
Chloride  ion. The results were similar to those seen for
COD. The data representing storm flow almost always
lay within the extremes of the dry weather data. The
frequency distributions  for  wet  and dry  weather
Kjeldahl  Nitrogen  loadings were very similar.  Those
comparing wet and  dry weather phosphate and chlor-
ide loadings suggested that during storm periods some
dilution of those constituents occurs.
                                                              Figure 20.  Flow Meter Recorder Chart
                                                                         Showing Effect of Rainfall on
                                                                         the Minneapolis Northwest
                                                                         Interceptor
                                                   23

-------
              1000
 ALL HOURS:
 Observ.   4138
 Interval
 Max
 Min.
 Mean
 S.D.
 10.0
176.88*
  1.08
 43.56
 17.96
800
 RAIN HOURS:
 Observ.   274
 Interval   10.0
 Max
 Min
 Mean
       600
 s.n.
176.88*
 14.07
 43.78
 17.51
'This value is not
 significant and
 should have been
 thrown out.
               400
               200
                                                FREQUENCY DISTRIBUTION
                                   NO. OF OBSERVATIONS ALL HOURS
                                                  NO. OF OBSERVATIONS RAIN HOURS
                                                                                                     100
                                                                                      80
                                                                                              60
                                                                                                     40
                                                                                                     20
                         15.00   30.00
                                 45.00    60.00   75.00   90.00   105.00  120.00  135.00  150.00
                                 COD LOAD, 1000's of pounds per hour.
               Figure 21. Plant Raw Wastewater COD Loadings-All Hours and Stormflow Hours
   Concentration data for some specific rainfall events
 were reviewed also. An example of some of this data is
 presented  in Figure 22.- A rainfall amount of 0.67 in.
 occurring over a five hour period  was measured at the
 Weather Bureau gage in South Minneapolis on Septem-
 ber  8,  1968,  a Sunday.  The flow data and COD
 concentration data at the Minneapolis N.W. and S.W.
 Meters for this day were compared  to those occurring
 on a dry Sunday two weeks earlier. The concentration
 values are  unaffected by the storm water in the system.
 Similar studies of other concentration data as well as
 COD data  for other rains at other sampling stations was
 inconclusive.
   Very likely what is measured may be a combination
 of three possibilities. In some parts of the sewer system
 storm  water is  probably  diluting wastewater while
 elsewhere  a  first  flush of  sewer  deposits may be
 occurring  and still elsewhere  the stormwater  may be
 polluted about  as  much  as the  wastewater it mixes
 with.  This would especially seem likely at the Treat-
 ment  Plant where the arriving flow is such a composite
                                               of wastewater and stormwater from such a large area.
                                                  Because  the  net  effect  of  stormwater on  the
                                               strength of the combined flow at the treatment plant is
                                               small, one must  conclude  that combined  sewer over-
                                               flows can  affect the  river.  At  least  some of  the
                                               combined sewer overflows must be as polluted or more
                                               polluted than the dry weather flow that is usually
                                               diverted to the interceptor.

                                                 Some of the large mass of data accumulated on hour-
                                               ly wastewater strength and flow have been analyzed. All
                                               the data are,  however,  stored  on magnetic  tape and
                                               punched  cards  and  can  be  analyzed using data
                                               processing techniques. Most of the data is also stored
                                               with the FWQA STORET system. Hopefully important
                                               use of this data is yet to come.

                                                  One  immediate use of the data by the MSSD has
                                               been in  establishing bypass priorities for use  in system
                                               operation and flow routing when  storm flow  condi-
                                               tions are such that interceptor capacity is limited.
                                                    24

-------
       AUG. 25 I	II  (DRY DAY)
                                            SEPT. 8
                                                              (RAIN DAY)
   100
    80
 Q
 U
    60
    40
    20
     0
 HOOO
cc

Ul
o
o
o
o
o
o
   800
   600
   400
   200
     0
   100
    80
 O

 5
 g~ 60
 O
 u.
    40
    20
     0
 ,1000
Z  800
O
H
a  600
Z
UJ
u
I  400
U
o
8  200
               SW
       111****
               SW
               NW
               NW
       Illlllllll
                   METER FLOWS
                     fill**1
                   METER COD's
O 111 miniiiimiiiiiiiiiiii I****"
                   METER FLOWS
                   METER COD's
                                              •.,.•:.,..•:,,.••.,..
                                              ;';•;••:; 0.67" i;V:X
                                                   mini
                                                          linn
                                                 """•iiiiiiiiiiiiinii
                                                                        iiiiiniiiiii
                                                                                     "IIIIIIIIIII
                                                                                    iim
                                                                                         ilium
                                                                                                 .25 «
                                                                                                     u.


                                                                                                 .50 I
     10           12            14           16            18            20           22
                                                 TIME
         Figure 22. Sewage Flow and COD Concentration Compared for a Dry Day and a Wet Day
                                                                                               24
                                                25

-------
IV.  INSTALLATION OF GAGING SYSTEM
In Sewer Work by Regulator Crew
   The Sanitary District's consultants had, as part of
earlier studies,  successfully  used  a  gaging system
employing a pressure-conveying tube installed in the
sewer, an air-supply cylinder,  a  bubbler and a local
recorder. A  similar approach was  planned for use with
the Regulator Demonstration  Program system with a
pressure  transducer and telemetry replacing the local
chart recorder. The installation of bubbler  elements
and  pressure transmitting tubing in the sewers was
accomplished entirely by the Sanitary District's Regu-
lator  Crew.  Much of the work was done at night to
take advantage of low flow depths, making possible the
installation of most  air discharge elements at or near
the sewer invert. Originally, all the elements were made
from  1/4" flexible copper tubing with several rows of
holes spaced at 90° intervals to  provide multiple air
discharge ports. Some of these  have since been replaced
with galvanized pipe elements.
   Forty elements were installed at regulators, usually
with  three  per regulator:  one  upstream  from the
regulator in  the trunk sewer, one downstream from the
regulator  in the outlet pipe carrying diverted flow to
the interceptor and one downstream from the diversion
dam in the storm water overflow  pipe. Eight elements
were  installed in the interceptors. Existing  level moni-
toring elements were already in use at three  interceptor
locations  and three  regulator  locations.  All  elements
were  terminated in a safe location in a  manhole well
above maximum flow depths and  1/4" I.D.  plastic
tubing used  to transmit air pressure to the surface. The
Regulator Crew has since maintained the tubing and
elements,   making  repairs  and  replacements  as
necessary.

Contract 664A
   In order to avoid  expensive construction in existing
congested urban  locations and to  provide a  "clean"
location  for  electrical  and  mechanical equipment,
pre-cast concrete vaults, identical  to those used by the
power and telephone industries, were specified as part
of Contract  664A. Fourteen vaults were  installed near
regulators where revisions were to be made for control
and monitoring.
   Seven  pole-mounted  cabinets (Figure 23)   were
installed  near interceptor monitoring stations where
less space for equipment  was  required. One vault was
used  for interceptor monitoring equipment to serve a
station  located  within a  city  park where an  above-
ground cabinet was aesthetically undesirable.
   The concept of a sealed "umbilical" connection to
the hazardous  environment of the sewer  was devel-
oped. The facilities entering the sewer atmosphere were
restricted  to non-hazardous low  pressure air, oil hy-
draulic and low voltage direct current (five volts at very
low current).
   Contract 664A also included the electrical service,
vault to stub pole telephone wires and bubbler panels
for flow depth measurement.  The contract was com-
pleted during the summer of 1967. Early installation of
electric  services  facilitated later work  under  other
contracts  and  toy  Sanitary  District personnel. The
bubbler cabinets contained a high-pressure air  mani-
fold, pressure reducing valve and bubbler, an electric
24-hour clock and a purge timer and a solenoid valve to
switch  high-pressure purge air  into the tubing to purge
the  element once every  24  hours.  Purge  time is
adjustable  and  a  10  second  purge has been  used
satisfactorily. Originally compressed air cylinders were
used to supply bubbler air, but these were all replaced
by  small compressor-tank outfits. This has eliminated
the problem of changing heavy cylinders periodically in
poorly accessible locations.

Rain Gages
   The  original  rain gage network  was comprised of
eight rain gages; a ninth rain gage was added during the
summer  of 1969.  The  rain  gages  used  are Belfort
Model  5915 Rainfall Transmitters.  The  principle of
operation is weight measurement, with the scale  pan
movement transmitted mechanically to  a single turn
precision potentiometer. The power supply for potenti-
ometer  excitation  is  supplied  from  the  telemetry
equipment  and  the voltage  analog  proportional  to
rainfall  equivalent weight is transmitted to the central
station. One of the gages is shown in Figure 24 with
collector and case removed to show the collector pail
and weighing mechanism.
           Figure 23. Pole-mounted Cabinet
                                                    27

-------
   Figure 24. Transmitting Weighing-type Rain Gage

   Seven of  the eight  original  rain gages constitute
separate  remote stations. The eighth original gage is
connected  to  a river monitor  station  interrogation
matrix and did not require a separate station telemetry
installation.
   Rain gage locations were  chosen by the project
staff. Initial  letters  requesting cooperation regarding
installation of  rain gages and telemetry equipment at
various municipal sites were sent during the summer of
1966. Among the requirements  for a site were: (1) a
suitable rain  gage location where representative rainfall
measurement could  be made, (2) an indoor location
within reasonable distance  from the gage site where
telemetry equipment could be installed and connected
to  110  volt  AC power  and telephone wires, (3)  the
capability of installing wires between the gage and the
telemetry equipment, and (4) a location where vandal-
ism  would  not be anticipated.  In general,  locations
around the periphery of the Twin Cities were chosen so
that storm fronts could be observed as they moved
toward  combined sewered areas. Surprisingly, it was
quite difficult to find sites meeting the general require-
ments on publicly owned facilities. An attempt was
made to  site  a gage on the  plant  grounds of large
corporation,  but  approval could  never  be  obtained
from the  various corporate departments  and  the
attempt  was abandoned. The  locations finally used
were the  following:  Minneapolis  Water  Treatment
Plant, St. Paul Water Department Reservoir, St. Paul
Water  Department Pumping  Station,  University  of
Minnesota Agricultural Plots, Suburban City Halls (2),
and  Metropolitan Airports  Commission Maintenance
Garage.
   The ninth gage was  installed  at a regulator control
and  monitoring station and connected to the spare
telemetry address there. This location is near the center
of the Twin Cities. The locations of the nine rain gages
are shown on Figure 25.
   Installations of the eight initial gages were made by
a local electrical contractor. Direct burial cable  and
conduit were  used  in connecting the  rain gage to the
building  in which  the telemetry equipment was lo-
cated. At one site  direct burial cable was replaced by
buried conduit after gophers chewed the insulation off
the wires. The ninth gage was installed by the Sanitary
District electricians.
  The gages seemed susceptible to lightning damage
and  several  of the precision  potentiometers were
damaged  and  needed  replacement.  Several  standard
replacement  potentiometers,   which  were  specially
wound for Belforte, were purchased  at a cost of about
$90 each. Later, a stock item potentiometer that could
be used  with slight modifications was found to  be
available for about  $20, with no noticeable difference
in precision or linearity. Fuses were installed on all rain
gage  circuits and several fuses subsequently have blown
while no  damage to the potentiometer occurred, so
they may  be a worthwhile addition.
   A  range of 0 to 5 volts was equivalent to  0 to 6
inches of  precipitation. Transmission "noise" appeared
to be causing variations in scaled readings of as much as
±  0.02  inches.  A change  of  0.04   inches  over  a
five-minute interval  would be equivalent to a  rainfall
rate  of 0.48  inches/hour, so this noise could produce
rainfall data unacceptable to the mathematical  model.
Repeatability  tests on the telemetry equipment  indi-
cated  that its performance was acceptable  and  that
improvements could not  likely  be made. Noise could
be introduced by induction in the cable connecting the
rain  gage or by the leased telephone lines. Improve-
ments were made  by changing the  transmission  sam-
pling procedure used at the central station to eliminate
the effect of any transient  noise. Now 16 readings of
the A/D  output are averaged.  If any  of the readings
differ by more than one per cent from the first reading,
a new set  of 16 readings are taken.
   To further insure accurate rainfall data the collector
area  of the rain gages was increased by a factor of three
by installation of a rain amplifier.  Each increment of
rainfall therefore became equivalent to three times as
large  a voltage  increment  as  before  and  any noise
reduced in significance  by 1/3. The rain gage  data  is
now   accurate to the nearest 0.01 inch.  The  rainfall
collector buckets must be emptied more often, but this
has not  been  intolerable, and a method of  remote
control emptying will likely be devised for future use.
  The  use of  remote  transmitting  rain  gages  has
elicited interest from meteorologists in the area.  The
State  Climatologist, along with University of Minnesota
scientists,  is  studying the micro-meteorology  of  the
metropolitan area,  and we are supplying him with data
from  the rain gage system. Data has also been supplied
to the public works departments of  the suburbs in
which gages are installed.
                                                    28

-------
NJ
(O
(D
ISJ
CJ1
                  NSSD REMOTE TRANSMITTING RAIN GAUGES
              UM     ST. PAUL CAMPUS
              VAD    LAKE VAONAIS
              HAZ    HAZEL PARK
              FIR    WEST SIDE
              VIL    GOLDEN VALLEY
              EON    EOINA
              MAC    AIRPORT
              WTP    MPLS.  WATER WORKS
              EUS    EUSTIS
        cc
        CO
        CD
        H
        3
        3
        Co

        I
        3D
        BJ
         -r   .^v
         r4-'  •'    '/        L ~   r">>*    I  T'
Vadnais    7-''     /        -T   jBireh- \ -^
Heights    IGem Lake/ White Bear P^^iwtJOdV I r
        A   /   Lake    ,   h      '-r
        /  1	If,            ^   '
     H Htop  I
    ' "1      f"]
ITPcoiutnb,.  J-rl

-------
V.  REGULATOR MODIFICATIONS
   The largest design and construction job involved in
the project was  the  revisions to the  regulators. In
addition  to  converting  existing gates  to  remotely
controlled  power operation  and  installation  of  a
variable position  diversion  device,  there were  some
modifications  to  existing concrete structures to im-
prove flow conditions. Before plans and specifications
were  completed,  the Regulator  Crew  assisted the
consultant's engineer  and  a  testing   laboratory in
obtaining measurements of the forces required to  move
existing gates. Field measurements were taken and all
the sites were inspected.
 Contract 677
    Contract 677 covered  modifications  to  existing
 regulators and  sewer structures. Work done included
 demolition of some existing structures and removal of
 existing  equipment,  concrete  construction  of  new
 sewer structures and constructing and providing regu-
 lator gates and  adjustable diversion devices with neces-
 sary fluid power systems for their operation.  The
 adjustable  diversion devices  at all  locations  where
 required  were  inflatable dams under the base  bid.
 Alternate bids  were  permitted  for the  adjustable
 diversion devices and either fabricated aluminum crest-
 type gates or wood fabricated rotating vane-type gates
 could have been bid.  No bid on either alternate was
 received.  Nineteen  locations  were  involved  in  the
 work—eight in  St. Paul and  eleven in  Minneapolis.
 Figure 26 shows the locations of modified regulators.

    At the Camden regulator,  (CAM), a side overflow
 weir was raised 1.1 feet and a new gate structure was
 constructed at the entrance to the outlet pipe leading
 to  the interceptor. Two new fabricated wooden gates
 with oil  hydraulic cyclinder operators were installed in
 the new chamber on the dry weather outlet and on the
 storm water  bypass.  Because  the sewers have little
 cover there, an  above-ground  structure was necessary
 to  enclose the hydraulic gate operators. The location is
 on  Minneapolis Park Board property so an attractive
 structure was designed and built.

    At Portland  and  Washington and at  Second  and
 Main, (PW and M2), the entire interceptors are diverted
 by  master regulators. At these two  locations and at
 26th and Seabury, 39th  and Minnehaha, 38th  and
 Edmund, and Oak Street (SEA, M39, E38 and  OAK,
 respectively), hydraulic operators were installed on the
 regulator gates  and  a  rubberized fabric  "Fabridam"
 replaced the existing weir.
    At 31st Avenue N.E. and  Randolph  Street N.E.
 (RAN),  a deformed  corregated metal pipe had been
 placed in the invert of the existing trunk sewer as part
 of  a Minneapolis sewer separation project. A Fabridam
 was installed  here anyway  since, under storm condi-
 tions, combined flow exceeding the capacity of the
 corrugated  metal  pipe  occurred,  and  higher  head
 conditions would increase diversion capacity.
    Flow from  each of the Minneapolis Interceptors is
 metered, primarily for purposes of cost apportionment.
 At all  three interceptor  meter stations,  flow  is mea-
 sured by parallel Venturi meters. Two meters are in use
 at all times at  the Southwest and East Interceptors. At
 the Northwest, prior to the Regulator Demonstration
 Program, only  one Venturi meter was used. At each of
 the three structures overflow  relief  to  the  River  is
 provided as well as by-pass gates. There were gates  in
 the channels upstream from the Venturi meters at the
 Southwest and East also.
   Bypass gates were installed in 1938 and were ex-
tremely difficult to operate. A full day was required to
operate one gate using a  large crew. As part of Con-
tract  677,  new bypass gates with hydraulic cylinder
operators were  installed. The gates were fabricated of
creosoted wood with teflon bearing plates to decrease
friction of the  gate against the new aluminum guide
channels.
   New gates and  guide channels were installed in the
 channel just upstream of the Northwest meters and the
 second,  inactive  Venturi meter  was prepared  for
 service.  Figure 27 shows the two new gates at the
 Northwest Interceptor. Now more flow  can  be con-
 veyed from the Northwest Interceptor to the inverted
 siphon river crossing during storm flow periods if there
 is capacity  available downstream  in  the  Joint Inter-
 ceptor.
Figure 27.  New Venturi Meter Control Gates at
           Minneapolis Northwest Interceptor
   Modifications to  regulators in St.  Paul, in  general,
were more  extensive.  At the Rice-Rondo and Trout
Brook regulators (RR and TB), the only modifications
were the hydraulic  gate operator and  the Fabridam.
The regulator gate and Fabridam  at  the Trout Brook
                                                    31

-------
00
           CtM
           PW
           SEA
           M39
           E38
           RAN
           N2
           OAK
           RR
           TB
CONTROL AND MONITORING REGULATORS
CAMDEN PARK
PORTLAND AND WASHINGTON
26TH AND SEABURY
39TH AND MINNEHAHA
38TH AND EDMUND
3IST AND RANDOLPH
2ND AND MAIN
OAK STREET
RICE AND RONDO
TROUT BROOK
PHALEN CREEK
BELTLINE
OTIS AND MARSHALL
ST. PETER  AND KELLOGG
EUSTIS - EAST
EUSTIS - WEST
   ^ Vadnais
-N   ' Heights    »Gem Lake/ White Bear
'\  V            /\      /    Lake
     ~
                 I      2       3      4MILES

-------
regulator are shown  in Figure 28 and Figure 29.  The
gate is 32 inches wide and has a stroke of 21 inches.
The Fabridam, installed in an old stone block sewer,
has a height of 75 inches above the concrete base dam
and  is seen in this  photograph from the  back  side,
looking upstream. The air inflation pipe, seen along the
left  side  of  the photograph,  disappears into  the
concrete base dam and enters the Fabridam  near the
center of its front attachment.
Figure 28.  Hydraulic Power  Operated  Gate at the
           Trout Brook Regulator
   At the Phalen Creek regulator  (PC) a new, enlarged
regulator access chamber was constructed and  in it a
new gate and operator on the dry weather outlet was
installed. In the trunk sewer a Fabridam was installed.
A downstream secondary orifice  regulator on the dry
weather outlet pipe was eliminated, so a great improve-
ment was made in  the flow  conveying structures. At
this particular regulator, the outlet capacity  is small in
relationship to the flow arriving in the  trunk  sewer,
especially  in  the  summer   months. Before  modifi-
cations, dry  weather overflow must have frequently
occurred since, during 1969, high dry weather flow was
completely diverted only when the Fabridam was fully
inflated and a high head condition existed at the outlet
pipe entrance.
   At the Beltline regulator (BL), a new manhole was
installed to permit installation of the  Fabridam and the
existing diversion channel  was widened  to improve
flow conditions at the entrance to the outlet pipe. A
new fabricated wooden gate and hydraulic operator
were installed.
   At  the  Otis and Marshall regulator (KD),  a  new
power-operated gate replaced an  existing regulator
gate, the diversion tunnel was enlarged and a Fabridam
was installed in the trunk sewer.  Figure 30  shows the
upstream  side of  the Fabridam in a view  looking
downstream. The opening to  the dry weather outlet is
visible on the right side of the Fabridam, which is 60
inches high.  A power-operated  gate is now used  to
control the flow.
Figure 29.  Inflated  Fabridam Diversion Weir at the
           Trout Brook Regulator
Figure 30.  Inflated Fabridam and Dry Weather Outlet
           at Otis and Marshall Regulator.

   At the  St. Peter-Kellogg regulator (SPK), a Fabri-
dam was installed  and  work  done on  a complicated
regulator structure to  improve  its  capacity. This in-
                                                    33

-------
eluded lining 73 feet of previously unlined tunnel and
the removal of a cast orifice regulator. The Fabridam is
pictured in Figure 31.
Figure 31.  Inflated Fabridam at St. Peter and Kellogg
           Regulators
           Figure 32. Hydraulic Power Units
   Wabash and Cromwell regulator (EUE and EUW) is
a  newly constructed  regulator provided as part of a
Storm Water  Drainage  project in St. Paul. Here two
Fabridams were installed in each of two trunk sewers
and  equipment added to permit remote control as well
as local automatic control  of two existing pneumatic
cylinder operated gates.
   Table 7 summarizes the Regulator Modifications.

   Table 8 is  a tabulation of old diversion weirs and
Fabridam  data. In every instance, the fixed concrete
base upon which  the Fabridam rests is no higher than
the old fixed height weir so that there  is no uncon-
trolled loss of overall capacity to convey flow at any
regulator.

   Figure  32  shows package hydraulic  pump units,
commonly used in industry, which  provide power for
the cylinders which operate gates. These units are all
the same  size  for easy replacement and maintenance.
As  a consequence the  smaller gates  can be  quickly
positioned and the larger gates move more slowly. Gate
travel times for half stroke vary from 10 seconds to 10
minutes. The  slowest operating gates are the bypass
gates and  Venturi channel  gates at the Minneapolis
meter structures.  The hydraulic pump units are instal-
led  in  the vaults serving   the  regulators and are
connected  to  the gates in the sewers by stainless steel
tubing.

   Fabridam  inflation  is accomplished  by  small  air
blowers. The  blowers used  have  a rated capacity of
about 40 CFM when operated at 3000 rpm against two
psi outlet  pressure. For  uniformity, interchangeability
and  ease of  replacement and  maintenance,  all  the
blowers  are  identical although  different  pulley sizes
were  used  at  a few  of  the smallest dam  locations to
reduce blower speed and delivery. Inflation times vary
from  about four minutes to  about eight minutes to
bring  a  completely deflated dam  to full inflation
pressure. Most dams can be fully inflated in about  five
minutes. Recommended maximum working inflation
pressures vary from 2.0 psi  to 2.6  psi. The dams
become  quite rigid at  much  lower  inflation pressures
and  under  most flow  conditions will function ade-
quately at pressures around 0.5 psi.

   The  Fabridams are attached to the walls of the
 sewer and to  the reinforced concrete base constructed
 on the  floor of the sewer by rows of 12" long anchor
 bolts. Bolt diameter varies with the size  of the dam;
3/4",  7/8"  or  1"  diameter  bolts  are used.  The
specification required  by Firestone for these bolts  is
that  the "anchorage develop  the full  strength of the
bolt." Apparently this is quite important.
Post-Construction Evaluation
   In  June, 1968, less than a month after installation,
one dam was badly  torn during a rain storm when the
fabric pulled  from the clamping apparatus and caught
the rushing flow of water like a sail. The center portion
of the fabric was ripped on  both sides from front to
rear.  If  the  original  installation  of  the  clamping
apparatus was good, and Firestone  is quite  confident
that the  checking procedures used by the installer are
adequate, then the  probable cause of this was a slight
                                                   34

-------
                         Table 7
TABULATION OF ADDITIONS AND MODIFICATIONS - CONTRACT 677
      REGULATOR GATES
ADJUSTABLE DIVERSION DEVICE
Location
Camden

Portland &
Washington

26th & Seabury
N.W. Meter


S.W. Meter
River Crossing
Outlet

39th &
Minnehaha
38th &
Edmond
31st&
Randolph
2nd & Main
Oak
E. Meter
Rice & Rondo
Trout Brook
Phalen Creek

Belt Line
Otis&
Marshall
St. Peter &
Kellogg
Wabash &
Cromwell
Number
of
Gates
2
-

2
-
1
3
-
-
1
3
-
-

11

11

None
1
1
1
1
1
1

1

1

1
2
-
Type
of
Gates
New wood
New wood

Exist.
Exist.
Exist.
Exist.
New wood
New wood
Exist.
Exist.
Exist.
Exist.

Exist.

Exist

Required
Exist.
Exist.
Exist.
Exist.
Exist.
New wood

New wood

New wood

New wood
Exist.
Exist.
Size Cylinder
of Stroke
Opening (Inches)
4'0" x 5'0" 60"
5'-0%"x 4'0" 48"

28"
28"
28"
66"
7'-iy2"x7'0" 84"
7'-V/2"x7'0" 84"
78"
60"
60"
60"

16"

16"

This Location.
28"
16"
78"
Exist. 36"
21"
3'0" x 3'0" 33"

2'0" x 2'0" 24"

3'0" x 3'6" 42"

6'0" x 2'6" 72"
Exist. 36"
Exist. 36"
No.
of
Devices
None


1

1
None


None
None



1

1

1
1
1
None
1
1
1

1

1

1
2
-
Sewer
Section
Required


Arch.

Cir.
Required


Required
Required



Horseshoe

Cir.

Cir.
Horseshoe
Cir.
Required
Arch.
Arch.
Semi-
Elliptical
Transition
Semi-
Elliptical

Transition
Square
Square
Maxi-
mum
Width
This


7'-6"

8'-0"
This


This
This



11 '-0"

10'-0"

8'-6"
7'-0"
8'-0"
This
8'-0"
10'-0"
9'-3"

9'-0"



5'-11"
7'-6"
5'-0"
Max. Height
of Device
Above Invert
Location


5'-4"

5'-4"
Location


Location
Location



6'-10"

6'-8"

5'-8"
4'-8"
5'-4"
Location
5'-6"
7'-0"
6'-3"

7'-8"



4'-8"
5'-0"
3'-4"
                          35

-------
                                                Table 8
                               TRUNK SEWER DIVERSION  WEIR DATA
Location
Portland & Washington
26th & Seabury
39th & Minnehaha
38th & Edmund
31st & Randolph
2nd Bt Main
Oak Street
Eustis-1
Eustis-2
Rice & Rondo
Otis-Marshall
Phalen Creek

Trout Brook
Kellogg - St. Peter
Belt Line
Height of
Pre-Project
Diversion Weir
31 1V"
I /2
2'
3' 6"
2' 6"
12"
2' 2"
3'
-
-
-
r 8"
3'

r 8"
r 3"
"
Height of
Fixed Dam
3' 1%"
2'
11'/2"
2' 6"
12"
2' 2"
r 3%"
-
-
-
12"
1.1'

9"
r 3"
12"
Height of
Fabridam
2' 6"
3' 4"
5' 101/2"
4' 2"
4' 8"
2' 6'/i"
3' 7%"
5' 3%"
3' 6"
5' 6"
5' 0"
5' 1-7/8"

6' 3"
3' 5"
6' 8"
Section
Horseshoe 7'6" x 7'6"
Circular 8'0"
Horseshoe 10'6" x 10'3"
Circular 10'0"
Circular 8'6"
Horseshoe 7'0"x 7'0"
Circular 8'0"
Square 7'6"
Square 5'0"
Horseshoe 8'0" x 9'6"
Semi-elliptical 8'2yj"x 9'0"
Semi-elliptical 9'8-11/16"x
9'4-1/32"
Horseshoe 10'0" x 10'6"
Elliptical 5'8"x7'2"
Elliptical 9'0" x 11 '7"
vertical  movement of anchor bolts having inadequate
anchorage which  allowed the fabric to pull from  the
clamping apparatus. The anchor bolts were all removed
and  replaced  before  a  new  dam  was  installed.  No
further  problems with this particular installation have
occurred. During the spring and summer of 1969, three
additional  Fabridams were damaged  after being re-
leased from the clamping.  Two were damaged to  the
extent that replacement of  the dam was necessary and
one  could be  repaired and  reinstalled. Testing of  the
anchor bolts confirmed that bolt withdrawal was likely
the  cause  of  the  failures.  The  three dams were
reinstalled during the spring of 1970 after  all  anchor
bolts had been replaced.

   Some of the Fabridams have developed air leaks
that  were not  present immediately  after installation.
Some of these were caused by a small movement under
the clamping bar. Repairs were made at  two locations
where this  had occurred,  but both dams later failed
catastrophically, so it is safe to presume that the leaks
had  developed because of anchor bolt movement.
Another leak  that  was fixed had originated  at  the
location where the inflation air enters the dam.  During
installation,  overly  large bolt  holes were cut  in  the
fabric for  the flange bolts and the holes  elongated
under fabric stress and produced leaks.  This leak was
repaired  in  May,  1969,  and the air leaking  almost
stopped. Great  care was  taken  on reinstallations to
insure that bolt holes were cleanly cut to the minimum
size  so  that this  problem should  not reoccur. The
installer provided by Imbertson and Associates, Inc. of
Burbank, California, engineers for the initial Fabridam
installations, is no longer employed by them.

   Revisions to  the clamping  apparatus  have been
suggested  by  Firestone and  Imbertson that  should
make  installations somewhat   less  dependent  on
strength of bolt anchorage. The revision would  permit
clamping together  the top and bottom clamping bars
separately from the anchorage of the clamping  bars to
the sewer. This should provide more uniform clamping
pressures as well as reduce the likelihood that a small
movement  of  an  anchor  bolt would  result in cata-
strophic damage to the fabric dam.

   Inspection of several dams has revealed  no discern-
able  deterioration  of  the rubberized  fabric material
after  about  two  years  of  service in the  sewers.
Although  there  have been  some  problems, the local
contractor  and  Firestone have  been  cooperative  in
correcting problems at no cost to the Sanitary District.
The Sanitary District's Regulator Crew provided most
of the  labor, under the  supervision  of the contractor
and the  Imbertson installer, for the reinstallation  of
three  Fabridams in  1970,  however. The  experience
                                                   36

-------


gained by them  has  increased their ability to handle
maintenance of the Fabridams in the future.
   The original  installation of the Fabridams  included
 no provision for draining water from the interior of the
• dam.  Water  can enter through  punctures or other
 wounds to the fabric, through  leaks or breaks in the
 inflation pipe, by leaking through the attachment, and
 by condensation. The first problem with water was
 encountered following repair of a section of  inflation
 pipe  that was torn out during a rain storm. After the
 repair, the dam  could be  inflated, but not deflated,
 because the water in the dam could rise  in a vertical leg
 of the inflation  pipe until the column of water was
 equal to the inflation  pressure in the dam.  Rather than
 pierce the fabric to install a drain, a "Y" fitting was in-
 stalled near the  point where the inflation pipe enters
 the concrete base under the Fabridam.  With a plug in
 the "Y" the dam can be  inflated and then, when the
 plug  is removed, the air pressure in the dam forces the
 water out. Several  inflation-deflation cycles have been
 sufficient to  remove  almost all the water  from a dam
 and  restore deflation control.  The extra fitting  was
 installed  on  all  the  Fabridam  inflation  pipes so de-
 watering can easily be accomplished.  This  also permits
 deflation of  the dam without  requiring  operation of
 the solenoid valve in the vault.
   In  addition to the blower, there are several other
devices installed  in the vaults to control and protect
the Fabridams. Figure 33 shows the control valves and
piping  at  the blower. A  one-inch  normally closed
solenoid valve is used to  control deflation. The valve
can be energized and  opened locally or remotely via
the telemetry system. A one-half inch  normally open
solenoid valve will cause the dam to deflate automati-
cally  upon power failure to insure that,  if control is
lost,  the  dam   will   be   in  the safest  position. A
pneumatic check valve prevents air flow back  through
the  blower.  The  dam  must  be protected  against
over-pressurization, apparently  to avoid  damage. A
spring-loaded pressure relief valve was  provided  to
allow  air release when pressure exceeds  a set limit.
These valves have not been very reliable. Once opened,
they  tend  to remain cracked open slightly and do not
always seal  after  pressure has  been relieved. They
apparently do not  provide  fast  enough pressure relief
either.

   A   final   precautionary  device  to  prevent over-
pressure is  a water column whose height equals the
maximum  allowable  internal  pressure.  Ostensibly the
water column was  intended to be a redundant device.
 However,  it was  found  that  severe   storms  did
occasionally cause Fabridam pressure  to  be relieved
by the water column rather  than  by the mechanical
valve.  Air  entering  at  the base of the column forms a
piston which discharges most of the water from the top
of column quite suddenly, and then  the dam deflates
Figure 33.  Control  Valves,  Piping, and  Blower  for
           Inflatable Dam

entirely. The dam cannot then be reinflated until the
column is again filled with  water. This has happened
frequently  on only a few dams but is a disadvantage
since post  peak runoff often could  be diverted if the
dam could be reinflated. On two locations where  this
has  occurred,  a pressure  switch  has been installed
which,  when pressure in  the pipe terminated in the
water column exceeds a set value somewhat less than
the height of the water column, will energize and open
the solenoid  valve until the pressure is relieved. Since
this  more  positive  acting pressure relief measure has
been  in use, several  rainstorms  have occurred during
which pressure  relief was adequate and the column not
discharged. Pressure switches will be installed at other
locations where the water  column emptying has re-
duced  the  effectiveness of the system. Another mea-
sure that could  be taken would be to replace the 4-inch
diameter water column with  a larger one through
which  air  could be  discharged  without  causing  the
column to be emptied.
   A pressure sensing line runs inside the 1-1/2" or 2"
inflation pipe  into the  interior  of the  dam.  This
pressure is supplied to a pressure  switch installed in the
electrical circuit which operates the air blower. The
switch  is set to open the circuit  and  stop blower
operation when  pressure in the dam reaches a set limit.
The  use of  the  tubing  permits the blower to be
controlled by a near static air pressure because of the
large volume in  the dam.
   The  original installation of the Fabridams did  not
provide any means of determining remotely the inter-
nal pressure. Sanitary District personnel later obtained
                                                    37

-------
and  installed  pressure transducers  which enable the
dam inflation pressure  to  be monitored. Small film
type potentiometers manufactured  by Computer  In-
struments Corporation were used and have been found
to be extremely reliable.
   All the piping appurtenant to the inflatable dams is
PVC plastic. There has been no real disadvantage to its
use except that the portions installed in the sewer have
been damaged and torn out at two locations. This has
not recurred since reinstallation included protection
from debris carried by storm flow. The use of securely
anchored  metal pipe would probably  have made this
protection unnecessary.  The inflation  pipes enter the
concrete base dams on their downstream sides so that
they are protected by the dam except  under severe
storm flow conditions. This also permitted the work to
be done on the dry side of the dam during installation.
   Tank gages are used to measure  gate position. Gate
movement  produces  a shaft rotation into a  sealed
potentiometer  which transduces gate position  into an
analog voltage. Several of the gages have been removed
for cleaning  and  several potentiometers have been
replaced. The service is severe, however—several can be
submerged  during  rainstorms—and  this  maintenance
should probably be expected.
                                                    38

-------



VI.  PROCESS CONTROL AND COMMUNICATIONS
EQUIPMENT
Processing System
   Specifications were prepared by Pioneer Service and
Engineering  Co.  for "Installation  of  Air  Pressure
Transducers,  Telemetering  and  Supervisory Control
Equipment, and Computer-Contract 664B."

   The system described in the specifications had to be
capable of collecting, processing and  logging data  on
sewer  flow, gate status, rainfall and  river quality. Data
from remote station instruments was to be input to a
computer, scaled to engineering units and logged both
as a printed record and in machine processible form.
Provision  for  controlling  gates at remote stations was
also required. Computer time had  to be shared for
logging treatment plant data. The computer had to  be
capable of performing other computations during  its
idle time  in addition to the primary on-line functions.
In other words, the computer was to be provided with
foreground-background time sharing capabilities. The
requirements  were  essentially for a  computer-based
supervisory control and telemetering system.

   Specifically, the work  under this  Contract 664B
consisted  of  furnishing  and  installing 44  pressure
transducers,  telemetering  equipment  for  38 remote
stations,  supervisory control equipment, and a com-
puter  with  certain peripheral equipment. The  con-
tractor was  responsible  for the  compatibility  and
operation  of  the  complete system   from  pressure
transducer to  and including the computer and com-
puter logging.

   Bids were  received from Badger Meter  Manufac-
turing Company, The Bristol Company, General Elec-
tric  Company, Automatic Electric  Company, Control
Data Corporation  and CAE Industries. The contract
was awarded to the low  bidder. Badger Meter Manu-
facturing Company.

   Certain optional  alternates were bid and  during
contract negotiations the following were approved:
1. Substitution of  a Potter Chain printer  for  one
   output  typewriter.
2. Addition  of 8000  words to the core memory  to
   increase its size to 16K.
3. Substitution of a  disk  file for a  Random  access
   memory file.

   Later by change order, two other items were added
to the contract. One was the inclusion  of an operator's
panel  for the out-plant  sewer control system.  This
panel would  allow  manual addressing, remote  opera-
tion of gates, and  monitoring of outplant data even
though the  computer might be out  of service.  The
operator's panel was also  of value for maintenance and
checkout purposes. The other item was the inclusion of
two  DEC TC02 magnetic tape transports. This pro-
vided a more reliable system, backup mass memory in
case  of problems in the disc file, and easy exchange of
manufacturer prepared programs. Both have proved to
be very valuable additions.

   The digital computer used with  the  system is a
PDP-9  manufactured by the Digital Equipment Corpor-
ation.  The PDP-9, pictured  in  Figure 34, is a  single
address, 18-bit word length, parallel binary computer.
It is a completely self-contained machine, not requiring
special air conditioning or humidity  control.  The
central processor's  core memory  has a 1.0  micro
second cycle  time.  Originally  the PDP-9  used had
16,384 words;  the core memory  size was later in-
creased to 24,576  words as part  of  an agreement
modifying the contractor's software obligations.
Figure 34.  Digital  Equipment
           Computer
Corporation  PDP-9
   The standard  PDP-9  I/O  facilities  include a 300
 character-per-second paper tape reader, a 50 character-
 per-second paper tape punch and  a  10 character-per-
 second console teleprinter.  In addition,  the Minne-
 apolis-Saint Paul  Sanitary District system has a high
 speed  disk, two  magnetic tape drives, a line printer,
 two additional console teletypewriters and  a logging
 teletype.
   The magnetic tape system, called  DECtape, em-
 ployes small  reels of 3/4", 1 mil mylar sandwich tape
 with an information capacity of 2.7 x 106 bits per 260
 foot reel.
   The disk storage drive is a Control Data 853, shown
 in  Figure 35. This  disk drive uses  industry standard
 removable disk packs. Each disk pack has six disks with
 ten usable disk surfaces and 100 tracks per surface. The
 total  capacity per disk pack  is  24,576,000 bits or
 1,365,333  18-bit  words.  The maximum  one-track
 access time is 30 milliseconds.
                                                   39

-------
       Figure 35. Control Data 853 Disk Drive

   The line  printer,  Figure 36, is a Potter  HSP-3502
Chain Printer having a printing speed of 315 lines per
minute. Its features include  132 character output per
line and reformattable page control.
           Figure 36. Potter Line Printer


   The  teletypewriters on  line with  the  system are
standard ASR  33 and ASR  35. Both types employ the
10 character-per-second  output and compatible inter-
facing. The ASR 35 is simply a heavy duty model of
the  ASR 33  which  enjoys  longer life  and quieter
operation.
   The controllers and special hardware requirements
were designed and  built specifically for the  Minnea-
polis-Saint  Paul  Sanitary District's  system. The disk
controller, provided by Digital Logic Corporation as a
subcontractor  of Badger Meter Manufacturing Com-
pany, caused  serious problems for a  while but  since
being  revised  has  been  highly  dependable. Other
                                                      interfacing  was accomplished by  Badger Meter. This
                                                      included the line printer handler, the operator's control
                                                      panel,  digital displays, and  strip  chart  recorders, all
                                                      mounted in a single interface cabinet. Figure 37, along
                                                      with the disk controller and  the base station telemetry
                                                      equipment.
            Figure 37. Interface Cabinet


   The software requirements were to be met by the
systems library software developed  by the computer
manufacturer,  by special  programs supplied by the
contractor, and Sanitary District prepared  programs.
Digital  Equipment  Corporation's standard  software
package for the  PDP-9  is called Monitor  V4B. Its
features include:
1. ASA FORTRAN  IV
   This compiler meets all ASA  specifications and
   standards.
2. MACRO-9
   A  basic assembler  language  comparable to  BAL
   (IBM's processors) or OSAS (CDC's processors)
3. PIP
   Peripheral Interchange Program for manipulation of
   data files and programs.
4. EDIT
   Systems program  used  in program preparation and
   modification.
5. DDT9
   Dynamic Debugging Technique. A modified loader
   which can debug  a user's program while  in opera-
   tion.
6. LINKING LOADER
   A  program  which loads any absolute or relocatable
   user program and all called subroutines.
As special  options, the POP system also has:
   Extended Arithmetic Element
   Memory Extension Control
                                                   40

-------
   Power Failure Protect
   Memory Protection
   Automatic Priority Interrupt
   In  addition  to  its program library, the Sanitary
District has had access to all other services extended by
DECUS  the  Digital  Equipment Corporation  Users
Society. Project personnel  have attended technical and
educational  meetings held  by  DECUS.  Additional
software for the PDP-9 is automatically received as it is
entered into or updated by the  DECUS library.
   Badger  Meter Manufacturing Company was to have
provided  an executive time-sharing  routine, operator
Input/Output routines and other programs for trend
recorder selection and output, acquisition of outplant
data,  and gate selection  and  control.  The Sanitary
District was to supply  the real  time  programs  for
scanning, scaling data and logging data, obtaining the
data from contractor supplied subroutines.
   The original FWPCA  project termination date was
July 1, 1969, with the year of operation for evaluation
complete in 1968. When it was realized that Contract
664B  would not be completed in time to start system
operation  during  1968,  the Sanitary  District applied
for and  received from  the FWPCA  a project time
extension  of one  year  with  no  increase  in  project
funding. This meant that operation had to begin in the
spring of 1969. In January, 1969, the Project Engineer
felt that the best interests of the Sanitary District and
the FWPCA required that  Badger Meter Manufacturing
Company's work toward meeting the requirements of
Contract 664B  be interrupted so that the system could
be  prepared for  operation  during  the 1969  rainfall
season.  The contractor's efforts at  this time were
largely  directed,  through a  subcontractor,  toward
completion  of the executive time-sharing  software.
Provisions of the contract clearly permitted the Sani-
tary District to take this action. The contractor, with
the understanding that it would be provided a further
opportunity  to complete its  obligations  when  the
system's utilization  was  less critically needed, coopera-
ted by providing system  maintenance and support.

   Fortunately, almost all of the system was usable and
the Sanitary District was  able to prepare programs in
time to operate in  1969. The system was operated as a
dedicated  system and could not, while scanning, be
utilized further.
   In  May,  1969,  negotiations began  regarding com-
pletion of  remaining  contract obligations by  Badger
Meter Manufacturing Company. By this time, operating
programs for the system had been  prepared and most
of  the system  hardware had  been  successfully used.
The contractor proposed  to complete  all  unfinished
hardware  related work  and  have  separate  hardware
acceptance testing as the criterion for  hardware accep-
tability. They also proposed that a  soon-to-be-released
DEC  foreground/background monitor system would be
better  for the  Sanitary District than  the still  not
completed  executive  time-sharing  routines. All pro-
gramming by the Sanitary District since January  had
utilized the current DEC monitor system so the use of
DEC's  foreground/background  monitor had  definite
advantages.  The  proposal   also included  additional
hardware  items,  another I/O teletype with associated
logic and  8,000 words  additional core memory,  and
cash contribution toward writing the  handlers  and
making other modifications required to adapt the DEC
monitor for use on the Sanitary District system.
   On  October 15, 1969, an agreement between the
Minneapolis-Saint Paul  Sanitary District  and  Badger
Meter Manufacturing Company was signed. Essentially
the agreement relieved the contractor of the software
obligations in Contract 664B  in  exchange  for  the
additional  hardware  and  cash  contribution  toward
adaptation of the DEC foreground/background system.
Subsequently,  the contractor returned, completed un-
finished hardware work  and obtained complete hard-
ware  acceptance.  The  Sanitary District  received  a
preliminary  foreground/background  monitor  system
from DEC and had its systems consultant begin writing
necessary handlers. At the time of writing, June, 1970,
the general  and  complete foreground/background
monitor had not been released by DEC. This release is
anticipated soon.  Based  upon experience gained using
the preliminary version,  it is  expected that only a  few
months will be required  to implement the foreground/
background  time-sharing system once  the  finished
version is received.

Leased Line System
   The telemetry equipment provided under Contract
664B to permit information to  be returned from the
outplant sites and  control  to be exercised from the
central station is dependent upon  a  system of leased
telephone  lines   as the medium of  communication.
Fairly  early in the project a  meeting  was held between
representatives of Northwestern Bell Telephone Com-
pany,  Badger Meter Manufacturing Company and the
Sanitary District.  It  was agreed that seven  circuits
leading to the central station would  result in the  best
application  of  the communications  equipment.  The
telephone  company would  in the leased  lines incor-
porate high  impedence bridges with  sufficient  gain
added  to  make the lines appear as an 8 db loss at 1,000
cps. The  frequency  spectrum  to   be  used for the
communications  system was 400 to 2700 cycles per
second. Figure 38 shows the leased  line arrangement.
   Northwestern Bell  Telephone Company began work
on providing the leased  lines in December,  1967. Many
of the circuit drops  were  made by January but not
without  problems and much communication between
Sanitary District and telephone company personnel.
Many  of the remote sites  were underground  vaults,
some at sites  not easily described. The river monitors
                                                   41

-------
                       SEWER CONTROL SYSTEM
                               DATA POINTS
Figure 38. Schematic Diagram of Outplant Stations and Leased Line System

-------
were not in place by that time so completion of some
of the circuit legs had to be delayed. Rain gages were
located near public buildings and the locations of the
telemetry equipment located in any available space in
the nearby structures,  which sometimes caused com-
plications. Street addresses were used to help locate the
various remote  stations,  but calls  from  telephone
company personnel  asking questions such as,  "Are you
sure you  want  this circuit terminated at  the shoe
store?" were not uncommon.
   After the  circuits had  been  physically  prepared,
they  were tested  for attenuation  losses at various
frequencies and corrections were  made where required.
Progress was  somewhat hindered by a strike of  tele-
phone company  workers from  April  18  to May 6,
1968.  Since  telephone company work preceded  the
installation of communications equipment  in the field
by  Badger Meter Manufacturing Company, some equip-
ment connections were not made until the  first part of
June,  1968, because the telephone company had to
extend their  lines from  one termination  point  to
another within reach of our equipment.
   Installation charges for the seven leased  line circuits
service 37 outplant sites were $1,000. The monthly
charge for the seven circuits is $548.


Communications System
   Part of Contract 664B  called for provision of a
communications system consisting of supervisory con-
trol  and  telemetering equipment.  The   supervisory
control equipment permits  sequential selection, inter-
rogation, and verification  of remote station transducer
signals for connection to  telemetering channels and
also control of gates at remote stations in the system.
The telemetering equipment transmits data  from  the
remote stations  to  the central control station where
analog  to  digital  interface conversion permits data
input to the computer.
   The air pressure transducers  provided  are Robert-
shaw  Model 115 Microsen Pressure Transmitters which,
by  means of a  bellows pressure element,  convert air
pressure to an  analogous  4 to  20 milliamp signal.
Forty-two  pressure transducers  in  ranges of  0-40",
0-72", 0-120", and  0-144"  of water are used, and one
spare in each range was supplied.
   The supervisory control and telemetering equipment
is that manufactured  by Noller Control  Systems, a
wholly owned subsidiary of Badger Meter Manufactur-
ing Company.
   The remote station  telemetering equipment had to
transmit data from the pressure to current transducers
and from rain gages, gate position transducers and river
monitors. At the present time there are  139 unique
addresses from which information can be obtained.
   Remote station  addresses are three or four  digit
octal  numbers which  in binary form a selection code
 indicating which of the seven  leased telephone lines is
 to be accessed, which group  of two transmitters the
 central station will use, and  which of three possible
 states each of the two transmitters will be in. Table  9
 shows the numbers of points of each description.

                    Table 9
       NUMBER  OF MEASUREMENT  AND
             CONTROL FUNCTIONS
1
Function Lo
Level Measurement— Interceptor Sewers . .
Level Measurement — Trunk Sewers . . . . .
Level Measurement— Outlets to Interceptors
Gate Positions and Controls
Rain Gages 	
River Quality Monitors
Alarms and Spares . ... - 	

Mo. of
cations
12
15
12
17
9
5
18
(a)
No. of
Points
12
18
12
34
9
30
24
139
(a) Total number of locations of telemetry equipment is 37
   due to overlapping functions at certain stations.
   The remote stations are of two types insofar as the
 telemetry equipment is concerned. One type»has only a
 single  point such as an  interceptor monitoring station
 or a rain gage. The other type has up to eight points at
 a single location, although most have only five. At the
 single  point stations only a single receiver and trans-
 mitter are used to complete the  unique point address
 and produce a "confirmation" code back that verifies
 selection  of the  proper point.  The  other stations
 employ a  pair  of  receivers and transmitters.  Trans-
 mitters are three-state frequency-shift devices,  so the
 pair arranged  as a three by three matrix can produce
 nine possible  combinations. One  combination is not
 used, so a pair of tone transmitters and receivers can be
 used for up to eight points. Data transmission is analog.
 Transducers produce or  have signals conditioned  to 0
 to  5 volt analogs, which are converted to pulses, and
 then  to  a frequency-shift keying  tone  transmitter
 signal. Central station receiving equipment reproduces
 the direct current transducer signal.  All components
 are  self-contained on  circuit  boards which are  edge
 mounted in card cages.
   The Sanitary District has been able to maintain the
 operation  of the telemetry  system without  assistance
 for over a year now. A short period of instruction was
 provided  by the contractor and  initial checkout and
 adjustments accomplished by him.
   Some spare components have been bought to assist
 in maintaining continuous operation from all outplant
 stations. MSSD has  not yet performed  repair of any
 components; components  replaced because of  prob-
 lems have been sent to Badger Meter Manufacturing
 Company for repairs.
   MSSD  has found that maintenance of the communi-
 cations system involves more  than the  ability to
 maintain one's own equipment. It has been helpful to
 be able to define telephone  line problems  to some
 extent when reporting such problems to the telephone
                                                   43

-------
company. All complaints regarding telephone lines are
handled at the closest exchange to the central station.
This has worked well since only a single person need be
contacted rather  than  someone at each of  the many
telephone central  offices involved  in the system. Some
special  equipment  has  been  required to  assist  in
maintaining the  communications  system.  Two  fre-
quency selective  voltmeters, two AC  voltmeters and
two electronic counters, one each for  the remote site
mobile maintenance vehicle and for the central station,
were obtained.
   Maintenance of  the system  has been  under the
direction of the Project Engineer who has also handled
the  central  station end   of  many  calibration  and
troubleshooting operations. The outplant end has been
handled  by the  electrician assigned to the  project.
Other  plant  electricians  have  been trained, and the
maintenance  electricians  are now able to  handle the
communications equipment alone.
   Periodically, perhaps two or three  times per year,
adjustments in tone transmitter and receiver sensitivity
levels  may be required to maintain the proper levels
and operating margins. This is necessitated by seasonal
variations in the characteristics of the leased telephone
lines.

Calibration Procedures
   The use of a computer with the data logging system
makes calibration  simple. All calibrations are from the
input  to the primary  sensing  element to  the scaled
value in engineering units appearing on the printed
page.  An example of a calibration is cited:  Personnel
who have  reached  the remote site to be calibrated
contact someone at the central station via radio. At the
central  station,  the  calibration  program,  named
CHKOUT for CHECKOUT, is  called.  This  program
samples the A/D output  and calculates the analog
voltage  equivalent and  prints this voltage on  the I/O
teletype. At .the remote site, a hand-operated pressure
calibrator is used to apply various levels of pressure to
the pressure transducer. The corresponding  voltage  is
measured  at  the  central  station for  each pressure.
Figure 39  illustrates the teletype commands  used  in
calibrating, with  the  CHKOUT  program.  A linear
scaling equation is then calculated from the pressure-
voltage  data. The  same  technique is  used for  the
inflated dam pressure transducer. Voltages correspond-
ing to gate  position are read also. With this data, which
can be obtained in a matter of minutes, four scaling
equations are calculated and entered into the appro-
priate  reference  list  to complete calibration of the
control and monitoring  station.
   Other stations  are  calibrated similarly. Rain gages
are calibrated using one-inch equivalent weights. River
monitors are calibrated  using  the ambient  reading
observed at the local readout meters and the zero point
for each parameter.  All calibrations are periodically
checked to see if scaling factors need to be updated.
 tC

 MONITOR  V4B

 $E  CHKOUT




 ADDRESS-CXXXX)

 1397
   (Operator types in request for
    CHKOUT program)
(Address 1307 is the regulator gate
at 39th and Minnehaha)
 FACTORS-*XXXX.XXX  XXXX.XXX)
                               (Factors used only when
                               scaled values are to be
                               reported)
 1307 2.845    0.000
 ADDRESS-(XXXX)
         (0% open - transducer
         output is 2.845 volts)
 1307

 FACTORS-CXXXX.XXX  XXXX.XXX)
 1307 2*845    0.000
 ADDRESS-(XXXX)
1307

FACTORS-tXXXX.XXX XXXX.XXX)
 1307  4-640    0.000
ADDRESS-(XXXX)
       (100% open - transducer
        output is 4.640 volts)
1307

FACTORS-fXXXX.XXX XXXX.XXX)
1307  4.640    0.000
ADDRESS-(XXXX)
 Figure 39.  Calibration of an Outplant Address using
           the CHKOUT Program
 Inplant Data Acquisition System
   Unrelated to the Regulator Demonstration Program,
 but  included  in Contract 664B, was computer capa-
 bility and equipment to  permit plant monitoring and
 data  logging  for  certain parameters  in the sewage
 treatment plant as a secondary priority  function. This
 includes a station located in the Plant Operators' office
                                                   44

-------

where  plant data can  be displayed  and logged.  The
system provided  is hard-wire  with all of the analog
signals wired directly to the interface equipment.  One
combination of the three most significant bits of the
computer buffer register and the nine least significant
bits  is used  for  address codes for  the inplant  data
inputs. The output  buffer  register  is decoded using
binary to octal decoders  to give 241  individual address
signals which control 241 reed relays which switch the
various analog signals into tne A/D converter.  The
common  input channel  for the thermocouples  js a
thermocouple  wire referenced  to a cold junction  and
fed through a high gain amplifier and then to an  A/D
converter. All  other input signals are conditioned  to a
standard  DC  signal  before  being  switched  into  the
 common input-channel to the A/D converter.
   The  hardware provided  has been  tested and  ac-
cepted but little use of the plant data logging capabili-
ties has been made to date. It is expected that during
1971, after  the  new time-sharing monitor system  has
been implemented,  the  inplant data logging programs
will be prepared.
   The  equipment  located in the  Plant Operator's
office-the ASR35  logging teletypewriter, the ASR33
I/O teletypewriter, and the inplant interface cabinet-is
shown in  Figure 40. The strip chart recorders can be
assigned to  either  inplant or  outplant  points. Also
mounted in  the cabinet are receivers used to present
telemetering data  from the existing  interceptor and
plant flow meters to the computer.
                                Figure 40. Plant Operator's Logging Equipment
                                                    45

-------
VII.  RIVER MONITORING
   Since there has only been intermittant sampling of
Mississippi  River water in  the  metropolitan area at
frequencies varying between once  per week and once
per month, with no  consideration given  to time of
occurrence of rainfall, it was felt  that the effects of
sewer overflows were really  unknown.  To assist in
making an evaluation  of the effects of combined sewer
and storm sewer overflows, five river quality monitors
were  acquired  and sited  to intensively  monitor the
urban stretch of the river most  affected  by  sewer
overflow.

Site Selection
   In selection of sites, several factors were taken into
consideration. First, the general locations had to be
chosen such that the intention, measuring the extent to
which the river  is affected by overflows, could be
accomplished. A monitor  upstream of  most of the
metropolitan area, particularly upstream of any com-
bined sewer  outfall,  was  necessary. About half-way
through the cities is the confluence of the Minnesota
River, which differs considerably  in quality, with the
Mississippi  River. To  permit observation of the effect
on the  river  of the upstream 13  miles of storm and
combined sewer discharge, a monitor had to be located
upstream from the river junction. The Minnesota River
joins  the Mississippi River  in a section  of rather quiet
flow, and there is no  dam,  waterfall or rapids to aid in
mixing.  The two rivers appear to flow side by side in a
common channel for  some distance. To permit obser-
vation of the effect  of the Minnesota River on the
Mississippi  River,  a monitor  was  needed  somewhat
downstream from a point where adequate mixing of the
two waters has occurred. Fortunately, there are only a
few  sewer outfalls in the  section  of river  where this
mixing  takes  place  so this monitor  location  will
constitute  a  good measure  of   entering  condition,
upstream  from a substantial part  of St. Paul's sewer
outfalls. The next  general  location  needed is  down-
stream from the remaining St. Paul outlets but upstream
of the Minneapolis-Saint Paul Sanitary District Treat-
ment Plant outfall.
   The upstream monitor was sited at the Minneapolis
Water Treatment Plant which is located just north of
the Minneapolis city limits  in the suburb of Fridley. A
location in the raw water intake pumping station there
was  easily the  best  site  available. The total annual
water diversion is  over 28 billion gallons, averaging
about 77  MGD  or  121 cfs. The river geometry at the
pumping station is such  that this diversion should
provide a good representative sampling of the river.
   To insure,  as much as possible, that representative
samples  of  river water are  provided to  the  river
monitor,  the  river was cross-sectioned  and sampled
extensively at each  of the other general locations. The
contours of the river  channel were  roughly determined
by  soundings.  Some measurements of current were
made but were probably not extremely reliable since
they  were made with  the  current  meter suspended
from  a boat which was not  always absolutely station-
ary.  Samples were  taken  at two  or  three depths  at
several stations across each  cross-section  and also  at
any  feasible  pump-mounting site. Dissolved oxygen,
conductivity and  temperature were measured immedi-
ately, and samples were also run through the Techni-
con Autoanalyzer in the Sanitary  District laboratory.
The  automatic  analyzer determined concentrations of
chloride  ion, ammonia-nitrogen, total Kjeldahl nitro-
gen,   phosphate  and  urea.   In  each  case  the data
indicated  that  the  proposed  sampling  sites  would
provide representative samples.

Contract 666
   Specifications  were prepared for the  Minneapolis-
Saint Paul Sanitary District by Pioneer Service and
Engineering Company to provide water quality  instru-
mentation stations. Five river quality monitoring sta-
tions were  acquired in 1968  under contract  from
Fairchild  Camera  and  Instrument  Corporation  of
Syosset,  New York.  Four were  mounted in two-wheel
trailers,  the  type  used for temporary  construction
offices. The fifth  was provided for permanent installa-
tion  in the pumping station at  the Minneapolis Water
Treatment Plant.  The units measure pH, conductivity,
oxidation-reduction  potential, temperature, chlorides
and  dissolved  oxygen.  Fairchild  Camera  and Instru-
ment Corporation has since disposed  of its Environ-
mental Systems Operation which  is now Automated
Environmental Systems, Inc. of Woodbury, New York.
   The trailers  measure 8  feet by  22 feet and are
insulated and equipped with sinks, plumbing, electric
heating, ventilating,  fluorescent lighting systems, cabi-
nets and a work  counter. One sink is provided with a
small water reservoir and a pump as a clean  water
source.  The other  sink is  connected to  the  sample
plumbing and provides a water sample tap at the work
counter.  The trailers have one door and no windows  in
an attempt to make them as vandal-proof as possible.
The  water analyzer unit itself,  Figure  41,  is quite
compact  so a  smaller trailer could  have been used.
However, the dollar savings  in getting a smaller  trailer
would have been minimal.and it was felt that a larger
size  would increase their later utility for expanded
sampling and analysis work.
   The analyzer  has local  readout  meters  for each
parameter measured  and two electrical signal outputs,
one 0 to 5 volt range for telemetry and one 0-50 mv.
range for local  recording. The  monitors were bought
without  local recorders since a computer-based tele-
metry system was available. Since then, two multipoint
strip chart recorders have been acquired; one provides a
local  recording station at the Minneapolis Water Treat-
ment Plant because of their interest, and the other
provides  data recording  backup  in case of telemetry or
                                                    47

-------
telephone circuit problems at any of the monitors and
provides a  portable recorder for more intensive record-
ing of  readout  data  as  an aid to instrument trouble
shooting.
   External connections  to the trailer plumbing are an
influent pipe and an effluent pipe.
    Figure 41. River Quality Monitor Analyzer Unit
Installation, Operation and Maintenance
   The analyzers were  mounted in  the trailers at the
Fairchild factory in New York, transported over the
road without  incident and  delivered  to  the Sanitary
District.  Delivery of one of the trailers was delayed
because  of  its instability in high  winds at highway
speeds, but trailers have been demonstrated to be fully
transportable without damage to the analyzer. The first
monitor which arrived was put into operation for a test
period  on  plant  secondary effluent. The monitor
operated well  there and some effects of rainfall on the
final effluent were recorded.

   Peerless Dynaflow positive displacement submersi-
ble well  pumps were purchased to provide water flow
to the trailers. The same pump  is used by the FWPCA
on their river monitors here.  The trailers were delivered
in May, 1968  and were  installed immediately although
telemetering was not at that time ready. Figure 42
indicates  the   locations  of the  five  river  quality
monitors.
   The pumps are housed in strainers fabricated from
sections  of  well screen. These strainers  confer struc-
tural protection and keep the small  screen over the
pump intake from being covered by large debris. Figure
43 shows one of the strainers. The trailers were moved
to their  installation sites by commercial movers and
blocked by Sanitary  District personnel.  Blocking the
four covers of  the trailers  increased their stability,
allowed them to be closer to the ground and permitted
removal and storage of the wheels  and tires which
could  otherwise be  stolen.  Telephone and  power
connections were made to the trailers and  the pumps
and the appurtenant piping installed.
    Figure 43. River Quality Monitor Pump Screen

   The downstream monitor, RQ 1, pictured in Figure
44, was  installed on a dike at an industrial molasses
terminal. The pump was mounted  on a wooden pile
cluster dolphin used in mooring barges at the unloading
dock. Flexible polyethylene pipe was used to convey
the water to the trailer.  From the water's edge to the
trailer, the pipe was covered by rubber insulation under
which electric heating tape was installed to  prevent
freezing during the winter months if the pump should
be  off  for  a short  time.  The insulated  pipe was
protected from physical damage by a rigid pipe  running
up the bank to  the trailer. The effluent pipe was run
unprotected to the water's edge. This installation was
accomplished  by personnel from the Sanitary  District
maintenance shop.
    Figure 44.  Industrial Molasses River Monitor
                                                   48

-------
CO
          *
          RQI
          RQ2
          RQ3
          RQ4
          RQ5
MSSD RIVER  QUALITY  MONITORS
INDUSTRIAL  MOLASSES
FARM BUREAU
HIGH BRIDGE
FT. SKEUING
MPLS.  WATER WORKS
	l

-------
   Because of the inplant work load,  the other four
installations were made by a local plumbing contractor.
Monitor RQ 2 was installed at a fertilizer terminal with
the pump mounted from a concrete cell dolphin used
for mooring barges at their unloading dock. Because of
a flood wall between the trailer and the river, the pipe
and power cable were run on the river bottom through
an outlet  culvert leading into  a storm water pumping
station discharge wet well and  up to the trailer. The
pipe  above  water  was again  heated, insulated and
physically protected by a  rigid pipe where exposed.
The effluent pipe from the trailer discharges to a storm
drain  nearby.
   Monitor RQ  3 was installed  at a cooling water
intake pumping  station serving Northern States Power
Company's High Bridge Power Plant. The trailer was
parked  next to the  pump building  and the  pump
suspended into the wet well channel behind a traveling
water screen. The pipe was protected  outside  the
building in the same manner as at the other monitors.
   The monitor  upstream  from the Minnesota River
mouth, RQ 4, is located on City property, right-of-way
from  a  since-demolished highway bridge. Because of
the narrow width of the river and  the barge traffic, the
Corps of Engineers would not permit the installation of
a  pile cluster, even  a small distance from  the river's
edge.  Therefore,  the pump was mounted on the end of
a private marina  boat slip. Because of the low elevation
of the trailer site and the questionable pump location,
no effort was made to heat, insulate or protect the pipe
other than by direct burial.
   The pump  serving river  monitor RQ 5 is mounted
behind a traveling water screen in a pumping forebay at
the Water Treatment Plant. The analyzer was installed
in an old  meter repair shop now being used only for
storage.
   The primary operating difficulty  has been the need
to replace pumps periodically. The pumps run continu-
ously and are  subject to wear from silt and sand which
is often  encountered in the water. The average life
expectancy appears  to be about  four  to six months.
The pumps are  repaired as they are replaced, but the
repairs are fairly expensive. Repairs that  have been
made are replacement of motor  and replacement of
worn rotor and  stator. A few centrifugal pumps were
purchased and used, but proved even less reliable, so
now  only  the  positive  displacement  pumps  are
repaired.
   On two occasions, freezing of water in the pipe has
caused problems. After a  pump failure at RQ 4, new
pipes and power cable were  run over the snow and ice
and  the  frozen  pipe left until  spring.  We had  no
freezing problems with unprotected pipes  so long as
the water flow was continuous. On another occasion, a
power outage allowed ice to block a pipe.
   RQ 4  had to be moved  in October,  1968 when
flooding on the Minnesota River submerged the trailer
site.  Three trailers,  RQ 1, RQ 3 and RQ  4 had to be
moved in the spring of 1969 during the  near record
flood  on  the  Mississippi  River.  In  each  of these
instances the pumps and piping  remained in the river
and  only  the  trailers were moved. This permitted
speedy reinstallation after water levels receded.
   Maintenance of the  analyzer itself has been limited
to cleaning and calibration  of sensors. Electronically
the  equipment  has been  very  reliable.  During  the
winter months  when  water  temperatures  are  near
freezing,  maintenance visits for cleaning of sensors and
calibration checking at intervals of  two weeks were
adequate. Very little biological growth  in the sample
cups and on the sensors occurs and,  probably because
of the ice cover and absence of river traffic, much less
silt accumulates in  the sample chamber.  During the
summer  months, however,  maintenance  visits twice
weekly were the rule, primarily to accomplish cleaning
of the sensors and the  sample cups and drawers. With
the exception of the chloride parameter,  it has been
fairly easy to keep the equipment  in calibration, as
long as the sensor are kept clean. The chloride sensor is
installed  in a  baffled cup because the probe is velocity
sensitive. The baffled cup  quite easily becomes fouled
by biological growths  and by accumulations of sedi-
ment. This condition is usually followed by abnormally
high chloride ion readings. The chloride sensor  is  also
the most difficult to calibrate probably because of its
long thermal  time constant and its velocity sensitivity.
It does work, however, as evidenced  by the increase in
chloride  concentrations observed when spring  thaws
flush the heavy amounts of  ice  melting salt  into the
river. Figure 45 is an example of this phenomenon.
   Maintenance and calibration since the intial start-up
of the monitors has been performed almost exclusively
by  part-time  and  full-time  temporary student  em-
ployees.  Initial  calibrations  were  performed  by the
project's engineering staff with  the assistance of the
manufacturers local service group. The project's engi-
neers then trained the  student employees.  Major
calibrations utilizing full-range standards  are not be-
lieved to be frequently necessary. The usual procedure
is to  measure the ambient status and  make whatever
small adjustments  necessary. Samples are taken  and
titrated for DO  and  chloride ion  in the trailer. A
portable  pH meter  and a  portable conductivity meter
are laboratory calibrated and used as  secondary cali-
bration standards. Temperature standard is a measure-
ment by an ordinary laboratory thermometer. Oxida-
tion-reduction  potential  is  always  calibrated using
standard  solutions  prepared  in  the laboratory  and
carried  to the trailer.  When major calibrations are
performed, the only additions necessary are: zero DO
solution since the river is usually near saturation, high
                                                    50

-------
                                           FEBRUARY 22, 1969
                                                         '
28
                                                             .
               RIVES QUALITY MONITOR
               AT THE MINNEAPOLIS
               WATER INTAKE
                                                                 30 mg/1 During Thaw
            -
'
20 ' S
•• 2
. O
t K





, c
>j;r'
v *
_£*

.*£*




Dissolved Oxygen =12 MG/L
V
"A .
^•v
: "V
i 2 -. Chlorides (salt) concentration = '7"
16 'I H
: Z
HI
. . O
; o
i 0
12 i J"
; O
^»°'&
; o
8 	




4__
0— -r
	 . __
_







o



, . - * J
P4



	
~"9I

	 0
>M ;
k-i^—O-







^
jC***









^

^ •' '
y
^
jf
if-




i
': i
| 1
Li ,
! 1
i !
1 i
0.4
















, 6PM j
i <

; : ' •"•'* •... :
"v ' ' ; • ' • •
• '*
i. \ . *-* • ^ '
"^ • :



i
i
i



ME


*^ 1
'. *"*>• : i '
	 i . . • _ "t • '^**)fc : ''
. ™ • : ' " •£t'j*r ' * — "~

i i i • l* , "&?•'.
\ Without Thaw 7 mg/1 !
I J a I

j I
1 3PM 	 12 N
                                          »dnoD loisme
             •  •
                         Figure 45. Chloride Data from River Monitor on a Thawing Day
   and low chloride standards, two or three pH buffer
   solutions, two or three conductivity standards and ice.
      Some of the parameters were acquired with dual
range readout to increase accuracy of transmission and
readout. These are DO, temperature and conductivity.
The ranges available are illustrated in Table 10 along
with the normal full range of each parameter observed.
                                                  Table 10
                                 RIVER QUALITY MONITOR PARAMETERS
            Parameter                                Instrument Range
    Dissolved Oxygen  	   0-24 mg/1
                                                       0-12 mg/1
    Temperature  	   30 - 80° F.
                                                       50 - 100° F.
    Conductivity	   0 - 600 umhos
                                                       0- 1200 umhos
    Chloride  	   0-48 mg/1
    pH	   4-10
    Oxidation-Reduction Potential	   —600 • +600 mv
                           Approximate
                          Extreme Values
                     Maximum       Minimum
                     14.2 mg/1        5.8 mg/1
                      83.0° F.

                     620 umho

                     20.5 mg/1
                      8.7
                      585
 32.0° F.

220 umho

 4.2 mg/1
 7.4
 310
                                                  51

-------
Findings and Conclusions
   River monitoring data acquired during and immedi-
ately after rainfall events of sufficient intensity to have
caused appreciable  combined  sewer  overflow  to the
river were closely scrutinized. The effort was made in
an attempt to evaluate the significance of the effect of
pollution slugs  originating from combined and  storm
sewer outlets in Minneapolis and Saint Paul on the
Mississippi  River. It was thought that depletion in
dissolved oxygen concentration might occur because of
the BOD  entering the river in storm  and  combined
sewer discharge. Because of the low  chloride concen-
tration in the river  water (5-10 mg/1),  it was thought
that an increase in chloride concentration might  be
observed.
   During and  after most rainfall events, no changes
were observed  in any of the parameters measured at
any of the river monitoring stations. There were a few
instances when  some reductions in  DO were  noted
following rain, but they  were very slight in magnitude.
Diurnal variations in concentration of DO were often
in evidence so  this, plus the possible effect of cloud
cover variability, complicates the situation and  in-
creases the uncertainty that any slight  change  should
be  attributed to pollution  from  combined or  storm
sewer discharges.
   Extremely low river flow conditions did not occur
during the  summer of  1969 despite  the fact that
rainfall for the  Minneapolis-Saint Paul  area was rela-
tively low. Table 11 shows monthly  maximum, mean
and  minimum  Mississippi River flows at St. Paul for
April through  November  of   1969. The  minimum
recorded  discharge at the Mississippi River at St. Paul is
632 cfs. At more critically low river flows, the  effects
of combined and storm sewer flows would be more
pronounced.
                     Table 11
        MISSISSIPPI  RIVER DISCHARGE
           APRIL -NOVEMBER, 1969
      UNOFFICIAL RECORDS FOR ST. PAUL

                         Mississippi  River Discharge
                             at Saint Paul, cfs
                      Minimum
                        Day
April	  32,400
May  	  20,600
June	  10,500
July	   9,600
August	   3,400
September	   2,700
October	   3,400
November	   3,700
   In  1969, data  from the river monitoring system
indicates that  the dissolved oxygen  resources of the
Mississippi River in Minneapolis and St. Paul were not
adversely  affected  by  storm  and  combined  sewer
discharges.
   Bacteriological  effects  were  likely  more  pro-
nounced, but the only data available are from periodic
grab samples that are completely uncoordinated with
rainfall events. Anderson (4) attempted to analyze river
coliform count and BOD statistically for recent years
in an attempt to gage the effect of improved regulator
surveillance and modifications. The limited number of
observations makes it difficult to base  any significant
conclusions upon these data.

Mean
88,500
39,000
14,900
14,200
5,900
3,400
5,000
5,300
Maximum
Day
148,600
59,000
20,200
18,500
9,800
4,500
5,700
7,100
                                                   52

-------
VIII.  COMPUTER ANALYSIS OF STORM RUNOFF
AND  MATHEMATICAL  MODEL  OF THE REGU-
LATOR-INTERCEPTOR SYSTEM
   In  November,  1967, the Minneapolis-Saint  Paul
Sanitary District entered into an agreement with the
University of Minnesota St.  Anthony Falls Hydraulic
Laboratory to have prepared a mathematical model of
the existing interceptor sewer system. Work toward
development of  this program has been reported in four
status reports (5, 6, 7, 8).
   Verification of the program, modification for use on
the PDP-9 computer at the Sanitary District, and the
rainfall loss-rate analysis not complete at the time of
preparation of the earlier  reports have all since  been
completed. Verification of the program was accom-
plished using real data obtained from the outplant data
logging system during the summer of 1969.
   The final report on this project entitlted "Mathe-
matical  Model  of  Urban  Storm  Runoff  and  the
Interceptor Sewer System  of the Minneapolis-Saint
Paul Sanitary  District" has been submitted to  the
Minneapolis-Saint Paul Sanitary District and is included
as Part II of this report.
                                                 53

-------
IX. UTtMAMUlM Ul- IHt SYSItlVI
Data Acquisition Programs
Data acquisition or scanning is under control of a
group of programs collectively named "RTIME", short
for REAL TIME. The RTIME routine is called up by
issuing a one line command from the I/O teletype. The
operator then initializes the computer internal clock by
typing in the hour of the day, the minute of the hour,
and the day of the year. The computer internal clock
then maintains the time, which is entered on all reports
and stored with all data. RTIME will initiate scanning,
either under control of the computer internal clock, or
by receiving manually initiated signals from the inter-
face control panel. As each address is scanned, an
electrical analog signal from the telemetry is digitized
and read into the computer which referring to the
appropriate list of scaling factors (Figure 46), stores it
in a file, and prepares a written report. At this time
there are five outplant scanning routines used in
RTIME. Normal scanning employs four:
1. RIVMON which scans the river quality monitors.
2. REGCTL which scans the regulator control and
monitoring stations.
3. INCPTR which scans the interceptor monitoring
stations.
4. RAIN RP which scans the rain gages.
PIP V78
>T TT  - OK RAINRP LST
0 01 29000 0641 UM-RG .421 -.284
0 01 30030 1144 VAD-RG .449 -.431
0 01 31000 1141 HAZ-RG .460 -.520
0 01 32000 1644 FIR-RD .429 -.421
0 01 33000 1441 VIL-RG .437 -.435
0 01 34000 1641 EDN-RG .436 -.510
0 01 36000 1244 MAC-RG .420 -.416
001 41000 1511 WTP-RG .408 -.151
0 01 42000 0633 EUS-RG .477 -.384
0 00 77777 7777 END-TAP
Figure 46. Rain Report Scale Factor List
The fifth routine, ALARM, can be used to scan
extra addresses at almost all stations which are called
"alarm points." When any of these scanning routines
operate, the data is added to a file stored on the
magnetic disk. These files are usually allowed to
accumulate for 24 hours from 8:00 A.M. to 8:00 A.M.,
closed, and given a distinctive name be which they can
later be accessed. Figure 47 is an example of one of
these files that was printed on the line printer. The
names used are made up of 9 characters. For example
RISR70 140 is a name for a file of river quality
monitoring data (Rl) occurring in a sorted but un-
packed format (SR) for the period in the year 1970
(70) approximately 24 hours in length beginning at
8:00 A.M. on the 140th day of the year (140). Should
this file later be packed, a process by which its
magnetic storage requirements are reduced, it would be
renamed RISP70 140.
182 9 35 1426 CAM-TL 30.620
_J,8 2_1 JL_2_2__.1A2.6. CAMrJL 	 3 3 .045 	
182 12 22 1426 CAM-TL 32.970
._18_2_14._L2. 	 1.4 2 6 _ CAM- TL 	 3 2 . 3 74 ._
182 14 46 1426 CAM-TL 31.702
_1.8.2__L6._10.._1426 .CAM-TL --30.695 —
182 17 4 1426 CAM-TL 30.770
_1.8.Z_1.8 	 4 	 1.426 CAM-TL 	 30.832 —
182 19 20 1426 CAM-TL 32.672
__18-2..20_2.0 	 1426 CAM-TL- _... J3 .008 -
182 21 20 1426 CAM-TL 34.350
.-.182 22. 20_. 1426 - CAM-TL - -34.537 -
182 23 20 1426 CAM-TL 34.201
— 132_23 5.4 ._.. 1426 CAM-TL 	 33.754 .
183 0 4 1426 CAM-TL 35.432
183 018 ..1426 CAM-rlL 	 38.901--
183 0 28 1426 CAM-TL 42.855
.18.3.... 0...40— 1426-CAM-TL 	 43.862 -
i83 0 50 1426 C3M-TL 41.922
	 18.3 	 1 	 0._.1426 .CAM-TL 	 40.207 	
183 1 10 1426 CAM-TL 38.528
..._18_3.._L_20 	 1.426. CAM- TL 	 3 6. -73 8 -
183 1 30 1426 CAM-TL 35.880
	 183—1-40 1426 CAM-TL- 34.350-
183 151 1426 CAM-TL 33. 791
..— 1 83 	 2- — 1 	 142 6- CAM- T4_ 	 32 . 9 33
183 2 11 1426 CAM-TL 31.963
._..!& 3 — 2-2-1- -1426 CAM-TL 31.478
183 2 31 1426 CAM-TL 31.068
_183 — 2-4 l_.. 1-426 -CAM-TL- — -30 . 285 -
183 3 16 1426 CAM-TL 28.233
... 18.3 	 3-2-6 	 1 42-6- CAM-TL 	 28.494 -
183 3 36 1426 CAM-TL 27.189
183 	 3 46 1426 CAM-TI ?7 O4O.
183 3 56 1426 CAM-TL 27.152
— 18-3 	 4 — 6 	 1426 CAM-TL 	 26.-965-
183 4 17 1426 CAM-TL 27.040
	 18_3 	 4 27 1426 CAM TL 26 704
183 4 37 1426 CAM-TL 25.958
_.183 	 4-4.7—1426 :CAM-TL 	 26.406-
183 4 57 1426 CAM-TL 26.033
—1.8 J 	 5 	 .7—1426 -CAM-TL 	 25.883-
183 5 20 1426 CAM-TL 25.846
U3 5 30 1426 fAK-Tl 2S R3V
183 5 41 1426 CAM-TL 25.846
	 183. .6 11— 1426. CAM-TL 26.182
183 6 24 1426 CAM-TL 27.077
	 18J-_ 6-34_-1426~CAM-TL 27.077
183 644 1426 CAM-TL 27. 338
	 183 	 6-5-4 	 1-426- CAM-TL 	 27.450-
183 7 4 1426 CAM-TL 27.562
	 18J 	 7-15. ...1426 - CAM-TL 	 28. 159
183 7 28 1426 CAM-TL 28.308
._ 183 7 38 1426 CAM-TL 28.644
<•;•»• •er»;i^u»»-j-fc*r.-»«,sv»7^r* *?it " • -^rrv-vt' *•-*"'• •*i*-v»«vl.r" •,-•*— **«'~^"*
182 9 36 1435 CAM-OL 28.548
....182 11 23. 1435-. CAM-OL 30.407
182 12 23 1435 CAM-OL 30. 152
Figure 47. Data File as Stored on the Disk
55

-------
   The River Quality Monitor Report available immedi-
 ately after scanning is shown in Figure 48.

   The  Regulator Control Report is shown in  Figure
 49. The  station names  used   are familiar  to the
 operating  personnel and tables are available to assist a
 less familiar person  in interpreting the data. Camden
 Park through Oak Street are regulators in Minneapolis,
 and Rice  & Rondo through Eustis-West are regulators
 in Saint Paul. Their locations are referenced in Figure
 26. Trunk level (TL) is the depth of flow in the trunk
 sewer just upstream of the regulator. Outlet level (OL)
 is the depth of flow in the dry weather outlet pipe
 which carries diverted  flow from  the regulator gate
 controlling flow into the outlet pipe expressed as per
 cent open. Storm  gate (SG) is  the inflation pressure
 inside the Fabridam. At Camden Park regulator, where
 there is no Fabridam, SG represents per cent open for a
 bypass gate.

   The  Interceptor Report is shown in Figure 50. The
names used refer to the locations of the monitoring
stations, which are referenced on Figure 51. Level is
depth of flow in the interceptor  in inches. The section
sizes and shapes are referenced  in Table 12. Bypass
gates at the ends of the three Minneapolis interceptors
are  remote controlled and  their  positions monitored.
The other two gate positions on the report refer to
gates  at the entrances to the  two parallel Venturi
meters on the Northwest interceptor. Usually only one
meter is in use, but during rain periods the second gate
can be remotely opened and  two meters used. This is
accomplished  automatically when flow  through  the
other  Venturi meter exceeds  a  set point value. The
flow rates in MGD  and levels at Bypass and Headhouse
are  actually  inplant data received from old metering
and telemetry equipment in use  before the Regulator
Demonstration Program.
  The Rain Report is shown in  Figure  52. The rain
gage names  used  represent  their locations  and  are
referenced on Figure 25. In addition to the data from
                        MINNEAPOLIS-ST.  PAUL SANITARY  DISTRICT

                              RIVER  QUALITY  MONITOR  REPORT
YEAR  70   DAY  OF  YEAR  164 TIME   20  11
INDUSTRIAL  MOL  (RQM1)

FARM  BUREAU  (RQM2)

HIGH  BRIDGE  (RQM3)

FT. SNELLING   (RQM4)

MPLS.  WATER  WORKS  (RQM5)
CL
11.5
10.7
10.6
7.9
2.1
COND
492.0
476.7
409.9
288.0
198.5
DO
7.6
7.7
8.1
8.8
8.6
ORP
513.2
490.6
448.7
387.9
516.0
                                                                                  PH
                                8.2

                                7.9

                                8.3

                                8.0
TEMP

  76.1

  76.4

  77.1

  79.2

  74.8
                                  Figure 48.  River Quality Monitor Report
                                                56

-------
                  HIWHEAPOL3ES-ST.  PAUL SARITAHY DISTRICT
                            REGULATOR CONTROL
                                SEWER SCAN
YEAR 1970  DAY  146  TIME   9  33
CAMDEH PARK
PORTLAND + WASHIHGTH
26TH + SEABURY
39TH 4- MIWHEHAHA
38TH •!• EDMUiJD
31ST 4- RANDOLPH
2ND  + MAIN
OAK STREET
RICE + RONDO
TROUT BROOK
PHALEH CREEK
BELT LIKE
OTIS H- MARSHALL
ST. PETER + KELLOGG
EUSTIS - EAST
EUSTIS - WEST
TL
INCHES
29.1
16.4
13.7
20.1
16.7
17.5
28.1
17.0
3.S
2.6
44.7
1.7
10.4
0.9
2.4
1.8
01.
INCHES
29.2
39.7
2?.. 3
19.0
25.8
0.0
15.6
29.3
1.5
4.6
61.7
4.3
5.1
3.8


06
7. OPEN
92.5
95.6
85.9
91.5
91.4

77.8
94.1
102.2
98.0
104.2
101.4
97.5
100.1


SG
PS I •
-0.7
1.4
1.0
0.3
0.4
-0.3
1.6
1.4
0.2
0.1
0.6
0.1
0.5
0.5
0.1
-0.4
                       Figure 49. Regulator Control Report
                                                                     WL
                                                                   INCHES
                                           -2.7
                      MINHEAPOLIS-ST. PAUL SANITARY DISTRICT
                                  INTERCEPTOR REPORT
    YEAR  1970   DAY  146  TIME  9 39

                          LEVEL   BY-PASS  GATE  1
                            GATE 2
                              MGD
     19TH AVE  4- 2ND ST.
     FRANKLIN  + RIVER
     44TH 4-  HAWADAHA
     10TH +  MARSHALL
     ROBLYN  4-  FAIRVIEW
     PINE 4-  SPRUCE
     ST. CLAIR * 7TH
     STEWART + TUSCARORA
     OTIS 4-  MARSHALL
     MPLS.  NORTHWEST
     MPLS. SOUTHWEST
     MPLS. EAST
     MPLS. TOTAL
     BYPASS
     HEADHOUSE
     PLANT EAST
     PLANT  WEST
     PLANT  TOTAL
 26.5
999.9
 37.8
 30.1
 56.9
 64.2
 12.2
 25.7
 24.7
 18.0
 19.7
 50.0

 -0.2
  2.8
999.9
  0.8
  0.8
99.4
999.9
 44.9
 30.8
 35.7
111.5
                                      122.6
                                       8.3
                                      130.9
                            Figure 50. Interceptor Report
                                   57

-------
01
00
                •  INTERCEPTOR MONITORING STATIONS
              112   19TH AVE. AND  2ND  ST.
              FR   FRANKLIN AND RIVER RD.
              NDA  44TH AND NAWADIHA  BLVD.
              NE2  10TH AND MARSHALL
              JT4  ROBLYN AND FAIRVIEW
              ITS  PINE AND SPRUCE           B
              «71  ST. CLAIR AHD  «. 7TH
              W72  STEWART AND TUSCORQRA
              MR   OTIS AND MARSHALL
              NW   MPLS. NORTHWEST
         -n   S*   MPLS. SOUTHWEST
              ME   KPLS. EAST
;;    ,-.  <
 A           -        :       {
                       r-'"'
                                                                             HPLS.S.W.
                                                                             INTERCEPTOR mil  INTERCEPTOR
                                            V $ST+? «7-
                                             ' 7-1S
                                             "  n	L

-------
                        MINNEAPOLIS-ST.  PAUL SANITARY  DISTRICT
                                           RAIN  REPORT
YEAR  1970    DAY
PAST  READNG  DAY
284
284
TIME
TIME
15
14
51
51
ST.  PAUL  CAMPUS
LAKE VADNAIS
HAZEL  PARK
WEST SIDE
GOLDEN  VALLEY
EDINA
AIRPORT
MPLS.  WATER WORKS
EUSTIS
                   Table 12
     INTERCEPTOR LEVEL MEASUREMENT
            LOCATIONS AND SIZES
 Location                      Size
 MINNEAPOLIS:                4'-6"
 19th and 2nd Streets             4'-0" x  4'-6'
 Nawadaha                      3'-6" x  6'-0"
 S. W. Meter                    6'-0" x  6'-0'
 Franklin                       9'-0" x  9'-0"
 N. W. Meter                    10'-3" x 10'-3'
 East Meter                     6'-0" x  6'-0'
 Tenth and  Marshall              7'-0" x  7'-4'
PRESENT
READING
0.064
0.044
0.011
0.460
0.249
0.202
0.132
0.258
0.072
Figure 52.
PAST
READING
0.048
0.030
0.007
0.449
0.215
0.189
0.117
0.235
0.049
Rain Report
                                              RATE
                                             IN /MR

                                              0.016
                                              0.014
                                              0.004
                                              0.011
                                              0.034
                                              0.013
                                              0.015
                                              0.023
                                              0.023
                                                         INCHES
                                                         TOPAV

                                                         0.037
                                                         0.055
                                                         0.014
                                                         0.03*
                                                         0.100
                                                         0.0*1
                                                         0.04?
                                                         0.109
                                                         0.054
 SAINT PAUL:
 Mississippi River Boulevard
 Tuscarora
 Fairview
 Pine and Spruce
 St. Clair
       3'-9"x 5'-10
       2'-6"x 6'-0"
       3'-9"x5'-10"
the scan run, this report also contains the data from
the last  prior  scan, the  rainfall  rate during  the
intervening period in inches per hour,  and the total
accumulation since the file day began. The readings are
inches  of  water  equivalent to the weight  in  the
collector bucket.
   A routine called SCNFRQ for SCAN FREQUENCY
is a part of RTIME and can  be used to vary scan rates.
The operator calls  the  routine  by depressing  the
appropriate  switch on the  operator's  console. Any
interval between scans can be indicated in seconds; the
various routines can and often are scanned at different
rates.  Figure 53  shows the teletype commands that
would be used to change scan frequencies from hourly
to rates commonly used during rainfall.
                      Gates  and Fabridams  can  be contolled remotely
                    both under program control and manually through the
                    interface  cabinet. Any address can  be  manually ac-
                    cessed and the analog voltage read so monitoring and
                    control remain possible even without the use of the
                    computer. The program by which control functions are
                    normally   exercised   is called  GATCON for GATE
                    CONTROL.  It can be manually initiated or called by
                    the computer internal clock. As presently constituted,
                    the program  will access any gate or  Fabridam whose
                    address  is written   in a   list  called SETGAT  LST
                    (Figure 54),check the position of the gate or inflation
                    of the Fabridam, and brings the position or inflation
                    pressure into agreement with a set point that it also
                    reads from the list if the actual value differs from the
                    set point  by  prescribed limits. A GATE control report
                    is shown  in Figure 55. The remark "INOPERABLE" is
                    printed if the gate position reading does not reach the
                    set point value ± 2% within a period  of time specified
                    on SETGAT  LST.

                    Modes of Operation
                      During dry weather the system  is used  for data
                    acquisition,  data  reduction and processing, and for
                    programming. Normally, scans  run at hourly intervals.
                    The gate  control program, GATCON, runs every three
                    hours to  maintain all outlet gates 100% open and all
                    Fabridams inflated.  While most of the Fabridams are
                    not  effective in  or required  for  diversion of  dry
                    weather flow, they are better kept inflated because of
                    the  rather lengthy  time  that  would be required to
                    inflate 15 fully deflated dams—about 75 minutes—and
                    because they need be deflated only during very intense
                    rainstorms that do not occur frequently.
                      Each  day at  about  8:00  A.M.  the  data  files,
                    RIVMON DLY, REGCTL DLY, INCPTR DLY, AND
                    RAINRP  DLY, are renamed  with  a unique  name
                                              59

-------
MONITOR V49

SE RTIME


 ENTER- 1-13  MN DAY

 08 45 146


ENTER ID AND RATE-CX  XXXXX)


ID PROGRAM

 2 RIVh'R
 3 REGTR
 4 INCPT
 5 RAIN  •
 6 ALARM
 7 SETGT
 8 SLUDGE
 9 EXIT

ENTER ID AND RATE-CX  XXXXX)

2 01800

OLD RATE WAS   3600
ENTER ID AND RATE-CX  XXXXX)

3 00900

OLD RATE WAS   3609
ENTER ID AND RATE-CX  XXXXX)

A 00600

OLD RATE WAS   3600
ENTER ID AND RATE-CX  XXXXX)

5 00300

OLD PvATE  WAS   3600
ENTER ID AND RATE-CX  XXXXX)
PIP V78
>T
1
2
3
A
6
7
8
9
10
11
12
13
IB
19
20
21
22
23
24
31
32
33
35
36
38
39
77
TT CA)
000225
000227
000245
000247
000305
000307
000425
000427
000505
000507
000545
000547
000745
000747
001045
031047
001025
001027
001226
031307
001345
001347
001427
001547
001745
001747
007777
- OK SETGAT LST
M2-SG
M2-OG
PW-SG
PW-OG
OAK-SG
OAK-OG
SPK-SG
SPK-OG
RR-SG
RR-OG
TB-SG
TB-OG
KD-SG
KD-OG
BL-SG
BL-OG
PC-SG
POOG
NWI-G1
M39-OG
SEA-SG
SEA-OG
CAM-OG
RAN-SG
E38-SG
E38-OG

Figure 54. Fabridam and Gate
70 180
98 0.60
65 180
100 063
70 180
96 010
70 180
100 060
70 243
100 060
70 240
85 030
70 180
98 060
070 249
92 060
65 180
95 010
95 060
90 060
68 180
96 030
95 060
70 180
70 180
92 015

Control List
Figure 53. Use of SCNFRQ Routine to Control
       Scanning Frequency
                                60

-------
                                   MIHMEAPOLIF-ST.  PAUL SANITARY  DISTRICT
                                               GATE CONTROL  REPORT
           YEAR  1970   DAY  146   TIME    9 45
GATE NO
1
2
3
4
6
7
8
9
10
11
12
13
18
19
20
21
22
23
24
31
32
33
35
36
38
39
ADDRESS
0225
0227
0245
0247
0305
0307
0425
0427
0505-
0507
0545
0547
0745
0747
1045
1047
1025
1027
1226
1307
1345
1347
1427
1547
1745
1747
MM EN ON 1C 7.
M2-SG
f/2-OG
PW-SG
PW-OG
OAK-SG
OAK-OG
SPK-SG
SPK-OG
RR-SG
RR-OG
TB-SG
TB-OG
KD-SG
KD-OG
BL-SG
BL-OG
PC-SG
PC-OG
NWI-Gl
M39-OG
SEA-SG
SEA-OG
CAM-OG
RAH-SG
E38-SG
E38-OG
OPEW
69
97
70
99
71
94
68
98
72
101
70
84
72
98
77
90
67
95
94
89
68
93
95
76
71
89
WAS
67
87
70
96
72
95
86
98
93
101
93
81
86
98
91
90
80
95
94
87
79
93
96
99
87
90
DESIRED
70
98
65
100
70
96
70
100
70
100
70
85
70
98
70
92
65
95
95
90
68
96
95
70
70
92
                                                                            REMARKS
                                                                           INOPERABLE
                                                                           INOPERABLE
                                     figure 55. Gate Control Report
describing the file's content and its date. The next scan
run will begin  new daily files. A hard paper copy of
each file is then obtained from  the line printer. The
data files are usually quickly scanned to look for data
indicating problems—either real problems or problems
in monitoring  equipment. The  scaling routines will
record a 9's field, 9999.9, in place of any figure outside
a normal  range to make it easier to spot  unusual or
abnormal data.
   If the telemetry system does not successfully access
a specific address, it is noted by a failure of  the address
to "confirm." Confirmation of an address occurs when
the outplant station returns to the base station a pair
of  tones  corresponding  to  those sent by the base
telemetry  as part of the unique address. The failure of
an  address to confirm is indicated  by a 7's  field,
7777.7, so it is easily recognized, and this failure does
not generate misleading data.  The paper  copies are
bound serially and retained for later reference.
  The data is also reduced using a DATA ANALYSIS
program  named  DAIMALS.  This program  reads the
entire  data  file  and,  after  calculations,  prints the
address, the number of times the address was scanned,
the  maximum observed value of the parameter at that
address  and  the time  at  which  it  occurred, the
minimum value and the time at which it occurred, the
mean value, the standard deviation,  the  number of
times that address did not confirm and the number of
values  falling outside  a set range.  Figure 56  is an
example of a DANALS  Report. These reports are also
bound serially and retained.

  Renamed, closed data files remain on the magnetic
disk surface, "but must be removed so that adequate
room is available on the disk for new data files and
programs. Magnetic disk is too expensive to be used for
data storage and  the high speed capabilities of the disk
are  not required for data storage.
                                                  61

-------
MINNEAPOLIS-SAINT PAUL SANITARY DISTRICT
-P-*4*- A"AL*VIS RfMT
FOR 07/01/70







31
i
CJl
en
a
3
3
a.
3D
n
e









POINT
641
__11A^..
1441
	 J64V.-
. — 14 .4 I . .
	 1641 ..
	 1 2.4.4.. .
	 1SJ.L.
63J-.
._ . L.42A-.
	 1AJ5-
_. -1*27
14PS
	 Z4A_..
	 2Ji5-
	 i4i_
	 -245_
. 	 UL46_
11SS
J347. .
1345
i ins
	 111 3 ..
1307
1305
NAME
UM-RC

HA2-AG
F 1 R-fiO
VIL-8C
EON-ftC
-HAX-_aC
..UIP-ao
EUS-RC
CJLM-.-U. .
-CAH.-OL-
CAM-.OG
r«N-pp
.. PU-Ti. -
._£U_-01.-.
__EU^)iU_
_-.PU.-SG.
	 SEAr U_
^F H-QL
_. SEA-OC.
SEfl-SC

	 B3S.-JL.
Miq-nr.
H39-SC
NO.
- -READ—
. -56- -
56
56
56
56
56
56
.56
56
54
54
54
5t
54
54
	 St..
. 54
.- 	 54 -
5^
54
—5V
54
54
. 54.
54
MAX.
0.4
0.6
1.0
0.7
0. 5
0 .-6
0-. 9
0.5
4J.*
39.3
99. 0
1 . 2

- 62 . »—-
- - 96,9 	 -
- 1.5 -
- . 59.5 -
4), 5--
90. 7
— -4- 2 -
82 9
27.4
114 5— -
1. 1
TIME
OAT -MR HN
>83
183
183
183
183
183
183
183
183
183
183
182
183
183
4-83-
183
183
1-83
183
1-8-J
— HJi-
183

182
7
3
-7-
5
3
7
4
4
0
0
0
1 8
1
1
3
6
0
1
7
A

1
-1^7-
19
47
43
48
40
24
14
2
27
40
40
18
5
11
11
56
12
19
j
5
-15
—31
32
- 6-
22
HIM.
0.5
0.3
0.5
0.2
0.4
o.*-
0.7
0.4
25.8
21.0
V4 .0
-1 .4
15.0
23.9
94. >
0.9
12.7
. ^j — ^_
37.7
0 1
14" 6
14.6
1 0*- 8-
0.5
TIME
OAT HR UN
182 18
182 23
182 23
182 23
182 23
182 23
182 11
182 17
18 3 5
18 3 5
18 3 3
182 21
183 6
183 4
182 18
183 2
183 5
' 18 3 5
182 14
183" 2
- —is) - 7
183 6
	 183 7
182 14
1 3
40
40
40
49
30
33
14
20
41
36
21
24
28
5
32
21
4 2
48
4 2
1 6" "
56
-16 ~
48
PAGE

MEAN
0.7
0.5
0.8
0.6
O.S
0.6
0.9
0.5
31. 2
26. 3
97.0
-1.0
26.-J
37.9
95. 7
1 .2
21.5
27 9
86.6
0 4
- 39 [
21.2
110.8
0.8
I

S.D
0
0
0
0
0
0
0
0
4
4
1
0
11
11
0
0
10

10

22
4
1
0



. 1
. 1
.2
.2
.0
. 1
.1
. 1
.7
.2
.2
.5
.8
.6
. 7
.1
.8

.1
3
3
.2
. 8
.2




NOT 1 NO.
CON ORG
0
0
D
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0


3
.p
3
0
o
0
0
0
0
0
0
0
c
0
0
0
0
0
0
0

0


0
o
0

-------
   If it  is desired  to retain a file, it can be stored on
magnetic tape or on paper tape. It can be stored in its
initial format or it can be packed. A packing routine
has been prepared which  reduces the "space" require-
ment for data storage  by about  60%. The  magnetic
tape used  is  referred  to  as  DECTAPE,  and  the
computer is equipped with two DECTAPE  drives. A
dectape  library of  128  reels  is in use for program
backup   storage as  well  as data storage.  Each reel
contains 150,000 eighteen bit words on 260 feet of 3/4
inch magnetic  tape. Punch  paper tape  in use is the 1
inch fan fold type. A paper  tape  library contains
programs backup and diagnostic and test programs.
   Almost   all  data collected  in  the  first year  of
operation was stored on magnetic tape. It is impractical
to consider continuing this practice so, after  inspec-
tion, verification that a paper copy exists, and com-
 pletion of the data analysis program, some of the older
 files have been deleted from magnetic storage. All files
 containing data acquired during and immediately after
 rainfalls have  been  retained,  however, so further data
 analysis could easily be accomplished.

    Another  data processing  program that  has been
 developed is a data plotting routine named GRAPHP.
 This program  will plot data on the line printer from all
 or  any part  of any  particular file.  The time  axis
 (horizontal) plotting interval and the data axis (verti-
 cal) full scale  value are selected by the operator, so the
 plotting  routine  is quite  flexible.  Usually  only the
 rain-day  data was  plotted   since  it  was  carefully
 scrutinized as part of the project evaluation. Figure 57,
 Figure 58, and Figure 59 are examples of plotted data
 from the rainfall occurring on July 1 and 2, 1970.
                                                    1  IS FULL SCALE
             	o •
              line 20
                                    22  o
                                     112
0 0
181
1  0
IB]
)  o
tti
                           Figure 57.  Machine Plotted Rain Gage Data for July 1-2,1970
                                                    63

-------
                                       100 IS FULL SCALE
HUE 20  o     n  o
OAT   IA2        1*2
22  0     2] 0
 182       182
00       10       20       30      40
181        i»        183       193       19;
  Figure 58.  Machine Plotted Interceptor Sewer Flow Depth Data for July 1-2,1970
                                       130 IS FULL SOLE
fI ME 20  0     210
Dir   |»2       1S2
                    00       10      20
                    la)       103       183
                            30       to
                            183        183
       Figure 59. Machine Plotted Trunk Sewer Flow Depth for July 1-2,1970
                                       64

-------
   A  second   plotting  program  called  OGIVE was
developed  for our system. An ogive is a type of curve
used to graphically portray frequency distributions in a
cummulative manner.  It shows the per cent of time or
frequency  that a measurement of a certain magnitude
      a
                       or greater occurs. This technique was used to portray
                       utilization  of interceptors in studies carried  out in
                       1960 (2). Figure 60 is an ogive prepared from 1960
                       data and Figure  61 is an ogive prepared from 1969 data
                       at the same location in the interceptor.
                                             MAXIMUM STORM FLOW
                                                4-23-40  4.3'
                                         MAXIMUM DRY WEATHER FLOW
                                                9-23-60  3.3'
                                            6-0" X 6'-0" SECTION
                                        MINNEAPOLIS EAST INTERCEPTOR
                                                VENTURI MCTER  I
                         30                  aO   ^   60                 80
                                  %OF TIME DEPTH OF FLOW ECUALS OR EXCEft'S
                         Figure 60.  Ogive-1960 Data for Minneapolis East Interceptor
   OC1V6
  	PIUIOO ;OVERCD-«P«ll  H70
POIKI' 7J6 KE-II      PIPE MEICHT (INCHES!-  1> .0    NO. Of OSSERVtl IONS' 1*54
      LMC? io tv _ 7.2.000
                75.0

                72.C
                                                                MAX. LEVEL
    E_»M.rt IN INCHES Jb.O
                                                                                                     72.0"
         	0.0
                                         1 OF lint DEPTH OF FLOW EQUALS OR EXCEEDS
                         Figure 61.  Ogive-1970 Data for Minneapolis East Interceptor
                                                     65

-------
   Mounted in the  interface cabinets in the computer
 room  and in the Plant Operator's  office  are  eight
 two-pen miniature strip chart recorders. Data addresses
 assigned to these pens are recorded on the strip charts
 moving at one inch  per hour. The pens are updated at
 8-second intervals, so if all  16  pens are assigned, each
 address's data value is updated  every 128 seconds. The
 routine which is used to assign addresses to recorder
 pens   is  named   SEL   REC  for  SELECTION-
 RECORDER.  Figure 62 shows  the operator command
 string at the teletype which is used to assign addresses
 to the 16 recorder pens.
                     Table 13
  SCANNING TIME REQUIREMENTS - MINUTES
 tc

 MONITOR  V4B

 SE RTIME


  ENTER-  HR  MN DAY

  08  50  146


  ENTER  RECDR  AND POINT NO.-(XX  XXXX)

 01 1511

 NEXT

 02 1441

 NEXT

 03 1641

 NEXT

 04 1244

 NEXT

 99  (99 Exits Program)

 Figure 62.  Assigning Data Addresses to Recorder Pens
           Using SELREC
   The  RTIME  program can  be operated with  or
without the reports being concurrently  typed  out on
the logging teletype. When the report is typed, more
time is required to scan all points. Table 13 shows the
time required for each scan, with and without  output
to the teletype.
 RAINRP
 IIMCPTR
 RIVMON
 REGCTL
   TOTAL
      With         Without
Teletype Output  Teletype Output
        1             0.5
        2             1.5
        2             1.5
        6             4.0
       11
7.5
   Normal operation of the scanning program, RTIME,
employs the disk as the file device. The flexibility of
the system was demonstrated during the early part of
the summer of 1969 when various equipment bugs led
to use of the scanning program in various ways. When
the disk was unavailable, a modified program was used
which wrote the data file on magnetic tape. When the
Dectapes were unavailable,  a version of the scanning
program that used the ASR 33  TTY  as  the output
device  enabled  the teletype's  paper  tape punch  to
record the data in processible form.

Modes of Operation
   Operation of  the system  during rainfall events can
be  accomplished with or  without  the  use of the
mathematical  model. Personnel familiar with the sy-
stem and  its operation are usually on duty during
periods of rainfall. The mathematical model program
incorporates RTIME, and can handle the acquisition of
data. The model also  produces reports, records data,
and makes recommendations for operator  action. The
operator can then change gate positions either from the
teletype keyboard or manually from  the operator's
panel.  The  model  then checks  the  changed  gate
positions with the  current  data  and  will show the
results  of  operator action.  The model can  alert the
operator  when  excessive  flow at any  point  in  the
interceptor system is expected.
   Use  of  the  model  as an  operator aid  during  a
rainstorm would  likely be as follows. The operator has
loaded  the model program, and  it is scanning rain
gages. The model is requested to use the rain gage data
to calculate a  runoff  hydrograph anticipated at  each
regulator,  perform a diversion  analysis  at each regu-
lator, calculate volumes diverted to the interceptor and
to the  river, and route the flows through the  inter-
ceptor system, checking for excess flow at each inlet
and at other key places.
   The  operator  can choose to have any or all of this
data printed on the line printer. Most likely, he would
first check the results of the diversion analysis to see if
any overflow  to the  river is predicted. With this
information he is alerted as to which trunk sewers will
be carrying the  largest storm  hydrographs and  when
                                                  66

-------
the storm flow will be arriving. He can then check real
data  at appropriate  times  to  verify  if  Fabridam
deflation  is  required  to  prevent  surcharging of the
trunk sewer.

   The  routing portion of  the program will print the
flow hydrograph at any point in the interceptor system
where  the  predicted flow exceeds the capacity of the
pipe. Assume that  in this  storm, the model warns  of
excessive  flow  at  the Minneapolis East Interceptor
Meter Station. The operator,  noting the time at which
this will occur, checks a bypass priority table, decides
where  flow release  is preferred, and tests the result  of
lowering the  Fabridam using the model.

   After performing the diversion analysis and routing
with the changed  regulator condition,  the model  no
longer  predicts  excessive  flow, so  the  operator now
deflates the Fabridam.

   By  this time, the  model  has  accumulated  more
rainfall  data and  can update  its predictions.  The
operator can also check real  time flow data from the
regulators and interceptors.

   The system can also be operated manually, without
use of the model. The operator has assigned rain gages
to the trend recorders and can check rainfall accumu-
lation every  two minutes. Depending upon  the rain
gage results,  there will be  certain  trunk sewers where
depth data  should  be checked.  Operating experience
from 1969 has indicated which trunk sewers are  likely
to carry large flow hydrographs. Trunk sewer level data
can be  assigned to  trend  recorders so that locations
where  Fabridam deflation might be  required can be
watched. By  watching trend recorders and data logging
reports,  the operator can keep abreast  of what  is
happening  and control gates  and dams to accomplish
the best operation within limitations of the interceptor
system.

   Operation of the system during 1969  was usually
less routine.  The mathematical  model  had  not been
calibrated and  was not available for use. The  trend
recorders were not usable either. Early operation then
consisted  of insuring  that  data  scans were run fre-
quently and checking  rainfall  accumulation, trunk
sewer level, interceptor  level, and Fabridam pressure
data as they were logged on the logging teletype.
   The  only operator  actions taken  were to deflate
dams  when the depth of  flow  in  the trunk sewer
approached  the  sewer  crown. This was not required
very  often. During extremely  heavy  rainfall, flow
conditions at many regulators caused rapid increases in
Fabridam  pressure causing them to deflate through the
water column. It is always preferable to relieve air in a
controlled  manner  using the deflation valve,  but the
storm water hydrograph arrives very suddenly at some
regulators after intense  thunderstorms. Trunk level data
could be obtained every six minutes, but usually only
about every ten minutes because rain gage  and inter-
ceptor scans were also  run. Therefore, it was often not
possible to note a rapidly rising trunk level and react to
it  in   time to  prevent  the overpressuring  of the
Fabridam  and its subsequent automatic deflation.
   Better operation of  Fabridams is now possible, and
fewer instances of uncontrolled deflation should occur.
Controlled deflation from the base station will be more
timely because of the use of the rapidly updated trend
recorders. Also, local  pressure  sensing  switches have
been  installed  at  several  regulators  to open  the  de-
flation valve and relieve pressure before it has increased
to the level of the water column.
   This type  of  "automation"  could be  expanded
somewhat with  more local  control loops, but at this
time it is  not necessary. The computer and supervisory
control system are also used at several key locations to
automatically control gates and Fabridams. The second
Venturi meter  on  the  Minneapolis Northwest Inter-
ceptors is brought into  use when flow through the
other meter exceeds a set point, and  the Fabridam at
the Beltline Regulator  in Saint Paul is deflated when
the flow depth  measured at the head end of the dry
weather outlet pipe exceeds a set value. More gates and
dams  could be automatically controlled in this manner.
                                                    67

-------
X. SYSTEM EFFECTIVENESS
   Data accumulated in operation of the system during
rainfall events were analyzed to determine its effective-
ness  in  reducing combined sewer overflow pollution.
Effectiveness could be evaluated by several approaches.
One  used  was to enumerate the overflows prevented.
Another was the duration of overflow prevented, and a
third was  the volume of overflow prevented. All three
approaches were used.
   Three  distinct bodies of  data  were analyzed. A
number of rainfall  events occurring during the 1969
rainfall  season  were  well  described. This  data  was
analyzed  over the winter  season. Overflow  reduction
from the  1970 spring thaw runoff was also evaluated.
The  third  group of data was rainfall events occurring
during April and May, 1970. Data from  1969 and 1970
rains were combined and seasonal figures  were extra-
polated.
Rainfall Events Analyzed
   The  initial operating season used  for evaluation of
the project was that of 1969. Considerable effort was
required  to  ready  the system for  operation  and
acquisition of data, since full-time use of the system
was  not acquired from the contractor until January,
1969.  Scanning and data acquisition  programs  were
ready  in  time but  the winter weather  had  caused
further  complications. Heavy snowfall over most of
Minnesota resulted in predictions of record or near
record flood crests on  the  Mississippi River in Saint
Paul.   In  1965 the  greatest flood  on  record  had
occurred and inundated the Sanitary District plant site;
there was a very strong  possibility that  this would
recur.  Even flood levels much  lower  than predicted
would  have submerged the access road to the plant and
also  submerged the telephone  cable  leading  to the
plant.  It   was  certain  then  that  the flood would
interrupt  data acquisition for at  least two weeks. With
project plans already behind schedule, and not wishing
to have to wait until 1970 to acquire  operating data,
the  Sanitary  District  moved  the  equipment  to  a
temporary location on March 25, 1969. Although the
flood had passed and  plant operation had  restarted on
April 28,  1969, the move back to the Sanitary District
offices was delayed until August 4,  1969, because of
the fear that equipment damage during a  move might
prevent acquisition of  data  or  confuse the contract
negotiations  with  Badger   Meter  Manufacturing
Company.
   The  move  to   the  temporary  project  operating
headquarters had  apparently caused some equipment
problems. These problems  and  other  hardware  bugs
caused  some operating difficulties in the early part of
the  summer, and  some rains occurred for which no
data  was obtained.  The  flexibility  of  the  system
allowed  data acquisition  to  continue through  some
hardware  problems, which were all finally resolved on
July 10, 1969.
   Unfortunately,  the rainfall  occurring  during the
spring and  summer of 1969 was  less  than normal, so
there were fewer opportunities to operate the system
and  obtain  operating data. May  through September
rainfall was only 58 per cent of  normal. August and
September were particularly dry months. Precipitation
data for the last five years compared with the Weather
Bureau Climatological Standard Normals in  Table 14
shows that  1969 was the driest year of  recent record.
Table 15  shows  1969 monthly precipitation data in
comparison with the normal for each month. All 1969
precipitation amounts are those measured by the U.S.
Weather Bureau's official gage in Minneapolis. Because
of problems caused by the  flood  related move, opera-
tion in  April  was  incomplete,  so  the  months of
operation were May through November.  Only October
had higher than normal precipitation.
   Because of the importance of acquiring data and the
shortage  of  manpower, extensive analysis  of data
obtained  was delayed  until the conclusion of the
rainfall season. A review of all data obtained led to the
selection  of sixteen rainfall events  to be studied further
and  used for the project  evaluation. The first of these
sixteen rains occurred on  April  24 and the  last  on
November  16. During this period, 13.00 inches fell at
the Weather Bureau Station and 7.46 inches of this fell
during  the  16  rainfall  events  selected for  analysis.
Therefore, the evaluation is based upon 56 per cent of
the  rainfall  occurring  in   this  period, certainly  a
representative sampling. Table  16 shows total precipi-
tation recorded  at the Sanitary District remote trans-
mitting rain gages and the Weather Bureau gage for the
sixteen rains. That rainfall  amounts vary considerably
over the metropolitan area is apparent. Table 17 shows
maximum  60-minute intensities  for  each rain gage
during the sixteen storms. The official Weather Bureau
gage is reported only at clock hours, so the figure is not
the  actual  60-minute  maximum  intensity. Table  18
shows maximum 30-minute  intensities.
   Data acquired  during rainfall  events occurring in
April and May, 1970, have been similarly  analyzed.
Table 19 shows total precipitation recorded at the nine
system gages and at  the official Weather Bureau gage.
The average precipitation in all  system gages for these
rains varied from a low of 0.20 inches on April 15, to a
high of  1.49 inches on  May 21 and 22. Rains varied
from being  nearly  uniformly  distributed  (May 27,
A.M.) to having  a ten  to  one  ratio  in precipitation
totals (April 28). Great diversity in areal distribution of
rainfall such as occurred on April 28 are typical of
thunderstorms. Uniformly distributed rains are more of
a frontal  nature. It is  interesting to note, in these rains
and  the  sixteen  1969  rains analyzed,  that  areal  di-
versity  in rainfall amount is much more  common than
uniform or  near uniform distribution. Of the 27 rains
mentioned,  only 8 had  maximum gage to minimum
gage ratios less than 2 to 1.
                                                   69

-------
           Table 14
HISTORICAL PRECIPITATION DATA
May-September Per Cent Total Annual Per Cent
Year Precipitations, Inches of Normal Precipitation, Inches of Normal
Normal* 16.07 100 24.78 100
1969 9.34 58 19.39 78
1968 23.89 149 37.93 153
1967 12.92 80 25.44 103
1966 13.53 84 24.34 98
1965 25.50 159 39.94 161
"Climatological Standard Normals (1931-1960)
Table 15
1969 MONTHLY PRECIPITATION DATA
Month Normal*, Inches 1969, Inches Per Cent of Normal
January .70 2.05 293
February .78 .31 40
March 1.53 .90 59
April 1.85 1.55 84
May 3.19 1.98 62
June 4.00 2.93 73
July 3.27 2.95 90
August 3.18 .99 31
September 2.43 .49 20
October 1.59 2.53 159
November 1.40 .65 46
December .86 2.06 240
TOTAL 24.78 19.39 78
MAY-SEPT. TOTAL 16.07 9.34 58
APRIL-NOV.TOTAL 20.91 14.07 67
Table 16.
TOTAL RAINFALL AT EACH GAGE - 16 1969 RAINS

Day
116
141
148
173
176
196
205
207
211
218
242
267
274
285
303
320

Date
4-26
5-21
5-28
622
6-25
7-15
7-24
7-26
7-30
G- 6
8-30
9-24
10-1
10-12
10-30
11-16
US
W.B.
.82
.18
.24
.80
.90
.75
.25
T
.42
.71
.24
.28
.14
.69
.79
.05
Total Precipitation, In.
DM
.58
.23
0
.58
1.64
1.04
.42
.52
.60
.51
.22
.26
.06
.97
1.04
—
VAD
.54
.04
0
.67
2.02
.08
.12
1.52
.53
.56
—
.14
.05
1.30
1.00
.08
HAZ
.37
.22
.16
.66
.80
1.34
.16
.20
.32
—
—
.20
.12
—
.66
.17
FIR
.58
.16
.14
1.02
.92
.69
1.04
.04
.32
.36
.26
.32
.07
1.12
.60
.22
VI L
.62
.20
.10
.56
1.60
1.49
1.08
.46
.74
.06
.22
.24
.06
.96
.68
.18
EDN
—
.14
0
.58
1.18
1.30
.30
0
.56
—
.26
.28
.08
1.11
.66
.17
MAC
-
.10
.14
1.06
1.11
—
.26
0
.62
1.12
.26
.50
.08
1.12
—
.12
WTP
.56
.08
0
.61
1.27
—
.20
.44
.50
—
—
.24
.05
1.16
.80
.08
EDS











.16
.05
.80
.68
.16
              70

-------
                        Table. 17. MAXIMUM 60 MINUTE RAINFALL
                             AT EACH GAGE - 16 1969 RAINS
                               Maximum 60 Min. Rainfall, In.
Day

116
141
148
173
176
196
205
207
211
218
242
267
274
285
303
320
Date

4-26
5:21
5-28
6-22
6-25
7-15
7-24
7-26
7-30
8- 6
8-30
9-24
10-1
10-12
10-30
11-16
US
W.B.
.20
.07
.24
.20
.56
.59
.18
T
.23
.69
.12
.12
.08
.13
.10
.03

UM
.23
.10
0
.23
.79
.92
—
.50
.36
.50
.12
.11
.05
.25
.16
—

VAD
.26
.04
0
.20
1.06
.06
—
1.53
.37
.53
—
.09
.04
.22
.18
.04

HAZ
.16
.05
.17
.19
.65
1.15
—
.20
.15
—
—
.12
.12
—
.17
.07

FIR
.27
.05
.14
.49
.63
.61
-
.04
.27
.36
.13
.14
.07
.12
.11
.08

VIL
.30
.05
.11
.18
.38
.83
-
.30
.59
.06
.15
.11
.04
.20
.10
.05

EDN
.23
.06
0
.16
.93
1.06
-
0
.42
—
.14
.14
.08
.16
.12
.08

MAC
—
.07
.15
.39
-
-
-
0
.45
1.12
.13
.21
.08
.22
—
.07

WTP
.25
.05
0
—
-
-
-
.44
.34
-
-
.12
.05
.22
.16
.04

EUS











.09
.05
.18
.10
.05
                         Table 18. MAXIMUM 30 MINUTE RAINFALL
                             AT EACH GAGE - 16 1969 RAINS

                                 Maximum 30 Min. Rainfall, In.
                     US
Day       Date       W.B.   UM     VAD    HAZ    FIR    VIL    EDN    MAC   WTP   EUS

116       4-26         _______      __
141       5-21
148       5-28
173       6-22
176       6-25
196       7-15
205       7-24
207       7-26
211       7-30
218       8-6
242       8-30
267       9-24
274      10-1
285      10-12
303      10-30
320      11-16
                                          71
.06
0
.12
.50
.49
.50
.25
.49
.10
.06
.05
.13
.07
.04
0
.14
.70
.05
1.53
.22
.53
—
.07
.03
.17
.07
.04
.17
.11
.37
1.10
.20
.09
—
—
.07
.11
.07
.14
.03
.07
.30
.38
.48
.04
.14
.35
.08
.08
.05
.06
.06
.04
.08
.09
.24
.44
.30
.45
.05
.10
.06
.02
.12
.07
.04
0
.12
.58
.81
0
—
0
.12
.07
.08
.09
.06
.04
.14
.24
.76
-
0
.28
1.12
.09
.15
.05
.13
—
.05
—
-
—
0
.44
.21
—
—
.06
.05
.12
.09









.05
.04
.12
.08

-------
                            Table 19. TOTAL RAINFALL AT EACH GAGE
                                           11 1970 RAINS
                                        Total Precipitation, In.
Day No.
102-103
105
109
112
118
120-121
141-142
147
147-148
149
151
Date
April 12-13
April 15
April 19*
April 22
April 28
April 30-May 1
May 21-22
May 27
May 27-28
May 29
May 31
U.S.W.B.
.34
.11
1.38
.31
.14
.67
1.10
.44
.58
.41
.77
UM
—
—
—
—
-
.72
1.37
.43
.44
.83
.24
VAD
.42
.21
1.34+
—
1.56
.69
1.19
.41
.27
.85
.15
HAZ
.59
.11
1.36+
.31
.36
.82
1.73
.51
.51
.39
.11
FIR
.58
.20
1.02+
.27
.52
-
1.57
.47
.38
.57
.20
VI L
.70
.29
1.04+
.55
.73
.57
1.78
.57
.56
.35
.54
EDN
.58
.26
1.13+
.52
.63
.76
1.33
.44
.59
.11
.45
MAC
.48
.14
1.36+
.48
.16
-
1.25
.56
.69
.51
.84
WTP
.45
.31
1.46+
.50
1.45
.73
1.35
.49
.55
.17
.42
EUS
.26
.10
1.42
.62
.25
.67
1.84
.46
.42
.83
.43
"Rainfall weight in 7 gages exceeded transmitter range so exact total amount was not measured.
   Table 20 shows the maximum 60 minute accumula-
tions at the nine system gages. Intensity influences very
greatly the overflow losses that occur. A short high
intensity rainfall  can  cause  more  combined  sewer
overflow than a less intense rain having a much greater
total accumulation of rain.


Regulator Performance
   For each selected  rainfall  event, a compilation of
data was arranged. Listings of data files for regulators,
interceptors,  rain gages,  and  river  monitors were
generated  on the line printer.  A plotting routine has
been written for the system, and the data was plotted,
making  review and  interpretation  easier. Knowing the
depth of flow in the trunk sewer and the height of the
dam or diversion weir, one  can  determine whether
overflow occurred and its duration.  Volumetric flow
rate of  the  overflow  can  be  calculated using  a weir
equation.  The  analysis of data  proceeded  approxi-
mately as follows:

1.  Gate position and  Fabridam inflation pressure were
   checked since both have an effect on depth of flow
   in the trunk sewer during a storm flow period.

2.  If the Fabridam was inflated, the duration of the
   overflow, if any, was determined and noted.

3.  Assuming that the old  diversion weir was still in
   place, the duration of  overflow  that would have
   occurred was determined and noted.
4.  In  instances where the  Fabridam was not in an
   inflated  state during the  storm flow or  where  it
   deflated during the  storm  flow,  an  attempt was
   made to arrive at  an estimate of overflow duration
   based upon the trunk level data, the rainfall data,
   the  flow at other  regulators, and the past behavior
   of the particular regulator. Estimates were made and
                              Table 20. MAXIMUM 60 MINUTE RAINFALL
                                    AT EACH GAGE - 11 1970 RAINS
                                      Maximum 60 Min.  Rainfall, In.
Day No.
102-103
105
109
112
118
120-121
141-142
147
147-148
149
151
Date
April 12-13
April 15
April 19
April 22
April 28
April 30-May 1
May 21-22
May 27
May 27-28
May 29
May 31
U.S.W.B.
.05
.05
.38
.23
.11
.11
.88
.22
.32
.26
.40
UM
—
-
-
—
—
.17
.56
.16
.20
.54
.16
VAD
.12
.09
.56
—
—
.12
-
-
.14
.57
.07
HAZ
.12
.06
.51
.18
.21
.16
-
.18
.30
.29
.05
FIR
.06
.08
.42
.14
.42
-
.77
.21
.12
.52
.10
VI L
.16
.12
.40
.34
.69
.06
.61
.18
.23
.13
.45
EDN
.16
.12
.38
.41
.32
.12
.70
.20
.24
.04
.24
MAC
.13
.06
.44
.40
.13
-
1.01
.25
.38
.33
.41
WTP
.10
.10
.46
.31
1.34
.12
.59
.15
..22
.06
.15
EUS
.04
.05
.38
.52
.13
.14
.67
.16
.15
-
.32
                                                   72

-------
   used in overall data compilation only if conditions
   indicated  that  accurate  interpretation  could  be
   made.
   When  an inflated dam was  the diversion weir,
volumetric flow calculations for the old weir condition
could not  be  calculated using  the weir  equation.
Applying an orifice equation to the gate opening could,
however,  produce a  figure for additional flow volume
diverted because of  the higher head condition in  the
trunk sewer.
                               by reduction of incidence of overflow during the 16
                               1969 rains. The figures cited for incidence of overflow
                               are totals for all monitored regulators. Table 22 shows
                               regulator  performance as measured by reduction in
                               duration of overflow during the same rains. That many
                               of the per cent reduction figures are high confirms that
                               in the  past, with diversion weirs only slightly higher
                               than dry weather flow depth, regulators overflowed for
                               a fairly long period  following most rains. Table 23 and
                               Table 24  show the  reductions in overflow incidence
                               and  overflow  duration during  the  eleven 1970 rains
                               analyzed.
Day
116
141
148
173
176
196
205
207
211
218
242
267
274
285
303
320
Table 21 shows regulator performance as measured
                                           Table 21
                   SUMMARY OF 1969 OVERFLOW INCIDENCE REDUCTION
                                             Number of Regulator Overflows
                                                                Post-Project
                                                                 Conditions
  Date
April 26
May 21
May 28
June 22
June 25
July 15
July 24
July 26
July 30
August 6
August 30
September 24
October 1
October 12
October 30
November 16
TOTALS

 SUMMARY
Pre-Project
Conditions
    13
     8
     6
    14
    15
    15
    13
    11
    12
    13
     8
    10
     5
     9
     8
     7
   167
                                                                      1
                                                                      0
                                                                      1
                                                                      6
                                                                    13
                                                                    11
                                                                      3
                                                                      4
                                                                      6
                                                                      7
                                                                      1
                                                                      1
                                                                      1
                                                                      2
                                                                      1
                                                                      1
                                                                    59
Day
116
141
148
173
176
196
205
207
211
218
242
267
274
285
303
320
Date
April 26
May 21
May 28
June 22
June 25
July 15
July 24
July 26
July 30
August 6
August 30
September 24
October 1
October 12
October 30
November 16
Pre-Project
Conditions
37 00
33 00
7 15
50 00
51 45
43 15
21 30
15 45
41 00
31 30
17 15
20 15
8 00
69 00
61 30
19 30
               TOTALS
                                           Table 22
                              OF 1969 OVERFLOW DURATION  REDUCTION
                                         Total Regulator Overflow Hours
                                                               Post-Project
                                                                Conditions
                                                                   1  30
                                                                   0
                                                                   1  00
                                                                 10  30
                                                                 11  45
                                                                 12  45
                                                                   1  30
                                                                   2  45
                                                                   7  30
                                                                   6  15
                                                                   0  30
                                                                   1  30
                                                                   1  00
                                                                   2  30
                                                                   1  00
                                                                   0  15
                                      527 30                     62  15
i Reduction
    92
   100
    83
    57
    13
    27
    77
    64
    50
    46
    87
    90
    80
    78
    87
    86
    65
                                                                     i Reduction
                                                                         96
                                                                        100
                                                                         86
                                                                         79
                                                                         77
                                                                         71
                                                                         93
                                                                         83
                                                                         82
                                                                         80
                                                                         97
                                                                         93
                                                                         87
                                                                         96
                                                                         98
                                                                         99
                                                                         88
                                                 73

-------
                                              Table 23
                      SUMMARY OF  1970 OVERFLOW INCIDENCE REDUCTION
Day
102-103
105
109
112
118
120-121
141 - 142
147
147-148
149
151
       Date
     April 12-13
     April 15
     April 19
     April 22
     April 28
     April 30 - May 1
     May 21 - 22
     May 27
     May 27 - 28
     May 29
     May 31
     TOTALS
                                                 Number of Regulator Overflows
                      Pre- Project
                      Conditions
                            8
                            8
                           11
                           12
                           11
                            8
                           13
                           11
                           11
                           10
                           11
                         114
                            Post-Project
                            Conditions
                                 0
                                 0
                                 8
                                 6
                                 8
                                 3
                                11
                                 4
                                 7
                                 6
                                 5
                                58
                        > Reduction
                           100
                           100
                            27
                            50
                            27
                            62
                            15
                            64
                            36
                            40
                            55
                            49
Day
102-103
105
109
112
118
120-121
141 - 142
147
147-148
149
151
                                              Table 24
                      SUMMARY OF 1970 OVERFLOW DURATION REDUCTION
                                             Total Regulator Overflow Hours
  Date
April 12-13
April 15
April 19
April 22
April 28
April 30
May 21 •
May 27
May 27 - 28
May 29
May 31
TOTALS
• May 1
22
Pre-Project
Conditions
  67 00
  46 15
 111 15
  66 45
  37 00
 102 45
  66 00
  45 30
  47 30
  36 30
  29 15
 655 45
Post-Project
 Conditions
    0
    0
  17 30
   4 15
  11 00
   7 00
  18 15
   8 45
   9 45
   5 15
   3 15
  85 00
% Reduction
    100
    100
     84
     94
     70
     93
     72
     81
     79
     86
     89
     87
   Probably of greater interest than the effectiveness of
the program in  reducing overflow  incidence or over-
flow duration,  is a measure of its  effectiveness in
reducing  overflow  volume.  To  avoid  the  laborious
calculations that would be necessary to produce the
additional volumes  diverted and prevented from over-
flowing, use was made of the  mathematical model. As
part  of the model,  gate movements may be simulated
to evaluate effects of operations.

   The actual rainfall data from four 1969  rains  and
three 1970 rains were used as  input for the model with
diversion  weir height varied  to  simulate both fully
inflated  Fabridams and the  old fixed  height weirs.
Table 25  shows the results of this analysis for eleven
regulators.
                                       The overall volume  reduction  predicted by  the
                                     model for these seven rains is 51 per cent. Reductions
                                     for individual rains varied from 30 per cent to 98 per
                                     cent.  It is  to be  expected  that success in reducing
                                     overflow would be inversely related to the total runoff
                                     generated, and this is confirmed by the model. The rain
                                     of  Day 218  was very intense and  caused a compara-
                                     tively  large  runoff volume. Rains of  this  nature occur
                                     only a few  times a year. The other rains analyzed by
                                     the model are also of higher than  average amount, so
                                     the  gentle,   long  duration  rainfall  events are   not
                                     represented  in this analysis. It is in rains of this nature
                                     that the overflow can be completely  controlled at the
                                     modified regulators. Excluding the Day 218 rain, the
                                     overflow volume reduction for the other six rains is 73
                                     per cent.
                                                 74

-------
                                               Table 25
                          OVERFLOW  VOLUMES AT ELEVEN REGULATORS
                   FROM  MATHEMATICAL MODEL WITH ACTUAL RAINFALL  DATA
                                        Volume  of Overflow, CF
                                                                  Additional Volume
 Rainfall Day             Old Weirs             Inflated Dams            Diverted,  CF
 207-1969                181,100                   2,900                178,200
 211-1969              1,095,700                474,000                621,700
 218-1969              3,161,600               2,216,000                945,600
 267-1969                220,700                  27,100                193,600
 147-148-1970            568,500                108,000                400,500
 149-1970                397,600                  10,400                387,200
 151-1970                522,300                176,300                346,000
 TOTALS               6,147,500               3,014,700              3,072,800
 TOTALS
  (excluding Day 218)
        2,985,900
     798,700
   2,127,200
        Per Cent Reduction
           of Volume
               98
               57
               50
               88
               81
               97
               66
               51


               73
    It is interesting to compare the model's  predicted
 results for the eleven regulators with the data accumu-
 lated during  the actual rainfall event. Table  26 shows
 this comparison.  For some  rains, duration  reduction
 closely  parallels overflow volume reduction,  while for
 other rains, duration reduction is both greater than and
 less than volume reduction.

 Projected Seasonal Effectiveness
    A sufficient number of rains were analyzed that
 projections of normal seasonal effectiveness of the new
 regulator  control  system can be made. The greatest
 probable source of error involved in scaling up rainfall
 related  overflow  data  is in evaluating  how  repre-
 sentative  is  the data available of  the entire normal
 season. Very likely the rains for which data  were
 analyzed fall on the heavy side of the distribution of
 rainfall events. This system can admittedly effect good
 control only  for frequent return interval, low intensity
 rainfall events. Therefore, scaling  up measured  to
 seasonal overflow in the same proportion as  measured
                                    to seasonal rainfall, will indicate poorer than actual
                                    overflow reduction success on a seasonal basis. Statis-
                                    tical studies on rainfall return  interval  primarily deal
                                    with one year return interval storms and greater and
                                    little  information was available  for high frequency
                                    rainfall events.
                                       So it should be emphasized  that overflow control
                                    data presented on  a seasonal  basis are estimated.
                                    Because of the low number of rains analyzed volu-
                                    metrically with the mathematical model, the projection
                                    of seasonal overflow volumes and reductions is even
                                    more tenuous.  Other  methods of estimating overall
                                    season  overflow volumes were used in an attempt to
                                    improve the accuracy of the projections.
                                       The  previously  mentioned 27  rains occurring in
                                    both 1969 and 1970 represent  a total of 15.59 in. of
                                    rain. This figure is the sum of the averages of all system
                                    gages in use for each rain. Average total precipitation
                                    amounts for the 27 events ranged from 0.07 in. to 1.32
                                    in. The normal precipitation total for Minneapolis-St.
                                    Paul during April through October is 19.51 in.
                                             Table 26
                                      OVERFLOW CONTROL
                              MATHEMATICAL MODEL PREDICTION
                                             VERSUS
                                     REAL DATA INDICATION
                        Per Cent Reductions
                         Predicted by Model
Rainfall Day
    207
    211
    218
    267
    147
    149
    151
Overflow Volume
       98
       57
       30
       88
       81
       97
       66
Incidence of
 Overflow
    67
    43
    22
    50
    71
    89
    60
Overflow Duration
        81
        83
        75
        85
        80
        86
        89
Per Cent Reductions
Indicated by Real Data
              Incidence of
               Overflow
                  67
                  60
                  37
                  83
                  36
                  40
                  55
                                                  75

-------
   Table 27  indicates that 164 incidents of combined
 sewer  overflow discharge  and over  1000 combined
 sewer  overflow hours  were prevented during the 27
 rains,  reductions of 58  and 88  per  cent over  what
 would have occurred had  pre-project conditions existed
 at the  major regulators.  The  overflow  incidents and
 durations were multiplied by  the ratio of seasonal to
 measured  rainfall  to  produce  estimated   seasonal
 figures.

                    Table 27
     PROJECT EFFECTIVENESS IN REDUCING
  FREQUENCY AND DURATIONOF OVERFLOW
                                Total     Normal
                              27 Rains    Season
                               15.59
                                281
                                1183
19.51

 352
1480

 147
 184
                                            58
                                            88
Average Total Rainfall, In.
Pre-project overflows
 Number
 Duration, hours
Post-project overflows
 Number                        117
 Duration, hours                 147
Reduction (per cent)  in overflow
 due to project
 Number                         58
 Duration                        88
a measured values
k estimated values
   Results from computer  analysis  of actual rainfall
data  using  the mathematical  model  routines  have
already  been presented  in  Table 25.  At eleven regu-
lators during these seven  rains a total  of 3,072,800
cubic feet of  combined sewage was  prevented from
overflowing  and  was  intercepted.   Comparison  of
routed interceptor flow  hydrographs near the end  of
the  Joint Interceptor  for  simulated  pre- and  post-
project  conditions suggest that a   total  of about
4,850,000 cubic feet of combined sewage was captured
by the system  during these  rains. Scaling up this value
by the simple  ratio of seasonal  total rain to measured
rain  results in  an estimate of 31 million cubic feet  or
230  million gallons combined sewage overflow volume
prevented by the system on  a seasonal basis.
   The mathematical model, described in greater detail
in Part  II of this report, attempts to account for  all
flow which reaches the interceptor system, so that the
routing phase of the program will produce hydrographs
at the end of  the interceptor system that are repre-
sentative of actual  conditions.  Since  only the major
regulators and interceptor inlets were directly modeled,
the other smaller regulators and inlets were accounted
for by using coefficients representative of the ratio of
total tributary area to modeled tributary area for each
of the 15 catchments modeled. The model calculates
the total  inlet volume for any rainfall data  and then
calculates how each hydrograph is diverted at each of
the 15 modeled regulators.

   For the seven rains analyzed, the model calculated a
total inlet volume of 426 MG. Scaled up to a seasonal
figure  by the  ratio of rainfall amounts, the total inlet
volume becomes 2700 MG. This  represents the total
runoff volume  for the combined  sewered  areas of
Minneapolis and St. Paul, since the base or dry weather
flow was subtracted. This volume  is distributed among
approximately 150 combined sewered regulators, but,
based  upon studies of areas involved,  it is estimated
that about 70 per cent  of  the  runoff arrives  at the
controlled  major regulators for diversion.  Figure 63
shows  post-project and pre-project dispositions  of the
estimated total  seasonal  combined  sewer  runoff  vol-
ume.  It  should  be emphasized  that the  figures are
estimated by the model program which was used  in a
rough  state of  calibration.  Certain  of the numbers
shown in  Figure 63  are  not consistent  with  those
produced by  other  manipulations  of  data. For ex-
ample, the reduction in  overflow going to  the river
from  controlled  regulators  of 230 MG represents a
reduction of only 34 per cent. This is much lower than
the 51  per  cent figure  in  Table  25. This table  was
prepared from analysis of eleven controlled regulators
for which model predictions seemed most consistent
with  measured  data, so this data is  perhaps more
reliable.

   In  defense of the combined sewer system it should
be  pointed out  that a large fraction of storm  runoff
occurring each year is intercepted  and conveyed  to the
treatment plant.  Based  on  2700  MG  of storm  water
runoff from combined sewered area, 2160 MG runs off
the separate  storm sewered  area. Only 720 MG of
combined sewage reaches the river, 1/3 of the volume
discharged from  the separate  storm sewers. So com-
bined sewage would have to be 3  times as  polluted as
separate  storm sewer water to equal the effect of the
storm sewer discharges on the receiving water.

   Unfortunately the treatment  plant  is unable  to
handle flows  in excess  of  about  one half of  the
capacity  of  the  interceptor  at the plant.  Therefore
most of the additional volume intercepted is bypassed
untreated at the plant to protect the plant. Correction
of  this deficiency  will  be accomplished  in  future
construction at  the plant, and in the meanwhile the
pollution of the  urban stretch  of the Mississippi  River
is  much   less  frequent  and  of  shorter duration  to
increase dramatically the acceptability of the river for
boating, hiking and other recreation.
                                                   76

-------
            Post-Project Conditions

                  2700 MG
           Combined Sewer System
(70%)      1900MG
800 MG    (30%)
                                      Pre-Project Conditions

                                            2700 MG
                                                                        i
                                     Combined Sewer System
 (70%)     1900 MG
  800 MG
(30%)
     Controlled
     Regulators
Uncontrolled
  Regulators
    (77%)  1460 MG      520 MG  (65%)
440 MG
          280 MG
                  1980 MG
                 Interceptors
                      I
                  720 MG
                     i
               Mississippi River
  Large Regulators
Small Regulators
                                  1230 MG
                          520 MG
670 MG
                                           1750 MG
                                          Interceptors
                                               I
                                           950 MG
           280 MG
                                         Mississippi River
                       Figure 63. Estimated Disposition of Seasonal Runoff on Combined
                                 Sewered areas in Minneapolis and St. Paul.
                                                  77

-------
1970 Spring Thaw
   The control and monitoring system was maintained
year-round and hourly scanning continued through the
winter. Thaws occurring during the winter  and spring
were capable of causing combined sewer overflow, so it
was interesting to note the improvements effected  by
operation  of the system.  Operation to increase diver-
sion of combined thaw runoff and sewage is somewhat
simpler. Thawing in this climate is usually restricted to
daylight hours and usually occurs between 12:00 noon
and 5:00 P.M. Thaw flows slowly  and predictably rise
and fall, there being no equivalent of a  "cloudburst,"
so there is no need to be concerned about deflating the
Fabridams to  provide needed flow capacity.
   The  heavy snowcover  accumulated  in  December
remained on the ground most of the winter. Table  28
compares  December,  1969-March, 1970 snowfall as
inches  of  water  with  the  climatological standard
normal precipitation.  Figure 64 is the Venturi meter
chart for the Minneapolis  East Interceptor  for Febru-
ary 23, 1970, illustrating a typical thaw runoff pattern.

                    Table 28
      PRECIPITATION AS SNOWFALL FOR
             WINTER OF 1969-1970
                 Precipitation  Normal Precipitation
  Month         Inches-Water     Inches-Water
December  	  2.06             0.92
January	  0.47             0.81
February	  0.16             0.86
March	  2.05             1.52
Total	  4.74
4.11
   Figure 64.  Flow Meter Chart-Minneapolis East
              Interceptor (one of two identical
              parallel meters)
                  There were  four  days on which thawing  was
               extensive enough to cause at least one  overflow to the
               river. There were 26 days, however, on which thawing
               would  have  caused  overflow  at one or more of the
               same  regulators  had  pre-project conditions existed.
               Table 29 indicates the amount of thaw overflow  that
               was prevented during the  spring snowmelt of 1970.
               Most of the  overflow incidences prevented were prob-
               ably small and the volumes not great. Seven thaw days
               produced considerable runoff, however, and substantial
               volumes of combined flow were kept out of the river
               by the system.


                                    Table 29
                REDUCTION OF COMBINED SEWER OVERFLOW
                         DURING 1970  SPRING  THAW

                                     Pre-Project       Post-Project
                                     Regulator         Regulator
                                     Conditions       Conditions
               Number of Days
               Having Thaw
               Overflow	  26               4
               Number of Incidences
               of Regulator Over-
               flow 	  99               8


                  About 70 per cent of  the  incidences of overflow
               that would have occurred, had pre-project conditions
               existed, were at  four regulators. There were  a few
               regulators for which  no pre-project overflows would
               have occurred at all.
               Computer Analysis of Simulated Rainfall
                  Further use was made of the  mathematical model
               for evaluation of system effectiveness. Rather than use
               real  rain  data,  rainfall data  were simulated. As a
               simplification,  the synthetic  rainstorms all  were  of
               uniform distribution and intensity. While this type of
               rain  may  never  in  fact  actually  occur, the analysis
               provides a means of quantifying what the system may
               be capable of doing and pointing  out the variability of
               what can be done at various regulators. Each synthetic
               rain  data file  was  used  as  input to  the model
               twice—once  with  the  inflated  dams as the diversion
               weirs, and once  with  the old weirs as the diversion
               devices. Rainfall lastings 1/4, 1/2,  1, 2, and 3 hours was
               simulated. Rainfall intensities from 0.10 inches/hour to
               1.50 inches/hour were used.  Table 30 indicates the
               various combinations of  intensity  and  duration that
               were used.
                  Figure  65 and Figure 66 summarize the results of
               this  analysis.  Figure   65  shows that  reduction  in
               overflow  volume  is  influenced by both intensity and
               duration. As duration increases, some limiting overflow
                                                  78

-------
                Table 30
  SIMULATED RAINFALL DATA USED
FOR MATHEMATICAL MODEL ANALYSIS
   OF THE COMBINED SEWER SYSTEM
INTENSITY, INCHES/HOUR DURATI'
1/4 1/2
0.10. 	
0.20 ....
0.30 	
0.40 	
0.50 	
0.75 	
1.00 	
1.25 	
1.50 	


	 X

	 X
	 X
	 X X
	 X
	 X X

ON HOURS
1 2 3
XXX
XXX
X
X
XXX
X X
X X
X X
X
              reduction curve would be reached as steady state con-
              ditions develop and continue. Figure 66  shows, for
              rains of one-hour duration, the total overflow occurring
              at 12 controlled regulators for the two diversion weir
              conditions. Somewhere between intensities of 0.75 and
              1.0 inches/hour, the additional increment of flow that
              can be diverted appears to become constant,  which is
              what would be expected.

                In these analyses, interceptor capacity has not been
              considered to be a factor.  The model, in  its routing
              section, does indicate where it calculates excessive flow
              in the interceptor; that is, where an input or a routing
              results in  a flow rate exceeding that considered to be
              the  maximum for the interceptor at that location. In
              processing some of  the heavier simulated rainfall, the
              model did  indicate  flow volumes exceeding the capa-
              city of the interceptor, but these conditions are  not
              reflected in the figures.
     100
      80
    5
    D
    _1
    O


    I
    S60
    O
    O
    Q40
    O
    EC
  \
      20
                      .25
   .50             .75
RAINFALL INTENSITY, IN./HR
                                                                     1.00
                                                                                    1.25
                            Figure 65. Model Predicted Overflow Reduction
                                             79

-------
                            .25              .50             .75
                                  ONE HOUR RAINFALL INTENSITY, IN./HR

                                Figure 66. Model Data from Simulated Rainfall
                                                                            i.oo
                                                                                             1.25
Cost-Effectiveness of the Control System

   Using the extrapolated,  estimated figures obtained
from the mathematical  model,  approximately  30,000
combined sewered acres produce  approximately  720
MG  of  combined   sewer  overflow  each  year.  To
duplicate the reduction in  overflow accomplished by
the control system, 230 MG, by separation would then
require  complete separation of  almost  10,000 acres.
Estimates of the cost of separation  in St. Paul alone are
about 400  million dollars, at an  average cost of about
$20,000/acre. At that cost, the  1.7 million  dollar
demonstration  program is accomplishing the  same
effect as sewer separation work  costing about  200
million  dollars. Additionally, this hypothesized separa-
tion project would increase the separate  storm sewer
discharge  by about 350  MG  each  year,  thereby
increasing storm sewer  pollution. Possibly little head-
way would  be made by separating at this very  great
expense.
   The separation cost of $20,000/acre is probably not
an  overly  high  estimate either. Guarino et. al.  (23)
have  reported a detailed  cost  analysis for combined
sewer  separation in  Philadelphia. Their  analysis in-
cluded costs of internal plumbing conversions as well as
costs of separation  of the collection system  in the
streets. The total cost of complete separation on a per
acre basis was $55,000 to $66,000 per acre.
   So it  should  be assumed that the combined sewer
overflow control approach demonstrated  here on a
small number of major regulators  is capable  of more
than duplicating the results of  complete separation  at
less than one one-hundredth of the cost.
   This  is even  more impressive when one considers
that these results can  be accomplished  in two years. It
would be practically  impossible to  accomplish  com-
plete separation of 10,000 acres in 2 years regardless of
cost. And whatever the time schedule, separation work
would involve infinitely  greater  inconvenience  and
commercial  interference for the citizens of the cities.
                                                    80

-------
 XI. RECOMMENDATIONS
    At a  meeting in  Washington in the Fall of  1968,
 among other things, the District and FWPCA personnel
 agreed that  it would be  worthwhile to  empanel  a
 technical review committee representing various fields
 of engineering interest applicable to the project. The
 committee's  purpose  was, as stated in  the Board
 Minutes of October 14,1968, "to analyze project goals
 and  other  program  aspects   which  might  lead  to
 suggested modification."  The  committee,  authorized
 by  the  Board on  February 24, 1969, with members
 jointly selected  by FWPCA and District staff, met at
 the District's offices on March 17 and 18, 1969. The
 committee consisted  of  nationally known experts in
 research, design, and  operation of  pollution  control
 facilities.
    The committee members were most penetrating in
 their questioning of the project personnel.  In addition
 to the District project personnel, representatives of the
 St. Anthony Falls  Hydraulics Laboratory who  de-
 veloped  the  mathematical  model,  and the District's
 consulting  systems  analyst  participated  in  the  dis-
 cussion.  A  number  of changes in  the mathematical
 model and in data management resulted. Several new
 ideas  for evaluation of the effectiveness of  the system
 were obtained.
    Not all  of the recommendations of this committee
 were followed because of the need to proceed with the
 projects' limited objectives. Most of  the  recommen-
 dations are still very pertinent now,  as the  demonstra-
 tion period of the project ends and a final report  is
 prepared, and as the continued use  and  improvement
 of the system is anticipated. The entire  report of the
 committee follows.
     REPORTBYEVALUATING COMMITTEE
          IN SESSION MARCH 17-18, 1969
   The  Evaluating  Committee named below  met on
March 17-18,  1969,  to review  the  project  entitled
"Dispatching System for Control of Combined Sewer
Losses" financed jointly by the  Federal Water Pollu-
tion Control Administration and Minneapolis-St.  Paul
Sanitary District.
   The following members  of the Committee partici-
pated in the deliberations:
                               Mr. Srifel W. Jens
                               Rein & Jens,
                                Consulting Engrs.
                               Ill So. Meramec Ave.
                               St. Louis, Mo.  63105
                          Dr. E. Benjamin Wylie
                          Associate Professor of
                            Civil Engineering
                          University of Michigan
                          Department of Civil Engrg.
                          A /m A rbor, Mich.  48104
Mr. David R. Dawdy
Research Hydrologist
U.S. Geological Survey
345 Middlefield Road
Menlo Park, Cat. 94025

Mr. Carmen F. Cuarino
Deputy Commissioner
Ciry ofPhila. Water Dept.
15th and John F. Kennedy
  Blvd.
Philadelphia, Pa. 19100
Mr. V. A. Koelzer
Vice President
Harza Engineering Company
400 West Madison Street
Chicago, III.  60605
Mr. M. B. McPherson
Program Director
ASCE Urban Water Resources
  Research Program
Harvard University
Engineering Science Lab.
40 Oxford Street
Cambridge, Mass. 02138
                 Ben Sosewitz, Acting Chief
                  Maintenance and Operation
                 The Metropolitan Sanitary
                  District of Greater Chicago
                 100 E.Erie Street
                 Chicago, III. 60611
    The Committee elected as its Chairman, David R.
 Dawdy, and as its Secretary, Ben Sosewitz.
    The following project description was  reviewed and
 is taken from  the information sheet prepared by the
 Office  of Research  and Development, Federal Water
 Pollution Control Administration:

 DESCRIPTION OF PROJECT:  "Dispatching System
 for Control of Combined Sewer losses. "
   "The project includes preliminary studies to update
   historic  data, a  four-phase construction project
   consisting of installation of a gaging system, a data
   logger, five river monitors, telemetering  rain gages,
   regulator  modifications, and  a  post-construction
   program evaluation to include special studies by the
   University of Minnesota. Existing regulators will be
   replaced  with modem power operated gates  at 18
   key  diversion locations. A  supervisory system will
   be provided to telemeter gate  positions, flows, and
   levels in  sewers to  be controlled to a central point
   where a dispatching operator can observe  conditions
   and regulate flow accordingly.  Maximum  utilization
   of interceptor sewer capacity would be assured and
   overflow to the river will be minimized.  "
   The  above mentioned description defined initially
 the tools to be used in achieving  the objectives of this
program which are articulated in the last sentence of
 the project description.  It is the consensus of the
Evaluating Committee that the approach to achieve
 this end, as well as the  construction work completed
and the instrumentation provided, have been coordi-
nated  sufficiently  to demonstrate  successfully  the
capability of operating remotely and automatically the
regulating devices on  the interceptor sewer system.
   It has also been demonstrated that the collection,
storage, and analysis of data describing sewer elevation,
gate position, rainfall occurrence and waterways quali-
 ty  monitoring,  are  on  line in accordance with  the
project description.
   The Committee has chosen for its purpose to divide
 the basic  objectives of this project into the following
 three categories:
    1.   operational  benefits accruing  to  the Minnea-
        polis-St. Paul Sanitary District;
    2.   continuity of operational  results  and  data
        accumulation  for  future analysis,  evaluation
        and study; and
                                                     81

-------
    3.  program concepts and results usefully trans-
        ferable  to other municipalities facing similar
        problems.
   It  is the consensus of the Committee that opera-
tional benefits may be anticipated by the Minneapolis-
St.  Paul Sanitary District in the utilization of this
system to reduce overflows to the river. No numerical
results can  be  associated with  this  until  after an
operational season is completed. The full impact of this
system's  value can only be expected after the mathe-
matical model  being  developed is placed into service
and coordinated with the system and computer  in
performance tests. Several suggestions regarding the
above follow:
    (a) Development of the modeling program should
        be continued.
    (b) It would be desirable to obtain some measure-
        ments of discharge inside the system to verify
        the model.
    (c) Other auxiliary field measurements should be
        obtained to  verify the data being monitored
        by the system.
    (d) Pre-project conditions should be  simulated
        and evaluated in order to determine relative
        comparisons.
    (e) Economic  evaluation  to determine relative
        effectiveness  of the system should be planned
        for and pursued.
   Of greater concern to the Committee was the need
for crystalizing  data  and conclusions  which may be
transferable and of benefit to other communities. The
Committee  cautions  the Project Director  and the
Federal Water Pollution Control Administration not to
allow the unique opportunity, available as a result of
this project, to be  dissipated  because  of lack of
completeness or followup. It is  the opinion of the
Committee that the following steps should be included
in the present program or an extension of the program
in order to guarantee the value of the transferable data:
    (a) The model theory applied in both routing and
        analysis should be verified operationally and
        by field measurement.
    (b) Simulation ofpre- and post-project conditions
        should be accomplished.
    (c) Considerable  effort  should be placed in re-
        porting out  the details of the system used in
        achieving the operational  capability demon-
        strated.
    (d) Any   sophisticated  economic  evaluations
        should  be reinforced  by actual field verifi-
        cation.  The  Review Committee recommends
        that funds be sought for a new program or an
        extension of the present program  to  make
        such economic evaluations.
    It is the feeling of the Committee that prior to
final  submittal  of the report as outlined, a  second
project review committee should evaluate performance
over  the operational season,  which will commence
shortly,  together with the responsiveness to the recom-
mendations outlined herewith. Such a committee could
bring  additional insight  if it included independent
participants such as  a hydraulic modeler, hydrologic
modeler, planner, instrumentation specialist and speci-
fically operators from Seattle and Detroit because of
the similarity of the project to their activities.  In
addition, it is suggested  that the final report  author
avail himself of an ASCE report to the USGS (dated
April,  1969) which will shortly be available.
    In conclusion it is the consensus of the Committee
that the project direction has been handled expertly
and  that in  the Committee's opinion, if additional
funds  are  required  to establish  the continuity and
guarantee the value of the project, such expenditure is
strongly justified by the opportunity available to meet
the industry's needs at the present time.

DRD:rc
April 17, 1969

                 END OF REPORT
   Regarding the four steps recommended in the report
to be included in the present program or an extension
of the program, the following remarks can be made.
1. The  model theory  has  been  verified operationally
   with real data. This is reported upon in detail in Part
   II  of this  report,  the  report  of the Saint Anthony
   Falls  Hydraulics  Laboratory to  the  Minneapolis-
   Saint Paul Sanitary District.  There is, nevertheless,
   need for additional improvement, calibration, and
   further verification with more data that has and will
   be collected.  At  this  time, this is  not  possible
   because of lack of funding.
2. Simulation of  pre- and  post-project conditions has
   been accomplished to some extent using the existing
   mathematical model. This is reported upon  in both
   Part  I and Part II of this report. One benefit derived
   from this  simulation,  besides evaluation of system
   effectiveness, is that  areas of the model requiring
   improvement are  readily  identified in analysis of
   simulation results.
3. Hopefully,  this  report  has  adequately reported
   details  of systems  used in  the  Sanitary   District
   Regulator  Demonstration Program. If not, personal
   inquiries and inspection  visits will be as welcome in
   the future as were the many  occurring in the past.
4. Sophisticated economic evaluations have not been
   made.
   Many recommendations regarding the data  collec-
                                                    82

-------
tion aspects of  the  system have been  received.  The
following are selections from personal correspondence
to  the  Executive Director-Chief  Engineer  and  the
Project Engineer at the Minneapolis-Saint Paul Sanitary
District from review committee members and others.
   "Both you and your staff are to be congratulated on
   the progressive policy  which  you  are  following
   concerning water  pollution control. There are a lot
   of words being spoken  these days about necessary
   advances in our field, but not as many deeds. Your
   organization is matching your words with deeds."
      Carmen   F.  Guarino,  Deputy  Commissioner,
      Water Department, City of Philadelphia.  March
      19,1969.
   "You are indeed fortunate in  having pursued so
   vigorously this inovation in the control of overflow
   from a combined  sewer  system. The detail to which
   this project  has  been carried and the enthusiasm
   expressed by  its staff were particularly impressive."
      Vinton  W.  Bacon,   General  Superintendent,
      Metropolitan   Sanitary  District  of  Greater
      Chicago. March 28, 1969.
   "The opportunity for collecting the additional data
   by proper supplementary instrumentation and  net-
   work  design  for  one or more  of the catcnments
   involved in the Minneapolis-St. Paul demonstration
   project  ...  is so great  that to ignore it appears
   indefensible . . .
   Local operational and probable design benefits will
   accrue from  a verified  or modified  mathematical
   model."
      Stifel W.  Jens, Reitz and Jens, Consulting Engi-
      neers. April 12, 1969.
   "The  approach  taken  here  certainly does have
   applications  in the other metropolitan areas. If the
   older  cities are required to eliminate or minimize
   stormwater  overflows,  they  will  have  to choose
   between various alternatives and  combinations. The
   experience of the Minneapolis-St. Paul area will be
   extremely helpful  in this regard.
   To attempt separating sewers in large cities  would
   be a tremendous  undertaking and, needless to  say,
   very expensive .. .
   The  information  which  will be  collected by the
   system that Minneapolis-Saint Paul has installed .. .
   will be  helpful to determine the best methods of
   minimizing stormwater pollution, proper design of
   collector systems, and the practicality of routing
   storms through a collector system."
      Carmen  F.  Guarino,  Deputy  Commissioner,
      Water Department, City  of  Philadelphia. April
      13,1969.
   "Although  Mr. Sosewitz's report  emphasized the
   need for contining the data collection  program, I am
   so convinced  of the unique opportunity available in
   your project to obtain data that are otherwise sadly
   lacking, that  I  feel continued re-emphasis of this
   aspect  is essential. This  alone would justify future
   additional grants from FWPCA  or Office of Water
   Resources Research."
      V. A. Koelzer, Chief, Engineering and  Environ-
      mental Sciences,  National  Water Commission.
      May 1, 1969.
   "... a  point ...  was  made  by  our evaluating
   committee ... that, in the light of the very meager
   amount of data on quantity and  quality of waste-
   water flows,  the unique  installation which you have
   should  continue to generate information which will
   be of widespread  public value. It would be very
   disappointing, in   my view,  for  the information
   system  to be  used for purely operational purposes."
      V. A. Koelzer, Chief, Engineering and  Environ-
      mental Sciences,  National  Water Commission.
      December 22, 1969.
   To the  words of these people, the Project Engineer
can only add the statement that the Minneapolis-Saint
Paul Sanitary District has  been  deeply impressed by
these recommendations. Certainly, efforts  will con-
tinue to be made to accumulate much needed data, but
to insure transferability of data, guidance and direction
must come from without.

   A report to the United States Geological Survey by
the  American  Society of  Civil  Engineers also em-
phasized the importance of data similar to that which
the Sanitary District system can continue to collect.
   "Extensive construction of new  and replacement
   urban drainage systems  and facilities for reduction
   of pollution  from storm  discharge  and  combined
   sewer overflows over the remainder  of this century
   will involve significant national expenditures. Exist-
   ing  knowledge  is  inadequate  for  efficient and
   optimal planning and development of such facilities.
   Better  methods of analysis and more realistic design
   criteria  are needed for  engineering design of new
   drainage systems."
   "Improved methods are evolving very  slowly  be-
   cause of a dearth  of field measurements of rainfall-
   runoff   from  which logical,  reliable  models and
   quantification of  their  principal parameters can be
   developed. No municipality has  had  the financial
   and  manpower  resources to  mount  a  research
   program suited for  national  transferability of re-
   sults, and currently no prospect is in evidence of  a
   consortium of local governments financially pre-
   pared to undertake the task." (9)
   Very likely the Sanitary District system would have
to be modified  to increase  its value as a source of basic
data needs. Again, impetus must come from without.
   Other recommendations can be made. In all proba-
bility, systems similar to the one used at  the Minnea-
polis-Saint Paul Sanitary   District can  be used with
similar success elsewhere.
                                                   83

-------
   "Of the fourteen largest U.S. cities, ten are partially
   or wholly served by combined systems of sewerage:
   New York,  Chicago,  Philadelphia,  Detroit,  Cleve-
   land, Washington, St.  Louis, Milwaukee, San Fran-
   cisco and Boston. Of these, the following are served
   almost exclusively  by combined  systems:  Chicago,
   Detroit, St. Louis and San Francisco; and more than
   half of New York, Philadelphia  and Cleveland  are
   served by combined systems." (9)
   Several  years  ago, the  American  Public Works
Association  (21)  estimated the national  construction
cost for eliminating combined sewer systems  by sepa-
ration at $48 billion. Because of the extreme  cost and
the impracticality of  separation, along with  evidence
that urban storm  water discharges from separate storm
sewers  is  signficantly  polluted,  the FWQA  has pro-
moted  research and demonstration  studies of alterna-
tive  methods of  combined sewer overflow pollution
abatement, of which  the Minneapolis-Saint Paul Sani-
tary District Regulator Demonstration Program is one.
   The approach used here, one of control and routing,
would  seem to  be  applicable elsewhere to achieve
limited but  more immediate results  at proportionally
low cost and to complement other alternative methods
by which more complete abatement might be attained.
   Any  treatment facility built on a controlled regula-
tor system,  such as that in Minneapolis and St. Paul,
would accrue these benefits from the existing system:
Its size  could  likely be smaller; it could be accurately
sized  because of data already  available; and operating
and maintenance costs would be lower because of less
frequent utilization.
   Another  fact worthy  of mention  is that this was
really a "first  of a  kind" project.  Much  has been
learned  here, and in other similar projects elsewhere,
that should  permit subsequent systems, having greater
capabilities,  to  be installed elsewhere at  lower cost.
Engineers at Badger  Meter Manufacturing  Company
have indicated,  in private conversation, that the system
they  provided  the Sanitary  District with  could  be
acquired at  a lower cost today both  because of what
was learned during  this project  and  because of ad-
vancing  technology. The  central processor, or  com-
puter  power, of this system could  be provided today at
a much  lower cost than three years ago.
                                                   84

-------
XII.  REFERENCES
 1. Baker,  D. G.,  Haines, D.A. and Strub, J.H., Jr.,
    "Climate  of   Minnesota-Part  V.  Precipitation
    Facts,  Normals  and  Extremes,"  University of
    Minnesota-Agricultural Experiment Station  Tech-
    nical Bulletin 254, 1967.
 2. Toltz, King, Duvall, Anderson and Associates, Inc.,
    St. Paul, Minnesota, "Report on the Expansion of
    Sewage Works  in the Minneapolis-Saint Paul Met-
    ropolitan Area," Sponsored  by the Minneapolis-
    Saint Paul Sanitary District, Volume 3, September,
    1960.
 3. Anderson, J.J., Bruss, O.E., Hamilton,  L.A. and
    Robins,  M.L.,  "Application  of the AutoAnalyzer
    to Combined  Sanitary and  Storm  Sewer  Prob-
    lems,"  Presented at  the  Technicon  Symposium
    entitled Automation in Analytical Chemistry, New
    York, October  3, 1967.
 4. Anderson, J.J., "Computer Control  of Combined
    Sewers,"  Presented at American Society of Civil
    Engineers Annual and Environmental Engineering
    Meeting, Chicago, Illinois, October 13-17, 1969.
 5. Bowers,  C.E., Harris, G.S. and Pabst, A.F., "The
    Real-Time Computation   of  Runoff and Storm
    Flow in  the Minneapolis-St. Paul Interceptor Sew-
    er,"  University of  Minnesota, St. Anthony  Falls
    Hydraulic  Laboratory Memorandum  No. M-118,
    Prepared for Minneapolis-St. Paul Sanitary District
    and FWPCA, December, 1968.
 6. Harris, G.S., "Real-Time  Estimation of  Runoff in
    the   Minneapolis-St.  Paul  Metropolitan Area,"
    University of  Minnesota,  St. Anthony  Falls Hy-
    draulic  Laboratory Memorandum No.  119, Pre-
    pared for Minneapolis-St. Paul Sanitary District
    and FWPCA, December, 1968.
 7. Harris,  G.S.,   "Mathematical  Models  of Major
    Diversion Structures  in the Minneapolis-St.  Paul
    Interceptor Sewer System," University  of Minne-
    sota, St. Anthony  Falls Hydraulic Laboratory
    Memorandum  No. 120, Prepared for Minneapolis-
    St. Paul Sanitary  District and FWPCA, December,
    1968.
 8. Harris,  G.S., "Development of a  Computer Pro-
    gram to  Route  Runoff in the Minneapolis-St. Paul
    Interceptor Sewers,"  University of Minnesota, St.
    Anthony  Falls Hydraulic Laboratory Memoran-
    dum No. M-121, Prepared for Minneapolis-St. Paul
    Sanitary District and FWPCA, December, 1968.

 9. "An Analysis of National Basic Information Needs
    in  Urban  Hydrology, A  Report  to  the  U.S.
    Department of  Interior,"  by the American Society
    of Civil Engineers, April, 1969.
 10. "Dispatching System  for  Control  of  Combined
    Sewer  Losses,"  Interim Report to the Federal
    Water Pollution Control Administration Submitted
    by the Minneapolis-Saint Paul Sanitary District,
    May, 1969.
11. Anderson, J.J., "Dispatching and Routing of Com-
    bined Sewer  Storm Flows for Maximum  Inter-
    ceptor Utilization," Presented at American Society
    of Civil  Engineers Water Resources Conference,
    New  York, October 18, 1967.
12. Anderson, J.J., "Real-Time Computer Control of
    Urban Runoff," Presented at the ASCE Hydraulics
    Division  Conference, Cambridge,  Massachusetts,
    August 21-23, 1968.
13. Gallery,  R.L., "Management  of Pollution Data
    Using Modern Techniques," Presented at American
    Society of  Civil  Engineers Annual  and Environ-
    mental Engineering Meeting, Chicago, October 15,
    1969.

14. Request for Facilities Demonstration Grant Form
    Submitted by the Minneapolis-Saint Paul Sanitary
    District to the Federal  Water Pollution Control
    Administration, April, 1966
15. Green, S.R., "The Storage  and Retrieval of Data
    for Water Quality Control,"  U.S. Department of
    the  Interior,  Federal  Water Pollution  Control
    Administration, Public Health Service Publication
    No. 1263, August, 1966.
16. Freeny,  A.E.  and  Gabbe,  J.D.,  "A  Statistical
    Description of Intensive Rainfall," The  Bell Sys-
    tem  Technical Journal,  Volume 48, Number 6,
    Page  1789, July-August, 1969.
17. Horner, W.W. and  Jens, S.W.,  "Surface Runoff
    Determinations from Rainfall Without  Using  Co-
    efficients," Transactions, AM. Soc. C.E., Volume
    107,  Page 1039, 1942.
18. Hicks,  W.I.,  "A  Method  of Computing  Urban
    Runoff," Transactions, AM. Soc. C.E., Volume
    109,  Page 1217, 1944.
19. Mick, K.L., "The Minneapolis-Saint Paul Sanitary
    District Operation and Expansion," Journal of the
    Water Pollution Control Federation, Page  1684,
    October,  1967.
20. McKee, J.E.,  "Loss of Sanitary Sewage Through
    Storm Water  Overflows," Journal  of the Boston
    Society of Civil Engineers, Volume XXXIV, Num-
    ber 2, April, 1947.
21. "Report of Problems of Combined Sewer Facilities
    and  Overflows, 1967,"  Federal Water Pollution
    Control  Administration by the  American Public
    Works Association, December 1, 1967.
                                                  85

-------
22. Anderson, J.J., "Analysis of Operating Data from     23.  Guarino,  C.F., Radzuil, J.V., and  Greene, W.L.,
    a Full Scale  Primary  Sedimentation Plant  at the         "Combined Sewer Considerations By Philadelphia,"
    Minneapolis-Saint Paul Sanitary District," A thesis         Journal  of  the  Sanitary  Engineering  Division,
    submitted to the University of Minnesota in partial         ASCE, Vol. 96, No. SA  1., February, 1970.
    fulfillment of the requirements for  the Degree of
    Master of Science  in  Civil Engineering,  April,
    1967.
                                                   86

-------
XIII. ACKNOWLEDGEMENTS
   The author of the report wishes to acknowledge the
help and cooperation  of the Board and Staff of the
Metropolitan Sewer Board and the former Minneapolis-
Saint Paul Sanitary District.

Metropolitan Sewer Board Members
   MILTON C. HONSEY, Chairman, New Hope
   JULIUS SMITH, Vice Chairman,  Chaska
   HERMAN J. ARNOTT, Treasurer, Minneapolis
   JOSEPH C. COOK, Coon Rapids
   BERNON STENSENG, Stillwater
   WARREN SCHULTZ, Eden Prairie
   GEORGE RUTMAN, Saint Paul
   DANIEL R. BAKER, Saint Paul
Metropolitan Sewer Board Staff
   RICHARD J.   DOUGHERTY,  P.E.,  Chief Ad-
     ministrator
   MAURICE L.  ROBINS, P.E.,  Deputy  Chief Ad-
     ministrator
   GEORGE W. LUSHER, Chief  Superintendent of
     Operations
Former Minneapolis-Saint Paul  Sanitary  District Board
   Members
   ARTHUR E. NAFTALIN, Minneapolis
   JACK E. NEWTON, Minneapolis
   DONALD P. RISK, Minneapolis
   GEORGE J. VAVOULIS, Saint Paul
   ROBERT A. OLSON, Duluth
   ARNOLD J. IMSDAHL, Saint Paul
   THOMAS R. BYRNE, Saint Paul
   GERARD D. HEGSTROM, Minneapolis
   ROBERT F. PETERSON, Saint Paul
   CHARLES P. McCARTY, Saint Paul
   JENS CHRISTENSEN, Minneapolis
   ROBERT F. SPRAFKA, Saint Paul
   DONALD R. WATKINS, St. Cloud
   LOUIS P. COTRONEO, Saint Paul
   NATHAN HARRIS, Minneapolis
   VERN R. ANDERSON, Minneapolis

Former Minneaplis-Saint Paul Sanitary District Staff
   KERWIN L. MICK, P.E., Chief Engineer and Super-
     intendent, MSSD (Retired July, 1968)
   MAURICE L.  ROBINS, P.E., Executive Director-
     Chief Engineer, MSSD (July, 1968  to Present)
Principal Project Personnel
   ROBERT L. GALLERY, P.E.
     Project Engineer, July, 1969-Present
     Assistant Engineer, June, 1967-July, 1969
   JAMES J. ANDERSON, P.E., Project  Engineer,
     April, 1966-July, 1969
   LOUIS J. BARSCHER, JR., Data Analyst
EPA Project Officers
   DARWIN R. WRIGHT, Washington, D.C.
   LOUIS J. BREIMHURST, Minneapolis, Minn.
Consulting Programmers
   STEPHEN F.CARLSON, Systems Analyst
   RICHARD S. CARLSON, Programmer
Mathematical Model Program Personnel
   C.  EDWARD  BOWERS, Professor,  University  of
     Minnesota
   GARTH A.  HARRIS, Post  Doctoral Fellow, Uni-
     versity of Minnesota
   ARTHUR F. PABST, Research Fellow, University
     of Minnesota
Plans and Specifications
   Contract 664A-Pioneer Service  &  Engineering
                  Co., Chicago
   Contract 664B - Pioneer  Service  &  Engineering
                  Co., Chicago
   Contract 666  —Pioneer  Service  &  Engineering
                  Co., Chicago
   Contract 677  — Toltz,  King,  Duvall, Anderson &
                  Assoc., St. Paul
Contractors
   Contract 664A - George F. Cook Construction Co.,
                  Minneapolis
   Contract 664B - Badger Meter Mfg. Co., Milwaukee
   Contract 666  — Fairchild Camera and  Instrument
                  Co., New York
   Contract 677  - Acton Construction Co. and Ash-
                  bach Construction Co., a  joint
                  venture, St. Paul
Special Consultants
   Mathematical Model Program—Saint  Anthony Falls
     Hydraulics Laboratory, University of Minnesota,
     Minneapolis


Cooperation by others
   The  District has enjoyed excellent cooperation from
many  local,  state, and federal  agencies, numerous
private  firms, and  private individuals. A complete list
and all the details here would  be too lengthy, but a
brief summary follows with apologies for unintentional
omissions.
Federal Water Pollution Control  Administration
   This project  has been authorized, supported, and
financed  in  part,  by  the Department  of  Interior
pursuant  to the Federal Water Pollution Control Act.
   The  fine assistance and cooperation  from  the
personnel of the FWPCA  is greatly appreciated. Dis-
cussions on various matters with Washington and local
personnel  always  provided quick resolution to pro-
blems and yielded  information requested immediately.
The river quality monitors are as developed by  the
FWPCA,  with  considerable  help locally  and  from
Cincinnati. STORET personnel have been very knowl-
edgeable  and  helpful,  and  the  effectiveness of  the
system which the FWPCA has developed can hardly be
overstated.
                                                87

-------
   Without the financial  assistance from the FWPCA,
the project might not have existed.
U.S. Geological Survey
   The USGS  has furnished data for the project, both
locally and in Washington,  promptly  and exactly as
requested.
ESSA
   The  U.S.  Weather  Bureau has assisted  the daily
regulator surveillance by  providing direct rainfall in-
formation and the state climatologists have assisted in
the placement and checking  of rain  gages. They have
also made publications and rainfall data available, some
of which  has been placed on punch cards  for corre-
lation with MSSD data.
Corps of Engineers
   The Corps of Engineers has provided data and maps,
and computer programs developed by them have been
used in the project.
Cities of Minneapolis and Saint Paul
   Both Cities are providing financing for the District's
share of works constructed exclusively for one city and
the cost of the District's share of jointly used project
facilities is financed  by them through normal  District
procedures. The engineering departments in both cities
have provided plans and drawings and obtained neces-
sary approvals, in addition to maintaining  close liaison
on other matters.
   The Water  Departments in both cities have  allowed
installations of rain gages  on  their property. One river
quality  monitor is  installed in  a  building  at  the
Minneapolis Water Department.  The Saint  Paul  Fire
Department has cooperated in the installation of a rain
gage, as has the Metropolitan Airports Commission. A
river quality monitor is located at one of Saint Paul's
storm water pumping stations.
Private Firms and Individuals
   Northern States Power Company has allowed instal-
lation of a river monitor on their property at the High
Bridge Plant. They also  have  been most cooperative in
providing electrical services  at  all of the  locations,
including some in unusual locations.
   Northwestern  Bell  Telephone  Company  personnel
have been extremely helpful and have maintained close
liaison with District and contractor personnel. When
required, all parties have worked very effectively as a
team.
   Industrial Molasses Corporation and Svoboda Boat
Works have allowed river  monitors  to be  placed on
their property.
   Manufacturers,  vendors, sales representatives  and
contractors in  all fields have  taken an interest in this
unusual project and have provided help even when no
direct benefit to them could be realized.
American Society of Civil Engineers
   The ASCE  Urban Hydrology Research Council has
made available the results of extensive  research and
publications concerning  combined sewer overflow pro-
blems sponsored by the FWPCA and concerning urban
hydrology and water resources management  sponsored
by the USGS and OWRR, respectively. The interchange
of information at meetings and in the Division Journals
has been invaluable to the project.
                                                    88

-------
                   APPENDIX A
       An Example of Data, May 27-28,1970

   Presented as an appendix is a nearly complete set of
plotted data  for  a  single rainfall event. The data are
typical of the quality and quantity of data available for
27  rainfall events.  By the end  of the 1970  rainfall
season, this body of data  will  likely  have grown to
include over 50 rainfall events.
   With this large collection  of  high quality data, the
calibration of  the  existing mathematical model can
certainly  be improved. This data, which because of the
procedures that  have been  set up  can continue to
accumulate, will be of great value to. the Public Works
Departments  of the cities of  Minneapolis and St. Paul
or to their consultants  as they face  the continuing
problems of controlling pollution from  their combined
sewer systems.
   The data  are  presented  in  the form  of  plots or
graphs that were generated  on the  project system. The
time axis on all plots is 16 hours long, from 1600 hours
on Day  147 (May 27) to 0800 hours on Day 148 (May
28). Each  vertical grid line represents  one hour. The
vertical  axis  on all plots  is  labeled from  0 to 100,
representing 0 to  100 percent  of  the  full scale  value
that is printed at the top center of each  plot. Raingages
and inflatable dam pressures have full scale values of 2
in. and 2 psi respectively. Full scale values for depth of
flow may be 50, 100, or 120 in.
   On plots of trunk  level (TL) the horizontal row of
2's represents  the  height of the old  pre-project diver-
sion weir and the horizontal row  of  1's represents the
height of the new inflated dam.
   The rain for which the data are presented  was fairly
uniform. Rain  began  falling at about 2100  hours on
May 27  and lasted, on the average, about 3 hours. The
total rainfall  amount measured  at the 9  system  rain
gages varied from 0.27 in. to 0.69 in.
   Review of  this data  indicated that  there  were 7
overflows from control and monitoring regulators and
that there would  have  been  11 overflows had  pre-
project  conditions existed.  Total overflow duration
hours for pre-project conditions would have  been 47.5
and the actual total overflow hours were only 9.7, so at
controlled regulators  37.8  overflow hours were  pre-
vented.
   When the rainfall  data  were input to the mathe-
matical  model, the overflow volume reduction attri-
butable  to the  program at  eleven well modeled regu-
lators was from 568,500 cubic feet  to 108,000 cubic
feet, a reduction of 81 per cent.
                                                   89

-------
RAIN GAGES
                     The  total amount of precipitation
                     at the Water Treatment Plant rain
                     gage was .55 inches.
                     The  total amount of precipitation
                     at the Golden Valley rain gage was
                     .56 inches.
                     The total amount of precipitation
                     at  the  Edina  rain  gage  was .59
                     inches.
      90

-------
                The  total amount of precipitation
                at .the  University of  Minnesota St.
                Paul  Campus rain gage was  .44
                inches.
                The total amount of precipitation
                at the  Eustis  rain  gage  was .42
                inches.
                The total amount of precipitation
                at  the  Airport  rain gage was .69
                inches.
91

-------
mt V
                                      1*1    1*1     !*•
                                                                The  total amount of precipitation
                                                                at the Lake Vadnais rain gage was
                                                                .27 inches.
                                                                The total  amount of precipitation
                                                                at the Hazel Park rain gage was .51
                                                                inches.
                                                                The total  amount  of  precipitation
                                                                at  the West Side Fire Station  rain
                                                                gage was .38 inches.
                                                 92

-------
                                      REGULATORS
                                                            Camden Regulator

                                                            The increased depth of flow caused
                                                            by rainfall  is apparent in this plot.
                                                            The pipe height is 48 inches.
                                                        CAM
        I'M Elft-Bl       l»l IV (BU l(*ll


I I I.I 1 I.I I I.I I I.I I I.I I I.I I 1.1 I 1.1 I I.I I I.I I I.I I 1.1 I I.I I I.I I 1.1 1 1.1 1 •
                                                            The height of a side overflow weir
                                                            was raised from 35"  to  48". No
                                                            overflow occurred,  but before this
                                                            modification, overflow would have
                                                            occurred for 1.5 hours.
                                           93

-------
                Portland and Washington

                This is termed  "Master Regulator"
                because it  is  built into  the N.W.
                Interceptor, and regulates all flow
               : entering the interceptor  upstream.
                Although  increased  flow was re-
                ceived  here,   no overflow  would
                have occurred even before regulator
                modification. The pipe height is 90
                inches.
               The same proportionate increase in
               flow is apparent in the interceptor
               down stream from this regulator.
                The vertical  lines represent auto-
                matic inflation of the  'Fabridam'
                The pressure losses due to leakage
                cause  no  problems.  Within  the
                range of pressure shown, there is no
                change in height.
94

-------
                                             26th and Seabury
                                             This regulator would have overflow-
                                             ed for 3.5 hours prior to installa-
                                             tion  of  the  Fabridam.  The pipe
                                             height is 96 inches.
                                              This depth  is measured in the dry
                                              weather outlet chamber. The outlet
                                              pipe size is  36 inches, and submerg-
                                              ed entrance conditions existed for
                                              over 1.5 hours.
ifi-tc •      tin mi scut
                                              This plot  shows the automatic in-
                                              flation pressure adjustment at three
                                              hour  intervals  and  the  increased
                                              pressure   exerted  upon  the
                                              'Fabridam'   during  high  flow
                                              periods.
                               95

-------
                                    39th and Minnehaha
                                    The  depth of flow here exceeded
                                    the height of the old weir for 2.25
                                    hours.  The  trunk sewer height is
                                    123 inches.
S fULl SCILE
The  increased flow  in  the outlet
pipe leading  to  the  interceptor is
shown. The level sensor is located
within  a  parshall flume,  so  the
discharge can  be determined here.
                                           M39
                                    Due to damage the 'Fabridam' was
                                    not functional.
                    96

-------
m» 111-11      loons fyn
             0:11 FUll iC»L(
t?*i tH-iC       1. IS Ftlll SCUt
                                                   38th and Edmund
                                                   During this rainfall  event, flow ex-
                                                   ceeded the height of the  inflated
                                                   'Fabridam' for four hours. A much
                                                   greater volume  would  have been
                                                   lost over the old weir  which was
                                                   overflowed   for  six  hours.  The
                                                   height of the pipe is 120 inches.
The 36 inch outlet pipe here did
not  flow  full,  indicating that the
size  of the gate opening controls
the maximum diversion rate.
                                                    The  re-inflation every  third  hour
                                                    and  the  increased  flow pressure
                                                    against the 'Fabridam' can be seen
                                                    in this plot.
                                   97

-------
                31 stand Randolph
                Runoff would have caused overflow
                for  the  duration  of the  rainfall
                event prior to modification.  The
                pipe height is 102 inches.
                   • RAN.
                The plot shows re-inflation every
                third hour and the result of flow
                pressure on the 'Fabridam'. There
                was a  rapid leak at this  location
                which is shown in this plot.
98

-------
100'. IS fUll SClLt
                                        2nd and Main
                                        This is a "Master Regulator" for the
                                        Minneapolis East Interceptor. Over-
                                        flow would  have occurred  for  at
                                        least 4  hours prior to modification
                                        at  this  regulator.  Flow did exceed
                                        the height  of the  'Fabridam' for
                                        approximately  45  minutes.  The
                                        pipe height is 84 inches.
                                       The  trend  of  increased  flow  is
                                       shown  in the  interceptor  down-
                                       stream  from  this  regulator.  The
                                       pipe size is 42 inches.
                                       Re-inflation and  increased pressure
                                       on the 'Fabridam'  is shown.  The
                                       loss of pressure is very slow here.
                       99

-------
                Oak Street
                Runoff flow exceeded the height of
                the 'Fabridam' here for approxi-
                mately  3  hours.  (The  old  weir
                would have overflowed for over five
                hours and much more volume would
                have been lost.)
                       EAST INT.
                 Increased pressure due to the storm
                 flow  and deflation of the 'Fabri-
                 dam'   is   shown   in   this  plot.
                 Normally this dam is almost  leak
                 free,  but the pressure relief valve
                 apparently released air  during  the
                 runoff event.
100

-------
                Rice and Rondo
                The pattern of increased runoff is
                apparent  in this plot.  Because of
                high velocities, this data may not be
                correct.  The pipe  height  is  114
                inches.
                This plot of the level in the outlet
                section shows  the effect of rainfall
                runoff. The dry weather outlet pipe
                size is 36".
                      JOINT INT
   \
                Increased pressure from runoff flow
                and re-inflation every third hour is
                shown here. This dam has a known
                leak and rapid loss of pressure.
101

-------
                Trout Brook
                Overflow occurred here for approx-
                imately 15 minutes. Prior to 'Fabr-
                idam'  installation,  this  regulator
                would  have overflowed  almost  4
                hours.  The  pipe height  is  126
                inches.
                This data indicated a problem with
                the  bubbler  element  which  was
                later repaired.
                     JOINT INT.
102

-------
                Phalen Creek
                Runoff flow exceeded the height of
                the 'Fabridam' for a total of  1.25
                hours.  Dry weather  flow  would
                have  been  occurring  before  the
                event prior to modification.  Pipe
                height is 112 inches.
                A  submerged entrance condition
                occurred   continuously   before,
                during, and after the rainfall event.
                Deficient outlet  capacity is appar-
                ent. The pipe height is 30 inches.
                       JOINT INT.
                The increased pressure of additional
                flow and the  re-inflation of the
                Fabridam' is shown in this plot.
103

-------
                Belt Line
                Because  of high velocities,  trunk
                sewer level data was not reliable.
                This level is in the outlet chamber.
                The entrance to the 24 inch outlet
                pipe was submerged  for 3Vi hours.
                       JOINT INT.
                A very rapid pressure  loss is evi-
                dent. An air  leak in the attachment
                caused  this.  The  function  of the
                dam  is  therefore  somewhat un-
                certain.
104

-------
: A
                                   Kittsondale
                                   Overflow would have occurred for a
                                   total of 2.25 hours at this regulator
                                   prior to modification. No flow pass-
                                   ed  over  the inflated  dam. Pipe
                                   height is 108 inches.
                                    Increased flow in the outlet section
                                    is shown in this plot.
                                      • KD

                                      MISS. RIVER
                                      BLVD INT.
                                    Inflation pressure adjustment every
                                    third hour is shown in this plot.
                   105

-------
INTERCEPTORS
                      Minneapolis Northwest Interceptor
                      The rainfall almost filled the Minne-
                      apolis Northwest Interceptor at this
                      location at 1 a.m. The pipe height is
                      54 inches.
                      The runoff from the rainfall event
                      in  the northwest area of  Minne-
                      apolis that was captured is shown in
                      the  interceptor level at the end of
                      the   Northwest  interceptor.  The
                      pipe height is 123 inches.
    !«• •    It*
      106

-------
                Minneapolis Southwest Interceptor
                 Increased runoff filled the Minne-
                 apolis Southwest Interceptor at this
                 location for well  over two hours.
                 The pipe height is 72 inches.
                 Increased runoff flow in the Minne-
                 apolis  Southwest   Interceptor  is
                 shown  in  this plot of level at its
                 end.  The pipe height is 72 inches.
                          NDA
107

-------
                Minneapolis East Interceptor

                The plot of level in the Minneapolis
                East  Interceptor  at  Tenth  and
                Marshall  shows the effect of  in-
                creased rainfall. The pipe height is
                88 inches.
                The effect of rainfall runoff on the
                flow in the Minneapolis East Inter-
                ceptor is shown in this plot of level
                at its  end. The pipe height is  72
                inches.
                            EAST INT.
                                    MISS. RIVER
                                   BLVD. INT.
                 St. Paul Mississippi River Blvd.
                 Interceptor
                The St. Paul Mississippi River Blvd.
                Interceptor was filled to capacity
                around midnight  due  to the  in-
                creased runoff.  The pipe height is
                70 inches.
108

-------
               St. Paul W. 7th St. Interceptor


               The  level in St. Paul West Seventh
               St. Interceptor  shows the  increased
               flow during this  runoff event. The
               pipe height is 70 inches.
               Joint Interceptor

               Runoff  flow at  the  Joint Inter-
               ceptor at Pine and Spruce reached
               almost 130 inches. The pipe height
               is 157 inches.
                       JOINT INT.
                                           JT3
109

-------
               APPENDIX B
Selected Project Materials

a. Tables of Inputs to Interceptors
  Several tables of data abstracted from  that
accumulated  during the sewer  sampling and
analytical phase of the program are presented
to show examples  of  preliminary  analysis of
this information. Only average values are incor-
porated into the tables. Part C of this appendix
illustrates what the average value obscures.
b. Figures of Inputs to Interceptors
   Information  of  the  type included  in  the
tables  has been  applied to a schematic of the
interceptor  system to more graphically illus-
trate the sources of sewage load under average
conditions.
c. Three-Dimensional Plots
   The best way to illustrate the diurnal and
day to day variations in wastewater  flow or
composition  without  resorting  to  statistical
parameters was  found to be graphical. These
are photographs of three-dimensional models
constructed by  student-employees.
                                             111

-------
                                      HOURLY CONCENTRATIONS OF
                                       CHEMICAL OXYGEN DEMAND
In sewers
Tributary to the following:
     STATION
                           SAMPLE PERIOD
     Minneapolis Northwest Interceptor
     Camden                April, 1968-October, 1968
     Portland and Washington  July, 1967-January, 1968
     Northwest Meter         October, 1966-July, 1967
NUMBER OF
 SAMPLES
                                                          1005
                                                           262
                                                           651
CHEMICAL OXYGEN DEMAND MG/L

   MAX.      MIN.       MEAN
                   1340       173         344
                   1790       211         673
                   1700       150         578
 II   Minneapolis Southwest Interceptor
     39th and Minnehaha      January, 1967-January, 1968
     38th and Edmund        January, 1967-January, 1968
     Southwest Meter         October, 1966-July, 1967
III
IV
Minneapolis East Interceptor
31st-Randolph          April, 1968-October, 1968
Second-Main            October, 1967-October, 1968
Oak Street             April, 1968-October, 1968
East Meter             October, 1966-July, 1967
Joint Interceptor
Eustis
Wabash-Cromwell
Kittsondale
Rice and Rondo
West Seventh Street
St. Peter and Kellogg
Trout Brook
Phalen Creek
Belt Line
                           January, 1968-July, 1968
                           January, 1968-April, 1968
                           April, 1968-July, 1968
                           January, 1967-October, 1967
                           April, 1967-October, 1967
                           January, 1968-July, 1968
                           January, 1967-October, 1967
                           January, 1968-July, 1968
                           January, 1967-January, 1968
                                                           612
                                                           741
                                                           629
                   1630
                   1920
                   1420
              181
              183
              174
           426
           574
           466
811
1059
967
678
864
24
351
654
632
853
723
981
570
1990
1980
1910
1910
1950
1910
1950
1940
1870
1810
1240
1940
1660
184
177
169
168
187
504
186
227
178
179
179
187
180
593
606
408
622
625
1132
613
890
479
523
367
533
402
     Treatment Plant
     Plant Raw Waste
     Plant Primary Effluent
                      October, 1966-January, 1969
                      April, 1968-July, 1968
   4719
    501
   1930
   1310
148
181
600
405
                           NOTE:  Data contains wet weather as well as dry weather flows.
                                                    112

-------
III
IV
                                AVERAGE INPUTS OF CHEMICAL OXYGEN
                          DEMAND FROM KEY REGULATORS TO INTERCEPTORS
     STATION
                          SAMPLE PERIOD
Minneapolis Northwest Interceptor
Camden                May-August, 1968
Portland-Washington     August-October, 1967
Seabury                January, 1967-August-October, 1967
Northwest Meter         November, 1966-May, 1967

Minneapolis Southwest Interceptor
39th-Minnehaha         January, 1967/August-October, 1967
38th-Edmund           January, 1967/August-October, 1967
Southwest Interceptor    November, 1966-May, 1967

Minneapolis East .Interceptor
Randolph              May-August, 1968
2nd-Main              May-August, 1968
Oak                   April-July, 1968
East Meter              November, 1966-May, 1967

Mississippi River Boulevard Interceptor
Otis                   April-May, 1968
Mississippi River Blvd.
 AVERAGE
FLOW, MGD


    10.5
    35
    12
    50
 AVERAGE CHEMICAL
  OXYGEN DEMAND
MG/L            LB/DAY
     Joint Interceptor
     Eustis
     Rice-Rondo
     W. 7th Interceptor
     St. Peter-Kellogg
     Trout Brook
     Phalen Creek
     Belt Line
                                                                               345
                                                                               675
                                                                               500
                                                                               580
                      March-May, 1968
                      January—August, 1967
                      May-August, 1967
                      March-May, 1968
                      January-August, 1967
                      March-May, 1968
                      January—August, 1967
                          NOTE: Hourly concentrations and flow depths are available.
                  30,000
                 197,000
                   5,000
                 242,000
8
13.5
37
14
27
8
37
430
575
470
595
605
410
625
29,000
65,000
145,000
69,000
136,000
27,000
193,000
3.5
11
15
6
5
0.8
11.5
15
9.3
613

625
890
480
525
370
530
400
18,000

78,000
44,000
20,000
4,000
35,000
66,000
31,000
                                                 113

-------
                        PRELIMINARY ANALYSIS OF VOLUMETRIC AND COD LOAD
                   INPUTS INTO THE MINNEAPOLIS-ST. PAUL INTERCEPTOR SYSTEM
     Some flow data is estimated. COD concentration values are from Technicon AutoAnalyzer. COD concentrations
     in parentheses are calculated from differences in flow COD load.
 I   Minneapolis Northwest Interceptor
     Camden
     Other
     Portland and Washington
     Seabury
     Other
         TOTAL

 11   Minneapolis Southwest I nterceptor
     39th-Minnehaha
     38th-Edmund
     Other
         TOTAL

III   Minneapolis East Interceptor
     Randolph
     Other
     Second and Main
     Oak Street
     Other
         TOTAL

IV   Mississippi River Blvd. Interceptor
     Otis and Marshall
     Other
         TOTAL

 V   Joint Interceptor
     Eustis
     Rice and Rondo
     West Seventh
     St. Peter and Kellogg
     Trout Brook
     Phalen Creek
     Belt Line
     Other
         TOTAL
     Plant
 AVERAGE
FLOW, MGD

    10.5
    24.5
    35
    12
     3
     8
    13.5
    15.5
    14
    13
    27
     8
     2
     3.5
     7.5
           50
           37
           37
            11
AVERAGE
COD MG/I.

    345
   (820)
    675
    500
           580
   430
   575
   (395)
   595
   (620)
   605
   410
  (1800)
   613
  (1150)
           470
           625
           981
15
6
5
0.8
11.5
15
9.3
7.4
625
890
480
525
370
530
400
(1250)
          TOTAL
           70.0
          205
           608
           600
AVERAGE
COD # /DAY

  30,000
 167,000
 197,000
  50,000
 -5,000
  29,000
  65,000
  51,000
  69,000
  67,000
 136,000
  27,000
  30,000
  18.000
  72,000
                                           78,000
                                           44,000
                                           20,000
                                            4,000
                                           35,000
                                           66,000
                                           31,000
                                           77,000
           242,000
           145,000
           193,000
            90,000
           355.000
          1,025.000
                                                  114

-------
                                 AVERAGE INPUTS OF CHLORIDES FROM
                                  KEY REGULATORS TO INTERCEPTORS
     STATION
     SAMPLE PERIOD
 I    Minneapolis Northwest Interceptor
     Portland-Washington     August — October, 1967
     Seabury               January, 1967-August-October, 1967
     Northwest Meter        November, 1966-May, 1967

 II    Minneapolis Southwest Interceptor
     39th-Minnehaha        January, 1967/August-October, 1967
     38th-Edmund          January, 1967/August-October, 1967
     Southwest Interceptor   November, 1966-May, 1967
AVERAGE         AVERAGE
FLOW,MGD  CHLORIDES, MG/L       LB/DAY


    35               170              50,000
    12                80               8,000
    50               165              69,000
                                     8               110               7,000
                                    13.5              180              20,000
                                    37               175              54,000
III   Minneapolis East Interceptor
     East Meter             November, 1966-May, 1967
                                    37
                      90
                 28,000
IV   Joint Interceptor
     Rice-Rondo
     W. 7th Interceptor
     Trout Brook
     Belt Line
January—August, 1967
May-August, 1967
January—August, 1967
January—August, 1967
     6
     5
    11.5
     9.3
110
 65
 90
 90
6,000
3,000
9,000
7,000
                          NOTE: Hourly concentrations and flow depths are available.
     STATION
                                     AVERAGE INPUTS OF AMMONIA
                          NITROGEN FROM KEY REGULATORS TO INTERCEPTORS
     SAMPLE PERIOD
 I    Minneapolis Northwest Interceptor
     Portland-Washington     August - October, 1967
     Seabury               January, 1967-August-October, 1967
     Northwest Meter        November, 1966-May, 1967

 II    Minneapolis Southwest Interceptor
     39th-Minnehaha        January, 11J67 /August-October, 1967
     38th-Edmund          January, 1967/August-October, 1967
     Southwest Interceptor   November, 1966—May, 1967
 AVERAGE   AVERAGE AMMONIA
FLOW, MGD     NITROGEN, MG/L        LB/DAY


    35                15               4,400
    12                 8                 800
    50                17               7,000
                                     8                14                 900
                                    13.5              16               1,800
                                    37                20.5             6,300
III   Minneapolis East Interceptor
     East Meter              November, 1966-May, 1967
IV   Joint Interceptor
     Rice-Rondo
     W. 7th Interceptor
     Trout Brook
     Belt Line
January—August, 1967
May-August, 1967
January—August, 1967
January-August, 1967
                                    37
     6
     5
     11.5
     9.3
                           NOTE: Hourly concentrations and flow depths are available.
 21
  4.5
 11
 15
                                       2,500
1,100
  200
1,100
1,200
                                                    115

-------
MINNEAPOLIS NORTHWEST
                             MINNEAPOLIS EAST
                                           182,000 Ib./day
                                             625 mg/l
                                              35MGD
                         50MGO
                         580 mg/l
                      242,000 Ib./day
                                                                        JOINT
                                    | SWI | ]

                                    35MGD
                                   470 mg/l
                                 137,000 Ib./day
                      MINNEAPOLIS SOUTHWEST
11MGD
                                                 MISSISSIPPI RIVER BLVD.
                                                                                                 PLANT
                        Note: Hourly concentrations and
                             flow depths available.
                                                    , I |PRw|
                                                      205MGO
                                                      600 mg/l
                                                   1,025,000 Ib./day
                                    Input  of Chemical  Oxygen Demand from
                                     Minneapolis Meters to Joint  Interceptor
                                            under Average Conditions
        10.5MGD
  -OD   345 mg/l
  CUD 30,000 lb./day
   MINNEAPOLIS WEST
                             8MCO
                            430 mg/l
                          29.000 Ib./day
                                   MINNEAPOLIS SOUTHWEST
                                                                              Note: Hourly concentrations and
                                                                                   flow meter readings available.
                                      Input  of Chemical  Oxygen Demand
                                    from Major  Regulators to  Interceptors
                                           under Average  Conditions
                                                         116

-------
      MINNEAPOLIS WEST
                                 MINNEAPOLIS EAST
                                               28.000 lbs./day
                                                   90 mg/l
                                                   37MGD
                          50MGD
                          165 mg/l    |NWI
                       69,000 Ibs./day
                                       |SWI

                                       37MGD
                                       175 mg/l
                                    54,000 Ibs./day
                            MINNEAPOLIS SOUTHWEST
                                                                          JOINT
MISSISSIPPI RIVER BLVD.
                                                                                                              PLANT
                                                                                   Note: Hourly concentrations and
                                                                                        flow depths available.
                                                Input  of Chlorides from
                                       Minneapolis Meters  to Joint  Interceptor
                                              under Average Conditions
MINNEAPOLIS WEST
                          MINNEAPOLIS EAST
                    35MGD
                    170 mg/l
                 50.000 Ibs./day
                                                                                          9,000 Ibs./day
                                                                             6.000 lbs./day     90 '
                                                                               110 mg/l
                                                                                6MGD
                                                                                  RR
                                SEA

                               12MGD
                               80 mg/l
                             8,000 Ibs./day
                             [39-MJ-*
                             8MGD
                            110 mg/l
                          7,000 lbs./day
                                                MISSISSIPPI RIVER BLVD.


                                                  ' |  E38 |
                                                  13.5MGD
                                                   180 mg/l
                                                20,000 Ibs./day
                                       MINNEAPOLIS SOUTHWEST
                                                            9.3MGD
                                                      BL |   90 mg/l
                                                          7,000 Ibs./day
                                 5MGD
                                 65 mg/l
                              3,000 IbsVday
                                                   PLANT
                          Note:  Hourly concentrations and
                                flow meter readings available.
                                                Input of  Chlorides  from
                                                   Major  Regulators to
                                        Interceptors under Average Conditions
                                                            117

-------
     MINNEAPOLIS WEST'
                               .MINNEAPOLIS EAST
                                    37MGD
                                    20.5 mg/l
                                  6,300 IbsVday
                         50MGD
                         17 mg/l
                      7,000 lbs./day
                                                                                                         PLANT
                                                   MISSISSIPPI RIVER BLVD.
                        MINNEAPOLIS SOUTHWEST
                                                                                 Note: Hourly concentrations and
                                                                                      flow depths available.
                                         Input of  Ammonia  Nitrogen from
                                     Minneapolis  Meters to  Joint  Interceptor
                                              under Average Conditions
                        I MINNEAPOLIS EAST
MINNEAPOLIS WEST
                  P-W

                 35M6D
                 15 mg/l
              4,400 Ibs./day
                             1,100 Ibs./day
                               21 mg/l
                                6MGD
                                                                                           1,100 Ibs./day
                                                                                              11 mg/l
                                                                                             11.5MGD
                 JOINT
                         |39-M
                         8MGD
                         14 mg/l
                       900 lbs./day
MISSISSIPPI RIVER BLVD.

    E38|

  13.5MGD
   16 mg/l
 1,800 lbs./day
                                      MINNEAPOLIS SOUTHWEST
                                      5MGD
                                     4.5 mg/l
                                    200 IbsVday
                                                                                                              9.3MGD
                                                                                                              15 mg/l
                                                                                                           1,200 IbsVday
                                                                                                        PLANT
                              Note: Hourly concentrations and
                                   flow depths available.
                                         Input of  Ammonia  Nitrogen from
                                         Major  Regulators  to Interceptors
                                              under Average Conditions
                                                          118

-------
NOTE:
          Data includes wet and dry weather
          flows. Flow data available to calculate
          pounds per hour.
   Hourly  Frequency  Distributions of  Chlorides
   in  Plant  Raw  Waste,  Approximately  1800
   Observations.
                                                    600
                                                              	•	~\           300
                                                              v           400     -rinNMG'
                                                              *"   co«*»1BKI
NOTE:  Data  includes  wet and dry weather
        flows. Flow data available to calculate
        pounds per hour.
Hourly Frequency  Distribution  of Chemical
Oxygen  Demand at  Minneapolis  Northwest
Meters. Approximately 600 observations.
                                                          119

-------
                                                                                    NOTE:  Data  includes wet  and dry weather
                                                                                            flows. Flow data available to calculate
                                                                                            pounds per hour.

                                                                                     Hourly Frequency  Distributions of Ammonia
                                                                                     Nitrogen  in Plant Raw Waste,  Approximately
                                                                                     1800 observations.
NOTE:  Data  includes wet  and dry weather
        flows. Flow data available to calculate
        pounds per hour.

Hourly  Frequency Distributions of Chemical
Oxygen  Demand  at  Minneapolis East Meter,
Approximately 600 observations.
                                                               120

-------
                          Part 11

MATHEMATICAL MODEL OF URBAN STORM RUNOFF AND THE
           INTERCEPTOR SEWER SYSTEM OF THE
        MINNEAPOLIS-ST. PAUL SANITARY DISTRICT

                            by
               Arthur F. Pabst, C. Edward Bowers
                     and Garth S. Harris
                        Prepared at the
               St. Anthony Falls Hydraulic Laboratory
                     University of Minnesota

                            for

              The Minneapolis-St. Paul Sanitary District
                            and
              The Federal Water Quality Administration
                            121

-------
I.  INTRODUCTION
   The cities of Minneapolis and St. Paul are served by
a combined sewer system which is designed to carry
both  normal dry-weather sanitary  sewage  and storm-
water  runoff. Originally the  main  trunk lines  dis-
charged the combined flow directly  into the Mississippi
River. However,  in the 1930's interceptor sewers were
constructed to collect the flow that would normally be
discharged  into the river and convey it to a treatment
plant.  Diversion structures were built into the trunk
sewers to  divert the flow  into  the  new  interceptor
sewers. The interceptor sewers were designed to carry
the normal dry-weather sanitary flow  plus  a small
amount of  storm-water runoff, but  excess storm-water
still had to be discharged  into the river during large
storms.  Figure 1 shows the location of the interceptor
sewers, the main diversion structures,  and the inlets to
the interceptor lines.
   Originally most  of  the  major diversion structures
had  fixed-height  dams. Then  in  1968  remotely-
controlled  fabridams and gates were  installed ^t the
diversion structures. These controllable gates and fabri-
dams allow an operator to alter the system according
to current conditions in the sewer system.  With the
variable-control diversion structures, advantage can be
taken of the non-uniformity  of rainfall distribution to
obtain  the  maximum  capture  of  runoff.  Larger
amounts of  runoff can  be  directed  into the sewer
system at certain locations if it is known that little or
no  rain  is occurring  in other  areas of the  cities.
However,  it  is  difficult for  an   operator  to make
quantitative decisions about the amount of rainfall and
storm runoff in  the system. A mathematical  model of
the Minneapolis-St. Paul interceptor sewer system [1] *
was  therefore  developed to  aid  the  operators in
determining the  expected flow at critical points in the
system so  that gates and fabridams could  be manipu-
lated  to  retain  the maximum flow in  the  system and
release the least overflow to the Mississippi River.

'Numbers  in brackets refer to references.
                     V MISSISSIPPI RIVER
                                           LEGEND
                                      «-  Major Inlet
                                      •   Diversion Structure
                                 "i    •   Rain Gage
                                                                                 MMNEAPOLIS- ST. PAUL
                                                                                SEWAGE TREATMENT PLANT
   Figure  1.   Map of  Twin  City area showing location of interceptor sewers,  major  inlets, diversion
              structures, and rain gages.
                                                   123

-------
II. GENERAL CONCEPT OF MODEL
   The mathematical model (UROM-9, Urban Runoff
Model  PDP-9) is a deterministic runoff model capable
of continuous real-time operation. Input to the model
consists of rainfall readings and  postulated gate and
fabridam settings. Output  can range  from a simple
message to the operator, to complete  rainfall and loss
rate analysis, diversion analysis, and  predicted hydro-
graphs at points in the sewer system. The operator can
interact with the  model while it is operating in real
time to alter the sequence of operations, postulate new
gate  and  fabridam settings,  and change the model
output. In addition the model can be set to operate on
historical or  design rainstorms. The operation of the
model  is discussed in detail in Appendix A, Operator's
Manual for UROM-9.
   To  make  the  model  easier to understand and  to
work  with  it has been broken down  into a series  of
simple  tasks,  each of  which  is carried out  by  a
subroutine. As  each subroutine  completes its task  it
passes  the results to another routine  until the entire
model  has been executed. This process can be illustra-
ted as follows:  The rainfall data that  are given to the
model   may  contain various  errors  of electronic,
mechanical,  or  human origin. Thus the  model's first
task is to  screen  the rainfall data  for errors and make
reasonable corrections. This  task is  performed by  a
subroutine entitled RAINF. Since the rainfall data may
have been recorded at any random  time interval, the
next task  is  to  space the data at some regular interval
(usually 5 minutes); this task  is performed by  sub-
routine EVNTL.  The  next  step is to  determine what
portion of the rainfall can be expected to be "lost" to
the ground  and  what portion will  become runoff.
Subroutine SLOS performs this task. These are only a
few of the steps in the overall model.
   It can be seen that changes in methods are easily
implemented. If a different loss analysis is desired, for
example, the change can be made by simply replacing
the current version of  SLOS  with  a  new  version
containing the new loss analysis. The subroutines are
not concerned with  the particular area modeled, but
simply perform their functions on the data given them.
   The model is applied to a given area by supplying it
with input data describing the physical features of the
area.  These  input data include the number of sub-
watersheds  the total area is broken into, their areas,
their times to peak, and the base times  of their unit
hydrographs. All input data is given  in Appendix A. In
the preceding  example, subroutine  RAINF would be
fed the number of rain gages in the system, subroutine
EVNTL would be told at what regular interval to space
the rairrfall  data, and subroutine SLOS would be given
several parameters for each sub-watershed  which would
relate the amount of pervious and impervious area, the
initial  depression storage,  the infiltration  constants,
and other variables depending on  the particular loss
analysis used. Thus the methods used are governed by
the subroutine instructions, while the application to an
area is governed by the input data to the subroutines.
   Some restrictions on the general applicability of the
model are as follows:
1. Storage requirements set an upper limit on the size
   of  input data arrays which in turn limits such things
   as  the  number  of sub-areas that can  be modeled.
   The upper limit can change depending on computer
   storage available.
2. Variations in the sizes of data arrays may require
   minor changes in the format of the output data.
3. Certain aspects  of diversion structure  modeling are
   not general in nature, but apply only to  the area to
   which the model is being applied.
                                                   125

-------
   III. SPECIFIC MODELING TECHNIQUES
      For purposes of discussion the model can be divided
   into three phases: the hydrology phase, the diversion
   phase,  and the routing phase. The hydrology phase is
   the period between the time the storm water appears as
   rainfall  and the  time it arrives at the  inlets to  the
   interceptor sewer system. The diversion phase deter-
   mines,  for given positions of the rubber fabridam and
   variable gate, how much of the arriving flow will enter
   the interceptor sewers and how much will be diverted
   to the river. Finally, the routing  phase  predicts  the
   movement  of  the  flow  that enters the interceptor
   sewers.  Further discussion of each  of these phases is
   given below.

   A.  Hydrology Phase
      In this phase rainfall from a network of recording
   gages is converted to  runoff at the major inlets to  the
   interceptor system.  The first step in the process deals
   with the mechanics of input of the rainfall data (For
   additional information on these steps see Appendix C.)
   This step provides corrected rainfall  readings at each
   gage in the network at regular time intervals, usually 5
   minutes. From  these raingage  readings  the  average
   rainfall  over each sub-watershed is determined,  using
   the Thiessen method [2]; this  consists of weighting
   each raingage by a factor dependent on the location of
   the gage relative to  the sub-area. During  the  1969
   runoff  season  a network of eight  recording raingages
   supplied  data  to the model. Figure  2 shows  the
   Thiessen polygons  used  to determine  the weighting
   factors for an eight-gage network with 15 sub-areas. At
   the end of the year (1969) a ninth recording gage was
   added.   Future changes  may also  be made in  the
   locations of some existing gages.
                         LEGEND
                       — WATERSHED BOUNDARY
                        • RAIN GAGE        *
                       — THIESSEN POLYGON
                         - INTERCEPTOR SEWERS
Figure 2.   Map of Twin City Area showing major water-
           sheds and Thiessen polygons.

     Next  the  losses and the excess precipitation to be
  expected in each of the sub-areas must be found. The
  "loss" is that part of the total precipitation which is
retained in the sub-area,  while the "excess" is the
portion that  becomes  runoff and  enters  the sewer
system. Predicting the losses  that will occur during a
particular  rainfall  event  over  a  particular  area  is
probably the most difficult task of the model. Several
approaches have been considered. The current method
determines separately the losses  on the  impervious
surfaces (i.e.,  paved areas) and  those on the pervious
surfaces (grassy areas) and  then combines them, taking
into account  the relative amounts  of  pervious  and
impervious surface in the sub-area.

   These  relative  areas  were estimated  using  aerial
photographs of representative areas of the Twin Cities.
Photos  on a scale of either one inch equals 100  ft or
one inch equals 200 ft, covering areas of approximately
0.1 sq  mi and 0.4 sq mi, respectively, were overlaid
with a  1/4 inch or a 1/8  inch uniform grid, and  grid
squares enclosing  impervious surfaces were counted.
Only  effective impervious surface area was included;
where runoff from rooftops would  drain across lawns,
for example, the rooftops  were not considered imper-
vious area. The  proportion   of impervious area  was
found  to  vary  from 96  per cent for  a downtown
business area  to  26 per cent for  a  single-family
residential  area. The average  for Minneapolis and St.
Paul combined is 35 per cent.

   Impervious  Analysis—The   excess from  impervious
areas is assumed to depend on (1)  the rainfall, (2) an
antecedent index parameter,  and (3) three  constants.
The antecedent parameter is the accumulated loss—i.e.,
the moisture retained—on the  impervious surf aces; this
is  presently taken to be zero at the beginning of an
event. The first  constant is the  value of the accumu-
lated loss that must be reached before any runoff  will
occur. This value may take into account interception
by  trees and wetting of land surfaces. Zero is presently
used for this initial loss constant. Once the  initial  loss
has been satisfied, the loss is determined as follows:
                                                                    Loss = Rainfall x loss coefficient        (1)
                                                         where
                                                               Loss coefficient = Ce~k (accumulated loss)   (2)
and  C  and  k  are  constants  for a  sub-watershed
(assumed  to  be 1.0 and 2.0,  respectively.)  Figure  3a
shows this loss coefficient as a function of accumulated
loss  on  the  impervious area.  The  loss  coefficient
decreases  asymptotically from an initial value of 1.0
for completely dry conditions to  a value of zero for
wet conditions. Thus for dry conditions the entire rain
is  retained  as loss on  the impervious surfaces.  As
moisture builds up, depression storage is filled and the
loss  coefficient decreases,  allowing a  portion  of the
                                                      127

-------
rainfall  to  become runoff.  Figure 4a shows the actual
loss  rate  when this same  function  is  applied  to a
uniform one-inch-per-hour rain.
   The impervious excess is  then
         Impervious Excess = Rainfall — loss
(3)
   Loss
 Coefficient
          0       0.5      1.0      1.5      2.0     2.5
                 Accumuloted loss on impervious oreo (in)
Figure  3a.  Impervious loss function.
                 0.5      1.0      1.5     2.0
                Accumuloted loss on pervious Oreo (in)
                                                2.5
Figure 3b.  Pervious loss function.

   Pervious Analysis—The excess from pervious areas is
assumed to depend on (1) the rainfall, (2) an antece-
dent  index  parameter,  and  (3)  four constants.  The
antecedent  parameter is  the accumulated  loss on
pervious surfaces—i.e., the moisture retained by the soil
in a given  layer. This value is presently taken as zero at
the beginning of an  event.  The  first constant  is an
initial value of  the  accumulated loss that  must be
satisfied before any runoff can occur; a value of  0.25
in.  has been used here. Once the initial loss  has  been
satisfied, the loss rate is determined using the Holtan
infiltration  equation  as  modified  by  Muggins  and
Monke: [3]

         ,   ,   . /T  - (accumulated loss)\k      IA.
         f=fc + Al	-	1        (4)
where f is the potential infiltration rate (in. perhr); fc
is the constant equilibrium infiltration rate, taken to be
0.2  in. per hr; A is  the increase  in  infiltration rate
under dry conditions,  assumed  to  be  2.0  in.  per
hr—giving a total infiltration rate of 2.2 in. per hr when
the soil is dry; T is the total available storage in the soil
above the restricting layer, assumed to be 2.0 in.; and k
is a constant taken as 0.7.
   Figure 3b shows the potential loss rate as a function
of the accumulated loss on the pervious area. The loss
rate decreases from an initial value of 2.2 in. per hr for
completely  dry conditions to a value of 0.2 in. per hr
for wet  conditions. This same loss rate applied to a
one-in-.-per-hour rain is shown in Fig. 4b. Under these
conditions it may be noted that for the first 1-2/3 hrs
of the storm the rainfall rate is lower than the potential
loss rate, so the actual loss rate is equal to the rainfall
rate.  During the remainder of the storm the potential
loss rate is lower than  the rainfall rate. At 2-1/3 hrs the
potential  loss  rate has  decreased  to  its equilibrium
value (0.2 in.  per hr). Thus when the rainfall is  less
than the potential  loss rate, the actual loss rate is equal
to  the rainfall  rate;  but when the rainfall  equals or
exceeds the potential loss, the actual  loss rate is equal
to  the potential loss  rate, which  in turn  equals  the
equilibrium value fc. The excess on the pervious areas
is then given by
                                                                     Pervious excess = rainfall — loss
                                                          and the sub-watershed excess is given by
                                                          (5)
                                                         Basin excess = —= I impervious excess) + ——-—  (pervious excess)(6)

                                                         where  x  is  the  percentage  of  sub-area  that is im-
                                                         pervious.
                1.0
        Loss rate
         (in/hr)
                 .5
                                                                          (	Rainfall rote = 1.0 in/hr)
                  0123
                                  Time (hrs)
       Figure  4a.  Impervious loss rate for 1 in./hr. rainfall rate.
                i.o
        Loss rate
        (in/hr)
                                                                  .5
                                 \
                                 PERVIOUS AREA
                                  Time (hrs)
        Figure 4b.  Pervious loss rate for 1 in./hr. rainfall rate.
                                                     128

-------
    Figure 4c shows the results when the complete loss
 rate is applied to an area that  is 1/3 impervious with a
 one-in.-per-hour  rainfall.  The  area between the time
 axis and the loss curve represents the total loss over the
 sub-watershed, while the area  between the loss curve
 and  the  dashed  one-in.-per-hour line  represents the
 total  excess over  the sub-watershed. This excess will
 become runoff at the inlet to the interceptor sewer.
        1.0
Loss rate
 (in/hr)
         .5
                                 COMBINED AREA
                                 (33%  IMPERVIOUS)
          0        I          2        3
                          Time (hrs)

 Figure 4c.  Combined loss rate for 1 in./hr. rainfall on
             a watershed 1/3 impervious.

    An important feature of this loss analysis is that the
  independent  variable  is  accumulated loss  (moisture
  retained  in  the watershed), whereas  in  other  loss
  analyses—Morton's  [2], for example—time is the inde-
  pendent variable. Since under real conditions there are
  often  periods of low or zero rain during a  prolonged
  storm, an analysis based on time  requires  additional
  techniques, such as time condensation. In addition to
  simplifying the  process, the use of accumulated loss as
  the  independent  variable allows for a  recovery  or
  draining feature to be incorporated  into the analysis
  which  would describe the  return to dry conditions
  during periods of no rain. Presently no recovery feature
  is included in the model.
    The determination  of  the hydrographs at the inlet,
  is next. The excess water, determined in the above loss
  analysis,  arrives at the  interceptor sewer inlet; it  is
  assumed that this quantity of water is spread uniformly
  over the sub-watershed.  Several  approaches to  deter-
  mining the  hydrograph,  or  time distribution, of  this
  runoff have been proposed. The  unit hydrograph
  approach  used  here is based on the theory  that when
  the  water  drains to  the  watershed outlet (i.e.,  the
  interceptor inlet)  it does so  according to a given time
  pattern which  is assumed to depend on the physical
  features of  the area—its size, shape, slope, etc. The
  theory further  assumes that if different amounts of
  water occur  as runoff,  the  magnitude of the runoff
  changes proportionately,  but the time pattern does
  not change.
    The shape of the  unit hydrograph is most reliably
  determined  from recorded events  that have occurred
  on  the watershed. Since this information was  not
 available in the present case, some characteristics of the
 unit  hydrograph were estimated from  the physical
 features  of the sub-watersheds, and a triangular shape
 was assumed. The  estimated  characteristics were the
 time  to peak  Tp  and the base
  proximation  Tp  and
  relationship given by Kent [4]:
           Tb.  As a first ap-
Tb  were determined from the
                                                                            0.8
                                                           TD(hrs)  = -
                                                                                  9000  Sa
                                                                                                       (7)
where D is the duration of rainfall increment (hr), L is
the length of mainstream to farthest divide (ft), CN is
the SCS hydrologic soil-cover complex number, Sa is
the average slope of the watershed (per cent) along the
mainstream, and
                                                                         Tb (hr) = 2.67 Tp
                                               (8)
   Kent provides a nomograph for the solution of the
second term of Eq.  (7). The peak discharge of the unit
hydrograph would then be given by
                                  AR
             Qp (cfs) = 0.0000464 ^~
                         (9)
where A = area (ft^), R = rainfall excess (in.), and Tb =
base time (hr).
   The  initial values of Tp and Tb given  by Eqs.  (7)
and (8) have been modified  in some cases as recorded
data have been obtained by the data acquisition system
at the MSSD treatment plant. The Tp and Tb given by
Eqs. (7)  and  (8) have agreed reasonably well with
values obtained from recorded data.
   The  duration  of  the unit impulse is selected as 5
minutes.  Thus each  5-minute  rain  period  yields a
runoff  hydrograph.  These  incremental  hydrographs
are combined  linearly and base flow is added to form
the total inlet hydrograph. The mean annual base flow
is used at each  inlet with  average hourly variations
imposed on it.
   The  formation of the total inlet hydrograph, com-
pletes  the hydrology phase  of  the model. Additional
discussion  of  hydrology phase   may  be found  in
reference [5].


B. Diversion Phase
   In  this phase the total inlet hydrographs, as deter-
mined  in  the hydrology phase, are passed  through the
simulated diversion structures to provide an estimate of
how much of the flow will enter the interceptor sewer
and how  much will enter the Mississippi  River. The
amount  of  flow  that  enters  the interceptor sewer
system  will then be passed on  to the routing phase of
the model.
                                                    129

-------
   To accomplish the diversion simulation each major
diversion structure was analyzed and then represented
by a combination of orifices, weirs, and pipe sections
for which discharge  equations were prepared. Several
structures are similar in  geometric configuration, while
others are of unique design, occurring only once in the
sewer system. The most  common type of diversion
structure is composed of (1)  an outlet from the trunk
sewer to  the interceptor sewer controlled by a  hy-
draulically operated gate and (2) a tunnel to the river
controlled by a pneumatically operated fabridam. Such
a structure  is shown in Fig. 5. This configuration is
referred to as a TYPE 1 diversion structure. The gate
opening is modeled as a variable-area  orifice and  the
fabridam as a variable-height weir. [6]
Figure 5.  Typical Type 1 diversion structure.
                                                    130

-------
   The flow coming down the trunk sewer is assumed
to take one of two paths: either through the gate to
the interceptor  or over  the  fabridam to  the  river.
Actually,  some of  the  flow must go  into temporary
storage  in the  trunk  sewer,  but  this  flow  is  not
modeled.  The portions  of the trunk flow which go to
the interceptor  and to the river can be determined
from  the  gate opening  and the fabridam height. The
discharge  Q0 that will pass through the orifice with a
driving head equal to the fabridam  height can then be
found. If  Qo is greater than the flow coming down the
trunk, all  the flow will go through the orifice to the
interceptor sewer.  If Qo  is less than the trunk flow,
part of the flow  will go through the orifice and part
will go over the fabridam. In this latter case a common
head  must  be   found  such  that  the  sum of  the
discharges over the weir and through the orifice equals
the trunk flow.  This result is found using an iterative
procedure (method of bisection). Figure 6 shows two
complete  solutions of  a TYPE 1 diversion  with the
fabridam  both down  and  up,  and  each with three
different gate openings  (0.25 ft, 0.75 ft, 1.25 ft).  For
the fabridam up and the 1.25 ft gate opening, it can be
seen that  when the trunk  flow is less than 50 cfs the
entire  flow  will  go through the orifice to the inter-
ceptor. For discharges greater  than 50 cfs only part of
the flow  goes to the interceptor,  and the  remainder
goes over the fabridam to the river. Thus for a trunk
flow of 100 cfs the interceptor flow will be 60 cfs and
the river flow will be 40 cfs. Once flow has started over
the fabridam only  small increases in interceptor flow
occur for large increases in trunk flow. This is due to
the difference between the exponent  for flow over a
weir (1.5) and that for flow through  an orifice (0.5)
and the difference in datum for the heads.
   The  diversion structure thus acts  to  clip off the
higher flows, sending only a regulated amount to the
interceptor sewers.
   The  clipping  action of the diversion  structures
increases  the accuracy of the  model prediction.  If the
loss rate  analysis predicts rainfall  excess equal to or
greater than the actual values, the inlet hydrograph will
be  either correct  or  over-estimated.  The  diversion
structure  will  then  cause  the  excess  flow  to  be
discharged into the river. Thus the amount that enters
the interceptor, which  is of primary interest, is more
nearly correct, with the errors going to the river. Figure
7 illustrates this effect.  Curve 1 represents a computed
inlet  hydrograph that may be in error. Curve 2 shows
the portion of this hydrograph that would enter the
interceptor, Curve  3 represents the actual inlet hydro-
graph, and Curve 4 shows the  portion of this actual
hydrograph that would enter  the  interceptor. There-
fore,  while there are large differences between 1 and 3
(that is,  before diversion), there are only small differ-
ences between 2  and 4, after diversion. This, of course,
holds  only  when  the  true inlet  hydrograph is  large
enough to cause diversion to the river.
   120
   80
                    Fabridam up
                          Gole Opening (ft.)    1.25

                                             0.75
              40
                                         160
                                           200
                        80      120
                    Trunk Inflow (cfs)
Figure 6a. Typical  Type  1  diversion rating  curve with
           fabridam up.
   I20
  o
  £40
                    Fabridam down
                          Gate Opening (ft.)    1.25
              40       80       120
                    Trunk Inflow (cfs)
                                          160
                                                  200
Figure
6b. Typical  Type  1  diversion rating  curve with
    fabridam down.

      M^ Computed inlet hydrogroph
      (?) Computed diverted hydrograph
      (3) Actual inlet hydrograph
      (4J Actual diverted hydrogroph
    80
    60
   6 20
     0        I       2       3      4      5       6
                        Time (hrs)
Figure  7.  Illustration of flow clipping action  of diversion
           structure.

    The TYPE 1 diversion structure occurs ten times in
  the  sewer  system.  The model  includes nine  other
  locations where flow may  be diverted  to the  river.
  Three of these  are modeled as sharp clipping devices
  (i.e., all discharge above a fixed value is released to the
                                                   131

-------
river).  Another  three  occur at the Minneapolis dis-
charge meters and are modeled as variable bypass gates
only. The remaining three are each special cases and are
handled  individually.  A complete  discussion of each
structure is given in Ref. [7].
   The land area that contributes storm runoff to the
interceptor system at the 15 major inlets represents 61
per cent of the  total  area contributing  runoff  to the
system. The remaining area contributes storm water to
the system  through about 47 minor inlets. Diversion to
the river may occur at the 15 major inlets plus 4 more
structures such  as  venturi  meters.
   To  allow for the  minor  inlet flow  the contri-
bution of each  minor inlet is  assumed to  enter the
sewer  system at the  nearest major inlet.  All  model
computations through the diversion phase are based on
the actual  area  contributing at a major inlet.  After
diversion is complete  the  diverted  hydrographs are
corrected by multiplying  the  discharges by the inlet
lumping factor, which is taken as the ratio of the area
assumed  to  contribute to  a  major inlet to  the actual
area that contributes to the inlet. In other words, the
minor  inlets are  assumed to  behave in a manner which
is hydraulically similar to that of the major inlets. The
largest area represented in this manner  is at the Phalen
Creek inlet.  In this case the actual area contributing to
the inlet is 2.06 sq mi, and multiplying by the lumping
factor of 3.96 gives an assumed or lumped area of 8.15
sq  mi.  Most of the  lumped  area  here is  from the
Beltline  trunk  sewer. The  dry-weather connections
from the trunk sewer to the  interceptor sewer for both
the Phalen  Creek inlet and the Beltline  inlet have low
capacities.  Only a  small  portion of the total  trunk
sewers capacity can  enter  the  interceptor.  For this
reason the  lumping of such a large  area at  the Phalen
Creek  inlet does not adversely  affect the accuracy of
the model prediction.

C.  Routing Phase
   General  Discussion—The  third phase of  the  study
covers the routing of the flow  through  the interceptor
system.  Available  methods permit a  much  better
prediction  of the routing   process  than they do of
runoff or diversion. The "method of characteristics"
was selected for initial routing of the hydrographs. The
usual  methods of hydrologic routing are based primar-
ily  on the equation of continuity  with constants or
simple  relations used  to   account for momentum
effects. The method of characteristics is one of several
methods of  mathematical  routing  which  solve the
equations of momentum and continuity. This method
has been proven in the literature to accurately repre-
sent unsteady, gradually varied flow in  open channels.
   It  was found that for  the  large system of open
channels which  makes up the interceptor system, the
method-of-characteristics solution requires a consider-
able amount of computer time, and therefore a method
was  sought  which would produce  the  same answers
using less  computer time.  It was found  that the
"progressive  average-lag  method"   from  Ref.  (8]
accomplished this objective if the constants used were
determined  by comparing input and  output hydro-
graphs  with those  obtained from the  method of
characteristics.
   Method  of Characteristics-The  equations of  con-
tinuity and  momentum for unsteady, gradually varied
open channel flow are
        A 8
1 d-A
A 3t
                          3V
                                               (10)
and
where

 V  is the mean velocity at any instant t and position
     x
 y   is the depth of flow at any instant t and position
     x
 A  is the corresponding cross-sectional area
 x   is the distance measured along the channel
 t   is the time
 S   is the energy slope
 So  is the bed slope.
 g   is the acceleration due to gravity.

These  equations  can be written in finite difference
form  for  integration along what are known as the
characteristic curves as follows:
                                              (12)
                                               (15)
The first two equations refer to integration along the
C+ characteristics shown  in  Fig.  8, while the second
two refer to the C_ characteristic.
   The reader is referred to Ref. [9] for the derivation
of  these equations. The  subscripts  in  the  equations
refer to the points P, R, and  S  in Fig. 8. The values of
velocity and depth at R and S are found by interpola-
tion between known values at the grid points A, B, and
C. As  indicated  in  Ref.  [9], these equations can be
                                                   132

-------
"
A
_
Figi
t
L
t




c/

P
\c_




O A R C S B
jre 8. Forward and backward characteristics an
definition of grid points.
solved to determine the velocity and depth at all points
like  P which are At later than the points like C, at
which conditions  are  known from previous  calcula-
tions. In other words, the flow can be represented on
the x — t plane, as shown in  Fig. 8. At every point on
the  grid the velocity and depth must be calculated
using a procedure which is similar to extrapolation.
   The  boundary  conditions are  important  in this
process, and  they  give some  information in the x — t
plane about the depth and the velocity at all points on
the lines x = 0 and x = L, where L is the total length of
the  channel.  The  boundary  conditions, which are
discussed in Ref. [10], are
   a. Critical flow at the downstream end of a  channel
   b. Inflow  hydrograph  at  the upstream  end of a
      channel
   c. Inflow hydrograph in the middle of a channel
   d. Junction of a number of channels.
At the  end of the channel only  one pair  of character-
istics equations applies  (i.e., either  a C+  or a C_
characteristic—see  Fig. 9). The other equation in the
case of critical flow, for  instance, is
                                               (16)
   Thus in the case of critical flow  Eqs. (12), (13), and
   (16) can be solved to determine Vp and yp. The details
   of the derivation of the boundary conditions are given
   in Ref. [10].
     It was  found  that  the interceptor sewer  system
   could  be broken up into a series of 23 open channels,
   each with one of the above boundary conditions at its
   two ends. The critical flow condition commonly occurs
   at the downstream end of the channel, while input
   hydrographs  commonly  enter through  a  diversion
   structure and drop shaft either at  the upstream end of
   the  interceptor or in  the middle of it.  The junction
   conditions occurs only once.
     A typical solution for a channel is shown in Fig. 10.
   It can be seen that the hydrograph is attenuated as it
   flows downstream and is displaced along the time axis.
    Ditcharga
                                      ivo Avvroga Log
                                       had ol
                                        Chorocttrilticf
Figure 10.
Results of  routing irregular hydrograph by the
method of characteristics and by the progres-
sive average lag method.
   A       R                             C
Figure 9.   Characteristic at downstream boundary.
     Many different numerical tests have been conducted
   using  the  characteristics routing  method  on various
   channels; a resume of the results is as follows:
   1. There is a maximum allowable rate of rise or fall in
     the input hydrograph for a given channel subdivided
     into certain  reach  lengths.  This  rate is at present
     found by trial and error.
   2. The program has been compared with numerical and
     actual results from work at Colorado State Univer-
     sity  (sponsored in part  by the  Bureau of Public
     Roads)  as reported by Mitchell  in Ref. [11]. The
     numerical results calculated  in the present study are
     indistinguishable from those calculated  by Mitchell,
     and  both  of  these  agree  reasonably  well  with
     measured values.
   3. A comparison has been made between  results for a
     circular  cross section and  for  a  rectangular cross
     section  in a channel of fixed length using the same
     input hydrograph  for  both  configurations. The
     rectangular channel width was  the diameter of the
     circle, and  the  hydraulic radii of the two were the
     same when the circular channel was half full. The
     outflow hydrographs in the  two cases are so similar
                                                    133

-------
      that the routing process does not seem  to depend
      too closely on the cross-sectional shape for rela-
      tively minor differences in shape.
   4.  The attenuation of  the  peak of the hydrograph
      depends  on  the  shape  and  size  of  the  input
      hydrograph.
   5.  The rate of travel of  the peak of the hydrograph is
      dependent  on  the shape and  size of  the  input
      hydrograph.

   Progressive Average Lag Method
      As previously  mentioned, the method  of charac-
   teristics requires  a  comparatively  large amount  of
   computer  time, and a simpler  method  which would
   produce the same results was sought to facilitate rapid
   computation of hydrographs while rain is still falling.
   As  mentioned in  Ref.  [10],  quite  a  number  of
   different   methods  were  tried,  including the Musk-
   ingum. The method initially selected as a substitute is
   the progressive average-lag  method (Ref. [8]), and this
   has been incorporated in program UROM 9.
      The method has  several  variations,  but it is best
   described as shown in Fig. 11. The value of qnj on the
   routed hydrograph depends on the average of a number
   of  equally spaced values  of q  on the inflow hydro-
   graph. In the figure qnj is given by
   Generally
                   nl = 1/3(q,.2+qM
                     Qnl  = - I
                                                 (18)
DISCHARGE
                                               TIME
  Figure  11.   Progressive average lag routing.
  where n is the number of points to be averaged. The lag
  time  tg is determined empirically,  from known flow
  times, or from the method of characteristics.
     Successive use of Eq. (18) will determine the routed
  hydrograph at the end of a certain  reach. The process
  can be repeated over the next reach downstream with.
  the first routed hydrograph becoming the input hydro-
  graph.
     In this project  the values  of   n and tg and  the
number of reaches required for a given channel have
been  determined by  trial  and  error. The  input  and
routed  hydrographs  calculated  by  the method  of
characteristics for the particular channel  have been
taken as correct and different values of n and tg tried
until  the  routed  hydrographs coincided. Again refer-
ence  is made to  Fig 10, where  the  points plotted
represent  the routed hydrograph which best conforms
to  the  hydrograph   calculated  by  the method  of
characteristics.  It  can be seen  that  there is good
agreement between the two methods.  Similar results
were  obtained  for  other  conduits with   diameters
ranging from 4.44 to 13.83 ft and lengths from 2000 ft
to 23,500 ft. As many as three input hydrographs have
been  used as input data for  a particular channel, and
the same  routing constants gave good answers in each
case.
   It  was concluded  that  the progressive  average-lag
method of routing produces the correct routed hydro-
graphs for the channels investigated if the appropriate
routing constants are used. The progressive  average-lag
method  is considerably faster than the  method  of
characteristics,  and  hence it has  been used  in  the
real-time program UROM 9.
   In  addition to routing flow, the routing phase of the
model must  perform two additional  functions: the
combining of flow and the  diversion  of flow. Com-
bining of  flow takes place either where additional flow
enters  an interceptor at an  inlet  or where two
interceptors  meet.  Combining  is  accomplished  by
simple addition  of the flows; this follows from the
superposition principle of unit hydrograph theory.
   Diversion of flow sometimes occurs after the flow
has already entered the interceptor lines. This kind of
diversion  may occur at five points in the interceptor
system to prevent overloading of the main interceptor
lines.
   The  routing  phase  of  the  model  contains  the
subroutines necessary to  perform each of  the three
operations—routing, combining, and diversion. A table
of input data determines which of the three operations
should  be carried out and  provides an  index as to
which hydrographs are involved. The steps listed in the
table  are  carried out  in sequence until all  operations
have  been completed. The program instructions are
thus independent of the area modeled.  The data table
is  in effect a map that shows the layout of the sewer
system  and the  locations of the inlets and diversion
structures. Thus  by  adding  entries  to or  removing
entries  from  the table,  sewer lines  can be added or
removed from the system. Inlets and diversion struc-
tures can be added or removed in the same way.
                                                     134

-------
IV. DISCUSSION OF METHODS
   Throughout the model various methods are used to
simulate the runoff process. Some of these methods are
linear  and some are non-linear.  Linear  methods are
used where the  results  obtained are  believed to  be
within the degree of accuracy desired or where the
information is such  that  only  linear methods can  be
used.

   Initially the  precipitation data are given for each
raingage as a  function of time at some  irregular time
spacing. Linear  interpolation  is  used  to  determine
equivalent rainfall  data  spread  at  regular intervals
(usually 5 minutes). Since the model must  operate in
real time up to the latest rainfall reading available, the
interpolation can be  based only on readings taken prior
to the current  readings—the future  trend  being un-
known—and a linear  interpretation is justified.
   The area!  distribution  of precipitation  over each
sub-watershed is found  by the Thiessen method. An
alternate  method based  on  isohyetal  maps of the
precipitation,  according to which the isohyetal lines are
determined  by  linear  interpolation between stations,
has been found by  Kwan [12] to give  results which
agree well with those of  the Thiessen method, and this
agreement served to substantiate the use of the  more
rapid Thiessen method.
   Determination of the sub-watershed losses and the
resulting  excess  precipitation  is  a non-linear process
and has been modeled as such.
   The unit hydrograph method is used in this program
to  determine the   inlet  hydrograph  resulting   from
excess precipitation. It has been shown by Machmeier
and Larson  [13] and others that the runoff process is
actually non-linear. With increased excess precipitation
the magnitude of the peak  discharge increases non-
linearly.  In addition, the time to peak discharge de-
creases. Most of these studies were concerned with
basins possessing "natural" characteristics. For urban
sewered areas other conditions might cause the runoff
process to  behave quite differently.  Changing control
at different discharges can cause marked changes in the
runoff distribution.  This may result from constrictions
in the system,  clogging of  inlets,  or various other
factors. Also, a sewered area  is actually served by two
drainage systems, the primary system  (i.e.,  the storm
sewers)—which is designed to handle  flows up to  a
given  design value-and the secondary  system (i.e.,
natural or manmade surface drainage), which should be
available to handle flows above the design value of the
primary system. Factors such as these may well mask
out or reverse  the  commonly  expected "non-linear
effects."  Thus it was felt that at  least for the initial
approach to the formation of the inlet hydrograph, the
unit hydrograph method is justified.
   The diversion structure  analysis,  as  shown  pre-
viously, is non-linear when  the discharges in the trunk
sewer cause overtopping of the fabridam. The use of
the lumping  factor to  account  for  area not  actually
contributing to a major inlet assumes a linear relation-
ship between  the response of the actual contributing
area  and   the  total lumped  area.  This  assumption
provides  an  easy  method  of  accounting  for flow
entering at major inlets. The  use of the lumping factor
tends to overemphasize the effect of some  of  the
variable diversion structures.
   The progressive  average-lag routing method is en-
tirely linear.
                                                   135

-------
 V.  MODEL   PERFORMANCE   FOR  NATURAL
 EVENTS
    Rainfall for the months of June, July, August, and
 September  of 1969  was 7.36  inches. The  normal
 expected for this period would be 12.88 inches. The
 dryness of the 1969 runoff season, in combination with
 hardware and software problems of a new computer
 system, limited the number of runoff events that were
 recorded  during  1969.  From  the  data  that  were
 available, four events were selected on which to test
and fit the runoff model. These four events occurred
on  Julian  days 207,  211,  218,  and  267,  which
correspond to  July  26,  July  30, August 6,  and
September  24,  1969 respectively.  The precipitation
characteristics for each  of these storms are shown  in
Figs.  12a, b,  c,  and d, respectively. The areal distri-
bution of precipitation  for the entire storm is shown
by the isohyetal lines in the figures. The time distri-
bution is given  for four of the recording gages in the
network.
Figure  12a. Temporal and areal distribution of storm
           on  day  207 (July 26, 1969)
Figure  12b. Temporal and areal distribution  of storm
           pn day 211 (July 30,  1969)
Figure 12c. Temporal and areal distribution of storm
           on  day 218 (Aug. 6, 1969)
Figure  12d. Temporal  and areal distribution of storm
           on  day  267 (Sept. 24, 1969)
                                                  137

-------
       The first event (day 207) was centered north of the
    Twin Cities. Intensities of 2 in. per hour were recorded
    at gage number 8. The southern-most portions of the
    cities received  no  rain  at  all. The significant rainfall
    lasted about 20 minutes.  Rainfall data were recorded at
    about 5-minute intervals.
       The  second event (day 211) was centered  to the
    west. Moderate  rainfall  occurred throughout the area.
    Intensities reached 0.8 in. per hour in most areas. The
    rainfall  lasted about 1-1/2 hours. The rainfall data were
    recorded at about 5-minute intervals.
       The third event (day 218) was centered to the south
    of the  cities.  Intensities as great as  3.8  in. per hour
    were  recorded at  gage  number one. The northwest
    corner of  the cities received  only light showers.  The
    significant rainfall lasted  20 minutes or  less. Data were
    recorded at about 5-minute intervals.
       The fourth event (day 267)  was also centered to the
    south of the cities. Very mild  intensities occurred over
    a period of three hours.  Data  for the early part of the
    storm were recorded  at  one-hour intervals, and there-
    fore  5-minute intensities are shown as uniform  over a
    one-hour period.
       The effect of  each of these storms on the Minne-
    haha inlet will  be shown. This  inlet  is  located  in the
    southwest  corner of the  Twin Cities,  as shown in  Fig.
    1. Only a  portion of the sub-watershed is serviced by
    combined  sewers contributing to  the interceptor  sys-
    tem. The area serviced is  1.14 sq mi. The farthest point
    in the watershed  is about 3.7  mi from the interceptor
    inlet.
       Figures  13a, b, and c show the rainfall and runoff at
    the Minnehaha  inlet. The average precipitation  and
    calculated excess (darkened area)  is shown at the top
    of each  figure. The computed hydrograph at the inlet
    before diversion  and  after diversion is shown  at the
    bottom  of each figure.
       On  day 207  no rain occurred  over  the southern
    portion  of the cities. Since  only sanitary base flow
    would appear at the Minnehaha inlet, a  figure was not
    prepared for this day.
                                 Inlflt njdrogroph

                                 Diverted nydrograph. lob OOv

                                 OivtrlM ri.droOrOpri, fob up
           20    21     22    23
Figure 13b.  Inlet hydrograph and  hydrograph  diverted to
             the interceptor sewer  at Minnehaha  diversion
             for event of day 218.
                               inlet hydrogropri

                               Divined hydrogioph, lob do*i
Figure 13a. Inlet hydrograph  and hydrograph diverted  to
            the  interceptor sewer at Minnehaha diversion
            for event of day 211.
Figure 13c.  Inlet hydrograph  and hydrograph diverted to
             the  interceptor sewer at Minnehaha diversion
             for event of day 267.
     On  day  211  (Fig.  13a)  significant  precipitation
   excess occurred in two bursts over a period of about 1
   hour.  All  of  the  excess  for  this  event  came  from
   impervious surfaces.  During  the storm the initial loss
   on the  pervious  surfaces was satisfied,  but rainfall
   intensities  did  not  exceed  the  computed  pervious
   infiltration rate.  In  the  impervious analysis the loss
   coefficient had been  reduced to 0.44 by the end of the
   storm.
     On day 218 (Fig.  13b) a very  intense short storm
   occurred.  In this case excess came  from both pervious
   and impervious  surfaces.  The  pervious contribution
   came during two 5 minute  periods when the rainfall
   rate of 3.42 in/hr exceeded the infiltration rate of 1.97
   in/hr giving a  pervious excess of 1.46 in/hr, and  the
   rainfall rate of 3.79 in/hr exceeded the infiltration rate
   of 1.83 in/hr giving a  pervious excess of 1.94 in/hr. The
   impervious surfaces contributed excess during all per-
   iods of rainfall. At the end of the storm the impervious
   loss coefficient had been reduced to 0.28.
     For  the storm of  day 267 (Fig. 13c) all the excess
   came from the impervious surfaces. At the end of the
   storm  the  impervious loss coefficient had been reduced
   to  0.50. The  final pervious loss rate was 1.84 in/hr,
   much greater than any measured rainfall intensities.
     The  inlet hydrograph produced  from  the  total
   excess is shown by the solid  line in figures 13a, b, and
                                                       138

-------
 c. The precipitation  excess  for  days  211  and 267
 occurred over a period of 1-1/2 hours or more and
 produced hydrographs with peak discharges of 40 cfs
 and 33 cfs, respectively. In contrast, the large excesses
 for day 218 occurred  in only a 10 minute period, and
 produced a peaked  inlet hydrograph with a maximum
 discharge of 105 cfs.
    The clipping action of the  diversion  structures is
 also seen in these figures. Two postulated conditions
 are shown in  each  figure, fabridam up and fabridam
 down. On days 211  and 267  (Figs. 13a and 13c) with
 the fabridam up, the inlet hydrograph (	) and
 the diverted  hydrograph  (A  A) coincide, indicating
 that no  flow was released to the river. On the other
hand with the fabridam down (	) a portion of
the flow would be discharged to the river. Fig 13b (day
218} shows an inlet hydrograph with a peak discharge
of 105  cfs. For this large discharge in the trunk, flow
was released to the river even with the fabridam up.
   The  interceptor flow from the City of Minneapolis
is  metered at the  Minneapolis-St. Paul  city line. The
location of  the  meters  are  shown  in Figure 1. The
meters  are designated NW meter,  SW  meter,  and  E
meter for the  respective  interceptor sewers they  serve.
The  metered  (	)  and predicted  (	)
discharges for the SW  and E meters  are shown  in
Figures 14a, 14b, 14c, and 14d. The NW meter  is not
shown because on day 211 and 218 it was found that
         150
         100
Discharge
   (cfs)
         50
                               S.W.  Meter
            19        21        23
                          Time  (hrs)
        150
        100

Discharge
   (cfs)

         50
                               East   Meter
           19       21       23
                         Time  (hrs)
                      Figure 14a. Metered and model predicted flows for event of day 207.
         ISO
         100
Discharge
  (cfs)
          50
                               S.W.  Meter
         150
                                                                100
Discharge
 (cfs)
          50
                               East  Meter
            19        21       23
                            Time (hrs)
            19        21       23         I
                            Time (hrs)
                      Figure 14b. Metered and model predicted flows for event of day 211.
                                                    139

-------
 Discharge
   (cfs)
            150
            100
            50
                                       S.W. Meter
                                           150
                                           100
                                Discharge
                                   (cfs)
                                            50
                                                                                             East  Meter
               19
 21       23        I
      Time  (hrs)
                                        19       21        23        I
                                                      Time (hrs)
                       Figure 14c. Metered and model predicted flows for event of day 218.
           ISO
           IOO
Discharge
   (cfs)
           50
                                 S.W. Meter
                               Discharge
                                 (cfs)
                                          ISO
                                          IOO
                                           50
                                                                                       East  Meter
             20
22
    0
Time (hrs)
20
22
                                                                                      0
                                                                                  Time (hrs)
                       Figure 14d. Metered and model predicted flows for event of day 267.
   an abnormal backwater  condition affected the N\N
   meter discharge. This condition was caused by the then
   unknown slippage of a gate that controls the flow of
   the sewage from the west side of the river to the east
   side. The  SW meter discharges were  effected by this
   same condition but to a lesser extent.
      The results in Figures 14a, b, c, and d indicate that
   generally the runoff process is being modelled ade-
   quately  but that further adjustment of coefficients or
   parameters is necessary. The results  shown were ob-
   tained with only a few changes to the original values of
   the model parameters.
      The predicted  hydrograph at the  headhouse (near
   the MSS6  treatment plant)  is shown in Figures  15a, b,
   c, and d. It is of interest to note that a numeric average
   over the two cities  of  the raingage catches are 0.52"
   and 0.51"  for days 211  and 218 respectively. Thus, the
   same amount of rainfall occurring in different time and
   spatial  patterns (see Figures  12b and 12c) produced
   significantly different  runoff patterns  and volumes.
   The storm  for day  211  actually produced a slightly
   smaller volume of excess precipitation than day 218.
                                  However, for day 211 a significantly larger amount of
                                  the excess precipitation  reached  the  interceptor and
                                  was retained in the system. This illustrates the obvious
                                  effect that the shape of the hydrograph at the diversion
                                  structure has on the amount diverted to the river. For
                                  the same volume of hydrograph much less is lost to the
                                  river if a flatter peak occurs.
                                     In  Figures 15a, b, c, and d the predicted hydrograph
                                  at the headhouse is shown for two conditions. The
                                  solid  line (	)  shows the expected discharge
                                  with all gates open and fabridams  up. The dashed line
                                  (	) shows the expected discharge with all gates
                                  open and fabridams down. For the storms on days 207,
                                  211, 218  and  267, the  percent  increases of volume
                                  retained in the sewer system were  16%, 25%, 41%, and
                                  30%,  respectively. As stated  earlier the lumping  of
                                  some  areas at  major inlets  increases the  apparent
                                  impact of  the variable diversion structures. Thus  the
                                  percentages given above are probably somewhat higher
                                  than the true values, but it may be  stated that there is a
                                  significant  increase  in  the  volume  retained  in  the
                                  system when the fabridams are raised.
                                                     140

-------
         800
         600
 Discharge
   (cfs)
         400
         200
               Day 207
                                                            1200
                                                             900
                                                     Discharge
                                                        (cfs)
                                                             600
            20
                                                              300
                                                                    Day 211
                      22       0        2        4               19       21       23        I         3
                           Time (hrs)                                           Time  (hrs)
Figure 15a.  Effect of fabridam  on model predicted flows   Figure 15b. Effect  of  fabridam on model predicted flows
            reaching waste treatment plant for event of day              reaching waste treatment plant for event of day
            207.                                                   211.
         1200
         900
 Discharge
    (cfs)
         600
         300
                Day 218
                                	Fab. up
                                	Fab. down
                                                            800
                                                            600
                                                      Discharge
                                                        (cfs)
                                                            400 -
                                                            200
             19
                      21
    23
Time  (hrs)
                                                       	Fab. up
                                                       	Fob. down
Figure  15c.  Effect of fabridam  on  model predicted flows    Figure 15d. Effect  of  fabridam on model predicted flows
            reaching waste treatment plant for event of day               reaching waste treatment plant for event of day
            218.                                                    267.
                                                   141

-------
VI. CRITIQUE OF MODEL
   The material  in the previous sections describes the
initial  methods  used in the  model and  the overall
performance  of these methods.  Several assumptions
have  been  made  in developing  the  model.  These
assumptions will be  discussed  in  this  section and  in
addition suggestions for improvement of the model will
be given.
1. The modeling of  the diversion structures is com-
   plicated by  the rubber fabridams.  This device  is
   raised  and  lowered   pneumatically. A  record  is
   maintained of the air pressure  in the fabridam. This
   gives an indication of the  position of the fabridam
   which is modeled  as a variable height weir. Presum-
   ably a given fabridam pressure would be associated
   with a given fabridam height. This is not necessarily
   true as water  build-up behind the fabridam also
   causes a change in the fabridam pressure. Thus, it is
   difficult at  best  to  estimate  the fabridam  height
   from the fabridam pressure. In addition the rubber
   fabridam is deformable.  The fabridam crest is often
   not horizontal and may actually be depressed in a
   vee shape on the sides or in the center under partial
   inflation conditions.

2. The present spacing  of  the remote  indicating rain-
   gages could be improved.  The gages are generally
   spaced around  the Twin  City area. For modeling
   purposes a network centered within the area would
   be  more desirable. Presently gage spacing averages
   about 5 miles, allowing significant variations in areal
   precipitation to go undetected.
3. The present  analysis does  not properly model the
   system for relatively  large  rainfall events. An  urban
   area is actually  served by two drainage systems. The
   primary one, the sewer system, is designed to  handle
   a certain amount of runoff. If the  runoff exceeds
   the capacity of the sewer system, local  ponding and
   surface runoff  will occur.  This surface runoff will
   follow natural or man-made surface channels until it
   finally enters another sewer or some other surface
   drainage facility.  In the model all runoff is assumed
   to enter the sewer system and arrive at the diversion
   structure. This  is a valid assumption for small and
   moderate  amounts of runoff (up to design values).
   For larger amounts where ponding and "natural
   diversion"  occur  the model will predict excessive
   flows in the trunks. While  the diversion structures
   do tend to limit the  flow  to the interceptor there is
   still no upper limit on the amount they will predict
   that goes to the interceptor. Thus, for unreasonably
   large amounts of  trunk  flow the model will predict
   reasonable but  still too large  values of interceptor
   flow.
4. Each time the model is executed the initial moisture
   conditions in the area  are taken to be zero. This
   would be true only  when all surface depressions
   were dry and  the upper soil moisture was  at the
   wilting point.  Actual initial  conditions should be
   established before each real storm event. For moder-
   ate rains where only runoff from impervious areas is
   significant, a dry  initial  condition  is still a reason-
   able assumption.
5.  The use of separate pervious and impervious loss
   functions  is desirable.  Although  the constants  in
   each  function have  been selected by  judgement
   only, the analysis indicates that the most significant
   runoff  contribution in an urban area comes from
   the  impervious area. The pervious  areas contribute
   little runoff during the less intense rainstorms. The
   constants  in  the  loss functions  should  be  sub-
   stantiated by runoff data.
6.  The present model may be used  for various pur-
   poses; real  time  operation  aid, system evaluation,
   design evaluation, etc. As an  operational tool, the
   accuracy  and  reliability of  the  model  may be
   increased by the addition of a feedback correction
   based  on  measured water levels in the sewer system.
   The model in  its existing form does not utilize the
   recorded  water level data to  aid in  its prediction.
   This  feature  would  be  of  use during real time
   applications.  When  evaluating  postulated  storm
   events, recorded  data would  not exist  and the
   prediction would  not  include feedback correction.
7.  If excessive  flow  is indicated  in  the interceptor
   system, the console operator must decide what gates
   or fabridams should be manipulated to alleviate that
   condition. A  gate and  fabridam operation scheme
   may be added  to allow the machine to decide what
   the best remedy would be.'The operator could still
   decide whether  or not to  implement the recom-
   mended gate or fabridam movements.
8.  The addition  of  storage in the trunk sewer behind
   the  fabridams or at  other  locations  would  bd
   another desirable revision to  the  model. Presently
   storage in the  trunk sewers is reflected only in the
   shape  of  the unit  hydrograph.  Accounting for
   storage in the trunk sewers would be of importance
   only where the storage  volume is  significant com-
   pared with the volume of the runoff hydrograph.
9.  Only the 15 major inlets to the interceptor sewer
   are modeled. The flow that enters at other  points in
   the  system is assumed  to contribute at the nearest
   major inlet. It is  assumed that the additional flow
   lumped  at the major  inlets  will  occur  in direct
   proportion  to  the relative land areas contributing.
   That is, if the actual area contributing to a major
   inlet is x  sq mi, and an additional area of y sq mi is
   lumped at this inlet, the computed discharges based
   on the actual  area x will be increased by the  ratio

   -—¥ to represent the flows from the total lumped
    x
   area (x + y). This lumping correction is applied im-
                                                   143

-------
    mediately following  the diversion analysis at each
    inlet.
10.  For verification and fitting purposes it is important
    to have recorded runoff data corresponding to given
    rainfall events. Presently water levels are recorded in
    the  trunk and interceptor sewers in addition to
    venturi metered flows in the  interceptor sewers at
    the  Minneapolis-St. Paul city  line. These recorded
    values are needed to assess the discharges entering or
within the interceptor system. They are of less use
for assessing the  runoff discharges  in  the trunk
sewers. This is due to the problem of modeling the
fabridam  which acts as a control section in  these
lines.  For this  reason it would be very desirable to
install a fixed control section (weir) at some  point
in the trunk  sewer  that  would  enable the  total
runoff to  be measured. This would allow significant
improvement in the  loss rate and runoff  prediction.
                                                    144

-------
                REFERENCES
1. Bowers, C.E.; Harris, Garth S.; and Pabst, Arthur
   F., Status Report on "The  Real-Time Computa-
   tion  of Runoff and Storm  Flow in the Minne-
   apolis-St. Paul Interceptor Sewers," Memorandum
   No. M-118, St. Anthony Falls Hydraulic Labora-
   tory, University of Minnesota, December, 1968.
2. Handbook of Applied Hydrology, Ven Te Chow,
   Editor, McGraw-Hill, 1964.
3. Huggins, L.F.; and Monke, E.J. "The Mathematical
   Simulation  of the Hydrology  of Small Water-
   sheds",  Technical Report No. 1, Water Resources
   Research Center,  Purdue  University, August 1966.
4. Kent,  K.M.;  "A Method  for Estimating  Volume
   and Rate of Runoff in Small Watersheds," USDA,
   SCS-TP-149, January 1968.
5. Harris,  Garth S.,  Status  Report on  "Real-Time
   Estimation of Runoff in  the Minneapolis-St. Paul
   Metropolitan Area," Memorandum No. M-119, St.
   Anthony Falls Hydraulic Laboratory, University
   of Minnesota, December 1968.
6. Anwar, H.O.; "Inflatable Dams",  Journal of the
   Hydraulics Division, ASCE, Vol. 93,  No. 3, May
   1967.
7. Harris,  Garth S., Status Report on "Mathematical
   Models  of  Major  Diversion Structures  in  the
   Minneapolis-St. Paul Interceptor Sewer System,"
    Memorandum  No.  M-120,  St.  Anthony  Falls
    Hydraulic  Laboratory, University of Minnesota,
    December 1968.
 8.  Hydrograph Combining and  Routing, Hydrologic
    Engineering Center, U.S.  Army Engineer District,
    Sacramento, California, August 1966.
 9.  Streeter, V.L. and E. B. Wylie,  Hydraulic Tran-
    sients, McGraw-Hill, New York, 1967.
10.  Harris, Garth S., Status Report on "Development
    of a Computer Program to Route Runoff in the
    Minneapolis-St. Paul Interceptor Sewers," Memo-
    randum No.  M-121, St. Anthony Falls Hydraulic
    Laboratory,  University of Minnesota, December
    1968.
11.  Mitchell, James S., "Comparison of Mathematical
    vs. Experimental Flood Wave Attenuation in Part-
    Full  Pipes," Colorado State University Engineering
    Research Center, March 1967:

12.  Kwan,  J.Y.; Riley, J.P.; and Amisial,  R.A., "A
    Digital Computer Program to Plot Isohyetal  Maps
    and  Calculate Volumes of Precipitation," Pub. No.
    80,  IASH Symposium at Tucson, Arizona, Decem-
    ber  1968.
13.  Machmeier,  Roger  E.  and  Larson,  Curtis L.,
    "Runoff Hydrographs for Mathematical Watershed
    Model," Journal of the Hydraulics Division, ASCE,
    Vol. 94, No. 6, November 1968.
                                               145

-------
                 Appendix A

        UROM-9 OPERATOR'S MANUAL

Section I - Introduction
Section II - Operation of CREATE
   1. Input data file preparation
   2. Execution of CREATE
                                                           — Operation of Mathematical Model —
Section
UROM-9
   1. Input data
   2. Sense switch options
   3. Program control
   4. Gate editor
   5. Execution of model
   6. Loader Memory  map
     (UROM-9)
of CREATE and Model
                                             147

-------
I.  INTRODUCTION
   UROM-9 (Urban Run-Off Model)  is a digital com-
puter program  written for execution on the Minne-
apolis-St.  Paul  Sanitary District's  Digital Equipment
Corporation PDP-9 computer.  The  purpose of this
program is to predict discharges in  the Minneapolis-St.
Paul  Interceptor  Sewers  given rainfall  readings  at
various points around the Twin Cities.
   UROM-9 acquisitions its own data when operating
on  a real  time basis.  Instructions pertaining to  the
operation  of the  data-gathering section of  UROM-9
(RTIME) are not given in  this manual. References are
made to  RTIME  only  to clarify  its relationship  to
UROM-9.

1.  Method
   The rain that occurs at the raingages is first operated
on by a subroutine that determines rainfall losses. The
excess rainfall is then applied  to a unit hydrograph for
each of the catchments modeled.
   The resulting composite hydrograph represents the
flow that would arrive at the major diversion structures
of the inlets to the interceptor sewer system.
   The flow then enters the interceptor. Any flow not
able to pass into the  interceptor is discharged to the
Mississippi  River. The  flow entering the interceptors is
increased by a lumping parameter to account for flow
that does not really enter at one of the 15 major inlets
modeled.
   The flow in the interceptor is then routed and
combined  with  other entering flow until it reaches the
treatment  plant.

2.  CREATE Routine
   The working section of UROM-9  requires  many
items of  input data  that describe  the hydrologic,
diversion and routing  constants of the modeled area.
These data are  fixed for a given area and are handled
separately from information (such as rainfall  readings)
that pertain to a given storm  event.  CREATE is a
one-chain-execute  file, run prior to the actual model
program UROM-9, that prepares the  fixed data in a
form that is easily used by the model.
   From the given input data  file on the disk (DAT 3)
CREATE lists the  input data on the line printer  (DAT
5)  and then writes 4  binary output files on the disk
(DAT 8).   Each  of these  4 binary files is used by a
different chain of the model.
   CREATE also does a small number of calculations.
The data are not screened for errors other than  those
detected by the system routines.
   When new input data are to be run by the model, or
if old data have been changed,  it is necessary to then
execute CREATE  before  running the model.  Once
CREATE has been executed the model may be run any
number of times.
3. Structure of Model, UROM-9
   The  model  itself  (UROM-9)  is  composed  of 6
chains, or sections. (See Fig. 1) At any given time only
one  chain may be  loaded  into  core  storage. The
remaining chains are held  on disk storage until  called
by the chain in core. The data in unlabeled common
storage are retained in core from chain to chain.
a.  Chain 1
   This  chain  initializes  the  model  and  is  always
executed first.  Its functions  include initializing  vari-
ables and constants, loading the common storage area,
setting the  real time clock,  setting initial  loss rate
parameters to zero,  inputting the rainfall file to be
processed, and printing instructions pertaining to sense
switches  used by  the model. Control is always passed
to Chain 2 upon completion of Chain 1.
   Use of sense switches is explained in detail in  III.
b. Chain 2
   Chain 2  is the basic  control chain for  the  model.
After  any other chain is completed, control  is  always
passed to Chain 2. Chain 2 then determines the proper
chain  to  be called next. This decision is based  on the
previous  chain  processed  and  the  positions  of  the
various  sense switches selected  by the operator.  In
addition  the PROGRAM  CONTROL routine  within
Chain 2  allows the  operator to override  the  normal
processing  sequence.  Use  of the PROGRAM CON-
TROL options is explained  in detail in Section 4.
   If real time data acquisition is desired, the scanning
chain  (RTIME)  (Chain 9) is called each time Chain 2 is
entered.  Upon  completion  of  scanning,  control is
returned to Chain 2 for further model processing.
                                                  149

-------
   This chain also contains the GATE EDITOR rou-
tine. This routine is explained in Section 5.
c. Chain 3
   This chain  begins  the actual  model calculations.
When  processing  under  real  time conditions, initially
data 2  hours prior to the current real time are used.
Once this initial block is processed-.  Chain 3 updates
the  inlet  hydrographs based on  rainfall readings that
are being acquisitioned in real time.
   When historical data  is being  processed, processing
begins at the time specified by the operator, and each
prediction is updated  by exactly one hour of rainfall
data.
   Output from Chain  3 may include the loss rate
calculations and the inlet hydrographs at the 15 major
inlets to the interceptor sewers.

d. Chain 4
   Chain 4  performs  the  diversions  of inlet  hydro-
graphs to determine the flow to the interceptor sewers.
Fourteen of the inlet hydrographs are diverted by this
chain.  The  15th inlet hydrograph and the flow  at 4
additional  locations within  the  interceptors are di-
verted during the routing process of Chain 5.
   Chain 4 always uses the hydrographs produced by
the last run of Chain 3 for input data. Chain 4 may be
run  any number  of  times  with  different gate and
Fabridam settings without re-running Chain 3.
   Output  from  Chain 4  may include the diverted
hydrographs  at  the major inlets,  the diverted  hydro-
graphs after lumping of  additional flow at the  major
inlets, and the volumes of the undiverted and diverted
hydrographs.
e.  Chain 5
   Chain 5 generally  performs the routing and com-
bining  operations. In addition it may divert flow at 5
points  within the interceptor sewers.  Its output is the
interceptor flow at various locations.
f.  Chain9 (RTIME)
   This chain handles the acquisition of data. It is run
each time control transfers from one model chain to
the next. A  detailed  discussion of  the operation of
Chain 9 is not given in this manual.
                                                   150

-------
II.  OPERATION OF CREATE
1. Input Data File
   The input data file (INDAT8 LSI, INDAT9 LSI or
others) resides on the disk  (DAT 3). Changes to these
files should be  made with  the system  EDITOR Pro-
gram.  A  sample  listing  of  INDAT9 LST  is given in
exhibit  1.   In  general,  the constants  pertaining  to
hydrology  are given first,  then diversion,  and finally
routing.
   Each time CREATE  is executed a complete listing
of input  data and associated headings  is given on the
line  printer (DAT  5). This listing, given in the  next
                                                          part, contains the same information that appears in the
                                                          input data file,  but in a different format, and includes
                                                          titles and other  identifying information.
                                                             When  making changes to the  input data a copy of
                                                          the output from CREATE will  aid  in  identifying the.
                                                          parameters, but an  actual  listing of the  data  file  is
                                                          necessary where the proper format is in question.
                                                             INDAT8 LST is the standard input data file for the
                                                          8 raingage network.
                                                             INDAT9 LST is the standard input data file for the
                                                          9 raingage network.
                              Exhibit 1
       Sample input data file used by CREATE (INDAT9LST)
  900.
  IB 00.
  810.
  900.
  1000
  ueo
  1060
  1200
  1200
  1800
  1800
  1600
  11.00
  3600
  2000
       IS  100.
         7200.
         16200.
         2230.
         3300.
         10800.
          3960.
         2880.
         7800.
         18000.
         7200.
         moo.
          12000.
         10000.
          7200.
          6600.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0 .00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
 0.00 0.02  2.00 0.
57200000.
98900000.
29600000.
1)6400000.
75900000.
49100000.
25 100000.
5)400000.
2940000C.
36800000.
76600000.
31600000.
64600000.
61800000.
12800000.
 1244  641
 1641 1144
 1441 1141
 1511 1644
 1144 1441
 1 141 1641
 1644 1244
  641 1511
  633  6)3
                 95 0.31 0.
                 95 0.55 0.
                 95 0.30 0.
                 95 0. 33 0.
                 95 0.33 0.
                 95 0.36 0.
                 95 0.30 0.
                 95 0.52 0.
                 95 0.30 0.
                 95 0.39 0.
                 95 0.30 0.
                 95 0.38 0
                 95 0.29 0.
                 95 0.3 1 0
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
70 0.20
2.00 0.
2 . 00 0.
2.00 0.
2.00 0.
 .00 0.
 .00 0.
 .00 0.
2.00 0.
2.00 0
2.00 0.
2.00 0,
2.00 0
2.00 0
2.00 0.
2.00 0
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
0.25 0.00
1 .00
1 .00
1 .00
1 .00
 .00
 .00
 .00
 .00
 .00
 .00
 .00
 .00
 .00
 .00
 .00
           ).96
            1.06
            I. 58
            1.52
            1.0
            1.6
            2.85
            1.0
            1.82
            2.73
            1.2
            1. 13
            1.33
            1.48
            1.56
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
2.00 0.00
 0.0
 .2
 0.0

 0.4

 0.6

 0.2
 .2
 0.0
 1.
         0.)

         0.4

         1.0

         0.8

         0.6

         0.4

        0.1
        .3
        0.5
                                                       151

-------
.00
0.0
0.0
PH AL N
III GUI
HELLO
RICE
HARSH
EUST 1
OAK
IUNOO
18IH
HINNE
26TH
6IH
42 ND
2ND
EASin
NU N
SU M
1
200
0
168
)
0
J.
0.
0. b
.026
0.
6.
1.5
0.
0.
2.5
0.
0.
-1 .7
0.
0.
1.5
1. 7
9. 67
0.
0.
48
4
1
1
2
3
1
2
1
4
1
2
1
I
)
2
1
4
1
]
1
2
1
3
1
2
1
2
1
4
1
2
1
2
I
i
I








2
1
2
1
5.97
9.66
5. 17
6. 7
22.4
5. 8
1. 02
0. 70
1.20
1.11
.05
0.9
0.5

2 4
200 200
0 «4
272 242
1 0
0 0
].
1.
0.6
.026
.026
6.
1.5
1.5
0.
i.»
3.
0.
-2.S5
0.
0.
5.5
9.25
10.
10.
0.

15
15
15
15 14
15 0
15
IS 1}
15
12
li
12 1 1
12
12
15
15 12
15
10
10
10
10
10 9
10
10
10
10 «
10
15 10
15
7
I
1 6
7
15 7
IS
15 S
15
4
4
3
J
) 4
4
15 4
15
15 2
15
15 1
15
0.0007)
0.0007
0.00095
16.0
11 .5

0.91
0.69
I.2J
1.08
.05
.5
.0

5 6
7 }
0 0
10 40
2 1

J.I
3.
0.6
3.
.026
».
.6
1.5
0.
3.
2.75
0.
0.
0.
0.
9.5
2.5
5.T
10. 5
0.

15
15 2
15 2
15
15 12
15 6
15
15 2
12
12 6
12
12 a
12 17
15 16
15
15 2
10
10 6
10 14
10 6
10
10 15
10 1 5
10 6
10
10 )
15
15 2
7
7 5
7
7 2
15
15 2
15
15 2

2

4

3
15
15 B
15
15 2
15
15 8
0.012
0.012
0.012
1.8
10.2

0.43
0.73
1.25
1 .04
1 .

8

0

0

).
).




1.
1.

).
4.

-6
0.

4.

9.
8.
5.


-11
1

1
-7

0

-24

-I
7
7

1

-10
1
-4

1
7
0

0

0

1

1

-10

1

1

- 14

0

1

1

1



b.
16

0.
0.
1 .

0

9

0

1




026
026

5
5




.5


1






25
3

0
6

1

10

2
1
2

1

12

5

1
3
2

2

1

2

1

120

1

1

6

1

1

1

1



1
.0

98
82
26



10 1 1

18 0

0 1

3.
1.

.026
.026

1.5
1.5

3.
8.

-0.7
-0.6

4.25

9.75
7.5
0.

10.0
150.
250.
1290.
0290.
4290.
5360.
6360.
20.0
710.
8165.
9165.
10165.
11360.
12530.
13530.
5.0
1460.
1S60.
1660.
17125.
18125.
19125.
20125.
21200.
22200.
23600.
24600.

2519.3
2690.
2790.
28690.
29690.
30690.
31690.

3270.

3355.
34170.
3S170.
36820.
37820.
38820.
39820.
40900.
41900.



8.9
32.0

0.83
0.94
1.24



12 13 14 15 200

0000 209

11101

3. 3.
4.1 3.

1.5 .026
28.5 .026

.171 1.5
.6 1.5

0. 2.750
4. 4.

0. 0.
0. 0.

0.6 6.

7.5 9.0
0. 7.


42U-I
4 T Hill
6TMUA
6tMU«
6TWI
26S4U
26S4A
NUICT
nlNMi
38 IH
38 (•
SUM 1
SUNED
HUnfD
HE ICO
rflUH
RKLf 1
2NIMA
2NDHA
OAK
0»
ENEI
E1EIO
EUSI 1
EUSI 1
nEJUN
tIEJUC
HRBJN
jEFf 1
HUSH
HARSH
HR8JN
HRBJC
RICf
RICE
UIBIN
HELfll
UABJN
OAMII
VHOJN
UtBJC
UAB1N
UABIC
mriR
INIIC
IMIPH
IHIPC
INI«H



3.1 1.8
14.5 24.0

0.19 0.74
I.O5 1.14
1.20 1.15

46 44  39 37 34  30 29 24 20  17 10  9
                                            3  1
                                                                  152

-------
2. Execution of CREATE
   CREATE is an execute file that normally resides on
the disk unit. It is run from the system monitor by the
command

                  $E  CREATED)

(where    )   indicates  a carriage return).

   The message then appears on the console;

       ENTER NAME OF INPUT DATA FILE

one of the input data file names should then be typed
by the operator as,

                  INDAT9LSO

   If  the  name  given  by  the operator  is valid the
following message appears, otherwise another file name
is requested.

         ENTER TIME AND DATE (15AS)

   The time, date, and any other information may then
be entered on one line.  Whatever is entered appears on
the Line Printer listing for identification purposes.
   Upon  completion  the following  appears on the
console:

                  STOP 000000

and control then returns to the system monitor.
                            An  example of an execution of CREATE is shown
                          in exhibits 2 and 3.
                                             Exhibit 2
                                Sample teletype output from CREATE


                               MONITOR  V4B

                               SE  CREATE


                               ENTER  NAME OF  INPUT  DATA  PILE

                               RA

                               ENTER  NAME OF  INPUT  DATA  PILE

                               INDAT9 LST

                               ENTER  NAME OF  INPUT  DATA  PILE

                               INOAT9LST

                               ENTER  TIME AND DATE  C15AS)

                                12S42 OCT 8/1970

                               STOP   000000
                                               Exhibit 3
                                Sample line printer output from CREATE
   121 «2 OCI S.I970
   INPUT DM* FILE USl 0 - INDAT9LSI
   NO OF GAUGES  9
   NO OF CACHHENIS  15
   II IE  INCKEnENI OF RAINFALL HEADINGS
   TIKE  INCREMENT OF HTOROCRAPHS     ) Ou .
       100. SECS
      iECS
II HE 10 PEAK
900.
1800.
»3u.
too.
1000 .
1480.
1060.
\ 100.
woo.
laoo.
1 BOO.
)600.
11.00.
lt.00.
200(1.
BASE TIME
7200.
16100.
2230.
1)00.
10800.
I960.
2S80.
7AOO .
18000.
7200.
14400.
12000.
10000.
7200.
b600.
CACn

;







10
1 1
1 2
1 J
1 4
1 5
   LOSS DATE VARIABLES
   NO. TSTLOS RECOV All OECA! PE5V PfRVCF cHOK I BMC B.1L SILOS t'.ILO<°
                                                      SCLUSP COEF POKER ACCUSS
1    0.00   0.02
2    o.oo   o.o;
)    0.00   0.02
4    0.00   0.02
5    0.00   0.02
                      2.00
                      2.00
                      2.00
                      2.00
                      2.00
0. 95
o.v5
0.95
0.95
0. 95
0. )1
O.-S
J. 10
0. ))
0. 31
0. 10
0. 70
0. '0
J. 70
o. ro
3.20
J.20
0.20
2. 00
;. oo
•. oo
2. 00
2. 00
0.00
J.OO
0.00
0.00
0.00
0.00
3.00
o.oc
0.00
0.00
3. 25
0.2S
0. 25
0. 25
0. 25
0.00
0.00
0.00
0.00
0.00
1.00
1.00
1.00
1.00
1.00
2.CO   3.00
2.00   3.00
2.00   0.00
2.00   0.00
2.00   0.00
                                                 153

-------
6 0.00 0. 02 2.00 0.11 0.16 0.70 0.20 2.00 0.00 0.00 0.25 0.00
7 0.00 0.02 2.00 0.95 J.JO O./O 0.20 t . JO 0.00 0.00 3.25 0.00
8 0.00 0.02 2.00 9.95 0.52 O.fO 0.20 2.00 0.00 0.30 0.25 0.00
9 0.00 0.02 2.00 0.95 0.10 d. 70 0.20 £ . UO 0.00 0.00 0.2S 0.00
10 0.00 0.02 2.00 0.9V 0.39 0.70 0.20 2.00 0.00 0.00 0.25 0.00
11 0.00 0.02 2.00 3.95 0.30 0.70 0.20 .'.00 0.00 0.00 0. 2S 0.00
12 0.00 0.02 2.00 0.95 0.38 0.70 0.20 I . 00 0.00 0.00 0.25 0.00
11 0.00 0.02 2.00 0.95 0.29 0.70 0.20 .'.00 0.00 I). 00 : . 25 0.00
1* 0.00 0.02 2.00 0.95 0.31 O./O 0.20 J . 00 0.00 0.00 0.25 0.00
15 0.00 0.02 2.00 0.95 0.33 O./O 0.20 ' . OC 0.00 0.00 0.25 0 00
IREASISO FTI FKIOHS CUCH.NJ
57199488. .90 i
98899488. . 06 2
295997*4. .58 ]
1)6400000. .52 4
75899488. .00 5
49099744 . .00 b
25099872. .85 7
5 J 399744 . . 00 8
29199744. .32 9
1 6799744. . 7J 10
'6599488. .20 11
11599744. .1) 12
64599488. . JJ 11
61199488. .48 14
1 2799936. .58 15
RMNGACE ADDRESSES IN ORDER USED BT MODEL AND SCANNED
1 1244 641
2 1641 114
1 1441 114
4 1511 164
5 1144 144
6 1141 164
7 1644 124
8 641 1511
9 631 6))
RUN GAUGE PROPORTIONS
GAUGE NO 12 ] 4 5 6 7 B 9
0.000 0.000 0.000 0. 000 0.000 0.900 0.100 0.000 0.000 1
0.000 0.000 0.000 0.000 0.000 0.100 0.400 0.100 0.000 2
0.700 0.000 0.000 0.000 0.000 0.000 0.111(1 0.000 0.000 3
0.000 0.000 0.000 0.000 0.000 0.000 0. '00 0.000 O.JOO 4
0.000 0.000 0.000 0.000 0.000 0.000 0.500 0.100 0.400 5
0.000 0.000 0.000 C.OOO 0.000 0.000 0.000 0-001) 1.000 0
0.200 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.800 7
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.400 0.600 8
0.000 0.000 0.000 0.000 0.000 0.000 U.OOO 0.600 0.400 9
0.000 0.000 0.000 0. '00 0.000 0.000 C.OOO 0.200 0.1UO ;0
0.200 0.300 0.000 0.000 0.000 0.000 0.000 0. 000 0.500 11
1.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 12
0.000 0.050 0.050 0.003 0.000 3.000 0.000 0.000 0.900 13
0.000 0.000 0.500 0.000 0.000 0.000 D.OOO 0.000 0.500 14
0.000 0.000 0.000 1. 00.) 0.000 0.000 0.000 0. OOu 0.000 15
HTDKOCRAPH NO AND DIVERSION NO
1 1
2 2
4 1
5 4
6 5
8 6
9 7
10 a
11 9
12 10
11 11
14 12
15 1)
200 14
200 15
200 16
200 17
7 18
3 19
INDEX OF DIVERSION POINT
0 1
0 2
94 J
0 4
0 5
0 6
0 7
18 8
0 9
0 10
0 11
0 12
0 13
209 14
16> 15
272 16
242 17
30 18
40 19
.00 2.00 0.00
.00 2.00 0.00
.00 2.00 0.00
.00 2.00 0.00
.00 2.00 0.00
.00 2.00 0.00
.00 2.00 0.00
.00 2.00 0.00
i.OO 2.00 0.00
1.00 2.00 0.00
154

-------
ORIFICE ITPE
) 1
I 2
0 )
2 4
1 S
0 6
1 7
0 8
1 9
1 10
1 11
1 12
0 1 I
1 14
0 15
0 16
0

1
2
3
4
5
6
7
8
9
1 1
12
1 3
15
16
1 7

Cl i
17

2
2
3
2
2
2
2
Cl
99998
29998
99998
9999B
99998
99998
0.00000
2
2
2
4
0
0
0

99998
99918
99998
09996
60000
60000
60000


C2
0.02»00
2.19998
0
0
1
0
0
0
0
0
28
5
5
5

1 .69999
F«UR IDA
1
2
3






10
11
12
13
14
1 5
16
17





7
7
0
9
10
02600
02600
49998
02600
00000
02600
02600
02600
49957
99996
99V96
99993

C 14 9

I
0
\
i
0
1
0
1
1
1
0
0
0
0

C3
49999
60000
49999
41999
1 7800
49999
00000
49999
49999
41999
60000
00000
00000
00000

24984 C15
C.
J. 49997
. .99998
2.9)998
2 .99998
0 .00000
2.7*995
0.00000
2 . 99998
i. 9 99 95
7 .99990
1.9)995
0.00000
0 .00000
0.03000

2 .49997

-1
0
-5
-0
0
0
0
0
0
-0
0
0
0
0


C5
69998
00000
4999 3
70000
00000
00000
00000
00003
00000
60000
00000
COOOO
COOOO
00000


Co














1 . 49999 5

C7 ca














. 499 90 9 . 49990

OUnETEHS
66990
99184
69996
99990
74984
49990
99990
00000
99984
49984


























































































7.99990
7.49190
0
6
0
0
0
00000
99990
00000
00000
00000
UPERAMON TABLE.
OPE
















































R
1
2
3
4
5
6
7
a
9
10
1 1
1 2
1 3
14
15
16
1 7
IB
19
20
21
22
2 J
24
25
26
27
28
29
30
31
32
) 3
34
35
36
37
38
31
40
41
42
4 3
44
45
46
4 7
48










HTOROCRAPH
Hit) ENTER
4
1
I
2
3
1
2
1
4
1
2
1
3
3
2
1
4
1
1
1
2

1
1
2
1
2
1
4
1
2
1
2
1
2
1
4
1
4
1
2
1
2
1
2
1
2
1



















































































NOS AND KG'JIIHG CUKSIS
HID COMB
15
15
15
15
15
15
15
15
12
1 1
11
U
12
15
!<;
IS
13
13
10
10
19
10
13
13
10
10
15
15
J
I
r
i
15
15
15
15
,;
/,
1
t
1
i.
IS
s
«,
i;
'i
!i
















































HTO LE A Vc NST RL
0
3
3
14
0
0
1 3
0
0
0
11
0
0
0
12
0
0
0
0
0
9
0
0
0
H
0
13
0
0
U
6
0
/
•J
5
3
0
0
0
0
4
0
4
0
2
0
1
0
.5
15
|5
15
1-5 I
15
15
15
12
1 2
i2
1 2
1 2 1
1 5 1
15
i 5
10
13
10 1
10
10
10 1
0
2
2
0
2
6
0
2
0
n
0
a
7
6
0
2
0
o
i
6
0
•,
.0 15
.0
lO
10
15
1 5
r
r
t
i
15
1 •>
IS
15
4
*
1
I
4
4
15
15
15
t 5
15
;5
6
0
t
0
2
0
5
0
2
0
J
0
J
0
2
0
4
0
3
0
a
0
2
0
d
LAC
0
- 1 I
1
0
1
- 7
0
0
0
-2 <.
0

7
J
0
1
0
-10
1
- 4
0
I
7
J
0
0
3
0
0
I
0
1
0
- 1 0
0
1
0
1
0
-14
0
0
0
1
0
1
C
1
N?C HS
0
25
1
0
0
6
0
1
0
I 0
0
2
1
2
0
1
0
1 2
0
5
0
1
3
2
0
2
0
1
0
j
0
1
0
120
0
1
0
1
0
6
0
J
0
1
0
1
0
1
Ofl AX
I0.042N-I
50.04THNI
50.06IMUA
2)0. 06THUA
250. 06 1HUA
2S0.02t>SEA
360.026SEA
36U.ONUHEI
20.0MINMI
70. 0 J 8 1 h
165.038TH
165 . OSUME T
165.0SUKED
)LO. ONwntD
5 lO.OrtETCO
5 10.0METON
S.ORMM
60. 02 NOnA
60.02NCHA
60. OOAK
I25.00AK
l<5.0f IE!
125.0EME10
125. OEU^>; 1
200.0EUSI 1
200. Or.E JON
600.0ME JUC
600. OrtRB JN
0. OjEff I
19. } MAHS
lO.OMAR'.M
°0. OMRdJN
690. O.lfiHJC
610. OK ICE
6HO.OR ICE
6*rO. OUAfll N
0. OkELUI
'0. OWAUJN
O.OOAVIll
•)5.0UABJN
1 ro.OUABJC
170. OUABIN
SJO.OUABIC
820. 0 INT TR
820.0IN1IC
820. 01 NIPH
900. 01 NTPC
900. 01 NIHH
155

-------


d«SE HOU
1
2
3
4
5
6
7
a
i
10
ii
12
11
It
IS
OIAH : 5.97 SLOPE = O.U007J BANKINGS N = O.OU
OIAM = 9.66 ^LLOPE = 0.00070 MANNINGS N = O.U12
01 AH = 5.97 SLUPt = 0.00095 HANNINCS N = O.Sll

6.7
16.0
1.8
6.1
«.»
).l
1.6
Ji.«
11 .S
10.2
16.0
32.0
14. S
24.0
S.B
HOURL*  BASE  FLOU  f AC TORS
























NA KE









10
II
12
1)
14
15
HP
IIB
1
I
)
t.
5
6
;
a
9
10
11
12
13
14
15
16
17
16
19
20
21
22
23
24
INOEI
46
44
39
Jl
34
30
29
24
20
17
10
9
6
3
1
3
24
1.02
0.98
0.93
O.B8
0.91
0. 7f
0. 74
0. 70
0.69
0. 73
0.«2
0.94
I.OS
1 . 14
1 .20
1.23
1.25
1.26
1.24
1.20
1. 15
1.11
1 .06
1 .04
















6
54
                   3    10     S     4     4    II     6     6    12     S    12     7
                  II    30    13    10    26    60    24   48    40    31    24   22
                                                           156

-------
    OPERATION
    UROM-9
OF  MATHEMATICAL  MODEL-
1. Input Data
   All input data that are fixed for all  events are
processed by CREATE as explained in Section I. These
data  are  prepared as 4  binary  files. The first file
(STARTLIST)  is  used  by Chain  1  to initialize the
model.  The  remaining  files,  HYDROLIST,  DIVER-
LIST, ROUTELIST, are used by Chains 3, 4,  and  5
respectively.
   In addition to the fixed data the model also requires
input data  consisting  of  the rainfall  at the remote
gages, the gate and Fabridam settings to be used, and
the initial moisture conditions before the storm.
   When processing real time data the  model uses the
file RAINRP DLY to obtain its rainfall data. New data
are continually  being added  to this file by  chain  9
(RTIME).
   When historical data are being processed, rainfall
data  are  obtained from  any  file specified by the
operator.  If  new  real time  rainfall data  will also be
acquisitioned, they will be  put in the file RAINRP
DLY.
   The gate and Fabridam settings that the  model will
use are stored in a file GATESETTG. This file contains
the times and positions of gates and Fabridams, as well
as the maximum  and minimum values allowable. All
positions are in feet. This file may be changed by the
GATE EDITOR routine (part of the model) or by the
systems EDITOR program.
   The current  values of  basin moisture  and starting
loss  are found  in the  file BMCXSTLOS.  This file
contains  for each of the 15  major  catchments, the
Basin Moisture  Content (BMC)  and  the  starting  loss
(STLOSP) for the  pervious areas and accumulated loss
(ACLOSS) and  starting loss  (STLOS)  for the imper-
vious areas. These  are given in inches. The last entry of
the file is the accumulated time in seconds since the
DECAY (a recovery factor) has been applied to update
the moisture values.

2. Sense Switch Options
   Accumulator switches  located on  the lower right
side  of the central processor are used as sense switches
by the model. All  sense switches should be off (down)
at the beginning of execution of the model  unless the
operator is familiar with the program.  The use of each
sense switch is described below.

Sense Switch Number
# 0  Not presently used
# 1   Normally if all sense switches are down the only
      printed  output  on  the  line printer  will be  a
      message indicating when the model predicts that
      there will  be excessive flow at  a given point. If
      this switch is up this message will not be printed.
#  2  If this switch  is  up the model  will return  to
      "Program Control" if excessive flow  is predicted
      anywhere in the system. This then allows the
      operator to postulate possible corrective changes
      or to continue calculations as he desires.
#3  If it  is desired  to see  the  rainfall rates at the
      raingages  for each  5-minute period,  the rainfall
      rates over each  catchment,  the loss  rates, basin
      moisture  content,  and  the  rainfall  excess, this
      switch should be up.
#  4  If on, this switch  allows the latest inlet hydro-
      graph  for each  major watershed  to  be  printed.
      These represent  the flow arriving at the structure
      before diversion.
#5  If on, the diverted hydrographs for  each  major
      watershed and the  total  lumped flow entering the
      interceptor sewers  will be printed.
#  6  If on,  a  tabulation  of the volumes of  water
      before diversion, after diversion, and diverted to
      the river will be shown.
#7  If  on,  the hydrograph at  any   point  in  the
      interceptor  system where the  predicted flow  is
      greater than the allowable flow will be  printed.
#  8  If on, all  hydrographs will be printed as the flow
      is routed through the  interceptor system.
#  9  If on, only the inlet hydrographs before diversion
      will be computed.  Model   will  cycle  through
      chains  1, 2, 3, and 6 only  (no diversion  or
      routing).
#10  If on,  only the inlet hydrographs and diverted
      hydrographs will be computed. Model will cycle
      through chains 1, 2, 3, 4, and 6  (no routing).
#11  Not presently used.
#12  If on, the program will scan real  time data only
      (RTIME).
#13  If on, the program will not scan  any real time
      data.
#14  If  on,  the program  will interrupt  its normal
      processing and  return  to  program  control for
      operator  intervention. The  computations  inter-
      rupted will be lost.
#15  If  on,  the program will   process a historical
      rainfall file  specified  by the operator. The data
      will be processed  in one-hour blocks, starting at
      the time given by the  operator.
#16  If on, the instructions  normally printed out on
      the teletype will not appear.
#17  If on, the program will return to program control
      for operator intervention after completing the
      chain of  the model that it  is currently working
      on. No computations will be lost.
      The sense switches may be reset at any time.
Depending  on  the specific function of the switch  it
may have an immediate effect on the program or a
delayed effect.  When controlling  or interrupting the
                                                  157

-------
sequence of processing by switches #9, #10, #14, or
#17,  the  operator  must  be careful to  re-enter the
processing sequence at the proper point.


3. Program Control
   The model  is so constructed that an operator may
intervene  in the normal  processing sequence. This is
accomplished by the routine PROGRAM CONTROL.

   At any time during processing the operator may
turn on  sense  switch  (SS) #14 or #17 to return to
PROGRAM CONTROL.

   PROGRAM CONTROL  is  also entered at  other
 times when information is needed by the program.

   The options within PROGRAM CONTROL  avail-
able to the operator are given below.

#1. The model is reentered  at its origin. This gives the
    operator  a chance to  process a  new historical
    rainfall file or  to process real   rainfall  in file
    RAINRP DLY. Control is transfered to Chain 1.

#2. This allows the operator to restart computing inlet
    hydrographs. When processing real data the pro-
    gram always computes starting 2 hours prior to the
    present real  time  or at  the beginning  of the
    RAINRP  DLY  file,  whichever  is  later.   For a
    historical  file the model starts at the starting time
    given by the operator. This time is entered by the
    operator at the console. If  rain has occurred prior
    to the starting time given by the operator, the first
    two hours of  prediction will not reflect this rain.
    Control is transfered to Chain 3.

#3. This causes the  model to continue the processing
    by  adding on the latest rainfall  readings for real
    time  data or readings for one hour  for historical
    data. Control is transfered to Chain 3.

#4. When reentering at this point the  program takes
    the last determined inlet hydrographs and diverts
    them. Control is transfered to Chain 4.

#5. When reentering at this point the program  takes
    the  last  determined  diverted hydrographs  and
    routes them. Control is transfered to Chain 5.

#6. This allows the program to scan if it is the proper
    time. PROGRAM  CONTROL will always  be re-
    entered following this selection unless sense switch
    #12 is kept on. Control is transfered to Chain 9.

#7. This calls  the gate editor program.  The gate editor
    allows the operator to enter gate  settings to the
    model.

#8. This causes the program to stop and return control
    to the monitor.
 4.  Gate Editor
   The  GATE EDITOR  is  an option  called  from
 PROGRAM CONTROL  that  allows the operator to
 delete, insert or replace gate and Fabridam settings that
 are used by the model. The file that contains the model
 gate settings is GATESETTG.  If desired this file may
 be  changed with  the systems EDITOR  program, but
 this requires that the model be stopped.

   The GATE EDITOR  has 5 operations, explained
 below.
1. The first operation (L) provides a listing of all the
   gate and Fabridam settings.  In addition the maxi-
   mum and  minimum settings  for each structure are
   shown.

2. This operation (D) will delete a setting from the file.
   The settings for day zero and day 365 may not be
   deleted.

3. This operation (I) will insert a new setting.

4. This operation (R) will replace a given setting with a
   new value.

5. This operation (E)  causes the routine to return to
   PROGRAM CONTROL.
 Format:
         NGTHRMNDAY  X.XX  X.XX
where

    N   is operation to be completed (L, D, I, R, E)
   GT   is gate number
   HR   is hour of day (0-23)
   MN  is minute of hour (0-59)
 DAY   is day of year (1-364)
X.XX   is gate opening in feet
X.XX   is fabridam height in feet


   Exhibit 4 shows the teletype  message when GATE
EDITOR is used.


                   Exhibit 4
     Teletype message when Gate Editor is used.
 BATE EDITOR
 TYPE L FOR  L1ST1N6 OF PRESENT SATESETTINGS
 TYPI D KltH 6ATC NUMBER AND TIME-  TO DELETE A SETTING
 TYPE I KITH GATE NUMBER. TIME AND SETTING- TO INSERT SETTING
 TYP.C R KITH DATE NUMBER. TIME AND SETING- TO REPLACE SETTING
 TYPE C TO CXIT TO PROGRAM CONTROL
 N OT HR MN  DAY X.XX X.XX
 L
 N GT KR MN  DAY X.XX X.XX
                                                  158

-------
   Exhibit 5  shows  the  list  of  gate  and Fabridam
 settings.

                      Exhibit 5.
           List of gate and Fabridam settings.
DIVERSION
 NUKfltB
            CAIE «ND FASflDAn SEIIINGS
line
 HUN
CAIE   FABRI£AH
SEIIIHC  SETTING
nix
 GATE
0
1
1
I
I
)
1
4
I
S
s
6
6
7
7
9
9
10
10
11
11
12
\2
11
11
It
14
1%
IS
It
It
17
17
IB
18
19
19

PHAL
P HAL
IROUt
IRQUI
KELOG
KELOG
RICE
II ICE
HARSH
HARSH
EUS1E
EUSIE
OAK
OAK
J8IH
jaTH
H INNH
n INNH
26SEA
26SEA
PUASH
PUASH
42CAM
42CA1
2 HA IN
2NA1N
N UN
Nun
Sun
sun
E AS1N
EASin
EUSTU
EUSTU
42-8P
42-BP
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
165
0
165
0
165
D
165
0
165
0
165
0
165
0
105
0
165
0
165
0
165
0
165
0
165
0
J65
0
165
0
165
0
165
0
165
0
2
2
1
1
6
6
1
1
}
1
1
3
1
1
1
1
1
1
2
I
I
2
4
4
2
2
0
0
0
0
0
0
1
1
00 0.00
50 1.00
50
75
75
00
00
00
00
75
75
00
00
25
25
25
2S
25
25
.25
.67
.00
.25
.67
.00
.50
.67
.00
.50
.67
.00
. 1)
.50
.67
.50
.82
10 2.00
10 5. 11
10 1.12
10 5.67
00 0.00
00 0.00
10 2.16
10 4.67
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 0.00
00 1.50
00 1.50
0.00 0.00
0.00 0.00
0
2
2
1
1
6
6
1
)
1
1








2
2
2
2
4
4
2
2
0
0
0
0
0
0
1
1
0
0
00
50
50
15
f5
00
00
00
00
75
75
00
00
25
25
25
2 5
25
25
10
10
10
10
00
00
10
10
00
00
00
00
00
00
00
00
00
00
0.00
6.25
6.25
7.00
7.00
4 .67
4.67
5.50
5.50
6.00
6.00
5.67
5.67
5 .11
5.31
6.67
6.67
6.e;
6.82
5. 13
•>. 13
5.67
5 .67
0.00
0.00
4 .67
4 .67
0 . 00
0.00
0.00
0.00
0 . 00
0.00
1 .50
1.50
o.co
0.00
  c.  Gather real time data and process historical data
  d.  Process historical data only

  A description of each of these modes follows.
  a.  To  use  the  model to  acquisition data only,  SS
     #12 should  be  turned on. The operator will be
     requested to ENTER REAL TIME. Then control
     will pass to Chain 9 and will remain there as long
     asSS#12  is on.
  b.  To both gather and process real time data SS #12
     and  #13  should  be off. The operator  will  be
     requested to ENTER REAL TIME. Then control
     will pass to Chain  9 to allow data acquisition and
     on to Chain 3 to allow data processing.
  c.  To  gather real  time data and process historical
     data SS #12 and #13 should be off and SS#15
     should be  on. The operator will be requested to
     enter the name  of the  historical rainfall file to be
     processed.

ENTER NAME OF HISTORICAL RAINFALL FILE
                 (FILNAMEXT)

     At  this  point the  operator should  type  the 9
     character file name containing the rain data to  be
     processed followed by a carriage return. If the file
     is not present on the input device (DAT. 10)  an
     error will appear:
 5.  Execution of Model (UROM-9)
    MODEL is an execute file that normally resides on
 the disk. It  is run from  the system monitor by the
 command:

                    SE MODEL )
                              K
    The processing that follows depends on the selec-
 tion  of  the various sense switches. As  noted above
 all  sense switches (accumulator switches are  used as
 sense switches and are located on the lower right side
 of  the central  processor)  should be down (off) at the
 beginning of execution unless the operator is familiar
 with the program. When the sense  switches are down
 the program will print out important instructions at all
 points  where  the  operator  may  interact with the
 program.
    The following  message  will appear to inform the
 operator of  the  possible  choices he has available  for
 output and program branching.

    The MODEL may be used in any of the following
 modes:
    a. Gather real time data only
    b. Gather real time data and process real time data
                                         FILE XXXXXXXXX NOT PRESENT ON DEVICE 8


                                              and the operator will  have the opportunity of
                                              entering another file name.
                                              After allowing a scan to occur in Chain 9 control
                                              will return to PROGRAM CONTROL. At this
                                              time the operator should  exercise option 2. This
                                              will allow  him to then ENTER MODEL START-
                                              ING TIME.  Processing  will then be initiated at
                                              the specified time in the historical rainfall file.
                                           d. To  process historical data only, the procedure
                                              outlined in 3 should be followed except that SS
                                              #13 should be on.

                                           When  acquisitioning data  the scanning reports will
                                         appear on the line printer. Output from the model will
                                         appear depending on the settings of the sense switches.

                                           Exhibit 6 shows the teletype output for a run made
                                         using  an historical rainfall file (RASR70219).  After
                                         processing two hours of data the operator interrupted
                                         the  program,  made a listing of  the gate and Fabridam
                                         settings and returned to the system  monitor. Exhibit 7
                                         shows part of the rainfall data used.
                                                   159

-------
                     Exhibit 6
     Teletype output for sample model run.
MONITOR  \

SE MODEL
                         Exhibit 7
            Partial listing of input rainfall data
                      (RASR70219)
Shown  is the time of reading (day,  hr, min.),  raingage
number.and mnemonics and accumulated rainfall (inches)
SENSE SWITCHESfACCUM. SHS> »l -MM 7 ARE USED IN THIS PROGRAM
 IF ON(UP)  THEY HAVE THE FOLLOWING FUNCTIONS

 •I SURPRESS PRINTOUT OF EXCESSIVE FLOW LOCATIONS
 It PAUSE  IF EXCESSIVE FLOW
 • 3 PRINTOUT RAINFALL LOSS-RATE DATA
 *« PRINTOUT UNDIVERTED HYDROGRAPHS
 •5 PRINTOUT DIVERTED HYDROGRAPHS
 16 PRINTOUT VOLUMES OF HYOROGRAPHS
 n PRINTOUT HYDROORAPHS AT POINTS OF  EXCESSIVE FLOW
 IS PRINTOUT HYOROGRAPHS AT ALL POINTS
 •9 COMPUTE UNDIVERTED HYOROGRAPHS ONLY
 lie COMPUTE UNDIVERTED AND DIVERTED HYDROGRAPHS ONLY
 •12 SCAN  REAL TIME DATA ONLY
 •I] 00 NOT SCAN REAL TINE DATA
 • M RETURN TO PROGRAM CONTROL
 •15 PROCESS HISTORICAL RAINFALL FILE
 •16 SUPPRESS PRINTOUT OF PROGRAM INSTRUCTIONS
 •17 PAUSE AT END OF LATEST COMPUTATIONS

SET SWITCHES TO DESIRED POSITION AND THEN TYPE
CONTROL P   AFTER THE PAUSE
THE SWITCHES MAY SE RESET AT ANY TIME
PAUSE
tPENTER NAME  OF HISTORICAL RAINFALL FILE  (FILMANEXT)

 RASR7B 219

     FILE RASR7B 81  NOT PRESENT ON DEVICE 8

  ENTCfl NAME  OF HISTORICAL RAINFALL FILE  (FILNANCXT)

 RASR702I9

 ENTER REAL TIME
  HR MN DAY

  12 12 281

 PROGRAM CONTROL PREVIOUS CHAIN   I  MODEL TIME  £81 IE IB

 TYPE 1 TO RESTART WITH ALL INITIAL CONDITIONS
 TYPE 2 TO RESTART COMPUTING UNDIVERTED HYDROORAPHS
 TYPE 3 TO CONTINUE COMPUTING UNDIVERTED  HYDR06RAPHS
 TYPE 4 TO CONTINUE COMPUTING DIVERTED HYOROGRAPHS
 TYPE J TO CONTINUE ROUTING
 TYPE « TO CONTINUE SCANNING REAL DATA
 TYPE 7 TO CALL GATE SETTINGS
 TYPE 8 TO STOP
  ENTER MODEL STARTING TIME
  HR MM DAY

  19 86 219
  IOP34I
50
0
11
2J
11
44
56
6
16
27
50
2
13
21
11
641 UH-RC
641 Un-RG
641 un-RG
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
64
l>4
64
64
64
64
64
64
UF1-RG
UM-RG
un-HG
un-RG
UM-RC
Utl-RC
UM-RC
UN-DC
un-RC
UN-RC
UM-RC
UM-RC
un-RC
un-RC
un-RC
UM-RC
un-RC
un-RC
UM-RC
UM-RC
un-RG
un-RU
UM-RC
UM-RC
un-RC
un-RC
un-RC
un-RC
un-RC
un-Rc
un-RC
UM-RC
UM-RC
UM-RC
un-RC
UM-RC
UM-RC
UM-RC
un-RC
un-RC
UM-RC
un-RC
UM-RC
UM-RC
UM-RC
UM-RC
UK-SG
UM-RC
un-RC
un-RC
641 un-RG
641 UM-RC
641 UM-RC
641 UM-RC
-0.
-0.
064
008
-0.010
-0.
-0.
-0.
-0.
0.
-0.
0.
0.
0.
0.
0.
0.
0.
0.
0
0
0
0
0.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
021
004
Oil
008
002
006
021
058
065
099
128
265
291
124
111
1BI
413
418
449
457
476
484
494
492
500
518
520
515
519
521
545
512
544
S49
SIB
540
540
517
542
SIS
546
512
492
0. 529
o
0
0
0
0
0
0
0
0
0
541
S42
5!»
540
511
519
546
S44
541
541
  TYPE NUMBED AGAIN

  9

  TTPI NUMBER A8AIN
  PR06RAM CONTROL PREVIOUS CHAIN

  B

  STOP   777777
                                 2  MODEL TIMC  219 17   «
  MONITOR V49
                                                          160

-------
  Exhibits 8-12 show part of the line printer output for the same run shown in Exhibit 6.

                                               Exhibit 8
                                   Partial output of rainfall loss analysis.
The rainfall file name is shown at the upper right. The nine raingages and their corresponding 5 minutes intensities
for the period ending at  16:11 day  219 are given next. The remaining table shows the rainfall loss and excess on
each of the  15 basins. Unless specified all values are given in inches for the 5  minute period ending at the time
shown. The  definition of the column mnemonics are  given below.
I
RAIN
ISTLOS
PSTLOS
IINLOS
PINLOS
ILOSCOE
PPOTLOS
I LOSS
PLOSS
IEXS
PEXS
BASLOS
BASEXS
BMC
ft A I NCACE
GAGE

READINGS
1
1
I
>
4
5
6
7
8
9
10
1 1
1 !
1 1
11
15

READ INGS
1
1
I
)
4
5
6
7
8
9
10
1 1
1 t
1 1
It
15

«EAO INGS
1
I
2
1
*
5
b
7
a
9
10
— basin index
— Average rainfall over basin
— Accumulated value of initial loss on impervious surfaces.
- Accumulated value of initial loss on pervious surfaces.
- Contribution to initial loss on impervious surfaces.
— Contribution to initial loss on pervious surfaces
— Impervious loss coefficient-dimensionless.
— Potential loss on pervious surfaces.
— Loss on impervious surfaces.
— Loss on pervious surfaces.
- Excess on impervious surfaces.
— Excess on pervious surfaces.
- Average basin loss.
— Average basin excess.
— Accumulated losses on pervious surfaces.
ACCLOSS — Accumulated losses on impervious surfaces.
no.

AND HUE
HA IN
0.0092
0.0196
0.0090
0.0195
0.0217
0.02)5
0.0191
0.0278
0.0299
0.0092
0.01*9
0.0052
0.0222
0.0186
0.0000

AND IIME
RAIN
0.0092
0.0198
0.0090
0.0195
0.0217
0.02)5
0.019B
0.0278
0.0299
0.0092
0.01*9
0.0052
0.0222
0.0156
0.0000

AND llnE
RAIN
0.0087
0.0176
0.0075
O.Olbl
0.0192
0.0235
0.0198
0.0275
0.0294
0.0090
12 1 4 5 b 7
HAC-
0.0628b 0.
ISTLOS PSTLOS
0.0000 0.0792
0.0000 0.1198
0.0000 0.0690
0.0000 0.1495
0.0000 0.1817
0.0000 0.2500
0.0000 0.2498
0.0000 0.2500
0.0000 0.2)99
0.0000 0.1192
0.0000 0.1749
0.0000 0.0552
0.0000 0.2500
0.0000 0.2186
0.0000 0.0100
MAC-
0.06286 0.
ISILOS PSILOS
0.0000 0.0884
0.0000 0.1)91
0.0000 0.0180
0.0000 0.1689
0.0000 0.20)4
0.0000 0.2500
0.0000 0.2500
0.0000 0.2500
0.0000 0.2500
0.0000 0.1284
0.0000 0.1899
0.0000 0.0605
0.0000 0.2500
0.0000 0.2)72
0.0000 0.0700
hAC-
0.06285 0.
ISTLOS PSILOS
0.0000 0.0971
0.0000 0.157)
0.0000 0.0855
0.0000 0.1850
0.0000 0.222b
0.0000 0.2500
0.0000 0.2500
0.0000 0.2500
0.0000 0.2500
0.0000 0.1374
EON-
08512 0
1 1 «LOS
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
o.oouo
0.0000
0.0000
EON-
»IL-
. 16500
PINLOS
0.0092
0.0198
0.0090
0.0195
0.0217
0.0000
0.0198
0.0200
0 .0299
0 . 0 09 2
0.0149
0.0052
0.0000
0.0186
0.0000
V IL-
08572 0.16500
1 1 NLOS
0.0000
0.0000
0.0000
0 . 0 00 0
0.0000
0.0000
o.oouo
0.0000
0 . 0 00 0
0.0000
0.0000
0 .0 000
0.0000
0.0000
0.0000
EON-
PINLOS
0.0092
0.0198
0.0090
0.0195
0.021 '
0.0000
0.0002
0.0000
0.0101
0.0092
O.OH9
0 .005 2
o.oooo
0.0186
0.0000
VII-
0851 1 0. 14000
1 1 NLOS
0 .0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
PINLOS
0.0087
0.0176
0.0075
0.0 161
0.0192
0.0000
0.0000
0.0000
0.0000
0.0090
BTP-
0.00000

-------
                                               Exhibit 9
                              Partial listing of calculated inlet hydro-graphs.
          Points on hydrograph are given at 5 minute intervals from -34 to +50 periods relative to
          point zero which corresponds to 16:06day 219. Read horizontally left to right. Discharge
          in cfs.
UNDIVERTED HVDROCftAPHS
Y U IN 1
INTPH
-34
-22
-10
2
1 4
26
) 8
so
INIIR
-34
-22
-10
2
14
26
3 8
SO
DAVN 1
-14
-22
-10
2
14
26
3 8
SO
KELOI

-22
-10
2
14
26
38
50
HYODOCBAPH NO. 1
8.0-11
a.
8.2-21 8.
(
1
.4 -9
.3 3
9.6 IS
8.0 27
7 7 19
8.
11 .
9.
a.
j
7.4
HTDROCBAPH
19.2-31
1
2(
1.7-21
.0 -9
22.2 )
22.9 IS
21.3 27
18.0
H







H



OROGRAPH
.6-3)
.7-21
.7 -9
.2 3
.7 IS
.6 27
.4 39
.2
DROCUPH
.)-))
.5-21
.6 -9
62.9 3

7.6 IS
7.3 27

.0 19
NO
19.
19.
20.
22 .
22.
21.

NO
4.
4 .
4.
7 .
4.
4 .

NO
7.
7.
7 .
62.
7.
7.
7 .
0-)2
2-20
4 -8
5 4
S 16
0 2B
7 40

2
2-32
7-20
o -a
a 4
8 16
2 28

)
6-12
7-20
8 -8
9 4
7 16
6 28
4 40
4
3-12
5-20
6 -8
9 4
6 16
3 28
0 40
.0-1 1
.2-19
.4-7
1 .4 S
.117
.0 29
T 741


19.2-31
19.7-19
20.0 -7
23.) 5
22. 7 17
21.1 29


-) 1
. -19
- /
5
. 1 7
. 29


7. 3-3 1
7.5-19
7.7-7
54.4 5
7.617
7.) 29
7.041
a
8
a
II
9
6


19
19
20
21
22
21










7
7
7
45
7
7
I

0-30
2-18
4 -6
2 6
2 18
0 30
742


2-10
7-18
0 -6
7 6
6 18
1 10


6-10
7-18
a -6
a 6
7 18
6 10


3-10
5-18
8 -6
8 6
6 la
3 30
042

a
a
a
11
9
a


19
19
20
21
22
21










7
7
a
)7
7
/
7
0-29
2-17
4 -S
1 7
0 19
0 11


2-29
7-17
0 -5
8 7
6 19
0 31


6-29
7-1
8 -
2
7 1
: -


3-29
5-17
0 -S
4 7
6 19
1 31
0 41
PUINTS ARb AT
8.0
8.2
8.5
10.9
8.9
a.o


19.2
19.7
20.0
21.7
22.5
20.9


4.6
4. 7
4.8
4.8
4.7
4 .6


7.1
7.5
a. )
29.1
7.6
7.3
7.0
-2 8
-16
- 4
8
20
12


-28
-16
- 4
a
20
)2


-28
- lo
- 4
B
20
32


-28
-16
- 4
8
20
32
44
8
8
8
10
a
8


19
19
10
23
22
20










7
7
8
21
7
7
7
S MlNUTt
0-27
2-15
5 - 1
7 9
7 21
0 ) 3


2-27
7-15
1 -3
7 9
4 2 1
8 11


6-27
7-15
9 - 1
8 1
7 2 1
6 11


3-27
5-15
9 -1
1 9
6 2 1
1 11
0 45
8
a
8
10
8
8


19
19
20
23
22
20










7
7
INTERVALS
.0-26
.2-14
.6 -2
.6 10
.6 22
. 0 34


.2-26
.7-14
.2 -2
.6 10
.) 22
.7 )4


.6-26
.7-14
.2 -2
.8 10
.7 22
.6 14


.1-26
.5-14
10 .9 -2
14
7
7
7
.6 10
.6 22
.3 )4
.0 46
.0-25
.2
.8
1 .4
. 4
.0


19.2
19.7
20.)
23.5
22.2
20.7


.6
.7
.7
.8
.7
.6


7. 3
7.5
15.)
10.0
7.6
7.)
7.0
-1 1
- 1
I 1
2)
35


-25
-1 3
-1
1 1
2)
)5


-25
-1 )
- 1
11
23
35


-25
-1 )
-1
11
23
15
47


1




19
20
20
21
21
19










7
7
25
7
7
7
6
.2-24
.4-12
.2 0
.112
. 1 24
.7 36


.7-24
.0-1 2
.7 0
.1 12
.5 24
.8 16


.7-24
.7-1 2
.4 0
.712
.6 24
.436


.5-24
.6-12
.1 0
.6 12
.3 24
.0 36
.848
a
8
.2-23
.4-1 I
9.8 1
10
a
7


19
20
21
2)
21
19


.0 1 3
.0 25
.7 37


.7-23
.0-1 1
.1 1
.0 1 3
.4 25
.7 37


4. 7-2 1







7
7
38
J
7
7
6
.7-1 1
.1 1
.7 1 J
.6 25
.437
2 49


.5-23
.6-1 1
.7 1
.6 1 1
.3 25
.037
.8 49
.2
.4
1 .7
.1
.0
.7


19.7
20.0
21 .7
21.0
21.4
H.6


4.7
4.7
t .0
4.7
4.6
4.4


I.S
7.6
5*. 8
r.6
7.3
1.0
4.1
6.8
                                               Exhibit 10
                        Partial listing of hydrographs diverted into the interceptor at
                            major inlets before lumping parameters are applied.
DIVERTED HTDHOCRAPHS
POINT 0 REPRESENTS PRESENT MODEL lint
INTPH HrOROCRAPH NO. 1
-14
-22
-10
2
1 4
26
3 8
50
INIIR
-34
-22
-10
2
1 4
2 6
38
50
DAVN 1
-22
-1 0
2
1 4
26
1 8
50
KELOI
-14
-22
-10
2
1 4
26
18
50
8.0-11
8.2-21
8.4 -9
11.) 3
9.6 IS
8.0 27
7.7 19
7.4
8.0-)!
8.2-20
8.4 -B
11 .5 4
9.5 16
8.0 28
7.7 40
HTDRDGRAPH NO.
19.2-13
19. 7-21
20.0 -9
22.2 J
22.9 IS
21.3 27
19.5 19
18.0
19.2-12
19 .7-20
20.0 -8
22 .» 4
22.8 16
21.2 28
19.5 40

xrDROCRAPH NO.
4.7-21
4.7 -9
B.2 )
4.7 IS
4.6 27
4.4 19
4.2
4 .7-20
4.8 -S
7.9 4
4.7 16
4 .6 28

HIDROCRAPH NO.
7.1-3)
7.5-21
7.6 -9
41.2 1
7.6 15
7.) 27
7.0 )9
6.8
7.3-1;
7 .5-20
7.6 -8
43.3 4
7.6 16
7.) 2»
7.0 4D

ft
e
8
11
9
8

2
19
19
20
23
22
2 I
19

3
4
4
6
4
4

4
7
7
7
37
7
7
7

0-> 1
2-19
4 - 7
4 5
1 1 7
0 29


2-1 1
7-19
0 - 7
1 5
7 1 7
1 29
4 4 1


7-1
B -
7
7 1
6 29


1-1 I
5-19
7 -7
4 5
6 1 7
1 21
0 4 1

16
6.0-10
U
a
11
9
d


19
19
20
21
22
2 1
19









7
7
7
17
7
7
t

2-ia
4 -6
2 6
2 ia
0 10


2-JO
1-18
0 -6
7 6
6 18
1 10
1 42


7-18
6 -6
B 6
' IB
6 30


)- 10
5-H
8 -6
2 6
6 1 8
. 1 10
.0 42

6 219
8.0-
8.2-
8.4
II. 1
9.0
a.o


19.2-
19. 7-
20.0
Jl.t
12. t,
21.0
19.2


4.7-
4.8
5.2
4. 7
4.6


7. )-
DISCHARGE
29
1 7
-5
7
19
11


29
1 7
-5
7
19
3 1
4 1


1 7
-S
7
19
11


29
.5-1 7
.0
I .0
. 6
. 3
.0

-5
7
19
11
4 1

tt . 0
8.2
8.5
10.9
8.9
8.0


19.2
19.7
20.0
21.7
22.5
20.9
19.2


4 . 7
4.8
4. B
4 . 7
4.6


7. 1
7.5
8.3
29. 1
7.6
7. 1
7.0

-28
-1 6
- 4
8
20
12


-28
-1 6
- 4
a
20
1 2
4 4


-16
- 4
8
20
32


-28
-16
- 4
a
20
12
4 4

AT 5 nlNUIE NTEHVALS
1!
8
8
10
8
8


11
19
20
21
It
2 0
11









7
7
B
21
7
7
7

0-! 7
2-15
5 -1
7 9
7 2 1
0 1 1


2-27
7-1 5
1 - I
7 9
4 I I
8 31
1 45


7-15
9 - 1
a 9
7 2 1
6 13


1-27
5-15
9 - J
1 9
6 2 1
1 13
.0 45

8
8
B
10
8
a


19
19
20
2 1
22
19









7
7
10
14
7
7
7

0-26
2-14
6 -2
6 10
6 22
0 14


2-26
7-14
2 -2
b 10
1 22
0 46


7-14
2 -2
a 10
7 22
6 14


1-26
5-14
9 -2
6 10
6 22
1 )4
0 46

8.0-25 8
8
B
10
B
8


19
19
20
23
22
2 0
IB


t
5
4
4
4


7
7
15
10
7
7
7

2-11
a - 1
4 1 1
4 23
0 35


2-2S
7-1)
1 - 1
S 1 1
2 21
947


7-11
7 -1
B 1 1
7 23
6 35


3-25
5-11
) - 1
0 11
6 21
3 35
0 47

8
9
10
8
7


19
20
20
2)
21
18









7
7
25
7
7
7
6

2-24
4-1 2
2 0
1 1 2
1 24
7 16


7-24
0-1 2
7 0
1 1 !
S 24
I 4B

7-24
7-1 1
4 0
7 1 2
6 24
4 36
2 48


5-24
t-12
1 0
6 1 2
1 24
0 16
8 48




1




19
2-2 3
4-1
a
0 1
0 2
7 3


7-2 3
20.0-1 1
21
21
21
18


4






7
7
36
7
7
7
6

1 1
0 1 1
4 25
1 49

7-21
-1 1
1
1 1
25
)7
2 49


5-2 3
6-1 1
1 1
6 1 1
3 25
0 3 7
849

8.2
a. 4
10.7
9.8
a.o
7.7


19.7
20.0
21 .7
21 .0
21 .4
18.0


4.7
8 .0
4 .7
4.6
4 . 4
4 . 2


.5
.6
1 .7
.6
. 3
.0
.8

                                                  162

-------
                                               Exhibit 11
                             Partial listing of flow diverted into interceptor
                                      after lumping factor applied.
DIVERTED H1DROGRAPHS
POINT 0 REPRESENTS PRIISENT  MODEL Tint
INTPH   HyOROCRAPH NU.
                                                DISCHARGE AI 5 M1NUIE INTERVALS
14
22
10
2
14
26
18
11.8-11
12.6-21
11.2 -9
44.9 1
38.2 15
11.8 27
10.5 19
12
1!
45
17
11
10
.8-12
.2 -8
.7 4
. b 16
.8 ii
.5 40
11 .8-1 1
12.6-19
11.2 - 7
45.0 5
36.9 17
31.8 2?
)0.5 41
1 1.8-10
12.6-18
11.1 -6
44 . 4 6
Ib. 1 13
11.! 30
30. 5 42
11.8-29
12.6-1 7
31. 1 -5
41.8 7
15.7 \^
11. t 11
30.5 41
11 .8-28
12.6-16
11.5 -4
41.1 B
15.1 20
11.8 12
10.5 44
31 .0-27
12.6-15
11.8 -3
14.5 21
11.8 13
10.5 45
11 .8-26
32 .6-14
14.1 -2
14 .0 22
11 .« )4
10 .5 46
11.8-25
32.6-1 1
34.9 -1
41.1 11
11 .8 15
10.5 47
126-24
11.2-1 2
16.4 0
40.1 12
10.5
29.5
36
43
32.6-
13.2-
19.0
19.5
11. B
30.5
29.5
•21
-1 1
1
1 1
25
1 7
49
13.2
W .5
18.8
11.8
10.5
29.5
lltTTR   HtOROCRAPH NO.
-J4
-22
-10
£
1 4
26
18
50
0«VN 1
-14
-22
-10
2
14
26
1 8
50
1.110 1
-14
~2 £
-1 0
t
\ 4
26
!8
50
20
20
21
2)
24
22
20
19
.4-31
.9-2 1
.2 -9
.6 3
.2 15
.6 27
.7 19
.0
20
20
21
24
24
22
20

MYDROGRAP
7
7
7
12
7
7
b
.2-11
.4-21
.5 -9
.9 1
.4 15
.227
. 7



1



HTDROCRAPH
11
11
65
11
11
10
10
.1-31
.6 -9
.7 J
.5 15
.1 27
.7 19
.1
11
1 1
11
65
11
1 1
10

4-12
9-20
2 -8
2 4
2 16
5 28
6 40

NO.
2-1J
4-20
5 -8
4 4
4 16
2 28

NO.
1- 32
6 -R
7 4
5 16
1 28
7 40

20
20
21
24
24
22
20

1
;
7
7
10
7
7

4
II
1 1
11
56
II
11
10

4-31
9-1
2 -
7
1 1
4 2
6 4


2-3 1
4-1 '
5 - 1
7 5
4 1 7
2 29


1-3 1
7 - 7
8 5
5 1 7
1 2^
7 4 I

20
20
21
25
24
22
20


Y
7
7
9
7
•


11
1 1
1 1
56
II
11
10

4- 10
fl-ia
I -6
1 6
a is
1 10
5 42


2-10
4-18
5 -b
2 b
> 18
2 10


1- 30
4-18
9 -b
b b
5 18
1 10
7 42

20
20
21
25
2 1
22
20


7
j
j
s
/
/


u
1 1
12
11
II
| ]
10

4-2*
9-17
2 -•>
1 7
1 11
2 31
4 4 1


2-29
4-1/
5 -•>
2 1
4 11
? 11


1-29
1 -5
1 7
5 19
1 11
7 41

20
20
21
25
21
22
20


7
7
7
7
7
7


1 I
12
&4
11
II
10

-28
-1 b
- 4
8
20
12
1 44


2-28
4-16
6-4
6 8
4 20
2 12


I -28
b -4
2 B
5 20
I 12
7 44

20
20
21
25
21
22
20







7


11
11
12
11
11
10

. -27
. -15
. - 1
9
. 2 I
. 3 1
.2 45


.2-27
.4-15
.6 - 1
.b 9
.4 2 1
.2 33


.1-27
.6 - 1
.*. 9
.5 21
.1 11
.7 45

20
20
21
25
2 1
22
20


7
7
8
7
7
7


11
16
22
II
11
10

4-26
9-14
4 -2
0 10
' 22
0 14
1 46


.2-26
4-14
2 -2
6 10
4 22
2 14


. -26
-2
10
. 22
14
46

20
20
21
24
21
21
20










11
23
15
II
11
10

4-25
9-1 3
5 - 1
9 11
6 2 1
9 35
1 47


2-25
4-1 3
9 - 1
6 11
423
2 15


1-25
1 - 1
2 1 1
5 2 1
1 15
7 47

20
21
2i
24
22
21
19


7
7
10
;
7
b


II
1 1
18
11
11
10
10

9-24
2-1 2
9 0
5 12
1 24
0 16
1 48


4-24
5-12
1 0
4 1 2
2 24
9 16


4-2 I
6-12
1 0
5 1 2
1 24
7 16
1 48

20
21
22
24
22
20
19


7
7
11
7
7
6


11
1 1
55
11
11
10
10

9-2 ]
2-1 1
1 1
4 1 1
7 25
9 37
2 49


4-2 3
5-1 1
3 1
4 I 1
225
9 1 7
7 49


4-21
6-1 1
2 1
5 I 1
I 25
7 1 7
349

20.9
21 .2
21.0
24.1
22 .7
20.8
19.1


7.4
7.5
12.7
7.4
7.2
6.9
6 . 7

11 .4
11.6
57. 1
11.5
II .1
10. J
10.1

                                              Exhibit 12
                            Summary of volumes before and after diversion.
INLET INOE1
1
2
1
4
5
b
7
8
9
10
11
12
1)
14
15
TOTAL VOLUME
TOTAL VOLUME
LUMPED INLET VOL
85991 1.
S5151 7.
192652.
456168.
1 5 78 6 9 .
171116.
49(705.
16811 6.
7 19502.
901750.
697681.
11 10914.
f. 1695 1.
1249695.
645588.
CF 10? 1661 4.
HC 76.42
LUMP ED 0 IV V GL
859911.
551517.
192652.
416791.
15(870.
164082.
354566.
720288.
719502.
901750.
695514.
1044109.
816-15 1.
1249695.
645588.
9912878.
74. 15
LUHPf D RIVER VOL
0.
0.
0.
19577.
-0.
7014.
14011 7.
4802 7.
-0.
0.
2167.
66809.
0.
0.
0.
101751.
2.27
INLET VOL
217151.
520126.
1 2191 1.
10024 4.
157868.
211950.
171582.
76811 4.
406322.
111045.
S8I407.
981127.
614256.
844199.
1801] 3.
6632271.
49.61
DIVERTED VOL
217151.
520126.
12193).
274206.
157B68.
227554.
12441 1.
720286.
406322.
111045.
571602.
924002.
614256.
844199.
18013 1.
6441706.
4 8.20
Rl VE ft VOL
0.
0.
0.
26018.
-0.
t 396.
49 172.
4B027.
-0.
0.
1806.
59124.
0.
0.
0.
188561.
1.41
                                                 163

-------
                           Exhibit13
                  6. Loader Memory maps of
             CREATE and MODEL (UROM-9)
MONITOR V48

SA OKA -6.-4.-1.3.19

ICHAIN
CHAIN V2A
>8UILD CREATE
.-CHAIN 1
*CREAT2. 1NPUTD. INPUTH.ROUTJN
>END
 CREAT2 55492
 INPUTO 54051
 INPUTH 52373
 ROUTJN 51460
        45462
        45074
        44541
        44530
        44461
        41432
        41 I 56
        41076
        41063
        40770
        40230
        40134
OKA.
LPA.
FILE
FLOAT
• DA
8CDIO
BINIO
• SS
STOP
SPNSG
FIOPS
OTSCR
INTEGE 37661
REAL   36625
•CB    40114

CHAIN* |
LOWEST 31537
COMSZE 00000
 BONI TOR V48

 IB PR

 MONITOR V4S

 ItJOS 8U1LO  MODEL.

 6UILO MODEL
 >CHAIN I
 >CHAIN1.SENSE.SET.TMADD
 >ENO
  CHAIN! 54545
  SENSE 54514
  SET    S4I67
  TMADD 53723
  OKA.  47725
        47372
        47323
 FILE
 .DA
 SCDIO  44374
 81NIO  44020
        43740
        43725
  • S3
  STOP
  PAUSE   43711
  SPHSO   43616
  FIOPS   430S6
  OTSER   4Z762
  INTEGE  42643
  REAL    41607
  .CB     41567

  CHAIN*  I
  LOWEST  41567
  COMSZE  07385
>CHA[N 2
>CHAIN2.CONTL.G«TEOT.EWNTL.TMAOD,SENSE.RAINF
.END
 CHAIN2 56410
 CONTL  55421
 BATEDT 5341 I
 EVNTL  52246
 TMADO  520Q2
 SENSE  SI 751
 RAINF  471 I I
 OKA.   43113
 LPA.   42525
 FILE   42172
 IABS   42156
 •DA    42107
                                                                     BCD 10
                                                                     BINIO
                                                                     .SS
                                                                     GOTO
                                                                     STOP
                                                                     SPMSG
                                                                     FIOPS
                                                                     OTSER
                                                                     INTEGE
                                                                     REAL
                                                                     .CB
        34751
        41633
        41553
        41523
        41512
        41 41 7
        40657
        40563
        40444
        33715
        40424
                                                                     CHAIN* 2
                                                                     LOHEST 33265
                                                                     COMSZE (7305
»CHA1N 3
>CHAIN3.HrDR.SLOS.SENSE.THADD
>END
 CHA1N3 55681
 HYDR   53435
        52152
        52121
        51655
        45657
        4S271
        44736
        44725
        44672
        44657
        44566
        44450
        44404
        44335
        41306
        41032
        40752
        40737
        40644
        40104
        37704
SLOS
SENSE
TMAOD
OKA.
LPA.
FILE
FLOAT
• BE
EXP
• EE
.EF
• EC
• DA
BCDIO
BINIO
• SS
STOP
SPMSG
FIOPS
OTSER
INTEGE 37565
REAL   36531
.CB    40064

CHAIN* 3
LOWEST 34777
COMSZE 07305
                                                                     >CKAIN 4
                                                                     >CHAIN4.DlV,Tri,Tr2.TY4.TY5.ARE4.BICC.TOP.CUES,SENSE.TRAPOZ,
                                                                     >THADD»FA88TS
                                                                     >ENO
                                                                      CHA1N4 54732
         53232
         52601
         52202
         51543
         50436
         50065
         47703
         47566
         47464
         47433
 DIV
 TYI
 TY2
 TY4
 TY5
 AREA
 B1CC
 TOP
 CRES
 SENSE
 TRAPOZ 47286
 TMADO  46762
 FABGTS 46521
 OKA.    42523
 LPA •    42 I 35
 FILE    41602
        41564
        41510
        41455
        41367
        41354
        41341
        41237
        41150
        41057
        40741
        40675
        40626
        34751
        40352
        40272
        40244
        40231
        40136
        34211
                                                                      ASS
                                                                      • BC
                                                                      • BE
                                                                      SORT
                                                                      SIN
                                                                      ATAN
                                                                      • EB
                                                                      • ED
                                                                      .EE
                                                                      • EF
                                                                      • EC
                                                                      BCOIO
                                                                      BINIO
                                                                      .SS
                                                                      GOTO
                                                                      STOP
                                                                      SPMSG
                                                                      FIOPS
                                                                      OTSER
                                                                      INTEGE 34072
                                                                      REAL   33036
                                                                      •CB    40022
          042
                                                                      CHAIN« 4
                                                                      LOWEST 32046
                                                                      COMSZE 37305
                                                               164

-------
>CHA1N 5
>CHAIN5.ROUTB.COMB,DI WlN.WALY,AREA,SENSE.TYI.TY7.FABOTSrTMAOO.
>CRES.TOP,BICC.TRAPOZ
>ENO
 CHAINS 55715
 ROUTS  54576
 COMB   34414
 OIVIN  53422
 VALY   53035
 AREA   52464
 SENSE  38433
 TYI    52808
 TY7    51636
 FABGTS SI 315
 TMADO  SI 131
 CRES   51027
 TOP    56712
 8ICC   59508
 TRAPOZ 50333
 OKA.   44325
 LPA.   43737
 FILE   434B4
 ABS    43366
 IABS   43353
 HOC    43326
 .BC    432S2
 .BE    43217
 SORT   43131
 SIN    43116
 ATAN   43183
 .EB    43801
 .ED    42712
 .EE    42621
 .EF    42503
 .EC    42437
 .DA    42370
 BCDIO  347SI
 BIN10  42114
 .55    42034
 GOTO   45006
 STOP   41773
 SPMSG  41708
 FIOPS  41140
 OTSER  41844
 INTEGE 40725
 REAL   33715
  •CB     40705

 CHAIN* S
 LOWEST 32364
 COHSZC 07305
 •CHAIN 9
 >SCKDL2«SCALER,RIVI10N,REGCTL.INCPTl)»RAlNRP.ALARHR.SETaAT.
 MNTRPT.SCNFRO.TINR.POINT.CONSII.TINE
 H) GATE/SENSE
 • IMD
 SCNOLP. 56432
 SCALD) SSS7S
 RIVMON S46B7
 REGCTL 53384
 INCPTH 52333
 RAINRP 51225
 ALARM) 50322
 SETBAT 47594
 INTRPT 47411
 SCNFRB 47B63
 TIM)    46727
 POINT  46455
 CONSW  46401
 TIME    4631B
 RGATE  46183
 SENSE  46051
 OKA.    42053
 LPA.    41465
 FILE    41132
 IABS    41116
 .BB     41040
 .DA     40771
 BCDIO  34751
 .SS     40711
 STOP    40676
 SPMSG  40603
 FIOPS  40043
 OTSER  34655
 INTECE 34536
 REAL    33592
 .CB     40023

 CHAIN* II
 LOWEST 33502
 COHSZE 07303
                                                                  165

-------
                  APPENDIX B
 Description of Labels Used in CREATE and UROM-9
   If the same label is  used in different routines to
represent different variables, the name of each routine
is shown with the appropriate description. An asterisk
(*)  indicates that the label may indicate a type other
than that implied by the  label name.
                                                             (TYPE 7)
Label
A
ACCUM
AG
AJJ
AJL
ALOSC
ALPH
ALPHA
AN
ANS
AO
AR
AREA
AVG
AVG1
A5
A6
BA
BB
BC
BD
BFFAC
BMC

BMSTL


CHECK
COUNT

CZ
C1
 Routine(s)
          (TYPE 4)
          (HYDR)
             Description
          Not used
          See AO
          See ARE A
 	    Not  used, accumulated  time
          since  loss rate recovery  para-
          meter has been applied.
 	    Temporary value
 	    Index in PAL routing
 	    Index in PAL routing
 	    Loss rate constants
 	    See NAMES
 	    See NAMES
 	    Manning's "n"
 	    SeeNSTRL
 	    Area of orifice opening (sq ft)
(CHAINS) See AREA
 	    Combined sewer area actually
          contributing to given  inlet (sq
          ft)
 	Average initial rainfall reading
 	    Average rainfall reading
 	    Area of orifice opening (sq ft)
 	    Area of orifice opening (sq ft)
 	    Temporary value
 	    Temporary value
 	    Temporary value
 	    Temporary value
 	    Base flow factors—hourly fac-
          tors to determine hourly base
          flow from mean daily
 	    Basin moisture content (total
          loss on pervious surfaces)
 	    File  name of  antecedent con-
          dition parameters
          (BMCXSTLOS)
          'Flag
          Number of valid rainfall
          readings
          Thiessen polygon factors
(RAINF)
                                          C10

                                          C11
                                          C12
                                          C13

                                          C14
C15
C2




C3



C4


	
(TYPE 1)
(TYPE 2)
(TYPE 4)
(TYPE 5)
(TYPE 7)
(TYPE 1)
(TYPE 2)
(TYPE 4)
(TYPE 5)
(TYPE 1)
(TYPE 2)
(TYPE 5)
(TYPE 1,
 TYPE 4)
(TYPE 2)
(TYPE 5)
                    CQ of Fabridams
                    CQ side channel weir
                    CQ weir A
                                          C5
                                          C6
                                          C7
                                          C8
                                          C9
                                          D
DDT
DELT
DF
DFA
OFF

DIFF


DIVC

DIVCC

DIVE
DR
DT

DTT

E
                    CQ orifice
                    Length  of  lower  side  weir
                    (Type 5)
                    CQ orifice (Type 5)
                    Elev. lower weir (Type 5)
                    Drop  from orifice to intercep-
                    tor (Type 5)
                    Length of upper side weir
                    (Type 5)
                    Width of orifice (Type 5)
                    1/2(32.2)  (CD
                    Length side channel weir
                    Fab. expon.
                    Length of Weir A
                    Width orif.
                    Fab. expon.
                    CQ gates
                    Discharge ratio
                    CQ orifice 0
                    Width orif.
                    Elev. weir
                    CQ Fab.
                    Elev. orifice invert (Type 1)
                    Expon. Fab (Type 5)
                    Area orifice D (Type 5)
                    Elev. upper weir (Type 5)
                    CQ side weir (Type 5)
(AREA,
 CRES)
(GATEDT)
(TOP)
(CONTL)
           Nominal opening of orifice
           Hollerith "D" Delete
           Nominal fabridam diameter
           See DTT
  	    Difference in time (hrs)
  	    Fabridam diameters (ft)
  	    Fabridam diameter (ft)
  	    Temporary value  fabridam di-
           ameter (ft)
  	    Difference between current and
           last reading; average rainfall in-
           crease in time period
  	    Diversion constants (C1  through
           C5)
  	    Diversion constants (C6
           through C15)
  	    Common block name
(EVNTL)   Difference of values
  	    Time  interval  of points on
           hydrograph (sec)
  	    Time interval of equally spaced
           rain data (sec)
(GATEDT)  Hollerith "E" Exit
                                                167

-------
F
FACT
GAGE     (RAINF)
GATES     	
          (TYPE 1)
H

HA
HF
HFA
HFF
HFFF
HFL
HFMAX
HFMAXL
HG
HGG
HGGG
HGL
HGMAX
HGMAXL
HS
H1
H5
H6
I
IA

IADDA
IADDR

IB
1C
ICN

ICONTA
ICONTB    	

ICONTC    	

ICONTD    	
ID         	
IDA        	
IDAY      	
IDAYH     	
IDD        	
          (DIVIN)
          (EVNTL)
Residual in bisection method
Lumping factors to account for
areas that are assumed to enter
interceptor at one of 15 major
inlet points
Gage name
File name for gate and Fabri-
dam setting file (GATESETTG)
Fabridam height (ft)
Gate opening (ft)
Gate opening (ft)
Fabridam height (ft)
Fabridam height (ft)
Fabridam height (ft)
Fabridam height (ft)
Fabridam height (ft)
Max Fabridam height (ft)
Max Fabridam height (ft)
Gate opening (ft)
Gate opening (ft)
Gate opening (ft)
Gate opening (ft)
Max gate opening (ft)
Max gate opening (ft)
Temporary value
Temporary value
Gate opening (ft)
Gate opening (ft)
Index
Index of diversion type
Temp index
Rain gage address
Valid   rain  gage addresses  in
order used by model
Index of meter
No longer used
Number of additional lag peri-
ods if NSTRL is  even
Flag indicating (0—don't print
instructions  to  operator,  1 —
print)
Index  of current operation in
operation table (JR)
Total number of steps in opera-
tion table (JR)
Index giving last chain executed
Time-days
Time-days
Real time day
Diversion model  time (days)
Time-days to  begin  forming
even intervals
IDT        	    Time interval (sec)
IDTM      	    Time interval minutes
IDTT    (RAINF)   See IDT
IE         	    Index of operation to be done
                    (1—routing, 2—combining, 3—
                    diversion, 4—begin on new line)
IGT        	    Gate number
IH         	    Time-hours
         (CHAINS)  Index  of  hydrograph  to  be
                    operated on
IHA        	    Time-hours
         (CHAIN 5)  Index  of  hydrograph  to  be
                    combined
IHH        	    Index of where to store  results
                    of operation
         (ENVTL)   Time-hours to begin forming
                    even intervals
IHR        	    Realtime-hour
IHRH      	    Diversion model time (hrs)
MB         	    Sense switch flags (0 through
                    17)
IJ          	    Flag for excessive flow
IK       (ROUTB)   Index
IM         	    Time minutes
IMA        	    Time minutes
IMIN       	    Real time minute
IMINH     	    Diversion model time (minutes)
IMM        	    Time-minutes to begin forming
                    even intervals
IN         	    Index
IN LAG     	    Index of hydrograph points to
                    be logged
IO         	    INDEX
IR       (HYDR)    See NNN (RAINF)
IRI         	    Index
IS         	    Time seconds
IS       (HYDR)    Index
ISEC       	    Real time seconds
ISS         	    Time-sec to begin forming even
                    intervals
1ST AT     	    File status flag
IT         	    Index
ITB        	    Number of time intervals for
                    base length of unit hydrographs
ITEST   (CHAIN 1)  File status flag
ITIME   (RAINF)   Temp,  storage oftimes of rain-
                    fall readings
(TIMES    	    Used by RTIME (to keep com-
                    mon size equal)
ITM      (GREAT 2) Largest  unit hydrograph base
                    time (DT  time intervals),  no
                    longer used
                                               168

-------
ITP        	    Number of time intervals for
                    time to peak of unit hydro-
                    graphs
ITQI     (GREAT 2) Not used
ITT        	    Index
IX         	    Index
IXD        	    Time days
IXH        	    Time hours
IXM        	    Time minutes
IXS        	    Time seconds
IXXX    (RAINF)    Dummy label
IY         	    Flag  (=),  no excessive flow;
                    @ 1, return to program control
                    at completion of routing)
IZ       (CONTL)   Parameter   indicating  where
                    control should be transfered in
                    Program Control
           	    Index
J          	    Index
         (GATEDT)  Flag indicating (0 file closed, 1
                    file open)
JADDR     	    Valid raingage addresses in or-
                    der scanned by RTIME
JDAY      	    Time days
JHR       	    Time hours
JIDA      	    Time days
JIHA      	    Time hours
JIMA      	    Time minutes
JISA       	    Time seconds
JJ         	    Index
JL         	    Index
JMN       	    Time minutes
JR         	    Operation table  for controlling
                    routing, combining and in line
                    division
JSEC      	    Time seconds
JTD       	    Time days
JTH       	    Time hours
JTM       	    Time minutes
JTS        	    Time seconds
JX        	    Index
K         	    Index  indicating  number  of
                    times file has been  searched in
                    GATEDT
KA        	    Flag
KD        	    Time days
KDAY    (TMADD)  Days before adding
KF        	    Index
KH        	    Time-hours
KHR      (TMADD)  Hours before adding
Kl         	    Index
KK

KKD
KKM

KKMM
KKS
KM
KMIN
KS

KSEC
KTD
KTH
KTHH
KTM


KTS
L
LAG
LD
LDAY
LF
LGT
LH
LHR
LIST

LIST A

LISTB

LISTC

LISTD

LISTX

LISTY

LL
(TMADD)

CHAIN 3)
(TMADD)
(GREAT 2)

(TYPE 1)
(TMADD)
(TMADD)
 (GREAT 2)
 (CHAIN 1)
 (GREAT 2)
 (CHAIN 1)
Number of  points needed  on
hydrograph for diversion
Time days
Number of time intervals base
flow factor should be applied
Number of time intervals base
flow factor should be applied
first time
Time seconds
Time minutes
Minutes before adding
Time-seconds
Index
Seconds before adding
Time days
Hour of day
Time hours
Time  minutes;  time into  the
past that base flow must be
established
Time seconds
Not used
Index
*Fabridam with (ft)
Number of 5 minute periods
flow should be lagged (Lag <0,
lag  time = |LAG| x 5 minutes;
Lag >0, lag time = (LAG +
k)  x  NRCHS  x  5  minutes)
where k = 0 if NSTRL is odd,
k = 1/2 if NSTRL is even
Time days
Days before and after adding
*Fabridam width (ft)
Gate Number
Time hours
Hours before and after adding
'File name  of  constants  for
Chain 1 (STARTLIST)
'File  name  of  constants  for
Chain3(HYDROLIST)
'File name  of  constants  for
Chain 4 (DIVERLIST)
'File name  of  constants  for
ChainB (ROUTELIST)
'File name of  valid  raingage
addresses (RAINGAGES)
'Not used

*Not used

Index
                                               169

-------
LLL        	    Dummy
LLOCI      	    Index of names (NAMES) for
                     inlet hydrographs
LM         	    Time minutes
LMIN     (TMADD)   Minutes before adding, minutes
                     after adding
LQ         	    Index of hydrograph related to
                     given diversion structure
LS         	    Time seconds
LSEC     (TMADD)   Seconds before adding, seconds
                     after adding
LSX      (CHAIN 1)  Time in  seconds which  model
                     back  dates  when  starting to
                     compute real time hydrographs
          (CHAIN 4)  Time in  seconds which  model
                     back  dates  when  starting to
                     compute diversions
M          	    Index
MIDAY     	    Model day
MIHR       	    Model hour
MIMIN      	    Model min
MISEC      	    Model sec
MQ         	    Index of maximum discharge at
                     diversion structure
MX         	    Dimension limit
N        (CONTL)   Dummy parameter
          (EVNTL)   Number  of  points unequally
                     spaced being past  to EUNTL
          (TRAPOZ)  Number  of  points on  hydro-
                     graph
NA        	    Index of orifice opening type
                     (rectangular, crescent, segment)
          (EVNTL)  Temp index
NAM      (GATEDT)  *Name in gate setting file
NAMES    	    *Names of various  points in
                    system
NC        	   Number  of  catchments  in
                    system
NCC       	    Index of catchments
ND        	   Numberof diversions in system
ND        	   Time days
NDTT      	   Number  of  points on hydro-
                    graph
NDTTB     	    Index  of  hydrograph   base
                     length
NDTTP     	    Index of hydrograph peak
NDTTP1    	    Index of hydrograph peak + 1
NG        	    Number of raingages in system
NH        	   Time hours
NM        	   Time minutes
NN       (EVNTL)   Number points generated so far
          (RAINF)   SeeNNN
NNG       	   Index of raingages
NNGG      	   Raingage index
NNN      (EVNTL)   Max number  of  points to  be
                    generated  by  EUNTL;  actual
                    number of pts generated
          (RAINF)   Number of rainfall intensities
NP         	   Number of points that can be
                    used in current interval
NPA       	   Number of points that can be
                    used in current  interval based
                    on  one  gate record   for  a
                    structure
NPB      	     Number of points that can be
                    used in current  interval based
                    on  2nd  gate  record   for  a
                    structure
NPP      (EVNTL)   Accumulated value NP
NPT       	   Index of last  point on hydro-
                    graph for which diversion is re-
                    quired
NP1        	   Index of first point on hydro-
                    graph for  which  a given gate
                    and Fabridam setting holds
NP2       	   Index of last  point on hydro-
                    graph for  which  a given gate
                    and Fabridam  setting holds
NQ        	   Index giving hydrograph num-
                    ber corresponding to diversion
                    number
NQI       (GREAT 2) No longer used
NQQ       	   Index   of   max   discharge
                    (QMAX) for Type 6 diversion
                    structures
NRCHS    	   Number of subreaches  section
                    should be broken into
NSTRL     	   Number of pts on hydrograph
                    to be averaged
NT       (EVNTL)   Time interval  of evenly spaced
                    points (sec)
NTER      	   Flag  indicating  (1-model  is
                    starting over  start from base
                    flow, GT. 1—add runoff to pre-
                    vious hydrographs)
NUMB     (RAINF)   Number of readings for each
                    rainfall gage
NX       (RAINF)   SeeNNN
N1         	   Index
p          	   Wetted perimeter
Q        (GREAT 2)  Not used
           	   Discharge before diversion (cfs)
         (ROUTE)   Temporary storage of discharge
                    hydrograph
QM        	   Max discharge at diversion
                                                170

-------
QMAX
QMIN

QN
QO

QP

QPB
QPP

QPPDP
QPPDX
QPPDZ
QS
QXXXDP
QZ
Q2

Q2S

03
Q5
R
(GREAT 2)
(GREAT 2)
(CHAIN 3)
(GATEDT)
(EVNTL)
(HYDR)
(VALY)
RA

RAIN

RAINDP
RAINDT
 RDG
 REX
 RF
(RAINF)
Maximum discharge at various
points  in  system,  additional
base flow at beginning of line
from suburbs. Upper  limit on
discharge for bisection method
Lower  limit on discharge for
bisection method
Last estimate of Q2
Discharge for which  depth is
desired (cfs)
Peak discharge of hydrograph
(cfs)
Mean daily base flow (cfs)
Hydrographs at  all  points for
all times
File name  of hydrographs at
inlets   before   diversion
(QPPXXDUMP)  saved for fu-
ture additions
File name  of hydrographs at
inlets    after     diversion
(QPPDIVERT)   passed   to
routing
File name  of hydrographs at
inlets   before  diversion
(QPPZZVERT)   passed   to
diversion
No longer used
No longer used
Seed
Discharge to  interceptor after
diversion
Discharge to  interceptor after
diversion 90" line
Portion of flow from 60" line
SeeQ2
Not used
Hollerith "R"
Values at uneven intervals
SeeRF
Hydraulic radius
Average rainfall  intensity  over
watershed
File name of rainfall file to be
processed
No longer used
File name of rainfall intensities
at  even intervals for HYDR
(IRAINDATA)
Raingage reading
Excess intensity (in/hr)
Rainfall intensity for  (DTT)
time period
                                          RO
                                          RR
                                          S
                                          SAVE

                                          SDATA

                                          SIMPSN
                                          SLOPE
                                          SO
                                          STLOS


                                          SUM
T
TB

TEMP
TEMPX
TF

TH
TMM
TP
          (HYDR)


          (HYDR)

          (RAINF)
(RAINF)


(CHAIN 4)
(GATEDT)


(GATEDT)
(GREAT 2)
TTB
TTP
TTT
TTTT
V
VOL A
VOLB
VOLC
X
XDECAY
XI
XINTL
XKD
XKH
XKKD
XKKS
XKM
XKS
XL

(GREAT 2)
(GREAT 2)
(CHAIN 1)
	
	
	
	
	
	
(HYDR)
(GATEDT)
	
	
	
	
(EVNTL)
	
	
(GATEDT)
(EVNTL)
                                                    XLS
See RF
Values at even intervals
Slope of interceptor at meters
Fraction of peak discharge on
unit hydrograph
Temp, storage of raingage read-
ings
File name of input data file to
GREAT
Temporary value
Rainfall rate (intensity, in/hr)
Bed slope at meters
Starting value of initial  loss
parameter  for  impervious sur-
faces
Total  of  valid  initial  rainfall
readings; total  of valid rainfall
readings
Total volume to river (cf)
Hollerith - operation
Base time  of unit hydrographs
(sees)
Temp file  name (TEMPOFILE)
Not used
Interpolation factor between
two values
Temporary value
No longer used
Time  to peak  of  unit hydro-
graphs (sec)
Temp, variable ITB
Temp, variable ITP
No longer used
Common block name
Flow velocity (fps)
Volume before diversion (cf)
Volume after diversion (cf)
Volume to river (cf)
Temporary value
Loss rate recovery parameter
Hollerith "I" Insert
Spacing of points
Time days
Time hours
Time days
Time seconds
Time minutes
Time seconds
Hollerith "L" listing
Index
Time seconds
                                                171

-------
XNP       	    Number of points that can be
                    used in current interval
XNT       	    Time interval - seconds
XRAIN     	    File name for real time rainfall
                    data file (RAINRPDLY)
XSUM      	    Total volume to river
XXKS    (EVNTL)   Time seconds
XXXI      	    Raingage name
XXX2      	    Raingage name
XI         	    C1 for one structre
X1Q       	    C1-
X11        	    C11
X12       	    C12
X2         	    C2 for one structure
X3         	    C3 for one structure
X4         	    C4 for one structure
X5         	    C5 for one structure
X6         	    C6
X7         	    C7
X8         	    C8
X9         	    C9
Y        (DIV)      Depth (ft)
           	    Gate opening (ft)
YMAX      	    Max   depth   for  bisection
                    method
YMIN      	    Min depth for bisection method
YY         	    Gate opening (ft)
Y1         	    See Y (DIV)
ZZ         	    Volumes of hydrographs before
                    and after diversion and vol. to
                    river (cf)
                                              172

-------
                                APPENDIX C
                              Listing of the Model
      The following is a complete listing of the model program. It is recognized that the
      programs in  the form listed are not always accomplishing maximum efficiency. The
      programs have not undergone refinement but are logically correct.

      PROGRAM  TO GREAT MAG  TAPE OF fiATA  FILES   27  SEP f,B
     REftL LISJ<2),LISTA(2),LIST3<2),LISTC<2).LISTD(2>
     REAL LISTX(2),LISTY<2>
     DIMENSION  RAINDP<2),DF(19),NQ( 19>,NQO( 19 > ,01 VCC4 10). 11 81 18) .
    11TP<15>.ITB(15>
     DIMENSION  QS<15),A(17),NA< 17), QPPDPt 2) .QXXXDP<2* ,L,_OCI (15)
     COMMON/BLOCKS/J«(* 8, 7) ,QHAX( 48 ) , D< 3 ) ,S <  3 ) ,QP B( 15 , 1) . &Ni, ALPH A < 4 8 ),
    1 ICONTC.LLOC1,BFFAC(24)
          COMMON/BLOCKL/L,ITQI,MQI,a<100,4)
     COMHON/BLOCKO/DF.NQ.NA.NQQ,A,DIVCC
    1,DIVC<17.5)
     COMMON/BLOCKH/  C2( 15 ,9 ).TP(15) ,TB<15),ALOSC( 15.15),A.
    1 FACT(15).IADDRC9). JADDR(9)
     DATA LISTA(1),LISTA(2)/5HHYDR0.4HLIST/
     DATA L1STB(  I),L ISTB(2)/5HDIWER . 4HL1ST/
     DATA LISTC(1).LISTC(2)/5HROUTE .4HLIST/
     DATA LISTD(1),LISTD(2)/5HRAING .4HAGES/
     DATA L1ST( 1),LIST(2)/5HSTART,4 ML IST/
     DATA QXXXDPC 1) ,QXX XDP ( 2 ) /5HQ XX XX. 4HDUMP/
     DATA QPPOP(  1),QPPDP(2)/5HQPP-XX .4HOUMP/
     DATA RAINDP(1).«AINDP<2)/5HRROIU,4HDUMP/
     DO  b 1=1,18
   6 IIB(I)=0
     CALL IMPUTHCNG.NC,DT.DTT)
     CALL (MPUTD(ND)
     CALL ROUTJN(TMM)
          DO  107  NCC=1,NC
          TFP-=TP(NCC)/DT
          IFP  (NCC)=TTP
     1F( (TTP-FLOAT( ITP(NCC) I ) .LT.O. S»GO  TO 104
103       ITP(NCC)=ITP(NCC)*1
104       TTB-(TB(NCC)-FLOAT(IT?(NCC))»DT)/DT
          1TB(NCC)=TTB
     IF(tTTB-FLOAT(iTB(NCC))).LT.O.S)GO  TO 106
105       ITB(NCC)= ITB(NCC)*!
106       ITB(NCC)=ITB(NCC)*ITP(NCCI
 107 CONTINUE
     URITE<5. 108) ITP.ITB
 108 FORMAT«5H ITP 15I5/5H  ITS  15151
           ITM-0
          DO  2 I=1,NC
           IF< IT8( I)-ITM)2,4,4
   4  ITM=IT8(I)
   2 CONTINUE
     CALL CLOSE  ( 3)
     CALL CLOSE  (8)
     CALL DLETEC8.L 1ST, I)
          CALL ENTER(8,LIST)
       WRITE(8)DTT,DT,IIB.LLOCI,ALPHA.L1ST A.LI STB.LISfC,LISTD.NG.NC
     CALL CLOSEJ8)
     CALL DLETE(8,LISTA,I)
          CALL ENTER<8,LISTA)
          WRITE(8)ITP,ALOSC,AREA,CZ,QP8,ITB,QPPDP.BFfAC
     CALL CLOSEC8)
                                   173

-------
   CALL DLETE<8.L IST8, I)
        CALL ENTER<8,LISTS)
   WRITE(8)ND,NC,DF,NO,NA,NQQ,DIV£.DIVCC, A.QMAX.FACT
   CALL CLOSE(8)
   CALL OLETE(8,LISTC.1)
        CALL ENTER (8.LISTC)
   WRITE <8)JR,OMAX,D,S, AN, I CONTC.OF,0 IVC,D I VCC
   CALL CLOSE(8)
   CALL DLE-TE(8.LISTO.I)
   CALL ENTER(8.L ISTO)
   URITE(S) IAOOR, JADO R
   CALL CLOSE(8)
        STOP
   END

   SUBROUTINE  INPUTDtND)
   DIMENSION C1(17).C2( 17J.C3U7) ,C4< 17).C5<17)
   CONMON/BLOCKQ/ OF( 19),NQ(19),NA<17),NQQ(19).A(I?>,01VCC(10)
  1 ,DIVC(17,5)
   REAO(3.8)A
   ND=19
 8 FORNAKA5)
   REAOO.DNQ
   REAOO.DNQQ
   REAOO.DNA
 1 FORHATC14I5)
   WRITE(5,2)
 2 FORMAK31H HYOROGRAPH NO AND DIVERSION NO)
   WRITE(5.3) (NQ( I),I .1 = 1,19)
 3 FORHATt    215,5X)
   URITE(5.4)
 4 FORMAT(/25H  INDEX  OF  DIVERSION POINT)
   WRITE(5,3) (NQQ(I), I, 1=1,19)
   URITE(5.5)
 5 FORMAT(/13H  ORIFICE  TYPE)
   WRITE(5.3)(NA( I),I . 1 = 1,17)
   READ(3,6)C1
   REAO(3.6)C2
   READ(3.6)C3
   REAO(3,6)C4
   REAO(3.6)C5
 6 FORMAT(7F10.5>
   REAO(3.6)C6.C7,C8,C9,C10,Cll,C12,C13.C14,C15
   WRITE(5.7)
 7 FORHAT(/9X ,3H C1.7X.3H  C2.7X.3H  C1.7X.3H  C4,7X,3H C5,7X,3H  C6
  U7X.3H C7.7X.3H  C8.7X.3H C9,7X,4H  C10.6X.4H C11.6X4H C12)
   00 25  1-1,17
25 URITE(5.24)I,C1(I).C2(I),C3(I) ,C4(I) ,C5( I)
24 FORMATC  15.5F10.5)
 9 FORHATJ  I5.5F10.5)
   WRITE(5.10)C6.C7.C8.C9.C10.C11.C12
10 FORMAT(55X,7F10.5)
   WRITE(5.11)C13 .C14.C15
11 FORMAT(4H C 1 3 , F 10 . 5 , 5 X . 4H C 1.4 , Fl 0. 5, 4H  C15.F10.5)
   REAO(3,6)DF
   WRITE(5, 12 )
12 FURMATI/19H  FA8RIOAM DIAMETERS)
   WRIT£< 5.13)( I,DF(I ) , 1 = 1,17)
13 FORMATl  IS.F10.S)
   DO  99  1=1, 17
   DIVC< I.D-CK I >

                               174

-------
      DIVC(I.2)-C2( I J
      DIVCCI.3)-C3( 1)
      DIVC(I,4)=C4( I J
   99 DIVC(I.5-) = C5( I >
      DIVCCt I) = C6
      DIVCC12) =C 7
      DIVCU 3)=C8
      DIVCC(4)= C9
      DIVCC(5)-C10
      DIVCC(6) -Cll
      OIVCC( 7) =C 12
      DIVCC( 8) =C 13
      DIVCC(9)=CU
      OIVCC(10)=C15
      RETURN

      END

      SUBROUTINE  INPUTH< NG,NC,DT,0TT )
      DIMENSION  SDATA<2)
      COMMON/BLCCKH/ CZ< 15 , 9 ),TP(15) , T8<15).ALOSC( 15.15),A(15) .F4CJ( 15)
      1 IADDRC9),JADDR<9)
C     READ QUANTITIES UHICH ARE  CONSTANT  FOR WHOtEl PROBLEM
C     NG-NO OF  RAIN GAUGES.NC-NO OF  CACHMENTS,DT-T I ME  INTERVAL  FGR
C     RAIN GAUGE  READINGS,OTT-TI ME  INCREMENT  FOR  HYDROGRAPHS
   12 URITE(b,10)
   10 FORMATOOH  ENTER NAME OF  INPUT  DATA FILE)
      READ(b,ll)SDATA
   11 FORMAT
C     LOSS COEFFS.XINLOSS-INITIAL  LOSS,FC-FINAL LOSS.RATE.FO  INITIAL
C     LOSS RATE.XK-EXPONENT.A A-A NTECEDANT INDEX
      READ13, 105)((ALOSC(I.J),J-1, 15 >. 1-1,NC)
  105 FORMAT(  15F5.2)
      WRITE(5.113)
  113 FORMAH//20H LOSS RATE  VARIABLES/
                                    175

-------
     193H NO.  TSTLOS  RECOV AX DECAY  PERV  PERVCF ENOK T8MC BMC  STLOS TSTL
     20SP STLCJSP  COEF POWER ACCLOSS    /)
      WRITE(5.106)(I.ULOSC(l.J>,J=l.l5).I=l,NO
  106 FORMATU5, 15F8.2)
      READ(3.3) (A( I),FACT< I ) , 1 = 1, NO
      WRITE( 5,107)
  107 FORMATU3H  AREAS(SQ FT ) , 5 X , 7HF ACTORS , 3 X , 8H CACH.NO)
      WRITE<5.108)(AU).FACT<[),I.I=1.NC)
  108 FORMAT(1X,F10.0.6X,F10.2, 16)
C  VALID RA1NGAGES  ADDRESSES  IM  ORDER  USED  BY  MODEL AND RTIME
      READ  ( 3,115) ( IADOR( I ),JADDR( I) ,.I = l.NG)
      WRITE (5,116)
  116 FORMAT(//54H  RAINGAGE ADDRESSES  IN  ORDER  USEO 3Y MDDEL AND SCANNED
     1  //)
      WRITE(5,114)(1,1ADDR(I),JAODR( I), 1=1,NG)
  114 FORMAT(2X,15,217)
  115 FORMAT(215)
C     RAIN  GAUGE  PROPORTIONS  FOR EACH  CACHMENT
      DO 109  1=1,NC
  109 READ(3,1)(C2(I,J),J=1,NG)
      WRITEC 5,110)(J,J=1 ,NG)

  110 FORMAT
      RETURN
      END

      SUBROUTINE  ROUTJN(TMM)
      COMMON/8 LOCKS/J«(48,7),QIAX «.8),D(3),S(3),QP8t15,1) . AN .ALPH( 48),
     11CONTC,LLOCI(15>.BFFAC(24)
      READC3.1)  ICONIC
      WRITE(5,5)
      DO 2  I=1.ICONTC
      READ ( 3,1 )( JR( I,J), J-l, 7),QMAX( I),A|_PH( I)
    2 WRITE(5.6)I.(JR(I,J),J=1,7).QMAX(I),ALPH(I)
    1 FORMAT(7I5,5X.F10. 2,A5)
    3 FORMAT(7F10.0)
      READ(3,15)(D(I),S(I),AN.I=1,3)
   15 FORMATC3F10.1)
      READ(3.3»QPB
      READ(3,3)BFFAC
    6 FORHAT(8I8,F10.1. A5)
    5 FORHATtSOH  OPERATION TABLE.HYDROCRAPH  NOS AND ROUTING  CONSTS
     1/56H  OPER     H1O  ENTER HYD  COMB HYD  LEAVE    NSTRL       LAG
     24X.18H     NRCHS  QMAX      >
      WRITE(5,9)(D(I).S(I),AN,I=1,3)
    8 FORMAT(15,F10.1)
    9 FORMAT*//(10X.7H DIAM =F10.2,9H  SLOPE  =F10.5,
     113H MANNINGS  N  -F9.3))
      WRITE(5,10)
   10 FORMAT(//10H  BASE  FLOW)
      WRITE(5.8)(I,QP8CI.I),!=!, 15)
      WRITE (5. 12 )( I, 8FFAC( I). I = 1,2 4)
                                       176

-------
  12 FORMAT(//5X,24HHOURLY  BASE FLOW FACTORS/ /< 5X , I*. F6. 2 »
     READ(3.20)LLOC I
  20 FORHATU5I3)
     URITE(5.11>( I.LLOC Id >, 1 = 1,15)
  11 FORHATI//11H  NAME  INDEX/( 1 5, 15 I >
     RETURN
     END

     CHAIN  11
     REAL  LIST! 2) .L ISTA.LISTB.L ISTC.L1STD.L ISTX(2),LlSTY(2) .NAMES
     DIMENSION QPP( 15, 1 20 > , I1B< 18 ) , LLOCI ( 15 ) .NAMES (48) ,LISTA( 2) .
    lRAIN(2>.XRAIN(2),LISTB<2).ulSTC(2),lISTO(2)
     DIMENSION OF( 19),NQ( 19) ,
    1NA(17).NQQ(19) ,OIVC(17,5),DH/CC(10),A(17).SAINOP(2),0(3),S(3)
     COMMON DTT,NDTT,IDT,NTER,DT, II 8.LLOC I,
    1 NAME S, L I ST A, L I STB . L I STC , L I S T 0 , KK , I CON T A , I CON TB ,
    2ICONTC.ICONTD, 1C, I Y, MI HR , M IM IN .M IDAY ,M I SEC , R AI N. NG. NC . QPP ,
    3ITIMES(12)
     DATA  LtST( 1) ,LIST( 2 ) /5HSTA «T .4 ML IST/
     DATA  XRAINl 1), XRA I N ( 2 > / 5HR A I NR .4HPDL Y/
     DATA  RAINDP( 1) ,«A I NDP( 2 ) / 5HRRO XX ,4HDUMP/
     RAIN( 1)-XRAIN( 1)
     RAIN(2)=XRAIN( 2)
     CALL  SEEK (8, LIST)
     REAO(8)DTT,OT, 118, LLOC I ,.MAME.S. LISTA, L I SI B,L [STC. LISTD, NG,NC
     CALL  CLOSE  (8)
      IDT=300
     NDTT=120
     KK=85
     MISEC^O
 900 FORMATtSX. A5, A4>
     NTER=1
     IY = 0
     ICOWTA=1
     ICONTB=1
     ICONTC=1
     ICONTD=1
     CALL  SENSEdb. I1B( 17) )
     IF(IIB(17) .EQ. 1>CO  TO  13
     WRITE16.2)
>     FORMATC61H  SENSE S W I TCH E5< ACCU M.  SJS)  "1-10. 17 ARE UiED  IN THIS P
    1ROGRAM/45H   IF  ON(UP)  THEY HAVE THE  FOLLOWING FUNCTIONS//
    250H  «1 SURPRESS PRINTOUT OF EXCESSIVE  FLOW LOCATIONS/
    328H  «2 PAUSE IF EXCESSIVE FLOW/
    437H  «3 PRINTOUT RAINFALL .OSS-RflTE  DATA/
    536H  «4 PRINTOUT UNDIVERTED HYOKOGRAPHS/
    634H  »5 PRINTOUT DIVERTED rtYDROCRAPHS/
    736H  »6 PRINTOUT VOLUMES OF HYOROGRAPHS/
    853H  «7 PRINTOUT HYDROGRAPiHS AT POINTS  OF EXCESSIVE FLOW/
    939H  «8 PRINTOUT HYDROGRAPHS AT ALL  POINTS)
     WRITE(6.4)
   4 FORMATCVOH   «9 COMPUTE UNDIVERTED HYDROGRAPHS OWLY)
   5 FORMAT!
    154H  »10  COMPUTE UNDIVERTED AND DIVERTED HYDROGRAPrfS  ONLY)
     WRITE(b.20D
 201 FORMATiiOH   «12 SCAN REAL  TIME  DATA  ONLY)
     WRITE<6«200>
                                   177

-------
200 FORMATC32H   «13 00 NOT SCAN REAL  TIME DATA/
   131H  "14 RETURN TO PROGRAM CONTROL'
   238H  »15 PROCESS HISTORICAL RAINFALL  FILE/
   347H  «16 SUPPRESS PRINTOUT OF  PROGRAM. INSTRUCTIONS)
    URITEC6.6)
  6 FORHAT(
   241H  "17 PAUSE  AT END OF LATEST  COMPUTATIONS)
    WRITE(6. 7)
  7 FORMAM/
   347H SET  SWITCHES TO DESIRE} POSITION  AND THEN  TYPE)
    WRITE(6,8)
  8 FORNATt
   427H CONTROL  P   AFTER THE PAUSE)
    URITE(b,9)
  9 FORHAT(
   538H THE  SWITCHES MAY BE RESET  AT  A. MY-. TIME)
    PAUSE
 12 CALL SENSEC15, 118(16))
    IF( 118(16) .EQ.O)GO TO 26
102 URITE(6.22)
 22 FORMATUX.51H  ENTER NAME OF HISTORICAL RAINFALL  F!L£  (FILNAMEXT))
    READ (6t23)RAIN
 23 FORMAT(A5. A4)
    GO  TO  25
 26 CONTINUE
    WRITE(b. 10)
 10 FORMATdX, 15HENTER REAL TIME/LX.10H  HR MN DAY)
    READC6.11) IHR, 1NIN. IDAY
 11 FORMAT(2I3 ,14)
    ISEC=3600»IHR*60»IMIN
    CALL SETdSEC.IOAYJ
    MISEC=0
     MIDAY= IDAY
     IF(II8(16) .ECJ. II GO TO 23
 21  CONTINUE
     LSX=((OTT-KK-1I»IOT
     CALL TMADO(MISEC.MIMIN.MIHR,MIOAY,LSX.O,0.0)
 20  TTT= 0.0
     GO TO  100
 25  CONTINUE
     CALL FSTAT(8.RA1N. ITEST)
     IF(ITEST.NE.O)GO TO 26
     WR1TE(6, 10DRAIM
101  FORHAT(5X,4HFILE,2X, A5.A4, 2X.2 3HNOT PRESENT  ON  DEVICE  8/ )
     GO TO  102
100  CONTINUE
     CALL CMAIN12)
     STOP
 13  ICONTA=0
     GO TO  12
     END
     SUBROUTINE  TMADD(L SEC ,LMI N ,L HR .LDAY, KSEC . KM I N , K«R, 
-------
      LDAY=LDAY  «•  KO AY
      IF(LSEC.GE .O.ANO.LSEC. IT. 60)30  10 I
      LMIN^LMI N*LSEC/60
      LSEC=LSEC-(LSEC/60)»60
      IFUSEC.GE .0)00 TO 1
      LSEC=LSE060
      LMIN=LMIN- 1
      IF(LMIN-GE .O.AND.LMIN.LT.6J)  GO TO 2
      LHR-LHR *•  LMIN/60
      LMIN=LMIN-'60
      IFUMIN.CE .0)  GO TO 2
      LM1N=LMIN  +  60
      LHR^LHR-1
      IFUHR.CE. O.AND.LHR.LT.24)  GO ID 3
      LOAY^LDAY  *  LHR/24
       IF(LHR.GE. 0)  GO TO 3
       LHR-LHR  *  2^
       LDAY=LDAY- 1
    3  RETURN
       END


C        CHAIN2
C       GET  RAINGUAGE READINGS,  SET GATES,  CHECK LEVELS
       REAL  LISTA.L IST8.L ISTC ,L IS TD . N AMES
       COMMON OTT,NDT T.IDT. NTER.D T, II S< 18) ,LL OC [(15 ) ,NAMES(4a ) ,
     1LISTA(2).LISTB(2).LISTC(2».LISTD(2).KK. ICONTA, ICONTB.
     2ICONTC.[ CONTD, 1C. I Y. MI HR , M I M IN . M [DAY , M I SEC. R A I Nl 2 ) . NG, NC ,
     30PPI 15.120), ITIMES( 12)
       CALL  SENSE(1«. IIb( 15))
       IF( I 18(15) .EQ. 1) GO  TO  3
       CALL  SENSE (13. I I3( 14 >>
       IF( 118(14) .EQ. 1>CO TO  6
       IF( ICONTO-LT.OGO  TO  6
       CALL  CHAIN (9)
       STOP
    b  CONTINUE
       ICONTDMABS( ICONTD)
       CALL  SENSE (9,  I 18(10) )
       CALL  SENSE( 10, IIB< 11 »
       CALL  SENSE (17. IIB( 18) )
       CALL  SENSE (?.,  118(3 ) )
       IF( IIBC 16) .EQ. 1.ANO.NTER.EQ..1) GO  TO 3
       IFUIBU8) .EQ. l.OR .1 IBU5) .EQ.l . OR . I CO NT D . EQ . 2)CO TO  1
       IF
-------
       CALL  CHAIN ( 5)
       STOP
       END


       SUBROUTINE  CONTL
       REAL  NAMES,LIST*,LISTB.LISTC.LJSTD
       DIMENSION  BMSTH2)
       COMMON OOT.NOTT.IDT.NTER.OT. I IB(18).LLOCK15).NAMES ( 48).L1ST A(2 ).
      1LISTB(2).L1STC(2).LISTD(2),KK,JCONTA,ICONTB, ICOflTC,ICONTD,IC, IY,
      2MIHR,MIMIN,MIDAY,MISEC,RAIN(2) ,NG,NC,QPP(15, 120J , IT I MESI 12>
       DATA  BMSTL(1),BMSTL(2)/5HBMCXS .4HTLOS/
   4 7 WRITE(6.99 HCONrO, MI DAY, MI HR , HIM IN
   99 FORMATd6H  PROGRAM CONTROL,15H  PREVIOUS  CHAIN, I4.12H   MODEL TIME
      1, 15,213)
       ICONTD=2
   27 IF( ICONTA.EQ.O) GO TO 46
       WRITE(6.28)
   28 FORMATC//
      146H TYPE  1  TO RESTART WITH  ALL -.INITIAL CONDI TtOflS/
      251H TYPE  2  TO RESTART COMPUTING UNDIVERTED  HYDROGRftPHS/
      352H TYPE  3  TO CONTINUE COMPUTING  UNDIVERTED  HYDROGRAPHS/
      450H TYPE  4  TO CONTINUE COMPUTING  DIVERTED HYDROGRAPHS/
      527H TYPE  5  TO CONTINUE ROUTING/
      638H TYPE  6  TO CONTINUE SCANNING REAL  DATA/
      729H TYPE  7  TO CALL GATE SETTINGS/
      815H TYPE  8  TO STOP/)
       ICONTA=0
   46 CONTINUE
   45 FORMATI12H  TYPE NUMBER)
   48 READ(6.35) IZ
   35 FORMAT(Il)
       IFUZ.LT.l .OR. I2.GT.8) GO  TO  50
       GO TO 191, 92,93,94,95,96,97.98 ), 12
   50 WRITEC 6.40)
   40 FORMATdBH  TYPE NUMBER AGAIN)
       GO TO 48
   91  CALL  CHAIN  ( 1)
       STOP
   92  NTER=1
       URITEC 6.100)
  100  FORMATtlX,25HENTER MODEL STARTING TI ME/I X.10H  MR  MM  DAY)
       READ(6.101>MIHR.MIMIN.MIDAr
  101  FORMATC2I3,14)
C      WRITE(5,102)MISEC,MIMIN,MIHR, Mi DAY
C 102  FORMAH8H  TIME 1  ,415)
       CALL  RAINF
C      URITE(5.103)MISEC.MIMIN.MIHR,MIDAY
C 103  FORMAT(8H  TIME 2 ,415)
       BMC^O.O
       STLOS^O.O
       CALL  DLETE(7,BMSTL,N)
       CALL  ENTER(7.BMSTL)
       DO 37 1=1.NC
       WRITE(7,36)  BMC.STLOS.BMC.STLOS
   37 CONTINUE
   36 FORMATdX, 4F5.2)
       WRITE(7.38)  BMC
   38  FORMATdX, F8.II
                                     180

-------
   CALL CLOSE(7)
   CALL CHAIN  ( 3)
   STOP
93 NTER=2
   CALL RAINF
   CALL CHAIN  (3)
   STOP
94 CALL CHAIN  (4)
   STOP
95 CALL CHAIN  (5)
   STOP
96 ICONTD=2
   CALL CHAIN(9)
   STOP
   GO  TO  46
97 CALL GATEDT
   GO  TO  47
98 STOP 777777
   END

   SUBROUTINE  GATEOT
   REAL NAM
   DIMENSION GATES(2).TEMP(2),IIB(18>
   DATA GATESi 1) , GATE 5>( 2 ) / 5HG AT ES .4HETTG/
   DATA XL. XI ,R,D,E/1HL, 1HI ,1HR ,1 HO.IHE/
   DATA TEMP( 1) , T EMP( 2 ) / 5HTEMPO , 4 HF I LF. /
   J = 0
   WRITE<6.2)
  2 FORMATU2H  GATE EDITOR/)
   CALL SENSE(16, I IB< 17) )
   IF< I 18(17) .EO. 1)  GO TO  3
   URITE<6,4>
  4 FORMAT(43H  TYPE L FOR LISTING  UF PRESENT CATESEIT INGS/
   154H  TYPE  D  WITH GATE  NUMBER  AND TIME- TO DELETE  A  SETTING/
   261H  TYPE  I  WITH GATE  NUMBER.  TIME  AND SETTING-  TO  INSERT SETTING/
   361H  TYPE  R  WITH GATE  NUMBER.  UME  AND SETING-  Tfl  REPLACE SETTING/
   434H  TYPE  E  TO  EXIT  TO PROGRAM  CONTROL/)
  3 WRITE(6.5)
  5 FORMATd X. 24HN GT HR  MN  DAT  X.XX X.XX)
   REflO(6,6)T,IGT,JH, 1M.ID.HG.HF
  6 FORMAT(A1,313.I4.2F5.2)
10 FORMAT(25X,26HGATE  AND  FABRIDAM SETTINGS//
   110X.9HDIVERSION8X4HT IMEIOX4HGA TE4X8HFABR I DAM 3X3MMAX6X3HMAX/
   211X6HNUMBER7X2HMR3X3HHMN2X3HDA Y<» X7HSET T I NG3 X 7HS€ T TI NG3 X4 HGAT E
   32X8HFABRIOAM/ )
   K=0
   1^0
   IFU.EO.XL  ) 1=1
   IF(T.EQ.D)1=2
   1FCT.EQ.XI  ) 10
   IF(T.EQ.R)1^4
   IF(T.EQ.E)1=5
   IFU.LT.l.OR.I .GT. 5)GO  TO  606
   IF(I.EQ.5)GO TO 500
   IF
-------
100  IF (J.EO.l)  GO TO 600
     CALL SEEK  (7,GATES)
     WRITE  (5.10)
 13  READ (7.11)  LGT.NftM, LH.LM, LD , H£L , HFL , HGM AXL, HFKAXL
     IF (LGT.EQ.20) CO TO  14
     WRITE  (5.23)  LGT.NAM.LH.LM.LD,HCL.HFL,HGMAXL,HFHAXL
     GO TO  13
 1*  CALL CLOSE  (7)
     GO TO  3
300  IF 
-------
       SUBROUTINE E VNTLC N, NNN, NT, ISS, 1MM.IHH, [DO)
C    TAKE  N UNEQUALLY SPACED EVENTSRAND GENERATE NN EtfENTSRR  WITH SPACING  0
C    STARTING AT TIME ISS . I MM. I HH, I DO.
       DIMENSION R( 35 ).RR(35) .IS( 35). 1M( 3'3 ) . I H( '35) . IDl 3 5 )
C      R  IS AVERAGE RATE  FOR  LAST  PERIOD UP  TO TIME  OF  R.
       COMMON/BKTL/R. IS. I M , I H , 10 , RR
C    GO  BACK IN RECORD TO  READIN3 NEJ(T PREVIOUS TO  ISS , IrtM, I HH , IOD
       NN-0
     2  J
       M^N-J+H
       IFXKM»60.*XKH»3600.+XKD»at>400.
       KS=IS( IA>
       KM-IM( IA>
       KH= IH( IA)
       KD= I0< 1A)
       CALL  TMAOD(KS,KM, K H ,KD,- I S S , - I MM.- I HH. -I DO)
C      WRITE( 5. 198)KS.XM, KH,KD
       IFtKD.LT -0)GO TO 3
       XKS-KS
       XKH^KH
       XKD=KD
       XNT-NT
       XKS=XKS*XKM»60.*XKH»3600. + XKD» 8*600.
       NP-XKS/XNT*! .
       XNP=NP
       XXKS-( XNP-1. 0) »XNT
       IF« XKS-XXKS) .EQ.XNT)NP = NP*1
       NPP = 0
       DO  4  L=l .NP
       NPP-NPP+1
       XL=L-l
                                     183

-------
       NN-NN+l
       DR=RUA)-R< I )
       TF = (XLS-XKS«-XNT»XL )/Xl_S
       RR ,NN, I ,R< l),R( U)
C   11  FORMAU20X.I4, 3F10 .5.2I4.2F10. 5)
       IF(NN.EQ.NNN)GO  TO  12
     4  CONTINUE
    11  CONTINUE
       XNP-NPP
       XKKS=XNP»XNT
C      HRITE<5.199)XKKS
C 199  FORMAT(1X,7H XK.KS  ^,F15.2)
       IFUKKS.LT. 131000. ) GO TO lb
       KKD=XKKS/86400.
       XKKD=KKD
       XKKS=XIUS-XKKD»86400.
    6  KKS-XKKS
   17  CflLL  TMADD ( ISS. IMM, IHH. IDD, KKS. 0,0. KKDJ
C      WRITE  <5. 10)XLS,XKS,NP. NN.KKS. M.J.I, NPP
C  10  FORHAT(2F1 0.1, 216, 110.4I&)
       IF(NN.EQ.NNN)GO TO  13
     3  CONTINUE
   13  NNN=NN
       KKS=-NT
       CALL  TMADDUSS.IMM, IHH.IDO.KKS .O.C.O)
C      HRITE(5.102) ISS.IMM. IHH.IDO
C 102  FORMAT(8H TIME 6 ,515)
       RETURN
   16  KKS=NPP»NT
       GO  TO 17
   95  NNN=0
       RETURN
       END

       SUBROUTINE RAINF
       INTEGER CHECK
       REAL  LISTA.L IST8.L I S TC , L I T, TO , N AMES
       DIMENSION RAINOT(2).SAVE(33,9) . I TIME ( 4 . 35, 9 )  . I ADDR( 9) .
     UAOORt 9).NUMB(9)
       COMMON /8KTL/R( 35), IS( 35) . I ri( 35 ).IH( 35).IOl35  ).RR(35)
       COMMON/TTTT/GAGE(9 ) ,RF (9)
       COMMON  DTT.NDTT.IDT.NTER.DT.'I IB< 18).LLOC I I 15),NAMES< 48 ),
     1LISTA(2).LISTB(2).LISTC(2),LISID(2).KK, 1 C UNT  A, ICONT8,
     2ICQNTC.ICONTD, IC.IY,M[HR.M1M.IN.MIDAY,MISEC,RAIN<2).NG.NC,
     3 SAVE, I TIME .IXXX<17 10), ITIMtSl 12)
       DATA  RAINOT(1).RAINOT(2)/5HIRA1N.4HOATA/
       CALL  SEEM 8.LISTO)
       CALL  OLETE(7,RAINDT. I)
       CALL  ENTER(7,RA/NDT)
       READ(8) IAOOR.JAOOR
       CALL  CLOSE (8)
       CALL  SEEK( 8,RA IN)
       DO  20 1=1. NG
   20  NUM8( I )=1
       CHECK=1
       IDTT=IDT
                                     184

-------
C      WRITE( 5,505)MIS£C, MIMIN.MI HR.M 1DAY
C 505  FORMAT(20X .6181
       1-1
       NNG=NG
  201  READ (8. 10) IDA, IHA. IMA.IAODA.XXK1.XXX2.RDG
   10  FORMAH3I3.I6. A5.A2.F9.3)
       IF(IADDA.CE.7777) GO  T 'J  210
       IF( 1AODA.EQ. JADOR(NNG) )  GO TO 203
  208  NNG-NNG+1
       CHECK=1
       IF1NNG.GT. NG) NNG = I
       00  101  1=1. NG
       NNGGM
       IF( IAOOR( I ).EQ.JADOR(NNG) >  GO TiJ 11
  101  CONTINUE
   11  CONTINUE
       GAGE(NNGG) -XXXI
C      HRITE(5.500)NNC,NNGG, JADDR
C 500  FORMAK 318 .5X, AS)
       1 = 1
       GO  TO  221
  202  REAO(8.10) IDA, IHA, IM A, I A DO A, X X Ml , XXX 2 , RDG
       IF(IADDA.GE.7777) GO  TO  210
       IF( IAOOA.NE. JAOOR(NNG) )  GO TO 208
  221  IF(ROG.LE. ( - . 5 ) .OR . ROG . GE . 3.0)  ROG-99999999.
  207  ITIME( l.l.NNGOMOA
       ITIHE( 2. l.NNGO-IHA
       ITIHEO. l.NNGG) = IMA
       ITINE(4.1.NNGG>=0
       SAVE(1.NNGG)=ROC
       1 = 2
       GO  TO  201
  203  IFIRDG.LE. (- .5 > .OR .RDG .GE. 3.0)  GO  TO 204
       IFCSAVEt I-l.NNGO.GT .3.0)  GO  TO 205
       DIFF=ROG-SAVE( I-l.NNGG)
       IF(OIFF.LT .(-.02)) GO TO 204
       JIOA=-ITIME( 1. I-l.NNGG)
       JIHA = -ITIH£(2. I-l.NNGG)
       JIHA=-ITIME(3. I-l.NNGG)
       JISA=0
       CALL THAOO(JISA,JIMA, JIHA. JIDA.O.INA .IHA. IDA)
       OELT=JIOA»2**JI«A*JIMA/60»JISA/3600
       SLOPE=DIFF/OELT
       IF(SLOPE .GT.6.0) CO  TO 204
       GO TO  205
  204  ROG=SAVE(  I-l.NNGG)
       IF(CHECK.EQ.O) RDG =99999999.
  205  IF(CHECK.EO.O) GO TO  206
       JIHA=-IHA
       JIMA=-IMA
       CALL T(1AOD(0, J IMA, JIHA, JIDA.O, MIMIN.MI
       IF( JIDA.GE.O)  GO  TO  207
       CHECK-0
   206 ITIHE< 1. I. NNGG) = IOA
       ITIME12, 1, NNGG)=IHA
       ITIMEt 3. I.NNGO-IHA
       ITIHE(4, I,NNGG)-0
       SAVE( UNNGG)=RDG
                                     185

-------
      NUMB  GO TO 210
      IF( IADDA.EQ. JAOOR(NNG) ) GO TO  209
      GO TO 208
  210 IF( I.LE.2)  GO  TO 300
      MX=35
      DO 211  I=1.NC
      N=NUMBd)
C     WRITE(5.503)GAGE( I ),NUMB( I )
C 503 FORMATC5X, A5.5X.I5)
      DO 212  J=1,N
      R(J)=SAVE< J, I)
      ID( J>- IT IMEU, J.I )
      IH< J)=ITIHE(2, J.I)
      IM( J)=ITIME<3, J.I)
      IS(J)=0
  212 CONTINUE
C     URITt(5.501>R
C 501 FORMAT<1X.10F13.3)
      CALL SENSE(3, IIB(4))
      NN=NDTT-KK
      IF( IIB(16) .EQ. I) NN=13
      IXS-MISEC
      IXM-HIMI N
      IXH-MIHR
      IXD=MIDAY
      CALL EVNTL(N,NN.IOTT. I XS , I XM , I XH, IXD )
      NUMB(I)=NN
      IF(NX.GT.NN)  MX-NN
      00 213  J-l.NN
  213 SAVEU. I >-RR( J>
C     URITE(5.502)NN
C     WRJTE( 5.50DRR
C 502 FORHATtllO)
  211 CONTINUE
      IF(NN.EQ.MX)  GO  TO  214
      ISEC=(MX-1)»IDTT
      CALL TMADOCMISEC. HIM IN , Ml HR, M I BAY . ISEC .0.0,0)
C     URITE( 5. 505) MI SEC, HIM IN. MI HR , M ID AY. I SEC
      GO TO 215
  214 MISEC=IXS
      MIMIN=IXH
      MIHR-IXH
      MIDAY=IXO
  215 SUM^O.O
C     URITE<5.505)MISEC. MI M IN. MI HR , M iOAY
      COUNT=0.0
      NN = HX
      DO 216  1=1 ,NG
      IF(SAVE(1. I) .GE.3. 0) GO TO 216
      COUNT=COUNT*1.0
  216 CONTINUE
      AVG-SUM/COUNT
      00 217  [=1.NG
                                    186

-------
       IF(SAVE(ltI) .GE.3.0) SAVE(1,I> =
  217  CONTINUE
       DO 218  1=2,NN
       SUM=0.0
       COUNT-0.0
       DO 219  J=1,NG
       IF(SAVE(I.J).GE.3.0) GO  TO  219
       SUM=SUM*SAVE(1.J>
       COUNT-COUNT* 1.0
  219  CONTINUE
       AVG1=SU«/COUNT
       DIFF=AVG1-AVG
       DO 220  J=l,NG
       IF(SAVE(I,J).GE.3.0) SAVE(b,J) -SAVE ( I- 1. JJ+0 IFF
  220  CONTINUE
  218  AVG^AVGl
       IF(IIB(4).EQ.O) GO  TO  70
       WRITE( 5,71)(I,1=1,NG)
   71  FORMATiIH1/1X.8HGAGE NO./9I9)
       DO 222  J=l,NN
  222  WRITE(5.72XSAVE
       IF(RF(I).LT.O.OJRF(IUO.O
   18  CONTINUE
       URITE(7)  RF
C      WRITE<5.100)  RF
C  100  FORMAH IOX.9F12.6 )
   19  CONTINUE
       CALL CLOSE(7)
       CALL CLOSE  (8)
       RETURN
       END


C       CHAIN  «3
       REAL I.ISTA,LISTS,L ISTC , L ISTD . N AMES
       DIMENSION ITP(15), ITB(15),ALOSC( 15,15) ,CZ( 15,9)
       DIMENSION QPB(15.1),QPPDP(2).Ri8),JJ(85)
       DIMENSION QPP( 15, 1 20 ) , 1 IB ( 18 ) , LLOCI ( 15 ) .NAMESCrS) ,LJSTA( 2),
     1LISTBC2),LISTC(2),LISTD(2),QPP02(2)
       COMMON  DTT.NOTT.IDT.NTER.Dr.II 8,LLOCI,
     1 NAMES, LIST A. LI STB, L I STC , L I STD, KK, ICONTA, ICONTB
     2, ICONTC. 1C ON TO, 1C, IY.MIHR, MI MI N,MIDAY, M I SEC , RA I>N( 2 ) , NC , NC , QP P,
     3  I TIMES* 12 >
       COMMON/BLOCKH/ITP,ITB.CZ,AR(15),8FFAC(2*).QP8
       COMMON/BLOCKT/ALOS C
       DATA QPPDZ(1),UPPDZ(2)/5HQPP2Z,^HVERT/
       ICONTD-3
       DO 203  I=1,KK
  203  JJ(l)--(NDTT-KK)*!
       CALL SEEK (8.LISTA)
       READ (8)  ITP,ALOSC,AR,C2.QPB,I 18,QPPDP,BFFAC
       CALL CLOSE(8)
C      READ RAIN GAUGES AND CALCULATE  HYDROGRAPHS
                                    187

-------
        IF(NTER-l) 100.101,100
101 CALL DLETE(8.0PPDP, IK)
    CALL ENTERCB.QPPDP )
    KTM=-(NOTT-KK) '(IDT/60)
    KTS=KTHH=KTD=0
    CALL TMADD(KTS,KTM,KTHH,KTD,MIS£C,MIMIN, MIHR.MIDAY)
    KKMMM 60-KTH)/ t IDT/ 60)
    DO 71 I=1,NC
    KKM-KKHM
    KTH=KTMH+1
    KS=1
 32 KF=KS*KKM-i
    DO 30 KI=KS,KF
    QPP(I.KI)-QP8( I.1)»BFFAC(KTH>
    IFUI.EQ.120)  CO TO 31
 30 CONTINUE
    KS=KF+1
    KKM=(3600/IDT)
     IF(KTH.GE.25)KTH=KTH-24
    GO TO 32
 31 CONTINUE
 71 CONTINUE
    WRITE <8)QPP
        CALL  CLOSE  (8)
100 CONTINUE
112  CALL HrDR>
     IFUIB15) .EQ.OJGO TO  80
     WRITEC5, 75 )MIHR,MI MIN, MIDA Y
 75  FORMftT(//lX.22HUNDIVERTED  HYDROGRAPHS/
   139H POINT  0  REPRESENTS PRESENT MODEL T I ME, 2 X, 2.1 3 , U . 5 X,
   232HPOINTS ARE  AT 5  MINUTE  INTERVALS)
     DO 74  1-1, NC
     LL-LLOCM1 >
     WRITE(5,72 JNAMES(LL)  . I
     WRITE (5. 73 > ( JJ< J) ,QPP( I, J) ,J=1 ,KK)
72
73
1
74
FORMATt/
FORMATI 1
13, F7. I,
CONTINUE
/I
X.
13

X, A5
13, F
,F7.

. 15H
7.1,
1.13

HtDROGR
I3.F7.1,
•F7.1.I3

APH
13. F
,F7.

NO., I
7. 1, I
1.13.

5)
3.F 7.
F7. 1.


1,1
13.


3.F7. 1.I3.F7.
F7. 1>


1


 80  CONTINUE
     CALL  CHAIN(2)
     STOP
     END


     SUBROUTINE HYDR
     REAL  L1STA.LISTB.LISTC,LISTO,NAMES
     DIMENSION OPP( 15,120).TEMPX(2) ,QPPOP(2 I,XOECAr(15)
     DIMENSION ITP( 15) . IT B( 1 5 ) , RA I NO! ( 2 ), GAGE ( 9)
                                  188

-------
      DIMENSION  RO(9).RF (9).CZ(15,9) , R(9),BMSTL(2)
      COMMON DTT,NDTT,IDT,NTER,DT,IIB<18).LLOCI(15)
     1, NAMES(48) ,LISTA(2),LISTB(2),HSTC(2),LISTD(2).
     2 KK, I CO NT A. ICOMTB, I CO NIC.ICONTD. IC.IY.MIHR.MIMIN.MIDAY
     3,MISEC.RAIN(2).NG,NC,OPP
      COMMON/BlOCKH/ITP,ITB,CZ,A(15) ,BFFftC(24) ,QPB(15. 1)
      COMMON/BLOCK!/ALOSC(15,15)
      DATA BMSTL(1),BMSTL(2)/5HBMCXS,4HTlOS/
      DATA RAINDTl 1) .RAINDTt 2J/5HIRA [N.4HDATA/
      DATA QPPDP(l),QPPDPl2)/5HOPPXX.4HDUMP/
      DATA TEMPX( 1), TEMPX(2)/5HTEMPO.4HRARY/
C     SETINIT1AL  RAIN GAUGE READING
      CALL SEEM 7.RAINDT )
   98 FORMATU8H  READINGS AND  T I ME , 9 F 10 .5 , 5 X , 2 I 3, I 4/>
C
C     TIME
C     TTT-TIME  AFTER  START OF  STROM, NTER-CORRESPONDI NC  RAINFALL
C     INCREMENT
C     NO
      CALL SEEK(8, BMSTL)
      DO  798  1=1,NC
      READC8.799) ALOSCU , 9 ) , ALOSU I , 10 ), AL OS Ct 1.12 ) . AL OSC( I , 15 )
      XDECAY(I)=ALOSC(I.4)*»(.25>
  799 FORMATC4F5 .2)
  798 CONTINUE
      READ(8.796)ACCUH
  796 FORMATIF8. 1)
      CALL CLOSE(8)
      1DTT=IDT
      CALL SENSEU, I 13< 4 ) )
      IF(IIBU).EQ.O)GO  TO 200
      URITE(S.201)RAIN, ( I , I - 1 , NG )
  201 FORMATdHl/51H  RAINGACE  READINGS AT  5 MINUTE  INTERVALS  IN (IN/HR)
     134H FOR  INTERVAL ENDING  AT  TIME  SHOHN, 10X,16HRAlNFALL  FILE -  ,
     2A5.A4//5X.8HGAGE NO.,3X,9110,8X.9HHR  MN DAY)
  200 CONTINUE
C     READ RAINGAUGES
      READ(7)IR.GAGE
      CALL SEEM 8. OPPOP)
      READ (8)OPP
      DO  129  IRI=1, IR
      ACCUM=ACCUM*DT
  199 READ ( 7)R
       IF(IIB(4).£Q.O)GO  TO 114
      JSEC=-(IR-IRI)»IDTT
      JMN=0
      JHR=0
      JDAY^O
      CALL TMADO(JSEC.JMN,JHR,JDAY.M 1SEC.MIM IN.MIHR,Ml DAY)
      WRITE* 5,97)(GAG£(I ), I = 1.NG>
   97 FORMAT(/18X.9(5X.A5))
      URITE< 5.98) (R( I),I = 1,9).JHR,JMN.JDAY
  114 DO  4 NGG=1.NG
           RF(NGG)=R(NGG>
           RO(NGG)=R(NGG)
     4  CONTINUE
           DO 21 NCC= KNC
                                    189

-------
    ALOSC
  6 CONTINUE
    1TT=NDTT-1
    DO 700 IT = 1, ITT
    IS=IT*1
700 QPP/<3600.0*12.0 >»DT«2.0/(FLOAT(ITB(NCC)-l).OT)
        NDTTP=1*ITP(NCC)
        NDTTB=1*IT8
    QPP(NCC.IT)=QPP(NCC.IT)*S»QP
  9 CONTINUE
 21 CONTINUE
129 CONTINUE
    CALL CLOSE(8)
130 CONTINUE
    CALL CLOSE(7)
    NTER=2
    CALL DLETE(8,BMSTL,LLL>
    CALL ENTERC8.BHSTL >
    DO 795 l=l,NC
    WRITE(8,794)ALOSC(It9),ALOSC(I.10).ALOSC<[.12).ALOSCCf.l5>
794 FORMAT(1X,4F5.2»
795 CONTINUE
    WRITE(8,793) ACCbM
793 FORMATilX, F8.1 I
    CALL CLOSE(8)
        RETURN
        END


    SUBROUTINE  SLOS(R, RE X, I, DT , 1 1)
    COMMON/BLOCKT/TSTLOS(15),RECOV < 15 ) ,AX( 15 ) ,DE CAK15) .XI HPERdSI .
   1XIMPCFC15),ENDK<15),TBMC(15),BMC(15),STLOS(15).TSTLSP<15),
   2STLOSPU5)  .COEFdS ).POWER( 15) , AC LOSS ( 15)
    RA=R»DT/3600.
    IMPERVIOUS  AREA
    Z=STLOS(I)
    STLOS(l)=STLOS(i)*RA
                                  190

-------
   IF(STLOS( I ) .GT.TSTLOS( I) >  GO Tfl  3
   X^O.O
   GO TO 19
 3 X=STLOSi l)-TSTLOS< I)
   STLOS( l)-TSTLOS(I)
19 XILIMP=STLOS( I )-Z
   XLSCOF = 1 .0/-TSTLSP( I )
 4 XILPRV = STLOSP( 1>-U
   SLOSX=ENOK( I)+AXU ) * (  (TBMC ( I )-BMC(I ) ) / TBMC ( I ) > » • 0 . 70
   POTLSS = SLOSX»DT/3600.
   ACTLSS=POTLSS
   IF(ACTLSS.GT. Y)  ACTLSS^Y
         = lf-ACTLSS
   BMC1 = BMC( I )
   BMCU )=BMC ( I J + TLSPER
   BASLSS=TLSPER»XIMPER(I )*TLS I MP Ml .0- XI MPER( I »
   BSEXCS= BMC ( I ) - TB MC ( I )
   IF( I I .EQ.01GO TO 14
   IF( I.GT. 1) GO  TO 11
   URITE( 5. 12 )
12 FORMAU128H        I   RAIN    ISTLOS    PSUOS   IINLOS  PINLOS ILOSC
  IDE  PPOTLOS    ILOSS   PLOSS    I E XS     PEXS    BASLOS    BASEXS   BMC
  2ACCLOSS. // )
11 WRITE(5.13)I,RA,STLOS
   EQUIVALENCE  < D I VC < 1 ) . Z Z( 1 . 1 ) ) ,  ,S S( 1) )
   DATA QPPDX ( 1) , QPPDXl 2>/5HQPPOI .4HVERT/
                                 191

-------
    DATA OPPDZ ( D.QPPDZm /5HQPPZZ .4 HVERT/
    ICONTD=4
    DO 200  1=1, KK
200 JJ( I)=-(NOTT-KK)«-1
    CALL SEEK  (8,L ISTB )
    READ(8)ND, NC.DF,NQ,N A, NQQ, DIVC ,D IVCC.FACT.SU M.'.iUM,QMAX, FACT
    CALL CLOSE (8)
    CALL SEEM 8.QPPDZ)
    REAO(8)OPP
    CALL CLOSEC8)
    DO 32 1=1, ND
    IF(NQ( I) .G6.16JCO TO 32
    CALL D1V( I )
    CALL SENSEU4, 118(15) )
    IF( IIBC15) .EQ. 1>GO  TO 500
 32 CONTINUE
    CALL SENSE(5. IIB<6»
    IF(IIB(6).EQ.O)  GO  TO 42
    UR I IE (5. 201) HI HR, HIM IN, MI DAY
    DO 42 1=1, NC
    LL-LLOCI (I )
    WRITE15, 72 )NAMES
-------
      IFdC.EQ. 2)GO  TO 303
      CALL SEEK(8,QPPDX)
      GO TO 302
  303 K=K*1
      00 44 1=1, NC
   44 ZZ(1, I )-ZZ(l, I )»FACT( I )
      WRITE(5»307)
  307 FORMAT<13H  INLET INDEX .18H LUMPED  INLET VOL ,
     118H   LUMPED DIV VOL  . 18H LUMPED  RIVER VOL .
     118H     INLET  VOL     ,18H   DIVERTED  VOL   ,14ri    RIVER  VOL  )
      DO 304  (=1 ,NC
      ZZ(K,I)=ZZ(1,I >-2Z (2,1 )
      zz(4,D=zz(i. n/FAcni >
      ZZ(S,I)=22(2,I )/FACTlI )
      22(6,1)^22(3, I )/FACT(I )
      DO 11 J=l,6
      SS( J)-SS(J) + ZZU, I >
   11 CONTINUE
  304 WRITE(5.305) I. (22( J, I>,J=1,6)
  305 FORMAU5XI2, 10XF 10 . 0 , 8XF 1 0 . 0 , 8 XF 10 . 0 . 8 XF 10. 0 . 8XF 10. 0 .
     18XF10.0I
      WRITE (5. 306MSSO) ,J=1.6>
  306 FORMAT(17H  TOTAL VOLUME  CF F 10 .0, 8XF 10 . 0 , 8XF 10.0 ,
     18XF10.0.8XF10. 0.8XF10. 0)
  308 FORMAT(17H  TOTAL VOLUME  MG 4 XF 8 . 2, 10XF8 . 2, 10XF8. 2 ,
     110XF8.2.10XF8.2.10XF8. 2)
      DO 12 1=1.6
      SSU >=SS([ )» 7. 48/1 000000.0
   12 CONTINUE
      WRITE(5.308) (SS ,J=1.6)
  211 CONTINUE
  500 CALL  CHAIN12)
      STOP
      END


      SUBROUTINE  OIV(ND)
C     AT EACH  INLET  POINT THE  HYDROGRAPH  IS  CALCULATED FOR THE  AMOUNT OF
C     RAIN  WHICH  HAS  FALLEN. THIS SUBROUTINE  THEN DETERMINES  THE AMOUNT
C     OF THE HYDROGRAPH WHICH ACTUALLY  FLOWS TO THE MAIN INTERCEPTOR  AT
C     THAT  POINT. (I.E. THE FLOOD WAVE WHICH  IS THEN ROUTED THROUGH  THE
C     SYSTEM)ND-DIVERSION NO,                DF-DIAM OF PIPE IN  WHICH
C     FA8RIDAM  IS FIXED, H-HGHT OF F A BR1DAM , NQ- NO OF HYDROGRAPH  FOR
C     DIVERSION  ND, NT-TIME NO.NA-NO  DES IGNAT ING GE OMET RIC AL  SHAPE  OF
C     ORIFICE. NQQ-IND£X OF POINT AT  WHICH  FLOW IS DIV£RTED(EG  METERS),
C     ALSO  INDEX  OF  INFLOW POINT FOR  ST PETER DIVERSION
C
C
      REAL  LISTA.LISTB.L ISTC ,L IS TO , N AMES
      DIMENSION  C1(17),C2(17),C3(17).C4(17).C5(17)
      DIMENSION  IIB( 18), L LOCI (15 I .NAMES (48 ) , L I STA ( 2 ) .L I STB( 2 ) ,
     1LISTU2) ,LISTD(2)
      DIMENSION  DF(19),NQ( 19),NA(17) .OPPl 15 , 12 0 ) , NQO( 1 9 )
C     FOR SPECIFIED  DIVERSION  STRUCTURES  GO  TO APPROPRIATE SUBROUTINE
      COMMON/TY/X1.X2.X3 ,X4,X5,X6.X7.X8,X9,X10.X11,X12,C13,C141C15
      COMMON DTT. NDTT.IOT.NTER.DT, MB. L LOCI. NAMES,
     1L ISTA.LISTB.LISTC, LISTD.KK, I CO NT A. I C ON TB , I CON TC. ICONTD,
     2IC,IY,MIHR,MIMIN, MIDAY.MISEC,RAIN(2),NG.NC,QPP
                                    193

-------
       COMMON /O I WE/ OF, NQ. NA,NQQ,C1,C2 .C3.C4.C5. QMAX<48) ,FACT< 15)
       Xl^CHND)
       X2-C2(NO)
       X3=C3(ND)
       X4=C4
      QPP
-------
32  CONTINUE
    IF(NP2.EO.KK»GO  TO  33
    NP = NP»IDTM«-IDTM
    CALL  TNADD(0, IMINH, IHRH.IDAYH, O.NP.0,0)
    GO TO 34
 33  CONTINUE
    GO TO 100
  4  LQ=NQ
-------
    6  LQ-NQ(NO)
       MO=NQQ
       00  181  K-l.KK
       IF(QPP
-------
       GO  TO  24
C3         WRITE(6,20l) NC,C4,HG
   23  CONTINUE
C01        FORMAT(2X.I2,2F10.5)
       AO=CRES(C4,HG>
C02        FQRMAT<2X,F10.5>
   24  IFUO.LE.0.0)AO-0.00001
       IF (Q.LE. 0.00) 0 = 0.000 001
       IF(C2»0»»2 ./AO»»2«-HG/2 .-H + C5M 1.11,2
    2  QMAX=Q
       L=TOP
-------
       SUBROUTINE T YPEM ND , Q, Q2 , H , HF ,      H A , HF A, OF ,DfA )
       REAL  LF
       COMMON/TY/C1.C2.C3
C      EUSTIS DIVERSION
C      Q-APPROACH ING FLOW , Q2-D I VE S F ED  F LOW, C 1 -COEF F OF  DISCHARGE OF
C      FABR IDAH.C2-EXPONENT  OF FA3RIDAM  01 SC H AR GE , H-H&H T  GF  GATE.HF-HGHT
C      OF  FABRIDAM,EQN F- C 1 »L F »( 0 2 » • 2 / ( 3 2 . 2 * A » » 2) + H/2-MF ) » »C2
C      A-AREA OF OR IF ICE, LF-L GHT  OF FAfcRlOAM
C      THE ABOVE APPLIES  TO  DIVERSION  OF  90INCH PIPE.HiE  SAME  E QN APPLIES
C      TO  DIVERSION OF 60 INCH  PIPE. THE LETTERA REFERS fO  THE LATTER DIVER
C      SI ON
C      Q3-APPROACHING FLOW  FOR 60INCH  PIPEIASSUMED TO BE  C3»Q
C
C
C
C
C      WRITE( 5.400)ND.a,C 1. C2.H.HF, C3.HA.HFA, DF.DFA
C 400  FORMAT(39H TYPE4 NO. 0, C 1 , C 2 , H , HF . C 4 . H A , H F A , DF , OF A/ I 5 . 1 OF 10. 5 )
       Q3=C3»Q
       Q=Q»(1 .0-C3)
       Kfl-0
     7  LF=TOP(DF, HF)
       A = CRES(3.,H)
C  ,    HRITE15.40DLF.A
C 401  FORMAT111H TYPE4 LF,A,2F10.5)
       IF(Q»»2/(32.2»A»»2)*H/2.-HF)l,l,2
     2  QMAX=0
       QMIN-0.
       K=l
     3  Q2=(QMAX+QMIN) 12.
       BB=Q2*»2/( 32.2»A»»2)*H/2.-HF
       IF(BB) 30.8,8
   30  86=0.000001
     8  F=Q-C1»LF»BB»«C2
C      WRITEC 5. 402)02, F.K
C 402  FORMAT(13H TYPE4 Q 2 , F , K , 2 F 1 0 . 5 , I 5 )
       CALL  BISECC(F,Q2,QMAX,QMIN,K,ND.QN)
       IF(K-50> 3, 4,4
     I  Q2 = Q
     4  IF(KA)5.5.6
     5  Q2S=Q2
C      60  INCH  DI VERS ION
       HSAVE^H
       DFF=DF
       DF^DFA
       HF-HFA
       KAM
       GO  TO  7
     6  Q2=02+Q2S
       H=HSAVE
       HF^MFSAVE
       DF^DFF
       RETURN
       END
                                     198

-------
      SUBROUTINE  TYPE5CND,Q,DF,HF
      l,C13,C14,H.Yl.Q5,C 15)
      REAL  LF
      COHMON/TY/Cl,C2.C3.C<>,C5,Cb.C7 .C8.C9.C10.Cll.C12
C     ST PETER  KELLOGG
C     NC-CACHMENT NO.Q-APPROACHING  FLOW.Cl-COEFF OF  DISCHARGE OF WEIR AT
C     A.C2-LENGTH OF ME IR AT A,C3-CO£FF OF DISCHARGE  OF  ORIFICE AT 0,
C     C7-AREA  OF  ORIFICE  AT D.C4-COEFF OF DISCHARGE  Of  F ABR I DA M, L F-L GTH
C     OF FABRIDAM CREST,C6-EXPONENT FOR FABRIDAM DISCHARGE  EON
C     1ST DIVERS ION EQN
C     F=Q-Q2-C1»C2»<02»» 2/l2'32.2»C3»'2»C7»»2-5>»»1.5-<4»LFMQ2'»2/2»
C     32.2»C3»»2»C7»»2-HF)»»C6
C     Q-APPROACH1NG FLOW  RATE.Q2-FLOW GOING TO  2ND 01 WERSION.DF-0I AM OF
C     FABRIDAM  P1PE.HF-HGHT OF FABRIDAM
C
C     2ND DIVERSION
C     C9-COEFF  OF DISCHARGE OF SIDE WEIR.C8-HGHT OF  UPPER  WE IR CREST.
C     C10-LGTH  OF LOWER SIDE WEIR  CREST,Cl1-COEFF  OF  DISCHARGE OF
C     ORIFICE,C12-HGHT OF LOWER  WEIR  CREST,C13-D IFF  IN  ELEV BETWEEN
C     ORIFICE  AND INTERCEPTOR.Cl*-LGHT GF UPPER WEIR CREST,AG~AREA OF
C     ORIFICE. Yl-DEPTH OF WATER  IN  INTERCEPTOR,H-HGHT OF ORIFICE GATEt
C     Q2-AS ABOVE.OS-FLOW DIVERTED  TO INTERCEPTOR
C     CIS-WIDTH OF ORIFICE
C     DETERMINE BRACKETTED  EXPRESSIONS
C     WRITE(5.500)ND,Q.C1,C2,C3,C4    .C6.C7.C8.C9.C10.C11.C12.C13.C14.
C     1H.Y1,C15.DF,HF
C 500 FORMAH39H  TYPES NO, Q, Cl, C 2, C 3 .C4, C6 ,C 7, C8, C9, CIO/15,1 OF 10 . 5
C     1/31H  C11.C12.C13.C14.H.Y1.C15.DF.HF/9F10.5)
      8A=Q»*2/(64.4»C3»»2*C7'»2)-5.
      BB=BA*5.-HF
C      IS THERE  FLOW OVER  FABRIDAM
       IF(8B) 1. 1.2
     1 88 = 0.
      GO TO 3
     2 BB=1.
     3 IF(BA)4.4.9
     4 BA-0.
      GO TO 10
     9 BA=1.
    10 IF(BA*B8)5,5. 11
    11 CONTINUE
      LF=TOPCOF,HF)
C     WRITE( 5.50DLF.8A, BB
C 501 FORMATU5H T YP E5 L F , BA ,BB. 3 F 1 0 . 5 >
      QMAX-Q
      QMIN^O.
      K = l
     7 Q2MQMAX+QMIN) /2.
       BC = Q2»»2/C64.4»tC3»C7)'»2) -5.
      BD = QZ»»2/(64.4 »(C3»C7)»»2) -HF
       IF(BC)22,23,23
    22 BC^O.OOOOOOl
    23 IF(BD)24.25,25
    24 BD-0.0000001
    25 F = Q-Q2-C1» C2»B A*(8C) • »1.5-C4.Lf *B8'(BD)*«C6
C      WRITE(5.502)02.F.K.BC.BD
C  502 FORMATU3H  T YPE5 02 . F . K, 2F 1 0 .5 . J 5 .2F 1 0 . 5 »
       CALL  BISECUF, Q2, QMA X, QM1 N , K  , NO, QN)
                                     199

-------
       IF(K-SO) 7. 8.8
C      2ND  DIVERSION
     5  Q2 = U
     8  IFCY1-C13-H/2. ) 12, 12,13
   12  HS=H/2.
       GO TO  14
   13  HS=Y1-C13
C      IS THERE  DISCHARGE OVER  WEIRS
   14  AG=C15»H
       8A = 1.
       88=1.
       IF(Q2»*2/( 64.4. (Cl 1»AG> »»2 ) -Cl 2+HSJ15, 15,16
   15  8A=0.
   16  IF(Q2»»2/( 64.4»18, 19, 18
   18  QMAX=Q2
       QMIN=0.
       K-l
C      WRITE<5,503)HS. AC, 8A.BB
C 503  FORMATU8H  TYPE5 HS . AG . BA , BB , 4 F10 .'5 )
   20  05=(OMAX*OMIN> /2.
       BC = Q5»»2/<64.4*(Cll»AG)»»2)-C12-»-HS
       BD=Q5»»2/t 64.4»
-------
       FUNCTION AREA  < D, Y Y )
       Y = Y Y
       IF> )
    20 IF-0.10)3,2,2
     2  K-K+1
       ON=QZ
       IFIK-50) 13, 14, 14
   13  IF(F)8.7,7
     7  QMIN-Q2
       RETURN
     8  QMAX-QZ
       RETURN
     3  K = 50
       RETURN
   14  WRITE(5,15)NC
   15  FORMAT127H  NOT  CONVERGING  CACHMENT NO, 15)
       RETURN
       END

       FUNCTION TOP(D,Y)
       IF(4.0»(0»Y-Y»YJ)1,2,2
C    1  WRITE(6,3)Y,D
     J  FORMAT15X.5H Y,0  . 2F 10.2)
     I  Y-D
     2  CONTINUE
       TOP=SORU4 .0*( a»Y- Y»Y) >
C      TOP IS TOP  WIDTH,D-DIAM.Y-3EP1H
       RETURN
       END

       FUNCTION CRES(D.H)
C      AREA  OF  CRESCENT
C      USES  FUNCT ION  AREA
       H1MD-H) 12 .
       A=AREA(D.H1)
       CRES=(D/2.)»'2'3.14159-2.»AREA<0,Hi)
       RETURN
       END
                                     201

-------
       FUNCTION  TRAPOZCN, XINTl., J)
       REAL NAMES.L ISTA.L ISTB.LISTC.L IS TO
       COMMON DTT.NDTT.IDT.NTtER.Df, 118(18) ,LLOC I (15). NAME S«;8>,
      1LISTA<2> ,L IiTe(2).L!STC(2) ,LIS TD(2).KK, ICCNTA. ICONT8.
      2ICJNTC. ICON TO, 1C,  I Y, Ml HR, Ml M [ M ,M I Dfi Y , M ['.,EC , R A I .M< 2J.NG.NC.
      JQPP<15.120)
       SIMPSN-(QPP(J, l)*QPP(j.N)) 12 .0
       N1=N-1
       00 2 1=2.Nl
     2 SIMPSN-SIMPSN+QPP(J,I)
       TRAPOZ-SIMPSN'XINTL
       RETURN
       END


       SUBROUTINE   FABCTS(ND, IMINH. 1HRH, IDAYH,HGC.HFF.NP)
       DIMENSION  GATES(2).DF(19),NQ(I'M
       COMMON/DIVE/OF.N3
       DATA GATES( I> , GATES(2>/5HGATES.4 HE TTG/
       LS^O
       CALL SEEK (7.GATES)
     4 READ(7,5) J.LH,LN,LD,HG.HF,hGMAX,HFMAX
       IF(J.NE.ND) GO  TO  2
       CALL TMADD(LS,LM,LH,LD,0,-IMINH.-IHRH,-IDAYH)
       IF(LD.GE.O) GO  TO  3
     2 HGG=HG
       HFF-HF
     5 FORMATC 13,5X,21.3, I4,4F!>.2)
       GO TO i,
     3 NP ••- LO'28d*LH»12>LM/5
       CALL CLOSE (7)
       RETURN
       END


C »»» »»CHAIN 5
C     PAL  METHOD OF  ROUT  ING
C     JR-CONSIANTS,ICONTB-OPERATION  NO,ICONIC  NO OF OPERATIONS
      REAL  LISTA.LISTB.LISTC.LISTD,NAMES
      DIMENSION JR(A8.7),OMAX(48),D( 3),S(3),QPPDX(2).Q(120),TEMP(2).
      10PP(15.120)
      COMMON DTT,NDTT.IDT,NTER.DT,tIB(18).LLOCI(15).
      1NAMES(48).LISTA<2),LISTB(2),LISTC(2) ,L IS TO(2),KK,ICONTA. ICONT8,
      2ICONTC.ICONTD, IC.IY.MlHR.MlMlN.MIDAY.MlSEC.kAtNt 2).NG,NC,OPP
      3, IT1MESC 12 )
           COMMON/BLCK5/JR.Q, IJ,0 MA X, JJ(120)
      COMMON/6KDI/O,S.AN,DF( 19),DlVC{17,5) .DI VCC<10)
      DATA  QPPDX(1),QP?DX(2)/5HQ?PDI.4HVERT/
      DATA  TEMP{ 1),TEMP(  2)/5HTEMPO,4 HFIL6/
       ICONTO=5
       IY = 0
      DO  70 1=1,NDTT
   70 JJ( I)=-(NDTT-KK»*I
   80 CALL  SEEM 8.LISTC)
      READ(8)JR,QMAX.D.S.AN,ICONIC.DF.DIVC.DIVCC
      CALL  CLOSE(8)
       ICONTB=0
      CALL  SEEM 8.QPPDX)
   60 READ(8)QPP
                                    202

-------
   CALL CLOSE(8)
 5 ICONTBMCONTB*!
   1H=JR( ICONTB,2»
   IHH=JR(ICONTB,4)
   IJ-O
   IE=JR( ICONTB,1 )
   GO T0(1,2.3,401.IE
 1 CALL ROUTB( 1H. I HH )
   GO TO  4
 2 CONTINUE
   IHA= JR< ICONTB, 3)
   CALL COHB( IH, I HA,IHH)
   GO TO  4
 3 CALL DIIHN(IH)
   GO TO  4
40 00 42  1-1,NDTT
42 QPP( I H.I)-QPP( IH, I )+QMAX
   IF(IIB(2).E0.1)GO TO 82
   WR1TE(5.12)NAMES(ICONTB)
82 IF(IIB(8).EQ.O.AND.1 18(9).EQ .0 ) CO TO 41
   IFUIB(8).EO.1.AMD.IJ.EQ.O.AND . JIB(9).EG .0)GO  TO  41
   URITE(5.8)NAME SU CONTB).IE,I CO NTB.QMAX( ICONTB).MIHR.MIMIN.MIDAY
 8 FORMATI//14H HYDROGRAPH  AT2X,A 5 , 5X,1 OH  OPERATION  13,7H  NUMBER  13,
  15X,5HQMAX  .F10.2,3X, 11HTIME  PI.  1 ,213,14)
   URITE(S.9>(JJ( 10).QPP(  IHH, 10), IU=1,NOTT)
 9 FORMAT(IX.I3.F7.l,I3,F7.1,I3,F7.1.I3.F7.1,I3.F7.1.I3.F7.1,
  1I3.F7.1,13,F7.1,I3,F7.1,I3,F7.1,I3,F7.1.I3,F7.1J
12 FORMATU18H EXCESSIVE FLOW  AT  2X ,A5, 2X ,9H»»» ».»»«»  /
  146H BOB  CALLERY'S BASEMENT  WILL  BE FLOODED»»*»»»»    )
41 CONTINUE
   IF( ICONTB- ICONTO5.6.6
 6 CONTINUE
   CALL CHAIN  (2)
   STOP
   END


   SUBROUTINE  ROUTB( I rt. IHH )
   REAL NAMES.L 1STA.L[STB,LISTC.LISTO
   DIMENSION  0(120),JR(48,7),QMAX(48),QPP(15.120)
   COMMON  DTT.NDTT,IDT.M'ER.OT,1I8<18),LLOCH15),
  1NAMES(48),LISTA{2),LISTB(2).LISTC(2),LISTD(2),KX,ICQNTA.ICONTB.
  2ICONTC.ICONTO, 1C, IY,MIHR,MIMIN , MI OhY,M1 SEC,RA IN(2),NG,NC,QPP
       COMMON/BLCK5/JR.Q,IJ.QMAX
    NRCHS^JR( ICONTB.7)
   LAG= JR( ICONTB,6)
   NSTRL= JRU CONTB.5)
   DO 56  1=1,NDTT
   IF(QPP
-------
C  40  FORMAK8I6)
       ANS=NSTRL
C   PERFORM   AVERAGING
       00  20  IZ=1.NRCHS
       DO  7  IX =l.NDTT
       J=IX-NSTRL 12
       JX=J+NSTRL-1
       0(1X1=0.0
       1F(J-l)2.3.3
    3  IF . t 'j . 0) GO  TO 6
       00  50  L-l,NDTT
   50  QPP( IH,L ) = 0(L)
       GO  TO  33
    6  DO  51  L=2,NDTT
   51  QPPl IH.L-1)-Qt L)
       QPP(IH.NDTT)-Q(NDTT)
   33  CONTINUE
   20  CONTINUE
       DO  32  1=1,NDTT
   32  Q(1) = QPP(tH. I  )
   41  FORMATt10F10.2 )
C      PERFORM   LAGGING
       ICN = 0
       I F.< LAG) 3 8, 2 1,2 1
   2 1  IF(MOD(NSTRL,2).60.0)ICN=NRCMS/2
       GO  TO  10
   38  CONTINUE
       NRCHS=1
   30  LAG=IABS(LAC)»NRCHS+ICN
       IF(LAG.GE. 0)GO TO  44
       LL=NDTT*LAG
       DO  42  IN=1.LL
       INLAG= IN-LAG
   42  OPP( IHH. IN)-Q( INLAC)
       LL=LL*1
   43  QPP( IHH.IN)=Q(NDTTI
       RETURN
   44  LL=LAG*1
       DO  9 IN=LL,NDTT
        INLAG=IN-LAG
    9  QPP( IHH. IN)=Q( INLAG)
       DO  10 IN=1.LAC
   10  QPP( IHH, IN)=Q( I)
       RETURN
       END
                                    204

-------
   SUBROUTINE  COMB ( I H , I Hfl . I HH )
   REAL NAMES.L ISTA.LISTB.LISTC.L.ISTD
   DIMENSION  JRI48.7)
   COMMON OTT.NDrT, IDT, NT ER. OF. 118(18). LLOCK15).
  1 NAMES ( 48>.LISTA(2).LlSTBl2),LlSIC<2).LlSrD(2).KK,lCONTA.lC3NTB.
  2 1 CONK. I CO NT 0. I C, 1 Y , M IHR , MI M I N , M 1 DA Y , M I SEC, R A I N< Z ) , NG . NC ,
  3QPP< 15.120)
       COMMON/BLCK5/JR, 0(120) , IJ , QMAX148)
   00 1 1=1 .NOTT
 1 QPP( IHH, I) -QPPUH, I ) +QFP( I HA , I )
   RETURN
   END

   SUBROUTINE  DIVIN(IH)
   REAL NAMESrLlSTA, LISTB.LlSrC.LlSTO
   DIME MS ION  JRU8,7)
       COMMON /BLCK5/JR, 0(1 20) , U.QMAXC^S)
   COMMON DTr.NDTUlDT.NTER.DT.IIBUaj.LLOCKl1;),
  INAMES< 48),LlSTA(2),LlST8{2).UISfC(
   VOLA=TRAPOZ(NDTT,DTT. IH)
   NP2 - 0
   NPT^NDTT
   IOTM- IDT/60
   LSX=-(NDTT-KK-1)»IDT
   IS=0
   IMINH- MI MI N
   IHRH^MIHR
   IDAYH-MIDAY
   CALL T MA DO (IS. IMINH. IHRH.IOAYH.LSX. 0,0.0)
 4 CALL FABCTS(ND, IMI NH, IHRH. 1DAY H. HCG.HFF. NP)
   IF(NPT .LE.NP)  CO TO 10
   NP2=NP
   GO TO  11
10 NP1=NP2*1
11 DO 22  K-NP 1.NP2
   GO TO  (1 . 1,1, 1.1,1. 7) , IA
 I CALL TYPE1  ( NO,DF( NO) .HFF.HCG, OV'P( Iri.K ) .02.1 )
   QPP( IH.K) =02
   GO TO  22
 7 IF(HGG.LE.O. 1 ) GO TO  22
   CALL VALY(0( IB) ,Y, S( IB),QPIJ( IH ,K).flN)
   CALL TYPEMND,Y,HGG,D1VC(NL>,1) ,DIVC(ND,2).QPP( l.H.K))
22 CONTINUE
   IF(NP2.EO. NDTT) GO TO  13
                                 205

-------
      NPT=NDTT-NP2
      NP=NP»IDTM«-IDTM
      CALL TMADD(IS, IMINH, IHRH,[OAYH,0,N?,0,0>
      GO TO 14
   13 CONTINUE
      IF(IIB(7).EQ.O)CO  TO 15
      VOLB=TRAPQZ(NDTT.OTT, IH)
      VOLC-VOLA-VOLB
      WRITE(5.16>VOLA,VDLB,VOLC
   16 FORMAT124H  VOLUME  BEFORE 0 IVERSI ON,F10.0/
     122H VOLUME  TO  INTERCEPTOR,F10.0/
     216H VOLUME  TO  RIVER,F10.0)
   15 CONTINUE
      RETURN
      END
      SUBROUTINE  V AL Y( D. Y . SO . Q'J, AN )
C     GIVEN  D-D1 AM.SO-BED SLOPE,QO-D ISCH.AND-MANNINGS N,  ROUTINE FINDS
C     NORMAL  OEPTH-Y FOR CIRCULAR  SE CIIOiM, US IN C  BISECTION.  AREA  FUNCTION
C     USED.
      K=0
      YMIN=0.0
      YMAX=D
    8 Y = (-YMIN*YMAX)/2.0
      K = K + 1
   10 FORMAH21H  NORMAL DEPTH USED  ft S. F10 . 3 , 34 H    DISCHARGE IS.GT.SECTI
     ION CARRY         >
      IF
-------
BIBLIOGRAPHIC:
   Minneapolis-Saint Paul Sanitary District, Dispatching Sys-
tem For Control of Combined Sewer Losses, Final Report
FWQA Grant No. 11020 FAQ, Maich, 1971.

ABSTRACT
   Impressive reductions in combined sewer overflow pollu-
tion of the Mississippi River in  Minneapolis  and  Saint Paul
have been effected  by a regulator control system. Working
entirely  within  the limits  of the existing interceptor  sewer
system, and with relatively minor modifications to selected
major combined sewer regulators, incidence of overflow was
reduced  by 66% and  duration of overflow by 88% during
most  of a rainfall season. Computer simulation  techniques
using  actual  rainfall  data indicate that the amount of
overflow  volume reduction achieved is  the equivalent of a
$200  million separation project. The efficiency of collection
was improved by about 20% at controlled regulators. The
reduction  in volume  of combined  overflow to the river is
estimated  to be  between  35% and  70% during  the runoff
season. The  unmodified combined sewer system captured
about 65% of the urban runoff.  Where modified,  the system
captured about 77% of the  urban runoff.
ACCESSION NO.


   KEY WORDS:
Combined Sewer
Regulator

Interceptor
Overflow
Control and Monitoring
Variable Diversion
Real Time Operation
Urban Hydrology
Stormwater Runoff
 BIBLIOGRAPHIC:
    Minneapolis-Saint Paul Sanitary District, Dispatching Sys-
 tem For Control of Combined  Sewer Losses,  Final Report
 FWQA Grant No. 1 1020 FAQ, March, 1971.

 ABSTRACT
    Impressive reductions in combined sewer overflow pollu-
 tion of the  Mississippi River  in Minneapolis and Saint Paul
 have been effected by a regulator control  system. Working
 entirely within  the limits of the  existing interceptor sewer
 system, and with relatively minor modifications to selected
 major combined sewer regulators,  incidence of overflow was
 reduced by  66% and  duration  of overflow by 88% during
 most  of a  rainfall season. Computer simulation techniques
 using  actual rainfall  data  indicate  that  the amount  of
 overflow volume reduction achieved is the equivalent of a
 $200 million separation project. The efficiency of collection
 was improved by about 20% at  controlled regulators. The
 reduction  in volume  of combined overflow to the river  is
 estimated  to be  between 35%  and  70% during the runoff
 season.  The unmodified  combined  sewer  system captured
 about 65% of the urban runoff. Where modified, the system
 captured about 77% of the urban runoff.
 ACCESSION NO.


    KEY WORDS:
 Combined Sewer
 Regulator
 Interceptor
 Overflow
 Control and Monitoring
 Variable Diversion
 Real Time Operation
 Urban Hydrology
 Stormwater Runoff
 BIBLIOGRAPHIC:
   Minneapolis-Saint Paul Sanitary District, Dispatching Sys-
 tem  For Control of Combined Sewer Losses, Final Report
 FWQA Grant No. 11020 FAQ, March, 1971.

 ABSTRACT
   Impressive reductions in combined sewer overflow  pollu-
 tion  of the Mississippi River in Minneapolis and Saint Paul
 have been effected  by a regulator control  system. Working
 entirely within  the liinils  of the existing interceptor  sewer
 system, and with relatively minor modifications to selected
 major combined sewer regulators, incidence of overflow  was
 reduced by 66% and  duration of overflow by  88% during
 most  of a rainfall season. Computer simulation techniques
 using  actual  rainfall  data indicate  that  the  amount  of
 overflow  volume reduction achieved is  the equivalent of a
 $200 million separation project. The efficiency of collection
 was  improved  by  about 20% at controlled regulators. The
 reduction  in volume  of combined  overflow to the river  is
 estimated  to be between  35% and  70% during the runoff
 season. The  unmodified combined  sewer  system captured
 about 65% of the urban runoff. Where modified, the system
 captured about 77% of the urban runoff.
ACCESSION NO.


   KEY WORDS:
Combined Sewer
Regulator
Interceptor
Overflow
Control and Monitoring
Variable Diversion
Real Time Operation
Urban Hydrology
Stormwater Runoff
                                                                       207

-------
   A mathematical model has been prepared that will, with
rain  gage data  as  input, perform  rainfall runoff analysis,
diversion  of combined sewer runoff hydrographs at regu-
lators, and  routing of diverted  hydrographs through  the
interceptor system. This model will  assist in operation of the
system  to retain  combined sewer flows and  utilize  the
maximum flow  capacity of the existing  interceptor  sewer
system. Part II of this report describes the model.
   The 1.75 million dollar project includes a computer-based
data acquisition and  control  system that permits remote
control of modified combined sewage regulators. Data from
rain  gages,  regulator  control  devices,   trunk  sewers and
interceptors,  and  river  quality monitors provide  real-time
operating information. Time varient quality  data from key
locations in  the  sewer  system were obtained by automated
analysis of numerous hourly samples.
   Work  on the "Dispatching System for Control of Com-
bined Sewer Losses" began in 1966, and  the system has been
operated  since April, 1969.
   This  report  was submitted  in  fulfillment  of Demon-
stration  Grant  11020  FAQ,  supported in part by  the
Environmental Protection Agency, Water  Quality Office.
    A mathematical model has been prepared that will, with
 rain gage  data  as  input,  perform rainfall  runoff analysis,
 diversion  of combined sewer runoff hydrographs at  regu-
 lators,  and routing of diverted  hydrographs through the
 interceptor system. This model will assist in operation of the
 system  to  retain  combined  sewer flows  and  utilize the
 maximum  flow capacity  of  the  existing  interceptor sewer
 system. Part II of this report describes the model.
    The 1.75 million dollar project includes a computer-based
 data  acquisition  and  control system  that  permits  remote
 control of modified combined sewage regulators. Data  from
 rain  gages,  regulator  control devices,  trunk  sewers and
 interceptors, and  river quality  monitors provide  real-time
 operating  information. Time varient quality data from key
 locations in the sewer system were obtained by automated
 analysis of numerous hourly samples.
    Work on  the  "Dispatching System  for Control of Com-
 bined Sewer Losses" began in 1966, and the system has been
 operated since April, 1969.
    This  report  was  submitted   in fulfillment of Demon-
 stration  Grant  11020 FAQ,  supported  in  part  by  the
 Environmental Protection  Agency, Water Quality Office.
   A  mathematical model has been prepared Ihat will, with
rain  gage  data as input,  perform  rainfall  runoff analysis,
diversion of  combined  sewer runoff  hydrographs at  regu-
lators, and  routing  of  diverted hydrographs  through the
interceptor system. This model will assist in operation of the
system  to  retain  combined  sewer  flows  and  utilize the
maximum flow capacity of the existing interceptor sewer
system. Part II of this report describes the model.
   The 1.75 million dollar project includes a computer-based
data  acquisition and  control  system  that  permits remote
control of modified combined sewage  regulators. Data from
rain   gages,  regulator control  devices,  trunk  sewers  and
interceptors,  and   river  quality monitors provide  real-time
operating  information.- Time  varient quality data from key
locations in  the sewer system were obtained by automated
analysis of numerous hourly samples.
   Work on the "Dispatching System for Control of Com-
bined Sewer Losses" began in 1966, and  the system has been
operated since April, 1969.
   This report was submitted  in fulfillment of Demon-
stration Grant  11020  FAQ,  supported  in  part by the
Environmental Protection Agency, Water  Quality Office.
         208

-------
     Accession Number
  w
                      Subject Field &. Group
                     04A,056
          SELECTED WATER RESOURCES  ABSTRACTS
                INPUT TRANSACTION FORM
     Organization
    Minneapolis-Saint  Paul Sanitary District,  St. Paul, Minnesota
     (NOW METROPOLITAN  SEWER  BOARD)
     Title
    DISPATCHING SYSTEM FOR CONTROL OF COMBINED SEWER LOSSES
 10
Authors)


Robert L. Gallery
16
Project Designation

EPA/WQO Demonstration Grant  11020 FAQ
                                  2]  Note
 22
     Citation
     Water Pollution  Control  Research Series  11020 FAQ 03/71,  206 pp,
       81 fig,  30 tab,  36 ref
 23
     Descriptors (Starred First)
     *Storm  Runoff,  *Sewers,  *0verflow, *Pollution Abatement,  *Mathematical
 Models, Diversion  Structures, Routing, Sewerage,  Sanitary  Engineering,
 Water Management  (APPLIED),  Simulation Analysis,  Model Studies, Water
 Pollution  Control
 25
Identifiers (Starred First)
*Combined  Sewer, *Regulator,  *lnterceptor,  *Control  and Monitoring,
 Urban Hydrology, Computer  Model,  Sewage Sampling  and Analysis
 27
     Abstract
     Results of initial operation of a computer-based data  acquisition and  control system
for a major urban combined  sewer system are presented.
     Impressive reductions  in combined sewer overflow pollution of the Mississippi River
in Minneapolis and St.  Paul have been effected.  Working entirely within the  limits of
the existing interceptor  sewer system, and with relatively minor modifications to selected
major combined sewer regulators, incidence of overflow was reduced by 66% and duration of
overflow by 88$ during  most of a rainfall season.  Computer simulation techniques using
actual rainfall data indicate that the amount of overflow volume reduction achieved is the
equivalent of a $200 million separation project.  The reduction in volume of  combined
overflow to the river is  estimated to be between 35% and 70% during the runoff season.
The combined sewer system captured about 77% of the urban runoff at the modified locations.
     A mathematical model has been developed ttiat will,  with rain gage data as input,
perform runoff analysis,  diversion of combined sewer runoff hydrographs at regulators,
and routing of diverted hydrographs through the interceptor system.
     Data from rain gages,  regulator control devices, trunk sewers and interceptors, and
river quality monitors  provide real-time operating information.  Time variant quality
data from key locations in  the sewer system have been acquired from analysis  of numerous
hourly samples.
         L.  Gallery
                        Watermation,  Inc.,2304 University Ave.,St.  Paul 55114
  WR:I02 (REV. JULY 1869)
  WRSIC
                       SEND WITH COPY OF DOCUMENT. TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                              U.S. DEPARTMENT OF THE INTERIOR
                                    "~         WASHINGTON. D. C. 20240
                                            209
                                                                          • GPO: I 970-360-930

-------
Continued from inside front cover....

11022 --- 08/67        Phase I  -  Feasibility of  a Periodic  Flushing System
                       for Combined  Sewer Cleaning
11023 --- 09/C7        Demonstrate  Feasibility of tfie Use of Ultrasonic
                       Filtration in Treating the Overflov/s from Combined
                       and/or Storm  Sewers
11020 --- 12/67        Problems of  Combined Sewer Facilities and Overflows,
                       19C7, (WP-20-11)
11023 — 05/68        Feasibility of  a  Stabilization-Retention Basin in Lake
                       Erie at  Cleveland, Ohio
11031 — - 00/68        The Beneficial  Use of Storm l/ater
11030 DNS 01/65        Water Pollution Aspects of Urban Runoff, (UP-20-15)
11020 Dili OG/69        Improved Sealants for Infiltration Control, (UP-20-13)
11020 DES 06/69        Selected Urban  Stonn Water Runoff Abstracts, (KP-20-21)
11020 — 06/6S        Sev/er Infiltration Reduction by Zone Pumping, (DAST-9)
11020 EXV 07/69        Strainer/Filter Treatment of Combined Sewer Overflows,
                       (WP-20-16)
11020 DIG 08/69        Polymers for  Sewer Flow Control, (WP-20-22)
11023 DPI 08/69        Rapid-Flow Filter for Sewer Overflows
11020 DGZ 10/69        Design of  a  Combined Sewer Fluidic Regulator, (DAST-13)
11020 EKO 10/69        Combined' Sewer  Separation Using Pressure Sewers,  (ORD-4)
11020 — 10/69        Crazed Resin  Filtration of Combined  Sewer Overflows,  (DAST-4)
11024 FKM 11/69        Storm Pollution and Abatement from Combined Sewer Overflows-
                       Bucyrus, Ohio,  (DAST-32)
11020 DWF 12/69        Control  of Pollution by Undenvater Storage
11000 — 01/70        Storm and  Combined Sewer  Demonstration Projects -
                       January  1970
11020 FKI 01/70        Dissolved  Air Flotation Treatment of Combined Sewer
                       Overflows, (WP-20-17)
11024 DOK 02/70        Proposed Combined Sewer Control by Electrode Potential
11023 FDD 03/70        Rotary Vibratory  Fine Screening of Combined Sewer
                       Overflows, (DAST-5)
11024 DMS 05/70        Engineering  Investigation of Sewer Overflow Problem -
                       Roanoke, Virginia
11023 EVO 06/70        Microstraining  and Disinfection of Combined Sewer
                       Overflows
11024 — 06/70        Combined Sewer  Overflow Abatement Technology

-------